E. coli and Campylobacter spp. strains of poultry origin previously characterized by whole genome sequencing (14, 15) were used for primers selection, qPCR optimization, analytical and diagnostic validation, as well as positive control for the analysis of the field samples. The list of the strains used in this study, including the bacterial species, the resistance profile, the sequence accession numbers, and in which stage of the validation they were used, is reported in supplementary material 1. E. coli strains were resuscitated on nutrient agar (Microbiol, Italy) and incubated at 37 °C ± 5 °C for 24 ± 2 h. Campylobacter spp. strains were streaked on blood agar (Microbiol) and incubated at 41.5 °C ± 5 °C for 48 ± 2 h under a microaerophilic environment (CampyGen, Oxoid, UK). Bacterial genomic DNA (gDNA) was isolated using the Invisorb Spin Tissue Mini Kit (Invitek Molecular, Berlin, Germany) following the manufacturer's instruction. gDNA quantity was assessed using the Qubit dsDNA High Sensitivity kit (Thermo Fisher Scientific, Massachusetts, USA).

Based upon a thorough search of the literature, a total of ten primer pairs (Table 1) used for the detection of genes tetA, tetB, and tetO by end-point PCR were identified (5, 16, 17). Each primer pair was first checked in silico using primer BLAST (https://www.ncbi.nlm.nih.gov/tools/primer-blast/) and then tested for specificity against strains harboring tetA (n = 5), tetB (n = 5), and tetO (n = 6). To assess the risk of false positive results, each assay was also tested against the gDNA isolated from the bacteria not harboring their target gene. End-point PCRs were carried out in a total reaction volume of 25 μl, consisting of 12.5 μl DreamTaq PCR Master Mix 2X (Thermo Fisher Scientific), 0.5 μl of each primer at a concentration of 10 pmol/μl, and 5 ng of genomic DNA as the template. The thermal cycling program started with an initial denaturation at 95 °C for 2 min (min), followed by 35 cycles of denaturation at 95 °C for 30 s (s), annealing at 60 °C for 20 s, and extension at 72 °C for 20 s. A final extension step was performed at 72 °C for 7 min using an Applied Biosystems 2720 Thermal Cycler (Thermo Fisher Scientific). The resulting amplicons were analyzed by agarose gel electrophoresis on a 2% gel run at 100 V for 60 min. Selected PCR products were purified using the QIAquickPCR Purification Kit (Qiagen), following the manufacturer's instruction, and purified amplicons were Sanger sequenced by an external laboratory (BMR Genomics, Padua, Italy). Chromatograms were visualized and analyzed using BioEdit (v7.7.1). Edited sequences were screened in BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi) for tet genes confirmation.

After the initial selection, the optimization of qPCR conditions was carried out for primer pairs tetA1, tetB3, tetO2, and tetO3 (Table 1). All amplifications were carried out in a LightCycler^® 480 Roche (Basel, Switzerland) and using the PowerUp^TM SYBR^® Green Master Mix (Thermo Fisher Scientific). Different concentrations of each primer were tested for each primer pair (300/300 pmol/ml, 600/600 pmol/ml, 600/900 pmol/ml, 900/600 pmol/ml and 900/900 pmol/ml for forward and reverse primer respectively), using 2.5 μl of end-point PCR amplicons diluted to a final concentration of 0.002 ng/μl. The optimization steps were performed using the following amplification protocol: initial incubation at 50 °C for 2 min, followed by 2 min at 95 °C, and 45 cycles at 95 °C for 10 s and 50 °C−60 °C for 40 s. A melting curve between 40 °C and 95 °C was determined by adding a dissociation step after the last amplification cycle at a temperature transition rate of 4.4 °C/s. qPCR data analysis was performed using LightCycler^® 480 software version 1.5 (Roche). All qPCR reactions were performed in triplicate.

Standard curves were constructed to assess the efficiency and the dynamic range of each primer set using nine ten-fold dilutions prepared starting from an initial concentration of 0.002 ng/μl of end-point PCR amplicons. The assays' limit of detection (LoD) was considered as the lowest target amount detected in at least 50% of nine independent replicates tested by two operators in two independent runs. The repeatability of the three qPCRs was evaluated by two operators testing each three target dilutions, corresponding to high, medium, and low dilution (one dilution above the LoD), in triplicate during three independent experiments, at weekly intervals. Intra- and inter-run repeatability was assessed by calculating the coefficient of variation (CV). A %CV equal to 5% was adopted as cut-off value to assess the robustness of the assays. The specificity was assessed by testing the assays against the dilution panels not containing their respective target.

The diagnostic validation was carried out using a panel of 69 bacteria of known resistance profile (Supplementary Material 1). The panel was prepared by one operator and blindly tested by two other operators at weekly interval. Each sample was tested in triplicate. Samples with at least two replicates showing cycle threshold (Ct) values within the dynamic range of the assay and the expected melting temperature (Tm) were considered positive. The diagnostic sensitivity (DSe), the diagnostic specificity (DSp), and the overall accuracy (Acc) were calculated for each assay.

The assays were used to assess the prevalence and to quantify their respective tet genes in 36 fecal, 11 cecal and 29 drinking water (DW) samples collected in previous studies (18, 19). Fecal, cecal, and DW samples were pre-treated as previously described (18, 19). Briefly, 25 g of feces were placed into a sterile 50-ml Falcon tube, mixed with 25 ml of phosphate-buffered saline (PBS), vortexed for 1 min, and centrifuged at 4,000 rpm for 10 min at 4 °C. DNA was then extracted from 250 mg of the resulting pellet using the DNeasy PowerLyzer PowerSoil^® kit (Qiagen, Hilden, Germany). Cecal contents were collected using sterile cotton swabs. Each swab was eluted in 1 mL of PBS, processed for 90 s at 25 Hz using a TissueLyser^® (Qiagen), the supernatant was recovered, and centrifuged at 4,000 rpm for 10 min at 4 °C. DNA was extracted from the resulting pellet using the DNeasy PowerSoil Pro^® Kit (Qiagen). Two liters of each water sample were filtered and DNA was extracted from the filter membranes using the PowerWater DNA kit (Qiagen). The list of the fields samples is reported in Supplementary Material 2. Field samples were tested in triplicates and samples with at least two replicates showing Ct values within the dynamic range of the assays and Tm matching the positive control were considered as positive. The 16S rRNA gene was also amplified (20) as internal process control and to normalize the target genes copy numbers. Positive (tet+) and negative controls (tet-) for each assay were included in each qPCR round.

Standard curves, assay efficiency, R² values, DSe, DSp, overall accuracy, and repeatability were calculated using GraphPad Prism v10.5.0. Descriptive statistics for prevalence and gene quantification, as well as graphical representations of the results, were generated using the same software.

The optimal primers concentrations were established as 600/600 pmol/ml for tetA and as 300/300 pmol/ml for tetB and tetO. The annealing temperatures for maximizing assays' performances avoiding nonspecific amplification were defined as 55 °C for tetA and tetB and as 56 °C for tetO. The sequences of the selected primers are reported in Table 1.

The qPCR assays exhibited efficiencies ranging from 80.7% for tetA to 93.6% for tetO, with all standard curves demonstrating excellent linearity (R² > 0.99; Figure 1). The limits of quantification were determined to be 2,174.3, 39.7, and 3.8 gene copies per μl for the tetA, tetB, and tetO assays, respectively, while the corresponding limits of detection were 271.4, 4.0, and 3.8 gene copies per μl. All assays demonstrated high robustness and repeatability, as indicated by %CVs consistently below the adopted threshold (5%) both within and between runs (Table 2). DSe ranged from 92.11% for tetA to 100% for tetB and tetO, whereas DSp was assessed as 91.38%, 98.41%, and 100% for tetB, tetO, and tetA, respectively. The overall diagnostic accuracy was 95.71%, 92.86%, and 98.57% for tetA, tetB, and tetO, respectively.

With the exception of one water sample, all tested samples were positive for at least one of the investigated tet genes, with overall prevalence ranging from 65.79% (95% confidence interval (CI): 54.88%−76.70%) for tetA to 93.33% (95% CI: 87.56%−99.11%) for tetO (Figure 2A). The prevalence of tetA varied across sample types, from 37.93% (95% CI: 19.15%−56.71%) in drinking water to 90.91% (95% CI: 70.65%−100%) in cecal content (Figure 2B). For tetB, the lowest prevalence was observed in water samples (62.07%; 95% CI: 43.29%−80.85%), while all fecal samples tested positive (100%) (Figure 2B). Both fecal and cecal samples were positive for tetO, whereas its prevalence in drinking water was 82.76% (95% CI: 68.14%−97.38%) (Figure 2B). The mean relative abundance of the three genes, expressed as log10 (tet gene copy number/16S rRNA gene copy number), ranged from −8.36 for tetA to −9.03 for tetB (Figure 2C). Specifically, tetA abundance ranged from −8.75 in cecal content to −8.11 in fecal samples (Figure 2D). tetB abundance ranged from −9.84 in drinking water to −8.45 in feces (Figure 2D). tetO exhibited the highest relative abundance in cecal samples (−8.18) and the lowest in drinking water (−9.95) (Figure 2D).