The type strain of K. pneumoniae subsp. pneumoniae KCTC 12385^T was used as the host bacterium for bacteriophage isolation. For the host range assay, 21 bacterial strains were tested. These included K. pneumoniae subspecies, Klebsiella alba, Klebsiella michiganensis, Raoultella ornithinolytica, Raoultella planticola, Acinetobacter baumannii, Escherichia coli, Pseudomonas aeruginosa, and Staphylococcus aureus. Some of the biological resources used in this research were obtained from the Korean Collection for Type Cultures (KCTC) or the Korean Agricultural Culture Collection (KACC), and the remaining strains were isolated from various samples (Table 1). Bacterial species were identified through 16S rRNA gene sequencing, with sequences compared against the EzBioCloud database (Chalita et al., 2024). The genes were sequenced using the bacterial-specific universal primers 27F, 337F, 785F, 800R, and 1492R. The sequenced fragments were assembled to full-length 16 rRNA gene sequences with SeqMan 5.0 program (DNASTAR) (Swindell and Plasterer, 1997). To identify the relationship between the used bacterial species, a phylogenetic tree was constructed using the sequences. The sequences were aligned by CLUSTRAL W software in BioEdit v5.0.9 (Thompson et al., 1994). Neighbor-joining (NJ), maximum-likelihood (ML), and maximum-parsimony (MP) algorithms were used with 1,000 bootstrap replications for tree construction using the MEGA v7.0.26 program (Kumar et al., 2016). Also, for better understanding of the taxonomical positioning of the genus Raoultella, a core bacterial gene set was extracted from the reference genomes of the species from the genera Klebsiella and Raoultella. The genomes were acquired from NCBI and reconstructed to a phylogenetic tree by UBCG (Up-to-date Bacterial Core Genes) v3.0 software (Na et al., 2018). Bacterial cultures were incubated in liquid Luria-Bertani (LB, BD Difco) medium or on solid LB medium containing 1.5% (w/v) agar at 37°C in an incubator.

Phage K14-2 was isolated from a water sample collected from the Jungnangcheon River (37.553267, 127.043876) located in Seoul, Republic of Korea. For phage isolation, the sample was centrifuged at 8,000 × g for 5 min at 4°C and then filtered through 0.45 and 0.22 μm PVDF syringe filters (GVS) to remove sediments and bacterial cells. Subsequently, the double-layer agar (DLA) method was used to form and isolate plaques. Briefly, 10 μl of the filtrate and 500 μl of an exponential phase of K. pneumoniae KCTC 12385^T culture were mixed in a 1.5 ml tube. The mixture was incubated at 37°C for 10 min to allow phage adsorption to the bacterial cells. After incubation, the mixture was added to 4.5 ml of 0.5% soft agar LB medium and overlaid onto a 1.5% hard LB agar plate. Both the soft and hard agars were supplemented with 5 mM CaCl₂. After overnight incubation at 37°C, the resulting plaques were isolated using sterile pipette tips and stored in SM buffer (100 mM NaCl, 8 mM MgSO₄⋅7H₂O, 50 mM Tris-Cl, pH 7.5). Three consecutive isolation procedures were performed to ensure the pure isolation of a single phage plaque. The isolated phage was stored at 4°C for future studies.

To establish optimal conditions for further experiments, the MOI between K. pneumoniae KCTC 12385^T and phage K14-2 was determined. MOI refers to the ratio of plaque-forming units (PFUs) to colony-forming units (CFUs). Phage and bacterial titers were obtained using the DLA method and colony counting method, respectively. Liquid host cultures were mixed with the appropriate concentration of the phage to achieve MOIs of 100, 10, 1, 0.1, 0.01, and 0.001 (Table 2). The mixed cultures were incubated at 37°C for 10 min to allow phage adsorption. The optimal MOI was determined based on the highest phage titer released from the mixture.

Additionally, a killing curve assay was conducted to assess the bacterial inhibition effect of phage K14-2. The method used was described by Das and Kaledhonkar (2024). A 100 μl aliquot of freshly incubated host bacterial culture and 100 μl of phage were mixed to achieve an MOI of 0.1 in 96-well plates. LB broth and uninfected host cultures were included as blank and control, respectively, then the plates were incubated at 37°C. The optical density of the bacterial cell was measured at 600 nm every 30 min for 300 min using a Synergy MX spectrophotometer (BioTek).

A one-step growth curve was plotted to determine the incubation period and burst size of phage K14-2, following the method described by Kropinski (2018) with minor adjustments. Briefly, mid-log phase bacteria (OD₆₀₀ = 0.4) in liquid culture were mixed with 1 × 10⁷ PFU/ml of phage and incubated at 37°C for 10 min for phage adhesion. The mixture was serially transferred to three flasks containing LB broth, resulting in 10^–2, 10^–3, and 10^–4 dilutions. The flasks were incubated in a 37°C water bath. Every 10 min, 0.1 ml of the cultures was removed from each flask, added to soft agar, and then overlaid onto hard agar plates. The plates were incubated overnight at 37°C, and the number of plaques was counted to plot the growth curve.

The sensitivity of phage K14-2 to thermal and pH stresses was evaluated using the method described by Das and Kaledhonkar (2024). For thermal stability, phage suspensions at a concentration of 1 × 10¹⁰ PFU/ml were incubated in a water bath at 40, 50, 60, 70, and 80°C for 60 min. Phage titers were then determined using the DLA method. Similarly, pH stability was assessed by incubating phage suspensions of the same titer in SM buffer adjusted to pH levels ranging from 1 to 14 for 60 min, followed by their measurement.

The host range of phage K14-2 was assessed using a spot test with 21 bacterial strains listed in Table 1. For the spot test, bacterial cultures were mixed with soft agar and overlaid onto agar plates to solidify. Subsequently, 10 μl of phage suspensions were spotted on to the surface of the plates. After overnight incubation at 37°C, plaque formation was observed. The plaques were scored according to the system described by Kutter (2009), as follows: 4, complete clearing; 3, clearing with a faintly hazy background; 2, turbidity throughout the cleared zone; 1, a few individual plaques; 0, no plaque formation. The host range assay was repeated three times for each bacterial strain tested.

The morphological characteristics of phage K14-2 were observed using a transmission electron microscope (TEM). A 10 μl drop of 1 × 10¹⁰ PFU/ml purified phage suspension was placed on a Formvar film (EMS) and stained with 2% (w/v) phosphotungstic acid (PTA) for 5 min. The stained film was observed with JEM1010 TEM (JEOL) at an accelerating voltage of 80 kV.

Genomic DNA of phage K14-2 was extracted from a 1 × 10¹⁰ PFU/ml purified phage suspension using the DNeasy Blood and Tissue Kit (Qiagen) with additional procedures. The phage suspension was first mixed with RDD buffer, DNase I, and RNase A, and incubated at 37°C for 90 min to eliminate external DNA or RNA. Following this, 0.5M EDTA and Proteinase K were added, and the mixture was incubated for an additional 90 min at 56°C. The remaining steps followed the manufacturer’s protocol. For library preparation, the Illumina TruSeq Nano DNA (350) library was utilized. Genome sequencing was performed using the Illumina NovaSeqX platform (Illumina). Prior to assembly, the quality of the raw sequences was assessed using FastQC v0.11.7 software,¹ and adaptor sequences were removed with Trimmomatic v0.38 (Bolger et al., 2014). The whole phage genome was assembled using SPAdes v3.15.0 (Bankevich et al., 2012).

The annotation of the K14-2 phage whole genome was performed using Prokka v1.14.6 (Seemann, 2014) and the Rapid Annotation using Subsystem Technology (RAST) server² (Brettin et al., 2015). To identify closely related Klebsiella phages, a BLASTn³ (Altschul et al., 1997) search against the NCBI database was utilized. The whole genome map was constructed using the CGView Server.⁴ Potential tRNA genes within the phage genome were identified using tRNAscan-SE (Chan and Lowe, 2019). Two similar phage genome sequences were analyzed for linear comparison with Easyfig program (Sullivan et al., 2011).

To construct a phylogenetic tree, amino acid sequences of the major capsid protein (MCP) from closely related phages were retrieved using the BLASTp program with the NCBI database. These sequences were then aligned using the CLUSTAL W program in BioEdit v5.0.9 software. The aligned sequences were subjected to phylogenetic tree reconstruction using the MEGA v7.0.26 program, employing the NJ algorithm with 1,000 bootstrap replications.

All data presented in the graphs were generated using GraphPad Prism v8.4.3 software. Error bars represent the standard error of the mean.

A novel lytic Klebsiella phage, K14-2, was isolated from a river water sample using K. pneumoniae as the host bacterium. After incubating the sample for 24 h at 37°C, the plaque morphology was observed, revealing clear, round plaques with diameters of 1–2 mm (Figure 1A). The morphology of the phage virion was visualized using TEM. The phage exhibited an icosahedral head (90.9 ± 3.8 nm), a long sheath (84.0 ± 0.9 nm), and a contractile tail (Figure 1B). The morphological characteristics indicate that the phage resembles features typical of myovirus.

The optimal MOI for phage K14-2 was determined to identify the most effective ratio of host to phage for infection. An MOI of 0.1, which resulted in the highest phage titer, was established as the optimal ratio (Table 2). The lytic activity of phage K14-2 was assessed using a killing curve assay (Supplementary Figure 1). The result showed that the phage inhibited the growth of the host bacteria and maintained the host concentration for 5 h, indicating that the growth of the K. pneumoniae was effectively inhibited, maintaining levels close to the initial state. Subsequently, a one-step growth curve was generated to determine the burst size and latent period of phage infection. The curve indicated a latent period was approximately 20 min and an outbreak period of approximately 50 min. The burst size of the phage was estimated to be an average of 32.9 PFU/cell, with a standard deviation of 13.3 (Figure 2).

Phages used in phage therapy should possess not only a broad host range but also the ability to maintain activity under diverse environmental conditions. Specifically, phage therapy intended for humans must be capable of withstanding fluctuations in body temperature and pH levels. Therefore, stability of phage K14-2 with respect to environmental factors, including temperature and pH, was evaluated. The temperature stability assay examined the phage’s resilience at temperatures of 30, 40, 50, 60, 70, and 80°C for 60 min. The assay revealed that the phage was stable at temperatures ranging from 30 to 70°C; however, its lytic activity significantly decreased at 70°C and was completely lost at 80°C (Figure 3A). The pH stability assay assessed the phage’s resilience across a pH range of 1 to 14 for 60 min. The results indicated that the phage retained lytic activity between pH 3 and 11, whereas no activity was observed at pH 1, 2, 12, 13, and 14 (Figure 3B). Collectively, similar to other Klebsiella phages, K14-2 was able to maintain activity between 30 and 70°C and pH 3–11 for 60 min. These findings suggest that K14-2 can be used in various environmental conditions, including those found in the human body. However, the assay was limited to a 60 min short incubation time, and further experiments are needed to assess its ability over longer durations.

Before the host range test, the relationships between the host bacterial species were identified using phylogenetic trees reconstructed with 16S rRNA gene sequences and bacterial core gene set extracted from the NCBI reference genomes (Supplementary Figures 2, 3). The resulting trees showed that the species of genus Raoultella clustered within the genus Klebsiella, indicating a close relationship between two genera. Thus, the host range of phage K14-2 was tested against 10 species and a total of 21 different bacterial strains, as detailed in Table 1. The results demonstrated that phage K14-2 show clearings to various species within the Klebsiella and Raoultella genera, including K. pneumoniae subsp. pneumoniae, K. pneumoniae subsp. ozaenae, K. alba, K. michiganensis, R. ornithinolytica, and R. planticola. Additionally, E. coli, another member of the Enterobacteriaceae family, exhibited cleared zone with turbidity. In contrast, other ESKAPE pathogens such as A. baumannii, P. aeruginosa, and S. aureus showed no lysis activity.

The complete genome of phage K14-2 was circular double-stranded DNA (dsDNA) molecule comprising 175,759 bp with a GC content of 41.8% (Figure 4). The annotated genome revealed 280 predicted protein-coding genes, including 184 hypothetical proteins. Of these, 96 functional protein-coding genes were classified into three distinct functional modules. Specifically, 41 genes were associated with structural proteins and the packaging system, 54 genes with replication and regulation, and 1 gene with lysis. The structural proteins predominantly included genes encoding the phage capsid, baseplate, sheath, and tail encoding genes. Genes related to replication and regulation encode proteins such as HNH endonuclease motif harboring protein, ligases, DNA polymerases, and helicases. The search for RNA genes within the phage genome revealed two tRNA genes. One of these tRNA genes encodes methionine and was conserved in 72 other phages with 100% similarity. The other tRNA gene is a pseudo type tRNA gene encoding leucine with an anticodon of CAA. Similar sequences of this predicted pseudogene were found in 73 phages, predominantly Klebsiella phages. The gene responsible for host lysis was annotated as holin, a gene commonly found in dsDNA phages. Holin encodes pore-forming proteins that create holes in the bacterial cell wall, releasing the assembled virions at the end of the lytic cycle. In contrast, genes associated with the lysogenic cycle, such as integrase, were not identified, indicating that phage K14-2 is a lytic phage. The activity of the lytic gene was further confirmed by the formation of the transparent plaques.

Homologous sequences of the phage K14-2 genome were identified using the BLASTn program. The results showed that the phage genome is highly similar to other K. pneumoniae phages within the Slopekvirus genus. The genome sequence with the highest similarity was that of Klebsiella phage vB_KpM-Milk, showing 98.66% similarity and 96% query coverage. According to the species demarcation criteria proposed by the Bacterial Viruses Subcommittee of International Committee on Taxonomy of Viruses (ICTV), novel bacteriophage species should have a similarity lower than 95% (Turner et al., 2021). Using the BLASTn program, the similarity value can be calculated by multiplying the percentage identity by the percentage coverage. Thus, the similarity value between phage K14-2 and vB_KpM-Milk is 94.71%, indicating that phage K14-2 represents a novel species within the Slopekvirus genus. Linear genome comparison was also performed with K14-2, vB_KpM-Milk and RCIP0008 phage genomes, showing that K14-2 was highly homologous with other two phages (Figure 5).

A phylogenetic tree was constructed using the MCP amino acid sequences to identify the evolutionary relationship of phage K14-2 with closely related species (Figure 6). The tree confirmed that phage K14-2 clusters with Klebsiella phages within the genus Slopekvirus of the family Straboviridae.