The triplex PCR was used to amplify the target gene segments in E. coli O157:H7 and Vibrio cholerae O139. In multiplex asymmetric PCR reaction, the variation of the ratio of forward primer to reverse primer and concentrations impacted the amplification efficiency obviously. Therefore, the signal intensities of hybridization were highly related to the ratio and amount of forward primer to reverse fluorescent primer for each gene. After the optimization, three PCR products in one tube had good specificity and average amplification efficiency. The concentration of primers was showed in Table 1, and then the result of agarose gel electrophoresis was shown in Figure 1.

The 149 strains of intestinal bacteria including various serotypes (parts as shown in Table 3) were detected under the optimal condition of triplex PCR and hybridization for testing the specificity of this oligonucleotide microarray. Two sets of triplex PCRs were performed to amplify bacterial DNA templates. Cy3 labeled PCR products were mixed and hybridized with the probe array fixed on the glass slides.

It was shown that, 8 strains of E. coli O157:H7 had positive signal on stx1, stx2, and uidA probes. Two strains of E. coli O157: non-H7 had positive signals on stx1 and stx2 probes. Additionally, 3 strains of E. coli O157: non H7 and another serotype of Escherichia coli had signal on stx1 probe; 2 strains of E. coli O157: non H7 had signal on stx1 and stx2 probe, and furthermore E. coli O157: non-H7 had no signal on uidA probe. The results indicated that all the detected E. coli O157:H7 had both stx1 and stx2 toxin genes, whereas E. coli O157: non-H7 and other intestinal bacteria had one or both of the stx1 and stx2 toxin genes. However, the uidA probe can be used for identification of the O157:H7 from other intestinal bacteria accurately.

Seventy four strains of Vibrio cholerae O139 had positive signals on ctxA, tcpA, LPSgt probes; Vibrio cholerae (Ogawa and 569B) had positive signals on ctxA, tcpA probes. The results showed that ctxA, tcpA toxin genes may be detected in Vibrio cholerae O1 or O139 subtype. However, the probe specific to the unique region of LPSgt gene of Vibrio cholerae O139 was a strictly discriminating probe for Vibrio cholerae O139.

The fifty five strains of other intestinal bacteria and three non O1 non O139 Vibrio cholerae produces no detectable signal on the oligonucleotide microarray. The above results were shown in Table 4. Meanwhile, the hybridization images were shown in Figure 5, from which we could observe that the positive signals emerged orderly at the position corresponding to two pathogens from pure bacterial cultures; at the same time the signals of high intensity appeared at the position of positive control probes, and furthermore, no signals emerged at the position of negative control probes and blank probes.

The standard strains of E. coli O157:H7 and Vibrio cholerae O139 were provided by the National Institute for the Control of Pharmaceutical and Biological Products of China. The local standard strains of Vibrio cholerae O1 (569B), Vibrio cholerae Ogawa, Vibrio cholerae O139 and E. coli O157:H7 and the standard strains of some common intestinal bacteria and the clinical specimens were provided by the Center for Disease Prevention and Control of Zhejiang province, China. Part of the standard strains was listed in Table 3. The pathogens that isolated from clinical specimens were identified by the conventional microbiological methods 2122. The double-blind 10 mock samples infected by E. coli O157:H7 were prepared. The genomic DNA from clinical samples was extracted by centrifugation as previously described 23. Meanwhile DNA of the standard strains was extracted with UNIQ-10 column Genomic DNA Minipreps Kit (Sangon Co., LTD Shanghai, China), according to the manufacturer's instructions. Purified genomic DNAs were stored at -20°C in TE.

Based on references and published gene sequences in GenBank, stx1, stx2 and uidA genes were selected for E. coli O157:H7, ctxA, tcpA, LPSgt genes for Vibrio cholerae O139. Various sequences of each gene were collected and aligned by software Vector NTI Suite 9/AlignX. Primers and probes were designed using software Primer 5.0. The specificity of sequences was analyzed by BLAST GenBank. The 10 candidate probes were designed for each gene (totally 60 candidate probes), 19–60 mer in length. The Tm values of six primers and according to probes were approximately identical, and then the length of each PCR product was less than 500 bp. 16S rDNA gene was chosen as an internal control for verifying that genomic DNA templates of pathogens were present and the PCR reaction carried out successfully. Based on the previous study, a 500-bp fragment was obtained through PCR amplification using 16S-F1 (CGCTGGCGGCAGGCCTAACACATGC) and 16S-R1 (CGCGGCTGCTGGCACGGAGTTAGCC), and then the bacterial universal probe (ACTGAGACACGGTCCAGACTCCTACGGGAGGCAGCAGTAGGGAATATTG) was chosen as a marker for checking validity of internal control 24. The oligonucleotide probes were modified with amino residue at 3' end. The reverse primer was labeled with fluorescent dye Cy3 at 5' end. The sequences of primers and probes were shown in Table 1 and Table 2 respectively.

A volume of 20 μL conventional PCR mixture contains: 0.1 μmol/L the forward primer, 1.0 μmol/L the reverse primer, 100 μmol/L dNTPs, 50–100 ng DNA template, 1× PCR buffer and 1.5 U Taq polymerase (Promega). PCR products were amplified in a thermal cycler (PTC-100™ programmable thermal controller, MJ. research Inc) under the following condition: initial denaturation(5 min at 94°C) followed by 40 cycles of denaturation (30 sec at 94°C), annealing(30 sec 62°C) and extension(30 sec at 72°C). A final extension step was carried out for 5 min at 72°C, and 4°C forever. Through many experiments the optimal final concentration of primers in triplex PCR reaction was determined. The PCR condition and concentration of each component for 16S rDNA gene referred to the previous study 24.

DNA templates were extracted from standard strains of E. coli O157:H7 W933 and Vibrio cholerae O139 M045 respectively, which were amplified using conventional PCR with unlabelled primers. The PCR products were purified, and then inserted into pGEM-T vectors (Promega). Consecutively, the recombinant products were transformed into Escherichia coli DH5α (Promega), and incubated overnight at 37°C with shaking. Recombinant plasmids were screened using the T7 and SP6 primers and extracted using Promega Plasmid Extracting Kit, which procedure was according to the manual. After purification, the insert DNA fragments were identified through sequencing using automatic DNA sequencer (CEQ-2000). The number of plasmid copies was counted as previously described 25. The plasmid was 10-fold serial diluted and stored at -20°C for use as PCR template.

The 3' end amino-modified probes were diluted to a final concentration of 50 μmol/L in spotting solutions (3 × SSC and 0.01% SDS) and were transferred into 384-well micro-titer plates in volumes of 10 μL. The probe solutions were spotted to aldehyde-coated glass slides (CEL Associates) with a microarray printer (Cartisan), which deposits 0.5 nL at each spotting site, resulting in spots of 200 μM in diameter. The humidity during spotting was 90% and the temperature kept at 23°C. After spotting, slides were incubated for another 2 h under the same conditions and stored at room temperature for at least 24 h before use.

The Cy3-labeled PCR products were mixed with the hybridization solution (6 × SSC, 1 × Denhard, 0.2% SDS), and 10 μL of the hybridization solution was transferred to the hybridization area on the glass slide. The slide was incubated in 42°C water bath for 1 h in a tightly-sealed hybridization chamber. After incubation, the slide was washed sequentially in washing solution A (1 × SSC, 0.2% SDS), washing solution B (0.2 × SSC) and washing solution C (0.1 × SSC) for 1 min each.

The glass slides were scanned using the GenePix 4000B (Axon), with excitation at 540 nm and emission at 570 nm (Cy3). Sixteen-bit TIFF images of 10 micrometers resolution were analyzed with software GenePix Pro 4.0. The signal intensity is displayed according to spot color, whose order from high to low is white, red, yellow, green, and blue. After the removal of local background, the average signal intensity of each probe was calculated. The cutoff value was evaluated in order to confirm whether the target was considered to be positive or not.

The sensitivity was evaluated according to serial dilution (10-fold) from 10⁴~10¹ copies/μL. Meanwhile, according to Jin et al24, the sensitivity about the amount of bacteria in clinical specimen was determined. Each dilution series of genomic DNA were mixed with a natural background of genomic DNA from stool samples of healthy volunteers. The genomic DNA of stool samples was extracted as shown above. After the whole operation, two kinds of sensitivity were determined through signal intensity according to cutoff value. Additionally, the sensitivity of real time PCR (In real time PCR, the fluorescent probe detected the same unique region in LPSgt gene as the detecting probe of the oligonucleotide microarray) (DaAn Gene Co., Ltd, Guangzhou, China) and conventional microbiological methods was determined in order to compare the sensitivity of this assay with other methods.

Statistical analysis in this paper was made with SPSS for Windows software (version 11.5, Chicago, Il, USA).