Multiplex PCR Assays for the Detection of One Hundred and Thirty Seven Serogroups of Shiga Toxin-Producing Escherichia coli Associated With Cattle

Escherichia coli carrying prophage with genes that encode for Shiga toxins are categorized as Shiga toxin-producing E. coli (STEC) pathotype. Illnesses caused by STEC in humans, which are often foodborne, range from mild to bloody diarrhea with life-threatening complications of renal failure and hemolytic uremic syndrome and even death, particularly in children. As many as 158 of the total 187 serogroups of E. coli are known to carry Shiga toxin genes, which makes STEC a major pathotype of E. coli. Seven STEC serogroups, called top-7, which include O26, O45, O103, O111, O121, O145, and O157, are responsible for the majority of the STEC-associated human illnesses. The STEC serogroups, other than the top-7, called “non-top-7” have also been associated with human illnesses, more often as sporadic infections. Ruminants, particularly cattle, are principal reservoirs of STEC and harbor the organisms in the hindgut and shed in the feces, which serves as a major source of food and water contaminations. A number of studies have reported on the fecal prevalence of top-7 STEC in cattle feces. However, there is paucity of data on the prevalence of non-top-7 STEC serogroups in cattle feces, generally because of lack of validated detection methods. The objective of our study was to develop and validate 14 sets of multiplex PCR (mPCR) assays targeting serogroup-specific genes to detect 137 non-top-7 STEC serogroups previously reported to be present in cattle feces. Each assay included 7–12 serogroups and primers were designed to amplify the target genes with distinct amplicon sizes for each serogroup that can be readily identified within each assay. The assays were validated with 460 strains of known serogroups. The multiplex PCR assays designed in our study can be readily adapted by most laboratories for rapid identification of strains belonging to the non-top-7 STEC serogroups associated with cattle.


INTRODUCTION
The polysaccharide portion, called the O-antigen, of the lipopolysaccharide layer of the outer membrane of Escherichia coli provides antigenic specificity and is the basis of serogrouping. As many as 187 E. coli serogroups have been described based on the nucleotide sequences of O-antigen gene clusters (DebRoy et al., 2016). Escherichia coli serogroups that cause disease in humans and animals are categorized into several pathotypes. The serogroups that carry Shiga toxin genes on a prophage are categorized as the Shiga toxin-producing E. coli (STEC) pathotype. As many as 158 serogroups of E. coli are known to carry Shiga toxin gene(s), which make STEC the most predominant E. coli pathotype ( Table 1). Illnesses caused by STEC in humans, which are often foodborne, range from mild to bloody diarrhea with life-threatening complications of renal failure and hemolytic uremic syndrome (HUS), and even death, particularly in children (Karmali et al., 2010;Davis et al., 2014). Seven serogroups of STEC, O26, O45, O103, O111, O121, O145, and O157, called "top-7, " are responsible for the majority of human STEC illnesses, including food borne-outbreaks (Brooks et al., 2005;Scallan et al., 2011;Gould et al., 2013;Valilis et al., 2018). However, STEC serogroups other than the top-7, called "non-top-7" have also been reported to cause human illnesses, more often as sporadic infections, although a few are also known to cause severe infections, such as hemorrhagic colitis and HUS (Hussein and Bollinger, 2005;Bettelheim, 2007;Hussein, 2007;Bettelheim and Goldwater, 2014;Valilis et al., 2018). In a recent systematic review done by Valilis et al. (2018), 129 O-serogroups of STEC were identified to be associated with clinical cases of diarrhea in humans.
Ruminants, especially cattle, are a major reservoir of STEC and harbor the organisms in the hindgut and shed them in their feces. A number of studies have reported on the fecal prevalence of the top-7 STEC in cattle because of the availability of detection methods. For these serogroups, culture method involving serogroup-specific immunomagnetic separation and media for selective isolation and PCR assays to identify serogroups of putative isolates have been developed, validated and widely used (Bielaszewska and Karch, 2000;Chapman, 2000;Bettelheim and Beutin, 2003;Noll et al., 2015a). A number of studies have reported shedding of non-top-7 STEC in cattle feces ( Table 2). However, not much is known about the prevalence of these STEC serogroups in cattle feces, in terms of their distribution and proportion of animals in a herd positive for various serogroups, largely because of lack of isolation and detection methods. Traditionally, identification of serogroups or serotyping of E. coli, conducted by agglutination reaction using serogroup-specific antisera, is restricted to a few reference laboratories that possess the required antisera. However, the method is time consuming and often exhibits cross-reactions with other serogroups (DebRoy et al., 2011a). A number of PCRbased assays, end point or real time, have been developed and validated for the detection of one or more clinically relevant serogroups of E. coli (Perelle et al., 2004;Monday et al., 2007;Fratamico et al., 2009;Bai et al., 2010Bai et al., , 2012DebRoy et al., 2011b;Madic et al., 2011;Luedtke et al., 2014;Iguchi et al., 2015b;Noll et al., 2015b;Sanchez et al., 2015;Shridhar et al., 2016a). However, only a few mPCR assays have been described to detect certain STEC serogroups that are non-top-7 (Iguchi et al., 2015b;Sanchez et al., 2015;DebRoy et al., 2018). In recent years, DNA microarray and whole genome sequencing have been widely used to identify E. coli serogroups and serotypes (Liu and Fratamico, 2006;Lacher et al., 2014;Joensen et al., 2015;Norman et al., 2015). However, mPCR assays targeting serogroup-specific genes to identify STEC is a    (246) To understand the ecology and prevalence of these STEC serogroups in cattle, it is essential to detect the non-top-7 STEC serogroups shed in cattle feces in order to determine their impact on food safety and human health. Therefore, the objectives of the present study were to develop and validate mPCR assays targeting serogroup-specific genes to detect 137 non-top-7 STEC serogroups known to be associated with cattle.

Design of the Assays
A total of 14 mPCR assays, each targeting 7-12 STEC serogroups were designed. The targeted genes to design primers for serogroup detection included: wzx, which encodes for the Oantigen flippase required for O-polysaccharide export (Liu et al., 1996), wzy, which encodes for the O-antigen polymerase required for O antigen biosynthesis (Samuel and Reeves, 2003), gnd, which encodes for 6-phosphogluconate dehydrogenase for O antigen biosynthesis (Nasoff et al., 1984), wzm, which encodes for transport permease for O antigen transport, and orf469 and wbdC, which encode for mannosyltransferase for O antigen biosynthesis (Kido et al., 1995). The primers were designed based on the available nucleotide sequences of the target genes for each of the STEC serogroups from the GenBank database. The sequences for each serogroup were aligned using ClustalX version 2.0. The primers were designed to amplify the target genes with distinct amplicon sizes for each serogroup within an assay for easier visualization. The forward and reverse primer sequences for these serogroups are provided in Supplementary Tables 1A-N.

Validation of PCR assays
The specificity of each assay was determined with pooled DNA of the positive controls from the other 13 sets and top-7 STEC plus O104 PCR assays. Additionally, each assay was validated with one or more strains of the targeted serogroups. A total of 460 STEC strains belonging to 137 targeted serogroups were used for the validation of the assays (Table 4; Supplementary Tables 2A-N).
The strains were obtained from our culture collection (n = 104), E. coli Reference Center at Pennsylvania State University (n = 223), Michigan State University (n = 42), University of Nebraska (n = 5), and Food and Drug Administration (n = 86). Strains stored in CryoCare beads (CryoCare, Key Scientific Products, Round Rock, TX) at −80 • C were streaked onto blood agar plates (Remel, Lenexa, KS) and incubated overnight at 37 • C. Following incubation, colonies from the blood agar plates were suspended in 1 ml of distilled water, boiled for 10 min, centrifuged at 9,300 × g for 5 min and the supernatant was used for the PCR assays.

RESULTS
Out of the 158 serogroups of STEC, only five, which include O36, O66, O95, O184, and O187, have not been reported to be present in cattle feces, beef or beef products ( Table 1). A total of 14 mPCR assays, each targeting 7-12 O-types of 137 non-top-7 serogroups, were designed (Table 3). Each set of mPCR assay contained primer pairs that generated amplicons of different sizes for each target serogroup that were readily differentiated using a capillary electrophoresis system (Table 3; Figures 1A-N). The PCR product size for all the assays ranged from 145 to 1,046 bp (Table 3; Figures 1A-N). The specificity of each assay was confirmed when only the genes of the targeted serogroups were amplified and none of the serogroups targeted by the other 13 sets and top-7 plus O104 PCR assays was amplified (data not shown). The assays were validated with 460 strains of known serogroups, and the results indicated that all the assays correctly identified the target serogroups ( Table 4). The 14 sets of mPCR assays did not include the following 14 serogroups: O14, O30, O36, O52, O57, O59, O95, O97, O104, O158, O183, O184, O185, and O187.

DISCUSSION
Of the known 187 serogroups of E. coli, 158 serogroups have been shown to possess genes that encode for Shiga toxin 1, 2 or both. Serogroups, O26, O45, O103, O111, O121, O145, and O157, are top-7 serogroups responsible for a majority of human STEC illness outbreaks (Scallan et al., 2011;Gould et al., 2013). Among the top-7, fecal shedding of the O157 serogroup has been studied extensively, but relatively fewer studies have examined fecal shedding of the other six non-O157 serogroups in cattle, particularly in the United States (Renter et al., 2005; Paddock et al., 2014;Dewsbury et al., 2015;Noll et al., 2015a;Cull et al., 2017). Among the six top-7 non-O157 serogroups, O26, O45, and O103 are the dominant serogroups in cattle feces with prevalence ranging from 40 to 50%. However, only a small proportion of these serogroups (2-6%) carry Shiga toxin genes (Noll et al., 2015a). Because Shiga toxin genes are located on a prophage, it is suggested that the serogroups lacking these genes either have lost the prophage or have the potential to acquire the prophage (Bielaszewska et al., 2007). A majority of the non-O157 top-six STEC have been show to carry Shiga toxin 1 gene . There is evidence that the type of stx gene carried by STEC in cattle is dependent on the age of the animal and season. Shiga toxin gene of STEC strains in adult cattle are predominantly of the stx2 type, whereas the strains from calves primarily possess stx1 type (Cho et al., 2006;Fernández et al., 2012). In a study on E. coli O157 in Argentina, strains of O157 detected in all seasons were predominantly of the stx2 type, the proportion of strains containing stx1 decreased and proportion of strains possessing both types increased in warm seasons (Fernández et al., 2009). Many PCR assays have been developed and validated, generally targeting top-7 STEC serogroups, and often in combination with major virulence genes (Shiga toxins 1 and 2, intimin, and enterohemolysin: Bai et al., 2010Bai et al., , 2012DebRoy et al., 2011b;Fratamico et al., 2011;Lin et al., 2011;Anklam et al., 2012;Paddock et al., 2012;Noll et al., 2015b;Shridhar et al., 2016a). There is limited development of PCR assays targeting the non-top-7 STEC in cattle feces. Individual primer pairs have been described and PCR assays have been developed for each of the 187 serogroups of E. coli (DebRoy et al., 2018). However, there are only a few multiplex PCR assays targeting non-top-7 STEC serogroups (Iguchi et al., 2015b;Sanchez et al., 2015). Sanchez et al. (2015) reported the development of three mPCR assays targeting 21 of the most clinically relevant STEC serogroups associated with infections in humans. The assays included, top-7 serogroups and O5, O15, O55, O76, O91, O104, O118, O113, O123, O128, O146, O165, O172, and O177. Iguchi et al. (2015b) designed primer pairs to develop 20 mPCR assays, with each set containing six to nine serogroups, to detect 147 serogroups that included STEC and non-STEC.
STEC serogroups other than the top-7 have been reported to be involved in sporadic cases and a few outbreaks of human illness (McLean et al., 2005;Espie et al., 2006;Buchholz et al., 2011;Mingle et al., 2012). Among the non-top-7 STEC, certain serogroups, such as O1, O2, O8, O15, O25, O43, O75, O76, O86, O91, O101, O102, O113, O116, O156, O160, and O165, specifically certain serotypes within these serogroups, have been involved in outbreaks associated with consumption of contaminated beef in the US and European countries (Eklund et al., 2001;Hussein, 2007). Many of the outbreaks included cases of hemorrhagic colitis and HUS. Serogroups O91 (mostly H21 and H14 serotypes) and O113 (mostly H21 serotype) have been associated with severe cases of hemorrhagic colitis and HUS in the US and other countries (Feng et al., , 2017. Obviously, the difference in virulence between serogroups and serotypes is attributable to specific virulence factors encoded by genes in the chromosome, particularly on large horizontally acquired pathogenicity islands, or on plasmids (Levine, 1987;Bolton, 2011).
In contrast to humans, cattle are generally considered to be not susceptible to STEC infections. Only new born calves, particularly those that are immunocompromised because of deprived colostrum, have been shown to exhibit E. coli O157:H7 infections characterized by bloody diarrhea and attaching and effacing lesions (Dean-Nystrom et al., 1998;Moxley and Smith, 2010). Other serogroups that have been associated with diarrheal diseases of calves include O5, O8, O20, O26, O111, and O113 (Mainil and Daube, 2005). The majority of the serotypes causing infections in calves carried only Shiga toxin 1 gene (Mainil and Daube, 2005). Moxley et al. (2015) have reported isolation of STEC O165:H25 from the colonic mucosal tissue of an adult heifer that died of hemorrhagic colitis.
Of the 158 STEC serogroups, 130 serogroups have been associated with clinical cases of diarrhea in humans (Mainil and Daube, 2005;Hussein, 2007;Valilis et al., 2018). Therefore, there are 28 STEC serogoups that have not been reported to cause human infections, which is interesting because Shiga toxins are potent virulence factors. Either these STEC have not yet been linked to an illness or they lack other virulence factors, such as those needed for attachment and colonization, necessary to cause infections. A further understanding and assessment of the virulence potential of these serogroups will require sequencing of the whole genome to obtain a comprehensive gene profile.
In conclusion, the multiplex PCR assays designed in our study, which can be readily performed in most microbiology laboratories, will allow for rapid identification of isolates belonging to the non-top-7 E. coli STEC serogroups that are prevalent in cattle feces, beef or beef products.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

AUTHOR CONTRIBUTIONS
JB, TN, and CD conceived and designed the experiments. JL and XS performed the experiments. XS, JB, CD, ER, RP, and TN contributed reagents, materials, and analysis tools. PS, CD, XS, JB, and TN wrote the paper. All authors contributed to the article and approved the submitted version.

FUNDING
This material is based upon work that is supported by the National Institute of Food and Agriculture, U. S. Department of Agriculture, under award number 2012-68003-30155. The funders had no role in the study design, data collection and analyses, preparation of the manuscript or decision to publish.