Detection and Categorization of Diarrheagenic Escherichia coli with Auto-microfluidic Thin-film Chip Method

Diarrheagenic Escherichia coli (DEC) causes human diarrhea symptom in both healthy and immunocompromised individuals. An auto-microfluidic thin-film chip (AMTC) instrument integrating one-step multiplex PCR (mPCR) with reverse dot blot hybridization (RDBH) was developed for high-throughput detection of DEC. The novel mPCR method was developed by designing 14 specific primers and corresponding probes. 14 indexes including an endogenous gene (uidA) and 13 pathogenic genes (stx1, stx2, escV, ipaH, invE, estB, lt, pic, aggR, astA, bfpB, sth and stp) of DEC were detected. This one-step mPCR + RDBH approach is useful for simultaneous detection of numerous target genes in a single sample, whose specificity and availability have been confirmed on the positive control of 11 DEC strains. In addition, with 300 diarrheal stool samples being detected by this method, 21 were found to contain five major DEC strains. Compared with monoplex PCR and previous one-step mPCR approach, this method could detect ipaH and estB, and compared with current commercial kit, the relevance ratio of DEC detected by the AMTC method was increased by 1% in stool samples. Furthermore, the novel integration AMTC device could be a valuable detection tool for categorization of E. coli.

Most E. coli strains are normal inhabitants of the human intestinal tract. Some strains have acquired virulence genes that confer pathogenicity and are clarified as DEC 1 . The major virulence factors are essential for the study of the epidemiology and pathogenicity of DEC infection, such as severe diarrhea, food poisoning and similar outbreaks worldwide. Infections with these kinds of pathogens have therefore aroused increasing concern in the clinical diagnosis of diarrheal disease in recent years. These virulent organisms can be classified into five major categories on the basis of the nature of their infections and pathogenic mechanisms [2][3][4] , namely, enterohemorrhagic E. coli (EHEC), enteropathogenic E. coli (EPEC), enteroinvasive E. coli (EIEC), enteroaggregative E. coli (EAEC), and enterotoxigenic E. coli (ETEC). However, current rates of infections by these important enteric pathogens are probably underestimated to a great extent, because the existing clinical diagnostic methods are unable to distinguish them from normal nonpathogenic flora. To achieve the goal of epidemic prevention and control of DEC, a more reliable procedure is required to identify and categorize DEC isolates.
Many clinical laboratories routinely perform only serotyping assays aiming at detection of DEC. By using primers of virulence genes, the current monoplex PCR or multiplex PCR assay offers the possibility of rapid diagnosis of DEC strains 5,6 . Nevertheless, screening of bacterial isolates for DEC strains requires a large number of individual PCR assays when single primer sets are used 7 . Qualitative detection of DEC is performed by the agarose gelelectrophoresis based on various distributions of different sizes of amplified fragments. So, the number of fragments that can be detected is limited, and the bands of small fragments cannot be recognized very well.

Detection of DEC Strains by Previous One-step mPCR and Electrophoresis. The above mentioned
eleven DEC strains were also detected by a commercial kit, in which one-step mPCR and electrophoresis were mainly used. As shown in Fig. 3, BP01 was identified as EHEC; BP03 was identified as EIEC; BP04 was identified as atypical EPEC; BP05 was identified as typical EPEC; BP08 and BP09 were identified as ETEC; BP10 and BP11 were identified as EAEC; and BP02, BP06 and BP07 were identified as E. coli. Some E. coli strains, positive and negative controls were identified ( Supplementary Fig. S2).
Detection of DEC Strains by the Novel One-step mPCR + RDBH. The detection results of DEC strains by a novel one-step mPCR + RDBH method on AMTC instrument are shown in Fig. 4. BP01 was identified as EHEC; BP02 and BP03 were identified as EIEC; BP04 was identified as atypical EPEC; BP05 was identified as typical EPEC; BP06, BP08 and BP09 were identified as ETEC; and BP10 and BP11 were identified as EAEC; only BP07 was identified as E. coli.

Detection of Diarrheal Stool Samples by Previous One-step mPCR and Electrophoresis. By
using the previous one-step mPCR and electrophoresis listed in the current commercial kit, 18 diarrheal stool samples were found to contain DEC strains among a total of 300 samples. The positive rate was 6%. Part of results are shown in Fig. 5, indicating that EHEC was identified positively in one sample; EIEC was identified positively in two samples; typical EPEC was identified positively in one sample; atypical EPEC was identified positively in five samples; ETEC was identified positively in four samples; and EAEC was identified positively in five samples.
Detection of Diarrheal Stool Samples by the Novel One-step mPCR + RDBH. These 300 diarrheal stool samples were also identified by the novel one-step mPCR + RDBH method on the AMTC instrument, Figure 2. Electrophoretograms of monoplex PCR for DEC strains. BP01 includes uidA, stx1, stx2, escV and astA, being identified as EHEC; BP02 includes uidA and ipaH, identified as EIEC; BP03 includes uidA, ipaH, invE and pic, identified as EIEC; BP04 includes uidA and escV, identified as atypical EPEC; BP05 includes uidA, escV and bfpB, identified as typical EPEC; BP06 and BP07 include uidA, but no virulence genes; BP08 includes uidA, lt and stp, identified as ETEC; BP09 includes uidA, lt, astA and sth, identified as ETEC; BP10 includes uidA, pic and aggR, identified as EAEC; and BP11 includes uidA, pic, aggR and astA, identified as EAEC. Figure 3. Electrophoretogram of one-step mPCR for DEC strains. BP01 includes uidA, stx1, stx2, escV and astA, being identified as EHEC; BP02 includes uidA, but no virulence genes; BP03 includes uidA, invE and pic, identified as EIEC; BP04 includes uidA and escV, identified as atypical EPEC; BP05 includes uidA, escV and bfpB, identified as typical EPEC; BP06 and BP07 include uidA, but no virulence genes; BP08 includes uidA, lt and stp, identified as ETEC; BP09 includes uidA, lt, astA and sth, identified as ETEC; BP10 includes uidA, pic and aggR, identified as EAEC; and BP11 includes uidA, pic, aggR and astA, identified as EAEC. where, 21 samples were detected to contain DEC strains. The positive rate was 7%. As shown in Fig. 6, EHEC was identified positively in one sample; EIEC was identified positively in two samples; typical EPEC was identified positively in one sample; atypical EPEC was identified positively in four samples; ETEC was identified positively in six samples; EAEC was identified positively in six samples; and ETEC+ atypical EPEC were identified simultaneously in one sample. Compared with current commercial kit, the relevance ratio of DEC detected by the AMTC method was increased by 1% in stool samples.

Discussion
With more detection indexes and improved detection rate of DEC, the AMTC method is able to detect 14 indexes in a single reaction. This would avoid the probability of undetected ipaH and estB. Compared with monoplex PCR, the AMTC method identified BP06 as ETEC, just owing to estB. Compared with commercial kit, the AMTC method identified BP02 as EIEC, and BP06 as ETEC, just owing to ipaH and estB. For the detection of diarrheal stool samples, compared with commercial kit, the AMTC method identified YP223, and YP248 as ETEC, just owing to estB; identified YP295 as EAEC, just owing to astA; and identified YP299 as ETEC+ atypical EPEC, just owing to estB.
The AMTC method could be used to distinguish sub-type of DEC. Some researchers reported that astA encoded the EAST1, but astA is widely distributed in DEC, and its role in the pathogenicity of this organism remains unclear 13 . Some researchers reported that a strain harboring astA was associated with a waterborne Figure 4. Qualitative detection of DEC strains. BP01 includes uidA, stx1, stx2, escV and astA, being identified as EHEC; BP02 includes uidA and ipaH, identified as EIEC; BP03 includes uidA, ipaH, invE and pic, identified as EIEC; BP04 includes uidA and escV, identified as atypical EPEC; BP05 includes uidA, escV and bfpB, identified as typical EPEC; BP06 includes uidA and estB, identified as ETEC; BP07 includes uidA, but no virulence genes; BP08 includes uidA, lt and stp, identified as ETEC; BP09 includes uidA, lt, astA and sth, identified as ETEC; BP10 includes uidA, pic and aggR, identified as EAEC; BP11 includes uidA, pic, aggR and astA, identified as EAEC.
SCieNtifiC REPORtS | (2018) 8:12926 | DOI:10.1038/s41598-018-30765-3 outbreak of diarrhea in Japan 14 . So, further studies are needed to evaluate the significance of EAST1 as a virulence factor of DEC. This study showed 15 clinical strains contained astA, while 7 of them contained 3 sth, 1 stp and 3 escV. Obviously, astA is common among EHEC, EPEC, ETEC and EAEC, and each sample should be tested separately with specific primers such as stx1, stx2, escV, estB, lt, pic, aggR, astA, bfpB, sth and stp by monoplex PCR method. However, these processes could be completed using the newly-developed AMTC method in a single reaction. Some researchers believed that EAEC possesses both pic and aggR present in the virulence plasmid pAA 15,16 . Some researchers identified a few pic-positive and aggR-negative strains, and vice versa 17 . Reportedly, some researchers used only aggR for detecting pAA 10 . However, we found a positive single virulence gene and coexistence of two positive genes in the detection of diarrhea samples about EAEC.
The AMTC method possesses high efficiency and high throughput. For example, EscV is common between EHEC and EPEC, and in order to distinguish these two pathogens, bfpB and stx1 + stx2 must also be run by monoplex PCR method. Meanwhile, when bfpB and stx1 + stx2 are negative, the results can be identified as atypical EPEC, with positive bfpB being identified as typical EPEC, and positive stx1 + stx2 identified as EHEC. The results obtained by us show this method is able to perform both simultaneous amplification of virulence genes from E. coli isolates and simultaneous differentiation of the 5 categories of DEC.
The AMTC method has the advantages of good repeatability and stability. Two new indexes, ipaH and estB basing on monoplex PCR, and a multiplex PCR diagnostic kit, have been added to this method. Compared with the previous one-step mPCR, segment length of the PCR products for the novel one-step mPCR was basically the same, thus it could reduce competition between primers. Compared with electrophoretic detection, the fragments of poor amplification could also be detected with such RDBH method, resulting in a stable detection. Furthermore, one-step mPCR integrating with RDBH allows it to improve repeatability and stability of the detection results. Accordingly, the AMTC method would be a valuable contribution to the routine diagnostic laboratory tests while providing important information for the epidemiological and other studies.  . Qualitative detection of DEC strains in diarrheal stool samples by RDBH on AMTC instrument. YP11 includes uidA, stx1, stx2 and escV, being identified as EHEC; YP55 and YP60 include uidA, ipaH and invE, identified as EIEC; YP74 and YP90 include uidA and escV, identified as atypical EPEC; YP108 and YP166 include uidA, astA and escV, identified as atypical EPEC;YP179 includes uidA, escV and bfpB, identified as typical EPEC; YP186 includes uidA, lt and stp, identified as ETEC; YP191 includes uidA, astA, lt and stp, identified as ETEC; YP192 includes uidA, lt, astA and sth, identified as ETEC;YP222 includes uidA, astA and sth, identified as ETEC;YP223 includes uidA and estB, identified as ETEC; YP248 includes uidA, astA and estB, identified as ETEC; YP255 and YP293 include uidA, astA and pic, identified as EAEC; YP276, YP277 and YP280 include uidA, pic, aggR and astA, identified as EAEC; YP295 includes uidA and astA, but no virulence genes; and YP299 includes uidA, estB and escV, identified as atypical EPEC and ETEC.     Probes and Primers used for Detection of DEC. The oligonucleotide probes used in the AMTC corresponding to the reverse primers in one-step mPCR of a newly developed kit are listed in Table 1. An aminolinker C6 was added as a 5′-modified group to provide spaces between probes and immobilization substrates. The oligonucleotide primers used in monoplex PCR were collected from the China National Standard GB 4789.6-2016, being listed in Table 2. The final concentration of each primer used in the PCR amplification system depends on the GB 4789.6-2016. The primers used in a multiplex PCR diagnostic kit (a commercial kit currently used) were the components in the kit. The oligonucleotide primers used in a newly developed kit were designed by Primer premier v5.0, as shown in Table 3. The 5′ ends of the reverse primers were modified by a biotin group. Among them, the proportion of forward and reverse primers for each pair primers was 2:3. The final concentration of each primer in mixed PCR primers should be 10 µM. The predicted length of PCR products was ranged between 100 and 300 bp. All the primers and probes were synthesized by Sangon Biotech (Shanghai) Co., Ltd. Microfluidic thin-film chip. DNA probes were spotted and immobilized on a nylon thin-film according to the surface design. This step can be performed manually or by a machine. Then, the nylon thin-film probes spot array was incubated in 80 °C for 1 h, and it was placed in a small cell with two small holes. The microfluidic thin-film chip consisted of three main parts: a square nylon thin-film, a small cell with two small holes, and two microfluidic tubes. The small cell was a small square plastic box with a sealing ring on the inner side of the lid. When a square nylon thin-film was placed into the small box, the lid was closed, forming a closed cell. PCR products were added to the reaction cell through the microfluidic tube. When the PCR products passed over the surface of the square nylon thin-film containing the probe array, the target DNA in the PCR products was captured by the sensing surface DNA base pairing between the probes and target DNA 12 .

Monoplex
Qualitative Detection of DEC with RDBH. The AMTC instrument consisted of four main parts: a machine shell, micro-reaction cell, sample needle with pump, and sample and waste liquid plate. The eight microfluidic thin-film chip cells were placed on the micro-reaction cell of the AMTC device (Sichuan Hua Hansan Bio Technology Co., Ltd. #39, Fucheng West Rd., Chengdu, 610041, China). PCR products were added as samples to the sample plate of the AMTC device, and then the single-stranded DNA (Positive-Oligo, 10 μmol/L) complementary to the positive control (PC) probe was added to each PCR product sample for quality control. The hybridization buffer, cleaning buffers, enzyme buffer, and dyeing agent (NBT and BCIP) were placed in sample plate, with the signal being visualized with streptavidin-alkaline phosphatase color development kit (ZSGB-BIO, China). The sample needle with pump operated with an up and down movement and took different samples by rotating the sample plate. Under the control of an automatic hybridization program, the AMTC instrument could add samples and perform the reaction, washing, and coloring steps. The running procedures and time of the AMTC instrument were listed in Table 4.  Statistical analysis. The criterion for classification of diarrhoea E. coli was shown in Table 5. According to the criterion for classification, positive samples were identified. Compared with the total number of samples, the positive rate of different methods can be calculated.