Specific Multiplex Analysis of Pathogens Using a Direct 16S rRNA

May 2, 2012 - To investigate specificity, those candidates were blasted in the NCBI Web site, and Ribosomal Database Project II was searched for seque...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/ac

Specific Multiplex Analysis of Pathogens Using a Direct 16S rRNA Hybridization in Microarray System Byeong Hee Hwang,† Hwa Hui Shin, Jeong Hyun Seo, and Hyung Joon Cha* Department of Chemical Engineering, Pohang University of Science and Technology, Pohang 790-784, Korea S Supporting Information *

ABSTRACT: For the rapid multiplex analysis of pathogens, 16S rRNAs from cell lysates were directly applied onto a DNA microarray at room temperature (RT) for RNA−DNA hybridization. To eliminate the labeling step, seven fluorescent-labeled detector probes were cohybridized with 16S rRNA targets and adjacent specific capture probes. We found that eight pathogens were successfully discriminated by the 16S rRNA-based direct method, which showed greater specificity than the polymerase chain reaction (PCR)-labeled method due to chaperone and distance effects. A new specificity criterion for a perfect match between RNA and DNA was suggested to be 21−41% dissimilarity using correlation analysis between the mismatch and the sequence according to the guanine−cytosine (GC) percentage or the distribution of mismatches. Six categories of food matrix (egg, meat, milk, rice, vegetable, and mixed) were also tested, and the target pathogen was successfully discriminated within statistically significant levels. Finally, we found that the intrinsic abundance of 16S rRNA molecules successfully substituted PCR-based amplification with a low limit of detection of 10−103 cells mL−1 and a high quantitative linear correlation. Collectively, our suggested 16S rRNA-based direct method enables the highly sensitive, specific, and quantitative analysis of selected pathogens at RT within 2 h, even in food samples. abundant copies (up to 104−105) of 16S rRNA from a cell14 could hybridize with the specific capture probes (CPs) on a solid surface without PCR amplification. The first reported direct detection method for an array consisted of hybridization of the fluorescence dye-labeled 16S rRNA, which was purified from cells, to the CP, which was immobilized on a gel-type array.11 The effects of various factors, such as the labeling method, RNA fragmentation, and hybridization temperature, were investigated to improve the efficiency of the direct detection.12 It was reported that the hybridization efficiency for direct detection was mainly affected by the secondary structure of the 16S rRNA target.13 The direct analysis method has been applied to environmental samples15 and has been combined with systems such as surface plasmon resonance (SPR)16 and electrochemical biosensors.17 The mechanism of direct analysis (Figure 1A) is the cohybridization of a fluorescent dye-labeled detector probe (DP), an immobilized CP, and the rRNA target.12 Many advantages of the direct analysis strategy come from the simple procedure, reduced total detection time, high sensitivity, live cell detection, precise quantification, reduced sequence error due to many copies of rRNA in cells, and elimination of the PCR-based amplification and labeling steps. However, the direct detection method also has some limitations, such as a fragility of RNA molecules, limited detection targets and CPs,

P

athogens, which can threaten the health of human beings with severe diseases, should be detected and quarantined to prevent the immense social and personal loss.1,2 Manifold studies have been performed to develop detection techniques to overcome current limitations of the established methods.3−5 Oligonucleotide arrays have been considered to be a potential solution because they can provide rapid and parallel analytical platforms.6,7 In addition, a practical application of an oligonucleotide microarray as a biosensor is another important issue in the microarray research field. However, there are many obstacles to be overcome, such as polymerase chain reaction (PCR) dependence, complex and nonstandard experimental procedures, mainly fluorescence-based detection, high cost, hybridization at high temperature, and lack of portability. Although the PCR-based method for an analytical microarray has many advantages, such as an enhanced sensitivity and specificity and a low limit of detection (LOD) due to the cycled amplification of a specific sequence, it also has several problems, such as a bias of PCR amplification and labeling efficiency, a sequence error that can be generated by PCR,8 a difficulty in determining the original amount of mixed templates from the nonlinear character of PCR amplification,9 and a necessity for a DNA template of high purity.10 Thus, due to the limitations of the PCR-dependent method, a new strategy is needed to satisfy the requirements of high sensitivity and to make the procedure simple by eliminating amplification and labeling steps. As an alternative solution to overcome the problems of the PCR-based method, the direct hybridization using 16S rRNA molecules in array biosensors has been suggested.11−13 The important principle of this direct detection method is that the © 2012 American Chemical Society

Received: February 17, 2012 Accepted: May 2, 2012 Published: May 2, 2012 4873

dx.doi.org/10.1021/ac300476k | Anal. Chem. 2012, 84, 4873−4879

Analytical Chemistry

Article

food matrixes. Various approaches were also tested for the prevention of a RNA decomposition and the hybridization at RT. In addition, based on the sequence data analysis, the specificity criteria were suggested for the design of DPs and CPs.

■

EXPERIMENTAL SECTION Pathogenic Strains. The 10 pathogenic strains and their culture methods were same with our previous report.18 Details are in the Supporting Information. Design and Synthesis of DPs. Four types of probes were used for the direct analysis: artificial standard capture probe (ARST), positive control capture probe (POCO), CPs, and DPs. The ARST, POCO, and CPs were identical to those used in our previous reports.18,19 The ARST was used as a positioning anchor and a standard for correcting chip-to-chip variation.19 To design new DPs, the positions of POCO and all of the CPs were mapped in the 16S rRNA (see Supporting Information Figure S-1). First, the front and rear sequences adjacent to each CP were set as candidates to take advantage of the chaperone effect.12 Then, Primer Premier 5 (Premier Biosoft International, Palo Alto, CA, U.S.A.) was used to determine the optimal length of candidates by considering the thermodynamic properties such as a melting temperature and the existence of stable secondary structures. To investigate specificity, those candidates were blasted in the NCBI Web site, and Ribosomal Database Project II was searched for sequence matches. The designed candidates were evaluated based on the following criteria, in order of importance: an ability to stably hybridize with 16S rRNA targets, a specificity in the level of genus, and a melting temperature above 59 °C. Finally, five sequences were selected as suitable specific DPs (Table 1). Two universal DPs were also designed for general applications to several bacteria. The sequences for the universal DP candidates were designed from conserved sequences, which are located adjacent to the positions of CPs. Because the fluorescence intensity can be affected by the distance between the DPs and the CPs, the conserved regions for DPs were selected from two domains adjacent to eight and three CPs, respectively (see Supporting Information Figure S-1). Considering the effects on secondary structure, thermodynamic properties, proximity, and universality, universal 1 and universal 2 were finally selected as universal DPs (Table 1). Next, seven (five specific and two universal) selected DPs were chemically synthesized and then their amine groups were conjugated with the NHS ester group of Alexa Fluor 647, followed by a

Figure 1. Schematic diagrams of (A) direct analysis through direct cell lysis and hybridization using 16S rRNA target, detector probe, and specific capture probe and (B) repeated array format.

and different specificity criteria for DNA−RNA hybridization. Therefore, for a successful pathogen detection microarray system, the direct analysis method should be a simpler, faster, and less biased substitutive technique compared to the traditional PCR-based method. For the rapid and specific discrimination of multiple foodborne pathogens, we applied 16S rRNA-based direct analysis strategy to our previously constructed 11 pathogen detection microarray system, which uses 16S rDNA information as an analysis principle.18 To simulate a practical usage, a direct hybridization was performed at room temperature (RT) within 2 h using 16S rRNA molecules from cell lysates with diverse

Table 1. Oligonucleotide Sequences Employed As Detector Probes and Their Thermodynamic Properties detector probe

length (bp)

antisense sequences (5′−3′) b

Tm (°C)a

rating

corresponding specific capture probe

25 25 25 31

60.1 61 59.6 58.6

86 100 86 92

ESCOO2 LIMO1 and 2 SACH2 STAU2

YEEN1f universal 1

/A647/TCTTCCTGTTACCGTTCGACTTGCA GCCCATCTGTAAGCGATAGCCGAAA/A647/ /A647/CCTGGAATTCTACCCCCCTCTACAA TCCCTAATAACAGAGTTTTACGATCCGAAGA/ A647/ GAAAGTGCTTTACAACCCGAAGGCC/A647/ /A647/GACATTACTCACCCGTCCGCC

25 21

60.5 60.1

83 100

universal 2

TTCYGTGGATGTCAAGACCAGGTAAG/A647/

26

58.2−60.1

YEEN1 and 2 ESCOO1 and 2, SHDY2, VIPA1 and 2, VIVU1, VICH1 and 2 ESCO, SAEN, SHDY1

ESCOO2f LIMO2b SACH2b STAU2f

91

a

Melting temperature was calculated by the nearest-neighbor two-state model with parameters (Allawi, H. T.; SantaLucia, J., Jr. Biochemistry 1997, 36, 10581−10594). bA647: Alexa Fluor 647 with C6 spacer. 4874

dx.doi.org/10.1021/ac300476k | Anal. Chem. 2012, 84, 4873−4879

Analytical Chemistry

Article

Lumonics). The acquired raw data of fluorescence intensities were transformed into two-dimensional (2D) plots, with a gray gradation, by using a Matlab program in a same manner shown in our previous work.22 For the quantitative analysis, the cell lysate RNA sample was serially diluted 10-fold with 1 M sodium phosphate buffer (pH 7.2). For Vibrio parahemolyticus, the extent of dilution was ranged from 104−107-fold, which was equivalent to 6.5 × 104 to 65 colony forming units (CFU). For the Vibrio cholerae food sample, the extent of dilution was ranged that it was corresponding from 2.2 × 104 to 22 CFU. Each concentration was plotted against the fluorescence intensity. The LOD was determined at the intersection of the log−log linear regression plot and the horizontal line of variance. Food Sample Test. Six categories of food matrix were tested: milk, rice, vegetable, egg, meat, and mixed food. Approximately 20 g of each food sample was incubated in 200 mL of LB broth with shaking at 37 °C for 20 h. Samples for each category were prepared by different pretreatment methods. The milk sample was centrifuged at 13 000 rpm for 1 min. After discarding the supernatant, the pellet was washed three times with PBS (pH 7.4). The rice and vegetable samples were centrifuged at 500 rpm for 2 min to remove the major solids, and after recentrifugation of supernatant at 13 000 rpm for 1 min, the remaining pellet was washed three times. Because the egg sample had two solid layers, after centrifugation at 13 000 rpm for 2 min, the top layer was resuspended, and the supernatant was collected. After centrifugation at 13 000 rpm for 1 min, the pellet was washed three times. For the meat sample, the bacteria were impossible to be separated from the solid portion that the pellet was just washed three times. Dried seaweed rolls (Korean rolls), which was composed of rice, tuna, mayonnaise, carrot, pickled radish, onion, sugar, salt, sesame, pepper, ham, cucumber, toasted laver, and imitation crab meat, were used as a mixed food sample. The well-mixed solution was collected and centrifuged at 500 rpm for 2 min. The supernatant was centrifuged at 13 000 rpm for 30 s. After discarding the supernatant, remaining pellet was washed three times. Finally, after these pretreatment steps, pathogenic bacteria of known concentrations were added to each food pellet, and the mixture was centrifuged at 13 000 rpm for 30 s; the supernatant was discarded. The same steps were followed to prepare the direct cell lysate RNA samples.

purification step with high-performance liquid chromatography (Integrated DNA Technologies, Coralville, IA, U.S.A.). Preparation of the DNA Microarray. The DNA microarray was prepared with the same format (Figure 1B) and with the same procedure as our previous report to compare the direct analysis results with the PCR-based results.18 Details are in the Supporting Information. Target Preparation. For Escherichia coli O157:H7, an asymmetric 16S rDNA target was amplified from the chromosome following the previous protocol18 due to biohazard concern. The PCR samples were separated by electrophoresis in 1% agarose gel with TAE buffer (40 M Tris acetate and 1 mM EDTA, pH 8.0). The single-sense strand band was excised, and the 16S rDNA was extracted by using a gel extraction kit (Qiagen GmbH, Hilden, Germany). The concentration and purity were evaluated by measuring A260/ A280 with a UV−vis spectrometer. For other pathogenic bacteria, two types of RNAs (purified RNAs and cell lysate RNAs) were prepared. First, the total RNAs were purified with the RNeasy tissue kit (Qiagen) to optimize the experimental conditions. The total RNAs were harvested from cells in the exponential growth phase (OD 0.6− 0.8) to obtain the maximum amount of 16S rRNAs.20 A UV− vis spectrometer evaluated the concentration and purity of the purified total RNAs. Second, the cell lysate RNAs were directly and instantly obtained for the practical sample experiments. The cell pellet was washed three times with the chilled phosphate buffered saline (PBS; pH 7.4) and stored in a deep freezer (−80 °C) for at least 30 min. Then, it was resuspended in 10 μL of lysis buffer (1 M NaOH, 0.1% Triton X-100, and 2 mM EDTA in 20 mM Tris−HCl (pH 8.0), 1 mg mL−1 lysozyme for Gram-negative bacteria or 20 mg mL−1 lysozyme for Gram-positive bacteria, and 0.2 μm filtered), and then incubated at RT for 5 min. After the incubation, 190 μL of 1 M sodium phosphate buffer (pH 7.2) was added to the viscous bacterial lysate and mixed thoroughly. Then, the prepared cell lysate RNAs were used immediately for the hybridization step. Most of reagents and apparatus were sterilized to minimize the RNase contamination. Hybridization and Analysis. The hybridization buffer (20 μL total) consisted of 4 μL of 20× SSPE (autoclaved), 1 μL of 50× Denhardt’s solution, 6 μL of 100% formamide, 7 μL of target sample, 1 μL of 100 μM Alexa Fluor 647-labeled DP, and 1 μL of 1 μM Alexa Fluor 647-labeled artificial standard target. To prevent RNA degradation, DEPC-treated water or RNase ZAP (Sigma, St. Louis, MO, U.S.A.) was tested in the hybridization solution. DEPC-treated water was prepared based on the protocol from the Molecular Cloning manual.21 The hybridization reagents were immediately and independently added to the double-oligonucleotide microarray covered with a coverslip (autoclaved, 10 mm × 10 mm). Then, the hybridization chamber was agitated at different sets of temperature and time conditions as follows: at RT for 1 h, at RT for 3 h, or at 4 °C for 12 h. After hybridization, the microarray was washed three times for 30 s each time with the following sterilized solutions in order: (1) 1× SSC (450 mM NaCl, 3 mM trisodium citrate, and 1.5 M N,N,N-trimethyl glycine (betaine; Sigma), pH 6.6, final c = 1.5 M) with 0.2% sodium dodecyl sulfate (SDS), (2) 0.1× SSC with 0.2% SDS, and (3) 0.1× SSC. Then, the hybridized microarray was dried by centrifugation at 1500 rpm for 3 min, scanned by a confocal laser scanner (ScanArray Lite; GSI Lumonics, Wilmington, MA, U.S.A.), and quantified using the software (QuantArray; GSI

■

RESULTS AND DISCUSSION Design and Evaluation of DPs. Seven DPs were selected from the 40 candidates located adjacent to the CPs (Table 1). The criteria for a probe selection were the specificity level, melting temperature, and the secondary structure stability of the 16S rRNA region complementary to each DP.13 Among the seven DPs, five DPs were chosen as specific probes with specificities at the genus level, and two DPs were selected as universal probes with conserved sequences. The predicted melting temperatures of the DPs were between 58 and 61 °C for a stable hybridization. For a high hybridization efficiency, an unstable secondary structure of 16S rRNA region complementary to the DP was selected between the front and back adjacent DPs (see Supporting Information Figure S-1). Probe mapping showed that all of the DPs could interact with the 17 CPs, which could discriminate all 11 target pathogenic bacteria. Analysis of E. coli O157:H7 Using an Amplified DNA Target. Due to biohazard concerns regarding E. coli O157:H7, we used extracted chromosomes, instead of living cells, for 4875

dx.doi.org/10.1021/ac300476k | Anal. Chem. 2012, 84, 4873−4879

Analytical Chemistry

Article

Figure 2. Direct analysis of 11 target pathogens. (A) Scanned raw images, (B) 2D visualization of specific spots for 16S rRNA-based direct discrimination, and (C) 2D visualization of specific spots for PCR-based discrimination: (a) E. coli ATCC 25922, (b) E. coli O157:H7, (c) L. monocytogenes, (d) S. choleraesuis, (e) S. enteritidis, (f) S. dysenteriae, (g) S. aureus, (h) V. cholerae, (i) V. parahemolyticus, (j) V. cholerae, and (k) Y. enterocolitica.

treatment effectively blocked the activity of RNases during the hybridization step, this method was used for all subsequent experiments. To maximize the sensitivity, the conditions of hybridization temperature, hybridization time, and target fragmentation were optimized (see Supporting Information Figure S-3). To optimize the hybridization temperature and time, experiments were performed under different conditions. The fluorescence intensity for 2 h at RT was the highest among tested conditions. This result might indicate that the CPs or DPs could not hybridize to the highly stable secondary structure of the 16S rRNA target at 4 °C and the 16S rRNA target or the fluorescent dye may be cleaved or decomposed at 50 °C (see Supporting Information Figure S-4). In the case of target fragmentation, the fragmented 16S rRNA target had higher fluorescence intensity than the intact 16S rRNA target had. Fragmentation of the 16S rRNA target might change its accessibility by reducing the steric hindrance and stability of the secondary structure, which can affect the hybridization efficiency. On the other hand, in the case of the 16S rRNA samples that were obtained by lysis of two Vibrio species, fragmentation of the 16S rRNA showed improved sensitivity but decreased specificity (see Supporting Information Figure S5; yellow arrows indicate nonspecific spots). The specificity was decreased because the fragmentation of the 16S rRNA target with universal 1 DP could relieve the secondary structure and

detection. Thus, the asymmetric PCR-amplified 16S rDNA of E. coli O157:H7 was cohybridized with ESCOO2 CP and ESCOO2f DP. The hybridization results showed similar patterns with reduced fluorescence intensity (Figure 2B, part b) as those of the PCR-labeled target method (Figure 2C, part b). The decreased intensity of specific spots may have been caused by a low labeling or hybridization efficiency; each DP was labeled by one fluorescent dye, whereas the PCR-labeled DNA target was labeled by ∼11 fluorescence dyes (usually one dye per 125 bp). In addition, the mismatched spot of SHDY2 could be explained by the close proximity (7 bp) to ESCOO2f (see Supporting Information Figure S-1) and 0% dissimilarity (the percentage of sequence difference) between the E. coli O157:H7 target and SHDY2 CP. Establishment of Conditions Using Purified Total RNA Targets. Unlike DNA, RNA is very unstable at RT. Therefore, when the purified total RNA target was applied to the DNA microarray, weak fluorescence intensity was observed in the scanned raw image (see Supporting Information Figure S-2A, part a). Thus, to increase the spot intensity, we tested RNase inhibitors, such as DEPC-treated water or RNaseZAP-treated glassware, in the hybridization step. The DEPC-treated water increased the intensities by 2−3-fold, and the RNaseZAPtreated glassware increased the intensities by 5-fold over those of the control experiment (see Supporting Information Figure S-2A, parts b and c, and S-2B). Because the RNaseZAP 4876

dx.doi.org/10.1021/ac300476k | Anal. Chem. 2012, 84, 4873−4879

Analytical Chemistry

Article

(Figure 2B, part j). These mismatches might be caused by low stringency due to hybridization at RT,22 the higher melting temperature between RNA and DNA hybridization,23 and the close distance between the DP and the mismatched CPs in 16S rRNA. For example, Figure 2B, part c, shows strong mismatched spots on STAU1 CP because only 13.6% sequence dissimilarity (3 bp difference, AACTAGCTAATGCAGCGCGGAT) combined with relatively close 14 bp distance could not be sufficient to block the hybridization between the L. monocytogenes target and STAU1 CP in RT. To verify the specific criteria between RNA and DNA, seven mixed DPs were applied to 10 pathogens (data not shown). For VICH1 CP, no signal was detected for E. coli, S. dysenteriae, and S. enteritidis with four mismatches (21% dissimilarity); meanwhile, there were nonspecific signals for V. parahemolyticus and Y. enterocolitica with four mismatches (21% dissimilarity). These results indicated the position effect of mismatches. Additionally, a very high GC content of the matched region affected the RNA−DNA hybridization criteria. S. choleraesuis and S. enteritidis with nine mismatches (41% dissimilarity) were detected by YEEN 1 CP. Therefore, a dissimilarity of 21−41% seems to be the specific criterion for RNA−DNA hybridization at RT for short oligonucleotide, which is greater than the required dissimilarity of 10−15% for DNA−DNA hybridization at 60 °C.24 Direct Pathogen Detection with Food Samples. For practical applications, we investigated the direct detection of S. choleraesuis using five categories of food matrix (egg, meat, milk, rice, and vegetable). The fluorescence intensities of SACH2 CP are listed in decreasing order as follows: control (a) ≅ rice (e) > milk (d) > meat (c) > vegetable (f) > egg (b) (Figure 3). The

enable hybridization with adjacent mismatched CPs. Consequently, the optimal conditions for the direct detection experiment were established: hybridization with whole (without fragmentation) 16S rRNAs at RT for 2 h. Analyses of Direct Pathogen Discriminations Using 16S rRNAs. For the rapid and simple discrimination of live pathogens, the 16S rRNA target from cell lysate was applied directly for hybridization. We found that eight species (E. coli ATCC 25922, Salmonella enterica serotype choleraesuis, Salmonella enterica serotype enteritidis, Shigella dysenteriae, Staphylococcus aureus, V. cholerae, V. parahemolyticus, and Yersinia enterocolitica) were successfully and directly detected using six DPs among 10 pathogens (previous eight species, Listeria monocytogenes, and Vibrio vulnificus) (Figure 2, parts A and B). As described above, we did not use living E. coli O157:H7 in the 16S rRNA-based direct detection experiment due to safety issues. Scanned raw images of Figure 2A show bright white specific spots marked with white rectangles. To provide overall data at a glance and to discriminate subtle differences easily, we employed a 2D visualization tool with quantitative gradient intensity (Figure 2B). Notably, when the newly acquired results were compared with our previous results from PCR-labeled 16S rDNA targetbased detection (Figure 2C),18 16S rRNA target-based direct analysis displayed better specificity. The direct method clearly discriminated between E. coli and S. dysenteriae (Figure 2B, part a vs part f) and between S. choleraesuis and S. enteritidis (Figure 2B, part d vs part e) by perfect match analysis, which was not possible with the PCR-based method (Figure 2C, part a vs part f, and part d vs part e). This result agreed with the tendency in the previous reports:12,13 adjacent DP or CP could change the secondary structure of 16S rRNA, like a chaperone, which could affect hybridization efficiency. The fluorescence intensity was reduced as increase of distance between CP and DP.12 Therefore, the stable structure without adjacent probes might reduce hybridization with SHDY2 CP and SACH2 CP, or universal 2 DP and remote distances (871 and 321 bp, respectively) from DP might reduce fluorescence intensity. Thus, these factors could minimize a distant false-positive binding and improve a specificity. In addition, we found that the fluorescence intensities were inverted for two CPs on one target pathogen in the direct analysis results compared with the PCR-based results. The fluorescence intensities of SACH1 CP and SACH2 CP were inverted for S. choleraesuis and SACH2b DP (Figure 2B, part d); similarly, the fluorescence intensities of VIPA1 CP and VIPA2 CP were inverted for V. parahemolyticus and universal 1 DP (Figure 2B, part i). Signals of two CPs were inverted for the same reason of proximity, as SACH2 (0 bp) and VIPA2 (3 bp) were located closer to each DP than SACH1 (18 bp) and VIPA1 (11 bp) were. SACH2b DP and universal 1 DP might also have affected the secondary structure of each 16S rRNA target which were complementary, not to SACH1 and VIPA1, but to SACH2 and VIPA2 by the “chaperone effect” of the DP. Significant new mismatches (cross reactivity) appeared on STAU1 CP with the L. monocytogenes target and LIMO2b DP (Figure 2B, part c). In addition, mismatches were also found on SHDY2 CP (1 bp difference, GACTCAAGCCTGCCAGTTTCGA) with the S. choleraesuis target (Figure 2B, part d), VIVU1 CP (2 bp difference, AAACAAGTTTCTCTGTGCTGCCGC) with the V. cholerae target (Figure 2B, part h), and VICH2 CP (4 bp difference, CTCTACCGGGCAATTTCCCA) with the V. vulnificus target

Figure 3. Direct analysis of S. choleraesuis with SACH2b DP in five food matrixes. Fluorescence intensity plot of SACH2 spots for (a) control (pure pathogen), (b) egg, (c) meat, (d) milk, (e) rice, and (f) vegetable.

pathogens were well-separated by the centrifugation method for rice, milk, and vegetable but not for egg and meat (see Supporting Information Figure S-6). This finding indicated that pathogen separation was difficult due to the soft and mixable pellets of egg and meat. Although there may be many ingredients that degrade RNA, interfere with hybridization, or show strong backgrounds (e.g., the meat sample), the presence of S. choleraesuis target was statistically identified in all cases. Next, we detected the V. cholerae pathogen in the Korean rolls, a mixed food, with universal 1 DP and compared these results with those of the pure cultures. V. cholerae was successfully identified in the mixed food sample: the 4877

dx.doi.org/10.1021/ac300476k | Anal. Chem. 2012, 84, 4873−4879

Analytical Chemistry

Article

fluorescence intensities of VICH1 and VICH2, two CPs of V. cholerae, were quantitatively similar without any interference by the food ingredients (Figure 4). In the food samples, the

system using pure V. parahemolyticus. However, PCR-amplified detection in food sample requires additional steps for DNA purification because the PCR reaction could be inhibited by food ingredients.25 It is important to note that the analysis result showed no significant interference from the food ingredients. The higher sensitivity in the food samples might be explained by the hybridization efficiency between the 16S rRNA and the CPs, which is mainly affected by the secondary structure of the 16S rRNA.13 This LOD is much better than previously reported values: 500 ng of total RNA corresponds to ∼7.5 × 106 cells12 and 10 ng of 16S rRNA15 corresponds to ∼1.6 × 105 cells based on 72 000 copies per cell.14 The superior sensitivity of our direct analysis system could be explained by accumulation of several differences. The imaging equipment of a CCD-based imager (the limit of 2.4 × 109 molecules)12,26 might be the major difference compared to our system of a confocal laser scanner with the limit of 3.9 × 106 molecules.27 Additionally, the properties of capture probes (e.g., VIPA vs VICH) and optimized hybridization conditions, such as temperature (RT vs 50 °C), time (2 vs 4 h), and a RNase inhibitor (RNaseZAP vs DEPC),15 could also make notable differences (see Supporting Information Figure S-4C). Our direct analysis is close to the imaging equipment limit of 5.4 cells. To improve the sensitivity further, the use of enrichment cultures, more efficient method for lysis of Gram-positive cells, better cell separation schemes for food matrixes, or the introduction of nanoparticle labeling may be employed.28

Figure 4. Direct analysis of V. cholerae with universal 1 DP in mixed food matrix. Fluorescence intensity plot of positive control (POCO) and specific spots for control (pure pathogen) and food samples.

intensities of the POCO, ESCOO2, and SHDY2 spots were increased. The significant increase of the POCO spot intensity might indicate that other Gram-negative bacteria existed in the food sample; similarly, the increases in the ESCOO2 and SHDY2 spots might also indicate that bacteria similar to E. coli or S. dysenteriae were in the food sample. Sensitivity of the Direct Analysis. We investigated the sensitivity of the 16S rRNA-based direct analysis system using pure culture and using food samples. The cell lysate of V. parahemolyticus pure culture (open circles in Figure 5) was used

■

CONCLUSIONS

■

ASSOCIATED CONTENT

Our 16S rRNA-based analysis method enabled the facile and rapid discrimination of selected multiple pathogens with high sensitivity and specificity at RT within 2 h without extra labeling, amplification, or purification steps. Eight pathogens were successfully and directly discriminated through specific spot match analysis. Importantly, this direct method displayed more specific detection than the PCR-labeled 16S rDNA-based method did, which could be explained by the chaperone effect of the DP and the distance effect between the DP and the CP. Although a few new mismatches, due to the close distance and insufficient sequence dissimilarity for RNA−DNA hybridization, were observed, the perfect match criterion for RNA− DNA hybridization was estimated to be 21−41% dissimilarity by mismatch sequence analysis. In six food categories, pathogens were successfully detected at the statistically significant levels. The outstanding detection limit of the direct method was determined from high quantitative linear correlation to be approximately 10−103 cells mL−1, even in food samples. Thus, the proposed 16S rRNA-based direct analytical tool proved its ability to classify species and even subtypes of target pathogens in a reliable, rapid, sensitive, and accurate manner. This direct tool might be successfully exploited in the fields of food industry and clinical applications, where rapid and accurate detection of pathogens is important.

Figure 5. LOD of direct analysis. Detection of pure pathogen sample of V. parahemolyticus (open circles) and food sample containing V. cholerae (closed circles) with universal 1 DP, expressed by a log−log correlation plot between fluorescence intensity and cell concentration.

to determine the LOD to be approximately 102−103 cells mL−1, which was approximately 10 times lower than the value (10− 102 cells mL−1) of the PCR-based system.18 The LOD of the direct analysis system was sufficient to detect infectious doses of E. coli, Salmonella, Yersinia, and other Vibrio species, but it may be insufficient to detect infectious doses of S. dysenteriae or V. vulnificus. Interestingly, the LOD of the direct system using the V. cholerae-contaminated food sample was determined to be 10−102 cells mL−1 from the high quantitative linear correlation (closed circles in Figure 5). This LOD value is similar to that of the PCR-amplified system and is higher than that of the direct

S Supporting Information *

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. 4878

dx.doi.org/10.1021/ac300476k | Anal. Chem. 2012, 84, 4873−4879

Analytical Chemistry

■

Article

(21) Sambrook, J.; Russell, D. W. Molecular Cloning: A Laboratory Manual, 3rd ed.; Cold Spring Harbor Laboratory Press: New York, 2001. (22) Zhang, D. Y.; Chen, S. X.; Yin, P. Nat. Chem. 2012, 4, 208−214. (23) Casey, J.; Davidson, N. Nucleic Acids Res. 1977, 4, 1539−1552. (24) Eom, H. S.; Hwang, B. H.; Kim, D. H.; Lee, I. B.; Kim, Y. H.; Cha, H. J. Biosens. Bioelectron. 2007, 22, 845−853. (25) Lampel, K. A.; Orlandi, P. A.; Kornegay, L. Appl. Environ. Microbiol. 2000, 66, 4539−4542. (26) Call, D. R.; Brockman, F. J.; Chandler, D. P. Int. J. Food Microbiol. 2001, 67, 71−80. (27) Hesse, J.; Jacak, J.; Kasper, M.; Regl, G.; Eichberger, T.; Winklmayr, M.; Aberger, F.; Sonnleitner, M.; Schlapak, R.; Howorka, S.; Muresan, L.; Frischauf, A. M.; Schutz, G. J. Genome Res. 2006, 16, 1041−1045. (28) Lee, G.; Cho, Y. S.; Park, S.; Yi, G. R. Korean J. Chem. Eng. 2011, 28, 1641−1650.

AUTHOR INFORMATION

Corresponding Author

*Phone: +82 54 279 2280. Fax: +82 54 279 5528. E-mail: [email protected]. Present Address †

Department of Chemical Engineering, University of California, Santa Barbara, CA 93106, U.S.A. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Technology Development Program for Agriculture and Forestry (No. 109187-3) from the Ministry for Agriculture, Forestry and Fisheries, the Brain Korea 21 Programs from the Ministry of Education, Science and Technology, and the Satellite Research Laboratory Grant from the Young Woong Environmental and Biological Technology.

■

REFERENCES

(1) Mead, P. S.; Slutsker, L.; Dietz, V.; McCaig, L. F.; Bresee, J. S.; Shapiro, C.; Griffin, P. M.; Tauxe, R. V. Emerging Infect. Dis. 1999, 5, 607−625. (2) Lynch, M.; Painter, J.; Woodruff, R.; Braden, C. MMWR Surveill. Summ. 2006, 55, 1−42. (3) Lazcka, O.; Del Campo, F. J.; Munoz, F. X. Biosens. Bioelectron. 2007, 22, 1205−1217. (4) Iqbal, S. S.; Mayo, M. W.; Bruno, J. G.; Bronk, B. V.; Batt, C. A.; Chambers, J. P. Biosens. Bioelectron. 2000, 15, 549−578. (5) Hwang, B. H.; Lee, J. W.; Cha, H. J. Appl. Biochem. Biotechnol. 2010, 162, 1187−1194. (6) Ferguson, J. A.; Boles, T. C.; Adams, C. P.; Walt, D. R. Nat. Biotechnol. 1996, 14, 1681−1684. (7) Schena, M.; Shalon, D.; Davis, R. W.; Brown, P. O. Science 1995, 270, 467−470. (8) Acinas, S. G.; Sarma-Rupavtarm, R.; Klepac-Ceraj, V.; Polz, M. F. Appl. Environ. Microbiol. 2005, 71, 8966−8969. (9) Koh, C. G.; Tan, W.; Zhao, M. Q.; Ricco, A. J.; Fan, Z. H. Anal. Chem. 2003, 75, 4591−4598. (10) Gray, J. P.; Herwig, R. P. Appl. Environ. Microbiol. 1996, 62, 4049−4059. (11) Guschin, D. Y.; Mobarry, B. K.; Proudnikov, D.; Stahl, D. A.; Rittmann, B. E.; Mirzabekov, A. D. Appl. Environ. Microbiol. 1997, 63, 2397−2402. (12) Small, J.; Call, D. R.; Brockman, F. J.; Straub, T. M.; Chandler, D. P. Appl. Environ. Microbiol. 2001, 67, 4708−4716. (13) Chandler, D. P.; Newton, G. J.; Small, J. A.; Daly, D. S. Appl. Environ. Microbiol. 2003, 69, 2950−2958. (14) Amann, R. I.; Binder, B. J.; Olson, R. J.; Chisholm, S. W.; Devereux, R.; Stahl, D. A. Appl. Environ. Microbiol. 1990, 56, 1919− 1925. (15) Peplies, J.; Lachmund, C.; Glockner, F. O.; Manz, W. Appl. Environ. Microbiol. 2006, 72, 4829−4838. (16) Nelson, B. P.; Grimsrud, T. E.; Liles, M. R.; Goodman, R. M.; Corn, R. M. Anal. Chem. 2001, 73, 1−7. (17) Liao, J. C.; Mastali, M.; Gau, V.; Suchard, M. A.; Moller, A. K.; Bruckner, D. A.; Babbitt, J. T.; Li, Y.; Gornbein, J.; Landaw, E. M.; McCabe, E. R.; Churchill, B. M.; Haake, D. A. J. Clin. Microbiol. 2006, 44, 561−570. (18) Hwang, B. H.; Cha, H. J. Biotechnol. Bioeng. 2010, 106, 183− 192. (19) Hwang, B. H.; Cha, H. J. Biosens. Bioelectron. 2008, 23, 1738− 1744. (20) Ruimy, R.; Breittmayer, V.; Boivin, V.; Christen, R. FEMS Microbiol. Ecol. 1994, 15, 207−213. 4879

dx.doi.org/10.1021/ac300476k | Anal. Chem. 2012, 84, 4873−4879