Rssc and non-Ralstonia strains used in this study are listed in Table 1. R. solanacearum strains RUN302 and UW551 were used as hosts for propagation of phage RsoM2USA. The bacterium was grown overnight in casamino acid peptone glucose (CPG) medium (Hendrick and Sequeira, 1984) at 28°C from a single colony streaked from a frozen stock, and its inoculum prepared in sterile water using OD₆₀₀ as an initial measurement of cell density, followed by 10-fold serial dilution plating to confirm final inoculum concentration (Ahmad et al., 2017).

A pure phage designated RsoM2USA was isolated from soil obtained from a tomato field infested by R. solanacearum strains in Florida, United States, using R. solanacearum strain RUN302 as a bacterial host and the triple phage purification process described by Addy et al. (2019). The only difference is that CPG containing 0.35%, not 0.45%, agar was used as the top layer for the plaque assay to facilitate the isolation of the jumbo phage. The pure phage stock of RsoM2USA was also made, stored, and its titer determined using the method of Addy et al. (2019).

To characterize the morphology of the jumbo phage, phage RsoM2USA particles were treated with an equal volume of chloroform and centrifuged at 9,391 × g for 10 min at 4°C. The upper layer containing the phage particles was transferred into a new tube and purified by ultracentrifugation at 109,000 × g through a 30% sucrose cushion for 2 h at 10°C. The phage pellet was dissolved in 500 μl of SM buffer containing 50 mM Tris/HCl at pH 7.5, 100 mM NaCl, 10 mM MgSO4, and 0.01% gelatin, and used for negative staining with sodium phosphotungstate (Dykstra, 1993) before observation under a Hitachi HT7700 transmission electron microscope. At least 10 phage particles were used to estimate the phage’s morphometrics using the open source image processing program ImageJ 1.50i (Abramoff et al., 2004).

To determine the infection cycle of phage RsoM2USA, one-step growth experiment was performed based on Ellis and Delbruck (1939) with modifications. Two hundred microliters of the overnight culture of R. solanacearum strain RUN302 was transferred into 9.8 ml of CPG broth and grown at 28°C with shaking until the culture reached the OD₆₀₀ of 0.05 (5 × 10⁷ CFU/ml). Phage RsoM2USA was then added at a MOI of 1 and allowed to adsorb for 15 min at 28°C. Any non-absorbed phage particles were removed by centrifugation, followed by washing with 10 ml of CPG and centrifugation again at 6,000 × g for 5 min at 4°C. The pellet was resuspended in 10 ml of CPG, diluted 10,000-fold, and incubated at 28°C without shaking. An aliquot of 500 μl was taken every 30 min for 7.5 h, filtered through 0.45 μm membrane, diluted, and subjected to the plaque assay described by Addy et al. (2019) using RUN302 as a host to estimate phage titers. There were three replicates for each time point, and the experiment was repeated three times.

To determine the host specificity of the jumbo phage RsoM2USA, the purified phage was subjected to the spot test (Ahmad et al., 2017) using 16 Rssc and three outgroup bacterial strains (Table 1). Briefly, a double-layered CPG plate was made first by pouring a top layer containing a mixture of 3.5 ml of CPG, 0.35% agar and 250 μl of each Rssc strain (OD₆₀₀ of 0.1) on top of a solidified CPG plate containing 1.5% agar. After the top layer was hardened for 15 min, 3 μl of each of a serial dilution (10⁰–10^–6) of the phage RsoM2USA suspension (10⁸ PFU/ml) was spotted on top of the double-layer CPG plate, and incubated overnight at 28°C. The formation of plaques (lysis zones) on the plate indicated that the bacterial strain was susceptible to the phage.

To find out the lethal temperature and to determine the effect of temperature on the stability of phage RsoM2USA, a thermal stability test was conducted by incubating the phage at temperatures ranging from 4°C to 90°C as described by Addy et al. (2019). Briefly, the purified phage was diluted to 1 × 10⁸ PFU/ml in SM buffer, followed by incubation of 1 ml of the diluted phage suspension at each of the designated temperatures for 1 h. To estimate phage numbers after incubation, the phage suspension was serially diluted in SM buffer and subjected to plaque assay using R. solanacearum RUN302 as a host. There were three replicates for each temperature and the experiment was repeated once.

Phage DNA was extracted from purified phage particles using either a phenol-chloroform method (Sambrook and Russell, 2001) or the Phage DNA Isolation kit (Norgen Biotek Corp, Canada). The phage DNA was sequenced on an Illumina MiSeq with 2 × 150 bp reads, and the genome sequence assembled using Spades v3.11 commercially by SeqMatic (Fremont, California). Potential open reading frames (ORFs) larger than 50 amino acids (aa), and putative tRNAs in phage RsoM2USA were identified using PHASTER (Arndt et al., 2016), GeneMarkS (Besemer, 2001), DNASTAR (DNASTAR Inc., United States), and tRNAscan-SE 2.0¹ (Chan and Lowe, 2019). Homology searches for each identified ORF were performed using BLAST/PSI-BLAST against NCBI’s protein databases. An e-value threshold of e–4 or less was used for two proteins to be considered a match. Functional annotation and pathway identification were done using KEGG Orthology (KO) (Kanehisa et al., 2016) and Balst2go (Götz et al., 2008). Codon usage frequencies in the jumbo phage RsoM2USA and R. solanacearum strain RUN302 were calculated using the Codon Usage program available at https://www.bioinformatics.org/sms2/codon_usage.html. Complete genome sequences of 47 jumbo phages in the family of Myoviridae with a genome size more than 200 kb were downloaded from GenBank (Supplementary Table 1) and their genome sequences were compared using a dotplot generated in Gepard ver. 1.40 by calculating the similarity of the genome sequences and displaying similar DNA fragments with default parameters (word length of 10) (Krumsiek et al., 2007). The Average Nucleotide Identity (ANI) value was calculated using OrthoANI Tool version 0.93.1 to measure the overall similarity between genome sequences (Lee et al., 2016) and the heatmap was generated by choosing Color Scales for the conditional formatting in Microsoft Excel. For phylogenetic analysis, amino acid sequences of the major phage capsid protein, terminase large subunit protein, and portal vertex protein were first aligned using MUSCLE (MUltiple Sequence Comparison by Log-Expectation), followed by construction of phylogenetic trees using the Maximum Likelihood method implemented in MEGA-X (Kumar et al., 2018) version 10.0.5² with 1,000 bootstrap replications. The terminase large subunit protein tree was built for 25 jumbo phages, while the major capsid or portal vertex protein tree was constructed for 24 and 19 phages including RsoM2USA, respectively. This is because annotations for the major capsid protein in Klebsiella phage K64-1, and for the portal vertex protein in Edwardsiella phage pEtSU, Prochlorococcus phage PSSM2, and Ralstonia phages RSF1, RSL2, RP12, and RP31 were not found (Supplementary Table 2).

Purified phage particles (5 × 10¹⁰) were denatured by mixing with 4 × Laemmli sample buffer and heating at 95°C for 5 min. After cooling down on ice, 30 μl of the sample were subjected to SDS-polyacrylamide gel electrophoresis (PAGE) (12% wt/vol polyacrylamide) according to the methods of Laemmli (1970). Protein bands were visualized with Coomassie Brilliant Blue R250 stain reagent (Thermo Fisher Scientific, United States). The most abundant bands were excised from the gel and sent to ProtTech, Inc. (Phoenixville, Pennsylvania) for protein identification using liquid chromatography-tandem mass spectrometry (LC-MS/MS). The mass spectrometric data is used to search against NCBI’s most recent non-redundant protein database, and against each of the predicted protein sequences of RsoM2USA using ProtTech’s ProtQuest software suite.

Tomato plants (Lycopersicon esculentum Mill. cv. “bonnie best”) were grown, transplanted and inoculated as described previously (Ahmad et al., 2017), except that for plant inoculation, 30 ml of R. solanacearum strain RUN302 (10⁸ cells/ml) was first poured into each pot. This was followed immediately by pouring either 30 ml of phage RsoM2USA suspension (10⁸ PFU/ml) for a MOI of 1, or with water as a non-phage treatment control. Negative control plants were inoculated with 60 ml of water. Inoculated plants were rated daily using a disease index (DI) of 0–4 (Roberts et al., 1988). There were five plants for each treatment and the experiment was repeated three times.

Means of disease index between untreated (wild type) and jumbo phage RsoM2USA-treated R. solanacearum RUN302 strains were analyzed for significant differences using the t-test in Microsoft Excel.

The complete genome sequence of Ralstonia phage RsoM2USA was deposited to GenBank under the accession number MG752970 (Supplementary Table 1). The accession numbers for the genome and megaplasmid sequences of R. solanacearum strain IBSBF1503 (alternative ID of RUN302) are CP012943.1 and CP012944.1, respectively.

A phage was isolated from a soil sample collected from a Rssc-infested tomato field in Florida, United States. The phage produced small and clear plaques with a diameter of approximately 1–2 mm on the top layer containing 0.35% agar of a double layered CPG plate using Rssc strain RUN302 as a host. The phage has an icosahedral head of 142 ± 7 nm (n = 10) in diameter, and a long tail with a length of 125 ± 5 nm (n = 10) (Figure 1). The phage also has a baseplate and tail fibers of approximately 70 nm in length (Figure 1B). Since the morphology of the phage is typical for members of the family Myoviridae, the phage was designated Ralstonia phage RsoM2USA by using our systematic phage naming approach (Ahmad et al., 2018), since it is the second Rssc strain-infecting phage belonging to the family Myoviridae that was isolated from the United States after Ralstonia phage RsoM1USA (Addy et al., 2019). Compared to RsoM2USA, phage RsoM1USA has a much smaller icosahedral head of 63 nm × 66 nm but a longer contractile tail of 152 nm in size (Addy et al., 2019).

Sixteen Rssc strains originally isolated from different geographic regions of the world belonging to different biovars/phylotypes/sequevars were tested for their susceptibility to Ralstonia phage RsoM2USA. Phage RsoM2USA infected 14 of the 16 tested Rssc strains in all tested biovars (1, 2, 2T, 3, and 4) and in all the three different Ralstonia species—R. solanacearum, R. pseudosolanacearum, and R. syzygii strains originated from different countries, indicating a wide host range of the phage (Table 1). The phage, however, did not infect tested Xanthomonas campestris and Pseudomonas syringae strains, indicating its specificity to the three Ralstonia species (Table 1). RsoM2USA has a lytic infection cycle, which was determined to be 360 min, with a latent period of 270 min, the longest latent period determined so far for jumbo phages (Monson et al., 2011; Abbasifar et al., 2014; Yoshikawa et al., 2018), followed by a 90-min rise period with a burst size of 32 ± 3 particles per infected cell (Figure 2A). A similar burst size of approximately 30 was found for jumbo phage XacN1, although its latent period was determined to be 90 min and growth cycle completed within 240 min (Yoshikawa et al., 2018). The phage was stable from 4 to 40°C, since no significant difference in its titer was found under this temperature range (Figure 2B). Significant loss in the phage titer, however, was observed at 50 and 60°C, and no phage particles were detected after the phage was incubated at 70, 80, and 90°C for 1 h, suggesting that the lethal temperature for the phage is approximately 70°C (Figure 2B).

The complete genome of the Ralstonia phage RsoM2USA was determined to be 343,806 bp in size (GenBank accession no. MG752970), resulting in the classification of RsoM2USA as a jumbo phage. Such genome size also makes RsoM2USA the largest Ralstonia phage sequenced and reported so far, since it is larger than previously reported jumbo Ralstonia phages RSF1, RSL1, RSL2, RP31, and RP12 with their genome sizes ranging from 222,888 to 279,845 bp (Yamada et al., 2007, 2010; Bhunchoth et al., 2016; Matsui et al., 2017). RsoM2USA also represents the third largest phage infecting plant pathogenic bacteria after Agrobacterium virus Atuph007 and Xanthomonas phage XacN1, and the 23rd largest phage reported to date (Buttimer et al., 2017; Casey et al., 2017; Yuan and Gao, 2017; Attai et al., 2018; Yoshikawa et al., 2018; Al-Shayeb et al., 2020). The G + C content of the RsoM2USA genome is 41%, significantly lower than that of its Ralstonia host genomes (e.g., G + C content of 67% in Rssc strain IBSBF1503 (accession number of NZ_CP012943.1 in GenBank) and 66.97% in strain GMI1000 (NC_003295). A total of 486 potential open reading frames (ORFs) were identified (Supplementary Figure 1 and Supplementary Table 3). Among them, 239 had no significant similarity with any of the protein sequences in the searched databases, 167 were annotated as conserved, conserved hypothetical or unnamed proteins with no assigned functions, and only 80 were predicted proteins shared homology with other phages or bacteria with assigned functions based on KEGG analysis (Table 2 and Supplementary Table 3). This is similar to other jumbo phages, in which even though many more proteins were predicted from the genomes of jumbo than from smaller phages, the majority of the jumbo phages’ proteins has no matches in the current databases with undiscovered functions that prevented detailed comparisons among jumbo phages.

Based on ICTV’s 2020 release of virus taxonomy,³ the family of Myoviridae consists of 8 subfamilies (Emmerichvirinae, Eucampyvirinae, Gorgonvirinae, Ounavirinae, Peduovirinae, Tevenvirinae, Twarogvirinae, and Vequintavirinae) with 64 genera and 294 species (Kuhn et al., 2013). In addition, 153 genera with 331 species are classified directly under the family of Myoviridae for a total of 217 genera and 625 species. In addition to Xanthomonas phage XacN1 and Serratia phage PCH45, species-undefined jumbo phages in Myoviridae, only 44 of the genera in Myoviridae contain jumbo phages with genome sizes more than 200 kb (Hendrix, 2009; Supplementary Table 1). Genomic relationships between Ralstonia jumbo phage RsoM2USA and 46 representative jumbo phages in the family of Myoviridae was therefore determined using the whole genome dot plot (Supplementary Figure 2). This was done by comparing the genome sequences of RsoM2USA with Xanthomonas phage XacN1, Serratia phage PCH45, and representative phages from each of the 44 jumbo phage-containing genera of Myoviridae, including five previously reported Ralstonia jumbo phages RSF1 and RSL2 (Bhunchoth et al., 2016), RSL1 (Yamada et al., 2010), RP12 and RP31 (Matsui et al., 2017; Supplementary Figure 2). Our dot plot result indicated that RsoM2USA is a novel phage because RsoM2USA did not show similarity pattern (indicated by absence of diagonal lines) with any of the 46 phages including the five Ralstonia jumbo phages (Supplementary Figure 2). Further analysis using the Orthologous Average Nucleotide Identity (OrthoANI) value revealed that phage RsoM2USA shared OrthoANI values ranging from 56.16 to 63.28% to only 14 of the 46 jumbo phages including Ralstonia virus RSL1 (Mieseafarmvirus) and Xanthomonas phage XacN1 (Figure 3 and Supplementary Table 4). The fact that the shared OrthoANI values of RsoM2USA with the 14 jumbo phages are less than 95% suggests that the jumbo phage RsoM2USA belongs to a new species in Myoviridae, since the OrthoANI value of 95% is used for demarcation of species (Lee et al., 2016).

To determine evolutional relationships between RsoM2USA and other Ralstonia and non-Ralstonia jumbo phages, phylogenetic analysis was conducted for RsoM2USA, five other Ralstonia jumbo phages in Myoviridae, and 19 jumbo phages representing all species with OrthoANI values of more than 56% (Figure 4 and Supplementary Tables 2, 4). Since jumbo phages are highly divergent, no set of genes is present in all phages. Three “signature gene products” including the major capsid protein, terminase large subunit protein and portal vertex protein were used in this study, since they are conserved in many jumbo phages and generally used for phylogenetic analysis of jumbo phages (Attai et al., 2018; Yoshikawa et al., 2018). A phylogenetic tree based on the predicted amino acid sequences of major capsid proteins showed that phage RsoM2USA is more closely related to Xanthomonas phage XacN1, but not to any of the five Ralstonia and the other 19 non-Ralstonia jumbo phages (Figure 4A). The phylogenetic tree based on the predicted amino acid sequences of terminase large subunit or portal vertex proteins also consistently placed RsoM2USA in the same clade with Xanthomonas phage XacN1, not with any other Ralstonia and non-Ralstonia jumbo phages used for the comparison (Figures 4B,C). Ralstonia virus RSF1 and Ralstonia virus RSL2 in genus Chiangmaivirus were group together in the same clade, so were Ralstonia virus RP12 and Ralstonia virus RP31 in genus Ripduovirus (Figures 4A,B). The four Ralstonia jumbo phages in the two genera were more closely related to each other than to Ralstonia virus RSL1 in genus Mieseafarmvirus, and to RsoM2USA, as revealed by phylogenetic trees based on the predicted amino acid sequences of the major capsid and terminase large subunit proteins, respectively (Figures 4A,B). Major capsid protein and terminase large subunit protein trees all support the grouping of similar phage clusters according to the relatedness of the jumbo phages (Figures 4A,B). One cluster is well-known containing Rak2-like phages previously reported by Attai et al. (2018) and Yoshikawa et al. (2018) that include Klebsiella virus Rak2 (and K64-1 in Figure 4B) (genus Alcyoneusvirus), Escherichia phages 121Q and PBECO4 (genus Asteriusvirus), Serratia virus BF and Yersinia virus Yen9-04 (genus Eneladusvirus), as well as Cronobacter virus GAP32, and Pectinobacterium virus CBB (genus Mimasvirus). A cluster consisting of Vibrio phages nt1, ValKK3, and KVP40 formed another cluster for jumbo phages belonging to the genus Schizotequatrovirus in the subfamily of Tevenvirinae. The portal vertex protein tree also supported similar grouping of phage clusters, although missing the cluster of Ralstonia jumbo phages in genera Chiangmaivirus and Ripduovirus, since no annotation for portal vertex protein has been found for the four Ralstonia jumbo phages. Interestingly and unexpectedly, RsoM2USA is grouped in the same cluster as the species-undefined Xanthomonas phage XacN1 by all three trees (Figure 4), suggesting that the two jumbo phages are distantly related to the Rak2-like phages as previously found for Xanthomonas phage XacN1 (Yoshikawa et al., 2018) and prompting a proposal of a new genus for the two jumbo phages. Comparisons of the general characteristics of the two phages and their best hit protein homologs are summarized in Supplementary Table 5.

Eighty of the 486 ORFs were predicted to function in the following categories: replication, recombination and repair; translation, ribosomal structure and biogenesis; transcription; posttranslational modification, protein turnover, and chaperones; nucleotide transport and metabolism; amino acid transport and metabolism; signal transduction mechanisms; cell wall/membrane/envelope biogenesis; structure; integral component of membrane; coenzyme transport and metabolism; as well as mobilome: prophages and transposons (Table 2). BLASTn and manual searches for the shell protein and tubulin genes in the RsoM2USA genome yielded no orthologs to those in Pseudomonas and Serratia jumbo phages (Chaikeeratisak et al., 2017a,b; Malone et al., 2020), although orthologs were found in other Ralstonia jumbo phages including Ralstonia virus RP12, Ralstonia virus RP31, Ralstonia virus RSL2, and Ralstonia virus RSF1, suggesting that RsoM2USA may use a different strategy to evade CRISPR-cas DNA targeting by its bacterial host.

To study the effect of the jumbo phage RsoM2USA on the virulence of its susceptible Rssc strain RUN302, we compared the virulence of the wild type RUN302 to that of the phage-treated RUN302. The wild type strain RUN302 started to wilt tomato plants 6 days after inoculation with an average DI of 0.4 + 0.2 (n = 15), and 14 of the 15 inoculated plants were completely wilted 2 weeks after inoculation (Figure 6). When tomato seedlings were inoculated with the jumbo phage RsoM2USA-treated RUN302 strain, however, only two of the 15 inoculated plants displayed wilt symptom with an average DI of 0.13 + 0.23 even 15 days after inoculation (Figure 6).

Our virulence result is similar to those of Ralstonia jumbo phages RSL1 (Fujiwara et al., 2011) and RSF1 (Bhunchoth et al., 2016), as well as non-jumbo phages P4282 (Tanaka et al., 1990), PE204 (Bae et al., 2012), and ϕRSM3 (Addy et al., 2012), showing either total loss of virulence (RSL1, PE204, and ϕRSM3) or reduced virulence (RSF1 and P4284) of their susceptible Ralstonia strains in tomato or tobacco (P4282) plants under greenhouse conditions. Recently, three lytic phages were isolated from environmental water in Spain, and the phages either alone or in combination were found effective to control diseases caused by R. solanacearum (Álvarez et al., 2019). The effect of the other three Ralstonia jumbo phages (RP12, RP31, and RSL2) on virulence, however, remains unknown (Bhunchoth et al., 2016; Matsui et al., 2017). Different from RsoM2USA, Ralstonia phage RsoM1USA (Addy et al., 2019) had no significant effect on disease symptoms, although both are myoviruses and isolated from soil in Florida, United States, suggesting that different phages may play different ecological roles in the environment.

Ralstonia phages isolated from Japan, Korea and Thailand are only known to be active against R. pseudosolanacearum (Álvarez and Biosca, 2017; Addy et al., 2019), so it is hard to assess their biocontrol potential to other Ralstonia species. Recently, Ralstonia phage RsoP1EGY from Egypt has been found to be specific to only the race 3 biovar 2 strains of R. solanacearum (Ahmad et al., 2018), while RsoP1IDN from Indonesia (Addy et al., 2018), RsoM1USA from United States (Addy et al., 2019), and the three lytic phages from Spain (Álvarez et al., 2019) are all active against both R. solanacearum and R. pseudosolanacearum. The ability of jumbo phage RsoM2USA to significantly reduce the virulence of its bacterial host R. solanacearum strain RUN302, the wide host range RsoM2USA displayed, and the specificity of RsoM2USA against R. solanacearum, R. pseudosolanacearum, and R. syzygii make it worthy of future study to determine its potential as biocontrol, either alone or in combination with other compatible phages, against bacterial diseases caused by Rssc strains.