Evaluation of Enzyme Immunoassays and Real-Time PCR for Detecting Shiga Toxin-Producing Escherichia coli in Southern Alberta, Canada

Abstract

Two immunoassays (Shiga Toxin Chek and Shiga Toxin Quik Chek) and real-time PCR were used to detect Shiga toxin-producing Escherichia coli. For enriched culture, the sensitivity and specificity of the three methods ranged from 80.0% to 98.2% and 98.0% to 100.0%, respectively. STEC isolates were identified in 2.6% of the 784 samples.

INTRODUCTION

Non-O157 Shiga toxin-producing Escherichia coli (STEC) is an emerging cause of enteric and systemic illness and account for 50% of STEC infections (1, 2). Serotypes O104, O121, O26, O145 and O157 (1–6) have been linked to outbreaks. STEC-related disease outcomes can result in hemolytic-uremic syndrome (HUS) (7–9) followed by other complications (10–13) affecting various organs (9, 12, 14–17). STEC can be transmitted via foods (3, 4, 18–20), water (21), animals (22–25), and from person to person (26–28). Ruminants are natural carriers of STEC and are considered the main reservoirs for these pathogens (29).

Conventional culture methods focus mainly on the O157 serotype, and non-O157 STEC serotypes are underreported (30). The Centers for Disease Control and Prevention guidelines from October 2009 recommend simultaneous culture of stool samples and detection of Shiga toxins and/or their genes for all STEC isolates (31). Amplification and enzyme immunoassay (EIA) kits for STEC detection are commercially available (30, 32–34).

The objective of this study was to evaluate the performance of two EIAs, Shiga Toxin Chek and Shiga Toxin Quik Chek (TechLab, Inc., Blacksburg, VA), along with our in-house stx₁ and stx₂ real-time PCR in determining the prevalence of STEC in the Lethbridge region in southern Alberta, Canada.

The Shiga Toxin Chek assay was performed using the DS2 automated enzyme-linked immunosorbent assay (ELISA) system (Dynex Technologies, Inc., Chantilly, VA). Shiga Toxin Quik Chek is a rapid membrane EIA for simultaneous detection of Stx1 (subtypes 1a, 1c, and 1d) and Stx2 (subtypes 2a, 2c, 2d, and 2e). Stool samples (n = 784) were collected in containers from 1 June to 31 August 2012 at the Chinook Regional Hospital in Lethbridge, and all duplicate patient samples were removed. The list of bacteria for full routine enteric workup is shown in Table 1. BBL CHROMagar O157 (Becton Dickinson, Inc., Mississauga, ON, Canada) agar was used to detect E. coli O157, and results were further confirmed by direct antibody agglutination (BD Difco, Burlington, ON, Canada). Stool samples were directly tested for STEC using both EIAs (Shiga Toxin Chek and Shiga Toxin Quik Chek). Stool culture enrichment was performed by inoculating 4.5 ml of MacConkey broth with 200 μl of watery/mucoid stool or a pea-sized solid/semiformed stool and incubating it for 16 to 24 h at 37°C (30). The culture was not treated with antibiotics. A 200-μl aliquot of the broth culture was used for the EIAs, and DNA extraction for real-time PCR. DNA template was prepared (35) and amplified as described previously (30, 36). Positive- and negative-control STEC stool samples were included for extraction and amplification; and positive DNA and water controls were incorporated for amplification. Real-time PCR was conducted on enriched cultures only at the Provincial Laboratory for Public Health. EIAs were performed according to the manufacturer's instructions. Specificity of both EIAs and real-time PCR was determined using a panel of bacteria as previously described (30). The limits of detection (LOD) for all three assays were determined by performing replicates on three different days using cell suspensions (E. coli O157 Sakai strain) containing from 10⁸ cells to 1 cell. The last dilution that showed a positive result was determined to be the LOD for that assay.

TABLE 1.

Culture results from clinical specimens

Organism	No. of positive samples (n = 74)	% of positivity
E. coli O157	7	9.5
Aeromonas	4	5.4
Campylobacter	41	55.4
Salmonella spp.	19	25.7
Shigella sonnei	1	1.4
Yersinia	1	1.4
Plesiomonas	1	1.4

Figure 1 shows the isolation of STEC strains from positive samples using chromogenic agar (BBL CHROMagar O157 or Colorex O157 [Dalynn Biologicals, Calgary, AB, Canada] and Colorex STEC [Alere Canada, Ottawa, ON, Canada]), as identified by any of the above-described assays, and were further subtyped by PCR (37). Briefly, positive samples were plated on BBL CHROMagar O157, Colorex O157, and Colorex STEC solid agar media. Five individual mauve colonies were picked from each of the above-mentioned plates. DNA was extracted by resuspending individual colonies in rapid lysis buffer and used as a template for the TaqMan real-time PCR to confirm the stx status. If no mauve colonies were found, five randomly picked blue colonies from Colorex O157 agar were created as a single pool for stx real-time PCR, and a total of four pools were included. Once a positive pool was identified, single-colony PCR was performed to identify the positive STEC isolates. All mauve colonies were tested for the O157 genotype by O157 direct antibody agglutination (BD Difco, Burlington, ON, Canada). If negative, serotyping was done at the National Microbiology Laboratory in Winnipeg, Canada.

FIG 1.

Isolation of STEC-positive samples identified by immunoassays (Shiga Toxin Chek or Shiga Toxin Quik Chek) or real-time PCR.

There were 74 positive samples detected using routine culture. The CHROMagar O157 agar plate was used for STEC detection; seven positive O157 stools were identified, but the remaining 13 non-O157 STEC isolates went undetected. Campylobacter infection ranked highest at 55.4%, followed by Salmonella (25.7%) and E. coli O157 (9.5%) (Table 1). If non-O157 STEC isolates were included as part of routine screening, 87 (11.1%) samples were positive for enteric bacteria, including 20 STEC infections (27.0%), ranking it second to Campylobacter infection. Among the STEC strains isolated, 65% were non-O157.

Both EIAs showed no cross-reactivity to the panel of bacteria included. The LOD for both EIAs was at 7.0 × 10⁵ CFU/reaction, compared to 3 and 30 CFU/reaction for stx₁ and stx₂ real-time PCR, respectively.

Sensitivities and specificities are shown in Table 2. A result was considered to be a true positive when at least two of the three assays were positive. Using Shiga Toxin Chek on stool samples, the sensitivity and specificity were 70.0% and 99.4%, respectively, while the positive and negative predictive values (PPV and NPV, respectively) were 73.7% and 99.2%, respectively. With enriched culture, the sensitivity increased to 80.0%, with a specificity of 98.2%, while the PPV and NPV were 53.3% and 99.5%, respectively. The Shiga Toxin Quik Chek performed similarly to the Shiga Toxin Chek in stool samples but with a PPV and NPV of 93.3% and 99.2%, respectively. When enriched culture was used with the Shiga Toxin Quik Chek, the sensitivity increased to 85.0%, with 100% specificity and a PPV and NPV of 100% and 99.6%, respectively. Subtypes 2b, 2f, and 2g were not identified by the EIAs. The hands-on time for the Shiga Toxin Chek was approximately 10 min for 50 samples. An average of 10 min for 25 samples is required for the Shiga Toxin Quik Chek assay.

TABLE 2.

Comparison of immunoassays and real-time PCR

Assay	Result (%) for assay
	Sensitivity	Specificity	PPV	NPV
Shiga Toxin Chek
Stool	70.0	99.4	73.7	99.2
Broth	80.0	98.2	53.3	99.5
Shiga Toxin Quik Chek
Stool	70.0	99.9	93.3	99.2
Broth	85.0	100.0	100.0	99.6
Real-time PCR (broth)	95.0	100.0	100.0	99.9

Real-time PCR on enrichment broth culture had a sensitivity and specificity of 95.0% and 100%, with a PPV and NPV of 100% and 99.9%, respectively. This assay detected 23/24 positive STEC isolates in a total of 784 samples. Our in-house real-time PCR assay failed to detect one STEC sample that contained an stx_1c gene. The setup time for real-time PCR is approximately 20 min for 25 samples.

Subtyping results are shown in Table 3. Eleven of the 24 (45.8%) positive isolates carried both stx_1a and stx_2a genes: seven were E. coli O157:H7, and the remaining four were serotypes OR:H7, O22:H2, O27:H8 and O111:H8. The next most frequent subtype was stx_1a (29.2%), followed by stx_2a (12.5%) and stx_2c (8.3%). There was a single stx_1c isolate associated with O174:H8.

TABLE 3.

Patient demographic data and characterization of STEC strains in this study^a

Serotype	Age (yr)	Gender	Clinical presentation	stx1/2	stx subtype(s)
O157:H7	3	Female	Bloody diarrhea	+/+	1a, 2a
O157:H7	9	Female	Watery diarrhea	+/+	1a, 2a
O157:H7	18	Male	Watery diarrhea	+/+	1a, 2a
O157:H7	21	Male	Watery diarrhea	+/+	1a, 2a
O157:H7	16	Female	Diarrhea with abdominal pain	+/+	1a, 2a
O157:H7 and ORough:H7	52	Female	Bloody diarrhea	+/+ and +/+	1a, 2a and 1a, 2a
O157:H7	58	Female	Bloody diarrhea	+/+	1a, 2a
O22:H2	58	Female	Bloody diarrhea	+/+	1a, 2a
O27:H8 and O111:H8	20	Male	Diarrhea for 5 days	+/+ and +/+	1a, 2a and 1a, 2a
O174:H8	76	Female	Watery diarrhea	+/−	1c
O68:HNM and ORough:HNM	1	Female	Diarrhea	−/+ and −/+	2c and 2c
O145:HNM and O145:H37	10	Male	Bloody diarrhea	−/+ and −/+	2a and 2a
O121:H19	5	Male	Bloody diarrhea	−/+	2a
O26:H11	2	Male	No history	+/−	1a
O26:H11	72	Female	Watery diarrhea	+/−	1a
O26:H11	1	Male	Watery diarrhea	+/−	1a
O26:H11	2	Male	No history	+/−	1a
O111:HNM (Salmonella coinfection)	1	Male	Diarrhea with fever	+/−	1a
O111:HNM (P. shigelloides coinfection)	43	Male	Watery diarrhea, travel to Mexico	+/−	1a
O128ab:H2	1	Female	Bloody diarrhea	+/−	1a

^aNote that four patients were positive for two different STEC serotypes, and two patients had coinfections with Salmonella or Plesiomonas shigelloides.

The positivity rate of STEC infection is 2.6%. Twenty patients were positive for STEC: 7 (35%) isolates were O157, and 13 (65%) were non-O157 (Table 3). Four stool samples were positive for two different STEC serotypes (O157:H7 and ORough:H7, O111:H8 and O27:H8, O68:HNM and ORough:HNM, and O145:HNM and O145:H37). Two patients had coinfections with other bacterial species (O111:HNM and Salmonella; O111:HNM and Plesiomonas shigelloides). There was a 1:1 gender distribution ratio. Patient age ranged from 1 to 76 years, with a mean of 23.5 years, and 53% were less than 10 years old.

Our study evaluated the performance of two EIA kits and real-time PCR for detecting all serotypes of STEC. Shiga Toxin Chek on the DS2 ELISA automation system can easily be adopted for routine testing in a high-volume laboratory. The assay is robust, with minimal hands-on-time, and the data are easily interpreted. The Shiga Toxin Quik Chek is a lateral flow device, is easy to perform and interpret, and does not require special equipment. It is most appropriate for a low-volume testing laboratory. Both assays can be used directly on stools or on overnight enriched cultures and performed better than another commercially available system for Shiga toxin detection—the ImmunoCard STAT! assay (30). Assay performance is also improved with enriched culture. Although the real-time PCR assay has 95.0% sensitivity and 100.0% specificity, implementation may not be possible in laboratories that lack staff with molecular training and without amplification platforms. With their ease of utilization, the EIAs would be more practical for implementation in frontline screening.

The prevalence of STEC infection in Lethbridge, southern Alberta, is 2.6%, slightly higher than the 2.08% reported in 2012 (30). In northern Alberta, the prevalence was at 0.9% in 2011 (34); this lower rate might be due to the smaller number of livestock feedlot operations in this area. Of the 2.6% STEC infections identified, 35% were O157, while the remaining 65% were non-O157. We have identified three (O26, O111, and O145) out of the top six non-O157 STEC serogroups (O26, O45, O103, O111, O121, and O145) reported by the United States as adulterants in meat by the Food Safety Inspection Services (FoodNet). Coinfections with different serotypes or another bacterial species were detected. We further noted that patients infected with STEC might not present with bloody diarrhea as suggested by previous studies (30, 34, 36).

In conclusion, the EIAs evaluated in this study are viable alternatives to amplification assays for frontline microbiology laboratories as a primary screening method for STEC. However, the challenge still remains for the reference laboratories to find a less labor-intensive method to isolate and identify the specific type of non-O157 STEC isolates.