Enhanced Identification and Characterization of Non-O157 Shiga Toxin–Producing Escherichia coli : A Six-Year Study

Abstract

Non-O157 Shiga toxin–producing Escherichia coli (STEC) are emerging pathogens with the potential to cause serious illness and impact public health due to diagnostic challenges. Between 2005 and 2010, the Wadsworth Center (WC), the public health laboratory of the New York State (NYS) Department of Health, requested that Shiga toxin enzyme immunoassay (EIA)–positive stool enrichment broths and/or stool specimens be submitted by clinical and commercial reference laboratories testing NYS patient specimens. A total of 798 EIA-positive specimens were received for confirmation and serotyping, and additionally a subset of STEC was assessed for the presence of six virulence genes (stx1, stx2, eaeA, hlyA, nleA, and nleB) by real-time polymerase chain reaction. We confirmed 591 specimens as STEC, 164 (28%) as O157 STEC, and 427 (72%) as non-O157 STEC. Of the non-O157 STEC serogroups identified, over 70% were O103, O26, O111, O45, O121, or O145. During this time period, WC identified and characterized a total of 1282 STEC received as E. coli isolates, stool specimens, or EIA broths. Overall, the STEC testing identified 59% as O157 STEC and 41% as non-O157 STEC; however, out of 600 isolates submitted to the WC as E. coli cultures, 543 (90%) were identified as O157 STEC. This report summarizes a 6-year study utilizing enhanced STEC testing that resulted in increased identification and characterization of non-O157 STEC in NYS. Continued utilization of enhanced STEC testing may lead to effective and timely outbreak response and improve monitoring of trends in STEC disease epidemiology.

Introduction

S higa toxin–producing Escherichia coli (STEC) are considered major pathogens contributing to foodborne illness in the United States (Scallan et al., 2011). The clinical laboratory can easily identify O157 STEC (Schmidt et al., 1999); however, there are difficulties in detecting and identifying the more than 100 serogroups of non-O157 STEC (Coombes et al., 2008), including those that can cause severe disease (Brooks et al., 2005). However, in 1995, an enzyme immunoassay (EIA) capable of detecting any STEC was approved by the U.S. Food and Drug Administration (FDA) and introduced for use in clinical laboratories. When this Shiga toxin EIA is performed at the clinical laboratory and is combined with confirmation at a public health laboratory, enhanced identification and characterization of STEC is possible (Schaffzin et al., 2012). Increased implementation of this test may have contributed to the change observed between published estimates of foodborne infections in 1999 and 2011. In 1999, the burden of acquired foodborne infections attributed to STEC was estimated at 100,000 illnesses yearly in the United States (Mead et al., 1999), with non-O157 STEC contributing approximately 30% of these infections (Mead et al., 1999). More recent updates estimate that 175,000 illnesses each year are due to foodborne STEC infections, with an increased number of these infections attributed to non-O157 STEC (64%) (Scallan et al., 2011).

In 2009, the Centers for Disease Control and Prevention (CDC) recommended routine culture for O157 STEC on selective and differential media, as well as testing for non-O157 STEC through the use of a Shiga toxin EIA or a molecular assay that detects the Shiga toxin genes (stx1 and stx2) in the clinical laboratory (Gould et al., 2009). Additionally, it was recommended that positive STEC samples be forwarded to local or state health department laboratories for confirmation and characterization, serogrouping, and pulsed-field gel electrophoresis (PFGE) of an isolate. In 2002, before these recommendations were published, Wadsworth Center (WC) began requesting EIA-positive stool specimens and broths to increase the identification and characterization of STEC.

The 2009 CDC recommendations also suggested that future methods assess the potential of the organism to cause severe disease due to the presence of known virulence factor genes (Gould et al., 2009). These genes reside on mobile genetic elements (MGEs), including pathogenicity islands, plasmids, and bacteriophages that carry the stx genes. To date, there is no method to differentiate high-risk non-O157 STEC serogroups from those posing little risk to humans (Coombes et al., 2008). We chose to characterize a subset of STEC isolated during this time period with a panel of virulence factor genes. Our panel of genes included stx1 and stx2, eaeA, hlyA, nleA, and nleB, and was based on known or predicted contributions to STEC virulence (Bugarel et al., 2010; Coombes et al., 2008; Deng et al., 2003; Kelly et al., 2006; Newton et al., 2009; Ritchie et al., 2003; Wickham et al., 2006).

We describe here our 6-year study of enhanced identification and characterization of STEC, including the methodology and algorithms utilized, to continue where previous studies concluded (Brooks et al., 2005).

Materials and Methods

Algorithm for testing Shiga toxin EIA–positive specimens submitted for confirmation

The laboratory algorithm for testing EIA-positive broth prepared from stool specimens at the clinical laboratory is presented in Figure 1. Briefly, EIA enrichment broths were plated directly on two differential and selective media, Sorbitol MacConkey Agar (SMAC) and cefixime-tellurite (CT-SMAC), then incubated overnight at 37°C.

FIG. 1.

Wadsworth Center algorithm for isolating Shiga toxin–producing Escherichia coli (STEC) from EIA-positive enrichment broths and/or stools. ^aTypically, 10 individual colonies are tested for non-O157 STEC and 2–4 individual colonies are tested for O157 STEC. ^bAdditional colonies or pools of colonies may also be tested by stxPCR if first selections are negative. EIA, enzyme immunoassay; SMAC, sorbitol MacConkey agar; CT-SMAC, cefixime-tellurite sorbitol MacConkey; StxPCR, Shiga toxin gene real-time polymerase chain reaction.

Non-sorbitol fermenting (colorless) colonies were tested for the O157 antigen using the O157 Latex agglutination assay (Oxoid, Basingstoke, Hants, UK). If the latex test was positive, individual colonies were screened for the presence of Shiga toxin genes (stx1 and stx2) with a real-time polymerase chain reaction (PCR) assay (stxPCR). If the latex test was negative or if only sorbitol-fermenting colonies were visible on the SMAC/CT-SMAC, then individual colonies were combined into pools, and screened using stxPCR; the individual colonies were simultaneously subcultured. If a pool of colonies tested positive by PCR, the individual colonies from that pool were retested using the stxPCR. When a pool was negative, a portion of the original enrichment broth or the stool incubated in an enrichment broth was subjected to total nucleic acid extraction and tested using the stxPCR. The stools were incubated if needed with an in-house prepared E. coli broth (120 g of tryptone, 30.6 g of lactose, 24 g of K₂HPO₄, 9 g of KH₂PO₄, 30 g of NaCl in 6 L of millipore H₂O; pH 6.9) and directly plated on SMAC and CT-SMAC plates. If the enrichment broth was stxPCR positive, the enrichment broths were diluted and plated on SMAC and CT-SMAC to test additional pools of colonies. This screening process was repeated until a single Shiga toxin–positive organism was isolated.

A stxPCR-positive isolate was identified biochemically as E. coli using triple sugar iron (TSI), urea, indole, citrate, and lactose, sucrose, and sorbitol fermentation characteristics. The confirmed STEC isolates were serogrouped by a real-time PCR specific for the top six non-O157 serogroups (described below) and screened for O-groups (O26, O91, O103, O111, O128 and O145) using the Dryspot E. coli Seroscreen and Serocheck Kits following manufacturer's instructions (Oxoid). If the Seroscreen was positive, the Dryspot Serocheck was used for confirmation. Additionally, Statens Serum Institut (SSI) E. coli antisera was utilized to assess for O5, O45, O69, O88, O113, O118, O121, O165, and O178; isolates that were unable to be serogrouped at WC were forwarded to the CDC for analysis.

Total nucleic acid extraction from enrichment broth

As part of the testing algorithm, when a pool was negative by stxPCR, the stool enrichment broth was tested directly. Enrichment broths were washed with PCR buffer. Briefly, a 1-mL aliquot of broth was centrifuged at 14,000 rpm for 2 min, the supernatant was discarded, and the pellet was resuspended in 1 mL of 1× PCR buffer by pipetting, the process was repeated, and the washed pellet was resuspended in 1 mL of 1× PCR buffer. A volume of 500 μL of the washed broth underwent total nucleic acid extraction using the NucliSENS^® easyMAG^® automated extraction apparatus (bioMérieux, Inc., Durham, NC) using the Generic 2.0.1 extraction protocol according to the manufacturer's instructions with on-board lysis.

Real-time PCR

From 2005 to 2009, Shiga toxin genes (stx1 and stx2) were detected with an in-house developed and New York State (NYS) Clinical Laboratory Evaluation Program (CLEP) approved TaqMan^® real-time PCR assay (stxPCR). Primers and probe sequences were as follows: Shig1F-GCGTGGCATTAATACTGAATTGT, Shig1R-GAAGGAAACTCATCAGATGCCATT, Shig1Probe-Vic-TCATCATGCATCGCGAGTTGCCA-TAMRA, Shig2F-TCACTGTCACAGCAGAAGCCTTAC, Shiga2R-CCGGAAGCACATTGCTGATT, Shig2Probe-FAM-TCGTCAGGCACTGTCTGAAACTGCTCCT-TAMRA. Pure colonies or pools of colonies were placed in 1× PCR buffer and heat-treated at 95°C for 15 min prior to real-time PCR. Utilizing a 25-μL final volume, the reaction contained 1× LightCycler Fast Start DNA master hybridization probe mix (Roche Applied Sciences, Indianapolis, IN), 2 mM MgCl_2, 900 nM of each primer, 250 nM each of two dual-labeled oligonucleotide TaqMan^® probes, and 5 μL of template DNA.

The primers and probes were redesigned in 2008 using Primer Express v. 3.0 (Applied Biosystems, Inc. [ABI], Foster City, CA) to detect the majority of currently known gene variants; the new primer and probe sequences are in Table 1. The Shiga toxin variants known to be detected are Shiga toxin 1, 1c, and 1d as well as Shiga toxin 2, 2c, 2d, 2e, and 2g. It was not possible to identify a “universal” primer-probe set consisting of one primer pair and associated probe. A variety of forward, reverse, and probe oligonucleotides were utilized as opposed to designing degenerate primers to multiple variant target loci. These primers and probes were then combined. In 2010, the original stxPCR assay was replaced with the new primer and probe set that was also CLEP approved. Real-time PCR reactions were prepared with 1× LightCycler Fast Start DNA master hybridization probe mix (Roche Applied Sciences, Indianapolis, IN) utilizing a 25-μL final volume. Each reaction contained 3 mM MgCl_2, 450 nM of each primer (or primer mix), 250 nM of each probe (or probe mix), and 5 μL of template. Thermocycling conditions were as follows: 1 cycle at 95°C for 10 min, followed by 40 cycles at 95°C for 15 s, and 60°C for 1 min. Thermocycling, fluorescence data collection, and data analysis were performed with the ABI 7500 Fast Real-Time PCR System (v. 2.0.2) according to the manufacturer's instructions, with the passive reference dye, ROX, turned off.

Table 1.

Primer and Probe Sequences of the Multiplex Virulence Factor Gene Real-Time Polymerase Chain Reaction (virPCR)

Target	Name	5’-3’ Sequence with reporter and quencher dyes	Location
stx1	stx1F	ATT CTG GGA AGC GTG GCA TTA	Bacteriophage
	stx1Fa	ATT CTG GGT AGC GTG GCA TTA
	stx1R	CGG GCA CAT AGA AGG AAA CTC A
	stx1Ra	CGG ACA CAT AGA AGG AAA CTC AT
	stx1Rb	GGG CAC ATA GAA GGA AGC TCA T
	stx1P	Texas Red - TCA TCA TGC ATC GCG AGT TGC CAG - IBRQ^a
	stx1Pa	Texas Red - TCA TCA TGC ATC ACG AGT TGC CAG AA - IBRQ
stx2	stx2F	CAA CGG ACA GCA GTT ATA CCA CTC T	Bacteriophage
	stx2Fa	CAG CAG TTA TAC CAC GCT GCA A
	stx2R	AAC GCC AGA TAT GAT GAA ACC A
	stx2P	FAM - TTC CGG AAT GCA AAT CAG TCG TCA CTC - BHQ-1^b
hlyA	hlyAF	GGT CTG ATC ACA TCG GCT GTT AT
	hlyAR	TGC TTA GCT CGC TCA AAT TTA TCT G
	hlyAP	JOE - TGG CTA TCA GTC CTC TTT CTT TCC TGG CTG - BHQ-2	Plasmid
eaeA	eaeF2	GGA TCG TAT CGT CTG GGA TGA	Locus of enterocyte effacement (LEE) pathogenicity island
	eaeF2a	GAT CGT GTC GTC TGG GAT GA
	eaeF2b	GGA TCG TAT TGT CTG GGA TGA TA
	eaeR2	TTG CAC ATA AGC AGG CAA AAT AG
	eaeR2a	GCA CAT AAG CCG GCA AAA TAG
	eaeP	Texas Red - TCT TGT GCG CTT TGG CTT CCG CTA T - IBRQ
	eaePa	Texas Red - TCT TGT ACG CTT TGG CTT CCG CTA TGC - IBRQ
	eaePb	Texas Red - CT TGT GCG CTT TGG CTT CCG CCA T - IBRQ
nleA	nleAF	CAT TGG ATA TGT TTG GTT TAT TCA ATG	Genomic pathogenicity islands (O-islands)
	nleAFa	CGT TGG ATA TGT TTG GTT TAT TCA AT
	nleAFb	CAT TGG ATA CGT TTG GTT TAT TCA ATG
	nleAR	TCT GGT TTT ACA ATG TTC ATC TGA ACA
	nleARa	TCT GGT TTT ACT ATG TTC ATC TGA ACA
	nleARb	CTG GTT TTA CAA TGT TTA GCT GGA CA
	nleAP	FAM - TCA GAT GGC ATT AGA AAT A - MGB
	nleAPa	FAM - TCA AAT GGC ATT AGA GAT AC - MGB
nleB	nleBF	CAT TGA TGA AAA AAC GCC TAT TCT T	Genomic pathogenicity islands (O-islands)
	nleBR	CAT TAT CTA AAT AGG GAT GCT GCT TAG TAT
	nleBP	JOE - ACC TCA TCT TTC TTA TAT CGT TCA GGA TTA GGT TCA AAC - BHQ-2
	nleBPa	JOE - ACC TCA TCC TTC TTA TAA CGT TCA GGA TTA GGT TCA AAC - BHQ-2

IBRQ, Iowa Black^® Dark Quencher (IDT, Coralville, IA).

BHQ, Black Hole Quencher (Biosearch Technologies, Novato, CA).

Additionally, primers and probes used for a novel two-step multiplex real-time PCR assay, capable of rapidly serogrouping the six most prevalent non-O157 STEC, are described in Table 2. This assay is currently utilized after a single stx-positive isolate is identified and STEC O157 has been ruled out. The first three-plex reaction targets the wzx gene of STEC O26 and STEC O103, and the wbd1 gene of STEC O111. The second three-plex reaction detects the wzy gene of STEC O45, STEC O121, and STEC O145. Real-time PCR reactions were prepared with 1× LightCycler Fast Start DNA master hybridization probe mix (Roche Applied Sciences, Indianapolis, IN) utilizing a 25-μL final volume. Each reaction contained 4 mM MgCl_2, 900 nM of each primer (or primer mix), 250 nM of each probe (or probe mix), and 5 μL of template.

Table 2.

Primer and Probe Sequences of the Multiplex Real-Time Polymerase Chain Reaction to Detect the Six Most Prevalent Non-O157 Shiga Toxin–Producing Escherichia coli (STEC)

Target	Name	5’-3’ Sequence with reporter and quencher dyes
STEC O26 wzx	STEC O26 FOR	CGC GAC GGC AGA GAA AAT T
	STEC O26 REV	AGC AGG CTT TTA TAT TCT CCA ACT TT
	STEC O26-P	Texas Red-XN – CCC CGT TAA ATC AAT ACT ATT TCA CGA GGT TGA - 3BHQ_2™
STEC O103 wzx	STEC O103 FOR	TTT TAG CCA TAT CCT CAT CGT TGT T
	STEC O103 REV	GCG CCT TTG AGG CCA AA
	STEC O103-P	VIC - AAA CTA AGC CCA CCA TAG A - MGBNFQ™
STEC O111 wbd1	STEC O111 FOR	AAG GCG AGG CAA CAC ATT ATA TAG T
	STEC O111 REV	CAT CTG GGA GAT TCA ATT CAC TTT T
	STEC O111-P	6-FAM - CTT TGT TAC ACA CTG AAA G - MGBNFQ™
STEC O45 wzy	STEC O45 FOR	TTT CTG TCG AGT TTG CCT TCA G
	STEC O45 REV	GCA AGG CCT GCC GTA ACA
	STEC O45-P	6-FAM – CTT CAC ATG GTT TTA TGT CT - MGBNFQ™
STEC O121 wzy	STEC O121 FOR	TCT CTA TTG CTT CTT TTC TCC ATT TTG
	STEC O121 REV	CGG TCC TCA TTG CCC ATT T
	STEC O121-P	VIC – CTG CAC TCA TTG AAG C - MGBNFQ™
STEC 0145 wzy	STEC O145 FOR	ATA TTG GGC TGC CAC TGA TGG GAT
	STEC O145 REV	TAT GGC GTA CAA TGC ACC GCA AAC
	STEC O145-P	Texas Red-XN – ACC ATG CTG TGC GCG AAC CAC TGC T - 3BHQ_2™

Virulence factor gene real-time PCR

A random subset of 288 STEC received during the six-year study were assessed to determine if selected virulence factor genes were present with a multiplex virulence factor gene real-time PCR (virPCR). The retrospective study included both sporadic and outbreak related isolates from both O157 and non-O157 serogroups; additionally some isolates were tested prospectively at special request by the submitter or for epidemiological purposes. Samples were analyzed utilizing two three-plex TaqMan^® real-time PCR assays. The first multiplex assay targeted stx1, stx2, and hemolysin gene (hlyA), while the second multiplex assay targets the genes for intimin (eaeA) and non–locus of enterocyte effacement (LEE)–encoded effector proteins nleA and nleB. The assays were developed to detect all the currently known variants of the six virulence genes, except for stx2f. For many of the primers and probes, a variety of forward, reverse, and probe oligonucleotides were utilized as opposed to designing degenerate primers for amplifying multiple variant target loci.

Real-time PCR reactions were prepared with 1× LightCycler Fast Start DNA master hybridization probe mix (Roche Applied Sciences, Indianapolis, IN) utilizing a 25-μL final volume. Each reaction contained 3 mM MgCl_2, 450 nM each primer, 250 nM of probe, and 5 μL of template. All real-time PCR assays were run on the ABI 7500 Real-Time PCR System, utilizing the following cycling conditions: one cycle at 95°C for 10 min, followed by 40 cycles at 95°C for 15 sec, then 60°C for 1 min.

Results

EIA-positive samples submitted to WC, 2005–2010

In total, 798 specimens were tested over the 6-year period of 2005–2010 (Table 3); the number of specimens received per year (approximately 130) remained consistent. STEC was isolated from 74% (591/798) of the specimens. The most frequent serogroup identified was O157 STEC, isolated from 20% (164/798) of the total specimens or 28% (164/591) of the specimens with an STEC isolation. Non-O157 STEC was isolated from 53% (427/798) of the total specimens or 72% (427/591) of the specimens with an STEC isolation; no single serogroup was more prevalent than O157. Culture-negative, stxPCR-positive specimens represented 4% (32/798) of the total specimens received. Additionally, 22% (175/798) of the specimens could not be confirmed to contain STEC by stxPCR or culture methods.

Table 3.

Summary of Enzyme Immunoassay (EIA)–Positive Broth and Stool Testing, 2005–2010

Serogroup	2005	2006	2007	2008	2009	2010	Total non-O157	Total
O157	44	25	29	30	20	16		164
O103	8	13	18	14	12	32	97	97
O26	8	11	9	13	20	18	79	79
O111	8	7	12	10	8	11	56	56
O45		9	10	9	11	8	47	47
O121	2	3	1	7	3	4	20	20
O145	1	1	3		1	7	13	13
O165	1	1		1	3	5	11	11
O5					3	1	4	4
O69		1		1	1	1	4	4
O118		2	1			1	4	4
O166		2				2	4	4
O88	1		1		1		3	3
O91	1	1				1	3	3
O128		1	1	1			3	3
O178		1	2				3	3
O113		1			1		2	2
O119			2				2	2
O130		1			1		2	2
O174		1				1	2	2
O179		1		1			2	2
O6			1				1	1
O8			1				1	1
O22			1				1	1
O28						1	1	1
O71					1		1	1
O76					1		1	1
O77		1					1	1
O79					1		1	1
O80					1		1	1
O84				1			1	1
O126				1			1	1
O143						1	1	1
O146						1	1	1
O153			1				1	1
O172					1		1	1
STEC^a		5	7	8	9	22	51	51
Stx PCR positive only^b	6	3	6	10	1	6		32
Negative^c	19	29	31	49	32	15		175
Total STEC identified	74	88	100	97	99	133	427	591
Total	99	120	137	156	132	154	427	798

STEC are Shiga toxin–producing Escherichia coli isolates for which a serogroup was determined to be serogroup undetermined or rough, or results were not available from Wadsworth Center (WC) or Centers for Disease Control and Prevention (CDC) at the time of our report.

Stx PCR positive only are EIA broths that were positive for stx1 or stx2 by real-time polymerase chain reaction (PCR), but for which no stx positive isolate was obtained.

Negative are STEC-positive EIA broths that could not be confirmed by real-time PCR or culture as STEC.

The serogroup distribution among the 427 non-O157 isolates is summarized in Table 3. Over the 6-year period, the top six non-O157 serogroups identified were O103, O26, O111, O45, O121, and O145, and they accounted for over 70% of the non-O157 organisms that were identified. Fourteen other STEC serogroups were identified sporadically over the years, and 15 serogroups were identified only once. Twelve percent (51/427) of the isolates were identified as STEC serogroup undetermined or rough, or had no serogroup determined by the WC or CDC.

Total STEC identified by WC, 2005–2010

As a point of reference, in the 6-year time period, WC identified and characterized 1,282 isolates that were positive for STEC (Table 4). Of these STEC, 59% were O157 STEC and 41% were non-O157 STEC. Six hundred of these were submitted to the WC as E. coli culture isolates. O157 STEC was identified in 543 (90%), while the other 10% were identified as non-O157 STEC. The other 682 STEC isolates identified were either from primary stool specimens or STEC-positive EIA broths (591 were from STEC-positive EIA broth). Interestingly, non-O157 was identified in 68% of the primary stool/STEC-positive EIA broth submissions, while O157 STEC was only identified in 32%.

Table 4.

Summary of Shiga Toxin–Producing Escherichia coli (STEC) Identified and Characterized by the Wadsworth Center (WC), 2005–2010

	Number of STEC-positive specimens
Specimen source^a	O157 STEC (% total by source)	Non-O157 STEC (% total by source)	Total
Broth/stool (EIA-pos)	216 (32)	466 (68)	682
Culture isolate	543 (91)	57 (9)	600
Total	759 (59)	523 (41)	1282

The source indicates how the specimen was submitted to the WC laboratories, either as a culture isolate or as a primary stool or STEC-positive enzyme immunoassay (EIA) broth.

Retrospective analysis of STEC for virulence genes

A subset of 288 STEC isolates was tested by virPCR, including 132 O157 STEC and 156 non-O157 STEC (Table 5). Resultant data are displayed for O157 STEC, each of the top six non-O157 STEC serogroups, and other non-O157 STEC (combined). The majority of the O157 STEC tested was found to contain stx2 (98%); however, a subset of these contained both stx1 and stx2 (31%), and 2% contained stx1 only. Overall, the non-O157 isolates were more likely to contain stx1 (81%), with smaller percentages containing both stx1 and stx2 (8%) and stx2 only (18%). For the other four virulence genes assessed, we found that all of the O157 STEC isolates contained hlyA, eaeA, nleA, and nleB. We also found the top six non-O157 serogroups tested to contain on average more of the virulence genes (98% hlyA, 97% eaeA, 100% nleA, and 62% nleB) in comparison to the other non-O157 serogroups tested (56% hlyA, 26% eaeA, 24% nleA, and 13% nleB).

Table 5.

Incidence of Virulence Factor Genes Identified in a Subset of O157 and Non-O157 Shiga Toxin–Producing Escherichia coli (STEC)

		No. positive (%) by real-time polymerase chain reaction
Escherichia coli serogroup	Total no.	stx1	stx2	hlyA	eaeA	nleA	nleB	Percentage male	Mean age (range), years
O45	34	34 (100)	0 (0)	34 (100)	33 (97)	34 (100)	34 (100)	74	37 (6–84)
O111	49	49 (100)	6 (12)	47 (96)	47 (96)	49 (100)	2 (4)	40	16 (0–79)
O103	9	9 (100)	1 (11)	9 (100)	9 (100)	9 (100)	7 (78)	43	37 (26–70)
O26	8	8 (100)	1 (12)	8 (100)	8 (100)	8 (100)	0 (0)	71	23 (3–59)
O121	4	0 (0)	4 (100)	4 (100)	4 (100)	4 (100)	0 (0)	50	20 (3–42)
O145	6	4 (67)	2 (33)	6 (100)	6 (100)	6 (100)	5 (83)	33	36 (19–63)
Other non-O157^a	46	23 (50)	27 (59)	26 (56)	12 (26)	11 (24)	6 (13)	34	37 (0–79)
Total non-O157	156	127 (81)	41 (26)	134 (86)	119 (76)	121 (78)	74 (47)	48	28 (0–84)
Total O157:H7	132	44 (33)	129 (98)	132 (100)	132 (100)	132 (100)	132 (100)	38	25 (1–85)

33 additional non-O157 STEC serotypes were assessed in addition to the top six non-O157 STEC.

Discussion

STEC causes an estimated 175,000 illnesses each year in the United States, and it is predicted that at least two-thirds of these illnesses are due to non-O157 STEC (Scallan et al., 2011). The introduction of Shiga toxin EIA tests has improved a laboratory's ability to detect non-O157 STEC. The use of the Shiga toxin EIA along with confirmatory testing at a public health laboratory likely contributed to the increased burden of disease attributed to non-O157 STEC in 2011 as compared to estimates from 1999 (Mead et al., 1999; Scallan et al., 2011). This report summarizes our utilization of enhanced identification and characterization of EIA-positive broths between 2005 and 2010.

During this study, a higher incidence of non-O157 STEC (72% [427/591]) as compared to O157 STEC (28% [164/591]) was observed in the 591 STEC-confirmed EIA-positive specimens (Table 3). Table 3 indicates that 72% of the O157 STEC identified at the WC were submitted as pure culture isolates, while only 28% of the O157 STEC were submitted as EIA-positive broths, suggesting the submitting laboratory's use of culture to detect O157 STEC reduces the number of EIA-positive broths identified as O157 STEC at the WC.

Additionally, 4% (32/798) of the EIA-positive specimens were culture negative but positive by real-time PCR. While this number is low, it may indicate a slight increase in sensitivity of the PCR and EIA methods over culture or a lack of viability of the organism in these specimens. Of interest was the finding that 22% of the EIA-positive specimens could not be confirmed by real-time PCR or culture as STEC. Previous studies have indicated that other infectious organisms may cross-react with the STEC EIA, including norovirus and Pseudomonas aeruginosa (Beutin et al., 1996; CDC, 2001, 2006). We have not been able to correlate this finding to a specific EIA kit, a particular submitting laboratory, or anomalies that may occur in storage or transport time. Further investigation to better understand this discrepancy may help elucidate the issue to prevent it in the future.

The top six non-O157 serogroups identified by the WC from 2005 to 2010 were O103, O26, O111, O45, O121, and O145, which is consistent with national non-O157 STEC isolations. These non-O157 serogroups are also the same six serogroups that were declared adulterants by the U.S. Department of Agriculture (USDA) in September 2011. In 2005, expanded STEC serogroup testing was implemented, and this additional testing has allowed for the real-time identification of 29 other serogroups. At the WC, identification of non-O157 STEC has increased from the two to three identified in 2000–2001, i.e., prior to this study (data not shown), to 95 definitively identified in 2010, as a result of this enhanced STEC testing approach. Of note, utilization of this enhanced testing approach allowed for the identification of the serogroup and PFGE type as part of non-O157 STEC outbreak investigations that otherwise might not have been supported by laboratory data (Schaffzin et al., 2012). Our laboratory has also implemented real-time PCR methods to detect the top six non-O157 serogroups, which has improved timeliness of identification and outbreak investigation (data not shown). The continued development and implementation of molecular STEC serogrouping methods like these may one day eliminate the need to stock O-grouping antisera and allow rapid serogrouping capability for STEC.

The World Health Organization has called the rapid identification of virulent non-O157 STEC a public health priority (WHO, 1998). Recent recommendations from the CDC suggest that future STEC methods should include an assessment of the potential of the organism to cause severe disease, possibly by detecting virulence factor genes. These recommendations also propose that comparative genomic studies performed on existing and newly sequenced STEC strains may identify gene targets that will likely aid in the identification or predictions of pathogenicity in the future (Gould et al., 2009). Based on these suggestions, we chose to design a virulence gene panel to assess if certain gene profiles identified in previous studies (Coombes et al., 2008; Newton et al., 2009; Ritchie et al., 2003; Wickham et al., 2006) were more common in some STEC serogroups or were detected more frequently.

We found that 98% of the O157 STEC contained stx2 and 81% of the non-O157 contained the stx1 gene. When assessing the other virulence genes in these STEC, we found that 100% of the O157 STEC tested harbored all four of the additional gene targets. The top six non-O157 serogroups harbored a high level of the hlyA, eaeA, and nleA genes (97–100%); however, we found nleB was positive for only 62% of the top six non-O157 STEC. The non-O157 STEC, other than the top six, had a much lower incidence of the virulence factor genes (13–56%) overall. NleA and NleB are members of a group of non-LEE encoded effector genes that are involved in the type III secretion system (Coombes et al., 2008). It has been hypothesized that NleB may be a molecule that also contributes to low infectious dose (Wickham et al., 2006). The finding that other non-O157 STEC tested showed a 13% nleB positivity rate compared to the 62% and 100% nleB positivity rates among the top non-O157 and O157 STEC tested, respectively, is suggestive of a virulence advantage for nleB-positive strains. Our study also found that while all O157 STEC and the top six non-O157 STEC contained nleA, the other non-O157 tested had a 24% nleA positivity rate. The presence of nleA may also provide a virulence advantage in O157 STEC and non-O157 STEC, causing the majority of human disease in our study, and this supports the finding that bacteria lacking nleA are unable to cause mortality in mice (Gruenheid et al., 2004). Future studies utilizing molecular methods to assess genetic markers of virulence should continue to elucidate the role of nleA and nleB as diagnostic tools for STEC virulence assessment.

We propose that early characterization of a virulence gene profile may provide insight into the potential for severe disease. Additionally, this type of profiling, or an expanded one, may be useful in early stages of non-O157 STEC outbreak investigations, or perhaps in food and environmental surveillance efforts (Gonzales et al., 2011). Anecdotally, we have found that the addition of the virPCR to the testing algorithm for suspected food sources is useful in identifying and isolating STEC containing the same profile as the patient isolate (data not shown). Rapidly screening a number of suspect bacteria isolated from a food can allow a more timely selection of the same bacteria in the food source that was isolated from ill patients. Used in this way, the virPCR provides a rapid test that can be utilized earlier in the testing algorithm for both food and patient specimens before testing such as PFGE can be performed.

This 6-year study highlights the contribution of non-O157 STEC to disease burden in NYS. Overall, we found 41% of the STEC identified in our laboratory to be non-O157 STEC, highlighting the importance of EIA testing for STEC in the clinical laboratory, as 72% of the EIA-positive broths received were determined to be non-O157. Without the submission of EIA-positive broths, these STEC would likely have been missed (over 85% of the non-O157 STEC isolated were submitted to WC as EIA-positive broths). The receipt of EIA-positive specimens allowed for enhanced identification and characterization of STEC, provided insight to non-O157 outbreak investigations, and helped to continue to reveal the true burden of STEC disease. Implementation of enhanced STEC testing will provide a better understanding of STEC disease, and will ultimately lead to identification of new and emerging serotypes and allow effective and timely outbreak response.

Footnotes

Acknowledgments

We would like to thank Dale Morse, Perry Smith, Shelley Zansky, NYS Epidemiology, and EIP staff members for their long-term support of the project. Finally, we would also like to thank Mehdi Shayegani, Jonathan Hibbs, Robyn Atkinson, and many past and present members of the Wadsworth Center Bacteriology Laboratory for support of this project. This work was in part supported by the Centers for Disease Control and Prevention, National Center for Infectious Diseases, Epidemiology and Laboratory Capacity for Infectious Disease Cooperative Agreement (grant U50/CCU223671), the Centers for Disease Control and Prevention Emerging Infections Program Cooperative Agreement (grant 3U01CI000311-05), and the Centers for Disease Control and Prevention Public Health and Emergency Preparedness Grant (grant U90TP216988).

Disclosure Statement

No competing financial interests exist.

References

Beutin

, Zimmermann

, Gleier

. Pseudomonas aeruginosa can cause false-positive identification of verotoxin (Shiga-like toxin) production by a commercial enzyme immune assay system for the detection of Shiga-like toxins (SLTs) Infection, 1996; 24:267–268.

Brooks

, Sowers

, Wells

, Greene

, Griffin

, Hoekstra

, Strockbine

. Non-O157 Shiga toxin–producing Escherichia coli infections in the United States, 1983–2002. J Infect Dis, 2005; 192:1422–1429.

Bugarel

, Beutin

, Fach

. Low-density macroarray targeting non-locus of enterocyte effacement effectors (nle genes) and major virulence factors of Shiga toxin–producing Escherichia coli (STEC): A new approach for molecular risk assessment of STEC isolates. Appl Environ Microbiol, 2010; 76:203–211.

[CDC] Centers for Disease Control and Prevention. University outbreak of calicivirus infection mistakenly attributed to Shiga toxin–producing Escherichia coli O157:H7—Virginia, 2000. MMWR Morb Mortal Wkly Rep, 2001; 50:489–491.

[CDC] Centers for Disease Control and Prevention. Importance of culture confirmation of Shiga toxin–producing Escherichia coli infection as illustrated by outbreaks of gastroenteritis—New York and North Carolina, 2005. MMWR Morb Mortal Wkly Rep, 2006; 55:1042–1045.

Coombes

, Wickham

, Mascarenhas

, Gruenheid

, Finlay

, Karmali

. Molecular analysis as an aid to assess the public health risk of non-O157 Shiga toxin–producing Escherichia coli strains. Appl Environ Microbiol, 2008; 74:2153–2160.

Deng

, Vallance

, Li

, Puente

, Finlay

. Citrobacter rodentium translocated intimin receptor (Tir) is an essential virulence factor needed for actin condensation, intestinal colonization and colonic hyperplasia in mice. Mol Microbiol, 2003; 48:95–115.

Gonzales

, Kulow

, Park

, Kaspar

, Anklam

, Pertzborn

, Kerrish

, Ivanek

, Döpfer

. A high-throughput open-array qPCR gene panel to identify, virulotype, and subtype O157 and non-O157 enterohemorrhagic Escherichia coli. Mol Cell Probes, 2011; 25:222–230.

Gould

, Bopp

, Strockbine

, Atkinson

, Baselski

, Body

, Carey

, Crandall

, Hurd

, Kaplan

, Neill

, Shea

, Somsel

, Tobin-D'Angelo

, Griffin

, Gerner-Smidt

. Centers for Disease Control and Prevention (CDC). Recommendations for diagnosis of Shiga toxin–producing Escherichia coli infections by clinical laboratories. MMWR Recomm Rep, 2009; 58:1–14.

10.

Gruenheid

, Sekirov

, Thomas

, Deng

, O'Donnell

, Goode

, Li

, Frey

, Brown

, Metalnikov

, Pawson

, Ashman

, Finlay

. Identification and characterization of NleA, a non-LEE-encoded type III translocated virulence factor of enterohaemorrhagic Escherichia coli O157:H7. Mol Microbiol, 2004; 51:1233–1249.

11.

Kelly

, Hart

, Mundy

, Marchès

, Wiles

, Badea

, Luck

, Tauschek

, Frankel

, Robins-Browne

, Hartland

. Essential role of the type III secretion system effector NleB in colonization of mice by Citrobacter rodentium. Infect Immun, 2006; 74:2328–2337.

12.

Mead

, Slutsker

, Dietz

, McCaig

, Bresee

, Shapiro

, Griffin

, Tauxe

. Food-related illness and death in the United States. Emerg Infect Dis, 1999; 5:607–625.

13.

Newton

, Sloan

, Bulach

, Seemann

, Allison

, Tauschek

, Robins-Browne

, Paton

, Whittam

, Paton

, Hartland

. Shiga toxin–producing Escherichia coli strains negative for locus of enterocyte effacement. Emerg Infect Dis, 2009; 15:372–380.

14.

Ritchie

, Thorpe

, Rogers

, Waldor

. Critical roles for stx2, eae, and tir in enterohemorrhagic Escherichia coli–induced diarrhea and intestinal inflammation in infant rabbits. Infect Immun, 2003; 71:7129–7139.

15.

Scallan

, Hoekstra

, Angulo

, Tauxe

, Widdowson

, Roy

, Jones

, Griffin

. Foodborne illness acquired in the United States—Major pathogens. Emerg Infect Dis, 2011; 17:7–15.

16.

Schaffzin

, Coronado

, Dumas

et al. Public health approach to detection of non-O157 Shiga toxin–producing Escherichia coli: Summary of two outbreaks and laboratory procedures. Epidemiol Infect, 2012; 140:283–289.

17.

Schmidt

, Scheef

, Huppertz

, Frosch

, Karch

. Escherichia coli O157:H7 and O157:H(-) strains that do not produce Shiga toxin: Phenotypic and genetic characterization of isolates associated with diarrhea and hemolytic-uremic syndrome. J Clin Microbiol, 1999; 37:3491–3496.

18.

[WHO] World Health Organization. Zoonotic non-O157 Shiga toxin–producing Escherichia coli (STEC) Report of a WHO Scientific Working Group Meeting, Berlin, Germany, 1998; 1–35. http://whqlibdoc.who.int/hq/1998/WHO_CSR_APH_98.8.pdf. 2012 August 15.

19.

Wickham

, Lupp

, Mascarenhas

, Vazquez

, Coombes

, Brown

, Coburn

, Deng

, Puente

, Karmali

, Finlay

. Bacterial genetic determinants of non-O157 STEC outbreaks and hemolytic-uremic syndrome after infection. J Infect Dis, 2006; 194:819–827.