Escherichia coli

Eco-hemX-GoST dendogram fragment
Typing Services

Drop us a line and we'll contact you with more information.

*We will never sell or share your email address.

$70/60/55/50 for 1/2/3/4 or more samples

Most E. coli strains are harmless, even modestly beneficial, commensals of our GI tracts. Other strains, however, are important foodborne pathogens, and a third group are agents of extraintestinal (particularly urinary tract) infection. For the latter, the pan-E. coli typing services described below can be used to investigate healthcare-associated infections.  However, the primary focus here is on a subgroup of the foodborne pathogens: Shiga-toxin producing E. coli (STEC). The prophage-encoded Shiga toxin (aka Verotoxin) along with the locus of enterocyte effacement (LEE) confer upon these enterohemorrhagic E. coli (EHEC) strains the ability to cause bloody diarrhea and, following bloodstream invasion, hemolytic-uremic syndrome (HUS) which can lead to kidney failure. Children and the elderly are at greatest risk.

STEC contamination has been documented in a wide variety of foods, including raw produce (e.g., spinach and alfalfa sprouts), raw milk, and unpasteurized juice. Ground beef, however, is the most common culprit since cattle serve as asymptomatic STEC hosts. Although most infections are sporadic, well-characterized (and publicized) outbreaks include one in 1993 traced to hamburger from Jack in the Box restaurants in the Pacific Northwest, and multiple outbreaks in 2006 traced to lettuce served at Taco Bell or Taco John restaurants or bagged fresh spinach from a California farm that had leased land from a cattle ranch.

STEC strains vary in their virulence; the most notorious is serotype O157:H7. The H antigen derives from the flagellar protein, while the clinically more relevant O antigen derives from the oligosaccharide repeat component of lipopolysaccharide.  Roughly 50% of STEC infections are O157 mediated, and of these about 8% progress to HUS.  In the U.S., most of the remaining infections are caused by serotypes O26, O45, O103, O111, O121, and O145 – dubbed the “Big 6”.  In Europe, an additional serotype, O104, was responsible for a major 2011 outbreak associated with consumption of sprouts from a farm in Germany, and surveillance detected 6 cases in the U.S. as well.

As with other foodborne pathogens, typing systems play a central role in identifying outbreaks and tracking down their sources, and in surveillance for emerging strains. In the clinical lab, EIA and PCR-based assays are increasingly used to diagnose STEC infection. Culture, however, remains essential for strain identification, initially by serotyping and subsequently by DNA-based methods that provide the resolution required for outbreak investigation and surveillance. For the latter, pulsed-field gel electrophoresis (PFGE) has been the gold standard, particularly in the U.S. where the PulseNet system ( provides the standardization, training, strain databases, and analysis capabilities that are needed to make this a viable typing system.

Multilocus variable number of tandem repeats analysis (MLVA) is an alternative typing system more widely used outside the U.S., although recently implemented by CDC/PulseNet as a complementary method for E. coli O157 and a few other foodborne pathogens ( To develop these MLVA methods, large numbers of repeat (VNTR) loci were tested by multiple laboratories on thousands of total isolates representing either O157 alone (1,2) or O157 plus non-O157 (3,4).

For each type of MLVA, a single VNTR locus emerged in these studies as substantially more informative than any other: TR2 for O157, and CVN014 for O157+non-O157. Both involve 6 base pair repeats within coding regions. Additionally, the non-coding repeat CRISPR1 (clustered regularly interspersed short palindromic repeat), has been shown to have value as a typing target for STEC strains, although insertion sequences and rearrangements preclude its use in some strains (5,6).

MicrobiType services for E. coli typing represent PLST extensions of these well-validated loci. EcMT1 (spanning the CVN014 VNTR) provides pan-E. coli  typing. To illustrate this, genomic sequences from all O157 and Big 6 strains included in the recent studies of Yin et al. (6) were used to generate an EcMT1 dendrogram. While most strains cluster by serotype (particularly O157) and epidemiological relatedness (e.g., spinach outbreak isolates), many others are well resolved. Other E. coli serotypes and non-typeable/non-STEC strains are not included in the dendrogram for space reasons, but are also typeable with EcMT1. Thus, this service is ideal for initial typing.

EcCRISPR1 is similarly useful for initial typing, as shown in the EcCRISPR1 dendrogram using the same strain set analyzed above. Again, most strains cluster by serotype and epidemiological relatedness. As noted above, however, approximately 10% of STEC strains lack an intact CRISPR1 locus.  

EcMT2 (spanning the TR2 VNTR), on the other hand, is designed for high resolution typing of O157 (and O121/O55) isolates; its locus is absent in other serotypes. As illustrated in the EcMT2 dendrogram, isolates from the 2006 Taco Bell, Taco John, and spinach outbreaks form well resolved clusters. Moreover, the 12 spinach outbreak isolates (from human, bovine, or bagged spinach) are themselves resolved into one major cluster and 5 additional branches, all well separated from the epidemiological outlier strain EC4115. Importantly, this EcMT2 analysis is highly consistent with the SNP analysis of Eppinger et al. (7) based on whole genome sequencing. Thus, this service provides an affordable alternative for O157 surveillance and outbreak analysis.

For all three services, results are reported as dendrograms and sequence alignments that illustrate the relatedness of the submitted isolate to GenBank database strains and to previously or concurrently submitted isolates from your lab.

(1) Noller et al., 2003, J. Clin. Microbiol. 41:5389
(2) Keys et al., 2005, J. Appl. Microbiol. 98:928
(3) Lindstedt et al., 2007, J. Microbiol. Methods 69:197
(4) Manges et al., 2009, J. Microbiol. Methods 79:211
(5) Delannoy et al., 2012, J. Clin. Microbiol. 50:4035
(6) Yin et al., 2013, Appl. Environ. Microbiol. 79:5710
(7) Eppinger et al., 2011, PNAS 108:20142

(MicrobiType services are for research/investigational use only)

© 2017 by MicrobiType LLC. All rights reserved. Terms of Service and Privacy Policy apply.