Isolation of Novel Carbapenem-Resistant Klebsiella pneumoniae (CRKP) ST 147 Phage RMN1

The rapid global spread of carbapenem-resistant Klebsiella pneumoniae (CRKP) has become a critical public health concern, particularly in hospital-acquired infections (HAIs). CRKP has emerged as a severe challenge for clinical management, considering increased mortality rates and elevated healthcare costs (Cassini et al., 2019). World Health Organization (WHO) has classified CRKP, as “critical“ priority pathogens requiring urgent research into alternative therapies (Tacconelli et al., 2018).

The global dissemination of CRKP sequence type 147 (ST147); a high-risk clone associated with multidrug resistance (MDR) and hospital outbreaks, poses a critical threat to public health systems worldwide (Peirano et al., 2020). ST147 strains exhibit enhanced virulence and plasmid-mediated resistance to last-resort antibiotics, limiting treatment options and increasing mortality rates in clinical settings (Wyres et al., 2020). The difficulties in treatment options, necessitating alternative approaches such as phage therapy, which utilizes bacteriophages to specifically target and lyse bacterial pathogens (Strathdee et al., 2023).

Bacteriophages offer several advantages over antibiotics, including high specificity, self-replication at infection sites, and minimal disruption to the host microbiota (Lin et al., 2017). Despite their therapeutic potential, the isolation and characterization of phages against CRKP ST 147 remain underexplored, particularly in clinical settings. While numerous phages with lytic activity against MDR K. pneumoniae have been reported (Fausti et al., 2025; Parra et al., 2025); nevertheless, phage efficacy against CRKP remains limited due to restricted host range (Li et al., 2024). Commonly isolated CRKP-specific phages often display narrow tropism, typically targeting only specific sequence types (e.g., ST258, ST11) or capsule variants (e.g., KL64, KL107) (Kim et al., 2024). Furthermore, environmental CRKP strains frequently harbor temperate phages, which may contribute to lysogenic conversion and potential enhancement of bacterial virulence or antibiotic resistance (Bleriot et al., 2020). Therefore, discovering novel, strictly lytic phages with broad-spectrum activity is crucial for advancing phage therapy.

Several recent studies on bacteriophages targeting CRKP have primarily focused on non-ST147 lineages (Li et al., 2024; Natarajan et al., 2024; Zhang et al., 2025), leaving a significant gap in our understanding of phage therapeutics for this high-risk clone. While several phages such as vB_KpS_GP-1 and vB_KpP_GP-2 have demonstrated efficacy against CRKP, their activity against ST147 (KL128) strains remains largely unexplored (Ponsecchi et al., 2024). Notably, ST147-specific phages are exceptionally rare in literatures, despite this clone’s global dissemination and association with carbapenem resistance (Dey et al., 2023). The few reported phages active against ST147, such as vB_KpP_GP-4 showed limited host range, highlighting the need for broader host spectrum (Ponsecchi et al., 2024). The ST147 clone includes multiple K-types, defined by variations in the cps loci, which influence phage susceptibility (Loh et al., 2021; Wakinaka et al., 2023). The objective of this study was to to isolate and characterize a novel bacteriophage, RMN1, capable of lysing CRKP clinical isolates. We assessed its morphological features and genomic stability to evaluate its therapeutic potential. The obtained findings contribute to the growing body of researches on using phage-based alternatives to combat antibiotic-resistant infections, offering a potential solution to the escalating crisis of CRKP.

The clinical bacterial strain used in this study was isolated from a urine specimen obtained from a hospitalized patient at Dr. Soetomo Regional General Hospital, Surabaya, Indonesia. A bacterial stock was maintained on Tryptic Soy Agar (TSA; BD Difco™, USA) supplemented with 1.5% (w/v) agar with overnight incubation at 37°C. Mugo et al. (2025) protocol was followed for phage screening with slight modifications. The protocol started with the double agar overlay technique that employed TSA containing soft agar (0.7% agar in TSB) for both spot testing and phage titration through plaque assays. For broth cultures, individual colonies were inoculated into 3 mL of Tryptic Soy Broth (TSB; Oxoid, USA) and incubated at 37°C with continuous shaking at 150 rpm.

A bacteriophage was isolated from wastewater treatment plant of a general hospital in Jakarta, Indonesia, following established protocols with slight modifications (Mardiana et al., 2022). The isolation procedure involved initial sample processing by centrifugation (6.000 x g, 10 min), followed by filtration using 0.22 µm pore-size membranes (Millipore filter) to remove particulate matter. For phage enrichment, the filtered supernatant was combined with 1 mL of exponential-phase bacterial culture and 15 mL of 10× concentrated TSB, followed by an overnight incubation at 37°C with agitation (150 rpm). Following enrichment, the lysate was clarified by centrifugation (6.000 x g, 10 min) and sterile filtration (0.22 µm). Initial phage detection was performed using the spot assay technique reported by Mugo et al. (2025) with slight modifications, where 5 µL aliquots of filtrate were applied to bacterial lawns prepared with the soft agar overlay method (0.7% agar in TSB). High-titter phage lysates (~1010 PFU/mL) were concentrated by centrifugation ((18.000 x g, 2 h, at 4°C) and the pellets were suspended in 1 mL Saline Magnesium (SM) buffer (0.05 M Tris-HCl, pH 7.5 containing 0.1 M NaCl, 0.008 M MgSO4•7H2O, and 0.01% gelatin).

Phage quantification was conducted using the standard double-layer agar technique with modifications (Mardiana et al., 2022). Briefly, serial dilutions of phage lysate were prepared in SM buffer and mixed with log phase bacterial culture in molten soft agar (0.7%), which was subsequently overlaid onto TSA plates. Plates were incubated at 37°C for 18–24 h to allow plaque development for titer determination.

Phage morphology was characterized using high-resolution transmission electron microscopy (HR-TEM) following established protocols with minor modifications (Mardiana et al., 2023). For sample preparation, 10 μL aliquots of high-titter phage lysate (≥1010 PFU/mL) were adsorbed onto 200-mesh copper grids coated with Formvar film. The grids were subsequently negatively stained with 2% (w/v) uranyl acetate and air-dried at room temperature. Imaging was performed using a Talos F200C G2 transmission electron microscope (TEM, Thermo Fisher Scientific, USA). Viral capsid diameters were quantitatively analyzed using ImageJ software (version 1.54g; NIH) following the standardized protocols for nanoparticle characterization (Schneider et al., 2012). Capsid diameters were measured using ImageJ software (version 1.54g; NIH) following the standardized protocols for nanoparticle characterization (Schneider et al., 2012), with triplicate measurement performed for each particle to ensure measurement precision. This analytical approach enabled comprehensive morphological characterization of phage RMN1, including determination of its mean capsid diameter with associated standard deviation (±SD).

Genomic DNA extraction from the selected bacterial host was performed using Qiagen DNeasy blood and tissue kit (Qiagen, Carlsbad, CA, USA) following the manufacturer’s protocol. Whole-genome sequencing (WGS) was conducted on the Illumina NovaSeq 6000 PE150 coverage 100X platform (Illumina, USA) (Modi et al., 2021). Quality-controlled reads were assembled de novo using the automatic bacterial isolate assembly, annotation, and analyses (ASA3P) (https://github.com/oschwengers/asap) pipeline (v1.3.0) with Klebsiella pneumoniae HS11286 (NC_016845.1) as the reference genome (Schwengers et al., 2020). Subsequent genomic characterization was carried out using the PathogenWatch (https://pathogen.watch/) platform for comprehensive antimicrobial resistance such as genome length, GC content, sequence type (ST), K locus, capsular type determination using wzi gene sequencing, O locus, predicted O type, and antimicrobial resistance genes. wzi gene is one of six conserved genes (i.e., galF, orf2, wzi, wza, wzb, and wzc) located in cps locus for K typing (Argimón et al., 2021).

High-purity phage genomic DNA was extracted from purified phage particles (≥1010 PFU/mL) using the Phage DNA isolation Kit (Norgen BioTek Corp., Thorold, ON, Canada). Phage WGS was conducted using Illumina NovaSeq 6000 PE150 coverage 100X (Illumina, USA). Bioinformatic analysis was performed using Geneious Prime 2025.0.2 build 2024-10-30 and its plugins (Kearse et al., 2012). Bioinformatics processing included quality trimming using BBDuk adapter and de novo assembly using SPAdes v4.0.0 (Bankevich et al., 2012). Genome annotation was carried out using RASTtk (https://rast.nmpdr.org/) pipeline (McNair et al., 2018). Comparative genomics analysis was performed using BLASTn (https://blast.ncbi.nlm.nih.gov/Blast.cgi) against the National Center for Biotechnology Information (NCBI) non-redundant nucleotide database (nr/nt) for phylogenetic analysis and taxonomic classification, with matched genome database evaluated based on percent identity and query coverage thresholds (i.e., identity >90%, coverage >80%). Presence of integrase, tRNAs, transmembrane proteins, virulent factor, and antimicrobial resistant genes in the genome were detected with Phagescope ( Wang et al., 2024). Phage genome packaging mechanisms and genome termini were determined through PhageTerm (Garneau et al., 2017). Phage genome map was visualized via SnapGene Viewer 8.0 (http://www.snapgene.com/). Genome comparison with its highly similar phages (top five of highest identity and coverage threshold reference genome) was performed using MUMMER protein-based comparison, visualized by pyGenomeViz (https://pygenomeviz.streamlit.app/) to highlight regions of synteny and divergence. Phylogenetic analysis and Viral Proteomic Tree (VIPTree) 4.0. (https://www.genome.jp/viptree/) were used to generate a proteomic tree of phage genome based on genome-wide sequence similarity computed through tBLASTx with Drulisvirus-specific filters (Nishimura et al., 2017).

Two bacterial isolates namely 032U and 118U were collected. Phenotypic characterization using the automated VITEK® 2 COMPACT system (bioMérieux, USA) confirmed both isolates as extended-spectrum β-lactamase (ESBL)-KP, which showed sensitivity to meropenem and imipenem. The bacterial genome analysis revealed 032U belonged to ST 147 and 118U belonged to ST 469. Therefore, 032U was selected as the host strain in this study. Genomic analysis using Pathogen Watch revealed that this isolate belonged to ST 147 with an unknown (KL128) capsule type and possessed an ompK35 mutation associated with carbapenem resistance. Based on this genotype, the isolate was classified as CRKP. Using VITEK® 2 COMPACT system (bioMérieux, USA), this isolate displayed resistance (100%) to several antibiotics; mainly third-generation cephalosporins, fluoroquinolones, penicillin, sulfonamides, tetracycline, and trimethoprim (Supplementary Table S1). The genome length was 5,452,315 bp with 57.3% GC content. This isolate expressed the presence of the blaCTX-M-15 resistance gene, confirming its identity as an extended-spectrum β-lactamase (ESBL)-KP isolate. Carbapenemase gene was not detected in this isolate.

The novel phage RMN1 was successfully isolated from wastewater treatment plant of General Hospital in Jakarta, Indonesia. Initial plaque assays demonstrated that phage RMN1 (1010 PFU/mL) produced distinct lytic zones (i.e., small and clear plaques) with a mean diameter of 0.25 ± 0.01 mm (Figure 1a). TEM analysis revealed that the viral particle possessed an icosahedral capsid architecture with an average diameter of 56 nm (Figure 1b), consistent with morphological characteristics of the Podovirus. This structural confirmation established phage RMN1 as a tailed phage with a podoviral morphology.

Whole-genome sequencing revealed that phage RMN1 possessed a 44,226 bp double-stranded DNA genome with a GC content of 53.9%, characteristic of the Drulisvirus genus. Genomic termini analysis identified 278-bp direct terminal repeats (DTRs) with redundant ends (Supplementary Figure S1). BLASTn analysis showed that phage RMN1 shared between 93.73% and 90% nucleotide similarity with Klebsiella phage vB_KpnP_Dlv622, P252, vB_KpP_AttikonH1, Pone, and VLC5 (Supplementary Table S2).

The phage RMN1 genome involved 58 putative open reading frames (ORFs), including 26 annotated and 32 hypothetical proteins. Functional annotation of the 58 predicted ORFs revealed five functional modules: (i) replication (phage DNA helicase, phage DNA-directed DNA polymerase, phage DNA-directed RNA polymerase, phage peptidase, and phage primase), (ii) packaging (endonuclease, exonuclease, phage phosphoesterase, HNH endonuclease, and phage terminase small and large subunits), (iii) structural assembly (capsid, major capsid protein, and non-contractile tail fiber protein), (iv) infection (internal virion protein A and phage non-contractile tail tubular protein), and (v) lysis (endolysin, holin, and spanin). Some ORFs exhibited combined functions, such as assembly and infection (collar and head to-tail connector protein) or lysis and infection (DNA ejectosome component and phage internal virion B) (Figure 2 and Supplementary Table S3), which were shared by several T7-like tail components.

The absence of virulence factors (e.g., toxin genes, immune modulators, or secretion system effectors) and antimicrobial resistance genes (particularly β-lactamases or qnr-type quinolone resistance) in phage RMN1’s genome aligns with World Health Organization (WHO) safety guidelines for therapeutic phages. Notably, the lack of tRNA genes further reinforced the phage’s obligatory lytic nature, minimizing risks of lysogenic conversion or host fitness enhancement.

Comparative analysis showed high similarity and gene synteny between the tested genomes, with some differences in hypothetical proteins, HNH endonuclease, and non-contractile tail fiber protein, suggesting divergent host recognition mechanisms and reinforcing phage RMN1’s host specificity (Figure 3a). A proteomic tree generated by VIPtree, included phage RMN1 and all available Klebsiella phages in database clustered phage RMN1 within the Drulisvirus genus (Figure 3b).

The complete genome sequence of phage RMN1 was deposited in the NCBI GenBank database under accession number PV804878, providing full open access to the genomic data.

Genomic analysis of isolate 032U revealed a chromosome of 5,452,315 bp with 57.3% GC content, dimensions that aligned closely with the typical genomic architecture of Klebsiella pneumoniae, which generally ranged from 5.0 to 5.7 Mbp with a GC content of 57-58%, confirming a high-quality genome assembly suitable for accurate genotyping (Yao et al., 2017). The isolate’s predicted resistance profile spaned multiple antibiotic classes, including carbapenems, β-lactams (via blaCTX-M-15), and fluoroquinolones, characteristic of multidrug-resistant ST147 strains (Park et al., 2023). Although no carbapenemase gene was detected, an ompK35 porin mutation was identified, supporting a carbapenem-resistant phenotype mediated by non-carbapenemase mechanisms (Cox et al., 2025). Collectively, all these attributes established 032U as a well-characterized, resistant ST147-KL128 variant, offering a stringent and clinically relevant host for evaluating the novel phage RMN1.

Successful isolation of bacteriophage RMN1 from hospital wastewater highlighted the potential of environmental sources as reservoirs for novel phages with therapeutic applications. Our findings demonstrated that phage RMN1 formed clear plaques, indicating strong lytic activity against its CRKP host. This aligns with recent studies showing that wastewater from healthcare facilities harbors diverse phages capable of infecting clinically relevant pathogens (Lu et al., 2024).

The successful isolation of podovirus RMN1 from hospital wastewater represented a substantial addition to the limited repertoire of phages active against CRKP ST147. The observed plaque morphology (0.25 ± 0.01 mm clear plaques) suggested robust lytic activity, comparable to other therapeutic podoviruses such as TC6 (0.5 mm plaques) (Tang et al., 2018), and it had an icosahedral head diameter of about 54 nm. The icosahedral capsid morphology (56 nm diameter) placed phage RMN1 within the typical size ranged for podoviruses infecting Enterobacteriaceae (55–60 nm) (Jamal et al., 2019). Notably, the observed clear plaque phenotype indicated strictly lytic behavior; a crucial characteristic for therapeutic applications, as it minimizes the risks of lysogenic conversion and horizontal gene transfer (Altamirano and Barr, 2021).

The isolation of phage RMN1 from an Indonesian hospital wastewater expanded the geographical diversity of characterized K. pneumoniae phages, which had been predominantly isolated from European and American sources (Castillo et al., 2014). This finding supported the hypothesis that environments may harbor unique phage diversity due to distinct microbial ecosystems, as suggested by recent metagenomic study (Weinheimer et al., 2023). The podoviral morphology of phage RMN1 is particularly promising, as members of this family have demonstrated enhanced stability under physiological conditions compared to myoviruses, considered as an important consideration for pharmaceutical formulation (Smolarska et al., 2018). These characteristics positioned phage RMN1 as an excellent candidate for further development against the globally emerging ST147 clone.

The CRKP clinical isolate 032U characterized in this study represented a particularly concerning ST147 strain with an unusual KL128 capsule type and extensive MDR profile, including resistance to carbapenems and third-generation cephalosporins. This isolate carried the blaCTX-M-15 gene and an ompK35 porin loss mutation. This genetic profile is a high-risk due to its potential to developing carbapenem resistance. This finding aligns with recent global reports of ST147 as a high-risk clone demonstrating remarkable rapid dissemination of resistance determinants (Park et al., 2023). The presence of CTX-M-15 gene detected in our isolate had been strongly associated with hospital outbreaks and treatment failures worldwide (Mshana et al., 2015). Notably, the KL128 capsule type identified in our strain differed from the more commonly reported KL64 and KL10 variants in ST147 (Rodrigues et al., 2021). This correlation between capsule type and sequence type observed in this study may be attributed to the impact of local clonal expansions of this pathogen (Park et al., 2023). Klebsiella capsule type is the main determinant for phage tropism (Haudiquet et al., 2024). Our genomic analysis identified capsule-type diversity within ST147, specifically KL128 variant. Variation in capsule type directly influences phage receptor recognition, thereby impacting therapeutic efficacy (Cheetham et al., 2024).

The identified genomic variations in phage RMN1, particularly in non-contractile tail fiber proteins and hypothetical proteins of unknown function, suggested evolutionary adaptations to specific K. pneumoniae hosts. This is consistent with the “arms race” co-evolutionary model between phages and bacterial capsular polysaccharides (Huang et al., 2025). Phage-host coevolution drives selective pressure for minor genomic changes that confer host specificity, even within highly conserved genomes (Holtappels et al., 2023). A structural study confirmed that even single-amino-acid changes in tail fibers can alter binding affinity to specific capsule types (Taslem Mourosi et al., 2022). Thus, while core genomic synteny was preserved, the critical tail fiber variations in phage RMN1 represented functional microevolution; –a hallmark of phage adaptation to bacterial surface diversity (Burmeister et al., 2023).

The combination of CTX-M-15 with resistance to last-resort antibiotics such as carbapenems creates particularly challenging clinical scenarios, as noted in intensive care unit (ICU) outbreaks where mortality rates exceed 40% (Souli et al., 2010). Our isolate’s resistance profile and ST147/KL128 characteristics provided a stringent model for evaluating phage therapeutics, as successful lysis demonstrated the phage potential against one of the most treatment-refractory CRKP variants currently emerging worldwide.

The genome shared < 94% identity with Drulisvirus Klebsiella phage vB_KpnP_Dlv622, P252, vB_KpP_AttikonH1, Pone, and VLC5 (below the 95% species threshold). Its distinct genetic signature and proteomic profile justified its classification as a novel phage with potentially unique therapeutic applications, confirming phage RMN1 as a novel member of the Drulisvirus genus (O’Connell et al., 2024). Meanwhile, the high degree of synteny conservation in core genomic regions supported the phage RMN1 functional similarity with these reference phages (Ismail et al., 2023), and the observed variations in tail fiber proteins and HNH endonuclease domains likely accounted for its unique host range specificity. These findings align with a recent study demonstrating that even minor genomic differences in tail fiber genes can dramatically alter host recognition patterns, particularly against diverse K. pneumoniae capsule types (Wang et al., 2024).

The presence of 278-bp DTRs is particularly noteworthy as this genomic architecture, also observed in the closely related vB_KpnP_Dlv622 phage (Gorodnichev et al., 2021), enhanced DNA packaging efficiency and genomic stability (Casjens and Gilcrease, 2009). Of the 58 predicted ORFs, the 26 annotated genes included essential structural and lytic components, while the 32 hypothetical proteins presented intriguing targets for future functional studies. Crucially, the absence of integrase, virulence factors, antimicrobial resistance genes, and tRNA molecules satisfied key safety criteria for therapeutic phages (Dhungana et al., 2024).

The phylogenetic positioning of phage RMN1 among closely related Klebsiella phages underscored its evolutionary proximity and supported its taxonomic assignment. Phage RMN1 clusters were tight with other phages such as Klebsiella phage vB_KpnP_Dlv622 and VLC5, forming a coherent clade characterized by short branch lengths and high bootstrap support–evidence of a shared recent ancestor (Lemoine and Gascuel, 2024). The congruent genome sizes (~43–45 kb) and GC content (~54%) across this clade reinforced a conserved genome backbone and affirmed phage RMN1’s classification within this phylogenetic framework of Klebsiella podoviruses, similar to the KP-Rio/2015 bacteriophage (Meira et al., 2016). This close phylogenetic relationship suggested conserved mechanisms of host recognition, possibly mediated by capsule-specific interactions, as phage RMN1 targeted ST 147 KL128-capsulated K. pneumoniae.

The genomic features of phage RMN1 positioned it advantageously within the current phage therapeutic landscape. Its 53.9% GC content showed remarkable adaptation to its K. pneumoniae ST147 host (57–58% GC), potentially facilitating more efficient gene expression during infection (Dey et al., 2023). Compared to other Drulisvirus members, phage RMN1 demonstrated conserved synteny in structural gene clusters, including capsid, major capsid protein, and non-contractile tail fiber protein while maintaining unique variations in tail fiber proteins, likely responsible for its specific host recognition (Wagemans et al., 2020). These molecular characteristics combined with its strictly lytic nature, made phage RMN1 an exceptionally promising candidate for clinical development against CRKP, particularly as it targeted the high-risk ST147 lineage.

The genomic architecture of phage RMN1 revealed a sophisticated lytic toolkit, particularly noteworthy for its dual-component cell lysis system comprising holin and endolysin proteins. The holin protein formed pores in the bacterial cytoplasmic membrane at precisely programmed times, allowing endolysin to access and degrade the peptidoglycan layer; a mechanism shown to be highly effective against MDR K. pneumoniae (Cahill and Young, 2019). Endolysin protein displayed its function as an amidase, an enzyme that cleavesed the essential peptide cross-links within the peptidoglycan matrix (Pennone et al., 2019). This precise targeting of the bacterial cell wall’s integrity presented a remarkable advantage over conventional antibiotics, as it avoided the common resistance mechanisms while maintaining high specificity for its CRKP ST147 host.

Equally intriguing was phage RMN1’s complement of structural proteins governing host recognition and DNA delivery. The non-contractile tail tubular proteins and DNA ejectosome component worked synergistically to recognize the KL128 capsule polysaccharide and created a channel for genome injection; a process recently visualized in near-atomic detail for related podoviruses (Hardy et al., 2022). The presence of putative spanin proteins suggested that phage RMN1 employed a two-step lysis mechanism, while holin-endolysin disrupted the cell wall from within (holin forms a hole in the inner membrane, allowing endolysin to escape and destroy the peptidoglycan, resulting directly in bacterial lysis), spanins mediated an outer membrane disruption, ensuring complete bacterial host cell lysis (Cahill and Young, 2019). This comprehensive lytic system, combining precise host targeting with redundant lysis mechanisms, explained phage RMN1’s exceptional efficacy against CRKP ST147 and positioned it as a promising candidate for engineered phage therapeutics.

Recent studies emphasized the importance of capsule depolymerases in phage tropism (Beamud et al., 2023; Cheetham et al., 2024), which may explain phage RMN1’s specificity. However, we were unable to detect a depolymerase gene in phage RMN1. The strictly lytic nature of phage RMN1 enhanced its therapeutic utility (Anjay et al., 2025). The discovery of phage RMN1 expands the arsenal against CRKP ST 147, which increasingly evade last-resort antibiotics like carbapenems.

We recognized the limitations of this study. While phage RMN1 was able to lyse a single ST147-KL128 K. pneumoniae strain, its broader host range was not assessed. Furthermore, a depolymerase gene was not yet been detected in the phage RMN1 genome, suggesting an avenue for further investigations into its receptor-binding and host specificity mechanisms.

This study presented the first report of the isolation and characterization of a novel lytic Drulisvirus phage, which was selectively active against KL128 capsular CRKP. Its genomic safety, lytic efficiency, and capsule-specific targeting aligned with the urgent need for effective alternatives to antibiotics. We recommend that future studies should evaluate the structural mechanisms underlying phage attachment to the host, which determine its specificity for the capsule type rather than the sequence type. This specificity may be related to a receptor-binding domain. Combining phage RMN1 with other phages or antibiotics could further mitigate resistance development; a strategy gaining traction in precision medicine. Additionally, further in vivo studies and clinical trials are warranted to translate the current findings into therapeutic applications.

Thank you to wastewater plant treatment unit of Dr. Cipto Mangunkusumo General Hospital, Jakarta, Indonesia Team for helping the team during water sample collection and to all the technician at Genomic and CryoEM laboratory, the National Research and Innovation Agency of Indonesia.

This study isolated and characterized phage RMN1 as the first reported Drulisvirus isolated from wastewater treatment plant of General Hospital in Jakarta, Indonesia. This novel phage specifically targeted the Carbapenem-Resistant K. pneumoniae ST147 with the rare KL128 capsular type; a critical-priority pathogen. It highlights the phage RMN1’s specificity for the rare KL128 capsular type, its safety profile confirmed by genomic analysis, and its potential for phage therapy in clinical settings where antibiotic options are limited. The discovery of this phage RMN provides a crucial and locally sourced therapeutic candidate to combat globally circulating, high-risk bacterial pathogens in a region heavily affected by antimicrobial resistance.

NAS: Conceptualization, data curation, formal analysis, investigation, project administration, methodology, visualization, writing–original draft, writing–review and editing.

MM: Data curation, formal analysis, investigation, methodology, software, validation, writing–original draft.

YM: Formal analysis, software.

IR: Investigation, resources.

DS: Resources, validation.

YRS: Conceptualization, funding acquisition, resources, supervision, validation.

RS: Conceptualization, formal analysis, funding acquisition, methodology, resources, supervision, validation, visualization, writing–original draft, writing–review and editing.

All authors have read and agreed to the published version of the manuscript.

Ethical clearance for this research was granted by the Joint Research Ethics Committee of Faculty of Medicine, Universitas Indonesia and Dr. Cipto Mangunkusumo National Central General Hospital (Ethical Approval Code: KET-431/UN2.F1/ETIK/PPM.00.02/2024).

This study was funded by BIMA DRTPM (Directorate of Research and Technology in Education and Learning) supported by the Ministry of Higher Education, Science, and Technology (Decree Letter No. 0459/E5/PG.02.00/2024 and Contract Agreement No. 051/E5/PG.02.00.PL/2024); Rumah Program, Organisasi Riset Kesehatan 2024 supported by National Research and Innovation Agency of Indonesia (No. B-11155/III.9/TK.02.02/12/2023).

No generative AI or AI-assisted technology was employed in this study.

There is supplementary material associated with this article. Access the material online at https://dx.doi.org/10.17582/journal.nrmj/2026/10.1.37.51

The authors have declared no conflicts of interest.