Alginate encapsulation enhanced the gastrointestinal stability and bactericidal efficacy of an optimized Salmonella phage cocktail for oral delivery in poultry feed

Abstract

Salmonella is a leading cause of gastroenteritis, and it is transmitted via contaminated water and food sources, especially poultry products. Many serovars are resistant to frontline antibiotics, necessitating the need for alternative treatment strategies. Bacteriophage (phage) therapy offers a promising alternative to antibiotics for Salmonella infection control in poultry, however, oral delivery is limited by the highly acidic gastric environment. Thus, this study aimed to optimize a cocktail of encapsulated phages to preserve and enhance their shelf-life, viability, target release and activity in the digestive tract of chickens. The phage cocktail, consisting of three strictly lytic Salmonella phages (ST-W23, SE-W112, and ST-W139) isolated from wastewater, exhibited a broad host range, lysing ∼83 % (107/129) of 25 important Salmonella serovars identified from chicken farms. To enhance stability in acidic conditions, the phage cocktail was encapsulated in alginate–CaCO₃ jelly beads. Under simulated gastric conditions, a single application of the encapsulated cocktail reduced S. Typhimurium growth below detectable limits within 3 h post-treatment and inhibited further regrowth for 24 h. Furthermore, the encapsulated phages maintained high viability for at least a month at room temperature, eliminating the need for cold storage logistics. This study demonstrates an effective phage encapsulation technique for optimal phage storage, oral administration and therapeutic efficacy to control intestinal Salmonella in poultry and protect public health. Further work will focus on upscaling phage production and testing in poultry farms.

Introduction

Non-typhoidal Salmonella serovars, particularly S. enterica serovar Enteritidis and S. Typhimurium are the leading causes of salmonellosis in poultry and humans, globally (Hofer, 2021). These serovars are readily transmitted via fecal-oral routes, and pose a serious food safety risk to humans (Koutsoumanis et al., 2024). The widespread use of antibiotics for infection control and as growth promoters in poultry production has encouraged the emergence and spread of multidrug-resistant (MDR) Salmonella (Wang et al., 2024a). This significant threat has prompted recent regulatory bans on antibiotic usage in poultry hence, novel alternative control strategies are urgently needed (FDA, 2024). This study thus recognized the innovative therapeutic potential of bacteriophages (phages, viruses which specifically lyse bacteria) as an alternative strategy to replace antibiotics or reduce the over-reliance on these anti-infectives for Salmonella control in poultry.

Phage therapy has great promise for infection control in poultry and it has been shown to not only selectively eradicate Salmonella (including MDR serovars), but can preserve gut microbiome balance (Febvre et al., 2019; Wang et al., 2024c), and improve the health and growth of birds (Godsil et al., 2025; Upadhaya et al., 2021). Despite these advantages, however, oral phage delivery in poultry has its own challenges, as it is extremely limited by the fluctuating phage sensitivity to different pH ranges (Khan et al., 2024; Luo et al., 2025; Tao et al., 2021). Although the extremely acidic (pH range of 2.6–4.2) environment (Rodrigues and Choct, 2018) of the proventriculus and gizzard of poultry serve as a protective barrier against pathogenic bacteria, this condition is deleterious for optimal (pH 5–9) phage survival and activity (Ranveer et al., 2024). Thus, various strategies are being explored to ensure optimal therapeutic phage delivery through this gastric barrier at a stable and high infective viral particles in the gut.

The encapsulation process is widely employed in phage preservation and it is aimed at protecting phages against harsh environmental acidic stress and enabling release at a desired pH to maximize stability and therapeutic efficacy (Loh et al., 2021; Malik, 2021). Various encapsulation carrier materials, such as liposomes, fibers, nano-emulsions, and hydrogels, have been investigated, and each has its inherent advantages and disadvantages (Loh et al., 2021). Sodium alginate stands out as a superior gel-forming agent for phage encapsulation (Adamiak and Sionkowska, 2023). Derived from brown seaweed, this FDA-approved polysaccharide is widely used in sustained-release formulations of biopharmaceuticals due to its biocompatibility, non-toxicity, and ability to form hydrogels with divalent cations (Raghav et al., 2023). Various reports have shown the therapeutic potential of encapsulated phages for the control of pathogenic bacteria in livestock. For instance, encapsulation enhanced phage stability in the gastrointestinal tracts of a rat infection model, reduced Escherichia coli O157:H7 colonization and improved therapeutic outcomes (Yin et al., 2021). Pertinent to this study, treatment with an alginate-encapsulated phage cocktail successfully reduced Salmonella colonization in a chicken infection model (Zhang et al., 2022). Therapeutic studies on phage encapsulation were reported mostly from research groups in the West (Lorenzo-Rebenaque et al., 2022; Richards and Malik, 2021a). However, to our knowledge, only few studies have reported the stability and activity of naturally occurring phages (unencapsulated) at gastric pH of 2 (Dlamini et al., 2023; Kaczorowska et al., 2021; Mhone et al., 2022). Despite the leading role of Thailand in global poultry production, no information currently exists to show work conducted on phage encapsulation from Southeast Asia, particularly Thailand. This knowledge gap clearly highlights the need for encapsulation strategies to improve phage stability and therapeutic efficacy for Salmonella control and to mitigate the spread of MDR serovars in Thai poultry farms and indeed, across the globe.

In this study, three Salmonella phages were isolated from sewage samples collected from Thai farms. Whole genome sequencing confirmed that the phages are strictly virulent and lack any virulence genes. A phage cocktail consisting of the three isolated phages have broad host range activity against five major Salmonella serovars. Encapsulating the phages in an optimized alginate-based jelly beans effectively preserved the phages in simulated gastric conditions and enabled their release in alkaline-based intestinal fluids in vitro. Viability assays showed that the encapsulated phage cocktail eradicated S. Typhimurium in the simulated intestinal to below detectable limits and remained stable for over a month under storage conditions. These findings show the advantages of encapsulated Salmonella phages to enhance their shelf-life and therapeutic activity in poultry.

Materials and methods

Bacteria and growth conditions

The Salmonella and other bacterial strains examined in this study are listed in Table S1. S. Enteritidis strain 37 was used in host propagation for phage SE-W112, while S. Typhimurium strain 273 was used to propagate phages SE-W139 and ST-W23. Bacteria were cultured in trypticase soy agar (TSA) or broth (TSB) (Titan Biotech, Delhi, India) at 37°C for 18–24 h with agitation at 150 rpm. To obtain exponential-phase bacteria, overnight cultures were sub-cultured at 1:100 dilution in fresh TSB and incubated at 37°C for 2 h.

Isolation and purification of Salmonella phages from wastewater

Salmonella-specific phages were isolated from wastewater samples collected in Bangkok and Pathum Thani provinces, Thailand, using a previously described enrichment-based protocol (Phothaworn et al., 2020). S. Enteritidis strain 37 and S. Typhimurium strain 273 were used as hosts for phage enrichment, isolation and propagation.

Analysis of phage morphology by transmission electron microscopy

The morphology of the isolated phages was analysed using transmission-electron microscopy (TEM) following negative staining as previously described (Imam et al., 2019).

Phage host range assay and cocktail optimization

The phage host range analysis was conducted by testing 10⁸ PFU/mL of the phage lysates against the 129 Salmonella cultures using an optical density (OD)-based microplate reader (BioTek Instruments, Winooski, VT, USA) and spot test (Nale et al., 2020). The cultures were prepared from isolates representing S. Enteritidis (n = 20), S. Typhimurium (n = 20), S. Choleraesuis (n = 11), S. Hadar (n = 5), S. Virchow (n = 5), and other serovars (n = 68) (Table S1). Phage specificity and potential cross-reactivity was also tested on 35 non-Salmonella strains shown in Table 1. To measure the OD₅₉₅, phage cocktails were added at 1:10 (v/v) ratios to exponential-phase cultures at OD₅₉₅ ∼0.2 (10⁸ CFU/mL) in 96-well plates, and bacterial growth was monitored hourly over 12 h. Lytic activity was indicated by reduced OD compared with controls. For the spot assays, 10 µL of each cocktail was applied to Salmonella lawns, and lysis was evaluated using inhibition zones. Lysis assays were performed with three independent biological replicates. Lytic coverage (%) was calculated for each formulation and was determined by calculating the proportion of isolates that exhibited lysis relative to the total number tested. To optimize the phage cocktail, three combinations were developed from the three phages (ST-W23 at 10¹¹PFU/mL, ST-W139 10¹¹PFU/mL, and SE-W112 10¹⁰PFU/mL) at the following ratios: (i) ST-W23, ST-W139, and SE-W112 (1:1:2); (ii) ST-W23 and SE-W112 (1:2); and (iii) ST-W23 and ST-W139 (1:1). OD measurements were assayed as described above.

Table 1.

Lytic activity of individual phages and their combination as a cocktail against 20 different Salmonella serovars and other foodborne pathogens.

Bacteria			% Lytic activity (isolates)
			Single phage			Phage cocktails
		No of isolates	ST-W23	SE-W112	ST-W139	ST-W23 + SE-W112	ST-W23 + ST-W139	ST-W23 + SE-W112 + ST-W139
Salmonella serovars	Panama	1	100.00%	100.00%	0.00%	100.00%	0.00%	100.00%
	Anatum	2	(1/1)	(1/1)	(0/1)	(1/1)	(0/1)	(1/1)
			0.00%	0.00%	0.00%	0.00%	0.00%	0.00%
	Agona	9	(0/2)	(0/2)	(0/2)	(0/2)	(0/2)	(0/2)
			0.00%	100.00%	0.00%	100.00%	0.00%	100.00%
	Albany	7	(0/9)	(9/9)	(0/9)	(9/9)	(0/9)	(9/9)
			14.29%	14.29%	14.29%	14.29%	14.29%	14.29%
	Altona	1	(1/7)	(1/7)	(1/7)	(1/7)	(1/7)	(1/7)
			0.00%	0.00%	0.00%	0.00%	0.00%	0.00%
	Brancaster	1	(0/1)	(0/1)	(0/1)	(0/1)	(0/1)	(0/1)
			0.00%	100.00%	0.00%	100.00%	0.00%	100.00%
	Bredeney	1	(0/1)	(1/1)	(0/1)	(1/1)	(0/1)	(1/1)
			100.00%	100.00%	100.00%	100.00%	100.00%	100.00%
	Bovismorbificans	1	(1/1)	(1/1)	(1/1)	(1/1)	(1/1)	(1/1)
			100.00%	0.00%	100.00%	100.00%	100.00%	100.00%
	Falkensee	1	(1/1)	(0/1)	(1/1)	(1/1)	(1/1)	(1/1)
			100.00%	0.00%	0.00%	100.00%	100.00%	100.00%
	Give	7	(1/1)	(0/1)	(0/1)	(1/1)	(1/1)	(1/1)
			85.71%	0.00%	0.00%	28.57%	28.57%	28.57%
	I 4, 5, 12:i:-	10	(6/7)	(0/7)	(0/7)	(2/7)	(2/7)	(2/7)
			90.00%	90.00%	20.00%	100.00%	90.00%	100.00%
	Indiana		(9/10)	(9/10)	(2/10)	(10/10)	(9/10)	(10/10)
		1	100.00%	100.00%	0.00%	100.00%	100.00%	100.00%
	Kedougou	10	(1/1)	(1/1)	(0/1)	(1/1)	(1/1)	(1/1)
			80.00%	20.00%	40.00%	70.00%	90.00%	100.00%
	Kentucky	3	(8/10)	(2/10)	(4/10)	(7/10)	(9/10)	(10/10)
			100.00%	0.00%	66.67%	33.33%	100.00%	100.00%
	Mbandaka	1	(3/3)	(0/3)	(2/3)	(1/3)	(3/3)	(3/3)
			100.00%	0.00%	100.00%	100.00%	100.00%	100.00%
	Orion	4	(1/1)	(0/1)	(1/1)	(1/1)	(1/1)	(1/1)
			0.00%	25.00%	0.00%	25.00%	25.00%	50.00%
	Saintpaul	1	(0/4)	(1/4)	(0/4)	(1/4)	(1/4)	(2/4)
			0.00%	100.00%	0.00%	100.00%	0.00%	100.00%
	Schwarzengrund	3	(0/1)	(1/1)	(0/1)	(1/1)	(0/1)	(1/1)
			0.00%	33.33%	0.00%	33.33%	0.00%	33.33%
	Senftenberg	3	(0/3)	(1/3)	(0/3)	(1/3)	(0/3)	(1/3)
			0.00%	0.00%	0.00%	0.00%	0.00%	0.00%
	Stanley	1	(0/3)	(0/3)	(0/3)	(0/3)	(0/3)	(0/3)
			0.00%	100.00%	100.00%	100.00%	100.00%	100.00%
	Total	68	(0/1)	(1/1)	(1/1)	(1/1)	(1/1)	(1/1)
			48.53%	42.65%	19.12%	60.29%	45.59%	69.12%
			(33/68)	(29/68)	(13/68)	(41/68)	(31/68)	(47/68)

Acidic pH tolerance of non-encapsulated Salmonella phages

The susceptibility of each Salmonella phage to acidic pH conditions was evaluated by incubating the lysates in sodium-magnesium (SM) buffer (100 mM NaCl, 8 mM MgSO₄·7H₂O, and 50 mM Tris–HCl) at various pH ranges. Briefly, 100 µL phage suspension (∼10⁹ PFU/mL) was added to 900 µL SM buffer adjusted to pH 7 (control), 3.0, 2.5, or 2.0 using 1 M HCl (Sigma-Aldrich, USA). The mixtures were incubated at 40°C for 30, 60, or 180 min. Phage viability was assessed at these time points using the double-agar layer plaque assay (Nale et al., 2020).

Phage genomic DNA extraction, sequencing, assembly and annotation

The phage genomic DNA was extracted using a previously described phenol–chloroform–isoamyl alcohol extraction protocol (Phothaworn et al., 2020). Phage genomic libraries were prepared using the Illumina DNA Library Prep kit (Illumina, Inc., USA), sequenced using an Illumina MiSeq platform and annotated as described previously (Withatanung et al., 2024). Phage taxonomy and protein-based hierarchical clustering were performed using VIPtree (Nishimura et al., 2017). Genomic analysis included examining comparative gene synteny and genomic organization using Clinker v.0.0.25 (Gilchrist and Chooi, 2021). Furthermore, phage sequences were examined for virulence factors and antimicrobial resistance genes using PhageScope (Wang et al., 2024b). The whole genome sequences of ST-W23, SE-W112, and ST-W139 were submitted to the GenBank database under accession numbers respectively.

Alginate encapsulation of Salmonella phages

The phage stocks ST-W23 (10¹¹ PFU/mL), ST-W139 (10¹¹ PFU/mL), and SE-W112 (10¹⁰ PFU/mL) were mixed at the optimal 1:1:2 ratio (as determined above). To perform alginate encapsulation, sodium alginate (3 % w/v, Sigma-Aldrich) and CaCO₃ (1.2 % w/v, KEMAUS, Australia) were dissolved in SM buffer and autoclaved. The CaCO₃ serves as a pH-buffering agent that reduces the acidity around the beads during gel formation, therefore protecting phage viability. The phage cocktail suspension (combined, ∼10¹⁰ PFU as determined by spot assay on confluent mixed mid-log cultures of S. Enteritidis 37 and S. Typhimurium 273) was then added to the sterile alginate–CaCO₃ mixture (1:5 v/v) and stirred at room temperature for 3 min until completely dissolved. Subsequently, the mixture was extruded dropwise into 10 % (w/v) CaCl₂ (VWR Chemicals, Belgium) through a 22 G needle using a flow rate peristaltic pump (Shenchen LabV1, China). After 30 min of crosslinking, phage capsules were collected with a sterile mesh strainer, gently washed three times with SM buffer, stored at 4°C and used within 7 days. To assemble alginate–pea protein beads, CaCO₃ was replaced with 3 % (w/v) pea protein suspension in SM buffer, and mixed with alginate at the same 1:5 ratio.

Calculation of encapsulation efficiency

The encapsulation efficiency (EE) of the phages was determined by comparing the amount of phages released from the capsules (P_released) to the total amount of phages initially used for encapsulation process (P_total), and expressed as a percentage as follows: EE (%) = (P_released/P_total) × 100 (Zhou et al., 2022). P_released was measured by dissolving 0.1 g of capsules in 1 mL citrate–bicarbonate–Tris buffer (pH 7.5) at 40°C for 1 h under gentle shaking at 50 rpm (Moghtader, Egri and Piskin, 2017). Phage titers released from the beads were determined using the double-agar layer method, with a top bacterial lawn prepared by mixing 100 µL mid-log phase cultures of S. Enteritidis 37 and S. Typhimurium 273 (50 µL each) into 3 mL soft agar (0.35 % TSA with 5 mM CaCl₂) containing 100 µL the released phages and overlaying the mixture onto 1 % TSA plates, followed by incubation at 37°C for 18–24 h. The resulting plaques were counted and expressed as PFU/mL to assess the activity of the phage cocktail.

Acidic stability testing under simulated gastric conditions

The stability of the encapsulated Salmonella phage cocktail under simulated gastric conditions was determined as previously described (Colom et al., 2017), with some modifications. Briefly, phage cocktail-loaded alginate beads (3 beads equate to 10¹⁰ PFU/mL) were incubated in 1 mL simulated gastric fluid (SGF; pH 2) containing 3 mg/mL pepsin (Sigma-Aldrich, MO, USA) in 0.85 % (w/v) NaCl at 40°C with shaking at 50 rpm. At the indicated time points (30, 60, or 120 min after incubation), the SGF was carefully removed by pipetting, and the beads were subsequently dissolved in 1 mL 50 mM bile salts (Sigma-Aldrich, MO, USA) at 40°C for 1 h with gentle shaking (Colom et al., 2015). The volume of phages released was quantified using the double-agar overlay plaque assay with a mixed-host lawn.

Phage target release assay under simulated intestinal pH conditions

The release of phages from the alginate encapsulated beads in the simulated intestinal fluid (SIF) was quantified as described previously (Colom et al., 2017). Briefly, phage cocktail-loaded alginate beads were incubated in 1 mL SIF (pH 8) containing 0.85 % (w/v) NaCl, 1 mg/mL pancreatin (Sigma-Aldrich, MO, USA) and 10 mM bile salts at 40°C with gentle shaking at 50 rpm. At 30, 60, 120, 180, or 240 min after incubation, phages were enumerated using the mixed-host lawn method as described above.

In vitro gastric and intestinal transit simulation

The protective effect of encapsulation on phage viability during gastrointestinal transit was evaluated using a sequential exposure model which simulate gastric and intestinal transit. Encapsulated phage cocktails (in alginate–CaCO₃ beads) were first incubated in SGF (pH 2.0) for 3 h at 40°C with continuous shaking at 50 rpm. Afterwards, samples were adjusted to pH 8.0 and then transferred to SIF, followed for another 3 h incubation at 40°C with continuous shaking at 50 rpm. The pH was maintained at 8.0 throughout the intestinal phase. For the control treatment, which evaluated phage release in nonstress conditions, non-encapsulated phages were subjected to the same SGF and SIF conditions under the same time and temperature conditions. Phages viability were enumerated at 0, 30, 60, 120, and 180 min during SGF and SIF exposures, using the double-agar overlay method on a mixed-host Salmonella lawn as described above. The proportion of phages that were released into the digestion supernatant were calculated as described above.

Killing efficiency of alginate-encapsulated phage cocktail under simulated intestinal condition

The impact of the alginate-encapsulated phage cocktail on Salmonella growth in SIF was evaluated as described previously (Zhou et al., 2022) with some modifications. Essentially, 3 mL SIF was spiked with 300 µL of 10⁴ CFU/mL S. Typhimurium 273 and treated with different concentrations of the optimized phage-alginate encapsulated cocktail. Single phage application consisted of a low-treatment group (3 beads, 0.1 g, 1.5 × 10⁹ PFU/mL) and the medium-treatment group (8 beads, 0.3 g, 4 × 10⁹ PFU/mL) administered at 0 h. The sequential treatment group received 8 beads (0.6 g, 8 × 10⁹ PFU/mL) at 0 h, followed by an additional 8 beads (0.3 g, 1 × 10⁹ PFU/mL) each at 0.5 and 6 h time points. The non-encapsulated phage cocktail (10⁹ PFU/mL) was used as the positive control. All treatment groups were incubated at 40°C with shaking at 50 rpm, and samples were collected at 0, 1, 2, 3, 4, 5, 6, 12, 18, and 24 h to quantify Salmonella recovery by colony counting on TSA plate. Phages in the supernatants over time were also enumerated using the double-agar overlay assay. All testing of encapsulated phages were conducted 40 °C instead of 37 °C to mimic the normal body temperature of chickens.

Statistical analyses

Statistical analyses was performed using GraphPad Prism 9.0 software (GraphPad, San Diego, CA, USA). Student’s t-test was performed for between-group comparisons, and the one-way analysis of variance (ANOVA) was used to compare the variance among the treatment group means. P-values < 0.05 were considered statistically significant.

Results

All three isolated phages showed serovars-specific lytic activity against Salmonella

Three novel Salmonella phages ST-W23, SE-W112, and ST-W139 were isolated as a result of screening 50 wastewater samples from Thailand. Evaluating the host range of the phages revealed high lytic activity against 61 Salmonella isolates which represent the five most prevalent serovars (S. Typhimurium, S. Enteritidis, S. Choleraesuis, S. Hadar, and S. Virchow) from poultry sources (Fig. 1). The host ranges of ST-W23, SE-W112, and ST-W139 were 50.8 % (31/61), 63.9 % (39/61), and 49.2 % (30/61) of all the isolates examined, respectively (Fig. 1A). In addition, the lytic activity of the phages was also evaluated against 68 additional isolates across 20 other Salmonella serovars isolated from Thai poultry farms (Table 1). Further analysis of ST-W23, SE-W112, and ST-W139 against the 68 additional isolates demonstrated 48.6 % (33/68), 42.7 % (29/68), and 19.1 % (13/68), respectively (Table 1). Notably, the phages showed no activity against the 35 non-Salmonella isolates Escherichia coli (n = 10), Shigella spp. (n = 10), Vibrio parahaemolyticus (n = 5), Listeria monocytogenes (n = 5), and Staphylococcus aureus (n = 5) examined.

Fig. 1.

Lytic activity and killing kinetics of the Salmonella phages that were isolated in this study. (A) Heat map showing the lytic activity of the phages individually and when combined. Each of the phages ST-W23, SE-W112, and ST-W139 were tested for their lytic activity against five major Salmonella serovars in comparison with various combinations of the phages. The lysis percentage was calculated based on optical density at 595 nm (OD₅₉₅) reduction at 6 h post-infection relative to untreated controls. (B) Killing kinetics of the phage cocktail against Salmonella under different multiplicities of infection (MOI). Five major Salmonella serovars (10⁴ CFU/mL) were incubated with the phage cocktail at various combined MOIs. The bacterial density was monitored by OD measurements over 360 min time. Data are expressed as the mean ± standard deviation of three biological and technical replicates.

Although each phage showed lytic efficacy against the different Salmonella serovars, their complementary host range and lytic effects and ability to mitigate antibiotic resistance were also examined to further determine their therapeutic effectiveness. Consequently, three distinct phage cocktails of various phage combinations were formulated and evaluated for their lytic activity against 129 representative Salmonella isolates. Among the three combinations (ST-W23, SE-W112 and ST-W139, ST-W23 and SE-W112, and ST-W23 and ST-W139) tested, the three-phage cocktail exhibited the highest lytic efficacy, reducing Salmonella growth more than the other two-phage cocktails (Fig. 1A). Furthermore, the three-phage cocktail lysed 60/61 representative isolates of the five major serovars and 47/68 isolates of the 20 other Salmonella serovars, demonstrating broad lytic coverage (Table 1). The overall lytic efficacy reached 82.9 % (107/129 isolates), which is significantly higher than the average 50 %–60 % efficacy of the individual phages (Table 1).

To provide insight into potential dosing regimen, we challenged a representative isolate from each of the five major serovars using different multiplicities of infection (MOIs, phage ratio to bacteria) of the three-phage cocktail. S. Enteritidis, S. Typhimurium, and S. Choleraesuis were all eradicated to below detection limits within 6 h post-exposure across a wide MOI range (0.01–10,000) with no bacterial regrowth noticed till up to 24 h, demonstrating effective suppression of bacterial growth resistance (Fig. 1B). Moreover, moderate activity was observed against S. Virchow and S. Hadar, with the former showing greater susceptibility to the phage cocktail (Fig. 1B). These findings highlight the broad lytic activity and therapeutic potential of the phage cocktail.

Genomic analyses revealed that the phages are strictly lytic and encode no virulence and lysogeny genes

To further determine the therapeutic potential and suitability of the three phages for Salmonella control, whole-genome sequencing and analysis were conducted to identify any undesirable genes. All the three phages were strictly virulent and exhibited no lysogeny genes that could mediate horizontal gene transfer. Furthermore, the phages lacked any antimicrobial resistance and virulence genes in their genomes (Tables S2-S4). These findings indicate the their great potential of the phages for Salmonella biocontrol and thus suitability for therapeutic deployment.

Further analyses revealed that each of the three phages possessed a double-stranded DNA genome. As shown in Fig. 2A, BLASTn analysis revealed that ST-W23 showed a high nucleotide sequence similarity with Salmp-p7 (95.6 % average nucleotide identity [ANI] over 98 % of the genome length; PQ189589) and vB_SenS_BPS1 (97.7 % ANI over 98 % of the genome length; MW 355474), which originated from China and the UK, respectively (Chen et al., 2024; Rodwell et al., 2021). Similarly, ST-W139 showed a strong similarity to HMD-P1 (98.9 % ANI over 99 % of the genome length; PQ471667), and Season12 (95.8 % ANI over 96 % of the genome length; MK286578), which were isolated in Turkey and the USA, respectively (Guzel et al., 2024; Xie et al., 2021). Similarity, SE-W112 exhibited 99.9 % ANI over 100 % of the genome length of SE-W109 (NC_073170) isolated from Thailand (Phothaworn et al., 2020) and 92.2 % ANI over 95 % genome of YMFM0293 (OK357457), which was isolated in Korea (Lee, Kim and Kim, 2023). Overall, the three Thai Salmonella phages showed high similarities not only to phages isolated in other Asian countries but also to those isolated in the UK and the USA.

Fig. 2.

Genomic and morphologic characters of the Salmonella phages that were isolated in this study. (A) Comparative genome maps of phages ST-W23, ST-W139, and SE-W112 in relation to six reference genomes. Stars indicate the three Salmonella phage isolated in this study. The grey shaded bars represent the percent amino acid identity from 1 % to 100 %. (B) The proteomic phylogenetic tree of phages ST-W23, SE-W112, and ST-W139, along with 50 homologous sequences available from the GenBank database. Red stars indicate the locations of phages ST-W23, SE-W112, and ST-W139. Phage taxa are indicated on the right side of the tree. (C) Morphology of the isolated Salmonella phages as ascertained using TEM. All phages possess long, non-contractile tails. Scale bar represents 100 nm.

To determine the potential relationships between the novel phages and other Salmonella phages, a proteomic phylogenetic analysis was conducted on 50 closely related phage genomes available from the GenBank database using the VIPtree program (Nishimura et al., 2017). This analysis revealed that phages ST-W23 and ST-139 were categorized under the family Chaseviridae and genus Chivirus (clade highlighted yellow, Fig. 2B) but were classified into distinct lineages (Fig. 2B). In contrast, SE-W112 belonged to the family Autographiviridae and genus Jerseyvirus (clade highlighted in purple, Fig. 2B). TEM further showed that all three phages possessed long, non-contractile tails. ST-W23 had a head diameter of 65 ± 6 nm and tail length of 260 ± 6 nm; SE-W112 had a head diameter of 70 ± 3 nm and tail length of 133 ± 5 nm; and ST-W139 had a head diameter of 71 ± 3 nm and tail length of 238 ± 2 nm (Fig. 2C).

Non-encapsulated Salmonella phages were sensitive to an acidic environment

Phage stability in the simulated acidic gizzard environment (pH 2–4) is essential for oral delivery and biocontrol in chicken guts. Thus, the acid tolerance of the three naturally occurring and naïve phages was investigated. As shown in Fig. 3, the non-encapsulated phages demonstrated stability at pH 3.0 for up to 180 min, indicating higher acid resilience compared to typical phages being stable within a pH range of 5–9 (Ranveer et al., 2024). At pH 2.5, only SE-W112 remained viable up to 60 min; however, its abundance significantly decreased to 1.3 log₁₀ PFU/mL after 3 h of exposure (p = 0.0095). None of the three phages survived at pH 2.0, even with only 30 min of exposure (Fig. 3). These results demonstrate the limited viability of phages under highly acidic conditions and highlight the need for encapsulation approaches to ensure their stability and resistance during application.

Fig. 3.

Susceptibility analyses of non-encapsulated Salmonella phages to acidic conditions. Salmonella phages ST-W23, SE-W112 and ST-W139 were incubated in sodium–magnesium (SM) buffer adjusted to the pH range shown at 40°C. After exposure for 30, 60, or 180 min, the numbers of viable phages were determined by double layer plaque assay. Phages in pH 7.5 SM buffer were used as the control treatment. Assays were performed in duplicate and repeated twice. Data were expressed as mean ± standard error of mean. ** indicates p < 0.01.

Alginate–CaCO₃ formulation showed high encapsulation efficiency

To optimize the encapsulation of the phage cocktail, three different alginate encapsulation methods were examined (Table S5): (a) sodium alginate alone; (b) alginate–calcium carbonate (CaCO₃), a widely used composite of oral drug delivery systems; and (c) alginate–pea protein, where pea protein crosslinking was incorporated to improve the internal environment of the hydrogel beads.

Varying sodium alginate concentrations (1.8 %, 2.0 %, 2.5 %, and 3.0 %) of were examined to identify the optimal gelation percentage. Among these, 2.5 % alginate formed the most structurally stable beads, which exhibited the highest EE% at 51.5 ± 3.0 %.Thus, this concentration was used in the base composition of the phage cocktail for further optimization (Table S5). For the second encapsulation method, 1 % CaCO₃ was incorporated into the sodium alginate solution to create the alginate–CaCO₃ matrix. This optimized formulation of 2.5 % alginate and 1 % CaCO_3, significantly improved EE% to 93.4 % ± 1.4 % (Table S5). In the third approach, CaCO₃ was replaced with pea protein to explore the latter’s potential for creating a compatible and stable matrix for phage delivery. Pea protein not only enhances bead integrity through protein–polysaccharide interactions but may also create a more favorable internal environment for phage stability (Shen, Hong and Li, 2022). This composition yielded an EE% of 86.30 ± 3.75 %, which is slightly lower than the 2.5 % alginate–1 % CaCO₃ composition. However, the alginate–pea protein matrix performed better than the alginate only option (Table S5).

Encapsulation enhanced phage acidic tolerance and maintained high stability at room temperature

To explore the acidic sensitivity of the encapsulated phage cocktail, the three formulations were tested under acidic conditions. At pH 2.5, both the alginate–CaCO₃ and the alginate–pea protein compositions effectively preserved phage viability over the 3-h testing period, with the former demonstrating greater retention of phage viability (Fig. S1).

In contrast, the pure sodium alginate formulation was inactivated within 60 min post-exposure (Fig. S1A). Under pH 2.0, phage viability was preserved only with the alginate–CaCO₃ matrix beyond 120 min post-exposure, whereas in the pure alginate and the alginate–pea protein compositions were completely inactivated by 60 min and 120 min post-exposure to this pH, respectively (Fig. S1B). These results highlight the synergistic protective role of alginate–CaCO₃ in acidic conditions. The optimized formulation (2.5 % alginate and 1 % CaCO₃) not only achieved high EE% (93.4 % ± 1.4 %) but also enhanced phage stability, even at pH 2. Based on these findings, the 2.5 % alginate–1 % CaCO₃ formulation was selected for further use.

In evaluating phage storage conditions, the viability of the non-encapsulated phages remained unchanged after 30 days at 4°C, indicating excellent stability under refrigeration. At 25°C, titers decreased slightly by 0.3 and 0.6 log₁₀ after storage for 14 and 30 days, respectively (Fig. S2). These results demonstrate that encapsulated phages retained high stability under both refrigerated and ambient conditions, providing an adequately long shelf-life and low cold-chain dependency during distribution and application.

Stability and target-release of the encapsulated phage cocktail in an in vitro gastrointestinal model

Further evaluation of the stability and target release of the encapsulated phage cocktail was performed in SGF and SIF to mimic the digestive environment of birds. The alginate-encapsulated phages demonstrated strong acid tolerance under SGF (pH 2) exposure, maintaining phage viability for up to 120 min (Fig. 4A). In contrast, non-encapsulated phages showed high susceptibility to acidic conditions, exhibiting a significant, time-dependent reduction between 30 and 120 min. Thus, encapsulation significantly enhanced the stability of phage.

Fig. 4.

Acid stability and the release of the phage cocktail from alginate–CaCO₃-encapsulated beads in simulated gastric and intestinal fluids. The encapsulated phage cocktail in the form of jelly beads (3 beads) was incubated in (A) 1 mL simulated gastric fluid (SGF; pH 2) or (B) 1 mL simulated intestinal fluid (SIF; pH 8). The plate inset shows a representative image of the encapsulated beads (∼4.0 mm in diameter each). Phage titers were measured at various time points post-incubation. (C) Sequential exposure to SGF and SIF showed the cumulative release of phages from encapsulated beads during the in vitro simulated gut conditions. Data are expressed as the mean ± standard error of the mean of triplicate assays.

Effective intestinal Salmonella control requires rapid phage release for efficient pathogen clearance. To evaluate the target-release potential in the intestinal tract, the encapsulated phages were exposed to SIF at pH 8. The release of free phages was monitored over a of 240-min period. A gradual and sustained release of free phages was observed, with the phage count increasing continuously over time. The release of phages peaked at 120 min and remained stable up to 240 min, at approximately 90.1 % infective viral particles released. In contrast, the non-encapsulated phages maintained a stable number in SIF throughout the observation period, indicating the alkaline tolerance of the naive phages (Fig. 4B).

During sequential exposed to SGF and SIF, phage titers remained relatively stable across both phases, with minimal reduction during model gastric digestion exposure (Fig. S3) and an observed progressive phage number during the intestinal phase (Fig. 4C). At 240 min-post exposure, approximately 85 %–90 % of viable phages had been released from the beads, increasing to 98 %–99 % at 360 min (Fig. 4C). In contrast, the non-encapsulated phages exhibited a 4–5 log₁₀ PFU/mL reduction during the gastric phase (Fig. S3). These findings highlight the dual advantage of encapsulation by enhancing acid resistance in gastric environment and controlling phage release in the alkaline conditions of the intestine where Salmonella are present, thereby supporting targeted delivery within the avian gastrointestinal tract.

Therapeutic efficacy of the encapsulated phage cocktail on Salmonella growth

To identify the effect of the encapsulated phage cocktail on Salmonella growth, culture of S. Typhimurium was spiked into SIF solution and treated with three titers: two single-treatments with a low-treatment group (3 beads) and a medium-treatment (8 beads) administered at 0 h; and a high-sequential treatment of 8 beads at 0 h, followed by an additional 8 beads at 0.5 and 6 h. The low-treatment failed to reduce bacterial levels throughout the 24-h experimental period (Fig. 5A), whereas, the medium-dose treatment achieved complete S. Typhimurium clearance within 3 h post-treatment and prevented bacterial regrowth for 24 h (Fig. 5B). However, although the high-sequential treatment significantly reduced bacterial counts within 2 h, it failed to achieve complete bacterial eradication, with regrowth observed after 24 h (Fig. 5C). Notably, the killing efficiency of the medium-treatment group released phages was comparable to non-encapsulated control (Fig. 5D), indicating that encapsulation does not impair antimicrobial efficacy. Also, the advantage of encapsulation promotes phages survival under acidic conditions prior to reaching the intestinal tract to target colonized Salmonella.

Fig. 5.

Effects of phage treatment on the bactericidal activity of alginate-encapsulated phage cocktail against S. Typhimurium. Simulated intestinal fluid (SIF, 3 mL) spiked with ∼10³ CFU/mL of S. Typhimurium was treated with either single- or sequential-treatment regimens of encapsulated phage beads: (A) Low single application, 3 beads (0.1 g) at 0 h; (B) medium single application, 8 beads (0.3 g) at 0 h; and (C) sequential application, 8 beads (0.3 g) at 0 h followed by an additional 8 beads at 0.5 h and 6 h. A non-encapsulated phage cocktail (10⁹ PFU/mL) served as the positive control (D). Samples were taken over 24 h assayed for bacterial viability (CFU/mL) by colony plating and phage titers (PFU/mL) in the supernatant using the double-agar overlay method. Data are expressed as the mean ± standard error of the mean of three independent experiments, each performed in duplicate.

Medium-treatment induced the rapid release of phages at 2 h, achieving 91.8 % efficiency (9.3 ± 0.05 log₁₀ PFU/mL), which remained stable until 24 h (Fig. 5B). Suboptimal release of free phages was observed with the low- and high-regimen treatments at 2 h post-exposure at 4.9 % and 17.3 %, respectively (Fig. 5A and 5C), indicating underloading or overly dense encapsulation, respectively (Loh et al., 2021; Richards and Malik, 2021b; Yin et al., 2021). Nevertheless, the numbers of free phages remained stable without significant loss after 24 h of exposure (Fig. 5D) for all three phages, indicating adequate survival and stability in the poultry intestinal tract.

Discussion

A major challenge to oral phage therapy in poultry is the acidic gizzard and other gut environments, preventing the optimal and successful targeted delivery of infective viral particles to infection sites (Zhang et al., 2022). The low acidic pH compromises the stability and viability of viral particles, causing the rapid inactivation of therapeutic phages before reaching the distal intestine and ceca, the primary sites of Salmonella colonization (Lorenzo-Rebenaque et al., 2021). The present study demonstrates a strategic approach to circumvent sub-optimal phage delivery in chicken gut using alginate–CaCO₃ encapsulation to optimize a broad-spectrum Salmonella phage cocktail for oral administration in poultry (Colom et al., 2017; Dlamini et al., 2023; Zhou et al., 2022). Encapsulation successfully preserved the phages against acid degradation and to ensured phage release in the ceca and colon, the typical Salmonella colonization sites in the chicken intestine.

The optimized phage cocktail consisted of three strictly lytic phages, namely, ST-W23, SE-W112, and ST-W139, which exhibited broad host range, killing isolates of over 25 Salmonella enterica serovars with the highest lytic efficiency at 98.4 %, ensuring strong therapeutic coverage. The phage cocktail effectively targeted common poultry-associated serovars, such as S. Typhimurium and S. Enteritidis (Koutsoumanis et al., 2019), as well as less-studied and emerging serovars like S. Hadar and S. Virchow, all of which have been implicated in recent outbreaks in chicken products in the USA, Europe, and South Korea (Brandenburg et al., 2024; FDA, 2024; Kang et al., 2024). The broad host range of our phage cocktail against Salmonella demonstrates its strong potential for food safety and global trade applications, industries that increasingly prioritize wide-spectrum, safe therapeutics (Dhulipalla et al., 2025; Wang and Zhao, 2022).

In addition to the broad host specificity, whole-genome sequencing confirmed that all three phages were strictly lytic, lacking any lysogeny, virulence, or antimicrobial resistance genes, indicating their safety and regulatory compliance (Cook and Hynes, 2025). Altogether, these features make the phage cocktail well-suited for therapeutic deployment in poultry production, which has been plagued by infection outbreaks caused by diverse Salmonella serovars (Teklemariam et al., 2023).

Although phage SE-W112 demonstrated greater acid tolerance than ST-W23 and ST-W139, none could withstand prolonged exposure to pH 2, severely limiting their activity and therapeutic efficacy when administered orally (Vinner et al., 2019). The current study established alginate–CaCO₃ encapsulation as the optimal matrix for preserving the Salmonella phage cocktail compared with the pure alginate and alginate–pea protein formulations. The mechanism of alginate–CaCO₃ encapsulation relies on the formation of a semipermeable gel matrix by alginate to entrap the phages (Loh et al., 2021), while the embedded CaCO₃ buffered against gastric acid exposure by neutralizing H⁺ ions (Tan et al., 2022). This dual strategic approach reduced acid damage and inhibited pepsin/bile salt inactivation of the phages to an EE% of 93.4 % ± 1.4 %. These findings highlight its strong potential utility of alginate–CaCO₃ encapsulation as a feasible antimicrobial delivery system in the animal feed industry (Colom et al., 2017; Gomez-Garcia et al., 2021).

Compared with other encapsulation methods, such as alginate–carrageenan (Zhou et al., 2022) and alginate–chitosan beads (Abdelsattar et al., 2019), the alginate–CaCO₃ matrix used in this study showed superior acid protection (up to 3 h at pH 2) and highly effective intestinal release, resulting in complete Salmonella eradication within 24 h. These findings corroborate those of previous studies demonstrating the efficacy of alginate–CaCO₃ encapsulated Salmonella phages (UAB_Phi20, UAB_Phi78 and UAB_Phi87) in maintaining viability in SGF at pH 2.8 and exhibiting more potent antibacterial activity than free phages (Colom et al., 2017). In contrast, liposome-encapsulated phages showed reduced acid stability (Colom et al., 2015), while alginate–carrageenan microcapsules showed variable efficacy at reducing Salmonella counts (Zhou et al., 2022).

The in vitro intestinal model demonstrated that the single application of 8 beads of alginate–CaCO₃-encapsulated phages (approximately MOI 10⁵–10⁶) completely eradicated of S. Typhimurium below detectable levels (<1 CFU/100 µL) within 3 h after treatment and prevented bacterial regrowth for at least 24 h. This strong and sustained effect correlated with rapid phage release from the beads (>91.8 % within 2 h) (Bernela et al., 2025; Dlamini et al., 2023). Similarly, non-encapsulated phages also achieved complete lysis within 4 h and maintained bacterial suppression.

Interestingly, the two-treatment regimen showed lower efficacy, likely due to delayed phage release, which resulted in a suboptimal active viral particles. This reduced efficacy may be attributed to phage interference, where initial phage replication limits subsequent phage propagation, as had been previously demonstrated for sequential phage applications against Pseudomonas aeruginosa (Hall et al., 2012). Given the controlled-release properties of the formulation, the initial phage treatment likely delivered a sufficient number of active phages for effective bacterial clearance, reducing the need for a second phage application (Youssef et al., 2025). In addition, repeated phage administration can lead to rapid immune-mediated clearance (e.g., neutralizing antibodies), limiting efficacy (Berkson et al., 2024; Tan et al., 2024). These results indicate not only the greater efficacy, and cost-effectiveness of the optimized treatment (including fewer or single application) but also its improved safety and feasibility in practical settings (Cui et al., 2024). Although the two-treatment regimen in our study showed greater regrowth at 24 h compared to the single treatment, actual phage efficacy may relies on synergy with host immunity (Roach et al., 2017), where early bacterial reduction may support immune-mediated clearance (Zborowsky et al., 2025). Further in vivo studies are needed to assess the practical relevance of late-stage regrowth.

Notably, our encapsulated phage cocktail achieved superior bacterial eradication compared with similar previous studies which reported bacterial regrowth (Henderson et al., 2019; Kawacka et al., 2020; Martinho et al., 2024). This enhanced efficacy may reflect the combined virulence and complementary properties of the component phages and localized high phage release from the alginate–CaCO₃ matrix, which aligns with earlier reports on sustained E. coli O157:H7 suppression in rats (Yin et al., 2021). Nonetheless, further studies are needed to further confirm the synergistic properties of the phages in the cocktail examined here.

Furthermore, the encapsulated phage cocktail remained highly stable at room temperature for 30 days, indicating a stable shelf-life and viability for field use without requiring cold-chain logistics requirements. Although initially designed for oral delivery, the tested formulation shows strong potential for food applications. Its acid resistance, controlled release, and broad lytic range make it suitable for post-harvest applications or as a biocontrol agent in food packaging (Alves et al., 2019) to reduce Salmonella contamination during processing, storage, and distribution.

Conclusions

This study presents a promising oral delivery strategy for an optimized phage cocktail using alginate–CaCO₃ encapsulation. The encapsulated phage cocktail comprised of three strictly lytic phages (ST-W23, ST-W139 and SE-W112) which were isolated from wastewater and are active against 25 poultry-associated Salmonella serovars. The encapsulated phages were shown to be protected against acid degradation while maintaining sustained release in alkaline condition. Further testing in a simulated intestinal model revealed that the encapsulated phage cocktail completely eradicated S. Typhimurium within 3 h of treatment and prevented bacterial regrowth for 24 h signifying suppression of phage-resistant strains. The encapsulated cocktail is stable under room temperatures (25°C) for a month. These findings highlight the potential of this encapsulated phage cocktail as a practical antimicrobial feed additive for pre-harvest Salmonella control in poultry. It also has a strong potential post-harvest intervention to ensure safety of chicken meat products. Further in vivo and scale-up studies are needed to confirm efficacy in farms and under commercial conditions.

Funding

This study received funding from the Agricultural Research Development Agency (ARDA), Thailand (grant number CRP6405031030).

Ethics statement

All experiments conducted were approved by the Siriraj Safety Risk Management Taskforce of Mahidol University (approval number SI2021-020).

CRediT authorship contribution statement

Siwaporn Prachoochote: Writing – original draft, Investigation, Formal analysis, Data curation. Sujintana Janesomboon: Project administration, Investigation, Formal analysis, Data curation. Preeda Phothaworn: Writing – review & editing, Investigation. Patoo Withatanung: Writing – review & editing, Formal analysis. Janet Y. Nale: Writing – review & editing, Investigation. Edouard E. Galyov: Writing – review & editing. Martha R.J. Clokie: Writing – review & editing, Supervision. Sunee Korbsrisate: Writing – review & editing, Supervision, Funding acquisition, Conceptualization. Veerachat Muangsombut: Writing – review & editing, Writing – original draft, Methodology, Formal analysis, Data curation, Conceptualization.

Disclosures

Sunee Korbsrisate received financial support from the Agricultural Research Development Agency, Thailand. The other authors have no known competing financial interests or personal relationships to declare that could have appeared to influence the work reported in this paper.