Biosensors and Bioelectronics 64 (2015) 69–73
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Highly sensitive detection of a bio-threat pathogen by gold nanoparticle-based oligonucleotide-linked immunosorbent assay Sang-Hwan Seo a,1, Young-Ran Lee a,1, Jun Ho Jeon a,1, Yi-Rang Hwang a, Pil-Gu Park a, Dae-Ro Ahn b,c, Ki-Cheol Han b, Gi-Eun Rhie a, Kee-Jong Hong a,n a Division of High-Risk Pathogen Research, Center for Infectious Diseases, Korea National Institute of Health, Osong Health Technology Administration Complex, Cheongwon, Chungcheongbuk-do 363-951, Republic of Korea b Center for Theragnosis, Biomedical Research Institute, Korea Institute of Science and Technology, Hwarangno 14-gil 5, Seongbuk-gu, Seoul 136-791, Republic of Korea c Department of Biological Chemistry, KIST Campus, Korea University of Science and Technology, Hwarangno 14-gil 5, Seongbuk-gu, Seoul 136-791, Republic of Korea
art ic l e i nf o
a b s t r a c t
Article history: Received 25 May 2014 Received in revised form 7 August 2014 Accepted 19 August 2014 Available online 23 August 2014
Francisella (F.) tularensis causes the zoonotic disease tularemia and categorized as one of the highestpriority biological agents. The sensing approaches utilized by conventional detection methods, including enzyme-linked immunosorbent assay (ELISA), are not sensitive enough to identify an infectious dose of this high-risk pathogen due to its low infective dose. As an attempt to detect F. tularensis with high sensitivity, we utilized the highly sensitive immunoassay system named gold nanoparticle-based oligonucleotide-linked immunosorbent assay (GNP-OLISA) which uses antibody-gold nanoparticles conjugated with DNA strands as a signal generator and RNA oligonucleotides appended with a fluorophore as a quencher for signal amplification. We modified the GNP-OLISA for the detection F. tularensis to utilize one antibody for both the capture of the target and for signal generation instead of using two different antibodies, which are usually employed to construct the antibody sandwich in the ELISA. The GNP-OLISA showed 37-fold higher sensitivity compared with ELISA and generated very consistent detection results in the sera. In addition, the detection specificity was not affected by the presence of non-target bacteria, suggesting that GNP-OLISA can be used as a sensitive detection platform for monitoring high-risk pathogens thereby overcoming the limit of the conventional assay system. & 2014 Elsevier B.V. All rights reserved.
Keywords: Francisella tularensis Bacteria ELISA OLISA Detection sensitivity Gold-nanoparticle
1. Introduction Francisella (F.) tularensis, the causative agent of the zoonotic disease tularemia, is categorized as one of the highest-priority biological agents of concern by the U.S. Center for Disease Control and Prevention (CDC) because it has a very low infectious dose (10 colony-forming units (CFU) of F. tularensis) (Ellis et al., 2002; Froude et al., 2011; Franz et al., 2001). Vaccination is the most efficient weapon against F. tularensis infection, but no vaccine has been officially approved by the U.S. Food and Drug Administration to date. Treatment of patients with proper antibiotics during the early stages of infection and decontamination of pathogen-containing environments with selected agents are the only ways to the control outbreak to the disease and to reduce risk of public n
Corresponding author. Tel.: þ 82 437198271; fax: þ82 437198309. E-mail address: [email protected] (K.-J. Hong). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.bios.2014.08.038 0956-5663/& 2014 Elsevier B.V. All rights reserved.
health. Identification of F. tularensis bacteria in biological and environmental samples is essential for efficient clinical treatment and appropriate decontamination by public health official. Hence, a sensitive and specific detection method to detect tiny amounts of F. tularensis is required. The conventional detection method for F. tularensisis requires at least three days of bacterial cultivation and identification using specific media (Pohanka et al., 2008). As an alternative, real-time polymerase chain reaction (RT-PCR) has been utilized due to its selectivity, speed, and sensitivity. However, the limit of detection (LOD) using RT-PCR is 103 CFU/mL for F. tularensis and it does not meet clinical demand (Dauphin et al., 2011). In addition, RT-PCR can generate false-positive signals if contaminated or false-negative signals if the target gene is mutated or if the sample contains Taq polymerase inhibitors (Alaeddini, 2012). ELISA is another widely used detection method; however, ELISA has a detection sensitivity of only 103–104 CFU/mL for F. tularensis (Grounow et al., 2000).
70
S.-H. Seo et al. / Biosensors and Bioelectronics 64 (2015) 69–73
Several studies have reported higher detection sensitivity for bacterial pathogens using ELISA techniques, but there are additional steps and accessories required. Immuno-magnetic microbeads (Liu et al., 2001), electrode-based technologies (Liebana et al., 2009), and polyacrylonitrile coated with antibody (Chattopadhyay et al., 2013) have all been suggested as alternative methods for rapid detection of pathogens. Single-walled carbon nanotubes conjugated with horseradish peroxidase (HRP) have also been reported to be used as a labeling platform in ELISA for the sensitive detection of pathogen (Chunglok et al., 2011). Since the detection sensitivity of conventional methods is not sufficient to allow for the detection of the bacteria and diagnosis of infected patients in the early stage; therefore, improvement upon conventional methods or development of a novel sensing method with high sensitivity is urgently needed. Previously, we introduced a novel detection system called oligonucleotide-linked immunosorbent assay (OLISA). Instead of HRP producing the absorbance signal as in ELISA, OLISA uses DNA strands to generate the fluorescent detection signal (Han et al., 2013). In addition, we found that gold nanoparticles coated with multiple strands of the signal generator lead to additional amplification of the fluorescence signal. We also demonstrated that this gold nanoparticle-based oligonucleotide-linked immunosorbent assay (GNP-OLISA) could be used for quantitative analysis of cancer biomarkers with 70 to 100-fold higher detection sensitivity than ELISA (Han et al., 2012). In the present study, we modified the GNP-OLISA procedure to utilize a single antibody instead of the antibody sandwich requiring two different antibodies to detect a bio-threat bacterial strain, F. tularensis. The adjusted GNP-OLISA was used for detection of the bacteria not only in buffer but also in rabbit serum samples, which closely approximate clinical samples. The performance of GNP-OLISA was compared with ELISA to evaluate the detection sensitivity of GNP-OLISA. Finally, the specificity of the assay was examined by detecting the target bacteria in the presence of other non-target bacterial species a competitors.
2. Materials and methods 2.1. Bacteria F. tularensis (subspecies holarctica live vaccine strain) was cultured in 250 mL of LB broth containing IsoVitaleX (BD Biosciences, USA) or brain heart infusion broth (BD Bioscience, USA) for 72 h at 37 °C with shaking (200 rpm). To determine the CFUs of F. tularensis present, bacteria were diluted with phosphate buffed saline (PBS) and spread onto chocolate agar plates (HANIL, Korea). Plates were incubated at 37 °C for 72 h and colonies were counted. Bacillus anthracis (Sterne) was cultured in 25 mL of Leighton-Doi broth at 37 °C for 24 h with shaking (200 rpm). Stock preparation and calculations were performed as above, except that LB media instead of chocolate agar plates was used. 2.2. Preparation of F. tularensis LPS After culturing F. tularensis, the lipopolysaccharide (LPS) on the surface of F. tularensis was extracted and purified using a commercially available kit following the manufacturer′s recommendation (Intron Biotechnology, Korea).
PBS containing 0.05% Tween-20 3 times and 200 mL of 1% BSA in PBS was then added. After 1-h incubation at room temperature, the plate was washed and 100 mL of anti-F. tularensis LPS antibody (Abcam, ab2033, UK) was added to the wells and incubated for 1 h. After washing, 100 mL of HRP-conjugated rat anti-mouse IgG2a antibody was added to each well and the plate was incubated for another hour. After rinsing with PBS containing 0.05% Tween-20, 100 mL of TMB solution was added and the development reactions were stopped using 100 mL of 2 N H2SO4 solution. The optical density of each well was measured at 450 nm using a SUNRISE absorbance plate reader (TECAN, Austria). For fluorescence-based ELISA, 20 mL of Ampliflu™ Red (10 mM, Sigma-Aldrich, USA) was added instead of the TMB solution. Fluorescence intensity was measured using a SpectraMaxs M2e (Molecular Devices, USA), with excitation at 530 nm and emission at 585 nm, respectively. 2.4. Preparation of Ab-GNP-DNA complex An Ab-GNP-DNA probe complex was prepared as described previously (Han et al., 2012). Briefly, 2 mL of 10-nm gold colloid (British Biocell International, UK) was incubated for 15 min at room temperature with 20 mg anti-F. tularensis LPS (Abcam, ab2033, UK). After adding 5 mL of 10% Tween-20, 50 mL of 50 mM 5′-thiolated DNA oligonucleotide probe (5′-AACCACAGTG-3′, Bioneer, Korea) and 20 mL of 0.5 M phosphate buffer, pH 8.0, the solution was incubated overnight at 4°C. Then, the solution was salted by adding 30 mL of 5 M NaCl, and incubated at room temperature for 1 h. The Ab-GNP-DNA probe was stabilized by adding 500 mL of 1% BSA. The solution was centrifuged for 10 min at 18,000g, the supernatant was removed, and the pellet was resuspended with 1 mL of 1% BSA in PBS. All incubation procedures were conducted under gentle agitation. 2.5. Transmission electron microscope (TEM) analysis To prepare the negative-staining control samples, 5 μL of emulsion (F. tularensisþAb-GNP-DNA probe) were applied to a formvar-coated 200 mesh copper grid for 1 min; the emulsion was then removed using filter paper. The grid was allowed to dry for 10 min and then stained with 2% uranyl acetate for 10 s; the staining solution was removed using filter paper and dried for a further 10 min. To make microtome sections of each sample, pellets (F. tularensisþ Ab-GNP-DNA probe) were fixed with 2.5% glutaraldehyde in 0.1M phosphate buffer at 4 °C for 2 h, washed with 0.1 M phosphate buffer (3 times), and fixed with 1% osmium tetroxide in 0.1 M phosphate buffer for 2 h at 4 °C. After washing, the pellets were dehydrated by soaking with 50%, 70%, 80%, 90%, and 95% ethanol, sequentially, for 5 min (2 times each concentration) and 100% ethanol for 10 min (3 times). The residual ethanol in the pellets was removed by rinsing with propylene oxide for 20 min (3 times). For embedment, the pellets were dipped in prophylene-Epon 812 solutions (prophylene:Epon¼ 3:1 and 1:1 for 30 min; prophylene:Epon¼1:2 overnight) and pure Epon 812 resin for 2 h. The sample was moved to a silicon plate mold and polymerized in an oven at 60 °C for 48 h incubation. The sample was then trimmed and sectioned using ultra microtome (UC7, Leica Microsystems, Germany). The sectioned sample was moved onto 100 mesh copper grid and double stained with 2% uranyl acetate and lead citrate. Both of the negative stained and sectioned samples were analyzed using transmission electron microscope (Libra 120, Carl Zeiss, Germany) at 120 kV.
2.3. ELISA 2.6. GNP-OLISA PBS-diluted F. tularensis LPS, F. tularensis, or PBS were used to coat a 96-well plate (Nunc, Denmark), which was incubated at room temperature for 1 h. The plate was washed with 300 mL of
GNP-OLISAs were performed using a protocol modified from our previous report (Han et al., 2012). Briefly, F. tularensis LPS and
S.-H. Seo et al. / Biosensors and Bioelectronics 64 (2015) 69–73
F. tularensis were serially diluted with PBS or rabbit serum (Abcam, UK) and placed in a Maxisorp 96-well microplate (Thermo Scientific, Germany). After incubation for 1 h at room temperature, the wells were washed with 300 mL of PBS containing 0.05% Tween-20 and blocked with 200 mL of 1% BSA in PBS for 1 h at room temperature. After washing with PBS containing 0.05% Tween-20, 100 mL of 1:4 diluted Ab-GNP-DNA probes were added to each well and incubated for 1 h at room temperature. After washing with PBS containing 0.05% Tween-20, 100 mL of RNase H solution (4 mM MgCl2 [Sigma], 10 mM DTT [Sigma], 30 U RNase H [Takara], 1 mM F-RNA probe-Q (FITC-5′-CACUGUGGUU-3′-BHQ1) [Bioneer], and 0.4 U Protector Protease Inhibitor [Roche]) were added. Fluorescence intensities were measured using a SpectraMaxs M2e (Molecular Devices, USA), with excitation and emission wavelengths of 485 nm and 538 nm, respectively.
3. Results and discussion 3.1. Preparation of Ab-GNP-DNA complex for GNP-OLISA of F. tularensis The previously developed GNP-OLISA procedure uses the antibody sandwich format comprised of both capture and detection antibodies to sensitivity detect a cancer biomarker (Han et al., 2012). For many species of bacteria, including F. tularensis, there are few available matched-antibody pairs that can be used for capture and detection in a sandwich-type detection assay. Thus, we modified the GNP-OLISA to become a form of direct-detection ELISA as illustrated Fig. 1. Briefly, the bacteria are absorbed to the well of a plate and then detected by an antibody conjugated to GNP coated with thiolated DNA strands (Ab-GNP-DNA). To generate the detection signal, the DNA strands are released by treatment with dithiothreitol (DTT) and hybridized with the complementary RNA probe (F-RNA-Q) appended with fluorophore at one end and a quencher at the other end. RNase H is then added as it cleaves only the RNA probe portion of the DNA/RNA duplex and recovers the fluorescence from the cleaved quencher. To obtain a successful assay signal, the Ab-GNP-DNA complex should bind effectively to the bacterial surface. Before performing the GNP-OLISA, we characterized the binding ability of the AbGNP-DNA complex to the target antigen (LPS) on the surface of F. tularensis using transmission electron microscopy (TEM) (Fig. 2). The TEM image shows that the antibody complex bound strongly to the target antigen on the bacteria and that the amount of the antibody complex bound to the bacterial surface was sufficient for the generation of detection signal in the following GNP-OLISA.
71
3.2. Sensitivity of GNP-OLISA for detection F. tularensis in buffer solutions After preparation of the Ab-GNP-DNA complex, we first examined whether GNP-OLISA using the antibody complex could quantitatively analyze the target antigen. As shown in Fig. 3A, the fluorescence intensity increased linearly as the concentration of the antigen increased (triangles in Fig. 3A). To evaluate the performance of the GNP-OLISA in comparison with ELISA, we also carried out an ELISA for the detection of F. tularensis LPS in which HRP-conjugated secondary antibody was used to produce the analytical signal (diamonds in Fig. 3A). While the absorbance signal increased linearly depending on the concentration of the marker and was as reliable as the fluorescence signal in GNPOLISA, higher concentrations of bacterial antigen were required to distinguish the detection signal from the background in ELISA as compared to the GNP-OLISA. To assess the detection sensitivity of the GNP-OLISA, we determined the limit of detection (LOD) for the marker by interpolating the curve using the average value of the blank plus 3 times the standard deviation of the blank. The LOD values from ELISA and GNP-OLISA for F. tularensis LPS were 72.47 710.61 ng/mL and 0.0467 0.005 ng/mL, respectively. The approximately 1500-fold improvement in the LOD value for the GNP-OLISA compared with that of ELISA suggests that GNP-OLISA is a highly sensitive method for the detection of F. tularensis LPS. To confirm that the improvement in the LOD by GNP-OLISA was not due to the difference in detection modality (fluorescence for GNPOLISA vs. absorbance for ELISA), we a performed an additional ELISA experiment using a fluorogenic substrate for the HRP reaction. The detection sensitivity of F. tularensis and LPS by ELISA based on fluorescence signal was determined as similar detection level (LOD ¼17.99 71.74 ng/mL for LPS, 838.41 7169.63 CFU/mL for F. tularensis) as compared to the LOD observed for the absorbance-based ELISA assay. (Table S1 and Fig. S1 in Supplementary data). The GNP-OLISA method still showed a 391-fold higher sensitivity than the ELISA for F. tularensis LPS performed using fluorescence signals, indicating that the improved detection sensitivity in GNP-OLISA was not due to the detection modality used, but was likely instead due to the multiple signal-generating DNA strands on gold nanoparticles. After observing that the target antigen could be sensitively detected by GNPOLISA, we examined how well the assay performed in the detection of F. tularensis bacteria suspended in buffer solutions. The bacteria were adsorbed on microwells and incubated with the AbGNP-DNA complex. After the unbound complex was removed by a washing step, the detection signal was generated by the RNase H-mediated reaction in the presence of the RNA probe. The
Fig. 1. Schematic process of gold nanoparticle (GNP)-oligonucleotide-linked immunosorbent assay (OLISA). (a) Plate coated with bacteria is recognized by GNP-DNA probe linked specific antibody (Ab-GNP-DNA probe). The following components are added sequentially: dithiothreitol (DTT), RNA probe conjugated with fluorophore and quencher (F-RNA-Q), and RNase H. (b) DTT allows the release of DNA probes from Ab-GNP-DNA complex and (c) DNA probes are hybridized with complementary F-RNA-Q. (d) RNase H cleaves RNA in F-RNA-Q/DNA probe duplex which frees the fluorophore from the quencher. Finally, fluorescence intensities are measured using a fluorescence reader with the appropriate excitation and emission filters.
72
S.-H. Seo et al. / Biosensors and Bioelectronics 64 (2015) 69–73
Fig. 2. TEM analysis of binding of the Ab-GNP-DNA probe to F. tularensis LVS using sectioned (A) and unsectioned negatively-stained samples (B). Black dots are Ab-GNP-DNA complex (red arrows) bound to the thin layer of the bacterial cell wall (blue arrows). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Table 1 Limit of detection values from ELISA and GNP-OLISA using various F. tularensis samples. Sample type
Detection method
Limit of detection
F. tularensis LPS
ELISA GNP-OLISA ELISA GNP-OLISA GNP-OLISA
72.477 10.61 ng/mL 17.56 0.0467 0.005 ng/mL 3.84 872.457 196.97 CFU/mL 22.58 23.477 1.83 CFU/mL 7.8 29.637 2.54 CFU/mL 8.57
F. tularensis Rabbit serum spiked F. tularensis Rabbit serum spiked F. tularensisþ B. anthracis
GNP-OLISA
30.707 2.34 CFU/mL
%CV
7.62
bacteria than ELISA. It is worth mentioning that GNP-OLISA also showed more than one order of magnitude higher sensitivity than the RT-PCR assay reported in our previous study (Dauphin et al., 2011), although comparison with other assays in the literature is difficult due to the differences in experimental conditions. In addition, a semi-log quantitative graph of the result of GNP-OLISA for F. tularensis detection showed a linear correlation between the fluorescence intensity and F. tularensis concentration, even for low levels of F. tularensis (8.30 CFU/mL, Fig. 3B). This result indicates that GNP-OLISA can detect F. tularensis levels below 10 CFU in buffer solutions, which is sufficient for sensing an infectious dose of F. tularensis. 3.3. GNP-OLISA for F. tularensis in rabbit sera Fig. 3. Comparison of the sensitivity of ELISA and GNP-OLISA for the detection of F. tularensis LPS (A) and F. tularensis (B) model. The quantitative detection curve for F. tularensis LVS LPS or F. tularensis LVS determined utilizing ELISA (◆) and GNP-OLISA (▲) using FB11 antibody. LVS: live vaccine strain, LPS: lipopolysaccharide, GNPOLISA: gold nanoparticle-signal enhanced oligonucleotide-linked immunosorbent assay, O.D.: optical density. RFU: relative fluorescence unit. Each data point represents the mean 7standard deviations (SD) of values obtained from three experiments.
fluorescence signal proportionally increased with the number of bacteria in the solution. The LOD value of GNP-OLISA was determined to be 23.477 1.83 CFU/mL (triangles in Fig. 3B and Table 1). When compared to the LOD value of ELISA (872.45 7 196.97 CFU/mL, diamonds in Fig. 3B and Table 1), GNP-OLISA was calculated to be 37-fold more sensitive for the detection of the
To evaluate the detection sensitivity of GNP-OLISA for F. tularensis in biological sample matrices, we used GNP-OLISA to detect F. tularensis suspended in rabbit sera (Fig. 4A). Quantitative semi-log graphs obtained from analysis of GNP-OLISA for analysis of rabbit sera-diluted F. tularensis showed a similar linearity as that of the buffer-diluted bacteria (Fig. 4A). The LOD value for GNPOLISA of rabbit sera-spiked samples was 29.63 72.54 CFU/mL (Table 1), indicating that the detection sensitivity of GNP-OLISA for F. tularensis may not be affected by biological sample matrices such as sera. The background signal in the rabbit sera-spiked samples during the assay was slightly higher than that of PBSdiluted samples, presumably due to auto-fluorescence from serum components as reported previously (Han et al., 2012).
S.-H. Seo et al. / Biosensors and Bioelectronics 64 (2015) 69–73
73
in buffer solutions but also in biological matrices such as rabbit sera. In particular, GNP-OLISA enables detection of 23.47 71.83 CFU/mL of F. tularensis in 100 μL samples, which is below the infectious dose. Thus, this novel technology easily fulfills the need for the early identification of F. tularensis-in order to prevent bio-terrorism events or disease outbreak as well as to effectively monitor decontamination of the infected area. Although the GNP-OLISA method was used to detect only one pathogen in this study, multiplex GNP-OLISA for detection of multiple pathogens may potentially be developed by employing DNA strands and RNA probes with different sequences, which will expand the utility of the GNP-OLISA method for clinical applications. Moreover, since only one antibody is required for the GNP-OLISA system, this method would be very useful to sensitively detect various kinds of target bacteria that do not have an appropriate pair of specific antibodies to capture and detect antigen in a traditional ELISA assay. Thus, the GNP-OLISA protocol used in this study shows great potential for use in clinical diagnosis as well as research applications that require detection of low level of bacteria.
Acknowledgments This study was supported by Korea National institute of Health (2013-NG45003-00), and funded by the Korea Ministry of Health and Welfare. We would like to thank Ki-Ju Choi for technical assistance with TEM analysis. We also thank to Kyu Jam Hwang for bacterial characterization.
Appendix A. Supplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2014.08.038. Fig. 4. The detection sensitivity of GNP-OLISA for F. tularensis in biological samples. GNP-OLISA was performed using rabbit serum-diluted F. tularensis (A) and F. tularensis diluted in rabbit serum including 2.5 104 CFUs of Bacillus anthracis Sterne (B). Each data point represents the mean 7 standard deviation (SD) of the values obtained from three experiments.
3.4. Specificity of GNP-OLISA to detect F. tularensis Finally, we evaluated the detection specificity of GNP-OLISA for a specific target bacteria in the presence of other bacteria such as B. anthracis Sterne (2.4 104 CFU), another top-priority bio-threat pathogen categorized by the CDC in analytical serum samples (Fig. 4B). The LOD value of GNP-OLISA for F. tularensis in rabbit sera also containing B. anthracis Sterne did not vary significantly from that for the samples containing only the target bacteria, implying that the assay has a high sensing specificity for the target bacteria in biological samples representing mixed infections (Table 1).
4. Conclusions Compared with conventional sensing methods, GNP-OLISA confers enhanced sensitivity in quantitation of bacteria not only
References Alaeddini, R., 2012. Forensic Sci. Int. Genet. 6 (3), 297–305. Chattopadhyay, S., Kaur, A., Jain, S., Singh, H., 2013. Biosens. Bioelectron. 45, 274– 280. Chunglok, W., Wuragil, D.K., Oaew, S., Somasundrum, M., Surareungchai, W., 2011. Biosens. Bioelectron. 26, 3584–3589. Dauphin, L.A., Walker, R.E., Petersen, J.M., Brown, M.D., 2011. Diagn. Microbiol. Infect. Dis. 70, 299–306. Ellis, J., Oyston, P.C., Green, M., Titball, R.W., 2002. Clin. Microbiol. Rev. 15 (4), 631– 646. Franz, D.R., Jahrling, P.B., McClain, D.J., Hoover, D.L., Bryne, W.R., Pavlin, J.A., Christopher, G.W., Cieslak, T.J., Friedlander, A.M., Eitzen, E.M., 2001. Clin. Lab. Methods 21 (3), 435–473. Froude, J.W., Stiles, B., Pelat, T., Thllier, P., 2011. mAbs 3 (6), 517–527. Grounow, R., Splettstoesser, W., McDonald, S., Otterbein, C., O'Brien, T., Morgan, C., Aldrich, J., Hofer, E., Finke, E.-J., Meyer, H., 2000. Clin. Diagn. Lab. Immunol. 7 (1), 86–90. Han, K.C., Yang, E.G., Ahn, D.R., 2012. Chem. Commun. 48, 5895–5897. Han, K.C., Yang, E.G., Ahn, D.R., 2013. Biosens. Bioelectron. 50, 421–424. Liebana, S., Lermo, A., Campoy, S., Cortes, M.P., Pividori, M.I., 2009. Biosens. Bioelectron. 25, 510–513. Liu, Y., Che, Y., Li, Y., 2001. Sens. Actuators B 72, 214–218. Pohanka, M., Hubalek, M., Neubauerova, V., Macela, A., Faldyna, M., Bandouchova, H., Pikula, J., 2008. Vet. Med. 53 (11), 585–594.
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Highly sensitive detection of a bio-threat pathogen by gold nanoparticle-based oligonucleotide-linked immunosorbent assay Sang-Hwan Seo a,1, Young-Ran Lee a,1, Jun Ho Jeon a,1, Yi-Rang Hwang a, Pil-Gu Park a, Dae-Ro Ahn b,c, Ki-Cheol Han b, Gi-Eun Rhie a, Kee-Jong Hong a,n a Division of High-Risk Pathogen Research, Center for Infectious Diseases, Korea National Institute of Health, Osong Health Technology Administration Complex, Cheongwon, Chungcheongbuk-do 363-951, Republic of Korea b Center for Theragnosis, Biomedical Research Institute, Korea Institute of Science and Technology, Hwarangno 14-gil 5, Seongbuk-gu, Seoul 136-791, Republic of Korea c Department of Biological Chemistry, KIST Campus, Korea University of Science and Technology, Hwarangno 14-gil 5, Seongbuk-gu, Seoul 136-791, Republic of Korea
art ic l e i nf o
a b s t r a c t
Article history: Received 25 May 2014 Received in revised form 7 August 2014 Accepted 19 August 2014 Available online 23 August 2014
Francisella (F.) tularensis causes the zoonotic disease tularemia and categorized as one of the highestpriority biological agents. The sensing approaches utilized by conventional detection methods, including enzyme-linked immunosorbent assay (ELISA), are not sensitive enough to identify an infectious dose of this high-risk pathogen due to its low infective dose. As an attempt to detect F. tularensis with high sensitivity, we utilized the highly sensitive immunoassay system named gold nanoparticle-based oligonucleotide-linked immunosorbent assay (GNP-OLISA) which uses antibody-gold nanoparticles conjugated with DNA strands as a signal generator and RNA oligonucleotides appended with a fluorophore as a quencher for signal amplification. We modified the GNP-OLISA for the detection F. tularensis to utilize one antibody for both the capture of the target and for signal generation instead of using two different antibodies, which are usually employed to construct the antibody sandwich in the ELISA. The GNP-OLISA showed 37-fold higher sensitivity compared with ELISA and generated very consistent detection results in the sera. In addition, the detection specificity was not affected by the presence of non-target bacteria, suggesting that GNP-OLISA can be used as a sensitive detection platform for monitoring high-risk pathogens thereby overcoming the limit of the conventional assay system. & 2014 Elsevier B.V. All rights reserved.
Keywords: Francisella tularensis Bacteria ELISA OLISA Detection sensitivity Gold-nanoparticle
1. Introduction Francisella (F.) tularensis, the causative agent of the zoonotic disease tularemia, is categorized as one of the highest-priority biological agents of concern by the U.S. Center for Disease Control and Prevention (CDC) because it has a very low infectious dose (10 colony-forming units (CFU) of F. tularensis) (Ellis et al., 2002; Froude et al., 2011; Franz et al., 2001). Vaccination is the most efficient weapon against F. tularensis infection, but no vaccine has been officially approved by the U.S. Food and Drug Administration to date. Treatment of patients with proper antibiotics during the early stages of infection and decontamination of pathogen-containing environments with selected agents are the only ways to the control outbreak to the disease and to reduce risk of public n
Corresponding author. Tel.: þ 82 437198271; fax: þ82 437198309. E-mail address: [email protected] (K.-J. Hong). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.bios.2014.08.038 0956-5663/& 2014 Elsevier B.V. All rights reserved.
health. Identification of F. tularensis bacteria in biological and environmental samples is essential for efficient clinical treatment and appropriate decontamination by public health official. Hence, a sensitive and specific detection method to detect tiny amounts of F. tularensis is required. The conventional detection method for F. tularensisis requires at least three days of bacterial cultivation and identification using specific media (Pohanka et al., 2008). As an alternative, real-time polymerase chain reaction (RT-PCR) has been utilized due to its selectivity, speed, and sensitivity. However, the limit of detection (LOD) using RT-PCR is 103 CFU/mL for F. tularensis and it does not meet clinical demand (Dauphin et al., 2011). In addition, RT-PCR can generate false-positive signals if contaminated or false-negative signals if the target gene is mutated or if the sample contains Taq polymerase inhibitors (Alaeddini, 2012). ELISA is another widely used detection method; however, ELISA has a detection sensitivity of only 103–104 CFU/mL for F. tularensis (Grounow et al., 2000).
70
S.-H. Seo et al. / Biosensors and Bioelectronics 64 (2015) 69–73
Several studies have reported higher detection sensitivity for bacterial pathogens using ELISA techniques, but there are additional steps and accessories required. Immuno-magnetic microbeads (Liu et al., 2001), electrode-based technologies (Liebana et al., 2009), and polyacrylonitrile coated with antibody (Chattopadhyay et al., 2013) have all been suggested as alternative methods for rapid detection of pathogens. Single-walled carbon nanotubes conjugated with horseradish peroxidase (HRP) have also been reported to be used as a labeling platform in ELISA for the sensitive detection of pathogen (Chunglok et al., 2011). Since the detection sensitivity of conventional methods is not sufficient to allow for the detection of the bacteria and diagnosis of infected patients in the early stage; therefore, improvement upon conventional methods or development of a novel sensing method with high sensitivity is urgently needed. Previously, we introduced a novel detection system called oligonucleotide-linked immunosorbent assay (OLISA). Instead of HRP producing the absorbance signal as in ELISA, OLISA uses DNA strands to generate the fluorescent detection signal (Han et al., 2013). In addition, we found that gold nanoparticles coated with multiple strands of the signal generator lead to additional amplification of the fluorescence signal. We also demonstrated that this gold nanoparticle-based oligonucleotide-linked immunosorbent assay (GNP-OLISA) could be used for quantitative analysis of cancer biomarkers with 70 to 100-fold higher detection sensitivity than ELISA (Han et al., 2012). In the present study, we modified the GNP-OLISA procedure to utilize a single antibody instead of the antibody sandwich requiring two different antibodies to detect a bio-threat bacterial strain, F. tularensis. The adjusted GNP-OLISA was used for detection of the bacteria not only in buffer but also in rabbit serum samples, which closely approximate clinical samples. The performance of GNP-OLISA was compared with ELISA to evaluate the detection sensitivity of GNP-OLISA. Finally, the specificity of the assay was examined by detecting the target bacteria in the presence of other non-target bacterial species a competitors.
2. Materials and methods 2.1. Bacteria F. tularensis (subspecies holarctica live vaccine strain) was cultured in 250 mL of LB broth containing IsoVitaleX (BD Biosciences, USA) or brain heart infusion broth (BD Bioscience, USA) for 72 h at 37 °C with shaking (200 rpm). To determine the CFUs of F. tularensis present, bacteria were diluted with phosphate buffed saline (PBS) and spread onto chocolate agar plates (HANIL, Korea). Plates were incubated at 37 °C for 72 h and colonies were counted. Bacillus anthracis (Sterne) was cultured in 25 mL of Leighton-Doi broth at 37 °C for 24 h with shaking (200 rpm). Stock preparation and calculations were performed as above, except that LB media instead of chocolate agar plates was used. 2.2. Preparation of F. tularensis LPS After culturing F. tularensis, the lipopolysaccharide (LPS) on the surface of F. tularensis was extracted and purified using a commercially available kit following the manufacturer′s recommendation (Intron Biotechnology, Korea).
PBS containing 0.05% Tween-20 3 times and 200 mL of 1% BSA in PBS was then added. After 1-h incubation at room temperature, the plate was washed and 100 mL of anti-F. tularensis LPS antibody (Abcam, ab2033, UK) was added to the wells and incubated for 1 h. After washing, 100 mL of HRP-conjugated rat anti-mouse IgG2a antibody was added to each well and the plate was incubated for another hour. After rinsing with PBS containing 0.05% Tween-20, 100 mL of TMB solution was added and the development reactions were stopped using 100 mL of 2 N H2SO4 solution. The optical density of each well was measured at 450 nm using a SUNRISE absorbance plate reader (TECAN, Austria). For fluorescence-based ELISA, 20 mL of Ampliflu™ Red (10 mM, Sigma-Aldrich, USA) was added instead of the TMB solution. Fluorescence intensity was measured using a SpectraMaxs M2e (Molecular Devices, USA), with excitation at 530 nm and emission at 585 nm, respectively. 2.4. Preparation of Ab-GNP-DNA complex An Ab-GNP-DNA probe complex was prepared as described previously (Han et al., 2012). Briefly, 2 mL of 10-nm gold colloid (British Biocell International, UK) was incubated for 15 min at room temperature with 20 mg anti-F. tularensis LPS (Abcam, ab2033, UK). After adding 5 mL of 10% Tween-20, 50 mL of 50 mM 5′-thiolated DNA oligonucleotide probe (5′-AACCACAGTG-3′, Bioneer, Korea) and 20 mL of 0.5 M phosphate buffer, pH 8.0, the solution was incubated overnight at 4°C. Then, the solution was salted by adding 30 mL of 5 M NaCl, and incubated at room temperature for 1 h. The Ab-GNP-DNA probe was stabilized by adding 500 mL of 1% BSA. The solution was centrifuged for 10 min at 18,000g, the supernatant was removed, and the pellet was resuspended with 1 mL of 1% BSA in PBS. All incubation procedures were conducted under gentle agitation. 2.5. Transmission electron microscope (TEM) analysis To prepare the negative-staining control samples, 5 μL of emulsion (F. tularensisþAb-GNP-DNA probe) were applied to a formvar-coated 200 mesh copper grid for 1 min; the emulsion was then removed using filter paper. The grid was allowed to dry for 10 min and then stained with 2% uranyl acetate for 10 s; the staining solution was removed using filter paper and dried for a further 10 min. To make microtome sections of each sample, pellets (F. tularensisþ Ab-GNP-DNA probe) were fixed with 2.5% glutaraldehyde in 0.1M phosphate buffer at 4 °C for 2 h, washed with 0.1 M phosphate buffer (3 times), and fixed with 1% osmium tetroxide in 0.1 M phosphate buffer for 2 h at 4 °C. After washing, the pellets were dehydrated by soaking with 50%, 70%, 80%, 90%, and 95% ethanol, sequentially, for 5 min (2 times each concentration) and 100% ethanol for 10 min (3 times). The residual ethanol in the pellets was removed by rinsing with propylene oxide for 20 min (3 times). For embedment, the pellets were dipped in prophylene-Epon 812 solutions (prophylene:Epon¼ 3:1 and 1:1 for 30 min; prophylene:Epon¼1:2 overnight) and pure Epon 812 resin for 2 h. The sample was moved to a silicon plate mold and polymerized in an oven at 60 °C for 48 h incubation. The sample was then trimmed and sectioned using ultra microtome (UC7, Leica Microsystems, Germany). The sectioned sample was moved onto 100 mesh copper grid and double stained with 2% uranyl acetate and lead citrate. Both of the negative stained and sectioned samples were analyzed using transmission electron microscope (Libra 120, Carl Zeiss, Germany) at 120 kV.
2.3. ELISA 2.6. GNP-OLISA PBS-diluted F. tularensis LPS, F. tularensis, or PBS were used to coat a 96-well plate (Nunc, Denmark), which was incubated at room temperature for 1 h. The plate was washed with 300 mL of
GNP-OLISAs were performed using a protocol modified from our previous report (Han et al., 2012). Briefly, F. tularensis LPS and
S.-H. Seo et al. / Biosensors and Bioelectronics 64 (2015) 69–73
F. tularensis were serially diluted with PBS or rabbit serum (Abcam, UK) and placed in a Maxisorp 96-well microplate (Thermo Scientific, Germany). After incubation for 1 h at room temperature, the wells were washed with 300 mL of PBS containing 0.05% Tween-20 and blocked with 200 mL of 1% BSA in PBS for 1 h at room temperature. After washing with PBS containing 0.05% Tween-20, 100 mL of 1:4 diluted Ab-GNP-DNA probes were added to each well and incubated for 1 h at room temperature. After washing with PBS containing 0.05% Tween-20, 100 mL of RNase H solution (4 mM MgCl2 [Sigma], 10 mM DTT [Sigma], 30 U RNase H [Takara], 1 mM F-RNA probe-Q (FITC-5′-CACUGUGGUU-3′-BHQ1) [Bioneer], and 0.4 U Protector Protease Inhibitor [Roche]) were added. Fluorescence intensities were measured using a SpectraMaxs M2e (Molecular Devices, USA), with excitation and emission wavelengths of 485 nm and 538 nm, respectively.
3. Results and discussion 3.1. Preparation of Ab-GNP-DNA complex for GNP-OLISA of F. tularensis The previously developed GNP-OLISA procedure uses the antibody sandwich format comprised of both capture and detection antibodies to sensitivity detect a cancer biomarker (Han et al., 2012). For many species of bacteria, including F. tularensis, there are few available matched-antibody pairs that can be used for capture and detection in a sandwich-type detection assay. Thus, we modified the GNP-OLISA to become a form of direct-detection ELISA as illustrated Fig. 1. Briefly, the bacteria are absorbed to the well of a plate and then detected by an antibody conjugated to GNP coated with thiolated DNA strands (Ab-GNP-DNA). To generate the detection signal, the DNA strands are released by treatment with dithiothreitol (DTT) and hybridized with the complementary RNA probe (F-RNA-Q) appended with fluorophore at one end and a quencher at the other end. RNase H is then added as it cleaves only the RNA probe portion of the DNA/RNA duplex and recovers the fluorescence from the cleaved quencher. To obtain a successful assay signal, the Ab-GNP-DNA complex should bind effectively to the bacterial surface. Before performing the GNP-OLISA, we characterized the binding ability of the AbGNP-DNA complex to the target antigen (LPS) on the surface of F. tularensis using transmission electron microscopy (TEM) (Fig. 2). The TEM image shows that the antibody complex bound strongly to the target antigen on the bacteria and that the amount of the antibody complex bound to the bacterial surface was sufficient for the generation of detection signal in the following GNP-OLISA.
71
3.2. Sensitivity of GNP-OLISA for detection F. tularensis in buffer solutions After preparation of the Ab-GNP-DNA complex, we first examined whether GNP-OLISA using the antibody complex could quantitatively analyze the target antigen. As shown in Fig. 3A, the fluorescence intensity increased linearly as the concentration of the antigen increased (triangles in Fig. 3A). To evaluate the performance of the GNP-OLISA in comparison with ELISA, we also carried out an ELISA for the detection of F. tularensis LPS in which HRP-conjugated secondary antibody was used to produce the analytical signal (diamonds in Fig. 3A). While the absorbance signal increased linearly depending on the concentration of the marker and was as reliable as the fluorescence signal in GNPOLISA, higher concentrations of bacterial antigen were required to distinguish the detection signal from the background in ELISA as compared to the GNP-OLISA. To assess the detection sensitivity of the GNP-OLISA, we determined the limit of detection (LOD) for the marker by interpolating the curve using the average value of the blank plus 3 times the standard deviation of the blank. The LOD values from ELISA and GNP-OLISA for F. tularensis LPS were 72.47 710.61 ng/mL and 0.0467 0.005 ng/mL, respectively. The approximately 1500-fold improvement in the LOD value for the GNP-OLISA compared with that of ELISA suggests that GNP-OLISA is a highly sensitive method for the detection of F. tularensis LPS. To confirm that the improvement in the LOD by GNP-OLISA was not due to the difference in detection modality (fluorescence for GNPOLISA vs. absorbance for ELISA), we a performed an additional ELISA experiment using a fluorogenic substrate for the HRP reaction. The detection sensitivity of F. tularensis and LPS by ELISA based on fluorescence signal was determined as similar detection level (LOD ¼17.99 71.74 ng/mL for LPS, 838.41 7169.63 CFU/mL for F. tularensis) as compared to the LOD observed for the absorbance-based ELISA assay. (Table S1 and Fig. S1 in Supplementary data). The GNP-OLISA method still showed a 391-fold higher sensitivity than the ELISA for F. tularensis LPS performed using fluorescence signals, indicating that the improved detection sensitivity in GNP-OLISA was not due to the detection modality used, but was likely instead due to the multiple signal-generating DNA strands on gold nanoparticles. After observing that the target antigen could be sensitively detected by GNPOLISA, we examined how well the assay performed in the detection of F. tularensis bacteria suspended in buffer solutions. The bacteria were adsorbed on microwells and incubated with the AbGNP-DNA complex. After the unbound complex was removed by a washing step, the detection signal was generated by the RNase H-mediated reaction in the presence of the RNA probe. The
Fig. 1. Schematic process of gold nanoparticle (GNP)-oligonucleotide-linked immunosorbent assay (OLISA). (a) Plate coated with bacteria is recognized by GNP-DNA probe linked specific antibody (Ab-GNP-DNA probe). The following components are added sequentially: dithiothreitol (DTT), RNA probe conjugated with fluorophore and quencher (F-RNA-Q), and RNase H. (b) DTT allows the release of DNA probes from Ab-GNP-DNA complex and (c) DNA probes are hybridized with complementary F-RNA-Q. (d) RNase H cleaves RNA in F-RNA-Q/DNA probe duplex which frees the fluorophore from the quencher. Finally, fluorescence intensities are measured using a fluorescence reader with the appropriate excitation and emission filters.
72
S.-H. Seo et al. / Biosensors and Bioelectronics 64 (2015) 69–73
Fig. 2. TEM analysis of binding of the Ab-GNP-DNA probe to F. tularensis LVS using sectioned (A) and unsectioned negatively-stained samples (B). Black dots are Ab-GNP-DNA complex (red arrows) bound to the thin layer of the bacterial cell wall (blue arrows). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Table 1 Limit of detection values from ELISA and GNP-OLISA using various F. tularensis samples. Sample type
Detection method
Limit of detection
F. tularensis LPS
ELISA GNP-OLISA ELISA GNP-OLISA GNP-OLISA
72.477 10.61 ng/mL 17.56 0.0467 0.005 ng/mL 3.84 872.457 196.97 CFU/mL 22.58 23.477 1.83 CFU/mL 7.8 29.637 2.54 CFU/mL 8.57
F. tularensis Rabbit serum spiked F. tularensis Rabbit serum spiked F. tularensisþ B. anthracis
GNP-OLISA
30.707 2.34 CFU/mL
%CV
7.62
bacteria than ELISA. It is worth mentioning that GNP-OLISA also showed more than one order of magnitude higher sensitivity than the RT-PCR assay reported in our previous study (Dauphin et al., 2011), although comparison with other assays in the literature is difficult due to the differences in experimental conditions. In addition, a semi-log quantitative graph of the result of GNP-OLISA for F. tularensis detection showed a linear correlation between the fluorescence intensity and F. tularensis concentration, even for low levels of F. tularensis (8.30 CFU/mL, Fig. 3B). This result indicates that GNP-OLISA can detect F. tularensis levels below 10 CFU in buffer solutions, which is sufficient for sensing an infectious dose of F. tularensis. 3.3. GNP-OLISA for F. tularensis in rabbit sera Fig. 3. Comparison of the sensitivity of ELISA and GNP-OLISA for the detection of F. tularensis LPS (A) and F. tularensis (B) model. The quantitative detection curve for F. tularensis LVS LPS or F. tularensis LVS determined utilizing ELISA (◆) and GNP-OLISA (▲) using FB11 antibody. LVS: live vaccine strain, LPS: lipopolysaccharide, GNPOLISA: gold nanoparticle-signal enhanced oligonucleotide-linked immunosorbent assay, O.D.: optical density. RFU: relative fluorescence unit. Each data point represents the mean 7standard deviations (SD) of values obtained from three experiments.
fluorescence signal proportionally increased with the number of bacteria in the solution. The LOD value of GNP-OLISA was determined to be 23.477 1.83 CFU/mL (triangles in Fig. 3B and Table 1). When compared to the LOD value of ELISA (872.45 7 196.97 CFU/mL, diamonds in Fig. 3B and Table 1), GNP-OLISA was calculated to be 37-fold more sensitive for the detection of the
To evaluate the detection sensitivity of GNP-OLISA for F. tularensis in biological sample matrices, we used GNP-OLISA to detect F. tularensis suspended in rabbit sera (Fig. 4A). Quantitative semi-log graphs obtained from analysis of GNP-OLISA for analysis of rabbit sera-diluted F. tularensis showed a similar linearity as that of the buffer-diluted bacteria (Fig. 4A). The LOD value for GNPOLISA of rabbit sera-spiked samples was 29.63 72.54 CFU/mL (Table 1), indicating that the detection sensitivity of GNP-OLISA for F. tularensis may not be affected by biological sample matrices such as sera. The background signal in the rabbit sera-spiked samples during the assay was slightly higher than that of PBSdiluted samples, presumably due to auto-fluorescence from serum components as reported previously (Han et al., 2012).
S.-H. Seo et al. / Biosensors and Bioelectronics 64 (2015) 69–73
73
in buffer solutions but also in biological matrices such as rabbit sera. In particular, GNP-OLISA enables detection of 23.47 71.83 CFU/mL of F. tularensis in 100 μL samples, which is below the infectious dose. Thus, this novel technology easily fulfills the need for the early identification of F. tularensis-in order to prevent bio-terrorism events or disease outbreak as well as to effectively monitor decontamination of the infected area. Although the GNP-OLISA method was used to detect only one pathogen in this study, multiplex GNP-OLISA for detection of multiple pathogens may potentially be developed by employing DNA strands and RNA probes with different sequences, which will expand the utility of the GNP-OLISA method for clinical applications. Moreover, since only one antibody is required for the GNP-OLISA system, this method would be very useful to sensitively detect various kinds of target bacteria that do not have an appropriate pair of specific antibodies to capture and detect antigen in a traditional ELISA assay. Thus, the GNP-OLISA protocol used in this study shows great potential for use in clinical diagnosis as well as research applications that require detection of low level of bacteria.
Acknowledgments This study was supported by Korea National institute of Health (2013-NG45003-00), and funded by the Korea Ministry of Health and Welfare. We would like to thank Ki-Ju Choi for technical assistance with TEM analysis. We also thank to Kyu Jam Hwang for bacterial characterization.
Appendix A. Supplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2014.08.038. Fig. 4. The detection sensitivity of GNP-OLISA for F. tularensis in biological samples. GNP-OLISA was performed using rabbit serum-diluted F. tularensis (A) and F. tularensis diluted in rabbit serum including 2.5 104 CFUs of Bacillus anthracis Sterne (B). Each data point represents the mean 7 standard deviation (SD) of the values obtained from three experiments.
3.4. Specificity of GNP-OLISA to detect F. tularensis Finally, we evaluated the detection specificity of GNP-OLISA for a specific target bacteria in the presence of other bacteria such as B. anthracis Sterne (2.4 104 CFU), another top-priority bio-threat pathogen categorized by the CDC in analytical serum samples (Fig. 4B). The LOD value of GNP-OLISA for F. tularensis in rabbit sera also containing B. anthracis Sterne did not vary significantly from that for the samples containing only the target bacteria, implying that the assay has a high sensing specificity for the target bacteria in biological samples representing mixed infections (Table 1).
4. Conclusions Compared with conventional sensing methods, GNP-OLISA confers enhanced sensitivity in quantitation of bacteria not only
References Alaeddini, R., 2012. Forensic Sci. Int. Genet. 6 (3), 297–305. Chattopadhyay, S., Kaur, A., Jain, S., Singh, H., 2013. Biosens. Bioelectron. 45, 274– 280. Chunglok, W., Wuragil, D.K., Oaew, S., Somasundrum, M., Surareungchai, W., 2011. Biosens. Bioelectron. 26, 3584–3589. Dauphin, L.A., Walker, R.E., Petersen, J.M., Brown, M.D., 2011. Diagn. Microbiol. Infect. Dis. 70, 299–306. Ellis, J., Oyston, P.C., Green, M., Titball, R.W., 2002. Clin. Microbiol. Rev. 15 (4), 631– 646. Franz, D.R., Jahrling, P.B., McClain, D.J., Hoover, D.L., Bryne, W.R., Pavlin, J.A., Christopher, G.W., Cieslak, T.J., Friedlander, A.M., Eitzen, E.M., 2001. Clin. Lab. Methods 21 (3), 435–473. Froude, J.W., Stiles, B., Pelat, T., Thllier, P., 2011. mAbs 3 (6), 517–527. Grounow, R., Splettstoesser, W., McDonald, S., Otterbein, C., O'Brien, T., Morgan, C., Aldrich, J., Hofer, E., Finke, E.-J., Meyer, H., 2000. Clin. Diagn. Lab. Immunol. 7 (1), 86–90. Han, K.C., Yang, E.G., Ahn, D.R., 2012. Chem. Commun. 48, 5895–5897. Han, K.C., Yang, E.G., Ahn, D.R., 2013. Biosens. Bioelectron. 50, 421–424. Liebana, S., Lermo, A., Campoy, S., Cortes, M.P., Pividori, M.I., 2009. Biosens. Bioelectron. 25, 510–513. Liu, Y., Che, Y., Li, Y., 2001. Sens. Actuators B 72, 214–218. Pohanka, M., Hubalek, M., Neubauerova, V., Macela, A., Faldyna, M., Bandouchova, H., Pikula, J., 2008. Vet. Med. 53 (11), 585–594.
