One hundred samples were collected from infected Arabian horses (aged 10 days −23 years) from different stations in Egypt and from private cases in Sharkia and Ismailia governorates over one-year, from August 2022 to August 2023. Swabs were collected from respiratory infections, wounds, and diarrhea. Uterine fluid was obtained from cases suffering from abortion and endometritis (Supplementary Table S1). The samples were collected before initiating antibiotics treatment. Swabs were transported to the laboratory in an ice container (0–4 °C) and transported to the laboratory within an hour of collection for immediate bacteriological examination. A specimen from each case was incubated overnight in Brain Heart Infusion (BHI) broth (Oxoid, United States).

Respiratory samples were plated on BHI agar, MacConkey agar, and Edward’s medium. Swabs from infected wound cases and pus from pyogenic infections (abscesses and open wounds), as illustrated in Supplementary Figures S1A–D, were cultured on mannitol salt agar (MSA), MacConkey agar, and BHI agar. Rectal swabs from diarrheal cases were enriched in Rappaport Vassiliadis broth at 41 °C for 24 h, followed by plating on xylose-lysine-deoxycholate (XLD) agar and incubation at 37 °C for 24 h. All media were sourced from Oxoid, United States.

The isolates were examined microscopically using Gram staining, and bacterial identification was performed using standard biochemical tests; for example, catalase and coagulase tests were performed to identify Staphylococci (16). Presumptive Gram-negative bacteria were identified through different biochemical tests such as urease, citrate, oxidase, lysine decarboxylase, and triple sugar iron tests (Oxoid, United Kingdom) (17).

Antimicrobial susceptibility of the recovered isolates was performed using the disk diffusion method according to Clinical and Laboratory Standards Institute (CLSI, 2024) guidelines (24) and interpretative criteria. A suspension of the tested isolate was prepared by adjusting the turbidity with 0.5 McFarland solution and then smeared on the surface of Muller-Hinton agar (MHA, Oxoid, Cambridge, United Kingdom), and the antibiotic disks (Oxoid, UK) were placed (Supplementary Tables S5A–E). Antibiotics to which susceptibilities of isolates were recorded to determine MDR, XDR, and PDR phenotypes were selected according to Magiorakos et al. (10) and Nocera et al. (25). The plates were incubated at 37 °C for 24 h, after which the inhibition zones diameters were measured and interpreted as sensitive (S) or intermediate (I) and resistant (R) according to CLSI documents (24) and if there is lacking information the guidelines from the European Committee on antimicrobial susceptibility testing (EUCAST, 2021) (26) were used (Supplementary Tables S5A–E). Quality control was assessed using the following strains E. coli ATCC 25922, P. aeruginosa ATCC 27853, K. pneumoniae ATCCBAA-1705, S. aureus ATCC 25923, S. pyogenes ATCC 12344, S. equi ATCC 33398, A. baumannii ATCC 17978, and P. mirabilis ATCC 29245.

The multiple antibiotic resistance (MAR) index for each isolate was calculated by dividing the number of antimicrobial agents to which an isolate exhibited resistance by the total number of antibiotics tested (27).

Manure samples from horses were collected, and bacteriophages were isolated using the double agar plate method (30). The samples were centrifuged at 21,840 × g for 10 min, and the supernatant was filtered through 0.45 μm syringe-driven filters. To inactivate residual bacterial cells and other microbial contaminants prior to phage enrichment, 50 mL of the filtrate was mixed with 0.5 mL of chloroform (final concentration ~1% v/v) and incubated for 20 min at room temperature, to preserve the infectivity of most tailed dsDNA phages while reducing bacterial contamination. Subsequently, 20 mL of 2 × Luria–Bertani broth (LB, Oxoid, Cambridge, United Kingdom) and 5 mL of the bacterial culture were added, and the mixture was incubated at 37 °C for 24 h. Broth cultures after 24 h were pelleted by centrifugation at 7600 × g at 4 °C for 20 min, and they were then filtered using 0.45 μm syringe-driven filters. The filtrate was then examined for lytic phage using the Adams double-layer agar method (31). The filtered samples were collected in sterile containers and stored at 4 °C (32).

Plaque assays were used to determine phage titers. Briefly, serial dilutions of the phage suspension (10–1 to 10–6) were mixed with host bacteria and overlaid with semi-solid nutrient agar (Oxoid, Cambridge, United Kingdom) on nutrient agar plates. After incubation at 37 °C for 10 h, plaques were counted and PFU/mL was calculated using the standard formula (33).

The host range of the two bacteriophages was assessed using the spot-test against the antimicrobial-resistant Gram-positive and Gram-negative isolates. Briefly, 100 μL of an overnight bacterial culture was spread onto NA plates and incubated for 3 h at 30 °C to allow partial growth. Then, 10 μL of each phage lysate was spotted onto the surface. Plates were incubated overnight at 30 °C, air-dried, and examined for lysis zones (34). To improve visualization and accommodate isolates with slower growth rates, a modified double-layer agar technique was also employed. Specifically, 100 μL of fresh broth culture was added to 50 mL of nutrient broth and incubated for 3 h at 37 °C. Then, 100 μL of this culture was mixed with 3 mL of molten top agar and overlaid onto nutrient agar plates. Ten microlitres of bacteriophage lysate were spotted onto the surface, and plates were incubated overnight. Lytic activity was recorded based on the appearance of clear zones.

Adsorption of the phage onto the host cell was determined as previously described (38). The phage was added to an exponentially growing bacterial culture and incubated at room temperature without stirring. At several time points (1, 3, 5, 7, 9, 12, 15, 18, 21, 24, 27, and 30 min), 100 μL aliquots of the phage-bacterial combination were obtained and immediately centrifuged at 16,000 x for 30 s. Phage titers in the supernatant were then calculated by counting the plaques on overlay agar plates.

According to a previously described method (39), the bacteriophage latent time and burst size were assessed using a one-step growth technique. Fifty milliliters of broth culture were incubated until OD = 0.4 (3.0 × 10⁸ CFU/mL). The bacterial cells were then pelletized. The pelleted cells were resuspended in LB broth. To adsorb bacterial cells, approximately 0.5 mL of phage filtrate (3 × 10⁸ phage/mL) was added. The mixture (MOI = 1, calculated as the ratio of 3 × 10⁸ PFU/mL phage to 3 × 10⁸ CFU/mL bacterial cells) was centrifuged at 14,000 rpm for 30 min to remove free unabsorbed bacteriophages. The resulting pellet was resuspended in 100 mL LB broth and incubated at room temperature with continuous shaking at 120 rpm. Samples were collected from the culture flask every 3 min, and the phage titer was determined using the plaque assay.

DNA was extracted from each purified high-titer phage stock suspension using the DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany), following the manufacturer’s instructions (40). Phage genomic DNA was digested with different restriction enzymes namely, hinfI (Cat. No. FD0804), hind III (Cat. No. FD0504), and haeIII (Cat. No. FD0154) according to the instructions of the manufacturer (Thermo FastDigest®). This process was performed at 37 °C for 15 min using a thermoshaker (Biometra, Germany). The 5 μL of restriction fragments was loaded onto a 1.0% agarose gel stained with ethidium bromide and visualized on a FluorChem gel documentation system (Alpha Innotech, San Leandro, CA, United States).

The study population consisted of 12 horses presenting with purulent discharge from wound lesions at various body sites. Before the initiation of bacteriophage therapy, bacteriological examination was done on the swab samples obtained from the lesions. Horses were excluded from the study if they exhibited no clinical signs of infection and if no bacterial growth was detected. Upon admission for treatment, the 12 horses were assigned to three groups. Four horses in the first group received gentamicin (either topically or by intravenous injection at 6.6 mg/kg IV q24h (41) based on wound characteristics). Four horses in the second group received topical gentamicin combined with 4 mL bacteriophage cocktail (vB_Pae_LS225 and vB_Pae_LP125 which have wide host ranges). The final group of four horses received intravenous gentamicin alongside the topical application of the same bacteriophage cocktail.

Bacteriophage cocktail preparation (with approximately 2.2 ×10⁹ pfu/ml for vB_Pae_LP125 and 1.5×10⁸ pfu/ml for vB_Pae_LS225) was applied directly to the infected site in a 4 mL volume once every 24 h for 14 consecutive days.

A blind assessment was done for examination of wound closure on days 3, 5, 7, 10, 12, and 14- post treatment. Bacterial clearance was examined by culturing swab samples from wound lesions. The horses were examined also on day 21 for recurrence evaluation. The wound closure percentage was calculated using the following formula: Wound closure % = [V0 − Vt / V0] × 100.

Where V0 represents wound size at time zero and Vt is the wound size at time t (days).

The study protocol was approved by the Zagazig University Institutional Animal Care and Use Committee (approval no. ZU-IACUC/2/F/172/2024). Written informed consent was obtained from the owners for the participation of their animals in this study.

All experimental procedures were performed in triplicate in independent experiments. The results are expressed as the mean ± standard error of the mean (SEM). The normality and homogeneity of variance among the data were determined using Shapiro–Wilk’s and Levene’s tests, respectively. Data were analyzed using SPSS version 26 (IBM Corp, Armonk, NY, United States). The chi-square test was used to study the variations in the prevalence of different bacterial species from different origins, and to assess the differences in the antimicrobial resistance patterns of the recovered isolates from various sources. One-way ANOVA and Tukey’s post-hoc tests were used for analysis results of phage stability studies (thermal, pH, UV, one-step growth, and adsorption curves). Additionally, the independent sample t-test was performed to detect if there is a significant difference between the effect of two phages (two independent normal distributed groups) at each point (thermal, pH, UV, one-step growth, and adsorption curves). Analysis of phage treatment results were done using ANOVA and Tukey’s post hoc test. The significance level was set at p < 0.05. All graphs were generated using GraphPad Prism software version 8 (San Diego, CA, United States) and R-software version 4.0.2.¹

As depicted in Table 1, S. equi was the most predominant pathogen isolated from respiratory manifestation cases (17/29; 58.6%), followed by K. pneumoniae and P. aeruginosa (9/29; 31.03%, each), E. coli (7/29; 24.13%), and S. aureus (4/29; 13.79%). The least detected pathogens were S. pyogenes and C. ovis biovar equi (2/29; 6.89% each), A. baumannii, and CoNS accounted for 1/29; 3.44%, each. A significant difference in bacterial prevalence was observed in respiratory manifestation samples (p < 0.0001; Table 1). Polymicrobial infections were found in 62% (18/29) of cases. S. equi was found in combination with E. coli (3/29), K. pneumoniae (6/29), P. aeruginosa (4/29), Staphylococcus spp. (2/29), and C. ovis biovar equi (2/29). K. pneumoniae was also mixed with E. coli and S. pyogenes in one case and with S. equi and P. aeruginosa in another one. Additionally, A. baumannii was detected in combination with P. aeruginosa and S. aureus (1/29). S. aureus was mixed with P. aeruginosa in one case and with E. coli in another.

Streptococcus equi was isolated from all foals suffering from guttural pouch empyema (GPE). K. pneumoniae (2/5; 40%) and S. aureus (1/5; 20%). Notably, both S. equi and K. pneumoniae were found in two cases and S. equi with S. aureus in other case.

Staphylococcus aureus was the most frequently isolated bacteria from pyogenic infections (9/18; 50%), followed by C. ovis biovar equi (6/18; 33.33%), E. coli (5/18; 27.78%), and CoNS and S. pyogenes (3/18; 16.67%, each). The least isolated bacteria were P. aeruginosa (2/18; 11.11%), and S. equi and K. pneumoniae (1/18; 5.56%, each). There is a highly significant difference (p < 0.01) in prevalence of bacteria causing pyogenic infections (Table 1).

Approximately 44.44% (8/18) of pyogenic infections were mixed infections. Notably, E. coli was mixed with S. equi, C. ovis biovar equi, K. pneumoniae, and CoNS in one case and found with S. aureus in three other cases. C. ovis biovar equi was often found in mixed infections with S. aureus (3 cases) and S. pyogenes (2 cases).

K. pneumoniae were isolated from cases of abortion at 6^th and 7^th months of pregnancy. Additionally, S. equi and A. baumannii were isolated from abortion at 3.5 months of pregnancy (1/4; 25%).

The most frequently isolated bacteria from diarrheic foals were E. coli (7/12; 58.33%), followed by K. pneumoniae and P. mirabilis (4/12; 33.33%, each). Approximately 25% (3/12) of the cases involved polymicrobial infections with K. pneumoniae and both P. mirabilis twice, and E. coli once.

Streptococcus pyogenes, E. coli, C. ovis biovar equi, K. pneumoniae and P. aeruginosa were isolated from the eye infection in a horse. Additionally, S. pyogenes and C. ovis biovar equi were identified in a case. While both K. pneumoniae and P. aeruginosa were the causative agent of ocular infection in another case. E. coli was the primary cause of eye infection in a horse.

Staphylococcus aureus exhibited the highest isolation rate in wound lesions from different body parts (10/24; 41.66%), followed by E. coli and C. ovis biovar equi (9/24; 37.5%, each), S. pyogenes (7/24; 29.16%) and S. equi, P. aeruginosa, and K. pneumoniae (6/4; 25%, each).

Polymicrobial wound infections were observed in 62% (15/24) of cases. A variety of bacterial combinations were identified. E. coli, S. aureus, and S. equi were found in two cases. C. ovis biovar equi, and P. aeruginosa were the causative agents in three cases. Moreover, C. ovis biovar equi and E. coli were found in two cases. S. pyogenes, C. ovis biovar equi, and E. coli were identified in three cases. Additionally, E. coli was found with S. aureus in three cases and with K. pneumoniae in one case. S. aureus was found with S. equi in two cases, and with K. pneumoniae, S. equi, P. aeruginosa and C. ovis biovar equi in one case. Also, S. aureus, K. pneumoniae, and S. equi were the primary cause of wound infections in some cases.

A single case of polymicrobial urinary tract infection (UTI) in a 21-year-old horse involved both E. coli and P. aeruginosa. Additionally, foals with navel ill (average age: 1.5 ± 0.5 months) experienced polymicrobial infections, commonly involving S. aureus and C. ovis biovar equi. In one case, S. aureus was found in mixed infections with S. pyogenes, S. equi, E. coli, and P. aeruginosa, respectively. Furthermore, swabs from rectal tear cases yielded E. coli and S. aureus isolates, with one sample exhibiting polymicrobial infections comprising both species (Table 1).

Among 100 cases, Streptococcus spp. was the most frequently isolated microorganism (45%), primarily from respiratory infections, wounds, pyogenic infections and GPE, and was less prevalent in umbilical infections, abortions, and ocular infections. E. coli isolates (33%) was predominantly found in wound samples, followed by respiratory, diarrheic samples, and pyogenic infections, with fewer instances in rectal tears, navel illness, UTI, and ocular infections (Figure 1).

Most Staphylococcus spp. isolates (31%) were recovered from wound samples, followed by pyogenic infections, and respiratory tract infections, with a minimal prevalence from foal-infected umbilicus, and GPE.

Klebsiella pneumoniae isolates (26%) were obtained from respiratory tract infections, followed by wound and diarrhea cases, with fewer instances in abortion and GPE cases (Figure 1). Pseudomonas aeruginosa isolates (20%) is primarily caused respiratory tract infections, followed by wound infections, and minimal instances of pyogenic infections, ocular, UTI, and umbilical infections. P. mirabilis isolates (4%) was identified exclusively in diarrheic cases. A. baumanni isolates (2%) was observed in abortion and respiratory tract infections.

Corynebacterium ovis biovar equi made up 20% of the cases and was most found in wound cases, followed by pyogenic infections. Its presence has been detected at a lower frequency in respiratory, navel illness, and ocular infections. All Corynebacterium spp. isolates were identified using direct sample PCR, and positive isolates were subjected to the amplification of narG gene for C. ovis biovar equi.

All 161 bacterial isolates were resistant to multiple antimicrobial agents. Among the AMR isolates 91/161 (56.52%), 69/161 (42.85%) and 1/161 (0.62%) showed MDR, XDR and PDR phenotype, respectively (Figure 2).

Among Streptococcus spp. 44 isolates were XDR with MAR index ranging from 0.57 to 0.93. Additionally, one MDR was found in a wound case, with an MAR index of 0.64. All Streptococcus spp. isolates were completely resistant to amoxicillin-clavulanic acid, amikacin, kanamycin, cefotaxime, and streptomycin (Table 2; Figure 3).

Furthermore, S. pyogenes showed complete resistance to penicillin and tetracycline, followed by ceftiofur (92.86%), ampicillin (85.71%), enrofloxacin (78.57%), meropenem, and trimethoprim-sulfamethoxazole (64.29% each), as well as imipenem and gentamycin (75.14% each). In contrast, S. equi showed complete resistance to Ceftiofur, followed by penicillin and tetracycline (93.55%), enrofloxacin and meropenem (78.10%), gentamycin (80.65%), and ampicillin (74.19%). However, the highest sensitivity was observed for imipenem (77.42%), followed by trimethoprim-sulfamethoxazole (61.29%). Additionally, there was a significant difference in the resistance of both Streptococcus spp. to imipenem (p = 0.028; Table 2).

All Staphylococcus spp. displayed MDR pattern with MAR index range (0.3–0.8). CoNS exhibited complete resistance to fosfomycin, followed by cefoxitin, fusidic acid, clindamycin, erythromycin, and tetracycline (75% each), and Quinupristin-dalfopristin and doxycycline (50%, each). However, high sensitivity was observed for tigecycline, ceftaroline, teicoplanin, vancomycin, ciprofloxacin, and linezolid, followed by gentamycin, rifampin, trimethoprim-sulfamethoxazole, and chloramphenicol (75% each) (Table 3; Supplementary Figure S2).

On the counterpart, S. aureus exhibited complete resistance to fosfomycin followed by clindamycin and cefoxitin (96.3% each), Quinupristin-dalfopristin (92.59%), tetracycline, fusidic acid, and erythromycin (81.48% each), rifampin (77.78%), chloramphenicol (66.67%), and trimethoprim-sulfamethoxazole (59.26%). In contrast, the isolates showed complete susceptibility to tigecycline, followed by teicoplanin, vancomycin, linezolid (96.3%), ceftaroline (88.89%), gntamycin (70.37%), doxycycline (66.67%), and ciprofloxacin (59.2%). There were no statistically significant differences (p > 0.05) in the resistance profiles among all Staphylococcus spp. isolates for all tested antimicrobials (Table 3).

Methicillin resistance was determined for Staphylococcus spp. based on their phenotypic resistance to cefoxitin. Of the 27 S. aureus isolates, 26 (96.3%) were methicillin-resistant S. aureus (MRSA). All MRSA isolates were MDR and susceptible to tigecycline. Additionally, 75% of CoNS were considered methicillin-resistant CoNS (MRCoNS). MIC₅₀ and MIC₉₀ of vancomycin for Staphylococci spp. were 0.5 and 2 μg/mL, respectively. Only one isolate was found to be VRSA (MIC = 16 μg/mL). VRSA also showed resistance to ceftaroline and cefoxitin. All Staphylococcus spp. isolates were susceptible to tigecycline, with MIC₅₀ and MIC₉₀ values of 0.25 μg/mL and 0.5 μg/mL, respectively.

Out of 63 Enterobacteriaceae isolates 49 isolates showed MDR pattern with MAR index range (0.4–0.8) and 14/63 isolates showed XDR pattern with MAR index range (0.8–0.9). They exhibited total resistance amoxicillin-clavulanic acid AMC, cefazolin, cefepime, ceftazidime, and fosfomycin. Statistically significant differences were observed in the prevalence of antimicrobial resistance among Gram-negative isolates (Supplementary Table S3). The resistance rates of all isolated Gram-negative bacteria to the tested antimicrobial agents are shown in Figure 4.

Among E. coli isolates 84.85% (28/33) were MDR and 15.15% (5/33) were XDR. E. coli isolates showed complete resistance to amoxicillin-clavulanic acid, cefazolin, cefepime, ceftazidime and fosfomycin, high resistance to ampicillin (96.9%), followed by ticarcillin-clavulanic acid (93.9%), cefoxitin (90.9%), colistin (84.8%), trimethoprim-sulfamethoxazole (75.7%), ceftaroline (66.6%), tetracycline and doxycycline (51.5%, each), aztreonam (48.4%), and chloramphenicol (42.4%). The total sensitivity was observed for tigecycline, imipenem and meropenem, followed by amikacin (93.9%), gentamycin (78.8%) and ciprofloxacin (60.7%).

A high proportion of K. pneumoniae isolates were found to be resistant to multiple drugs, with 80.77% being MDR and 19.23% (5/26) being XDR. The isolates exhibited high levels of resistance to amoxicillin-clavulanic acid, cefazolin, cefepime, ceftazidime and fosfomycin, ampicillin, cefoxitin and ticarcillin-clavulanic acid followed by aztreonam (96.2%), doxycycline, tetracycline, and trimethoprim-sulfamethoxazole (76.9%, each), ceftaroline (73.1%), gentamycin (61.5%), chloramphenicol (57.7%), ciprofloxacin (53.8%), and colistin (30.8%). Conversely, the isolates demonstrated susceptibility to imipenem and meropenem, with moderate susceptibility observed for tigecycline (84.6%) and amikacin (80.8%).

All P. mirabilis isolates showed XDR pattern. They demonstrated complete resistanceto amoxicillin-clavulanic acid, cefazolin, cefepime, ceftazidime and fosfomycin, ticarcillin-clavulanic acid, ampicillin, tetracycline, doxycycline, ciprofloxacin, trimethoprim-sulfamethoxazole, colistin, gentamycin, aztreonam, cefoxitin and ceftaroline and moderate resistance rate for tigecycline (50%). Amikacin and meropenem demonstrated complete sensitivity, followed by imipenem with 75% sensitivity.

Among P. aeruginosa isolates 50% were MDR with MAR index ranging from 0.4–0.6. Nine isolates (45%) were XDR with MAR index of 0.6–0.8 and one isolate was PDR with MAR index of 1.0. All isolates exhibited resistance to ticarcillin-clavulanic acid, ceftazidime, colistin, and Fosfomycin (100%), followed by aztreonam and cefepime (90%, each), gentamycin (40%), and ciprofloxacin (30%). The highest sensitivity (95%) was found for meropenem, followed by imipenem (85%) and amikacin (75%).

Both A. baumannii isolates showed XDR pattern with MAR index ranging from 0.7–0.8. They were completely resistant to ticarcillin-clavulanic acid, cefepime, ceftazidime, imipenem, colistin, tetracycine, doxycycline, trimethoprim-sulfamethoxazole, ampicillin-sulbactam. They demonstrated complete susceptibility for gentamicin, ciprofloxacin, and meropenem. Followed by AK (50%).

Testing MIC values of tigecycline for 63 Enterobacteriaceae isolates, revealed that 33 E. coli, 22 K. pneumoniae, and 2 P. mirabilis were susceptible to tigecycline. The remaining six isolates, including four K. pneumoniae and two P. mirabilis, were resistant to tigecycline. The tigecycline MIC₅₀ and MIC₉₀ values for E. coli isolates were 0.25 μg/mL and 1 μg/mL, respectively. For K. pneumoniae isolates, these values were 0.5 μg/mL and 16 μg/mL, respectively. Additionally, tigecycline MIC range for P. mirabilis was 0.0625–8 μg/mL.

Among 85 tested Gram-negative bacteria, there was only 28 E. coli, 8 K. pneumoniae, 20 P. aeruginosa, 4 P. mirabilis, and 2 A. baumannii (72.94%) exhibited colistin resistance. The remaining 23 isolates (27.06%), including 18 K. pneumoniae and 5 E. coli isolates were considered colistin-susceptible.

The colistin MIC₅₀ and MIC₉₀ for E. coli isolates, were 4 μg/mL and 64 μg/mL, respectively. For K. pneumoniae isolates, were 0.5 μg/mL and 16 μg/mL, respectively. In P. aeruginosa isolates the MIC₅₀ was 32 μg/mL and the MIC₉₀ was 64 μg/mL. The MIC range for P. mirabilis isolates was 16–64 μg/mL, whereas that for A. baumannii was 64 μg/mL.

Two lytic phages (vB_Pae_LP125 and vB_Pae_LS225) were isolated based on the spot and plaque assays targeting host bacteria (P. aeruginosa). The plaque morphology of the two isolated phages showed various sizes, with a clear plaque center and translucent surrounding area. vB_Pae_LP125 produced small single clear plaques (1 mm), while vB_Pae_LS225 produced medium size single clear plaques (2.5 mm). Based on TEM micrographs obtained from high-titre lysates, phage vBPaeLP125 exhibited a hexagonal head approximately 50 nm in diameter and a short tail measuring approximately 34 nm, consistent with podovirus-like morphology. Phage vBPaeLS225 displayed an isometric head of approximately 66 nm and a long (103 nm), flexible, non-contractile tail, resembling siphovirus-like morphology (Figures 5A,B).

Both phages had double-stranded linear DNA (dsDNA) and the genomic size for vB_Pae_LP125 is ~48 kbp and vB_Pae_LS225 is ~39 kbp (Figure 6A). Both phages DNA produced restriction banding patterns after hinfI, haeIII and hind III digestion (Figures 6B–D).

Table 4 showed the host range and the titer for each phage exhibited as PFU/ml. The host range for each phage was determined against P. aeruginosa, E. coli, K. pneumoniae, S. aureus, P. mirablis, A. baumannii, S. equi, and S. pyogenes strains by spot assay, then confirmed by plaque assays. Plaque morphologies of phage infections on semi-solid agar plates revealed host-specific plaque formation. There were distinct differences in plaque characteristics, including variations in size (small or large plaques) and morphology (clear, circular plaques or turbid, irregular plaques) (Supplementary Figures S3, S4).

Both phages were considered thermostable at temperatures ranging from 30 to 60 °C. However, phage inactivation was detected after exposure to 70 °C for 10 min for phage vB_Pae_LP125 and 80 °C for phage vB_Pae_LS225, rendering them ineffective (Figure 7A). There were significant differences (p < 0.0001) in the effect of different temperatures on the survival of vB_Pae_LP125, and vB_Pae_LS225 with the highest effect at the initial temperature, 30 °C, and 40 °C for both phages (Supplementary Table S7).

Similarly, the exposure of both phages to UV irradiation for more than 120 min rendered them inactive and lost their infectivity (Figure 7B). There were significant differences (p < 0.0001) in the effect of different time exposure to U. V. irradiation on survival of vB_Pae_LP125, and vB_Pae_LS225 phages isolates with the highest effect at the initial time and after 20, 40, and 60 min for both phages (Supplementary Table S8).

The pH stability of the phage was determined at various pH values for 24 h. Phages were highly stable at pH 7; therefore, the phages were considered neutral. Increasing the pH above 10 and decreasing it to below 4 revealed no plaque formation activity. The highest number of plaques were observed at pH 7, which were sustained up to pH 10 (Figure 7C). There were significant differences (p < 0.0001) in the effect of different pH on the survival of vB_Pae_LP125, and vB_Pae_LS225 phages isolates with the highest effect at the initial pH and pH 6, and 7 for both phages (Supplementary Table S9).

Both phages demonstrated effective adsorption durations when tested against P. aeuriginosa strain (host strain), as shown in Figure 8A. The adsorption rate constants were 99.3% for vB_Pae_LP125 after 4 min, and 99.9% for vB_Pae_LS225 after 3 min. There were significant differences (p < 0.05) in the effect of different adsorption durations of vB_Pae_LP125, and vB_Pae_LS225 phages isolates with the highest effect at the initial time for both phages (Supplementary Table S10).

In the one-step growth assay using the host strain, the latent period for vB_Pae_LP125 and vB_Pae_LS225 were approximately 10 min, with corresponding burst sizes of 120 PFU/cell and140 PFU/cell, respectively. Conversely, the rise periods for vB_Pae_LP125 and vB_Pae_LS225 were 15 min, as illustrated in Figure 8B. There were significant differences (p < 0.0001) in the effect of different incubation periods of vB_Pae_LP125, and vB_Pae_LS225 with the highest effect after 25, 30, and 35 min for both phages (Supplementary Table S11).

Upon admittance for therapy, four horses received antibiotic treatment (gentamicin), while the other eight horses received a combination of a bacteriophage cocktail (vB_Pae_LP125 and vB_Pae_LS225) and gentamicin for treatment wound infections (Figures 9–11). Bacteriological examination of wound lesions in the enrolled horses revealed the presence of S. aureus or P. aeruginosa. Complete clearance of these pathogens was observed after the treatment course. Wound evaluations were conducted after 3–5 days to assess the treatment’s effectiveness, revealing a reduction in exudate and signs of inflammation in all horses. By day 7–10 of treatment, the formation of a granulation tissue was noted, and by 12–14 days of treatment, the formation of scars was observed (Figures 9–11). Further treatment involved the use of steroid anti-inflammatory medications to decrease the size of the scar tissue. There were significant differences (p < 0.05) in the wound closure % among the gentamicin group and phage cocktail+gentamicin groups on days 3, 5, 7, 10, and 14. The highest significant wound closure % was detected among the phage cocktail+gentamicin injection group on days 3, 5, 7, and 10. On days 12 and 14, the highest wound closure% were detected among phage cocktail+gentamicin groups, with no significant difference between the two groups (Supplementary Figure S5).