Development of a multiplex loop-mediated isothermal amplification assay to detect shiga toxin-producing Escherichia coli in cattle.

A multiplex loop-mediated isothermal amplification (mLAMP) assay was developed for simultaneous detection of the stx1 and stx2 genes and applied for detection of shiga toxin-producing Escherichia coli (STEC) in cattle farm samples. Two target genes were distinguished based on Tm values of 85.03 ± 0.54°C for stx1 and 87.47 ± 0.35°C for stx2. The mLAMP assay was specific (100% inclusivity and exclusivity), sensitive (with a detection limit as low as 10 fg/μL), and quantifiable (R² = 0.9313). The efficacy and sensitivity were measured to evaluate applicability of the mLAMP assay to cattle farm samples. A total of 12 (12/253; 4.7%) and 17 (17/253; 6.7%) STEC O157, and 11 (11/236; 4.7%) non-O157 STEC strains were isolated from cattle farm samples by conventional selective culture, immunomagnetic separation, and PCR-based culture methods, respectively. The coinciding multiplex PCR and mLAMP results for the types of shiga toxin revealed the value of the mLAMP assay in terms of accuracy and rapidity for characterizing shiga toxin genes. Furthermore, the high detection rate of specific genes from enrichment broth samples indicates the potential utility of this assay as a screening method for detecting STEC in cattle farm samples.

Introduction

Shiga toxin-producing Escherichia coli (STEC) is a food-borne zoonotic pathogen that frequently causes human illness ranging from mild gastrointestinal problems to serious fatal complications [20,21]. It is estimated that approximately 176,000 cases of foodborne illnesses caused by STEC occur annually in the United States [18]. Among the 2,400 hospitalizations and 20 deaths caused annually by STEC, serotype O157 is implicated in roughly 35% of illnesses, 89% of hospitalizations, and 100% of deaths [18]. Because cattle are known to be a natural reservoir of STEC, asymptomatically infected cattle can transmit STEC to humans [14,22]. Therefore, investigating the prevalence of STEC in cattle and their environment is important for control and prevention of STEC transmission to humans.

STEC harbors essential virulence genes stx1 and/or stx2 [17]. STEC O157 is traditionally detected by a selective culture method based on the inability to ferment sorbitol [17]; however, detection of non-O157 STEC is difficult due to the lack of phenotypic characterization. Fast, sensitive, and specific nucleic acid-based amplification methods have recently been adopted to supplement the conventional culture method. The standard manuals established by the United States Department of Agriculture (USDA) and the United States Food and Drug Administration (FDA) recommend polymerase chain reaction (PCR) and real-time PCR (qPCR) as screening methods for detection of STEC by targeting various genes, including stx1 and stx2 [5,6].

The loop-mediated isothermal amplification (LAMP) assay is a promising novel nucleic acid amplification method in terms of its accuracy, simplicity, and rapidity [16]. The use of 4~6 specially designed primers (FIP, BIP, F3, B3, LF, and LB) results in high efficiency of the assay, which can amplify low titers of DNA templates [15]. The LAMP assay recognizes 6~8 distinct regions of the DNA template, thereby providing high specificity and sensitivity [15,16]. In addition, a special DNA polymerase, Bst polymerase or Gsp polymerase, reacts under isothermal conditions at 60℃ to 65℃ within 1 h [8,16]. The amplified product can be detected by simple inspection of insoluble magnesium pyrophosphate, gel electrophoresis, or measuring the turbidity of the product [7,16,23].

Several groups have used LAMP assays targeting stx1, stx2 [7,13,23], eae [23], rfbE [11], and serogroup-specific genes wzx and wzy [24] to detect STEC, and have shown that the LAMP assay is more sensitive than PCR or qPCR.

Despite the many advantages of the LAMP assay, multiplex LAMP (mLAMP) approaches are limited due to difficulty of differentiating more than two specific genes. Visual observation of LAMP products is indirect and cannot be used to distinguish multiple target genes. Gel electrophoresis after restriction enzyme digestion has been attempted [9,10,19], but is laborious and time-consuming. Moreover, the high sensitivity of the LAMP assay raises the concern of carry-over contamination during post-amplification analysis [12]. Probe-based detection is also not applicable due to the various types of cauliflower-like structures of LAMP products [12].

In this study, we developed a real-time mLAMP assay to detect stx1 and stx2 simultaneously based on different annealing temperatures determined by annealing curve analysis. To evaluate the applicability of the mLAMP assay to cattle farm samples, we assessed assay efficacy and sensitivity using cattle feces and environmental samples.

Materials and Methods

Bacterial strains

Bacterial strains used for inclusivity and exclusivity testing are listed in Table 1. Three STEC strains, ATCC 43984 (harboring both stx1 and stx2 genes), ATCC 43890 (stx1), and ATCC 43889 (stx2), were tested to evaluate the mLAMP assay. STEC ATCC 43894 was also tested to optimize the mLAMP assay and determine the detection limit.

Table 1

Bacterial strains used for inclusivity and exclusivity test

Sample collection

A total of 253 cattle farm samples (237 cattle feces, 13 farm soil samples, one raw milk, one water, and one forage sample) were collected from 15 cattle farms located in Gyeonggi-do, South Korea, from August 2012 to May 2013. Cow fecal samples were collected directly by rectal retrieval, and at least one environmental sample from each farm was collected. All samples were transported to the laboratory at 4℃ for microbiological tests and analyzed immediately upon arrival.

Isolation of STEC O157 strains

STEC O157 was isolated by conventional selective culture (conventional) and immunomagnetic separation (IMS) methods. For the conventional method, approximately 1 g of each sample was homogenized in 9 mL modified EC broth (mEC; Becton, Dickinson and Company, USA) supplemented with novobiocin (20 mg/L, Oxoid, UK) and incubated overnight at 37℃. Following incubation, one loop of mEC broth culture was streaked onto sorbitol MacConkey agar (Becton, Dickinson and Company, USA) supplemented with potassium tellurite (T-SMAC; 2.5 mg/L, Sigma-Aldrich, Canada) and incubated at 37℃ overnight. A maximum of six typical colonies were subcultured onto MacConkey agar (MAC; Becton, Dickinson and Company, USA) and CHROMagar O157 (CHROM; CHROMagar Microbiology, France) and then incubated overnight at 37℃. Typical colonies, pink in MAC and mauve in CHROM, were selected for testing using the E. coli O157 latex test kit (Oxoid, UK).

For the IMS method, Dynabeads MAX anti-E. coli O157 (Dynal; Invitrogen, USA) was used according to the manufacturer's instructions. The suspension of immunomagnetic beads was spread onto T-SMAC and incubated at 37℃ overnight. Up to four typical colonies were selected and identified by applying the same criteria as described in the conventional method. If the four colonies yielded no STEC O157 strain, up to four additional colonies were tested.

Isolation of non-O157 STEC strains

Non-O157 STEC strains were isolated from STEC O157-negative samples by PCR-based culture assay as previously described, with minor modification [3]. Briefly, one loop of enriched mEC broth culture was streaked onto T-SMAC and incubated overnight at 37℃. DNA was extracted from randomly swiped areas of heavy bacterial growth and tested for the presence of shiga toxin genes by multiplex PCR (mPCR) as described in the PCR assay section below. Plates possessing shiga toxin genes were subcultured onto MAC and incubated overnight at 37℃. The presumptive colonies were tested for the presence of shiga toxin genes by mPCR. Colonies possessing shiga toxin genes were assessed by conventional agglutination tests to determine the serotype of the isolates using commercial antiserum (Joongkyeom, Korea).

DNA extraction

The boiling method was used to extract DNA of gram-negative bacteria and enrichment broth culture. Briefly, a loopful of bacterial cells grown on tryptic soy agar (Becton, Dickinson and Company, USA) plates or a pellet of 1 mL enrichment broth culture was suspended in 1 mL normal saline and centrifuged at 6,000 × g for 3 min. The pellet was then resuspended with 200 µL sterile water, boiled for 10 min, and centrifuged at 6,000 × g for 3 min. The supernatant was used as a template for the PCR and LAMP assays. Instagene matrix (Bio-Rad Laboratories, USA) was used to extract DNA from gram-positive bacteria. The genomic DNA of three STEC strains (ATCC 43890, 43889, and 43894) was extracted with DNeasy blood and tissue kit (Qiagen, Germany) according to the manufacturer's instructions.

Primer design and the LAMP assay

To design primer sets targeting shiga toxins genes, the genomic sequences of stx genes from various serotypes of STEC were collected from GenBank. For the stx1 gene, the genomic sequences of STEC O26, O79, O103, O111, and O157 (GenBank Accession No. AP010953, FR875154, FE94195, AP010958, AP010960, and M19473) were collected and the genomic sequences of STEC O26, O157, and O178 (GenBank Accession No. FR850034, AB048240, X07865, FR850037) were collected for the stx2 gene. The consensus sequence of each gene was generated by multiple alignment and comparison using ClustalW method in MegAlign program (Lasergene 10.1.1, USA). Each primer set was designed from the consensus sequence using LAMP designer (Optigene, UK; Table 2).

Table 2

Primer sets used for the polymerase chain reaction (PCR) and the loop-mediated isothermal amplification (LAMP) assay

The mLAMP assay was evaluated using two primer sets targeting stx1 and stx2. To differentiate between targets, 100 ng/µL of DNA from three STEC strains (ATCC 43984, ATCC 43890, and ATCC 43889) were tested in triplicate. The mLAMP assay was performed on a Genie II instrument (Optigene, UK) in a total reaction volume of 25 µL containing 1× Buffer, 1 M Betaine, 4 mM MgSO4, 0.8 mM each deoxynucleoside triphosphate (dNTP), 0.8 µM each of FIPs and BIPs, 0.4 µM each of LFs and LBs, 0.2 µM each of F3s and B3s, 1× EvaGreen (SolGent, Korea), 8 U Gsp polymerase (Optigene, UK), and 4 µL template DNA. Distilled water (Invitrogen, USA) was used as a negative control. The LAMP reactions were carried out at 64℃ for 30 to 60 min, with annealing curve analysis was conducted from 98℃ to 80℃.

To determine the detection limit of the mLAMP assay, 10-fold serially diluted DNA templates with concentrations ranging from 100 ng/µL to 1 fg/µL were tested in triplicate. A standard curve was generated by plotting Tp values versus concentration of template DNA ranging from 100 ng/µL to 1 fg/µL on a log scale for each set, and the correlation coefficient of determination (R2) was calculated. The detection limit was evaluated and compared to mPCR described in the PCR assay section below.

PCR assay

Multiplex PCR (mPCR) targeting stx1 and stx2 was performed to determine the detection limit using serially diluted template DNA of STEC ATCC 43894 as for the mLAMP assay, and results of both assays were compared. mPCR was also used to identify STEC in STEC-positive enrichment broth cultures. For genotypic characterization of STEC isolates from cattle farm samples, stx1, stx2, eae, and uidA were tested, as described, on BAM (Table 2) [6].

The reaction mixture included 1× Emerald Master Mix (Takara Bio, Japan), 0.5 µM each primer, and 1 µL template. Distilled water was added to a final volume of 20 µL. The optimal reaction conditions were defined as follows: pre-denaturation at 95℃ for 5 min, followed by 30 cycles of 95℃ for 30 sec, 57℃ for 30 sec, and 72℃ for 45 sec, and then final extension at 72℃ for 5 min. mPCR was performed on a MyCycler thermal cycler (Bio-Rad Laboratories, USA) and the products were electrophoresed in 1.5% agarose gels.

mLAMP assay application in cattle farm samples

The mLAMP assay was used to verify the types of shiga toxin in all STEC isolates relative to the mPCR assay. To evaluate assay sensitivity, the presence of stx1 and/or stx2 in enrichment broth cultures was assessed by mLAMP and mPCR.

Results

mLAMP evaluation

To evaluate the utility of the mLAMP assay, three STEC strains with different shiga toxin types were tested. The results are shown with the Tp value, which represents the time taken to detect the amount of fluorescence required to produce a positive peak, and the Tm value (annealing temperature) determined by annealing curve analysis. For well 1 (ATCC 43894; stx1 and stx2 genes), the mean Tp value was 9:06 (min:sec) and Tm values (mean) were observed at 84.55 ± 0.32℃ and 87.22 ± 0.21℃. For well 2 (ATCC 43890; stx1), the mean Tp and Tm values were 8:57 and 85.59 ± 0.02℃, respectively. For well 3 (ATCC 43889; stx2), the mean Tp and Tm values were 9:23 and 87.71 ± 0.17℃, respectively. There was no product amplification in the negative control (Fig. 1).

Fig. 1

Discrimination of the stx1 and stx2 by distinct Tm values generated by annealing curve analysis of the mLAMP assay. The annealing curve shows temperature on the X-axis and fluorescence on the Y-axis. The peaks indicate Tm values generated by annealing curve analysis. Well 1 (red), stx1 and stx2 (STEC 43894); well 2 (green), stx1 (STEC 43890); Well 3 (blue), stx2 (STEC 43889); Well 4 (purple), NC (negative control).

Inclusivity and exclusivity of the mLAMP assay

To evaluate the inclusivity and exclusivity of the mLAMP assay, 48 bacterial strains were tested. mLAMP amplified and distinguished the shiga toxin genes in all 19 STEC strains according to their shiga toxin type. The mean Tp value was 10:12 (8:24 to 15:46) and the mean Tm value was 85.32 ± 0.60℃ for stx1 and 87.77 ± 0.38℃ for stx2, demonstrating 100% sensitivity of the mLAMP assay. Tm values were not generated for 29 gram-negative and gram-positive strains of non-shiga toxin-producing bacteria, demonstrating 100% specificity.

Detection limit of the mLAMP assay

The detection limit of the mLAMP assay was found to be 10 fg/µL in three independent experiments. At 1 fg/µL, stx2 was detected in all three experiments, but stx1 was detected in only one of three experiments. Fig. 2 shows a general amplification and annealing curve generated by the LAMP instrument. The DNA templates were amplified with mean Tp values ranging from 9:06 to 22:16 for concentrations ranging from 100 ng/µL to 10 fg/µL. The mean Tm values were 84.63 ± 0.30℃ for stx1 and 87.24 ± 0.25℃ for stx2. The quantification equation for the mLAMP assay was determined to be Y = -1.8145X + 23.671, with a correlation coefficient (R2) of 0.9313. For the mPCR assay, the detection limit was 1 pg/µL in three independent experiments.

Fig. 2

Detection limit and standard curve of the mLAMP assay targeting stx1 and stx2 genes of STEC. Ten-fold serial dilutions of STEC (ATCC 43894) template DNA from 100 ng/µL to 1 fg/µL tested in triplicate. (A) The amplification curve shows time in minutes on the X-axis and the amount of fluorescence detected (K) on the Y-axis. The amplification curve generated from 100 ng/µL to 10 fg/µL. (B) Discrimination of shiga toxin genes was based on the annealing curve showing 2 distinct Tm values (84.63 ± 0.30℃ for stx1 and 87.24 ± 0.25℃ for stx2). (C) The standard curve shows the amount of template DNA in log scale on the X-axis and Tp values (min:sec) on the Y-axis. DNA concentrations 1 to 8 correspond to template DNA from 10 fg/µL to 100 ng/µL in log scale.

A total of 253 cattle farm samples (237 cattle feces and 16 farm environmental samples) were tested for STEC. The conventional culture method yielded 12 (4.7%) STEC O157 strains from cattle fecal samples. The IMS method yielded 17 (6.7%) STEC O157 strains from cattle feces and farm soil samples. The IMS method yielded 17 (6.7%) STEC O157 strains from cattle feces and ground samples. Eleven (4.7%) non-O157 STEC strains were isolated from a total of 236 STEC O157-negative cattle fecal samples by the PCR-based culture method. As shown in Table 3, the non-O157 STEC isolates belonged to four different serotypes, O84 (2 isolates), O108 (5 isolates), O185 (1 isolate), and O119 (3 isolates).

Table 3

Characteristics of shiga toxin-producing Escherichia coli (STEC) isolated from cattle feces and environmental samples in Gyeonggi-do, Korea

*Twenty-eight STEC strains were isolated from a total of 253 cattle farm samples from 15 enrolled cattle farms by three different methods. For isolation of STEC O157, a conventional culture method based on SMAC agar media was used; while immunomagnetic beads coated with anti-O157 were used for the immunomagnetic separation (IMS) method. †A polymerase chain reaction (PCR)-based culture method was used for isolation of non-O157 STEC strains. From a total of 253 samples, 17 STEC-O157 positive samples were excluded from the PCR assay. ‡Two strains with different serotypic and genotypic characterizations were isolated from the same sample. §Shiga toxin types were determined by mPCR and the mLAMP assay, which were concordant. UT: untested.

Among 28 STEC isolates, seven (25.0%) and 13 (46.4%) harbored stx1 and stx2, respectively, while eight (28.6%) strains harbored both stx1 and stx2 (Table 3). All 17 STEC O157 isolates had both eae and uid genes, while non-O157 STEC did not harbor these genes. The shiga toxin types of 28 STEC isolates were tested by mPCR and mLAMP. The mLAMP determinations were identical to those of the mPCR assay (Table 3). The Tp values of 28 isolates ranged from 8:21 to 12:03 (mean Tp = 10:03), and the mean Tm values were 85.76 ± 0.42℃ for stx1 and 87.73 ± 0.29℃ for stx2.

Enrichment broth cultures that were positive for STEC were used to evaluate the sensitivity of the mLAMP assay relative to mPCR. Among the 12 enrichment broth cultures that were positive for STEC O157 by the conventional culture method, shiga toxin genes were detected in three (25.0%) by mPCR and in 12 (100.0%) by mLAMP. For the 17 enrichment broth cultures that were positive for STEC O157 by the IMS method, shiga toxin genes were detected in four (23.5%) and 15 (88.2%) samples by mPCR and mLAMP, respectively. For the ten enrichment broth cultures that were positive for non-O157 STEC, mLAMP detected shiga toxin genes in nine (90.0%) samples, while none were detected by mPCR assay (Table 4).

Table 4

Sensitivity of mLAMP and mPCR assays in STEC-positive enrichment broth cultures

*Positive STEC from enrichment broth (mEC broth) was categorized according to the method used for STEC isolation. A conventional culture method and immunomagnetic separation method (IMS) were used for isolation of O157 STEC strains, and PCR-based culture method was used for isolation of non-O157 STEC strains. †In enrichment broth culture, detection of stx1 and/or stx2 was considered positive in both mLAMP and mPCR assay.

Discussion

We developed a real-time mLAMP assay for simultaneous detection of the STEC stx1 and stx2 genes. This assay is highly specific, sensitive, rapid, and quantifiable. When applied to cattle farm samples, the mLAMP assay provided determinations of shiga toxin types within 30 min and provided a high detection rate from enrichment broth culture.

Our mLAMP assay was designed for simultaneous detection of two different types of shiga toxin genes. As shown in Fig. 1, two sets of primers successfully amplified two target genes, which could be distinguished by their different Tm values. The mean Tm values for stx1 and stx2 in all tested samples were 85.03 ± 0.54℃ and 87.47 ± 0.35℃, respectively, and were thus easily distinguished. Similarly, in a recent study of sex determination in the plant Carica papaya, two target genes were also differentiated by annealing curve analysis [8].

The limit of detection in our mLAMP assay was 10 fg/µL, which was 100 times more sensitive than mPCR. Kouguchi [12] reported a detection limit of 100 CFU/mL in STEC pure culture. In our study, the detection limit was measured at the DNA concentration level rather than the CFU level, so our results cannot be compared directly. However, in our preliminary tests, 6.2 ± 1.74 × 108 CFU/mL corresponded to 140.5 ± 54.7 ng/µL (data not shown). Thus, 10 fg/µL corresponds to 10 CFU/mL, making our assay 10 times more sensitive than the one reported by Kouguchi [12].

In this study, the mean Tp values ranged from 9:06 to 22:16 for the concentration range of 100 ng/µL to 10 fg/µL, which was much faster than previously reported for a LAMP assay conducted using a turbidimeter, in which 21.4 ~ 45.2 min was required to detect the stx2 gene from a 10 ~ 105 CFU/reaction [23]. These results are consistent with those of other studies designed to detect Vibrio parahaemolyticus or ammonia-oxidizing bacteria in which the detection time for the fluorescent-based LAMP assay was faster than that of the turbidity-based LAMP assay [1,2]. In previous mLAMP studies based on the post-amplification analysis of two target genes, detection limits were 10 times more sensitive than those of PCR for detecting Salmonella or Shigella spp. and 103 to 105 times more sensitive for differentiating Babesia spp. [10,19].

The quantitative capability of the LAMP assay was reported previously [2,23,24]. In the present study, a standard curve was generated based on triplicate Tp values (detection time) versus the concentration of template DNA. The standard curve had a linear relationship with R2 values of 0.9313 within the range of 100 ng/µL and 10 fg/µL, indicating that the mLAMP assay can estimate unknown STEC contamination levels in samples. Indeed, our result is unique in terms of its wide coverage of concentrations. In another study, the R2 values fell within a relatively narrow range of concentrations (102 to 105 CFU/mL), possibly due to the delayed detection time observed at lower concentrations [2,23].

Until now, only a few LAMP methods for detecting STEC have been used to analyze samples from beef, beef trimmings, lettuce, spinach, and human stool [23,24,25]. However, to the best of our knowledge, no LAMP assay has been used to screen for STEC on cattle farms. As a primary source of human STEC infection, cattle and their environment are important to public health and need to be monitored. To the best of our knowledge, this is the first report of use of a real-time mLAMP assay for detection of STEC targeting shiga toxin genes that was applied to cattle farm samples.

To evaluate the efficacy of the mLAMP assay, all STEC strains were examined for the presence of stx1 and/or stx2. We noted that the shiga toxin types of each isolate were identical when tested by mPCR and mLAMP, indicating high accuracy of the mLAMP assay. Moreover, the mLAMP assay detected shiga toxin genes in 11 non-O157 STEC strains, indicating that the mLAMP assay can be used to detect various serotypes of STEC. In addition, the mLAMP process was completed within 30 min (mean Tp value = 10:03), while a PCR assay generally takes 3 h, including the post-amplification processes. Overall, these findings indicate that the mLAMP assay could be useful for rapid and accurate characterization of various STEC serotypes.

The sensitivity of the mLAMP assay was evaluated in enrichment broth cultures that were found to be positive for STEC by conventional culture, IMS, and PCR-based culture methods. As shown in Table 4, the mLAMP assay detected shiga toxin genes in all 12 broth cultures that were positive for STEC O157 by the conventional culture method, which is regarded as the gold standard [5,6]. In addition, the mLAMP assay was able to detect shiga toxin genes in nine (90.0%) enrichment broth cultures that were positive for non-O157 STEC strains. These results indicate that the detection rate of the mLAMP assay is comparable to the culture-based detection method for STEC strains. Considering that the culture method is labor intensive and takes at least three days to complete, the application of the mLAMP assay to screening enrichment broth cultures would provide a detection rate similar to culture-based methods, but with less time and labor.

The STEC O157 strains were isolated by conventional culture and IMS methods. While 12 strains were isolated by the conventional method, five more strains were isolated by the IMS method. Considering IMS is known to be the most sensitive culture method [4], the contamination of these five samples with STEC O157 seems to be at a low level that may not be detected by conventional culture. The mLAMP assay detected shiga toxin genes in enrichment cultures of three of these five samples, demonstrating high sensitivity. Indeed, the assay could detect even levels of contamination below those that could be detected using the conventional culture method.

Our mLAMP assay may not detect shiga toxin genes in the presence of low amounts of STEC in enrichment broth cultures (under our detection limit, 10 fg/µL), but still showed greater sensitivity than the conventional culture method. In addition, the mLAMP assay was much more sensitive than the mPCR assay in enrichment broth culture, which was expected considering the high sensitivity of the LAMP assay over the PCR assay in pure culture. Conversely, it is possible that the mLAMP assay may be less sensitive than the mPCR assay to inhibitors in the sample matrix. Other studies have compared methods such as PCR, qPCR, or LAMP for detection of target genes from food or clinical samples and shown that the LAMP assay is less sensitive to inhibitors than the PCR or qPCR assay [19,23,25].

In conclusion, the mLAMP assay simultaneously amplifies the stx1 and stx2 genes, enabling detection of STEC, and enables identification of shiga toxin types more rapidly and accurately than current methods. Moreover, the high detection rate of specific genes from enrichment broth cultures indicates the potential utility of this assay as a primary screening tool for detecting STEC in cattle farm samples.