C. indologenes strain MA9 (accession no. CP075170), isolated from rotten roots of P. notoginseng, was provided by professor Ji of Yunnan Agricultural University and was used as a host in this study. Samples for screening phages were collected from different regions as follows: (1) The rhizospheric soils of healthy and rotten P. notoginseng were collected from Pingyuan Street, Yanshan County, Wenshan Prefecture, Yunnan Province in November 2020; (2) the rotten roots of P. notoginseng and P. notoginseng root soils were collected from Shilin Yi Autonomous Region, Kunming City, Yunnan Province in January 2021; (3) the soils of different vegetables and fruit trees were collected from the farmland near Kunming University of Science and Technology, Chenggong District, Kunming City, Yunnan Province in January 2021; and (4) tewage samples were collected from the First People’s Hospital of Yunnan Province in February 2021. C. indologenes MA9 was cultured in an optimized nutrient agar (NA) medium (10 g peptone, 5 g sodium chloride, and 3 g beef extract) at 160 rpm at 28°C.

First, we tried to isolate phages from the rhizospheric soil mixtures of healthy and decayed P. notoginseng infected with C. indologenes MA9; however, we could not find any phage. Subsequently, under the same culture conditions, we attempted to use other samples such as rotten roots, P. notoginseng root soils, and soils of various fruits and vegetables. However, the targeted phage was also not found. Finally, we isolated a C. indologenes strain MA9 phage from sewage samples of the First People’s Hospital of Yunnan Province, China, using the virus enrichment and double-layer agar plate methods (Pajunen et al., 2000). In brief, samples for phage isolation were filtered through a 0.22-μm filter; then 50 mL filtrate was added to a 250 mL conical flask containing 50 mL SM buffer (100 mM NaCl, 10 mM MgSO₄, 50 mM Tris–HCl at pH 7.5, and 0.01% gelatin) and kept static for 12 h. Subsequently, 40 mL suspension, 30 mL of SM buffer, 25 mL of NA liquid medium, and 2 mL of C. indologenes MA9 cultures were mixed and incubated overnight on a horizontal rotary shaker (160 rpm) at 28°C. The mixture was transferred to a sterile 50 mL centrifuge tube and centrifuged at 15,000 × g for 15 min at 4°C. Subsequently, the supernatant was filtered with 0.22-μm microporous membranes. In total, 5 mL filtrate was taken and mixed with 50 mL C. indologenes MA9 host cell cultures during the period of exponential growth phase [optical density at 600 nm (OD₆₀₀) =0.6–0.8] and then repeated 2–3 times to enrich phage MA9V-1. Successively, 200 μL of phage stock was serially diluted with SM buffer, and different dilutions were mixed with 300 μL of C. indologenes MA9, which were added to 4.5 mL semi-solid NA medium after being incubated at 28°C for 20 min, and finally, the solution was finally poured onto the surface of the NA solid medium. A single clear plaque was picked from NA double-layer agar plate incubated overnight at 28° C, placed into 1 mL of SM buffer, and vortexed vigorously before filtering through a 0.22-μm membrane. The above steps were repeated 2–3 times to obtain pure phage MA9V-1 particles, and the phage MA9V-1 titer was measured to be 10⁹ PFU/ml by the double-layer agar plate method. The isolates were stored at 4°C.

The optimal multiplicity of infection (MOI) is the ratio of phages to bacteria at the time of infection. The experimental method followed was according to the study mentioned in the reference (Sun et al., 2022). In brief, the CFU/ml of the host cell at the exponential growth phase was counted through the dilution coating method with serial dilutions of 10⁻⁴, 10⁻⁵, and 10⁻⁶. Phage stock of known titer and host cells at the exponential stage were mixed with a gradient of MOI ranging from 0.001 to 10 and cultured in a shaker (180 rpm/min) at 28°C for 12 h. The enrichment was centrifuged at 15,000 × g at 4°C for 15 min and filtered through a 0.22-μm filter membrane, and the phage titer was determined by the double-layer agar plate method with three replicates.

Adsorption rate, a value of phage particles adsorbed to the host cell at a particular time, was determined with minor modifications, as previously described (Sasikala and Srinivasan, 2016). In brief, 9 mL of C. indologenes MA9 cultures with absorbance of OD₆₀₀ at 0.6–0.8 was infected with 1 mL phage dilution (10⁶ PFU/ml) at multiplicity of infection of 0.01 and transferred to a sterile 50 mL conical flask to incubate at 28°C. Approximately 100 μL of culture was taken out at 4-min intervals (0, 4, 8, 12, 16, and 20 min) and added to sterile 2 mL tubes with 1.9 mL of pre-cooled NA medium, vortexed for 10–15 s, and then centrifuged at 15000 × g at 4°C for 15 min, to remove the cell-phage complexes. The titer of free phage in the supernatant was determined by the double-layer overlay method and was repeated three times (Stuer-Lauridsen et al., 2003).

To determine the parameters of the one-step growth curve, the protocol, as described earlier, was adopted (Wu et al., 2007) and modified according to the characteristics of the phage. C. indologenes MA9 was cultured up to the exponential growth phase (10⁸ CFU/mL), and 10 mL of culture was collected by centrifugation (11,000 × g, 8 min, 4°C) for harvesting, and the precipitate was resuspended with 2 mL of fresh NA liquid medium. Phage was added to the suspension at MOI = 0.01 and incubated at 28°C for 15 min without shaking. The mixture was, then, centrifuged at 11,000 × g for 8 min to remove the free phage particles, and the pellet was suspended with fresh NA medium to culture in a shaker at 180 rpm at 28°C after washed twice with 10 mL nutrient agar medium. In total, 200 μL of culture was taken out every 10 min until 130 min and mixed with 800 μL SM buffer. Finally, the mixture was centrifuged at 15,000 × g for 3 min at 4°C. The titers of the samples were immediately measured by the double-layer agar method. The burst size was calculated according to the following definition: Burst size is the ratio between the number of phage particles at the end of lysis and the number of the host cells at the beginning of the infection (Li et al., 2021). All assays were repeated in three independent experiments.

For determining the thermal and pH stability of phage MA9V-1, the experimental protocol was performed as proposed with slight modification (Sasikala and Srinivasan, 2016; Sun et al., 2022). In brief, 1 mL phage MA9V-1 lysate (10¹⁰ PFU/m) was treated under different temperatures from 30 to75°C at 5°C intervals for 1 h. Then, these treated samples checked the survival rate of phage MA9V-1 using the double-layer agar assay. Similarly, 100 μL of phage stock (10¹⁰ PFU/ml) was added to 900 μL different buffers of pH ranging from 3 to11 (component of citrate sodium, potassium dihydrogen phosphate, Tris–HCl, and sodium carbonate) for incubation of 1 h at 30°C. After treatment, the titer of phage MA9V-1 was evaluated using the same procedure. All assays were performed in triplicate.

The host range of Chryseobacterium phage MA9V-1 was tested against 11 different bacterial strains using the standard spot testing assay, including Chryseobacterium indologenes (n = 7), Bacillus cereus (n = 1), Pseudomonas syringae (n = 1), and Escherichia coli (n = 1), of which Chryseobacterium indologenes ATCC 29897, Pseudomonas syringae CGMCC 1.3070, and Escherichia coli ATCC 11303 were type strains purchased from Strain Preservation Center, while the others were isolates. Each 300 μL of bacterial cells was mixed with 200 μL of 10-fold series gradient dilutions (10⁻¹–10⁻⁷) of phage MA9V-1 suspension (10¹⁰ PFU/ml), then adding all mixtures to 4.5 mL semi-solid medium before pouring into the surface of double-layer agar plates, respectively, and incubated at 28°C overnight in triplicate.

The morphology and structure of pure phage particles were visualized using transmission electron microscopy (TEM). The phage particles purified with the density gradient centrifugation method were loaded on a carbon-coated copper grid to adsorb for 10 min, followed by negative staining with 1% phosphotungstic acid. The stained phage MA9V-1 was observed using transmission electron microscopy (Hitachi HT7820) at 120 kV (Johnson et al., 2011).

To improve the purity of the phage genome, 2.5 μL DNase I (1 U/μl) and 0.5 μL RNase A (1 mg/mL) were added to 500 μL phage stock to remove the host cell nucleic acid, and further experimental operation was performed using the phenol–chloroform protocol, as previously described in the study mentioned in the reference (Chen et al., 2019). Nucleic acid concentration of phage was determined by NanoDrop (Shanghai Baoyude Scientific Instruments Co., Ltd., BIO-DL, Shanghai) and imaged by 1% agarose gel electrophoresis. Sequencing of the phage genome was performed by the Illumina_Novaseq_PE150 platform (Shenzhen Huada Intelligent Manufacturing Technology Co., Ltd., Illumina, Shenzhen). First, phage genome was randomly broken into approximately 300–500 bp fragments using the Covaris M220 (Shanghai Tusen Vision Technology Co., Ltd., Shanghai, China) DNA fragmentation instrument, discarding adapters and trimming sequences with a length of less than 50 bp. Then, using the BBMap version 38.51 (Roy and Chanfreau, 2020),¹ comparisons were performed with the existing NCBI database to sift out the corresponding rRNA, host, and other bacterial genome sequences, to purify the phage genome. The assembly of clean reads was performed using the “De novo assembly” option of SPAdes v3.5.0 software (Prjibelski et al., 2020).² Subsequently, contigs assembled with reads were comparatively analyzed with BLAST (v2.10.0+)³ (Camacho et al., 2009) to virus-NT, NCBI nt, and viral RefSeq databases to match the most similar gene reference sequences. Finally, the assembled phage MA9V-1 genome was annotated using Proksee (Grant and Stothard, 2008).⁴ Functional protein bioinformatic annotation of ORFs was predicted manually using HHpred (Soding et al., 2005).⁵ The virulence-associated and antibiotic resistance gene potentially existing in the phage MA9V-1 genome was analyzed through the virulence factor database (Liu et al., 2019)⁶ and the comprehensive antimicrobial research database.⁷ The presence of transfer RNA genes was assessed using tRNAscan-SE (Lowe and Chan, 2016).⁸ The phage MA9V-1 genomic data were aligned by Clustalw v2.1 (Peng et al., 2022)⁹ and analyzed using the visualization bioinformatics software Easyfig v2.2 (Peng et al., 2022).¹⁰ MEGA v11 (Tamura et al., 2021)¹¹ was used to construct a phylogenetic tree for determining phylogenetic relationships between phage MA9V-1 and other strains. Furthermore, phage proteome phylogenetic tree was constructed by viral proteomic tree (ViPTree) server v1.9 (Nishimura et al., 2017).¹²

Approximately 7 mL of phage particle suspension with titer of 10⁹ PFU/ml purified by the polyethylene glycol (PEG) precipitation method was ultracentrifuged with CsCl density gradient centrifugation (HIMAC CP-100WX Ultracentrifuge, HITACHI, Japan) at 4°C at 150,000 × g for 8 h and extracted the separated layer of phage with a sterilized syringe. Then, 20 μL of ultra-purified phage particles was mixed with 2× SDS loading buffer, heated in boiling water for 15 min, and subjected to SDS-PAGE [12% (w/v) polyacrylamide] (Uchiyama et al., 2014).

The experimental material was the root of a healthy 2-year-old P. notoginseng plant collected from Shilin Yi Autonomous Region, Kunming City, Yunnan Province. After a series of pre-treatments, the slices of P. notoginseng root were placed onto a plate covered with wetted filter paper. In total, 1 mL of 24-h cultures of C. indologenes strain MA9 was transferred to a 50 mL conical flask and grown at 28°C with shaking until the culture reached the OD₆₀₀ of 0.6–0.8. Phage MA9V-1 and C. indologenes MA9 were mixed with optimal multiplicity of infection (MOI) of 0.001, 0.01, 0.1, 1, and 10, respectively, and incubated at 28°C for 10 min. In total, 100 μL of different MOI mixtures were was added to the root slices, sealed with tape, and cultured in an incubator at 26°C. This experiment was divided into seven groups with eight parallels in each group. Among them, the negative control was the C. indologenes MA9 reaching the exponential stage, whereas the positive control was the inactivated C. indologenes MA9 at the log phase after treating with the autoclave sterilization method. Under the same experimental condition, 100 μL of the mixture of different MOI was added to the experimental group at 24-h intervals, the negative control and positive control were added to the equal volume of treatment liquid, and the number of infected and uninfected P. notoginseng plants was counted after 7 days. In this study, one condition for determining root rot was the rotting and softening of the root interior. The disease incidence was calculated as the percentage of rotten roots to the total number of parallel experiments. The experiment was performed in triplicate.

To the best of our knowledge, phage MA9V-1 (formal name: vB_CinP_MA9V-1) of the C. indologenes strain MA9 is the first phage to have bacteriolytic activity against C. indologenes MA9. The results suggested that the phage MA9V-1 can form clear and round plaques with a diameter of approximately 0.5–1.5 mm on lawns of C. indologenes MA9 of the double-layered plate (Figure 1A). The morphological characteristics of phage MA9V-1 (Figure 1B) revealed that it has an icositetrahedron-coated protein “head” with a diameter of approximately 121.26 nm and a contractible tail length of approximately 170.34 nm with a width of approximately 44.62 nm. In addition, it also has a relatively short “neck” compared with other myoviruses such as Ralstonia phage RsoM1USA, a novel species of P2virus (Addy et al., 2019).

As shown in Figure 2A, phage MA9V-1 had the highest titer and was 1.25 × 10¹⁰ PFU/ml at an MOI of 0.001. The absorption curve revealed that phage MA9V-1 was capable of rapidly adsorbing onto the surface of the host cell membrane. Within 8 min, >75% of the phage particles had adsorbed, and after 12 min, approximately 90% were adsorbed, indicating that phage MA9V-1 had the ability of high adsorption efficiency (Figure 2B). The latent period, rise period, and burst size are three critical parameters to characterize the phage lifecycle; the results of the one-step growth curve suggested that these values are 20 min, 90 min, and 10 PFU/cell, respectively (Figure 2C). In the thermal stability experiment, Figure 2D shows that the temperature between 30°C and 50°C had no negative effect on phage MA9V-1 after a 1-h incubation. Even at 55°C and 60°C, the phage MA9V-1 also has the ability to infect C. indologenes MA9, but its viability is decreased to approximately 85 and 40% at 30°C, respectively. The infectivity of phage MA9V-1 was completely lost at 65°C and subsequent temperature of 70 and 75°C. Meanwhile, phage MA9V-1 had nearly no obvious variation by treating under different pH levels at 5 to 8 but showed strong sensitivity under extreme pH conditions and will be inactivated to no infectivity (Figure 2E). In addition, of 11 different bacterial cells examined, only 3 C. indologenes strains were found to be sensitive to the phage MA9V-1, namely Chryseobacterium indologenes MA9, Chryseobacterium indologenes 02, and Chryseobacterium indologenes 06, respectively (Table 1). When Chryseobacterium indologenes MA9 was considered as the host strain, phage MA9V-1 had the highest titer approximately two times than others.

Phage genomic analysis is crucial to determine the specific functional proteins in the genome and the safety of phage application. The results showed that the genome sequence of phage MA9V-1, a linear double-stranded DNA (dsDNA), was 213,507 bp in length and contained 35.99% of G + C content (Figure 3). The phage MA9V-1 genome contained 263 putative open reading frames (ORFs); among them, 257 and 6 ORFs were identified in the forward and reverse strands, respectively (Table 2). The lengths of the longest and shortest ORF gene sequences were 7,122 bp and 90 bp, encoding one hypothetical protein with 2,374 amino acids and one hypothetical protein with 30 amino acids, respectively. According to the BLAST results, 39 ORFs had homologs to genes encoding known functional proteins, and 14 of these clustered into 3 modules related to aspects such as phage structure, host lysing, and DNA replication, and the remaining 25 genes encoded other biofunctional proteins. The absence of integrase in the genome indicated that phage MA9V-1 was a virulent phage targeting C. indologenes MA9 not a lysogenic phage. In addition, concerned tRNA gene, lysogenic gene, drug resistance gene, and virulent genes were not found in the phage MA9V-1 genome, indicating that the phage MA9V-1 depended on host translation and satisfied several recommended criteria for safe usage in the prevention and control of C. indologenes MA9 (Bruessow, 2012; Han et al., 2017).

The phage MA9V-1 proteome is comprehensively analyzed through the ViPtree database and selected related genomes, which contain 2,201 virus families and 4,782 host groups. As shown in Figure 4A, a circular proteomic tree of phage MA9V-1 showed that its genome is located in the “others” gray module whatever for “virus family” or “host group.” In addition, we constructed a rectangular proteomic tree (Figure 4B), consists of 11 myovirus genomes, which revealed that the phage MA9V-1 was related to Sphingomonas phage PAU (accession no.NC_019521) and was branched together with Tenacibaculum phage PTm1 (accession no.NC_049340) and Tenacibaculum phage PT24 (accession no.NC_049383) (White and Suttle, 2013; Kawato et al., 2020). Moreover, comparative genomic analysis was performed with genome of phage MA9V-1 and other three similar phages to analyze genomic evolutionary relationship (Figure 4C). The result indicated that MA9V-1 was more closely related to PAU than others, showing a similar region between them mainly located in 7,970–40,762 nt and 172,015–200,768 nt region of MA9V-1 genome with a low similarity (< 50% in most regions), whereas other regions share no similarity with PAU (see Figure 4).

Structural protein expression analysis of purified phage particles by SDS-PAGE gel electrophoresis showed at least 15 bands of protein molecular weight ranging from 20 to 260 kDa. The protein bands were mainly distributed in 75–135, 45–65, and 25–35 kDa regions of SDS-PAGE gel, which were likely related to structural proteins, terminal family proteins, DNA replication proteins, and hypothetical proteins. According to SDS-PAGE results, we only obtained limited knowledge, and whole-genome analysis of phage MA9V-1 was required to better elucidate the role of all proteins (Lu and Breidt, 2015).

Based on the results of the proteomic tree, we constructed a phylogenetic tree analysis with four representative proteins and aimed to accurately validate the taxonomic classification of phage MA9V-1. Phylogenetic relationships were analyzed among phage MA9V-1 and other 15 members based on the amino acid sequences of four highly conserved proteins, namely terminase family protein (ORF262), DNA polymerase II (ORF135), capsid assembly protein (ORF149), and phage tail sheath protein (ORF001), respectively (Figure 6). The results are presented in Figures 6A,B,D. Phage MA9V-1 was more closely related to Sphingomonas phage PAU taking as “tree root” than others and grouped into a clade of an independent branch. Similar tree was also obtained using the deduced amino acid sequences of capsid assembly protein ORF149; we can observe that phage MA9V-1 was branched with Tenacibaculum phage pT24, which shared the similar results with above three proteins (Figure 6C). Taken together, phage MA9V-1 is similar to Sphingomonas phage PAU, Tenacibaculum phage PTm1, and Tenacibaculum phage pT24, but there is low identity between them, which was consistent with the conclusion of the proteomic tree analysis, indicating that it is a novel phage against C. indologenes MA9. According to the latest guidelines of the International Committee on Classification of Viruses (ICTV),¹³ in March 2021 (Turner et al., 2021), we only concluded that phage MA9V-1 is a new myovirus.

To investigate the growth inhibitory effect of phage MA9V-1 on the host strain, C. indologenes MA9, we first determined the absorbance value at OD₆₀₀ of C. indologenes MA9 with and without phage MA9V-1 for 13 h (Figure 7A). In the absence of phage MA9V-1 (MOI = 0), the bacteria reached a stable phase after 8 h of culture, showing a standard “S” growth curve. After adding different titers of phage MA9V-1 into C. indologenes MA9 cultures, as shown in Figure 7A, the curve at MOI of 0.001 and 0.01 was not significantly inhibited, until 5 h; the curve showed a downward trend and then regrew relatively rapidly. While using an MOI of 10, an opposite phenomenon to the earlier situation was observed, wherein the bacterial growth sharply reduced and then regrew slowly for 9 h. Meanwhile, the remaining MOIs inhibited the growth for 10 h. Overall, after treatment with phage MA9V-1 at different MOIs, the growth of C. indologenes MA9 could be inhibited. Therefore, we took advantage of phage therapy experiments to roughly estimate the therapeutic effectiveness of phage MA9V-1 on P. notoginseng root slice infected with C. indologenes MA9 for practical purposes (Figure 7B). Among them, the enlargement picture of second sample of MOI = 0.01 and fifth sample of MOI = 10 more intuitively reflects the difference between healthy and rotten samples (Supplementary Figure S1). According to the results shown in Figure 7C, we can observe that exceeded 85% of the root samples of P. notoginseng had been rotted in the negative control, which verified the strong pathogenicity of C. indologenes MA9. While all groups treated with phage MA9V-1 at different MOIs had samples with only slight or even no symptoms, especially at MOIs of 0.1 and 1, both of them decreased the incidence of disease of P. notoginseng root slice up to 50% or more and delayed the occurrence of root rot disease. Preliminary exploratory experiments suggested that the phage MA9V-1 is a good candidate for controlling root rot disease as its inhibition in the development of disease symptoms caused by C. indologenes MA9.