https://orcid.org/0000-0001-6476-5932
Figures
Abstract
Avian pathogenic Escherichia coli (APEC) can cause colibacillosis in poultry, characterised by localised or systemic infections. Colibacillosis is considered one of the leading causes of economic losses in the poultry industry due to reduced performance, increased mortality, treatment costs and carcass condemnations. A live attenuated Escherichia coli O78 aroA gene mutant is widely used to prevent disease. However, no effective strategies to differentiate the vaccine strain from field strains are available, hampering follow-up of vaccination campaigns. In the current study, we report a PCR-based method to simultaneously detect the vaccine strain by targeting the vaccine-specific mutation in the aroA gene, as well as the wild type E. coli strains by targeting the xanQ gene. The specificity of this PCR was evaluated using 123 E. coli isolates, form which 5 WT aroA auxotrophic strains (WT strains with a natural aroA deficiency), as well as 7 non-Escherichia isolates. The PCR showed 100% sensitivity of the xanQ primers for E. coli detection and 100% sensitivity of the ΔaroA primers for the vaccine strain. In order to allow quantification of the vaccine strain in complex samples containing many different E. coli strains and other related organisms, such as chicken faeces, a probe-based duplex qPCR was developed. The limit of detection (LOD) of this duplex qPCR method was 8.4*103 copies/g faeces. The specificity of the duplex qPCR was confirmed by determining both the vaccine strain levels, and the total E. coli load in intestinal digesta from both vaccinated and non-vaccinated birds. E. coli could be detected in both vaccinated and non-vaccinated birds. The duplex qPCR was specific for the vaccine strain as this strain was detected in all vaccinated birds, whereas no signal was detected in non-vaccinated birds. The duplex qPCR is helpful in monitoring colonization and shedding of the vaccine strain.
Citation: Leurs K, Goossens E, Christensen H, Mainil JG, Vancraeynest D, Ducatelle R, et al. (2022) Development of a duplex qPCR for the differentiation of a live attenuated Escherichia coli aroA mutant vaccine strain from field isolates in chickens. PLoS ONE 17(12): e0278949. https://doi.org/10.1371/journal.pone.0278949
Editor: Peter Gyarmati, University of Illinois College of Medicine, UNITED STATES
Received: June 23, 2022; Accepted: November 24, 2022; Published: December 19, 2022
Copyright: © 2022 Leurs et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: This work was supported by a research grant from Zoetis. This funder provided support in the form of salaries for authors K.L. and D.V., but did not have any additional role in study design, data collection and analysis, or preparation of the manuscript. The funder approved the publication of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.
Competing interests: This work was supported by a research grant from Zoetis. This funder provided support in the form of salaries for authors K.L. and D.V. There are no patents, products in development or marketed products associated with this research to declare. This does not alter our adherence to PLOS ONE policies on sharing data and materials.
Introduction
Escherichia (E.) coli is commonly present in the gastrointestinal tract of vertebrates. While most E. coli strains are benign, some are virulent and capable of inducing disease. Avian pathogenic E. coli (APEC) strains are the causative agents of colibacillosis, an economically important disease in domestic poultry [1]. The most common form of the disease begins either as a respiratory tract infection (causing airsacculitis and pneumonia) or salpingitis combined with peritonitis. When the bacteria reach the bloodstream and spread systemically, this can lead to colisepticaemia, characterized by fibrinous lesions in different organs (pericarditis, peritonitis, perihepatitis, salpingitis, cellulitis and splenitis) [2–6]. In the past, control of severe colibacillosis outbreaks has relied heavily on antibiotic treatments. This approach, however, is now under pressure because of major issues of antimicrobial resistance in E. coli [7]. Therefore, vaccination has become a common strategy to control APEC infection in commercial poultry. One such vaccine commonly used in the field is Poulvac® E. coli (Zoetis, Parsippany, New York, USA), a live attenuated strain of E. coli derived from a virulent O78:K80 APEC strain by mutating the aroA gene through allelic exchange. This defined aroA deletion mutant is deficient in the biosynthesis of aromatic acids and therefore cannot survive in the environment [8–10].
Differentiation of vaccine strains from field strains is recommended by various international organizations involved in the control of animal diseases such as the World Organization of Animal Health (WOAH) and the Food and Agriculture Organization of the United Nations (FAO). It is even a legal obligation for live Salmonella vaccines intended for vaccination of laying hens in the EU (regulation 1177/2006/EC). This is not yet the case for E. coli, but considering the ubiquitous presence of E. coli in the intestinal tract of animals, a similar strategy is highly recommendable. Identifying a specific bacterial strain can be problematic, mainly when derived from complex samples such as intestinal content. For example, the classical selective and indicative culture media usually allow selective growth and identification up to the family or genus level. Detection and identification of a single bacterial strain in a complex mixture typically rely on a PCR-based technology. To date, the differentiation of the commercial aroA mutant vaccine strain from other E. coli strains is achieved either via (i) the absence of growth on minimal agar whilst presenting growth on MacConkey agar [5, 11]; (ii) discrimination based on pulsed-field gel electrophoresis (PFGE) band patterns [11]; or (iii) a PCR developed by La Ragione et al. (2013) [10], in which the amplicon length between vaccine or field strain DNA shows only a marginal difference in amplicon length, which complicates the interpretation of the result.
Therefore, the current study aimed to develop a more robust PCR approach to discriminate between the aroA mutant and field strains. In addition, the primers designed in this study were used to establish a duplex quantitative PCR (qPCR) to simultaneously quantify aroA mutant and total E. coli levels in the intestinal tract of poultry, a tool which can be useful in the follow-up of vaccination campaigns.
Material and methods
Bacterial strains and culture conditions and genomic DNA extraction
A live attenuated aroA mutant vaccine strain, derived from a virulent O78:K80 APEC strain, hereafter referred to as ΔaroA, was used throughout this study (Poulvac® E. coli, Zoetis, Charles City Manufacturing site, USA). In addition, 132 E. coli strains including 5 aromatic amino acid auxotrophic strains (E. coli strains isolated from chickens that harbor a WT aroA gene, but were found to be unable to synthesize aromatic amino acids and therefore are phenotypically indistinguishable from the ΔaroA vaccine strain using standard plating methods [11]), as well as seven non-Escherichia strains were used to optimise and validate the PCR (S1 Table). All strains were routinely grown according to their growth preferences (aerobic or anaerobic) at 37°C. E. coli strains were grown on MacConkey agar, Clostridium perfringens on Columbia blood agar, Salmonella strains on Brilliant Green agar, Bacillus on Luria-Bertani agar and the Lactobacillus strain on De Man, Rogosa and Sharpe (MRS) agar. Genomic DNA was isolated from all strains using the alkaline lysis method by resuspending a single bacterial colony (or the pellet from 100μl culture) in 20 μl lysis buffer (0.25% SDS, 50 mM NaOH). After boiling at 95°C for 5 min, 180 μl of nuclease-free water was added and centrifuged for 5 min at maximum speed. The DNA was stored at -20°C until further use.
The analytical sensitivity of the developed duplex qPCR was assessed by comparing the ΔaroA count with the qPCR results. Therefore, the ΔaroA strain was grown overnight in Luria-Bertani Broth (LB) at 37°C with gentle shaking. A 10-fold dilution was made from this overnight culture and titrated on LB agar plates and colony-forming units (CFU)/ ml were counted the day after. From this 10-fold dilution, 100 μl culture was pelleted and the pellet was used to isolate genomic DNA using alkaline lysis as described above.
Target selection, primer and probe design
To develop a PCR that can discriminate between wild type (WT) E. coli and ΔaroA, both the wild type aroA gene (GenBank accession number CP004009.1) and the aroA mutant gene as created by La Ragione et al. (2013) were targeted [10]. To develop a duplex PCR to simultaneously determine the amount of ΔaroA relative to the total amount of E. coli (both WT and ΔaroA), the xanQ gene was targeted in addition to the aroA gene. The usability of this xanQ gene for the reliable detection of E. coli was previously described [12].
Primers and primer/probe sets were designed using the PrimerQuestTM software (IDT, Integrated DNA Technologies Inc., Coralville, IA, USA). The specificity of the designed primers and probes towards E. coli was confirmed using Primer Blast analysis [13]. 6-carboxyfluorescein (FAM) and hexachlorofluorescein (HEX) conjugated to the 5’ ends of the probes were used as fluorescent reporter dyes to detect amplification products specific for aroA (FAM) or xanQ (HEX). Because of its high fluorescent signal, FAM is known to be a suitable dye to detect low-copy transcripts. As E. coli is a commensal, the total amount of E. coli (both WT and ΔaroA) in intestinal content samples will exceed the amount of ΔaroA. Therefore, it was opted to couple a FAM signal to the probe for the aroA mutation (minority part) and a HEX signal to detect the xanQ gene (majority part). To obtain optimal signal detection, with minimal background fluorescence, all probes were designed as double-quenched probes, using a 3’ Iowa Black®FQ and ZEN quencher. The ZEN quencher has been reported to increase the signal sensitivity by decreasing the background fluorescence. The Iowa Black® FQ quencher was selected because it has a broad absorbance spectrum ranging from 420 to 620 nm with a peak absorbance at 531 nm and therefore overlaps with the emission spectra of FAM (excitation wavelength of 495 nm and emission wavelength of 520 nm) and HEX (excitation wavelength of 538 nm and emission wavelength of 555 nm). All the different primers and probes developed in this study were synthesized by IDT (Coralville, IA, USA) and are listed in Tables 1 and 2. Fig 1 is a graphical representation of these primers and probes on the genes.
(A) The aroA mutant gene (1279 bp) is schematically represented by the blue line on which the mutation is indicated in red. Underneath this line, the amplicons generated by the different primer pairs (arrows) are represented. Above the schematic gene representation, the location of the probes is indicated with a star. Since this figure depicts the aroA mutant gene, the primer pair aroA_WT cannot be represented. (B) A similar representation for the xanQ gene (1401 bp) in blue with probes (above) and primers (below).
PCR assay to determine the primer specificity
The specificity of the different PCR primer pairs was tested using a broad range of E. coli strains (mainly, but not exclusively, APEC), including 5 aromatic amino acid auxotrophic strains (E. coli strains harboring the WT aroA gene, but that are unable to synthesize aromatic amino acids), as well as other non-Escherichia strains (S1 Table). PCR was performed in a 20 μl total reaction mixture using 2 μl template DNA, 0.5 μM forward and 1 μM reverse primer in case of aroA primers; 0.5 μM in case of any other reverse primer and 2x biomix (Bioline, London, UK). Thermocycling was carried out in an Eppendorf Mastercycler Pro (Eppendorf, Hamburg, Germany) with the following parameters: 1 cycle at 94°C for 10 min followed by 30 cycles at 94°C for 45 s, 60°C for 30 s and 72°C for 45 s followed by 1 cycle at 72°C for 5 min. Agarose gel electrophoresis of the PCR products was performed on a 1.5% (w/v) agarose gel stained with MidoriGreen (Nippon Genetics, Tokyo, Japan) and visualized with UV-light. A 100-bp DNA ladder (GeneRulerTM, Thermo Scientific, Waltham, MA, USA) was used as a molecular weight marker.
Determination of the efficiency and specificity of selected primers in a simplex qPCR reaction
To determine the efficiency of the developed primer pairs, a standard curve was generated for the mutated aroA gene and the xanQ gene using respectively the standard fragment primer pairs aroA_SF or xanQ_SF (Table 1) and DNA from the ΔaroA strain. The resulting PCR products were purified (MSB Spin PCRapace, Stratec Molecular, Berlin, Germany) and DNA concentration was determined spectrophotometrically (Nanodrop Technologies, Wilmington, DE, USA). The concentration of the linear dsDNA standard fragments was adjusted to 108 - 100 copies/μl, with tenfold dilution steps.
The specificity and efficiency of each of the primer pairs were assessed using a SYBRgreen qPCR assay. Each 12 μl qPCR reaction consisted of 2μl template DNA, 6 μl SensiMix™ SYBR® & Fluorescein Kit (Bioline), 0.5 μM forward primer and 0.5 μM reverse xanQ primer or 1 μM reverse aroA primer. Cycling was performed on a real-time PCR thermal cycler (Biorad, Hercules, CA, USA) and conditions were as follows: 95°C for 10 min, followed by 40 cycles of 95°C for 45 s and 62°C for 1 min. The fluorescent products were detected at the last step of each cycle. To confirm the specificity of the reaction, PCR products were subjected to melt analysis using a dissociation protocol comprising 95°C for 15 s, followed by 0.5°C incremental temperature ramping from 65°C to 90°C.
To develop a duplex qPCR reaction targeting both the total amount of E. coli as well as the amount of ΔaroA in a sample, the xanQ gene (total E. coli) and the aroA mutant gene (vaccine strain) need to be quantified using probes in different fluorescence channels. Therefore, primer pairs yielding good efficiency (90%-110% efficiency) in the SYBRgreen qPCR assay set-up were selected for further development in a probe-based qPCR assay. Each 12 μl of reaction mix consisted of 2 μl template DNA, 6μl 2X IQ Supermix (Biorad); 0.5 μM forward and 0.5 μM reverse xanQ primer or 1 μM reverse aroA primer and 0.2 μM probe. Cycling was carried out on a real-time PCR thermal cycler (Biorad) and conditions were: 95°C for 10 min, followed by 40 cycles of 95°C for 45 s and 62°C for 1 min with fluorescence detection of the products at the last step of each cycle.
For both the SYBRgreen qPCR assay and the probe-based qPCR assay, the efficiency of the assay was calculated by plotting the CT values of the standard fragment dilution series against the log standard fragment amount and determining the slope of the resulting standard curve. From the slope, efficiency was calculated using the following formula:
The limit of detection (LOD) of the assay was reported as the lowest template copy number yielding the correct melting temperature, whereas the limit of quantification (LOQ) was defined by the last valid Cq of the standard curve.
Development of a probe-based duplex qPCR for detection of ΔaroA and WT E. coli in a single reaction
Two different primer-probe combinations targeting both the aroA mutant gene and the xanQ gene (total E. coli) were selected for further duplex probe-based qPCR development (Duplex qPCR A and B). The efficiency, LOQ and LOD of the qPCR reactions were determined in technical triplicates of a tenfold dilution series of both the aroA standard fragment and the xanQ standard fragment (108–100 copies/μl). Each 12 μl qPCR reaction contained: 2μl template DNA, 6 μl 2X IQ Supermix (Biorad); 0.5 μM forward, 0.5 μM reverse xanQ primer or 1 μM reverse aroA primer and 0.2 μM of each probe. Cycling was done on a real-time PCR thermal cycler (Biorad) with the following conditions: 95°C for 10 min, followed by 40 cycles of 95°C for 45 s and 62°C for 1 min, with fluorescence detection of the products at the last step of each cycle.
Determination of possible effects of a faecal matrix on the final duplex probe-based qPCR
Since the duplex qPCR aims to discriminate ΔaroA from WT E. coli strains in faeces or intestinal contents from vaccinated chickens, the final selected primer-probe mixture (Duplex qPCR B) was further checked for any possible matrix effects. To assess the efficiency of the duplex qPCR assay, tenfold dilutions of the ΔaroA cultures were spiked into 100 mg of fresh chicken faeces, resulting in 109 CFU/g-100 CFU/g faeces, after which DNA was extracted with the CTAB method as described previously [14, 15]. The quality and the concentration of the DNA were examined spectrophotometrically (NanoDrop, Thermo Scientific, Merelbeke, Belgium), diluted to 50 ng/μl and stored at -20°C until further analysis.
To further assess the usability of the probe-based duplex qPCR to detect the ΔaroA strain, samples from both vaccinated and non-vaccinated birds were tested. Therefore, a total of 216 male Ross308 broilers were obtained from a local hatchery on day of hatch (Day1, 1-24h after hatching) and randomly allocated to 18 concrete floor pens (housed on wood shavings, 1m2, 12 birds per pen) separated by solid walls. On arrival, birds from 9 pens (vaccinated group) were vaccinated using the live attenuated ΔaroA vaccine strain (Poulvac® E. coli, dose: 108 CFU/bird) in the drinking water. The nine other pens did not receive the vaccine (control birds). A commercial standard wheat/soy-based starter feed (Farm Mash 1, Versele Laga, Deinze, Belgium) and drinking water were provided ad libitum. At three days post-hatch, one bird per pen was euthanized to collect ileal content samples. All samples were stored at -20°C until further analysis. DNA was extracted and diluted to 50 ng/μl before performing the duplex qPCR as described above for the spiked faeces. Statistical analysis of the difference in total E. coli load between vaccinated and non-vaccinated birds was assessed using an unpaired t-test on the log-transformed data. The study followed the guidelines of the ethics committee of the Faculty of Veterinary Medicine, Ghent University, in accordance with the EU Directive 2010/63/EU.
Results and discussion
Development of a PCR technique to discriminate WT from ΔaroA E. coli
At the start of this study, there was one known primer pair by La Ragione et al. (2013) that could differentiate between WT and the ΔaroA E. coli. This primer pair (aroA_LR, Table 1) targets the aroA gene, spanning the mutation, thereby resulting in a PCR amplicon of 1236 bp for the WT E. coli or 1161 bp for the ΔaroA E. coli [10]. Although this PCR has been used to discriminate WT from ΔaroA before [11], the small difference in amplicon length makes it hard to distinguish the vaccine strain from commensal E. coli strains (Fig 2D). Therefore, a novel PCR assay was developed to unambiguously differentiate the ΔaroA vaccine strain from commensal E. coli strains. The known mutation in the aroA gene was used to design the new primer pairs in the current study. Multiple primer sets for detecting ΔaroA and one pair to detect WT E. coli were generated (Table 1). The specificity of both the ΔaroA or WT PCR assays was assessed by screening different E. coli strains (mainly, but not exclusively, APEC) as well as other non-Escherichia strains (S1 Table). The aroA_WT primer set successfully amplified all WT E. coli strains, whereas no amplicon was observed for the ΔaroA vaccine strain, the non-Escherichia strains nor the water control (Fig 2B). All the aroA mutant primer sets in turn exclusively rendered an amplicon for ΔaroA and not for any other strain, also not the aromatic amino acid auxotrophic strains (particular E. coli strains that harbor a WT aroA gene, but are unable to synthesize aromatic amino acids), nor the water control (Fig 2A). Aromatic amino acid auxotrophic strains have previously been isolated from chicken intestinal content, and were shown to be phenotypically indistinguishable from the ΔaroA vaccine strain using standard plating methods [11]. Therefore, genetic methods, such as the aroA mutant primer sets developed in the current study, are needed to discriminate the ΔaroA vaccine strain from all commensal strains. Moreover, by performing a duplex PCR targeting both the E. coli-specific xanQ gene and the aroA-mutation, E. coli identification and detection of the vaccine strain can be performed in a single reaction (Fig 2C) [5, 12]. Moreover, inclusion of the xanQ gene in the duplex PCR serves as a control, as absence of a xanQ amplicon might indicate that the isolate is not an E. coli strain, or the DNA extraction or PCR reaction was not successful. As such, the PCR primers developed in our study give an easy, rapid and relatively cheap solution to discriminate the ΔaroA vaccine strain from other E. coli strains. Since the primers indicate whether the mutation is either present or not, the PCR results are unambiguously interpretable (Fig 2).
The specificity of the primers targeting either the E. coli ΔaroA gene (A), the WT E. coli aroA gene (B), the previously published primers targeting both the WT and ΔaroA gene (total aroA) (D) or simultaneous detection of the ΔaroA and xanQ gene (C), was screened against a total of 132 E. coli strains and 7 non-Escherichia strains. (A) PCR specific for the ΔaroA gene. Amplification only occurred for the vaccine strain, while none of the commensal E. coli strains or non-Escherichia strains was amplified. The picture shows the results from primer pair aroA_3, which results in a band at 281bp when the aroA mutation is present. All other tested ΔaroA specific primers showed the same specificity. (B) The aroA primers targeting the WT gene (aroA_WT) result in a band at 660 bp in commensal E. coli, whereas no band occurred when the aroA mutation was present or in the non-Escherichia strains. (C) Duplex PCR with aroA_3 and xanQ_2 primers. The E. coli-specific xanQ primers result in a band at 97 bp, whereas the band at 281 bp indicates ΔaroA. Only for the ΔaroA E. coli strain, the two bands are present. (D) Primer pair aroA_LR (La Ragione et al., 2013, Table 1) is not specific for the ΔaroA mutation and results in an amplicon of 1236 bp for wild type E. coli, or 1161 bp for the ΔaroA vaccine strain. The pictures show a representative selection of the total number of strains used to assess the primer specificity. Samples: 1: blank control 2: E. coli O78:K80 ΔaroA (vaccine strain), 3: E coli O78:K80 WT, 4: E. coli Nissle 1917, 5: APEC_1 (chicken intestine), 6: APEC_2 (turkey intestine), 7: APEC019 (chicken intestine), 8: APEC023 (chicken intestine), 9: 220115–1 (chicken intestine, aromatic amino acid auxotrophic E. coli), 10: 220115–10 (chicken intestine, aromatic amino acid auxotrophic E. coli), 11: 06114–21 (chicken intestine, aromatic amino acid auxotrophic E. coli), 12: 7-35s (chicken intestine, aromatic amino acid auxotrophic E. coli), 13: 3-44s (chicken intestine, aromatic amino acid auxotrophic E. coli), 14: Salmonella Enteritidis, 15: Salmonella Typhimurium, 16: Clostridium perfringens NE18, 17: Clostridium perfringens CP20, 18: Bacillus subtilis DSM 29784. The complete list of tested strains can be found in S1 Table. Electrophoresis was performed on 1.5% agarose gel. GeneRulerTM 100 bp Plus DNA Ladder was used as a molecular weight marker.
Development of a probe-based duplex qPCR for quantitative detection of ΔaroA and WT E. coli in a single reaction
Primer design and validation in single PCR. To simultaneously detect the ΔaroA strain, as well as the total amount of E. coli in a sample, the xanQ gene was used as a target. This xanQ gene (also called ygfO) belongs to an evolutionarily conserved transporter family in E. coli [12], making it a functional target gene for an E. coli specific PCR. In total four different xanQ primer pairs were designed and tested for their specificity towards E. coli (xanQ_1 –xanQ_4; Table 1). All xanQ primer pairs were proven to be E. coli specific since no amplification occurred in non-E. coli DNA (Fig 2C).
Determination of the efficiency of selected primers in a simplex qPCR reaction. Different qPCRs were developed to quantify the amount of ΔaroA as well as the total amount of E. coli present in an individual sample. The four primer pairs for the aroA and the four pairs for the xanQ gene were first used in a SYBRGreen based qPCR to determine the efficiency of the qPCR reaction (Table 3). Therefore technical triplicates of a 10-fold dilution series of the standard fragments (108−100 copies/μl) generated with either aroA_SF or xanQ_SF were used (Table 1). The assays showing an efficiency between 90% and 110% in the SYBRgreen based qPCR were selected for further testing in a probe-based simplex set-up. This resulted in four aroA primer pairs and only two primer pairs for the xanQ gene. In none of the tested qPCRs amplification occurred in the negative controls (both water as well as non-specific DNA samples). The Cq values of the simplex probe-based qPCRs were comparable to those of the SYBRgreen assays (Table 3), indicating that the probe-based qPCR assay has the same working range as the SYBRgreen detection. Both aroA and xanQ probe-based qPCR assays had a comparable range of quantification, which was 102−108 copies/μl for the ΔaroA specific qPCR assays. The LOQ of the xanQ qPCR assay was slightly higher, with a quantification range of 102-108 copies/μl. For the xanQ_2 primers a higher LOQ (103 copies/μl) was observed. The LOD was 104 copies/μl. Since xanQ is E. coli specific and E. coli is a commensal, a higher LOQ has no impact on the usability of the qPCR set-up. The presence of a small amount (<104 copies/ml) of E. coli in a certain sample will not be detected, but in the scope of this qPCR (detection of E. coli in faeces wherein E. coli–as a commensal- is always present in large amounts) it is negligible.
For each primer pair (Table 1) the efficiency, limit of quantification (LOQ) and limit of detection (LOD) were assessed using a 10-fold dilution series of a standard fragment (108−100 copies/μl).
Only those primer pairs with an efficiency falling in the accepted range (90–110%) were withheld to perform a duplex probe-based qPCR, resulting in two aroA (aroA_1 and aroA_3) and two xanQ (xanQ_2 and xanQ_4) primer/probe combinations (Table 4).
For each of the two primer/probe combinations the efficiency, limit of quantification and limit of detection were assessed using a 10-fold dilution series of a standard fragment (108−100 copies/μl). The choice for these specific combinations of primer pairs for both genes was based upon limited primer/probe interactions.
Development of a probe-based duplex qPCR for simultaneous detection of ΔaroA and WT E. coli
A duplex qPCR was developed which allows the simultaneous quantification of both ΔaroA and commensal E. coli in the same sample, limiting possible technical errors or effects. At first, the two remaining aroA primer pairs (aroA_1/3) and the two remaining xanQ primer pairs (xanQ_2/4) were tested for possible primer/probe interactions by adding mismatches (e.g. aroA_1_fw with xanQ_2_rev) of primer/probes to a PCR reaction and see whether or not amplification occurred. Based upon these results two duplex primer/probe combinations were withheld, namely the combination of aroA_1 with xanQ_4 primers and aroA_3 and xanQ_2. The choice for those specific combinations of primer pairs for both genes was based upon the following criteria: (i) no amplification of the negative control; (ii) no non-specific interaction between the primers and probes of the two genes of interest. By performing qPCRs with altered primer/probe combinations, interactions between selected primers and probes for both genes were excluded. Cq values between simplex and duplex set-up differed less than 1 Cq, making it a successful duplex (Tables 3 and 4). The final choice to continue with the aroA_3 and xanQ_2 primer/probe combination was made based on the efficiency of the qPCR reaction. Because the aroA forward primer binds on the mutation whilst the aroA reverse primer binds on both the WT and ΔaroA E. coli, an increased need for reverse primer exists, which is solved by adding a double amount of reverse primer (1 μM as compared to 0.5 μM).
Specificity and analytical sensitivity of the duplex qPCR. The analytical sensitivity was assessed by testing the LOD of the assay. Therefore, either spiked chicken faeces or 10-fold serial dilutions of genomic DNA from the ΔaroA strain was used. To quantify the copy numbers, a standard curve was generated from a 10-fold dilution series (108−100 copies/μl) of a standard fragment. The lowest template copy numbers yielding specific melting temperature in the SYBRGreen assays were considered as the detection limit of the assay (Table 3). To accurately detect the ΔaroA strain in a faecal sample, a minimum of 104 CFU/g faeces is required for the duplex qPCR assay (Table 4 and Fig 3). In Fig 3, the amount of detected ΔaroA copies/ml versus the original input is depicted for both spiked faeces as well as for genomic DNA from the ΔaroA strain and for all three types of qPCR (SYBRgreen based, simplex and duplex probe-based). From this, it can be concluded that there is not a one on one relation between input and detection but that the detection still falls within reasonable deviation and this for all qPCR methods and both input methods.
The amount of detected ΔaroA (copies/ml) for the three different qPCRs (the SYBRgreen simplex qPCR (black triangles), probe-based simplex qPCR (dark grey squares) and probe-based duplex qPCR (light grey circles) is depicted. (A) The input of the qPCR is genomic DNA originating from a 10-fold dilution series of an overnight culture from the ΔaroA strain. (B) The input for the qPCRs is DNA isolated from fresh chicken faeces spiked with a 10-fold dilution series of the ΔaroA strain.
To further confirm the usability of the developed duplex qPCR assay, the ΔaroA strain was quantified in ileal content samples from vaccinated birds. For this, birds were vaccinated with ΔaroA on day-of-hatch. On day 3 post-hatch, ileal content samples were collected from both vaccinated and non-vaccinated birds. Vaccination tended to reduce the total E. coli load in the ileum (p = 0.053). Moreover, no detection of vaccine strain occurred in non-vaccinated chickens, whilst the vaccine strain was detected in all birds in the vaccinated group (Fig 4), showing the specificity and the usability of the developed assay.
Birds from the vaccinated group received 108 CFU ΔaroA vaccine per bird at day of hatch. At 3 days of age (2 days after vaccination), ileal content was collected. (A) Simultaneous detection of the vaccine strain (targeting the mutant aroA gene) (dark grey bars) and total E. coli load (targeting the xanQ gene) (light grey bars) was obtained using the duplex qPCR. Total E. coli load in the ileum of vaccinated birds tended to be lower than in non-vaccinated birds (p = 0.053). (B) 92.4% of the total E. coli load in the vaccinated birds was attributed to the ΔaroA vaccine strain, whereas in the non-vaccinated birds, no vaccine strain was detected.
The sensitivity of the current duplex probe-based qPCR is adequate to prove the presence of ΔaroA in a particular sample in a quantifiable manner. The qPCR has a detection limit of 8.4*103 copies/g faeces, which proved to be sufficient to detect the vaccine strain in intestinal content from vaccinated birds. When detection of lower amounts of vaccine shedding would be needed, it should be possible to perform an E. coli enrichment step prior to the PCR, after which solely the presence of the mutant strain can be proven but this no longer holds any quantification possibility.
Conclusion
The PCR tools presented in our study are crucial in epidemiological investigations as well as for research purposes. They could also prove of value for monitoring the hygienic quality of meat or the presence of the vaccine strain on meat when optimized for this particular matrix. To develop a duplex qPCR for the simultaneous quantification of the ΔaroA vaccine and total E. coli load in a specific sample, this study made use of: (i) the genetic aroA mutation inserted by allelic exchange [10] and (ii) E. coli specific primers targeting the xanQ gene [12]. The qPCR has a detection limit for ΔaroA of 8.4 *103 copies/g faeces as was determined by spiking fresh chicken faeces with a 10-fold dilution series of ΔaroA, and it is specific for this vaccine strain as it did not pick up any signal in control intestinal samples from non-vaccinated chickens. The set-up was optimized for DNA samples from chicken intestinal content.
This fast, easy-to-use and relatively cheap tool for discriminating the Poulvac® E. coli vaccine strain from other E. coli strains provides specific identification and quantification of the aroA mutant, making it ideal for screening bacterial populations retrieved from chicken faeces or intestinal content. The duplex qPCR is helpful in monitoring colonization and shedding of the vaccine strain.
Supporting information
S1 Table. List of all used bacterial strains in this study.
https://doi.org/10.1371/journal.pone.0278949.s001
(XLSX)
S1 File. Raw gel images used to construct Fig 2.
https://doi.org/10.1371/journal.pone.0278949.s002
(DOCX)
Acknowledgments
The extended list of APEC strains used to validate the tool was a kind gift from Prof. J. Mainil, University of Liège (ULiège), Faculty of Veterinary Medicine, Department of Infectious Diseases, Bacteriology (Schouler et al., 2012). The authors would like to thank Mr Jean-Noël Duprez for his excellent technical assistance in culturing this extended list of APEC strains.
References
- 1. Lutful Kabir SM. Avian Colibacillosis and Salmonellosis: A Closer Look at Epidemiology, Pathogenesis, Diagnosis, Control and Public Health Concerns. Int J Environ Res Public Health. 2010;7: 89. pmid:20195435
- 2. Antão EM, Glodde S, Li G, Sharifi R, Homeier T, Laturnus C, et al. The chicken as a natural model for extraintestinal infections caused by avian pathogenic Escherichia coli (APEC). Microb Pathog. 2008;45: 361–369. pmid:18848980
- 3. Ewers C, Janßen T, Kießling S, Philipp HC, Wieler LH. Molecular epidemiology of avian pathogenic Escherichia coli (APEC) isolated from colisepticemia in poultry. Vet Microbiol. 2004;104: 91–101. pmid:15530743
- 4. Ozaki H, Murase T. Multiple routes of entry for Escherichia coli causing colibacillosis in commercial layer chickens. J Vet Med Sci. 2009;71: 1685–1689. pmid:20046043
- 5. Salehi TZ, Tabatabaei S, Karimi V, Fasaei BN, Derakhshandeh A, Jahromi OAN. Assessment of immunity against avian colibacillosis induced by an aroA mutant containing increased serum survival gene in broilers. Braz J Microbiol. 2012;43: 363. pmid:24031840
- 6. Kariyawasam S, Wilkie BN, Gyles CL. Construction, Characterization, and Evaluation of the Vaccine Potential of Three Genetically Defined Mutants of Avian Pathogenic Escherichia coli. 2004;48: 287–299. pmid:15283416
- 7. Poirel L, Madec J-Y, Lupo A, Schink A-K, Kieffer N, Nordmann P, et al. Antimicrobial Resistance in Escherichia coli. Microbiol Spectr. 2018;6. pmid:30003866
- 8. Mombarg M, Bouzoubaa K, Andrews S, Vanimisetti HB, Rodenberg J, Karaca K. Safety and efficacy of an aroA-deleted live vaccine against avian colibacillosis in a multicentre field trial in broilers in Morocco. Avian Pathol. 2014;43: 276–281. pmid:24824589
- 9. Filho TF, Fávaro C, Ingberman M, Beirão BCB, Inoue A, Gomes L, et al. Effect of spray Escherichia coli vaccine on the immunity of poultry. Avian Dis. 2013;57: 671–676. pmid:24283136
- 10. la Ragione RM, Woodward MJ, Kumar M, Rodenberg J, Fan H, Wales AD, et al. Efficacy of a Live Attenuated Escherichia coli O78∶K80 Vaccine in Chickens and Turkeys. 2013;57: 273–279. pmid:24689185
- 11. Christensen H, Nielsen C. Safety of the live Escherichia coli vaccine Poulvac® E. coli in layer parent stock in a field trial. Vet Microbiol. 2020;240: 108537. pmid:31902490
- 12. Deshmukh RA, Bhand S, Roy U. A novel molecular quantitative method for rapid and sensitive detection of: Escherichia coli from roof-harvested rainwater. Analytical Methods. 2019;11: 3155–3167.
- 13. Ye J, Coulouris G, Zaretskaya I, Cutcutache I, Rozen S, Madden TL. Primer-BLAST: a tool to design target-specific primers for polymerase chain reaction. BMC Bioinformatics. 2012;13: 134. pmid:22708584
- 14. Griffiths RI, Whiteley AS, O’Donnell AG, Bailey MJ. Rapid method for coextraction of DNA and RNA from natural environments for analysis of ribosomal DNA- and rRNA-based microbial community composition. Appl Environ Microbiol. 2000;66: 5488–5491. pmid:11097934
- 15. Aguirre M, Vuorenmaa J, Valkonen E, Kettunen H, Callens C, Haesebrouck F. In-feed resin acids reduce matrix metalloproteinase activity in the ileal mucosa of healthy broilers without inducing major effects on the gut microbiota. Vet Res. 2019; 1–14. pmid:30795808