Genomic characterization of novel lytic bacteriophages targeting carbapenemase producing multidrug resistant Klebsiella pneumoniae from Ethiopia: a molecular study

Abstract

Klebsiella pneumoniae (K. pneumoniae) is a critical pathogen responsible for a wide range of severe infections. Its escalating resistance to frontline antimicrobials, particularly carbapenems, severely compromises existing therapeutic options and undermines current treatment strategies. This clinical challenge is further exacerbated by the emergence of multidrug-resistant (MDR) strains and hypervirulent lineages, which substantially complicate infection management and worsen patient outcomes. Ultimately, this alarming convergence of resistance and virulence presents a severe public health threat that necessitates the urgent development of novel therapeutic approaches; in this regard, phage therapy is emerging as a promising strategy. Bacteriophages and their derivatives have garnered increasing recognition as targeted antibacterial agents, demonstrating potent activity against carbapenemase producing and MDR K. pneumoniae in vitro models. They effectively lyse MDR strains, disrupt biofilm formation, and have shown promising therapeutic efficacy across various in vivo models. To ensure safety in clinical applications, genome sequencing is essential to identify and exclude undesirable genes, thereby determining therapeutic suitability. Here, we have sequenced and analyzed the genomic characteristics of seven lytic bacteriophages targeting carbapenemase-producing MDR K. pneumoniae. Genomic evaluation confirmed that these phages are strictly lytic and revealed the presence of distinct lysis modules across the genomes. Phylogenetic analysis showed that these phages cluster within four genera of Caudoviricetes: Taipeivirus, Drulisvirus, Przondovirus, and Webervirus. The phage genomes ranged from 15,773 to 166,437 base pairs (bp) in length and encoded core structural and replication modules. Furthermore, distinct lytic modules predicted to have function in bacterial capsule degradation were identified across the genomes. Importantly, the absence of known virulence or toxin genes guarantees therapeutic safety and satisfies rigorous genomic criteria for clinical translation. These findings underscore the potential of these phages as a promising alternative for combating MDR and hypervirulent K. pneumoniae infections. Future in vivo investigations and clinical trials are essential to validate their therapeutic efficacy, safety profile, and optimal treatment protocols.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12866-026-05410-z.

Background

Antimicrobial resistance (AMR) represents one of the most pressing global public health threats, rendering currently available antibiotics ineffective [1, 2]. ESKAPE pathogens epitomize this crisis, as they are the most formidable multidrug-resistant (MDR) bacteria, rendering currently available antibiotics ineffective. Among these, K. pneumoniae emerges as a leading cause of MDR infections. K. pneumoniae is an opportunistic Gram-negative encapsulated bacillus and a key member of the ESKAPE group of critical MDR pathogens, posing a major global public health threat [3, 4]. K. pneumoniae commonly found in the gastrointestinal tracts, upper respiratory tracts, skin of humans, livestock, and wildlife, as well as environmental niches like soil, water, plants, and sewage, this clinically important pathogen causes a diverse range of severe infections including pneumonia, urinary tract infections (UTIs), bloodstream infections (BSIs), septicemia, and wound infections [5, 6]. Its remarkable capacity to acquire and disseminate antimicrobial resistance (AMR)-driven by mechanisms such as production of antibiotic-inactivating enzymes (e.g., beta-lactamases and carbapenemases), efflux pumps, and plasmid-mediated gene transfer-has transformed this treatable commensal into a deadly pathogen [7]. This renders last-resort antibiotics, such as carbapenems, ineffective. The emergence of hypervirulent strains further exacerbates this threat, significantly increasing the challenge of effective treatment [8].

The incidence of MDR K. pneumoniae strains continues to rise alarmingly. Although multiple molecular pathways underpin this resistance, the production of extended-spectrum β-lactamases (ESBLs) and carbapenemases remains the dominant mechanism driving therapeutic failure in this pathogen [7, 9, 10]. Notably, K. pneumoniae strains producing ESBLs and carbapenemases pose major clinical challenges due to their extreme resistance profiles and severely limited treatment options [11–13]. This bacterium also poses significant pathogenicity concerns, driven by the synergistic action of multiple virulence factors that promote host colonization, immune evasion, and systemic dissemination. Key contributors include lipopolysaccharides, iron acquisition systems, polysaccharide capsules, mucoviscosity-associated exopolysaccharides, and adhesins.

Evidence shows that K. pneumoniae isolates are increasingly exhibiting MDR profiles, accompanied by rising global rates of carbapenem resistance [7, 14]. This convergence of escalating virulence and resistance to existing antibiotics underscores the urgent need for novel therapeutic strategies [15, 16]. The World Health Organization (WHO) has classified carbapenem resistant K. pneumoniae (CRKP) as a pathogen of top priority. This designation is attributed to its significant resistance to a broad spectrum of antibiotics, its prevalence in both healthcare settings and community environments, and the limited treatment options available for infections caused by this organism [14, 16, 17].

Alongside ESβ-producing strains, CRKP poses considerable challenges in clinical environments worldwide, contributing to heightened morbidity, mortality, and healthcare expenditures. The WHO’s 2024 Bacterial Priority Pathogens List highlights carbapenem-resistant K. pneumoniae as the most critical pathogen, emphasizing the urgent need for new therapeutic strategies to address this threat [18–20]. In this regard, phages have shown promising therapeutic potential. Recent research indicates that phages can effectively target MDR K. pneumoniae, including strains resistant to carbapenems and those producing carbapenemases and ESβLs [16, 21]. This growing body of evidence highlights phage therapy as a valuable tool for managing infections caused by these challenging pathogens [11–13]. Numerous studies have demonstrated that bacteriophages can effectively target multidrug-resistant K. pneumoniae strains in both in vitro and in vivo contexts [15, 22–24].

Whole-genome sequencing ensures the safety and efficacy of bacteriophages for clinical applications by verifying lytic cycles, identifying key host-targeting proteins, excluding undesirable genes, and facilitating taxonomic classification. This comprehensive genomic characterization ensures safety and therapeutic suitability and meets regulatory standards for therapeutic use. In this study, we present the whole-genome sequences and detailed functional annotations of lytic bacteriophages isolated from diverse environmental sources in Ethiopia. These phages specifically target multidrug-resistant K. pneumoniae, providing essential molecular data for the development of locally optimized phage therapies to address the antimicrobial resistance crisis in Ethiopia.

Materials and methods

Bacteriophage isolates and host strains

Lytic bacteriophages targeting K. pneumoniae were isolated from wastewater, hospital sewage, and soil samples (Table 1). These phages exhibited significant in vitro lytic activity and therapeutic efficacy against MDR strains of K. pneumoniae. A comprehensive biological characterization, encompassing molecular identification, isolation, and classification, is provided in our preceding studies [25]. The molecular characterization and AMR profile of the bacterial isolates used in this investigation are described in [9].

Table 1.

Isolation sources and characteristics of study bacteriophages

S.no	Phage isolate	Isolation source	PCR−classified phage genus	Host	Host Virulence and AMR gene profile	Host AMR Profile
1	KFM-TASP06	River	Taipeivirus	TASP06	blaCTX−M, blaNDM, blaSHV, blaTEM, MrkA, MdtK, gyrA, ArmA	MDR
2	TTM-TASP117	Sewage	Taipeivirus	T ASP117	blaCTX−M, blaNDM, blaSHV, MdtK, gyrA	MDR
3	ZHS-TASP18	Soil	Drulisvirus	TASP18	blaNDM, blaSHV, blaTEM, MrkA, gyrA, ArmA	MDR
4	AKM-TASP09	River	Webervirus	TASP09	blaCTX−M, blaSHV, MrkA, mrkD, mdtK, rmpA	MDR
5	BYM-TASP37	River	Drulisvirus	TASP37	blaCTX−M, blaNDM, blaSHV, gyrA, rmpA	MDR
6	GKM-TASP92	River	Taipeivirus	TASP92	blaCTX−M, blaNDM, blaSHV, FimH	MDR
7	MHM-TASP32	Sewage	Przondovirus	TASP32	MrkA, mrkD, mdtK, blaCTX−M, blaNDM, blaSHV	MDR

Phage genome extraction and quality analysis

Genomic DNA was extracted following a previously established protocol [22]. Briefly, the phage lysates were treated with DNase and RNase (1 h, 37 °C) to degrade extracellular nucleic acids, followed by enzyme inactivation with 20 mM EDTA. Proteinase K digestion (3 mg/mL) was performed at 56 °C for 90 min, followed by phage DNA purification using the EZ-10 Spin Column Bacterial Genomic DNA Miniprep Kit (Omega Bio-tek, USA). The concentration and purity of the extracted DNA were assessed using a Nanodrop 2000/2000 C Spectrophotometer (Thermo Scientific™, USA), and its integrity was confirmed by running a 1% agarose gel electrophoresis.

Whole genome sequencing and bioinformatics analysis

DNA libraries were prepared from purified phage DNA using the Nextera XT DNA Library Prep Kit (Illumina) following the manufacturer’s instructions. Sequencing was conducted in paired-end mode (2 × 150 bp) on the Illumina NextSeq 550 platform at the Armauer Hansen Research Institute (AHRI) in Addis Ababa. The quality of raw sequencing reads was assessed using FastQC (v0.12.0) [26], and low-quality bases and adapters were trimmed using fastp (v0.23.0) [27, 28]. The quality of raw sequencing reads was assessed using FastQC (v0.12.0), and low-quality bases and adapters were trimmed using fastp (v0.23.0) [27]. Taxonomic classification was performed using Kraken2 (v2.1.2) [29] to categorize reads at the genus and species levels and to filter out potential contaminants. De novo assembly was performed with MEGAHIT (v1.2.9) [30], and contigs from each sample were scaffolded to improve assembly quality. QUAST (V5.0.2) [31] was used for assembly quality assessment, while CheckV (v1.5.0) [32] was used to evaluate genome completeness. Phage-specific downstream bioinformatics analysis, including completeness assessments, lifestyle predictions, structural gene annotations, and taxonomic classification, was performed using PhageScope [33], which provides comprehensive annotations including completeness assessments, phenotypic annotations, structural annotations, taxonomy, and functional annotations. The nucleotide sequences have been deposited in the National Center for Biotechnology Information (NCBI) database and are available under the accession numbers provided.

Molecular phylogenetics

Phylogenetic analysis was performed using the nucleotide sequences of the major capsid protein (MCP) and the terminase large subunit (TerL) to evaluate the evolutionary relationships among the bacteriophages [34]. Additional sequences, including isolates from various countries, were retrieved from NCBI. For each query, MCP and TerL sequences with the best hits (e-value near zero and ≥ 90% sequence identity) were downloaded. To avoid geographic bias, no more than four phages per country were included when available. Sequences were aligned using the ClustalW algorithm in MEGA 12 [35]. Phylogenetic trees were constructed using the Maximum Likelihood method in MEGA 12, employing the Tamura-Nei substitution model to accurately estimate evolutionary distances by accounting for multiple substitutions and varying transition/transversion rates. MCP sequences were employed as the primary phylogenetic marker for six phages. For one phage lacking adequate MCP sequence data, the terminase large subunit (TerL) sequence was used as an alternative. The resulting phylogenetic trees were visualized and annotated with metadata using iTOL (V.6) [36].

Results

Genome sequencing and bioinformatics analysis

Seven phages were sequenced for further characterization. The quality of the sequencing data was assessed and verified using FastQC to ensure the suitability of reads for downstream analysis. A total of 829,073 sequence reads were generated, with GC contents varying from 41 to 51%. Taxonomic classification using Kraken2 confirmed the presence of the target phage sequences while removing contaminant reads. The complete genome sequences of all phages were deposited in NCBI GenBank under accession numbers PV588648, PV588649, PV588650, PV588651, PV588652, PV588653, and PV588654. All the sequenced bacteriophages had a linear double-stranded DNA genome (Table 2). The bacteriophages belong to the genera Taipeivirus, Drulisvirus, Przondovirus, and Webervirus (Table 2). The genomes of the bacteriophages ranged in size from 15,773 to 166,437 bp, and the G + C content ranged from 37.71 to 52.85% (Table 2).

Table 2.

Summary of seven bacteriophages sequenced in this study

Phage	Gene bank accession number	Genome size (bp)		Genes	GC content (%)	Genus	Family	Life style	Taxonomy
KFM-TASP06	PV588648		166,437	225	44.25	Taipeivirus	Ackermannviridae	Virulent	Caudoviricetes
TTM-TASP117	PV588649		167,792	227	44.87	Taipeivirus	Ackermannviridae	Virulent	Caudoviricetes
ZHS-TASP18	PV588650		47,475	57	52.85	Drulisvirus	Ackermannviridae	Virulent	Caudoviricetes
AKM-TASP09	PV588651		15,773	33	51.97	Webervirus	Drexlerviridae	Virulent	Caudoviricetes
BYM-TASP37	PV588652		17,666	26	44.50	Drulisvirus	Ackermannviridae	Virulent	Caudoviricetes
GKM-TASP92	PV588653		68,457	116	37.71	Taipeivirus	Ackermannviridae	Virulent	Caudoviricetes
MHM-TASP32	PV588654		38,888	51	52.78	Przondovirus	Studiervirinae	Virulent	Caudoviricetes

Molecular phylogeny

Phylogenetic relationships of the isolated phages were inferred using the major capsid protein (MCP) gene sequences for six phages, while the large terminase subunit gene was used for one phage, AKM-TASP09 (PV588651.1). The MCP-based phylogenetic analysis for six phages showed that phage BYM-TASP37 (PV588652.1) clusters closely with K. pneumoniae phage Kpn BHU3 (OL926437.1) and other members of the genus Drulisvirus (Fig. 1).

Fig. 1.

Phylogenetic analysis of six phages based on major capsid protein nucleotide sequences. The figure illustrates the phylogenetic relationships among the phages KFM-TASP06 (PV588648.1), TTM-TASP117 (PV588649.1), GKM-TASP92 (PV588653.1), ZHM-TASP18 (PV588650.1), and BYM-TASP73 (PV588652.1), based on nucleotide sequences of their major capsid proteins. The concentric rings indicate the following classifications: the inner colored ring represents phage genus classification; the middle ring indicates the country of origin; and the outer ring shows the sample source

Phage ZHM-TASP50 (PV588650.1) grouped with K. pneumoniae phage VLCpiA1a (ON602732.1). Phages KFM-TASP06 (PV588648.1) and TTM-TASP117 (PV588649.1) formed a distinct cluster with K. pneumoniae phage Madison (PQ368523.1) (Fig. 1), with their MCP sequences exhibiting 100% nucleotide identity. Further BLAST analysis of phage KFM-TASP06 and TTM-TASP117 resulted in 99% identity and 96% coverage. Phylogenetic analysis of phage AKM-TASP09 was performed using the large terminase subunit gene. This analysis revealed its closest evolutionary relationship with Klebsiella phages vB_KpnS (NC_049836.1) and mfs (OP611406.1) (Fig. 2).

Fig. 2.

Phylogenetic analysis of phage AKM-TASP09 (PV588651.1) based on TerL

Genome annotation

Comprehensive annotation through PhageScope provided detailed functional categorization of predicted genes (Fig. 3). All phage genomes contained multiple proteins essential for various stages of the viral life cycle, including those involved in infection, assembly, replication, packaging, lysis, regulation, and several proteins with hypothetical functions. While the number and identities of these transmembranes differed among phages, several membrane-associated hypothetical proteins were identified. All phages encoded a range of infection-related proteins, particularly tail and tail spike proteins, which are essential for host recognition and adsorption. Structural components, such as capsid proteins and head scaffolding proteins, were universally present. None of the predicted open reading frames (ORFs) encode lysogenic phage-related proteins, such as transposases or integrases. Additionally, no genes associated with antibiotic resistance or bacterial virulence were detected.

Fig. 3.

Overview of genome annotation for seven sequenced bacteriophages

Hierarchical clustering of predicted protein sequences from seven bacteriophages revealed several conserved functional modules across the genomes. Proteins were categorized into groups such as lysis, integration, replication, tRNA-related functions, regulation, packaging, assembly, infection, immune response, and hypothetical proteins based on sequence similarity (Fig. 4). Canonical lysis modules, including lysozymes/endolysins and holins, were detected in dedicated clusters, with most phages exhibiting co-localization of these genes along with spanins, indicative of holin-lysozyme cassettes. Structural divergences were primarily observed in tail fiber proteins, while core structural components showed high conservation (supplementary file 2).

Fig. 4.

Heatmap showing hierarchical clustering of predicted protein sequences from seven bacteriophages, grouped by functional modules

The annotated proteins were grouped into 10 functional categories: lysis, integration, replication, tRNA-related, regulation, packaging, assembly, infection, immune, and hypothetical, as indicated in Fig. 5 for phage ZHS-TASP18. Genome maps for the other six phages are provided in Supplementary file 1.

Fig. 5.

Annotated genome map of bacteriophage ZHS-TASP18

The Figure highlights the organization and distribution of protein-coding genes, categorized by their predicted roles in the phage lifecycle. The innermost rings indicate genomic coordinates, followed by GC content represented as a blue line graph. GC skew is shown with positive values in red and negative values in turquoise blue. The outermost ring depicts the identified open reading frames (ORFs). The genomes were analysed for proteins predicted to contain transmembrane helices (TMHs), revealing considerable variation among the phages. The number of proteins with transmembrane helices varied widely, from a single TMH protein annotated for BYM-TASP37 to up to 22 in KFM-TASP06. Notably, the proteins containing transmembrane helices were primarily holins and other membrane-associated proteins, as well as several hypothetical proteins. Circular genome maps of six phages, highlighting the locations of these TMH-containing proteins, can be found in Fig. 6.

Fig. 6.

Annotated genome map of seven bacteriophages with transmembrane proteins. A phage AKM-TASP09, (B) phage KFM-TASP06, (C) phage GKM-TASP92, (D) phage MHM-TASP32, € phage TTM-TASP117, (F) phage ZHS-TASP18, and (G) BYM-TASP37. The innermost rings indicate genomic coordinates, followed by GC content represented as a blue line graph. GC skew is shown with positive values in red and negative values in turquoise blue. The outermost ring shows proteins identified with transmembrane proteins.

Analysis of transcriptional terminators across the seven sequenced phage genomes revealed notable variation in their abundance (Fig. 7). Phages TTM-TASP117 and KFM-TASP06 exhibited the highest number of predicted terminators, with 46 and 45, respectively. In contrast, BYM-TASP37 contained only a single predicted terminator, indicating a simpler transcriptional architecture that may facilitate rapid gene expression or streamlined lifecycle processes. Other phages, such as GKM-TASP92 with 11 terminators and ZHS-TASP18 with 2, show intermediate terminator counts.

Fig. 7.

An annotated genome map of seven phages with transcription terminators: (A) phage ZHS-TASP18, (B) phage KFM-TASP06, (C) phage MHM-TASP32, (D) phage AKM-TASP09, (E) phage GKM-TASP92, (F) phage TTM-TASP117, and (G) phage BYM-TASP37

Discussion

In the current study, seven lytic bacteriophages targeting MDR K. pneumoniae were sequenced, revealing circular double-stranded DNA genomes, consistent with those of many tailed phages known to infect gram-negative bacteria [37]. All phages were confirmed to be strictly lytic, exhibiting no detectable genes associated with virulence factors or antimicrobial resistance. This finding is promising for their therapeutic potential, as temperate phages carry risks such as horizontal gene transfer and lysogenic conversion, which are not ideal for phage therapy. The phages were classified into the genera Przondovirus, Webervirus, Drulisvirus, and Taipeivirus. The sequence-based classification was in agreement with PCR-based identification [22]. Consistent with previous studies, the diversity of strictly lytic phages supports the genetic variability of bacteriophages targeting K. pneumoniae [38].

Whole-genome analysis of sequenced phages revealed conserved infection-related proteins, including tail and tail spike proteins, which are critical for host recognition and adsorption. These findings are consistent with previous research on phage host specificity [39]. We also identified fundamental structural elements, consistently observed across samples. Their ubiquitous presence underscores a conserved phage architecture essential for proper assembly and structural integrity. Additionally, the small terminase subunit, crucial for DNA packaging, was consistently annotated, supporting established roles in phage assembly [40]. The variable presence of endonucleases and exonucleases indicates diverse DNA processing strategies, similar to those found in other virulent phages. Importantly, no virulence factor or antibiotic resistance genes were detected, indicating the safety of these phages for therapeutic applications. This is in line with previous studies that emphasize the importance of strictly lytic phages, free from undesirable genes, for safe phage therapy. Similar genomic profiles have been reported by earlier studies in K. pneumoniae phages used against MDR strains [41, 42].

Sequence analysis of the lysis module indicated that the phages encode a diverse array of lytic enzymes, including holins, endolysins (lysozymes), spanins, peptidases, and lytic transglycosylases. These enzymes collectively contribute to the efficient degradation of bacterial cell walls and the subsequent lysis of host cells. Consistent with previous studies on Drulisvirus and Taipeivirus genera, which encode multiple putative depolymerases and lytic enzymes capable of degrading the polysaccharide capsule and peptidoglycan of K. pneumoniae [38], our phages harbor multiple ORFs with lyase, lytic transglycosylase, and peptidase activities. This supports the presence of enzymatic depolymerases targeting the bacterial cell wall, aligning with the known depolymerase profiles within these genera [43]. This correlation supports efficient and coordinated gene expression, consistent with previous studies showing regulatory elements scale with genome complexity [44].

The hierarchical clustering analysis revealed a modular organization within the bacteriophage genomes, highlighting conserved functional groups across diverse phages. The high sequence identity (65–92%) among major capsid proteins suggests a shared icosahedral architecture, consistent with their fundamental role in capsid formation [45]. Conversely, tail fiber proteins displayed significant divergence (25–55%), reflecting adaptations for different host recognition and specificity, a pattern typical in phages that accommodate host variability [46]. The presence of canonical lysis modules, including lysozymes/endolysins and holins. Structural protein modules associated with assembly and infection showed moderate to high homology, indicating functional conservation despite sequence differences. The observed mosaicism, particularly in tail fiber genes, underscores the dynamic evolution of phage genomes driven by host adaptation and horizontal gene transfer, influencing host range and infection efficacy [47].

The identification of Rz-like spanins in our phages highlights a conserved lysis mechanism characteristic of Klebsiella phages. Spanins disrupt the outer membrane following peptidoglycan degradation by endolysins, thereby completing the lysis process in Gram-negative hosts. Additionally, the detection of lysis inhibitors (RIIB and RIIA) reflects regulatory elements previously reported to modulate lysis timing, optimizing phage progeny release while preventing premature host cell lysis [48]. Notably, the putative depolymerase enzymes identified, particularly those containing lyase and lytic transglycosylase domains, represent promising candidates for novel antimicrobial agents. Prior research has demonstrated that Klebsiella phages encoding multiple depolymerases can degrade diverse capsule types, thereby enhancing infectivity and therapeutic efficacy [49].

Our findings suggest these phages may exhibit broad capsular specificity or be engineered to target specific K. pneumoniae capsule types, a strategy supported by the successful use of engineered phage depolymerases to overcome capsule-mediated resistance [49–52]. In the current study, the capsular type of the K. pneumoniae strains was not determined, which restricts our ability to assess the capsular specificity of the identified depolymerases. Future work should focus on characterizing the capsule types of the bacterial strains and evaluating the depolymerase activity against diverse capsule variants to better understand and optimize their therapeutic potential.

All phages encoded numerous lysis-associated ORFs, including holins, endolysins, spanins, and peptidases, consistent with their virulent lifestyles. Genus-level differences were observed: Taipeivirus phages (KFM-TASP06, TTM-TASP117, and GKM-TASP92) encoded multiple redundant RIIA/B lysis inhibitors and diverse lysozyme variants, indicating sophisticated regulation of lysis timing. This feature parallels Ackermannviridae phages such as phiAS5, which optimize host exploitation through lysis timing genes [53]. Conversely, the Przondovirus phage exhibited a minimalist lysis system with few core ORFs, consistent with the streamlined genomes of Studiervirinae phages that favor rapid replication over complex regulation [53]. Taipeivirus phages also encoded multiple Rz-like spanins (2–3 copies), exceeding the single-spanin systems typical of model phages like T4 [54, 55]. This multiplicity may enhance the disruption of the complex outer membranes of Gram-negative bacteria.

The presence of both RIIA and RIIB lysis inhibitors in KFM-TASP06 is unusual. These proteins are typically encoded separately and function independently in many phage systems. Their co-occurrence suggests novel regulatory mechanisms in lysis control. RIIA and RIIB have previously been reported to regulate lysis [48]. The presence of transmembrane (TM) domains in holin and hypothetical proteins in the current Klebsiella phage sequences supports their roles as membrane-associated lysis factors. Holins, characterized by TM domains, form pores in the bacterial membrane to enable endolysin access for cell wall degradation, facilitating phage release. Additionally, hypothetical proteins with transmembrane (TM) regions likely represent other lysis components. These may include spanins or Rz-like proteins. Such proteins disrupt the outer membrane in Gram-negative hosts like Klebsiella. This clustering of multiple TM-containing proteins within phage lysis cassettes reflects a coordinated lysis mechanism, underscoring their critical function in efficient bacterial membrane disruption and phage replication [56].

Ackermannviridae phages consistently encoded lytic transglycosylases, whereas Drexlerviridae (e.g., AKM-TASP09) relied primarily on holin-endolysin systems, reflecting distinct lysis strategies that align with prior reports. Several ORFs predicted to encode lysozymes with conserved hydrolase activity were identified, specifically ORFs 26, 118, and 128 in KFM-TASP06; ORFs 25, 26, 115, and 127 in TTM-TASP117; and ORFs 45, 46, and 55 in ZHS-TASP18. These enzymes likely play a vital role in peptidoglycan degradation, a crucial step in the phage lytic cycle, thereby facilitating the release of progeny [55, 57].

Phage-encoded cell wall hydrolases, including lysozymes and lytic transglycosylases, facilitate the degradation of bacterial cell wall polymers, thereby compromising structural integrity and promoting bacterial lysis [58]. The presence of multiple hydrolase genes suggests a synergistic enzymatic system targeting different peptidoglycan bonds for efficient host cell wall breakdown, a hallmark of virulent phages [58]. Their identification underscores the crucial role of phage-encoded enzymes in bacterial cell wall degradation and highlights their promising potential as antibacterial agents for phage therapy and as valuable tools in biotechnology, given that these modules can be cloned and harnessed for various applications.

The variation in intermediate terminator counts among different phages, such as GKM-TASP92 with 11 terminators and ZHS-TASP18 with only 2, underscores the diversity in transcriptional regulation mechanisms employed by these viruses. Previous studies have shown that phages with more complex regulatory regions often possess a higher number of transcription factors and terminators, which facilitate precise control of gene expression [21, 59]. This suggests that GKM-TASP92’s numerous terminators may allow for fine-tuned transcriptional switches, enabling the phage to adapt to different stages of infection or host responses. Conversely, the lower number of terminators in ZHS-TASP18 indicates a simplified transcriptional architecture, potentially favoring rapid gene expression with less regulatory complexity, as observed in simpler phage genomes [59, 60]. These differences imply that transcriptional complexity is evolutionarily tailored to optimize phage survival and infectivity within specific ecological niches [61]. Understanding these regulatory variations not only provides insights into phage biology but also has implications for engineering phages for therapeutic purposes, where precise control of gene expression is crucial.

Conclusion

This study provides a comprehensive genomic evaluation of seven novel, strictly lytic bacteriophages targeting carbapenemase-producing MDR K pneumoniae, reinforcing their potential as targeted biotherapeutic agents. The phages exhibit conserved structural and functional modules, including depolymerases and lytic enzymes, essential for efficient bacterial lysis and capsule degradation. Importantly, the absence of virulence factors and antibiotic resistance genes confirms their safety profile for clinical application. Variations in lysis regulation mechanisms, such as the presence of multiple spanins and lysis inhibitors, suggest sophisticated control of lytic timing and membrane disruption strategies, which may enhance their efficacy against diverse K. pneumoniae strains. These findings support the continued development of these phages as targeted biotherapeutic agents, addressing the urgent need for innovative treatments against multidrug-resistant K. pneumoniae infections. In vivo studies and clinical evaluations are essential to validate their safety, efficacy, and potential for therapeutic deployment.

Authors’ contributions

AAA: Conceptualization, study design, experimental framework, methodology development, laboratory investigations, resource, sequence analysis, data curation, annotation, result interpretation, visualization, reporting, manuscript preparation, original draft writing, and review and editing.AGB: Conceptualization, study design, Resources, Supervision, Validation, review, and editingMB: Resources, Supervision, Validation, review, and editingDHA: Methodology, laboratory investigations, data curation, and experimental frameworkAY: Data analysis, data curation, visualization, reporting, review, and editing.YD: Laboratory investigations and data curationKT: Laboratory investigations and data curationLTM: Data curation, review, and editingTST: Conceptualization, study design, experimental framework, project administration, Resources, Supervision, Validation, review, and editing.

Funding

This study did not receive any financial support or funding from public, private, or nonprofit organizations.

Data availability

The genome sequences of the phages have been deposited in NCBI GenBank under accession numbers [PV588648](https:/www.ncbi.nlm.nih.gov/nuccore/PV588648), [PV588649](https:/www.ncbi.nlm.nih.gov/nuccore/PV588649), [PV588650](https:/www.ncbi.nlm.nih.gov/nuccore/PV588650), [PV588651](https:/www.ncbi.nlm.nih.gov/nuccore/PV588651), [PV588652](https:/www.ncbi.nlm.nih.gov/nuccore/PV588652), [PV588653](https:/www.ncbi.nlm.nih.gov/nuccore/PV588653), and [PV588654](https:/www.ncbi.nlm.nih.gov/nuccore/PV588654) . The raw sequence data have been deposited in the NCBI Sequence Read Archive (SRA) under BioSample accession numbers [SAMN53043033](https:/www.ncbi.nlm.nih.gov/biosample/53043033), [SAMN53043034](https:/www.ncbi.nlm.nih.gov/biosample/53043034), [SAMN53043035](https:/www.ncbi.nlm.nih.gov/biosample/53043035), [SAMN53043036](https:/www.ncbi.nlm.nih.gov/biosample/53043036), [SAMN53043037](https:/www.ncbi.nlm.nih.gov/biosample/53043037), and [SAMN53043038](https:/www.ncbi.nlm.nih.gov/biosample/53043038), associated with BioProject [PRJNA1355480](https:/www.ncbi.nlm.nih.gov/bioproject/PRJNA1355480) .

Declarations

Ethics approval and consent to participate

The study proposal was reviewed and approved by the Research Ethics Committee at the Institute of Biotechnology, Addis Ababa University (IoB/431/2016/2024). All methods were conducted in accordance with relevant ethical guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Consent to participate

The study involved environmental sampling for bacteriophage isolation. The requirement for informed consent was deemed not applicable, as the study did not involve human subjects, in accordance with the National Research Ethics Review Guideline (NRERG, No. 01/2020, October 2020) of Ethiopia, which allows exemptions for minimal-risk research involving environmental sampling without direct human involvement.