The study protocol involving imported beef cattle was reviewed and approved by the Research Ethics Committee of the Faculty of Veterinary Medicine, Universitas Gadjah Mada, Yogyakarta, Indonesia (Approval No. 28/EC-FKH/int./2025; approved on May 7, 2025). All sampling procedures were conducted in accordance with institutional ethical standards and national regulations governing animal handling and quarantine procedures in Indonesia. Rectal fecal sampling and environmental specimen collection were performed by trained personnel under veterinary supervision to minimize animal stress and avoid unnecessary discomfort. The study also adhered to the ARRIVE 2.0 guidelines for reporting in vivo animal research and was adapted to the operational procedures applicable to quarantine sampling of imported livestock in Indonesia.

This study was conducted from March to November 2025 at five quarantine facilities under the jurisdiction of the DKI Jakarta Quarantine Office in Indonesia, where imported Australian beef cattle were temporarily housed during mandatory quarantine before distribution. The study locations included facilities A (Cianjur, West Java), B (Sukabumi, West Java), C (Purwakarta, West Java), D (Serang, Banten), and E (Bandung Barat, West Java). These quarantine facilities were selected because they actively received consignments of imported beef cattle during the study period and served as major receiving points within Indonesia’s live cattle importation system. The geographic distribution of the study locations is shown in Figure 1, and the approximate geographic coordinates of each facility are provided in Supplementary Table S1. Laboratory analyses were conducted at the Veterinary Public Health Laboratory, Faculty of Veterinary Medicine, Universitas Gadjah Mada, Yogyakarta, Indonesia.

A cross-sectional study design was used to investigate the prevalence, virulence characteristics, phenotypic antimicrobial susceptibility, and determinants of AMR in E. coli O157:H7 isolated from imported Australian beef cattle during quarantine in Indonesia. Both fecal and environmental specimens were collected to evaluate the potential role of imported cattle as reservoirs of zoonotic pathogens and AMR determinants and to assess possible environmental contamination within quarantine facilities.

Sample collection was performed during the mandatory quarantine period immediately after cattle arrival at the quarantine facilities. The minimum required sample size for fecal sampling was calculated using the formula:

where P represents the expected apparent prevalence, Q = 1 − P, and L represents the desired absolute precision [16]. The expected apparent prevalence of Shiga toxin-producing E. coli (STEC) O157 in young Australian beef cattle was based on a national Australian survey reporting a prevalence of 8.4% [8]. The prevalence value was rounded to 9% (P = 0.09), resulting in Q = 0.91. Using a 95% confidence level and a desired precision of 5%, the calculated minimum sample size was 130 fecal samples. To improve representativeness and minimize sampling bias among quarantine facilities, the calculated sample size was multiplied by five [16], resulting in a target sample size of 650 fecal samples.

In the present study, fecal samples were collected from cattle distributed across five quarantine facilities, with 130 samples collected from each facility. In addition, six environmental samples were collected from each facility, resulting in a total of 680 samples included in this study.

A total of 680 samples were collected, comprising 650 rectal fecal specimens from imported Australian beef cattle and 30 environmental specimens, including soil (n = 20), drinking water (n = 5), and wastewater (n = 5). Fecal specimens (approximately 50 g) were collected rectally from individual cattle using sterile disposable gloves. Soil specimens (approximately 50 g) were collected from the corners of cattle pens, whereas drinking water and wastewater specimens (100 mL) were collected from drinking troughs and wastewater holding ponds within the quarantine facilities using sterile conical tubes.

All samples were transported to the laboratory in cool boxes containing ice packs and processed immediately upon arrival. Environmental sampling was incorporated to evaluate potential contamination associated with imported cattle during quarantine and to assess possible spillover pathways of zoonotic bacteria and AMR determinants within live-animal import systems.

Sample preparation was performed by homogenizing 10 g of fecal and soil specimens with 90 mL of Buffered Peptone Water (BPW; Oxoid CM0509, Oxoid, Hampshire, UK) using a stomacher. For water samples, an initial filtration step was performed using a 0.45 µm mixed cellulose membrane filter (Millipore, Merck Millipore, Darmstadt, Germany). Subsequently, 10 mL of the filtrate was transferred into 90 mL of BPW for enrichment. The homogenized samples were incubated at 37 ± 1°C for 18–24 h.

After enrichment, a loopful of culture was streaked onto Eosin Methylene Blue agar (EMB agar; Oxoid CM0069B, Oxoid, Hampshire, UK) to identify lactose-fermenting E. coli based on the characteristic green metallic sheen produced by acid precipitation of EMB dyes. A single well-isolated colony exhibiting a metallic sheen was selected and subjected to Gram staining. Microscopic examination confirmed Gram-negative short rod-shaped bacilli arranged singly or in pairs, consistent with the classical morphology of E. coli [17, 18]. When mixed colonies were observed, isolates were re-streaked to obtain pure cultures before further identification. Confirmed Gram-negative colonies were subsequently subcultured onto Nutrient Agar and subjected to standard biochemical identification procedures.

Biochemical confirmation included Indole medium (HiMedia M463-500G, HiMedia Laboratories, Mumbai, India), methyl red and Voges–Proskauer medium (HiMedia GM070I-500G), citrate medium (HiMedia M069-500G), and Triple Sugar Iron Agar (TSIA; Oxoid CM0277B, Oxoid, Hampshire, UK). Isolates showing an Indole, methyl red, Voges–Proskauer, and citrate (IMViC) pattern of (+ + − −) and TSIA reactions characterized by acid slant/acid butt with gas production and absence of hydrogen sulfide were classified as E. coli. This biochemical profile is consistent with established identification standards in veterinary microbiology [18, 19]. The reference strain E. coli O157:H7 American Type Culture Collection (ATCC) 35150 was used as a positive control throughout the isolation procedures, including EMB agar (Oxoid CM0069B, Oxoid, Hampshire, UK) and Sorbitol-MacConkey agar (SMAC agar; HiMedia M298-500G,) culture, to validate the expected morphological and biochemical characteristics.

Biochemically confirmed E. coli isolates were screened on SMAC agar (HiMedia M298-500G) to identify presumptive E. coli O157 isolates. Non-sorbitol-fermenting (NSF) colonies that appeared colorless or pale after incubation at 37°C for 24 h were classified as presumptive O157 isolates according to established diagnostic criteria [3, 20]. However, SMAC agar has limited specificity for E. coli O157:H7 because certain non-O157 STEC strains and other members of Enterobacteriaceae may also produce sorbitol-negative colonies, whereas some E. coli O157 strains may demonstrate variability in sorbitol fermentation [19, 21]. Therefore, molecular confirmation using polymerase chain reaction (PCR) targeting rfbO157 and fliCH7 genes was performed to accurately identify E. coli O157:H7 isolates.

Genomic DNA from all presumptive E. coli O157 isolates was extracted using the Quick-DNA™ Miniprep Plus Kit (Zymo Research, Irvine, CA, USA) according to the manufacturer’s instructions. Multiplex PCR assays were performed to confirm the E. coli O157:H7 serotype and determine the presence of virulence-associated genes. Four genetic targets were analyzed, including rfbO157 (O157 somatic antigen), fliCH7 (H7 flagellar antigen), and the Shiga toxin genes stx1 and stx2. Primer sequences and expected amplicon sizes are presented in Table 1 [22–24].

The multiplex PCR assay targeted rfbO157 and fliCH7 for serotype confirmation and stx1 and stx2 for differentiation of STEC E. coli O157:H7 strains, thereby enabling assessment of the virulence potential of isolates recovered from imported cattle and providing epidemiological information relevant for risk-based quarantine surveillance.

Each PCR reaction was prepared in a total volume of 50 μL consisting of 25 μL PowerPol 2× PCR Master Mix (ABclonal Technology, Wuhan, China), 1 μL of each primer pair, 19 μL nuclease-free water, and 2 μL genomic DNA template. PCR amplification was performed using an initial denaturation step at 98°C for 45 s, followed by 30 cycles of denaturation at 98°C for 10 s, annealing at 57°C for 30 s, and extension at 72°C for 30 s, with a final extension at 72°C for 5 min.

PCR products were electrophoresed on 1.5% agarose gels stained with SYBR® Safe DNA Gel Stain (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA) and visualized under ultraviolet transillumination. Isolates showing simultaneous amplification of rfbO157 (259 bp) and fliCH7 (625 bp) were confirmed as E. coli O157:H7. The presence of stx1 (637 bp) and stx2 (297 bp) indicated STEC strains, whereas isolates lacking both toxin genes were classified as non-STEC E. coli O157:H7. The reference strain E. coli O157:H7 ATCC 35150 was included as a positive control in all PCR assays. A non-template control (NTC) containing all PCR components except DNA template was included as a negative control to monitor contamination and nonspecific amplification.

Antimicrobial susceptibility testing was performed using the Kirby–Bauer disk diffusion method on Mueller–Hinton agar in accordance with the Clinical and Laboratory Standards Institute (CLSI) guidelines [25]. The antimicrobial agents tested included ampicillin (10 µg), tetracycline (30 µg), enrofloxacin (5 µg), streptomycin (10 µg), and trimethoprim–sulfamethoxazole (1.25/23.75 µg). Bacterial suspensions were prepared and adjusted to a 0.5 McFarland standard before inoculation onto Mueller–Hinton agar plates. Antibiotic disks were placed on the inoculated agar surfaces, then incubated at 37°C for 18–24 h.

After incubation, inhibition zone diameters were measured in millimeters and interpreted as susceptible, intermediate, or resistant [25]. The interpretive criteria used for each antimicrobial agent are presented in Table 2. Each isolate was tested in duplicate to ensure reproducibility, and the reported inhibition zone diameter for each antimicrobial represented the mean value of duplicate measurements. Interpretations were based on CLSI M100 Performance Standards for Antimicrobial Susceptibility Testing, 35th edition, for Enterobacterales.

AMR genes were detected by PCR to identify resistance determinants associated with the antimicrobial agents evaluated in this study. The selected genes represented major resistance determinants frequently associated with mobile genetic elements in livestock-associated E. coli, enabling identification of latent resistance potential even among phenotypically susceptible isolates. The target genes included β-lactam resistance gene blaTEM, tetracycline resistance gene tetA, aminoglycoside resistance gene strA, sulfonamide resistance gene sul2, and quinolone resistance gene qnrS. These genes were selected based on antimicrobial classes commonly used in livestock production and frequently reported among E. coli isolates recovered from Australian cattle populations [26, 27]. In addition, quinolone resistance genes were included because fluoroquinolones are categorized as critically important antimicrobials in human medicine [2, 28].

The same genomic DNA extracted using the Quick-DNA™ Miniprep Plus Kit (Zymo Research, Irvine, CA, USA) was used as the template for PCR amplification of AMR genes. Primer sequences, expected amplicon sizes, and annealing temperatures are presented in Table 3 [29–31]. PCR reactions were prepared in a total volume of 25 μL consisting of 2 μL genomic DNA template, 12.5 μL PowerPol 2× PCR Mix with Dye V2 (ABclonal Technology, Wuhan, China), 0.5 μL forward primer (10 μM), 0.5 μL reverse primer (10 μM), and 9.5 μL nuclease-free water.

PCR amplification was performed using an initial denaturation step at 98°C for 45 s, followed by 30 cycles of denaturation at 98°C for 10 s, annealing at temperatures specific for each primer pair (Table 3), and extension at 72°C for 30 s, with a final extension step at 72°C for 5 min. PCR products were separated by agarose gel electrophoresis and visualized under ultraviolet transillumination. An NTC containing all PCR reagents except DNA template was included in each PCR run to monitor contamination. Amplicon sizes were verified using a DNA ladder during electrophoresis.

A total of 680 samples were analyzed, comprising 650 fecal specimens and 30 environmental samples. Cultural and biochemical examination identified E. coli in 649 of 680 samples (95.44%), with the highest prevalence observed in fecal samples (96.15%; 625/650), indicating that feces constituted the primary source of E. coli detection in the sampled cattle population (Table 4). Colonies exhibiting a metallic green sheen on EMB agar were selected and confirmed as Gram-negative short rods through Gram staining. These isolates exhibited a characteristic E. coli biochemical profile, including Indole-positive, methyl red-positive, Voges–Proskauer-negative, and citrate-negative reactions, with an acid reaction on TSIA.

Biochemically confirmed E. coli isolates were screened on SMAC agar to identify NSF colonies. As shown in Figure 2, colorless colonies were presumptively identified as E. coli O157. Twenty-four isolates (3.8%) exhibited an NSF phenotype, all originating from fecal samples, whereas no environmental isolates showed the O157 phenotype at this stage.

Multiplex PCR amplification targeting rfbO157, fliCH7, stx1, and stx2 confirmed 16 of the 24 presumptive isolates as E. coli O157:H7. The distribution of confirmed isolates across quarantine facilities is summarized in Table 5. As illustrated in Figure 3, 16 isolates produced the expected O157 (259 bp) and H7 (625 bp) gene bands.

Of the 16 confirmed E. coli O157:H7 isolates, 11 (68.75%) were classified as STEC E. coli O157:H7 carrying stx1 and stx2, whereas 5 (31.25%) lacked both toxin genes and were classified as non-STEC E. coli O157:H7 (Table 5). Facility C showed the highest number of E. coli O157:H7 isolates (7/136), followed by facility E (4/136) and facility B (3/136), whereas no isolates were detected in facility D. The positive control (E. coli O157:H7 ATCC 35150) generated all target bands, whereas the NTC showed no amplification. Amplification patterns of the Shiga toxin genes are presented in Figure 4.

Antimicrobial susceptibility testing of the 16 E. coli O157:H7 isolates revealed variable resistance patterns across the five antibiotics tested (Table 6). Resistance was highest to tetracycline, with 5 of 16 isolates (31.25%) classified as resistant. Ampicillin resistance was detected in 1 isolate (6.25%). No resistance was observed to streptomycin, trimethoprim–sulfamethoxazole, or enrofloxacin. Intermediate susceptibility was observed for streptomycin in 3 isolates (18.75%), enrofloxacin in 2 isolates (12.5%), and ampicillin in 1 isolate (6.25%). All isolates were susceptible to trimethoprim–sulfamethoxazole. A representative antimicrobial susceptibility pattern is presented in Figure 5, illustrating isolate FLP 117, which exhibited resistance to tetracycline and ampicillin.

Resistant isolates were detected in facilities A, C, and E. Tetracycline resistance was identified in isolates from facilities A, C, and E, whereas ampicillin resistance was detected in only one isolate from facility C. No resistant isolates were detected in facility B or facility D. Analysis of resistance patterns showed that 11 of 16 isolates (68.75%) were susceptible to all tested antibiotics. Four isolates (25%) were resistant to a single antibiotic (tetracycline), whereas one isolate (6.25%) was resistant to two antibiotics (ampicillin and tetracycline).

PCR analysis revealed the presence of several AMR genes among the E. coli O157:H7 isolates (Table 4). The tetA gene was the most frequently detected, present in 7 of 16 isolates (43.75%), followed by strA in 6 isolates (37.5%). The sul2 and qnrS genes were each detected in 3 isolates (18.75%), whereas blaTEM was detected in 2 isolates (12.5%). Representative amplification patterns of the resistance genes are shown in Figure 6.

Overall, eight isolates (50.0%) carried at least one AMR gene. Five isolates (31.25%) carried multiple resistance genes. The most common gene combinations were tetA + strA (n = 3), tetA + strA + sul2 (n = 1), and tetA + qnrS (n = 1). Four isolates (25%) did not carry any of the resistance genes examined in this study. Although several resistance genes were detected, most isolates remained phenotypically susceptible to the corresponding antibiotics based on disk diffusion testing.

This study provides the first characterization of E. coli O157:H7 virulence profiles and AMR genes in imported Australian beef cattle during quarantine in Indonesia. The overall prevalence of E. coli O157:H7 detected in this study was 2.35% (16/680), with 68.75% of confirmed isolates carrying stx1 and/or stx2, indicating the presence of STEC among imported cattle entering the quarantine system. Importantly, AMR genes were detected in 50% of isolates despite predominantly susceptible phenotypes, suggesting the presence of latent AMR determinants in imported cattle populations.

The prevalence of E. coli O157:H7 observed in this study was relatively low compared with some reports from domestic cattle production systems. Based on fecal samples, the prevalence reached 2.46% (16/650), which falls within the range reported in surveys of Australian beef cattle populations. Previous studies have documented E. coli O157:H7 prevalence ranging from approximately 1%–13%, depending on feedlot management systems, geographic location, and sampling season [8, 9, 29]. More recent surveillance studies in Australian beef cattle similarly reported relatively low prevalence levels in most feedlot populations [9, 30]. Although the prevalence observed in the present study was at the lower end of this range, the detection of E. coli O157:H7 in clinically healthy imported cattle confirms that asymptomatic carriage remains common in bovine carriers and continues to represent a potential source of pathogen dissemination [15, 34].

Multiplex PCR analysis revealed that 11 of the 16 confirmed E. coli O157:H7 isolates (68.75%) carried stx1 and/or stx2, indicating the presence of toxigenic STEC strains. The detection of Shiga toxin genes is epidemiologically significant because these virulence determinants are strongly associated with severe human disease, including hemorrhagic colitis and hemolytic uremic syndrome [11, 32]. Similar patterns have been reported in Australian cattle populations, where the overall prevalence of E. coli O157:H7 is generally low, but detected isolates frequently harbor Shiga toxin genes, particularly stx2, which is associated with increased virulence [8, 9, 33].

In addition to STEC isolates, five strains (0.76%) were identified as non-STEC E. coli O157:H7, indicating the absence of both stx1 and stx2. The coexistence of toxigenic and non-toxigenic variants is well documented in bovine populations and reflects the dynamic nature of stx-encoding bacteriophages. Because stx genes are carried by lysogenic bacteriophages, they may be gained or lost during bacterial evolution [14, 35]. Consequently, non-STEC E. coli O157:H7 strains remain epidemiologically relevant because they retain the rfbO157 and fliCH7 genetic markers and may reacquire Shiga toxin genes through horizontal phage transfer.

The distribution of E. coli O157:H7 varied among quarantine facilities, with the highest frequency observed in facility C (5.38%) and facility E (3.08%), whereas no isolates were detected in facility D. Heterogeneous distribution of E. coli O157:H7 among cattle groups has been widely reported and is often associated with differences in animal origin, management conditions, and environmental exposure during transport or holding [15, 34]. Such variability highlights the importance of facility-level surveillance in quarantine systems receiving imported cattle consignments.

The distribution of STEC and non-STEC E. coli O157:H7 isolates also varied across quarantine facilities. STEC E. coli O157:H7 were most frequently detected in facility C (6/7 isolates), whereas facilities B and E contained a mixture of STEC and non-STEC variants. Facility A contained only STEC E. coli O157:H7 (2/2), whereas no E. coli O157:H7 isolates were detected in facility D. These findings indicate heterogeneity in the distribution of toxigenic strains among imported cattle consignments. Such variation may reflect differences in the origins of cattle batches, transport conditions, or management practices prior to export. Previous studies have reported that STEC E. coli O157:H7 shedding in cattle populations is often clustered within specific groups or cohorts rather than evenly distributed across herds [15, 34].

The results of environmental sampling from soil, wastewater, and drinking water collected from the quarantine facilities are presented in Table 7. Environmental sampling detected commensal E. coli in soil and wastewater, but E. coli O157:H7 was not detected in any environmental specimens (0/30). The detection of E. coli in soil and wastewater reflects the role of fecal shedding in contaminating the surrounding environment, a phenomenon widely reported in cattle production systems [2, 16]. However, the absence of E. coli O157:H7 in environmental samples may indicate limited environmental persistence or low shedding intensity during the sampling period.

Phenotypic antimicrobial susceptibility testing showed that most E. coli O157:H7 isolates remained susceptible to the tested antimicrobials, with resistance confined to a small subset of strains. Similar patterns of low AMR in cattle-derived E. coli O157:H7 have been reported in regions with strong antimicrobial stewardship policies [12, 26, 27]. Australian surveillance programs likewise reported relatively low rates of resistance among E. coli isolates from beef cattle [26, 27].

Despite the largely susceptible phenotypic profile, molecular analysis revealed the presence of AMR genes including tetA, strA, sul2, qnrS, and blaTEM. These genes were detected in several isolates, even though phenotypic resistance was not always observed. Such genotype–phenotype discordance has been documented in livestock-associated E. coli, where resistance genes may persist on mobile genetic elements without conferring detectable resistance under routine susceptibility testing conditions [2, 12]. Discordance between genotypic detection and phenotypic susceptibility has been widely reported in E. coli from livestock and environmental sources and may result from low-level gene expression, absence of promoter elements, plasmid copy number variation, active efflux systems, or silent carriage on mobile genetic elements [12–14, 36]. These latent resistance determinants may remain undetected during routine susceptibility testing but may be expressed under selective antimicrobial pressure, thereby facilitating the horizontal dissemination of resistance genes within bacterial populations.

The detection of plasmid-mediated resistance genes such as qnrS and blaTEM in phenotypically susceptible isolates underscores a previously under-recognized risk in the live animal trade: the silent introduction of transferable resistance determinants into importing countries. This observation highlights a potential biosecurity concern associated with international live-animal trade and supports integrating molecular AMR screening into quarantine protocols, offering a novel recommendation for biosecurity in high-volume import nations such as Indonesia.

Although the prevalence observed in this study falls within the range reported in Australian cattle production systems [8], the novelty of this work lies in its focus on the import–quarantine stage of the international livestock trade pathway. Most previous studies investigated E. coli O157:H7 in domestic feedlot or slaughterhouse settings within exporting countries. In contrast, the present study evaluated pathogen carriage and AMR gene profiles during quarantine in the importing country, a critical point at which pathogens and AMR determinants may be introduced into new production environments. Therefore, surveillance at quarantine entry points provides valuable biosecurity insight for monitoring the transboundary dissemination of zoonotic pathogens and AMR within a One Health framework [2, 4].

This study has several limitations. First, the number of confirmed E. coli O157:H7 isolates was relatively small (n = 16), which limited the ability to perform statistical comparisons among quarantine facilities. Second, plasmid replicon typing and conjugative transfer assays were not conducted, preventing detailed characterization of the genetic context of the detected resistance genes. Third, the study did not determine stx subtypes (e.g., stx2a, stx2c), which may influence virulence potential.

Despite these limitations, the present study provides novel data on virulence profiles and latent AMR genes in imported cattle during the quarantine phase of international livestock trade. Future studies involving longitudinal sampling during transport and post-distribution stages would further improve understanding of pathogen dynamics in the global beef supply chain.