High-resolution melting real-time PCR assays for subtyping of five diarrheagenic Escherichia coli by a single well in milk

Diarrheagenic Escherichia coli (DEC) is a kind of foodborne pathogen that poses a significant threat to both food safety and human health. To address the current challenges of high prevalence and difficult subtyping of DEC, this study developed a method that combined multiplex polymerase chain reaction (PCR) with high resolution melting (HRM) analysis for subtyping 5 kinds of DEC. The target genes are amplified by multiplex PCR in a single well, and HRM curve analysis was applied for distinct amplicons based on different melting temperature (Tm) values. The method enables discrimination of different DEC types based on characteristic peaks and distinct Tm values in the thermal melting curve. The assay exhibited 100% sensitivity and 100% specificity with a detection limit of 0.5–1 ng/μL. The results showed that different DNA concentrations did not influence the subtyping results, demonstrating this method owed high reliability and stability. In addition, the method was also used for the detection and subtyping of DEC in milk. This method streamlines operational procedures, shorts the detection time, and offers a novel tool for subtyping DEC.


INTRODUCTION
Escherichia coli (E.coli) are predominantly commensal bacteria that inhabit the environment, food sources, and gastrointestinal tracts of both humans and animals (Gomes et al., 2016).However, certain strains of E. coli can cause diarrheagenic illness in humans known as diarrheagenic Escherichia coli (DEC).DEC is a kind of prevalent foodborne pathogen that poses serious health risks to individuals across all age groups, particularly infants, young children, and old people (Shahbazi et al., 2021, Aziz et al., 2023).Based on their virulence factors, pathogenesis, and epidemiological characteristics, DEC can be categorized into 5 distinct groups: Shiga toxin-producing E. coli (STEC) or enterohemorrhagic E. coli (EHEC), enteropathogenic E. coli (EPEC), enterotoxigenic E. coli (ETEC), enteroinvasive E. coli (EIEC), and enteroaggregative E. coli (EAEC) (Leimbach et al., 2013).Among these categories, both ETEC and EPEC are the primary causes of infant mortality due to diarrhea (Farfán-García et al., 2016).EAEC is responsible for acute or persistent diarrhea in children and adults in developing countries (Aslani et al., 2011).STEC can induce bloody diarrhea, colitis, and hemolytic uremic syndrome (HUS) (Petrucci et al., 2022), while EIEC is capable of invading intestinal epithelial cells to cause dysentery diarrhea, making it a significant pathogen for children over 6 mo old (Abbasi et al., 2014).DEC is widely present in nature and can contaminate various types of food, including milk and other dairy products (Aijuka and Buys, 2019).DEC poses a significant threat to food safety and has adverse impacts on human health and the social economy (Kotloff et al., 2013).To prevent the spread of the epidemic and reduce the social and economic burden, it is crucial to conduct rapid subtyping and traceability of DEC.Therefore, there is an urgent need for a rapid and accurate method to subtype DEC.
E. coli have many kinds of antigens include 187 O-antigens, 80 Kantigens and 53 H-flagellar antigens (Yang et al., 2018, Liu et al., 2020, Nakae et al., 2021).In the early stage, E. coli was differentiated by phenotypic methods which include bacteriophage typing and serotyping method.However, these methods are generally labor-intensive, time consuming, costly, and not always accurate.Thus, molecular subtyping methods based on specific virulence genes are better alternative methods and they can enable the determination of the strain's pathotype and pathogenic potential.Polymerase chain reaction (PCR) combined with electrophoresis analysis was developed for the identification of specific pathogenic genes of DEC (Oh et al., 2014, Zhang et al., 2020).Nevertheless, it is difficult for electrophoresis to recognize nonspecific amplicons and impossible to distinguish PCR products with different nucleotide sequences but similar sizes (Shan et al., 2021).Researchers investigated the methods that combined PCR or isothermal amplification for target genes with lateral flow assay-based visualization of the genes to improve the accuracy of the detection of nucleotide sequences (Shan et al., 2021, Petrucci et al., 2022).However, opening tubes and transferring amplification products into lateral flow assay would lead to aerosol leakage which can contaminate reagents in the laboratory for several weeks, and can be promptly used as a template in subsequent reactions (Luo et al., 2022).Multiplex fluorescence PCR methods have the capability to detect multiple genes simultaneously, but they are limited in their ability to perform single well typing for DEC detection due to device constraints (Chen et al., 2014).High resolution melting (HRM) curve analysis which is a post-PCR analysis method performed on doublestranded (ds) DNA samples.HRM analysis exhibits high sensitivity toward any single-base alteration in the amplified targets, which was used for identifying genetic mutation or Single Nucleotide Polymorphism in nucleic sequences (Guion et al., 2008, Narimisa et al., 2022).Due to its simplicity, cost-effectiveness, and user-friendliness, the utilization of HRM has witnessed an upsurge in clinical diagnostics (Dehbashi et al., 2020) and food safety domains (Jin et al., 2012, Forghani et al., 2016).Harrison and Hanson used HRM analysis for rapid detection of sequence type 131 E. coli which strains have emerged as an international multiresistant high-risk clone (Harrison and Hanson, 2017).Some researchers used HRM curve analysis for detecting ETEC, E. coli O157:H7, E. coli O103 and O121 (Velez et al., 2022) (Singh et al., 2020) (Liu et al., 2018).However, those works focused on only a kind of DEC or a kind of gene sequence.When detected DEC by HRM technology, it still relied on 2 real-time multiplex PCR assays (Devi et al., 2018).
This study developed a method that combined multiplex PCR with HRM analysis for subtyping 5 kinds of DEC simultaneously in a single well.The target genes are amplified by multiplex PCR, and HRM curve analy-sis was applied for distinct amplicons based on different melting temperature (Tm) values.The proposed method enables the discrimination of different DEC types based on characteristic peaks and distinct Tm values in the thermal melting curve.The established method simplifies the detection procedure, improves efficiency, and has great potential in practical applications.

Bacteria strains and DNA extraction
All bacteria strains used in this study were summarized in the online supplemental materials (Table S1).The standard bacterial strains were obtained from the American Type Culture Collection (ATCC, Rockville, MD, USA), China Medical Culture Collection (CMCC, Beijing, China), and China Center of Industrial Culture Collection (CICC, Beijing, China).Other bacteria strains were from the culture collection of the Jiangxi Province Centre for Disease Control and Prevention.All strains were cultured in tryptic soy broth (TSB, Beijing Land Bridge Technology Co., Ltd., Beijing, China) on a shaking incubator at 180 r/m and 37°C overnight.The culture was then inoculated onto Mac-Conkey agar plates (Qingdao Hi-Tech Industrial Park Hope Bio-Technology Co., Ltd., Qingdao, China) using sterile loops and incubated at 37°C for 18 h.Bacterial genomic DNA was extracted using a crude method.The colonies selected from the plate and the cells from signal colonies were suspended in 200 μL of ultrapure water, heated at 98°C for 10 min, and centrifuged at 12000 r/m for 15 min, and the resulting supernatant was used as DNA template.The concentration and purity of the obtained DNA samples were determined using Nanodrop Lite Spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA).Ultrapure water used in this study was obtained from Ultrapure Lab Water Systems (Milli-Q ® systems).

Primers design
The target genes were detected in this assay as follows: β-glucuronidase gene (uidA) of E.coli, Gene encoding LEE-encoded type III secretion system factor (escV) of EPEC, Shiga toxin one (stx1) and Shiga toxin 2 (stx2) of STEC/EHEC, heat-labile enterotoxin gene (lt) and heat-stable enterotoxins initially discovered in the isolates from human gene (sth) of ETEC, aggregative adhesive fimbriae regulator gene (aggR) and protein involved in intestinal colonization gene (pic) of EAEC, invasive plasmid regulator (invE) of EIEC.The gene sequences were acquired from the NCBI GenBank database, and their corresponding accession numbers are listed in Table 1.Real-time PCR primers were designed using Primer Premier 5.0 software and tested for specificity and cross-amplification potential with the NCBI/Primer-BLAST tool.All primers used in this study were commercially synthesized by Sangon Biotech Co., Ltd.(Shanghai, China), with their respective sequences shown in Table 1.

Real-time PCR amplification and HRM assay for detecting genes
A 2-step analysis including real-time PCR amplification and HRM assay was performed in the Applied Biosystems 7500 Fast Real-Time PCR Instrument (Thermo Fisher Scientific, Waltham, MA, USA).The reaction mixture containing 10 μL MeltDoctor HRM Mix (Applied Biosystems), 0.6 μM primers, 2 μL of template, and ultrapure water adjusted to a final volume of 20 μL.The PCR protocol used for the amplification of the target gene consisted of an initial denaturation step at 95°C for 10 min, followed by 40 cycles of 95°C for 15 s, and 60°C for 60 s.HRM assay was performed at the end of the PCR amplification from 60 to 90°C with 0.04°C increments every second.The ultrapure water was used as a negative control.The melt curve analysis was performed using the Applied Biosystems HRM Software version 2.0.1.

Real-time multiplex PCR amplification and HRM assay
Five DEC strains (STEC CICC 10670; EAEC CICC 24186; EIEC CICC 24188; EPEC CICC 24189, ETEC CICC 10667) and E.coli (ATCC 25922) were incubated and then DNA templets were obtained as the method described in section 2.1.Then the DNA samples of the 5 DNA strains were detected by real-time multiplex PCR amplification and HRM assay.The reaction mixture containing 10 μL MeltDoctor HRM Mix, 0.12 μM each primer (Table 2), 2 μL of DNA template, and ultrapure water adjusted to a final volume of 20 μL.The PCR protocol used for the amplification of the target gene consisted of an initial denaturation step at 95°C for 10 min, followed by 40 cycles of 95°C for 15 s, and 60°C for 60 s.HRM assay was performed at the end of the PCR amplification from 60 to 90°C with 0.04°C increments every second.The ultrapure water was used as a negative control.The melt curve analysis was performed using the Applied Biosystems HRM Software version 2.0.1.The experiments were conducted 3 times with reproducible outcomes.Capillary electrophoresis (Qiagen GmbH, Hilden, Germany) was applied for the analysis of PCR-HRM amplification products.

Evaluation of the detection limit of PCR-HRM assay
The concentration of DNA templates was determined using Nanodrop Lite Spectrophotometer.The DNA samples from EAEC, ETEC, STEC, EPEC, and EIEC were diluted to a certain gradient as templates.The reaction system and steps followed the protocol outlined in section 2.4.Additionally, ultrapure water was used as a blank control to assess the detection limit of the PCR-HRM assay.

Specificity PCR-HRM assay
Nine standard non-E.colistrains (Staphylococcus aureus, Listeria monocytogenes, Vibrio parahaemolyticus, Pseudomonas aeruginosa, Proteus mirabilis, Enterobacter aerogenes, Bacillus cereus, Yersinia enterocolitica, and Salmonella typhimurium) were subjected to PCR-HRM assay.A standard E.coli strain was used as a positive control while ultrapure water served as the negative control to evaluate the specificity of this method.

Application of PCR-HRM assay in clinical isolates of DEC
The method was used to detect 45 clinical strains of known bacterial types isolated from stool samples of patients with diarrhea, and then the detection results were compared with the actual results to prove the effectiveness of the PCR-HRM method in detecting DEC in clinical specimens.

Detection and subtyping of DEC field isolates in milk
Milk samples were purchased at a local supermarket.Five DEC field isolates were individually inoculated (1-10 cfu) each in 25 mL milk samples confirmed free of DEC to verify the application effect of the method in the actual sample detection.Then, each sample was separately added to the TSB medium to 225 mL to drop on a 180 r/m shaking incubator and incubated for 12 h at 37°C.The cultures were subsequently scratched and streaked onto MacConkey agar plates using sterile inoculating loops for single colony isolation.The plates were incubated at 37°C for another 18 h.These obtained single colonies were then examined using the constructed method.

Singleplex PCR-HRM assay for detecting genes
A singleplex real-time PCR assay was designed using the DEC gene as the positive template and ultrapure water as the negative control.Following amplification, HRM analysis was performed by gradually increasing the temperature from 60°C to 90°C at a rate of 0.04°C/s.During heating, the amplified product (double-stranded DNA) began to denature, resulting in a decrease in fluorescence signal intensity.When the temperature reaches the Tm value of the amplification product, there is a sharp drop in the fluorescence signal.A melt curve can be generated by plotting the negative derivative of fluorescence versus (-dF/dT).As shown in Figure 1A-I, each target product exhibited a unique melting curve and its corresponding peak on the abscissa represents its melting temperature.None of the negative samples exhibited peaks, indicating the absence of non-specific amplification.With the target gene plotted on the abscissa and melting temperature on the ordinate in Figure 1J, it was observed that stx2 had the lowest Tm value while uidA had the highest.The amplification products of the 9 genes exhibited distinct melting temperatures.The small error of experimental results indicated excellent repeatability of PCR-HRM and the detailed data were presented in Table S2.The results of the singleplex PCR-HRM assay for the 9 genes proved that each gene amplification product has a unique Tm value.Therefore, this enabled us to differentiate the virulence genes and allowed for the construction of a multiplex PCR-HRM experimental approach.

Multiplex PCR-HRM assay for detecting strains
Five strains of DEC with distinct virulence genes and one strain of E. coli were identified and subtyped by real-time multiplex PCR combined with HRM analysis.As shown in Figure 2, DEC could be detected by our method, and the strains were identified according to the Tm of the target virulence genes.STEC has 3 melting curve peaks at 73.6°C, 78.6°C, and 83.7°C, which are stx2, stx1 and uidA (Figure 2A).ETEC has 3 melting curve peaks, they are sth (75.7°C), lt (77.8°C) and uidA (83.9°C), respectively (Figure 2B).EIEC has 2 melting curve peaks of invE (77.0°C) and uidA (84.3°C) (Figure 2C).EAEC has 3 melting curve peaks, which includes aggR (74.6°C), pic (80.5°C) and uidA (83.6°C) (Figure 2D).EPEC has 2 melting curve peaks at 81.3°C (escV) and 83.9°C (uidA).All 6 strains exhibited a melting curve peak at the same Tm position (83.9°C), which was the amplification product of the uidA gene.Moreover, E. coli displayed only one melting curve peak of  the products of the uidA gene (Figure 2E).The detailed data were presented in Table S3, and the average Tm value for each specific product was determined through 3 repeated experiments.
The products of multiplex PCR combined with HRM were also analyzed by capillary electrophoresis.As shown in Figure 3, the STEC lane exhibited 3 electrophoretic bands with lengths of 81 bp, 114 bp, and

Detection limit for multiplex PCR-HRM assay
To determine the detection limit of this method, a dilution series of extracted genomic DNA was as the template.As shown in Figure 4, the genomic DNA of DEC strains (0.5-100 ng/μL) was detected by PCR-HRM assay.Therefore, the proposed method has a detection limit of 0.5 ng/μL for stx1, stx2, sth, lt, aggR, pic, escV, and uidA.Besides, the detection limit of invE was 1 ng/μL.
Additionally, the melting temperature remains almost constant for different concentrations of genomic DNA within the reaction system (Figure 4 A2-E2).Therefore, the concentration of strains has minimal impact on results within a certain range, demonstrating this method has high reliability and stability.

Specificity of multiplex PCR-HRM assay
The specificity analysis was performed using 9 standard non-Escherichia coli strains.As shown in Figure 5, only E.coli as the template exhibited a melting curve peak for the uidA gene, while the other 9 samples and negative control did not show amplification.The results showed that the specificity of the developed multiplex PCR-HRM method is 100% (10/10).

Application of multiplex PCR-HRM assay in clinical isolates of E. coli
Forty-five clinical strains isolated from Jiangxi Province Centre for Disease Control and Prevention were subjected to evaluate the PCR-HRM assay, with 5 standard DEC strains serving as the controls.The melting curves of clinical strains were compared with control strains.As shown in Table 2, all strains could be accurately typed based on the Tm of their virulence genes, which is consistent with real results, indicating a 100% (45/45) sensitivity of the method.

DISCUSSION
Currently, the detection and subtyping methods of foodborne pathogens depend on enrichment culture combined with immunological assay (Mei et al., 2024), biochemical identification (Hinić et al., 2017) or molecular biological methods (Fratamico et al., 2016).However, it still has a challenge to subtype different strains of serotype.For example, E.coli was differentiated from other non-E.colistrains by MacConkey agar on the basis of color, colonies on MacConkey medium can further be presumptively identified based on their colonial appearances (shape, size, margin, time of growth, etc).It is very hard to identify the DEC and non-DEC cells (but E.coli) by conventional culture method.The selective chromogenic culture medium of CHROMagar for STEC can be used to detect STEC and non-STEC, but this medium is limited, some kinds of STEC cells cannot grow on it, and the specificity is not very high.Bacteriophage typing and serotyping methods were also used for subtyping E. coli by bacteriophage and antisera (Fratamico et al., 2016).However, these methods are generally time-consuming and not always accurate.
Thus, gene-based methods were more discriminatory for subtyping E.coli.In China, the GB4789.6-2016 is the standard method used for subtyping the DEC serogroups in food.This method is based on specific virulence genes and relies on at least 9 teams of single PCR to detect the 5 DEC serogroups, and then the PCR products are identified by electrophoresis analysis.The standard method is labor-intensive, time-consuming, and costly.Multiplex fluorescence PCR methods that used fluorescent probes could be used to detect multiple genes (Dong et al., 2019) (Di Domenico et al., 2017).However, they are limited in their ability to perform multiple tests due to device constraints, and the fluorescent probes were expensive (Zhang et al., 2020).PCR combined with HRM analysis has already used for subtyping DEC (Singh et al., 2020) (Guion et al., 2008).However, those works focused on only a kind of DEC, such as STEC (Liu et al., 2018) (Singh et al., 2020), ETEC (Wang et al., 2017).When detected DEC by HRM technology, it still relied on 2 real-time multiplex PCR assays (Devi et al., 2018).
In this study, we used the traditional culture method to enrich the target bacteria and then developed a method that combined multiplex PCR with HRM analysis in a single well for subtyping 5 kinds of DEC.First, the multiplex PCR primers for the 9 target genes were designed, uidA was as one E.coli-specific marker gene (Brons et al., 2020), escV was for the identification of EPEC (Chen et al., 2014), STEC harbor stx1 and/or stx2 genes (Wang et al., 2017), lt and sth were chosen for identifying ETEC (Zhang et al., 2020), main virulence genes of EAEC are aggR and pic (Zhang et al., 2022), the virulence gene of EIEC was invE (Fujioka  , 2013).We used a singleplex real-time PCR assay combined with HRM for selecting the primers that could amplify the target gene and obtain amplification products of the 9 genes that exhibited distinct melting temperatures (Figure 1).Such primers could be applied for the multiplex PCR-HRM experimental approach.Then real-time multiplex PCR combined with HRM analysis was developed, and 5 strains of DEC and one strain of E. coli were identified and subtyped by the new method (Figure 2).At the same time, we used capillary electrophoresis for identifying the products of multiplex PCR combined with HRM (Figure 3).These strains could be identified according to the Tm of the target virulence genes and the results of electrophoresis showed target genes could be amplified, and there were no non-specifical bonds, indicating the specificity of multiplex PCR amplification for gene fragments is high.
Furthermore, we used different concentrations of target bacteria genomic DNA as the templates of PCR-HRM assay, the results showed that the melting temperature remains almost constant for different concentrations of genomic DNA within the reaction system (Figure 4 A2-E2).Therefore, strains concentration has minimal impact on results within a certain range, demonstrating this method has high reliability and stability.Besides, only one kind of fluorescent dye was used, that made the method was inexpensive.

CONCLUSION
In conclusion, we developed a multiplex PCR-HRM method for the detection and typing of DEC in this study.The assay exhibited 100% (45/45) sensitivity and 100% (10/10) specificity with a detection limit of 0.5-1 ng/μL.The results showed that different DNA concentrations did not influence the subtyping results, demonstrating this method owed high reliability and stability.Compared with the conventional multiplex PCR combined with the electrophoresis detection method, our approach significantly reduces the time required for analysis while simplifying the overall process and can be performed in a single tube.It is simple, cost-effective, efficient, and has great potential for practical application.This method could aid in the early detection, diagnosis, and timely treatment of foodborne diarrheal diseases.Competing interests The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

ACKNOWLEDGMENTS
Shan et al.: High-resolution melting real-time… Shan et al.: High-resolution melting real-time…
Figure 3. Capillary electrophoresis of the products amplified by the multiplex PCR.

Figure 4 .
Figure 4. Detection limit of DEC by multiplex PCR-HRM assay.
This work was supported by Department of Science and Technology of Jiangxi Province (20232BAB21605), National Natural Science Foundation of China (82003467), Key Research and Development Program of Jiangxi Province (20192BBG70069), General Project of Jiangxi Natural Science Foundation (20202BAB206066), Science and Technology Fund Plan Figure 4 (Continued).Detection limit of DEC by multiplex PCR-HRM assay.

Table 1 .
Shan et al.:High-resolution melting real-time… Targets detected and primers used in this study

Table 2 .
Results of multiplex PCR-HRM assay for clinical isolates of E. coli.