Sediment samples were collected from a karst cave (107°47′29”N, 25°26′12″E) in Libo County, Guizhou Province, China, in July 2019. The average temperature was 16.8°C, the average CO₂ concentration was 480.58 PPM in the cave. Samples were obtained from the rock surface at intervals of every 10 meters from three different sites within the cave. Subsequently, these samples were then pooled, transferred into sterile tubes, chilled on ice and transported back to 4°C for storage for further bacterial and phage isolation in the laboratory.

Bacterial strain was isolated from sediment samples using the spread plate method with Luria-Bertani (LB) media and purified 3–5 times to obtain a strain with a consistent colony morphology, designated as YCR3A-1. Sterile glycerol was added to the strain at a final concentration of 25% and stored at −80°C for subsequent experiments. The 16S rRNA gene of YCR3A-1 was sequenced by Sangon Biotech (Shanghai) Co., Ltd., the sequencing data was submitted to the National Center for Biotechnology Information (NCBI).

Ten S. maltophilia strains (originating from clinical samples) were kindly provided by Professor Huahao Fan, College of Life Science and Technology, Beijing University of Chemical Technology (Han et al., 2021), five Stenotrophomonas spp. were isolated from caves in our laboratory. Totally 15 Stenotrophomonas spp. were used to test the host range of phage A1432 (Supplementary Table S1). Based on 16S rRNA gene sequence, a phylogenetic tree of YCR3A-1 with similar strains and the 15 strains of Stenotrophomonas spp. was constructed using the maximum likelihood (ML) method in MEGA software with 1,000 bootstrap replicates (Kumar et al., 2016).

All strains were cultured in LB broth or on LB agar plates at 37°C. AST was performed using the disk diffusion method as previously described with some modifications (Han et al., 2022a). Briefly, bacteria were incubated to the exponential phase (OD600 = 0.7) at 37°C, 1 mL of the bacterial culture was added to 10 mL of 0.7% agar-containing LB medium. This mixture was immediately poured onto Petri dishes. After solidification, antibiotic sensitivity disks (Hunan BKMAM Biotechnology Co., Ltd.) were placed on the plates and incubated overnight at 37°C. The antibiotics used are listed in Supplementary Table S1. Sterile antibiotic-free paper disks were used as blank controls, E. coli DH5α was used as a negative control under the same experimental conditions. If a strain is sensitive to a particular antibiotic, a clear zone of inhibition forms around the disk. The diameter of the zone is measured and recorded. The experiment was conducted three times.

Phage A1432 was isolated from a sediment sample from a karst cave, using the host strain YCR3A-1. Initially, 10 g of the collected sediments were homogenized in 90 mL sodium chloride-magnesium sulfate (SM) buffer (100 mM NaCl, 8 mM MgSO_4·7H₂O, 50 mM Tris–HCl [pH 7.5]). The resulting supernatant was filtered through a 0.22 μm poresize membrane (Millipore). The collected filtrate was then subjected to the double-agar overlay method to detect the presence of phages. Briefly, a series of dilutions of the filtrate containing the potential phages was prepared. A total of 1 mL of the host YCR3A-1 in the logarithmic growth phase in LB broth was added to 1 mL of appropriately diluted filtrate. After a 20 min incubation at 37°C, the mixture was poured onto solid LB agar, 8 mL of semi-solid medium (LB broth with 0.5% agar) melted at 50°C was added. Once dried, the dishes were kept upside-down and incubated overnight at 37°C. The appearance of plaques indicated a successful infection of the host bacterium by the phages.

In order to purify the isolated phages, a clear single plaque was individually transferred into SM buffer to obtain the first purified phage. This purification process was repeated three times for each phage, the fully purified phage was then mixed with a fresh culture of the host strain and incubated at 37°C with shaking at 180 rpm to obtain the phage proliferation solution. To remove cell debris, the proliferation solution was centrifuged at 8000 g for 30 min at 4°C and then aseptically filtered through a 0.22 μm filter membrane. The phage titer in the filtrate was determined by the double-layer agar method and expressed as plaque-forming units (PFU) per milliliter of filtrate. The proliferating solution was stored at 4°C for subsequent experiments. Sterile glycerol at a final concentration of 25% was added to the phage proliferating solution and stored at −80°C for long periods of time.

Phage supernatant was centrifugated at 12,000 g for 2 min and then filtered through a 0.22 μm membrane. A droplet of phage suspension (10⁹ PFU/ml) was applied onto a carbon-coated copper grid for 10 min. The grids were negatively stained with 2% (w/v) phosphotungstic acid (pH 6.8) for 5 min, air-dried, then observed with an HT7800 transmission electron microscope (HITACHI, Tokyo, Japan) at an accelerating voltage of 100 kV (Yi et al., 2022).

Fifteen strains of Stenotrophomonas genus were used to investigate the host range of phage A1432 (Supplementary Table S1). Five μL of the purified phage suspension (10⁸ PFU/mL) was inoculated onto the relevant bacterial culture plates and incubated at 37°C for 24 h to observe the formation of lysis zones at the phage suspension. All tests were performed in triplicate.

The MOI represents the ratio of phages to host bacteria during infection. To determine the optimal MOI for A1432, YCR3A-1 culture was mixed with the phage in different proportions: 100, 10, 1, 0.1, 0.01 and 0.001. The mixtures were incubated for 15 min at 37°C, centrifuged, the pellet was resuspended in 2 mL of fresh media. The samples were incubated for 3 h at 37°C, after which phage titer was determined using the double-layer agar method with A1432 as the host. The MOI that resulted in the highest titer was identified as the optimal MOI. Each experiment was performed in triplicate.

The one-step growth curve was performed. Briefly, the phage was mixed with the host bacteria according to the optimal MOI. After adsorption for 15 min at 37°C, the mixture was centrifuged and the pellets were resuspended in 50 mL of LB broth, followed by incubation at 37°C under shaking at 180 rpm. Samples were taken at 0, 10, 20, 30, 40, 70, 100, 130, 160, 190, 220 and 250 min after inoculation. The phage titer was determined using double-layer agar method. The experiment was repeated three times.

In the thermal assay, phages (10¹² PFU/mL) were incubated at 0, 4, 20, 30, 40, 50, 55, 60, 70 and 80°C for 60 min, respectively. In the pH assay, 100 μL of phage solution (10¹² PFU/mL) was added to 900 μL of SM buffer with different pH values and then incubated at 37°C for 60 min. The pH value of the SM buffer was adjusted from 2 to 13 using HCl or NaOH (Assefa, 2022). The phage titer was measured using a double-layer agar technique. Each experiment was performed in triplicate.

The sensitivity of the phage to chloroform is often used as an important reference for determining the presence or absence of lipid components in phage capsids (Kalidasan et al., 2018). Phage A1432 (1012 PFU/mL) was mixed with chloroform at a final concentration of 25% and incubated for 10 min at room temperature (Xiang et al., 2020). The supernatant was collected after centrifugation at 13,000 g for 1 min, the phage titer was assayed by the double-layer agar method. SM buffer instead of chloroform was used as a blank control.

The bactericidal activity of phage A1432 was evaluated in vitro as previously described, with a few modifications (Chen et al., 2023). Briefly, 100 μL of phage A1432 was mixed with host YCR3A-1 (OD₆₀₀ = 0.30) in the equal volume at MOI of 0.01, 0.1, 1, 10, respectively, incubated at 37°C for 24 h. Bacterial growth was monitored by measuring the OD600 at each time point using spectrophotometry. The experiment was conducted three times. A culture medium of host YCR3A-1 without phage was used as a positive control.

The purified phage lysate (10⁹ PFU/ml) was filtered through 0.22 μm pore size filter (Millipore). Then DNase I and RNase A (Takara, Japan) were added to 200 μL phage lysate (10⁹ PFU/mL) at final concentrations of 50 U/mL and 250 μg/mL, respectively, the mixture was then treated for 1 h at 37°C to remove from DNA and RNA. After digestion, genomic DNA of phage A1432 was extracted using the TaKaRa MiniBEST viral RNA/DNA extraction kit (Takara, Japan) following the manufacturer’s instructions. DNA was finally resuspended in deionized water and stored at −20\u00B0C. The concentration and quality of DNA were determined at 260 nm and 280 nm using Nanodrop spectrophotometry.

DNA libraries were constructed using the TruSeqTM DNA Sample Prep Kit (Illumina, San Diego, CA, United States) and sequenced in paired-end model by Personalbio (Shanghai Personal Biotechnology Co., Ltd., China) through the Illumina NovaSeq platform. The obtained FastQ raw reads were trimmed of adaptors and low-quality bases and short reads were filtered using AdapterRemoval v2.1.3 (Schubert et al., 2016). De novo assembly was performed using A5-MiSeq v20160825 and SPAdes v3.12.0 software (Bankevich et al., 2012; Coil et al., 2014). Pilon v1.18² was used to correct the results for obtaining the final genome sequence (Walker et al., 2014).

After obtaining the whole genome sequence, open reading frames (ORFs) in the phage genome were identified using the RAST server³ (Aziz et al., 2008). The putative function of each ORF-encoded protein was annotated by searching against the non-redundant protein sequences (NR) database with BLASTp,⁴ tRNAscan-SE was used to search for tRNAs.⁵ The presence of determinants of antimicrobial resistance was investigated using the comprehensive Antibiotic Resistance Database⁶ (Alcock et al., 2023) and the Virulence Factor Predictor⁷ (Liu et al., 2019). The genome circle map of phage A1432 was drawn by CGView server (Grant and Stothard, 2008).

The phage terminase large subunit (TerL) and DNA polymerase are highly conserved characteristic proteins that are frequently utilized in virus classification. The phylogenetic trees were constructed by Maximum Likelihood (ML) method using MEGA 7.0 software (Kumar et al., 2016) with the JTT matrix-based model and 1,000 bootstrap replicates based on the amino acid sequences of TerL and DNA polymerase I Moreover, a proteome phylogenetic tree of phage A1432 was generated using the Viral Proteomic Tree server⁸ (Nishimura et al., 2017). In these different analyses, BLASTp (TerL and DNA polymerase I) or BLASTn (proteomic tree) with an E-value cutoff of 0 (accessed 4 May 2024) were used to find the 20 best hit viruses in the GenBank database, duplicate and unassociated sequences were removed. Then, the same viral genome dataset containing the 20 best hit phages was finally used (only 19 viruses were selected to construct the TerL phylogenetic tree since the TerL gene was absent in the genome of S. maltophilia phage DLP4).

Genomic network analysis of phage A1432 was performed using the Prokaryotic Viral RefSeq211 Merged database (last updated in June 2022) in vConTACT (v.2.0) (Bin Jang et al., 2019) and the database of the top 20 best hit viruses. Duplicate and unassociated sequences are removed. Based on the number of shared protein clusters (PCs) between genomes, Markov cluster (MCL) and clustering with overlapping neighborhood expansion (ClusterONE) (v1.0) (Nepusz et al., 2012) were used to group the closely related genomes into virus clusters (VCs). Visual network map was drawn using Cytoscape software (v3.9.1v) (Shannon et al., 2003). The whole genome phylogenetic analysis was conducted using the VICTOR online server with default parameters (Meier-Kolthoff and Göker, 2017) based on the database of 20 best hit viruses and the 68 S. maltophilia phages (including A1432) with publicly-available genome sequences. Duplicate and unassociated sequences were removed. A pairwise comparison of nucleotide sequences was performed using the Genome-BLAST Distance Phylogeny (GBDP) method. Subsequently, a balanced minimum evolution tree with branch support was constructed using the FASTME with SPR postprocessing based on the intergenomic distances derived from equations D0, D4, D6.

The Virus Intergenomic Distance Calculator (VIRIDIC) heatmap was analyzed using a dataset consistent of viruses with the VICTOR analysis. VIRIDIC was utilized to calculate the pairwise average nucleotide identity (ANI) between the genome of A1432, its closely related phages, the 68 S. maltophilia phages. Then a clustering report based on ANI was formed (Moraru et al., 2020).

The whole genome sequence of phage A1432 was compared with that of similar phages using Mauve software (Darling et al., 2004). The phages were further compared in more detail at the gene level using Easyfig software (Sullivan et al., 2011). The default parameters were used.

All experiments were performed independently in triplicate. The data are expressed as the mean value and the standard deviation (SD). One/two-way ANOVA analysis of variance using GraphPad Prism. A p-value of less than 0.05 was considered statistically significant.

The 16S rRNA gene analysis revealed that strain YCR3A-1 had the highest similarity of 99.72% and coverage of 99% to S. maltophilia IH128R2A01 (MN829927). The phylogenetic tree constructed using the ML method (Supplementary Figure S1) showed that YCR3A-1 and the fifteen Stenotrophomonas spp. used for the host range test located on the same evolutionary branch as S. maltophilia, confirming its identification as S. maltophilia. The 16S rRNA sequence was submitted to the NCBI under the accession number OQ983545.

The AST assay demonstrated that YCR3A-1 was resistant to imipenem, cefuroxim and ampicillin, but sensitive to levofloxacin, minocycline, gentamicin and sulfamethoxazole.

Phage A1432, infecting S. maltophilia YCR3A-1, was isolated from karst cave sediment. This phage could form two types round and transparent plaques on LB double-layer agar plates. One type of plaque had no halo with mean diameter of 1 mm. The other type of plaque has a translucent halo that expands over time, while the size of the lysis zone remained constant (Figure 1). This plaque polymorphism was consistently observed in purification and subsequent experiments. TEM revealed that phage A1432 has an icosahedral head (70.40 ± 1.46 nm) and short tail (14.87 ± 1.84 nm) morphology.

Fifteen strains of S. maltophilis isolated from clinical samples and karst caves were used to evaluate phage A1432 host range. The results showed that phage A1432 could only lyse YCR3A-1 and produce clear phage spots, but not the other 15 tested strains. This suggests that A1432 has a very narrow host range.

At a MOI 0.1, the titer of A1432 peaked at 1.80 × 10¹² PFU/mL (Supplementary Table S2), indicating that the optimal MOI of phage A1432 is 0.1. The one-step growth curve (Figure 2A) of phage A1432 was determined at the optimal MOI. There was no significant change in the phage titer within the initial 10 min (p > 0.05), after which the titer gradually increased (p < 0.01), reaching a plateau at 100 min. The final count of released phage particles was 9.2 × 10⁹ PFU/mL. These results indicate that the phage had a latent period of approximately 10 min, a burst time of 90 min, an average burst size of 43.2 PFU/cell.

The stability of phage A1432 under various environmental conditions was characterized. The activity of A1432 remained almost constant when the temperature was between 0°C and 50°C (p > 0.05). However, the titer of the phage significantly decreased at 60°C (p < 0.01) and it was completely inactive at 80°C (Figure 2B). The pH stability test showed that the titer of phage A1432 was relatively stable within a pH range of 5 to 11 (p > 0.05). However, when the pH dropped below 4 or rose above 12, the phage titer gradually decreased (p < 0.01). Moreover, the phage was almost inactive at pH 3 and 13 (Figure 2C). These results indicated that phage A1432 can withstand a wide pH range, from weakly acidic to alkaline conditions. After treatment with chloroform, phage A1432 remained largely stable compare to the control group (Figure 2D; p > 0.05). This suggests that phage A1432 is resistant to chloroform.

The in vitro bacteriolytic activity of phage A1432 against S. maltophilia YCR3A-1 at various MOIs was assessed (Figure 2E). The OD600 of YCR3A-1 culture consistently increased within 24 h. However, when phage A1432 was added, the OD600 of the YCR3A-1 culture decreased (p < 0.01) in an MOI-dependent manner after 2 h of co-culture. After 12 h of co-culture, the OD600 of the culture increased significantly (p < 0.01). The killing curve suggested that phage A1432 was capable to inhibit the growth of host bacteria to varying degrees at different MOI.

The complete genome sequence of phage A1432 has been deposited in Genebank under the accession number NC_073027.1. The genome sequencing revealed that phage A1432 is linear double-stranded DNA (dsDNA) with a size of 61,660 base pairs (bp) and a GC content of 61.92%. The RAST annotation showed that the genome contains 79 ORFs. Among these ORFs, 42 are on the positive strand and 37 are on the negative strand. Majority of ORFs (65/79) present ATG start codons, while the remaining (14/79) start with GTG. The total length of the ORFs is 57,667 bp, with an average length of 699.41 bp, accounting for 90.74% of the whole genome. The online tRNAscan-SE tool analysis indicated the presence of a sequence encoding isoleucine-tRNA (72 bp) in the genome (Figure 3). In addition, no virulence, antibiotic resistance, or integrase genes were annotated in the A1432 genome.

The ORFs were functionally annotated by searching a non-redundant protein database using BLASTp. Out of the 79 ORFs, 31 were annotated as proteins with known putative functions, 43 were annotated as hypothetical proteins, 5 unique ORFs were annotated as hypothetical proteins with unknown functions. Based on the putative functions of the 31 ORFs encoding proteins, the genome sequence of phage A1432 can be classified into four modules including structural proteins (16 ORFs), DNA replication and modification (8 ORFs), DNA packaging (3 ORFs), lysis (3 ORFs). All annotated genes were represented in the genome-wide map (Figure 3), the specific grouping of the 79 ORFs was shown in Supplementary Table S3.

The whole genome sequence of phage A1432 showed the highest similarity with the Xanthomonas phage Xoo-sp2, a member of the Bradleyvirinae subfamily, with a nucleotide identity of 81.46% and a query coverage of 37%, indicating that the genome sequence of this phage is relatively novel.

The phylogenetic trees of TerL (ORF7) (Figure 4A), DNA polymerase I (ORF61) (Figure 4B) and proteomics (Figure 4C) showed that phage A1432 clustered with 20 phages of subfamily Bradleyvirinae, including genus Xooduovirus, Bosavirus, Elanorvirus, Docaquintavirus, Pamexvirus, Cinvestavvirus, Mallosvirus, Donnerlittchenvirus. Phage A1432 was evolutionarily related to Xanthomonas phage Xoo-sp2 (NC_052966) and Stenotrophomonas phage vB_Sm_QDWS359 (NC_073028) of the Xooduovirus genus, but formed a separate branch.

Genomic network analysis based on shared protein showed phage A1432 was found to be associated with 45 phages (dotted box) (Figure 5). Of these, 38 phages were classified into 6 virus clusters (VC_0_0, VC_1_0, VC_2_0, VC_5_0, VC_7_0, VC_35_0), 7 phages could not be classified into any virus cluster. Phage A1432 was classified into a virus cluster (VC_0_0) with the top 20 closely relative viruses.

VICTOR was used to perform phylogenetic analysis of the whole genomes of 85 phages, including those closely related to phage A1432 and 68 S. maltophilia phages (including A1432). The GBDP tree inferred based on formulas D0, D4, D6 shows average support rates of 49, 10, 53%, respectively. Therefore, formula D6 with the highest supported rate was selected for analysis. The 85 phages were grouped 64 clusters at the species rank, 31 at the genus rank, 10 at the family rank (Figure 6). Phage A1432 and the 20 subfamily Bradleyvirinae phages (according to the ICTV classification, accessed on May 6, 2024) clustered together. Phage A1432 is most closely related to S. maltophilia phage vB_Sm_QDWS359 of the Xooduovirus genus.

The VIRIDIC analysis heatmap using a dataset consistent with the VICTOR analysis showed that the 85 phages were grouped into 78 clusters at the species rank and 51 clusters at the genus rank. Phage A1432 formed a separate cluster at the genus rank (Supplementary Table S4), the ANI between phage A1432 and the phages ranged from 0 to 60.5%, with the highest ANI of 60.5% between phage A1432 and Xanthomonas phage FMYAK-P1 (NC_073032) of genus Bosavirus (Figure 7).

In addition, the VICTOR analysis of 68 S. maltophilia phages revealed that there is evidently a high degree of genetic diversity. In addition, the clustering of the 68 S. maltophilia phages was relatively scattered, including 10 clusters at the family rank, 31 clusters at the genus rank, 54 clusters at the species rank (Figure 6). These S. maltophilia phages genomes have a wide range of sizes, ranging from 5,819 bp to 319,518 bp, with many different sizes within this range. The average GC% content of the S. maltophilia phages (57.12%) is similar to that of the host S. maltophilia (66.14%) (Fang et al., 2023), although there is substantial variation in GC% ranging from 35.38% (phage TS-12) to 67.41% (phage vB_Sm_QDWS359 and phage Suso) (Figure 6). The VIRIDIC cluster showed that 68 S. maltophilia phages were grouped in 44 clusters at genus rank and 64 clusters at species rank (Supplementary Table S4). The ANI between phage A1432 and the other S. maltophilia phages ranged from 0 to 59.5%, with the highest ANI of 59.5% between phage A1432 and phage vB_SmaS-AXL_1 (Figure 7).

For further comparative genomic analysis, Xanthomonas phage Xoo-sp2 and FMYAK-P1 were selected, which showed the highest similarity to phage A1432, according to BlASTn and VIRIDIC results, respectively. Analysis using Mauve software revealed that each of the three phages had three local collinear blocks arranged in different orders (Figure 8A). Further analysis using Easyfig software showed that phage A1432 had an almost identical gene arrangement order with phage Xoo-sp2 and FMYAK-P1, with gene similarity ranging from 67 to 100% (Figure 8B). As a whole, these three phages share similar genes in DNA replication and regulation, structural proteins, DNA packaging, lysis, demonstrating that they may share similar evolutionary traits. In addition, phage A1432 contained a tRNA in its genome, which was absent from the other two phages.