The amplification of bacteriophages using Salmonella YB1 as a host and its therapeutic efficacy against salmonellosis in broilers

Abstract

The residual bacteria in bacteriophage (phage) preparations have hindered their application in poultry disease. We isolated eight phages from aquatic environment samples and selected two phages (vB_StyS_SP03 and vB_StyM_SP07) for amplification using the diaminopimelic acid (DAP)-dependent Salmonella YB1 strain. Both healthy and Salmonella-infected broilers were intraperitoneally injected with phages amplified by Salmonella YB1 and their cocktail to evaluate the safety and therapeutic efficacy. The results showed that Salmonella YB1 effectively amplified phages, whose form and function remained unchanged after amplification. Salmonella YB1 in the phage preparations did not proliferate without DAP supplementation. Phage administration caused no mortality, and did not significantly affect the weight gain, immune organ indexes, and gut microbiota composition in healthy broilers. In Salmonella-challenged broilers, phage administration significantly improved weight gain, reduced organ indexes, alleviated clinical symptoms and pathological changes, and restored the intestinal microbial community. Collectively, we have developed a method for amplifying phages using Salmonella YB1, which effectively addresses the problem of residual bacteria in phage preparations.

Introduction

Salmonella is a common zoonotic pathogen transmitted between animals and humans (Wójcicki et al., 2022), posing a significant threat to the poultry industry and food safety (Ferrari et al., 2019). Antibiotics serve as the primary agents for the prophylaxis and treatment of salmonellosis. However, their overuse has led to a concerning rise in antimicrobial resistance (Abd-El Wahab et al., 2023), toxic effects on poultry articular cartilage (Maślanka et al., 2009), and the risk of antibiotic residues in meat tissues, which negatively impact human health.

Increasing evidence suggests that phages are natural antibacterial agents and offer a promising alternative to antibiotics (Jończyk-Matysiak et al., 2019; Kaur et al., 2021). This is because phages are specific to both the species and the strain (Lin et al., 2017). Previous studies have shown that phages contribute to combating Salmonella, when applied as environmental disinfectants in poultry farms, including on drinking utensils, shavings, and plastic surfaces (Evran et al., 2022; Korzeniowski et al., 2022). Importantly, phages successfully treated Salmonella infection in poultry, significantly increasing weight gain and helping maintain intestinal barrier homeostasis (Agapé et al., 2024; Pelyuntha et al., 2024; Chu et al., 2025). Additionally, phages reduce the bacterial growth in raw chicken breast (Jung et al., 2023; Unverdi et al., 2024) and act as a biocontrol agent against Salmonella infection in table and breeding eggs (Wang et al., 2024; Wang et al., 2025).

Phage amplification typically involves co-culturing phages with Salmonella, followed by centrifugation to remove bacterial debris and membrane filtration to produce phage preparations (Mohammadi et al., 2025). However, bacteria may not be completely eliminated and can persist after centrifugation or filtration, which poses significant challenges for phage preservation and safety assessment due to potential pathogenic risks. To address this limitation, a genetically modified Salmonella YB1 strain was employed (Yu et al., 2012), incorporating an oxygen-regulated auxotrophic system. In this engineered strain, the asd gene is regulated by a hypoxia-responsive promoter, making bacterial growth strictly dependent on exogenous diaminopimelic acid (DAP) supplementation under normal aerobic conditions (Chen et al., 2019). This design ensures that residual bacteria do not cause mortality in poultry. Previous studies have focused on the application of Salmonella YB1 for tumor treatments (Li et al., 2017; Lin et al., 2021), but no reports have addressed phage amplification using Salmonella YB1.

Thus, this study aimed to develop a novel methodology for amplifying phages using Salmonella YB1. We characterized the growth of YB1 and its efficiency in amplifying phages at various DAP concentrations, and conducted preliminary evaluations of the safety and therapeutic effects of phages amplified by YB1 in broilers. These findings may provide fundamental insights into the application of phage therapy in combating bacterial infections in poultry.

Materials and Methods

Experiment 1: in vitro study

Bacterial strains

The strains of Salmonella Typhimurium ATCC14028, Salmonella Enterica CMCC (B) 50335, Salmonella Delbrueckii LWCC1148, Salmonella Paratyphoid A CMCC (B) 50093, Salmonella Paratyphoid B CMCC (B) 50094, Salmonella Dublin CICC21497, Salmonella Bovismorbificans CICC21499, Salmonella Senftenberg CICC21511, Salmonella Gallinarum CICC21510, Salmonella Anatum CICC21498, Salmonella Choleraesuis DSM4224 and Salmonella Arizonae CMCC (B) 47001 used in this study were obtained from our laboratory stocks. Salmonella YB1 was conventionally cultured in Luria–Bertani (LB) medium supplemented with chloramphenicol (25 μg/mL) and DAP (50 μg/mL).

Phage isolation, purification and titer assay

Water samples were collected from puddles and sewage surrounding a livestock farm in Shanwei, Guangdong Province, China. The samples were centrifuged at 3,500 rpm for 10 min at 4°C to remove sediment, and the supernatant was filtered through a 0.22 μm filter membrane to obtain the filtrate. The filtrate was added to LB liquid medium containing Salmonella in logarithmic phase, and incubated at 37°C for 12-16 h to facilitate phage collection. The enriched solution was then centrifuged at 7,000 rpm for 15 min at 4°C, and the supernatant was collected and filtered through a 0.22 μm filter membrane to yield the original phage solution. Phages were purified by three rounds of single plaque isolations, and the titer was calculated using by double-layer agar plate (Kropinski et al., 2009; Ács et al., 2020).

Phage amplification

Phages were added to the logarithmic-phase Salmonella culture medium and incubated at 37°C for 8 h. The solution was centrifuged at 10,000 rpm for 10 min, and the supernatant was collected by passing it through a 0.22 µm filter. For Salmonella YB1, the culture medium was supplemented with an additional 50 µg/mL of DAP and 25 μg/mL of chloramphenicol.

Transmission electron microscopy

The morphology of phages was characterized using transmission electron microscopy (TEM). A suspension of phages was added to a copper mesh and adsorbed for 15 min (Sevier Biotechnology, China). The phages were negatively stained with 2% phosphotungstic acid in the dark for 10 min and then observed under a transmission electron microscope (Hitachi, Japan) at an acceleration voltage of 80 kV.

Biological characteristics of phage

Genomic DNA of the selected phages was extracted from highly concentrated phage lysates using the virus DNA/RNA extraction kit (Tiangen, China) and digested with restriction endonucleases (Xba Ⅰ, Xbo Ⅰ, Hind Ⅲ, EcoR Ⅰ, and BamH Ⅰ; TaKaRa, Japan) according to the manufacturer’s recommendations. The enzyme-digested products were evaluated by agarose gel electrophoresis.

The host range of the phages was determined by spot tests as described previously (Lu et al., 2022). Circular spots on the plate indicated that the phage was capable of lysing the strain. The experiment was conducted in triplicate.

The multiplicity of infection (MOI) was conducted according to the previously described method (Cao et al., 2022) with minor modifications. Briefly, phage solutions were mixed with ATCC14028 at MOI ratios ranging from 10⁻² to 10⁵, and the phage titer was quantified. Triplicate experiments were performed to identify the optimal MOI.

The one-step growth curve was determined by infecting the host with phage at the optimized MOI and incubating the culture with shaking. Samples were measured and a growth curve was plotted.

Thermal stability was determined by culturing phages at 40°C, 50°C, and 60°C, as previously described and improved (Sui et al., 2021). The pH stability was determined by mixing phages in LB medium at varying pH values (ranging from 2 to 12, adjusted using NaOH or HCl solutions).

Phage sterility test

The phage was uniformly spread onto LB agar plates using a sterile spreader. Following inoculation, the plates were inverted and incubated at 37°C for 24 h in a bacterial incubator to detect residual viable bacteria in the phage preparation.

Bacterial growth curve

The growth curve of the Salmonella YB1 strain at different DAP concentrations was determined by measuring the optical density at 600 nm (OD₆₀₀). A single colony of Salmonella YB1 was inoculated into LB broth supplemented with 50 μg/mL DAP and 25 μg/mL chloramphenicol, and cultured at 37°C with shaking at 220 rpm until the logarithmic growth phase. The bacterial suspension was inoculated into LB broth containing a gradient of DAP concentrations (0 to 1000 μg/mL) at a 1% (v/v) inoculation ratio. All cultures were incubated at 37°C with shaking at 220 rpm, and samples were measured per 2 h until the logarithmic growth phase. The experiment was performed in triplicate.

Phage amplification efficiency

Phage amplification efficiency at different DAP concentrations was determined using Salmonella YB1 as the host strain. Logarithmic-phase Salmonella was infected with phages at the optimal MOI. The infected cultures were then resuspended in LB medium containing varying DAP concentrations (0 to 500 µg/mL) and incubated at 37°C, 220 rpm. Phage titer in the supernatant of each sample was finally determined using the double-layer agar plate. The experiment was performed in triplicate.

Experiment 2: in vivo study

Animals and experimental design

One-day-old healthy yellow-feathered broilers were obtained from a commercial hatchery (Nanhai Poultry, China), and used for research after being raised for 2 days (3 days old). Water and feed were provided ad libitum throughout the trial. The phage cocktail was prepared by mixing SP03 and SP07 at a ratio of 1:1. The experimental design and treatments are detailed in Fig. 1. The animal study protocol was approved by the Institutional Animal Care and Use Committee of Foshan University (FOSU198801).

Fig. 1.

Experimental design. Both healthy and Salmonella-challenged broilers received intraperitoneal injections of phages amplified Salmonella YB1 or their cocktail for safety and efficacy evaluation. NC: negative control. Created with BioRender.com.

Trail 1

In the study of the safety of Salmonella YB1, 140 healthy broilers were randomly divided into two groups with 70 chickens per group: Salmonella ATCC14028 group and Salmonella YB1 group. Each group was further divided into seven subgroups, with 10 chickens in each subgroup. Broilers in each subgroup were intraperitoneally injected with different doses of ATCC14028 or YB1 (10² to 10⁸ CFU/mL). Clinical symptoms and mortality of the broilers were monitored daily for 7 days.

Trail 2

To investigate phage safety, 80 healthy broilers were randomly divided into four groups with 20 chickens per group (n = 20): SP03 group, SP07 group, Phage cocktail group, and negative control (NC) group. The SP03, SP07, Phage cocktail groups were intraperitoneally administered with 0.2 mL of 10⁹ PFU/mL phage, while NC group received the same dose of saline. During the one‑week experimental period under the same rearing conditions, all broilers were weighed and their clinical signs were closely monitored and documented.

Trail 3

To investigate phage therapy, 40 healthy broilers were randomly allocated to four groups with 10 broilers per group (n = 10): NC group, Phage cocktail group, Salmonella grooup and Salmonella + Phage cocktail group. Broilers in the Salmonella group and Salmonella + Phage cocktail group were intraperitoneally injected with Salmonella ATCC14028 at the determined median lethal dose (LD₅₀), while those in the NC group and Phage cocktail group received an equal amount of saline. After two hours, the Phage cocktail group and Salmonella + Phage cocktail group were intraperitoneally administered 10⁹ PFU/mL phage cocktail, while the Salmonella group and NC group received the same dose of saline. Throughout the experimental period under the same rearing conditions, all broilers were regularly weighed, and their clinical signs were monitored and recorded.

Median lethal dose

Fifty chickens were randomly allocated into five experimental groups. Groups 1-4 received intraperitoneal injections of 0.2 mL Salmonella ATCC14028 at respective concentrations (10⁶ to 10⁹ CFU/mL), while Group 5 received an equivalent volume of saline. The LD₅₀ was defined as the bacterial concentration causing 50% mortality (Cao et al., 2022).

Weight gain and organ index

The broilers were weighed individually before the experiment and after the experiment. The incidence of disease among the broilers was recorded accordingly, and both clinical and gross pathological changes were observed. Liver and spleen samples were harvested to calculate the organ index using the formula: Organ index = (organ mass / body weight) × 100%.

Serum cytokine determination

Blood samples were collected from the jugular vein of eleven-day-old broilers and centrifuged to collect the serum. Interleukin-6 (IL-6) and interferon gamma (IFN-γ) were measured using commercial enzyme-linked immunosorbent assay (ELISA) kits (Enzyme-linked Biotechnology, Shanghai, China), according to the manufacturer.

Histopathological analysis

Tissue specimens (liver, spleen, and heart) were fixed in 4% paraformaldehyde at 4°C for 24 h, followed by dehydration through a graded ethanol series (70%, 80%, 90%, and 100%). Samples were embedded in paraffin wax (58–60°C melting point) and sectioned into 4-μm slices using a rotary microtome (Leica, Germany). For hematoxylin and eosin (H&E) staining, sections were deparaffinized in xylene, rehydrated through a decreasing ethanol gradients (100% to 70%), and stained with Mayer’s hematoxylin and eosin Y.

Bacterial load quantification in visceral organs

Hepatic and splenic tissues of broilers were aseptically weighed, homogenized and subjected to serial 10-fold dilutions. Aliquots from each dilution gradient were plated onto selective Salmonella-Shigella (SS) agar, and colonies exhibiting characteristic Salmonella morphology were counted. Bacterial loads were expressed as CFU/g tissue.

16 S rRNA sequencing for gut microbiota composition

The microbial genomic DNA from cecal contents was extracted by using a Stool DNA Kit (TianGen, China). The V3 to V4 regions of the 16S rRNA gene were sequenced using universal primers (515F: 5′-CCTAYGGGRBGCASCAG-3′; 816R: 5′-GGACTACNNGGGTATCTAAT-3′). Quantified libraries were pooled and sequenced on the Illumina HiSeq 2500 PE250 platform (Novogene, China), in accordance with the effective library concentration and the required data volume. Data splitting, sequence assembly, data filtration, and chimera removal were performed on the sequencing samples to yield effective tags. For the effective tags obtained, denoising was performed with DADA2 in the QIIME2 software (Version QIIME2-202006) to derive initial amplicon sequence variants (ASVs). The relative abundance of gut microbiota at the phylum and genus levels was provided.

Statistical analysis

The data were analyzed statistically using one-way ANOVA in GraphPad Prism software (version 9.5, San Diego, CA, USA) and the Chi-square test with the IBM SPSS Statistics 26.0 (SPSS Inc., Chicago, IL, USA). The results were presented as the mean and SEM of at least three biological replicates. Duncan's multiple comparisons were used to assess the significance of the differences between treatment means. P < 0.05 was considered statistically significant.

Results

Isolation, purification, micrograph and categorizations of phages

Eight phages were obtained from samples collected from the aquatic environment of a livestock farm through repeated purification (Fig. 2A). These phages were designated names, and their specific details are presented in Table 1. To classify the phages into the morphotype-specific groups, the morphology of each phage was analyzed using TEM (Fig. 2B). These phages belong to the tail phage family. SP01, SP02, SP03, SP04, and SP06 are classified as long tailed phages with non-retractable tails, while SP05, SP07, and SP08 are muscle-tailed phages with retractable tails (Table 2). The host range of the isolated phages was tested against 12 strains of Salmonella strains of different serotypes and E. coli strain stored in the laboratory. As indicated in Table 3, all phages could lyse ATCC14028, but not E.coli ATCC25922, with SP06 demonstrating the ability to lyse Salmonella strains at a high cleavage rate of 92%. However, SP01 lysed only three types of Salmonella, with a lysis rate of 23%. In addition, SP02 could lyse four types of Salmonella with a lysis rate of 31%. SP05 and SP07 lysed five types of Salmonella with a lysis rate of 38%. SP03 could lyse six types of Salmonella with a lysis rate of 46%. SP04 and SP08 could lyse five types of Salmonella with a lysis rate of 54%.

Fig. 2.

Phage isolation and purification. (A) Plaques formed on double-layer agar plates. (B) Morphology of phage under TEM.

Table 1.

Plaque and titer of phage.

Phage	Host bacteria	Plaque	Titer (PFU/mL)
vB_StyS_SP01	ATCC14028	Transparent circular, with a diameter of 1mm, no halo rings	4.2×1011
vB_StyS_SP02	ATCC14028	Transparent circular, diameter 1.5mm, no halo ring	3.3×1011
vB_StyS_SP03	ATCC14028	Transparent circular, diameter 0.8 mm, no halo ring	8.76×1012
vB_StyS_SP04	ATCC14028	Transparent circular, diameter 4mm, no halo ring	3.8×1011
vB_StyM_SP05	ATCC14028	Transparent circular, diameter 1.2mm, no halo ring	4.0×1011
vB_SboS_SP06	CICC21499	Transparent circular, diameter 0.8mm, no halo ring	8.63×107
vB_StyM_SP07	ATCC14028	Transparent circular, diameter 0.7 mm, no halo ring	9.3×1015
vB_StyM_SP08	ATCC14028	Semitransparent circular shape, diameter 3mm, with halo ring	3.4×1011

Table 2.

The morphology of the phage.

	Head	Head diameter (nm)	Tail	Tail length (nm)
vB_StyS_SP01	Equiaxed symmetrical head	53	long tail	207
vB_StyS_SP02	Equiaxed symmetrical head	70	long tail	225
vB_StyS_SP03	Equiaxed symmetrical head	75	long tail	235
vB_StyS_SP04	Equiaxed symmetrical head	60	long tail	240
vB_StyM_SP05	Icosahedral head	95	retractable tail	99
vB_SboS_SP06	Equiaxed symmetric head	64	long tail	162
vB_StyM_SP07	Hexagonal head	90	retractable tail	100
vB_StyM_SP08	Icosahedral head	122	retractable tail	118

Table 3.

Phage lysis profile.

Host bacteria	phage
	SP01	SP02	SP03	SP04	SP05	SP06	SP07	SP08
ATCC14028	+	+	+	+	+	+	+	+
CMCC(B)50335	+	+	+	+	+	+	+	+
LWCC1148	−	−	−	−	−	+	−	−
CMCC(B)50093	−	−	+	+	−	+	−	+
CMCC(B)50094	−	−	+	+	+	+	+	+
CICC21497	−	+	+	+	+	+	+	+
CICC21499	−	−	−	+	−	+	−	+
CICC21511	−	−	−	−	−	+	−	−
CICC21510	−	−	−	−	−	+	−	−
CICC21498	−	−	−	−	−	+	−	−
DSM4224	+	+	+	+	+	+	+	+
CMCC(B)47001	−	−	−	−	−	+	−	−
ATCC25922	−	−	−	−	−	−	−	−

Note: “+”: plaque formation; “−”: no plaque formation

Biological characteristics of phages SP03 and SP07

The newly isolated phages SP03 and SP07 were selected for subsequent experiments (Table 1). The optimal MOI of SP03 was 100, while the optimal MOI of SP07 was 10 (Fig. 3A). As shown in Fig. 3B, both SP03 and SP07 exhibited typical one-step virus growth curves. The incubation period of SP03 was approximately 10 min, with a burst period of around 80 min, and the burst size of 65 PFU/cell. In contrast, SP07 exhibited a longer incubation period of approximately 30 min, a burst period of 70 min, and a burst size of 78 PFU/cell. Both phages remained activity across a pH range of 5-11, with optimal activity at pH 7, but were inactivated under extreme acidic (pH 2-4) or alkaline (pH 12) conditions (Fig. 3C). The thermal stability of SP03 and SP07 was subsequently evaluated. Both phages retained most of their activity after exposure to 40°C, 50°C, and 60°C for up to 120 min, although a slight reduction was observed after prolonged incubation (Fig. 3D). Moreover, the phage genomes of both phages were identified as dsDNA (Fig. 3E). Restriction enzyme analysis revealed distinct digestion patterns: SP03 was sensitive to Xba I, Hind III, and EcoR I, whereas SP07 was only susceptible to EcoR I (Fig. 3F).

Fig. 3.

Biological characteristics of the phages SP03 and SP07. (A) MOI of SP03 and SP07. (B) One-step growth curve. (C) pH stability. (D) Thermal stability. (E) SP03 and SP07 nucleic acid types. (F) The digestion sites of SP03 and SP07. Data are means ± SEM form at least three replicate experiments.

Amplification of phages SP03 and SP07 using Salmonella YB1

Residual Salmonella was detected after centrifugation or filtration (Fig. 4A), indicating a potential risk of bacterial carryover in phage preparations. As shown in Table 4, broilers inoculated with Salmonella YB1 showed significantly higher survival rates than those infected with the ATCC14028 strain (P < 0.05). Notably, both SP03 and SP07 exhibited effective lytic activity against Salmonella YB1, as indicated by distinct plaque formation (Fig. 4B). TEM analysis confirmed that phage morphology remained unchanged after amplification (Fig. 4C). Residual Salmonella YB1 was detected after phage amplification, but no bacterial proliferation was observed in the absence of DAP (Fig. 4D). The bacteriolytic activity of SP03 and SP07 against Salmonella was further evaluated. The negative control showed a continuous increase in OD₆₀₀, whereas SP03, SP07 and phage cocktail treatments resulted in sustained OD₆₀₀ reductions within 8 h (Fig. 4E).

Fig. 4.

Phage amplification of SP03 and SP07 using Salmonella YB1. (A) Aseptic testing of phage. (B) SP03 and SP07 lyse Salmonella YB1. (C) Morphology of SP03 and SP07 amplified by Salmonella YB1. (D) Aseptic testing of phages amplified by Salmonella YB1 with (+DAP) or without DAP (-DAP). (E) Phage activity in inhibiting bacterial growth. YB1: Salmonella YB1. NC: negative control.

Table 4.

The safety of Salmonella YB1 for broiler.

	Survival	Death	n	χ2	P
ATCC14028	55	15	70
YB1	69	1	70	13.83	< 0.05

In addition, Salmonella YB1 strains reached growth equilibrium at approximately 8 h across DAP concentrations ranging from 2 to 1,000 µg/mL. However, Salmonella YB1 growth was markedly reduced at 0.4 µg/mL DAP, and proliferation was nearly halted at lower concentrations (Fig. 5A). Correspondingly, phage titers decreased significantly (P < 0.05) when DAP concentrations ranged from 0 to 50 µg/mL during amplification (Fig. 5B). No significant differences in phage titers were observed at DAP concentrations between 50 and 500 µg/mL.

Fig. 5.

DAP concentration controls bacterial growth and phage yields. (A) Growth curves of Salmonella YB1 at different DAP concentrations. (B) Production of SP03 and SP07 at different DAP concentrations. Lowercase letters (a, b, c, d and e) indicate significant differences at P < 0.05. Data are means ± SEM at least three replicate experiments. YB1: Salmonella YB1.

Safety assessments of phages amplificated by Salmonella YB1

To evaluate phage biocompatibility, 3-day-old healthy broilers were intraperitoneally injected with SP03, SP07, phage cocktail, or saline. After 7 days, no mortality or growth retardation was observed, with 100% survival in all groups (Fig. 6A). No significant differences were observed in body weight gain, growth rates (Fig. 6B), or organ indexes (liver and spleen) among the groups (Fig. 6C). Serum IL-6 and IFN-γ levels were analyzed, showing no significant differences in cytokine concentrations between each phage-treated groups (SP03, SP07, Phage cocktail group) and NC group (P > 0.05) (Fig. 6D). As shown in Fig. 6E, the dominant phyla at the phylum level were Firmicutes and Proteobacteria. Furthermore, phage markedly reduced intestinal Proteobacteria abundance compared to the NC group, while increasing Firmicutes abundance. At the genus level, the relative abundance of Escherichia-Shigella pathogens was significantly reduced after phage injection (P < 0.05). Crucially, no adverse disturbances to the intestinal microbiota were observed, highlighting the safety of phage amplified by Salmonella YB1.

Fig. 6.

The safety of broilers after injection of SP03 and SP07 amplified by Salmonella YB1. (A) Survival rate. (B) Weight gain and growth rate. (C) Liver and spleen indexes. (D) Cytokine level (IFN-γ and IL-6). (E) The composition of gut microbiota. NC: negative control. IFN-γ: interferon gamma. IL-6: interleukin-6. ^ns indicates no significant difference at P > 0.05.

Phages amplificated by Salmonella YB1 rescued chickens from the Salmonella infection

The LD₅₀ of Salmonella ATCC14028 was determined to be 10⁷ CFU/mL (Table 5), and was utilized to establish the infection model for subsequent experiments. As shown in Fig. 7A, the weight of broilers in the Salmonella group was significantly lower than that in the NC group (P < 0.05). However, there was no significant difference in weight between the Salmonella + Phage cocktail group and the Salmonella group (P > 0.05). The weight gain and growth rate in the Salmonella + Phage cocktail group were significantly higher than in the Salmonella group (P < 0.05), although the weight gain in the Salmonella + Phage cocktail group was lower than NC group. No significant difference in growth rate was observed between the Salmonella + Phage cocktail group, Phage cocktail group, and NC group (P > 0.05) (Fig. 7B). Gross pathological examination revealed distinct differences in visceral integrity among the groups. Salmonella-infected chickens exhibited severe hepatosplenic congestion, multifocal hepatic necrosis, gallbladder distension, and cardiac encapsulation by fibrinous exudates. In contrast, broilers in the Salmonella + Phage cocktail group showed attenuated splenic congestion and fewer cardiac fibrinous deposits compared to the Salmonella group. Broilers in the NC and Phage cocktail groups maintained normal organ architecture without pathological lesions (Fig. 7C). Moreover, the liver, spleen and heart indexes of the Salmonella + Phage cocktail group were lower than those in the Salmonella infected group (P < 0.05) (Fig. 7D). Histological analysis of the liver, spleen, and heart tissues from infected broilers showed denatured liver cell granules, numerous vacuoles of varying size in the cytoplasm, disorganized liver cell cords, and uneven cytoplasmic staining. The spleen structure was damaged, with congested red pulp, enlarged white pulp lymph nodes, and swollen cells. Myocardial cell degeneration, mild swelling, thickening or even rupture of myocardial fibres, and infiltration of inflammatory cells mainly composed of macrophages between myocardial fibres. Compared with the Salmonella group, tissues and organs in the Salmonella + Phage cocktail group returned to normal, although some lesions were still present (Fig. 8A).

Table 5.

Determination of LD₅₀ of Salmonella ATCC14028.

Bacterial concentration(CFU/mL)	Total number of chickens	Number of deaths	Mortality
1×109	10	10	100%
1×108	10	7	70%
1×107	10	5	50%
1×106	10	4	40%
Negative control	10	0	0%

Fig. 7.

Effect of SP03 and SP07 amplificated by Salmonella YB1 on physiological and pathological indicators of chickens from the Salmonella infection. (A) Weight of broilers before infection and autopsy. (B) Weight gain and growth rate. (C) Gross lesions of tissues and organs. (D) Liver index, spleen index and heart index. NC: negative control. ^{a, b, c} indicate significant difference at P < 0.05.

Fig. 8.

Histopathology, organ microbial load, and intestinal microbial structure analysis of broilers from the Salmonella infection. (A) Histopathological images of the representative tissues and organs. (B) Liver and spleen bacterial load. (C) The composition of gut microbiota. NC: negative control. *P < 0.05.

We then measured the Salmonella load in the spleen and liver of the Salmonella and Salmonella + Phage cocktail groups, and found that the Salmonella count was reduced following phage therapy (Fig. 8B). Infection with Salmonella altered the normal gut microbiota structure of broilers. The relative abundance of Firmicutes in Salmonella-challenged broilers significantly differed from that of the NC group (P < 0.05), while no significant difference was observed in the abundance of Proteobacteria compared to the NC group. However, in the Salmonella + Phage cocktail group, Firmicutes abundance improved and Proteobacteria abundance reduced in the gut microbiota compared to Salmonella group, with no difference compared to normal chickens (P > 0.05). However, there was no statistically significant difference in the abundance of Firmicutes between the Salmonella + Phage cocktail group and the Salmonella group. At the genus level, there were no significant differences of the relative abundances of top 10 genera in all groups. The intestinal microbiota of all groups was dominated by Clostridia (Fig. 8C).

Discussion

Bacteriophages (phages) offer a promising alternative to traditional antimicrobials (O'Flaherty et al., 2009; Brix et al., 2020). Host bacterial residues have been identified as a major cause of failures during the early days of phage application (Pirnay et al., 2018). Effective removal of bacterial residues remains a challenge for meeting regulatory standards for pharmaceuticals products (João et al., 2021; Mohammadi et al., 2025). Downstream processing of phage typically involves primary purification through centrifugation and filtration to remove cells and debris (Hyman, 2019; Saavedra et al., 2025). However, our aseptic testing revealed incomplete bacterial removal. This contamination compromises phage purity and hinders subsequent applications. Although advanced purification strategies (including density gradient centrifugation, ion exchange chromatography, and tangential flow filtration) effectively purify phage preparations, they require significant investment in specialized equipment and expertise (Pirnay et al., 2018; João et al., 2021; Mohammadi et al., 2025), rendering them unsuitable for small-to-medium enterprises and farms. This research aims to explore an inexpensive and effective phage amplification method to address bacterial residue.

Salmonella YB1 is a derivative of the attenuated Salmonella Typhimurium SL7207 strain, with its asd gene expression regulated by a hypoxia-responsive promoter. Under aerobic conditions, asd expression is suppressed, halting the growth of Salmonella YB1 unless exogenous diaminopimelic acid (DAP) is supplemented or anaerobic conditions are maintained (Yu et al., 2012). Importantly, this genetic modification optimizes gene expression without introducing exogenous toxic components. The amplified phages are non-genetically modified organisms, as no gene editing was performed on the phage genome during propagation. The biosafety of YB1 has been extensively validated in previous studies (Ning et al., 2017). Moreover, research related to YB1 has entered human clinical trials for tumor therapy (NCT06889675, ClinicalTrials.gov), further confirming its favorable safety profile. Our results align with the previous studies showing that Salmonella YB1 exhibited no lethality in broilers due to its growth limitation under aerobic organism. These findings indicate that Salmonella YB1 is suitable host for phage propagation. However, residual YB1 may be misidentified as Salmonella contamination in routine monitoring assays of commercial broiler flocks, although prior studies show that YB1 can be effectively cleared from normal organs within 7 days (Yang et al., 2021). Therefore, future studies should track YB1 colonization and clearance over the long term to strengthen in vivo safety evidence.

Intraperitoneal administration of the phage did not adversely affect growth performance, organ indexes, or systemic inflammatory markers, including IFN-γ and IL-6, which is consistent with previous findings in poultry (Huang et al., 2022) and mice (Wang et al., 2023). In addition, phage markedly alleviated damage caused by Salmonella infection. Salmonella is known to disseminate via the bloodstream, and colonize organs such as the liver, spleen, and heart, leading to congestion, splenomegaly, tissue necrosis, and inflammatory cell infiltration (Freitas Neto et al., 2007; Sarrami et al., 2023). Our study showed similar pathological change, but phage therapy reduced histopathological organ damage and significantly lowered bacterial loads in the liver and spleen, supporting previous research that phage reduced bacterial load in various organs (Hao et al., 2023). Although liver, spleen, and heart indexes showed substantial recovery following phage therapy, complete normalization was observed only in the spleen indexes. This differential recovery may reflect the role of the spleen as a secondary immune organ with rapid immune responsiveness, whereas resolution of hepatic and cardiac inflammation likely requires longer recovery periods (Swirski et al., 2009).

We further investigated the effects of phage on gut microbiota composition. Consistent with established avian microbiota profiles (Waite and Taylor, 2015), the broiler gut microbiota was dominated by Firmicutes and Proteobacteria. Our findings align with previous studies demonstrating that phages modulate gut microbiota composition without inducing broad microbial disruption, in contrast to antibiotic treatment (Gao et al., 2017; Huang et al., 2021; Li et al., 2024; Guitart-Matas et al., 2025). Salmonella infection significantly reduced Firmicutes abundance, supporting the notion that infection disrupts the anaerobic gut environment and suppresses anaerobic taxa (Wei et al., 2024). Although Proteobacteria abundance increased after infection without reaching statistical significance in our research, this trend is still characteristic of infection-associated dysbiosis (Shin et al., 2015). Notably, phage therapy increased Firmicutes abundance and reduced Proteobacteria abundance, restoring the gut microbial community toward a healthier state. Importantly, Firmicutes are involved in energy metabolism and food digestion, and play an important role in host growth performance (Xu et al., 2025). Despite the phage-induced enrichment of Firmicutes, intraperitoneal phage administration did not fully restore body weight gain or growth rates in Salmonella-challenged broilers, consistent with the previous study that the body weight gain was not affected by intramuscular phage injection in broilers (Huff et al., 2004). In contrast, dietary inclusion of phage contributed to weight gain in broilers (Sarrami et al., 2022). This difference may be attributed to administration routes, which likely influence the enrichment of metabolism-associated Firmicutes and the sustained lytic suppression of pathogenic bacteria within the intestinal tract (Upadhaya et al., 2021).

Previous studies have shown that a DAP concentration of 50 µg/mL supports the growth of Salmonella YB1 (Guo et al., 2015; Yu et al., 2015; Ning et al., 2017; Lin et al., 2021; Hua et al., 2023). Our results demonstrate that 2 µg/mL DAP is sufficient for bacterial propagation, and that reducing DAP concentration impairs the growth rate of Salmonella YB1. Interestingly, while YB1 can still grow at low DAP concentrations, phage yields decrease in a concentration-dependent manner, highlighting the potential of DAP as an unrecognized factor influencing phage production when using YB1 as host. This suggests that although Salmonella YB1 can sustain basic metabolic processes and cell division at low DAP levels, these conditions likely fail to support optimal phage replication. The phenomenon may arise because phage production requires a high level of bacterial metabolic activity, but bacterial cell wall synthesis and membrane integrity may be compromised under low DAP conditions, reducing phage attachment and replication efficiency. In addition, previous research has shown that media supplements, such as carbon and nitrogen sources (Oh et al., 2019), metal ions (KAY, 1952; Bourdin et al., 2014), and surfactants (Kim et al., 2021), can optimize phage production by improving bacterial growth and enhancing phage adsorption to host cells (Jo et al., 2024). In line with these findings, our results suggest that DAP, a nitrogen source, may also play a critical role in optimizing phage production. Future research should explore optimizing both factors, as this combined approach may lead to more cost-effective phage production.

Although the intraperitoneal injection model does not fully replicate natural Salmonella infection routes in poultry, our findings demonstrate the safety and therapeutic potential of phages amplified by YB1. Long-term tracking studies are needed to assess the clearance kinetics of YB1 in broilers, further supporting the safety of YB1. Our findings highlight that the YB1-based amplification strategy offers a cost-effective alternative. Given their safety profile and efficacy, these phage preparations represent a viable alternative to antibiotics for poultry health management.

Conclusion

In conclusion, this study demonstrated that Salmonella YB1 is an effective host for amplifying isolated phages, with phage yields being modulated by DAP concentrations. The phages amplified by Salmonella YB1 exhibited excellent safety profiles in broilers, while effectively combating Salmonella infection, significantly reducing organ indexes, alleviating pathological damage in the tissues of liver, spleen and heart, and modulating the gut microbiota. These findings highlight the potential of substituting conventional Salmonella hosts with Salmonella YB1 in industrial phage preparation for poultry applications. However, further studies are needed to optimize culture conditions for phage amplified by Salmonella YB1 and to validate the applicability of other phages.

CRediT authorship contribution statement

Zhichao Hu: Methodology, Investigation, Data curation, Conceptualization. Jinhui Kang: Writing – original draft, Methodology, Investigation, Conceptualization. Nanyan Lin: Methodology, Investigation. Huifang Chen: Methodology, Investigation. Fang Liu: Supervision, Conceptualization. Juan Li: Supervision, Conceptualization. Cui Zhu: Writing – review & editing, Supervision, Funding acquisition. Yinshan Bai: Writing – review & editing, Supervision, Funding acquisition, Data curation, Conceptualization.

Disclosures

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.