A total of 97 Escherichia coli strains were isolated from the diarrhea of pigeons, chickens, and geese in Xinjiang. The isolates were obtained using selective media followed by PCR confirmation. These bacteria were used for phage isolation and determination of lytic profiles. All bacterial strains were cultured in Luria-Bertani (LB, Qingdao Hope Bio-Technology Co., Ltd.) broth at 37°C.

Fresh sewage samples were collected from the drainage outlet of a pigeon farm in Hotan, China, and transported to the laboratory under refrigerated conditions (4°C), where they were stored at the same temperature for short-term use. For phage isolation, 20 mL of the sewage sample was combined with 10 mL of LB broth, followed by the addition of 3 mL of the host strain 3-32-2 a MDR APEC clinical isolate obtained from a pigeon with diarrhea. The host strain (OD₆₀₀ = 0.5–0.6) belonged to serotype O166 and sequence type ST646, and harbored both virulence genes (irp2, fyuA, fimH, iroN) and resistance genes (bla _TEM-20, tet(A), sul1). The mixture was incubated at 37°C with shaking at 180 rpm for 12–16 hours to enrich for phages.

After incubation, the culture was centrifuged at 11,000 rpm for 10 minutes, and the resulting supernatant was filtered through a 0.22 µm sterile membrane to obtain the phage-containing lysate. This lysate was then mixed with an equal volume of Escherichia coli 3-32-2 and molten semi-solid agar (0.5% agarose), and the mixture was overlaid onto LB agar plates using the double-layer agar technique. Plates were incubated at 37°C for 12–18 hours. Individual clear plaques were isolated and subjected to multiple rounds of purification using the same double-layer agar method until homogeneous and well-defined plaques were achieved. The purified phage was subsequently stored at 4°C for further characterization.

The phage morphology was examined using transmission electron microscopy (TEM). In brief, purified high-titer phage particles were deposited onto a copper grid and negatively stained with 2% phosphor tungstic acid (PTA). After gently removing excess liquid, the grid was observed using a HT7800 TEM (Hitachi, Ltd., Japan).

The host range of the phage was determined by spot assay with minor modifications from the previously reported method by Asghar et al., 2022. Briefly, 100 µL of each Escherichia coli suspension was mixed with 8 mL of LB semi-solid medium (containing 0.5% agar) and promptly overlaid onto LB agar plates. Once the overlay solidified, 5 µL of the phage suspension was spotted onto the surface of each bacterial lawn. The plates were incubated overnight at 37°C, and the formation of clear plaques was used as an indicator of phage-induced lysis.

The infection kinetics were analyzed following the method previously reported by Li et al., 2019 with slight modifications. The host strain 3-32-2 was mixed with serial dilutions of phage suspensions (ranging from 1×10⁶ to 1×10¹⁰ PFU/mL) at MOI values of 0.01, 0.1, 1, 10, and 100. The mixtures were incubated at 37°C with shaking at 180 rpm for 6 hours. Following incubation, samples were filtered through 0.22 µm membranes to remove bacterial cells. Phage titers were then determined using the standard double-layer agar assay, and plates containing 30 to 300 phage spots were selected for counting. The phage potency (PFU/mL) was calculated as the number of phage spots × 10 × inverse of dilution. The experiments were performed independently three times.

The phage replication cycle were examined with slight modifications based on the method described previously by Zhou et al., 2022. In brief, phages were mixed with the host strain at a MOI of 1 and incubated at 37°C for 10 minutes to allow adsorption. The mixture was then centrifuged at 11,000 rpm for 10 minutes to separate the unadsorbed phages, which were removed by discarding the supernatant. The resulting pellet was washed twice with Luria-Bertani (LB) broth and resuspended in 10 mL of pre-warmed LB medium. The culture was incubated at 37°C with shaking at 180 rpm, and samples were collected at 10-minute intervals over a 2-hour period. Phage titers at each time point were quantified using the standard double-layer agar method. All experiments were conducted in triplicate. The experiments were performed independently three times.

Phage thermal stability was assessed by incubating phage filtrates at various temperatures (4, 25, 37, 40, 50, 60, 70, and 80°C) for 1 hour. Samples were collected at 20, 40, and 60 minutes during incubation, and phage titers were determined using the standard double-layer agar method. All temperature stability tests were conducted in triplicate. The experiments were performed independently three times.

For pH stability evaluation, phage filtrates were mixed with LB broth adjusted to pH values ranging from 2 to 13 using HCl for acidic conditions and NaOH for alkaline conditions. The mixtures were incubated at 37°C for 1 hour, after which phage titers were measured using the double-layer agar assay. Each pH condition was tested in three independent replicates. The experiments were performed independently three times.

The phage filtrate was treated with chloroform at a final concentration of 10% (v/v), thoroughly mixed, and incubated in a water bath at 37°C for 30 minutes. After incubation, phage titers were assessed using the standard double-layer agar plaque assay. All experiments were conducted in triplicate to ensure statistical reliability. The experiments were performed independently three times.

The phage-mediated growth inhibition was evaluated with modifications to the previously described method by Yao et al., 2023. Host bacteria were co-cultured with phages at MOIs of 0.1, 1, and 10 in LB broth at 37°C with shaking at 180 rpm for 12 hours. A bacterial culture without phage treatment served as the negative control. Bacterial growth dynamics were monitored by measuring optical density at 600 nm (OD₆₀₀) (BioTek Synergy HTX, USA) at 2-hour intervals. All experiments were conducted in triplicate. The experiments were performed independently three times.

Phage genomic DNA was extracted using the Rapid Bacteriophage DNA Extraction Kit (Geneaid, Taiwan), and its concentration was measured with a NanoDrop spectrophotometer (Bio-Rad, USA). Whole-genome sequencing was carried out using the Illumina NovaSeq™ platform. Raw sequencing reads were assembled into a complete genome using SPAdes v3.5.0, followed by gap closure with GapFiller v1.11 and genome refinement using PrInSeS-G v1.0.0. The assembled genome was analyzed for the presence of virulence factors using the Virulence Factor Database (VFDB, http://www.mgc.ac.cn/VFs/main.htm). Antibiotic resistance genes using the Comprehensive Antibiotic Resistance Database (CARD, https://card.mcmaster.ca/). tRNA genes were predicted using tRNAscan-SE (https://lowelab.ucsc.edu/tRNAscan-SE/). Genome visualization was performed using the CGView Server (http://cgview.ca/). Comparative genomics and phylogenetic analysis were conducted using NCBI BLASTN (http://blast.ncbi.nlm.nih.gov). A neighbor-joining phylogenetic tree was constructed to compare the complete genomes of the phages with those of their 10 closest 10 Phapecoctavirus phage relatives, using MEGA v11 with 1000 bootstrap replications.

The therapeutic efficacy of phage treatment against APEC infection in pigeons was evaluated using an oral gavage administration protocol. Phage suspensions were administered at three different timepoints relative to APEC infection 3 hours before (prophylactic), at the time of infection (concurrent), and 3 hours after (therapeutic) to assess both preventive and curative effects. A total of forty-eight 30-day-old commercial white-feathered meat pigeons were randomly assigned to six groups (n = 8 per group): three treatment groups (prophylactic, concurrent, and therapeutic), an APEC-only group, a phage-only group, and a control group. APEC strain 3-32-2 was delivered via intraperitoneal injection (4 mL of 1×10⁹ CFU/mL), while phage vB_EcoM_P3322 was administered by oral gavage (4 mL of 1×10⁹ PFU/mL). Pigeons were monitored daily for 7 days to determine survival rates. Body weight and food intake were recorded throughout the experiment. At the end of the study, blood samples were collected from the wing veins, and serum was separated and stored at –20°C for subsequent analysis. Surviving pigeons were humanely euthanized on day 7 post-infection. Under sterile conditions, organs including the heart, liver, spleen, lungs, kidneys, brain, and duodenum were harvested and fixed in 4% paraformaldehyde for histopathological examination. Tissue sections were embedded, sectioned, and stained with hematoxylin and eosin (H&E) for microscopic analysis. To evaluate the impact of treatment on gut microbiota, duodenal samples from three randomly selected pigeons per group were subjected to metagenomic sequencing. Sequencing was carried out by Beijing Novogene Co., Ltd. Raw sequencing data were preprocessed using fastp (https://github.com/OpenGene/fastp) to remove low-quality reads and generate high-quality clean data. In cases of potential host contamination, reads were aligned to the host genome using Bowtie2 (http://bowtie-bio.sourceforge.net/bowtie2/index.shtml), and any reads matching the host were filtered out. Using For taxonomic annotation, the remaining high-quality reads were aligned to the Micro_NR database using DIAMOND (https://github.com/bbuchfink/diamond/) to identify metagenomic species. Principal coordinate analysis (PCoA) was performed using the ade4 package to assess differences in microbial community structure among groups. To identify species with significant differences between groups, both MetaGenomeSeq and Linear Discriminant Analysis Effect Size (LEfSe) analyses were employed. MetaGenomeSeq was used to perform statistical testing at various taxonomic levels, generating p-values for intergroup comparisons. LEfSe analysis was conducted using LEfSe software, with a linear discriminant analysis (LDA) score threshold of >2.0 to identify taxa with significant differential abundance.

Statistical analyses were performed using GraphPad Prism and R software. Comparisons between groups were conducted using the T-test. Statistical significance was denoted as follows: * P < 0.05 (significant), ** P < 0.01 (highly significant), and *** P < 0.001 (extremely significant).

Phage vB_EcoM_P3322 was isolated and purified from sewage using APEC strain 3-32-2 as the host. On double-layer agar plates, it produced clear, circular plaques approximately 1 mm in diameter with well-defined edges ( Figure 1A ).

TEM analysis revealed that vB_EcoM_P3322 has an icosahedral head measuring approximately 91 nm in diameter and a contractile tail sheath around 105 nm in length ( Figure 1B ).

The phage demonstrated lytic activity against 65 of the 97 Escherichia coli isolates derived from various animals, including pigeon (60/91), goose (3/3), and chicken (2/3). Whole-genome sequencing of five representative susceptible strains (based on different ERIC-PCR fingerprinting patterns and strong lytic ability) led to the identification of five predicted serotypes: O8:H10, O15:H18, O51:H20, O149:H20, and O166:H6 ( Table 1 ).

The host strain 3-32-2 was infected with phage vB_EcoM_P3322 at varying MOIs to determine the phage titer and identify the optimal MOI. The highest phage titer was observed at an MOI of 1, indicating this as the optimal infection ratio ( Table 2 ).

One-step growth curve analysis ( Figure 1C ) revealed that phage vB_EcoM_P3322 has a latent period of approximately 10 minutes, followed by a burst period of 80 minutes, with an average burst size of 252 PFUs/cell (Burst size = phage titer at the end of lysis/number of host strains at the start of infection). The phage enters the stationary phase at around 90 minutes. Stability assessments under different environmental conditions showed that vB_EcoM_P3322 remains thermally stable across a broad temperature range (4–50°C), as shown in Figure 1D . However, its titer declined progressively at temperatures above 50°C and was completely inactivated after 20 minutes at 80°C. In terms of pH stability, the phage remained active between pH 5 and 9 ( Figure 1E ). A sharp decline in activity was observed at pH 4, 10, and 11, and complete inactivation occurred at extreme pH values of 2, 12, and 13. Treatment with 10% chloroform at 37°C for 30 minutes resulted in a ~1 log₁₀ PFU/mL reduction in phage titer, indicating partial sensitivity to chloroform ( Figure 1F ). To evaluate its antibacterial efficacy in vitro, strain 3-32-2 was exposed to phage vB_EcoM_P3322 at MOIs of 0.1, 1, and 10. In addition, compared to the untreated control, the phage effectively suppressed bacterial growth for up to 6 hours at all tested MOIs ( Figure 1G ). These characteristics underscore its potential as a stable and potent biocontrol agent for managing APEC infections in poultry.

Whole-genome sequencing of phage vB_EcoM_P3322 revealed a genome size of 151,674 base pairs with a GC content of 39.14% ( Figure 2A ). Genomic annotation identified 279 open reading frames (ORFs), of which only 36 (12.9%) were assigned known functions, while the remaining 243 (87.1%) were annotated as hypothetical proteins. The functionally annotated ORFs were categorized into three groups: DNA packaging and replication proteins, structural proteins, and host lysis proteins. Additionally, ten tRNA genes were predicted within the genome. Importantly, no lysogeny-related genes were detected, indicating that vB_EcoM_P3322 is a strictly lytic (virulent) phage. In addition, screening against the Virulence Factor Database (VFDB) and the Comprehensive Antibiotic Resistance Database (CARD) confirmed the absence of virulence and antibiotic resistance genes, supporting the genomic safety of vB_EcoM_P3322 for potential clinical or veterinary applications.

BLAST analysis revealed that the genome of phage vB_EcoM_P3322 shares significant similarity with multiple phages isolated globally, suggesting complex evolutionary relationships among them. Phylogenetic analysis grouped these phages into two major clades ( Figure 2B ), with vB_EcoM_P3322 clustering closely with Escherichia coli phage vB_EcoM_Ro121c4YLVW, showing 98.44% sequence identity and 97% coverage, indicating a close genetic relationship. Comparative genomic visualization using Easyfig 2.2.5 ( Figure 2C ) showed that vB_EcoM_P3322 and vB_EcoM_Ro121c4YLVW share high homology (>76%) across most functional genomic regions, although with opposite regional alignments in some areas. Based on its genomic characteristics, vB_EcoM_P3322 is classified under the class Caudoviricetes and belongs to the genus Phapecoctavirus.

The protective efficacy of phage vB_EcoM_P3322 against APEC infection was evaluated in a young pigeon model. The phage was administered via oral gavage at three different timepoints relative to APEC infection: phage administered 3 hours before APEC challenge (prevention group), phage administered simultaneously with APEC challenge (synchronization group), and phage administered 3 hours after APEC challenge (treatment group). The survival rate in the APEC-only group was 25%, while pigeons treated with phage 3 hours before infection showed the highest survival rate at 87.5%. Survival rates in the 0 h and 3 h post-infection treatment groups were both 75% ( Figure 3A ), indicating that vB_EcoM_P3322 significantly improved survival following APEC challenge. Additionally, phage-treated groups exhibited higher body weights compared to the APEC-only group, suggesting that phage therapy helped mitigate weight loss associated with APEC infection ( Figure 3B ). Feed intake decreased in all infected groups one day post-infection but gradually recovered in the phage-treated groups (the prevention, synchronization, and treatment groups), aligning more closely with the uninfected control group over time ( Figure 3C ). Serum levels of interleukin-6 (IL-6), measured via enzyme-linked immunosorbent assay (ELISA), were significantly elevated in the APEC-only group compared to the control ( Figure 3D ). In contrast, IL-6 levels were significantly reduced (P < 0.05) in all three phage-treated groups, indicating that vB_EcoM_P3322 treatment effectively alleviated the inflammatory response induced by APEC infection.

The host strain 3-32-2 did not induce any observable pathological changes in the heart or brain tissues. In contrast, significant tissue damage was observed in the APEC-infected group. Liver histology revealed hepatocellular swelling with numerous lipid droplets of varying sizes, indicative of steatosis. In the spleen, endothelial swelling of the central artery was noted, along with a reduction in the size of surrounding lymphoid follicles. Lung tissues showed extensive hemorrhage, while kidney sections exhibited widespread bleeding and hemosiderin granule deposition in the tubulointerstitium. Duodenal tissue displayed marked submucosal edema, enlarged interstitial spaces, bleeding, and infiltration of inflammatory cells, with hemorrhage observed in the intestinal villi. In comparison, phage treatment at all three timepoints (the prevention, synchronization, and treatment groups) resulted in a noticeable reduction of lipid droplets in the liver and partial alleviation of pathological changes in the spleen, lungs, kidneys, and duodenum. These tissue structures more closely resembled those of healthy, uninfected pigeons. Collectively, these findings suggest that phage vB_EcoM_P3322 can effectively mitigate multi-organ damage caused by APEC infection ( Figure 3E ).

Petal plot analysis ( Figure 4A ) revealed a total of 159,006 operational taxonomic units (OTUs) identified in the duodenal microbiota of the young pigeons, with 155,996 OTUs shared across all groups. Unique OTUs were distributed as follows: 185 in the prevention group, 151 in the synchronization group, 162 in the treatment group, 2,143 in the APEC-only group, 173 in the phage-only group, and 196 in the control group. These results indicate notable differences in gut microbial composition among the six experimental groups. Notably, APEC infection led to an increase in distinct OTUs, whereas phage treatment effectively reduced the abundance of these characteristic OTUs, suggesting a modulatory effect on the intestinal microbiota.

To gain deeper insights into the alterations in intestinal microbiota composition, we analyzed the dominant taxa at the phylum, family, and genus levels ( Figures 4B-D ). At the phylum level, Pseudomonadota, Actinobacteriota, and Bacteroidota were predominant. Notably, Pseudomonadota showed a marked increase in relative abundance in the APEC-infected group, becoming the dominant phylum ( Figure 4B ). At the family level ( Figure 4C ), Enterobacteriaceae abundance was significantly elevated in the APEC group compared to the other groups. At the genus level ( Figure 4D ), Escherichia coli was substantially more abundant in the APEC-infected pigeons. However, its abundance significantly declined in all three phage-treated groups, indicating the effectiveness of the phage in reducing Escherichia coli proliferation within the gut microbiota.

PCoA based on Bray-Curtis distance at the OTU level ( Figure 4E ) revealed that the first two principal components accounted for 85.24% (PC1) and 14.07% (PC2) of the variance, respectively. The APEC-infected group was clearly separated from the other groups, while the prevention, synchronization, treatment, phage-only, and control groups clustered closely together. This indicates that phage treatment helped maintain microbial community structure and mitigated the disruption caused by Escherichia coli infection.

Further analysis using the MetaGenomeSeq method ( Figure 4F ) showed that the abundance of Escherichia coli was significantly reduced (P < 0.01) in the phage-treated groups (prevention, synchronization, and treatment) as well as in the control group compared to the APEC group, confirming the phage’s efficacy in suppressing Escherichia coli overgrowth.

Thirty-seven OTUs were analyzed using LEfSe, and the results indicated that there were significant differences among these OTUs from the gate to the species level ( Figure 4G ). Further analysis ( Figure 4H ) identified 89 taxa with statistically significant differences among the prevention, synchronization, treatment, APEC-only, and control groups. These included 1 kingdom, 2 phyla, 7 classes, 15 orders, 15 families, 17 genera, and 32 species. The findings indicate that Escherichia coli infection in young pigeons significantly alters the composition of the duodenal microbiota. In the APEC-infected group, the relative abundance of Enterobacteriaceae and Escherichia coli in the duodenum was notably elevated, underscoring the microbial imbalance caused by infection.

Duodenal metagenomic sequencing analysis ( Figures 4B-D ) revealed the presence of Pigeon Adenovirus and Pigeon Circovirus (PiCV) in some meat pigeons. Despite this, the affected pigeons exhibited normal behavior, feed and water intake, and showed no typical clinical signs such as lethargy, diarrhea, or respiratory distress. Histopathological examination also revealed no significant damage to intestinal or immune tissues, confirming that these pigeons were asymptomatic carriers of both adenovirus and circovirus.