Comparative Genomic Analyses of the Genus Photobacterium Illuminate Biosynthetic Gene Clusters Associated with Antagonism

The genus Photobacterium is known for its ecophysiological versatility encompassing free-living, symbiotic, and pathogenic lifestyles. Photobacterium sp. CCB-ST2H9 was isolated from estuarine sediment collected at Matang Mangrove, Malaysia. In this study, the genome of CCB-ST2H9 was sequenced, and the pan-genome of 37 Photobacterium strains was analysed. Phylogeny based on core genes showed that CCB-ST2H9 clustered with P. galatheae, forming a distinct clade with P. halotolerans, P. salinisoli, and P. arenosum. The core genome of Photobacterium was conserved in housekeeping functions, while the flexible genome was well represented by environmental genes related to energy production and carbohydrate metabolism. Genomic metrics including 16S rRNA sequence similarity, average nucleotide identity, and digital DNA–DNA hybridization values were below the cut-off for species delineation, implying that CCB-ST2H9 potentially represents a new species. Genome mining revealed that biosynthetic gene clusters (BGCs) involved in producing antimicrobial compounds such as holomycin in CCB-ST2H9 could contribute to the antagonistic potential. Furthermore, the EtOAc extract from the culture broth of CCB-ST2H9 exhibited antagonistic activity against Vibrio spp. Intriguingly, clustering based on BGCs profiles grouped P. galatheae, P. halotolerans, P. salinisoli, P. arenosum, and CCB-ST2H9 together in the heatmap by the presence of a large number of BGCs. These BGCs-rich Photobacterium strains represent great potential for bioactive secondary metabolites production and sources for novel compounds.


Introduction
Photobacterium (Vibrionaceae, Gammaproteobacteria) comprises a group of Gramnegative, facultative-aerobic, mostly halophilic, and motile bacteria [1]. The genus currently consists of 37 validly published named species (www.bacterio.net (accessed on 19 May 2022)), with P. phosphoreum as the type species [2]. All species of Photobacterium were originally thought to be luminescent. However, later studies found that only strains of P. angustum, P. aquimaris, P. ganghwense, P. kishitanii, P. leiognathi, P. mandapamensis, and P. phosphoreum display this characteristic [1,3]. Although members of this genus are ubiquitous in marine contexts, isolations of Photobacterium strains from non-marine habitats such as a saline lake, the rhizosphere of a terrestrial weed, and spoiled meat have been reported [4][5][6][7]. Photobacterium exhibits different lifestyles: some occur as free-living in seawater and sediments; others function as symbionts of the light organs of marine fish and squid, decomposers of dead fish, or pathogens for marine animals [1,8,9]. High diversity within Photobacterium has been reported, which could be linked to the lifestyles of the species [10,11]. Photobacterium has been studied for ecophysiological traits such as piezophilic, salt adaptation, bioluminescence, and motility [3,12,13].
Photobacterium also encompasses species that synthesize various secondary metabolites, including those with antibiotic activity [14]. Wietz et al. reported that a pyrrothine Genomic islands play an important role in bacterial genome evolution and adaptation as they are the probable horizontal origin of genes for pathogenicity, symbiosis, and metabolism [31]. A total of 13 genomic islands ranging in size from 10.2 to 35.8 Kb were detected in the CCB-ST2H9 genome (Table S2). The genomic islands contain a considerable number of hypothetical genes, as well as genes related to energy, cell membrane, and signal transduction (Table S3). In addition, five prophage-like elements were detected in CCB-ST2H9, one of which was classified as intact, three as incomplete, and one as questionable (Table S4).

Phylogeny of Photobacterium
Although 16S rRNA phylogeny has been widely applied in species classification, the limitations of using this gene as a phylogenetic marker in Vibrionaceae have been noted previously [10,11]. Meanwhile, multigene phylogenies provide better resolution between phylogenetically close strains, and the tree topology is less affected by recombination events [32,33]. In this study, a phylogenomic analysis was performed based on the bacterial core genes from 37 Photobacterium strains, using Vibrio cholerae ATCC 14035 as an outgroup (Figure 1). This tree showed that CCB-ST2H9 clustered with P. galatheae S2753, forming a distinct clade with P. halotolerans DSM18316, P. salinisoli LAM9071, and P. arenosum CAU 1568, distant from the other Photobacterium strains. This grouping appears to be related to environmental sources. With the exception of P. arenosum from marine sediment, this clade consists of Photobacterium isolated from non-marine habitats: CCB-ST2H9 from estuarine sediment, P. halotolerans from a saline lake, and P. salinisoli from saline soil [7,34]. The remaining Photobacterium strains were grouped into the clades Damselae, Leiognathi, Ganghwense, Profundum, Phosphoreum, and Rosenbergii, consistent with prior reports [35,36]. While this clade classification can facilitate the study of large genera by grouping together lines of descent, it is not a standard in nomenclature [8].
outgroup ( Figure 1). This tree showed that CCB-ST2H9 clustered with P. galatheae S2753, forming a distinct clade with P. halotolerans DSM18316, P. salinisoli LAM9071, and P. arenosum CAU 1568, distant from the other Photobacterium strains. This grouping appears to be related to environmental sources. With the exception of P. arenosum from marine sediment, this clade consists of Photobacterium isolated from non-marine habitats: CCB-ST2H9 from estuarine sediment, P. halotolerans from a saline lake, and P. salinisoli from saline soil [7,34]. The remaining Photobacterium strains were grouped into the clades Damselae, Leiognathi, Ganghwense, Profundum, Phosphoreum, and Rosenbergii, consistent with prior reports [35,36]. While this clade classification can facilitate the study of large genera by grouping together lines of descent, it is not a standard in nomenclature [8]. The genome evolutionary relatedness between Photobacterium strains was depicted based on the average nucleotide identity (ANI) and digital DNA-DNA hybridization (dDDH). In agreement with phylogenomic clustering, the highest ANI was shared between CCB-ST2H9 and P. galatheae (88.95%), P. halotolerans (84.26%), P. arenosum (84.25%), and P. salinisoli (84.25%) ( Figure 2). Accordingly, the highest dDDH percentage (32.7%) was between CCB-ST2H9 and P. galatheae (Table S5). Both the sharing ANI and dDDH values between CCB-ST2H9 and other Photobacterium strains were below the proposed The genome evolutionary relatedness between Photobacterium strains was depicted based on the average nucleotide identity (ANI) and digital DNA-DNA hybridization (dDDH). In agreement with phylogenomic clustering, the highest ANI was shared between CCB-ST2H9 and P. galatheae (88.95%), P. halotolerans (84.26%), P. arenosum (84.25%), and P. salinisoli (84.25%) ( Figure 2). Accordingly, the highest dDDH percentage (32.7%) was between CCB-ST2H9 and P. galatheae (Table S5). Both the sharing ANI and dDDH values between CCB-ST2H9 and other Photobacterium strains were below the proposed 95-96% and 70% cut-off of bacterial species definition [37,38]. Considering the 16S rRNA sequence similarity and ANI and dDDH results, strain CCB-ST2H9 is potentially a new species of the genus Photobacterium. 95-96% and 70% cut-off of bacterial species definition [37,38]. Considering the 16S rRNA sequence similarity and ANI and dDDH results, strain CCB-ST2H9 is potentially a new species of the genus Photobacterium.

Core and Pan-Genome of Photobacterium
A pan-genome is comprised of core genes found in all strains, accessory genes (shell and cloud) common to two or more strains but not all, and unique genes present only in one strain. The orthologous clustering of protein-coding sequences from the 37 Photobacterium genomes analysed identified a total of 21,786 consensus gene clusters defining the pangenome (Figure 3a). The size of the Photobacterium pan-genome would continue to expand with the increasing number of genomes incorporated in the analysis and can be considered open (Figure 3b). This pan-genomic feature is typical of taxonomic groups capable of colonizing diverse habitats and hence the opportunity to exchange genetic material with different sources [25,39]. one strain. The orthologous clustering of protein-coding sequences from the 37 Photob terium genomes analysed identified a total of 21,786 consensus gene clusters defining t pan-genome ( Figure 3a). The size of the Photobacterium pan-genome would continue expand with the increasing number of genomes incorporated in the analysis and can considered open (Figure 3b). This pan-genomic feature is typical of taxonomic groups c pable of colonizing diverse habitats and hence the opportunity to exchange genetic ma rial with different sources [25,39]. More than half of the gene sets constituting the pan-genome (14,638 genes, 67%) b longed to the cloud cluster, whereas the shell and the soft-core clusters accounted for 26 (5587 genes) and 7% (1561 genes) of the pan-genome, respectively (Figure 3c). Both t cloud and shell clusters are subsets of the flexible genome that could be applied to in an organism's evolutionary trajectory and lifestyle or habitat adaptation [40]. These tw clusters are hypothesized to have different rates of gene acquisition and deletion, with t cloud including rapidly gained and lost genes, and the shell containing slowly gained a More than half of the gene sets constituting the pan-genome (14,638 genes, 67%) belonged to the cloud cluster, whereas the shell and the soft-core clusters accounted for 26% (5587 genes) and 7% (1561 genes) of the pan-genome, respectively (Figure 3c). Both the cloud and shell clusters are subsets of the flexible genome that could be applied to infer an organism's evolutionary trajectory and lifestyle or habitat adaptation [40]. These two clusters are hypothesized to have different rates of gene acquisition and deletion, with the cloud including rapidly gained and lost genes, and the shell containing slowly gained and lost genes [41]. The core genome presented in all examined Photobacterium strains had 894 genes, representing approximately 19% of the total genes in a genome, which indicates significant gene conservation among Photobacterium. The number of core genes decreased rapidly with the addition of genomes but stabilized after the addition of the 20th genome, suggesting that the core genome was closed (Figure 3d). In a previous Photobacterium pan-genome study, a higher number of core genes was observed, likely due to the inclusion of genomes from less diverse species in the analysis and the slightly smaller number of genomes in the dataset [10].
COG functional assignments revealed that the Photobacterium core genome was enriched in functions involving translation, coenzyme metabolism, and nucleotide metabolism and transport (Figure 4a). These core genes encode housekeeping functions related to fundamental processes in the cell that have been conserved in all species throughout evolution. In addition to translation and coenzyme metabolism, the soft-core genome was also abundant in transcription function. The abundant COG category transcription, which consists of transcriptional regulators, could enable Photobacterium to regulate metabolic processes, contributing to their adaptability to the local environment. Compared to the core genome, the Photobacterium flexible genome was well represented with functions related to energy production and conversion, cell wall/membrane/envelope biogenesis, and carbohydrate metabolism and transport. The flexible genome was enriched with environmental genes, such as those that allow Photobacterium to respond to environmental changes by utilising different types of carbon and energy sources for surviving under various conditions. The flexible genome contributes to functional versatility, which in turn improves the ecological success of the bacteria in diverse environmental niches. Analysis of CCB-ST2H9 specific genes revealed that only 34.8% (201 genes) were assignable to COG functions ( Figure 4b).
These unique genes that evolved in CCB-ST2H9 likely complement strain-specific activity via unknown mechanisms. Besides function unknown (31%), COGs associated with replication and repair, carbohydrate metabolism and transport, and cell wall/membrane/envelope biogenesis were most abundant among strain-specific genes.
of CCB-ST2H9 specific genes revealed that only 34.8% (201 genes) were assignable functions ( Figure 4b). These unique genes that evolved in CCB-ST2H9 likely comp strain-specific activity via unknown mechanisms. Besides function unknown (31% associated with replication and repair, carbohydrate metabolism and transport, a wall/membrane/envelope biogenesis were most abundant among strain-specific g All Photobacterium except P. lipolyticum and P. sanguinicancri harbour a beta-lactone cluster. The CCB-ST2H9 beta-lactone cluster contained the biosynthetic gene encoding 2-isopropylmalate synthase, which is involved in the biosynthesis of beta-lactone [42] but shows no correlation with known clusters. CCB-ST2H9 also harbors a BGC for butyrolactone, a quorum sensing molecule related to the regulation of antibiotic production and morphogenesis [43]. The cyanobactin BGC uniquely found in the CCB-ST2H9 genome displayed no match to known clusters. Due to the diversity of gene products and cluster organization in BGCs, some predicted cluster regions do not resemble known database entries. P. indicum, P. frigidiphilum, P. ganghwense, P. lipolyticum, P. galatheae, P. arenosum, P. haloterans, P. salinisoli, and CCB-ST2H9 all have BGCs coding for ectoine that are reported to play a role in osmotolerance [44]. The lack of ectoine in other Photobacterium strains implies that they might utilize different strategies for osmoregulation in hypersaline environments. Although a RiPP-like BGC was detected in CCB-ST2H9, it did not show similarity to known clusters. All Photobacterium except P. lipolyticum and P. sanguinicancri harbour a beta-lactone cluster. The CCB-ST2H9 beta-lactone cluster contained the biosynthetic gene encoding 2isopropylmalate synthase, which is involved in the biosynthesis of beta-lactone [42] but shows no correlation with known clusters. CCB-ST2H9 also harbors a BGC for butyrolactone, a quorum sensing molecule related to the regulation of antibiotic production and morphogenesis [43]. The cyanobactin BGC uniquely found in the CCB-ST2H9 genome displayed no match to known clusters. Due to the diversity of gene products and cluster organization in BGCs, some predicted cluster regions do not resemble known database entries. P. indicum, P. frigidiphilum, P. ganghwense, P. lipolyticum, P. galatheae, P. arenosum, P. haloterans, P. salinisoli, and CCB-ST2H9 all have BGCs coding for ectoine that are reported to play a role in osmotolerance [44]. The lack of ectoine in other Photobacterium strains implies that they might utilize different strategies for osmoregulation in hypersaline environments. Although a RiPP-like BGC was detected in CCB-ST2H9, it did not show similarity to known clusters. Many bacteria have developed the ability to secrete iron-chelating molecules or siderophores that bind iron to form siderophore-iron complexes that are then liberated in- Many bacteria have developed the ability to secrete iron-chelating molecules or siderophores that bind iron to form siderophore-iron complexes that are then liberated internally in the cell [45]. Siderophores enable producers to scavenge dissolved iron from the environment and deprive competitors of it. Under iron-limiting conditions, siderophore-producing bacteria have a competitive advantage over other species that lack iron-chelating ability [46]. A siderophore BGC showing high homology (88%) to aerobactin from Aliivibrio fisheri ES114, which is a hydroxamate-type siderophore, was identified in CCB-ST2H9. The presence of the siderophore aerobactin in P. halotolerans MELD1 contributes to the protection of host plants from phytopathogens [47]. Similarly, the siderophore aerobactin in CCB-ST2H9 could contribute to its antagonistic potential. The identified NRPS in CCB-ST2H9 showing 100% similarity to BGC in the database include xenotetrapeptide of Xenorhabdus nematophila ATCC 19061. While X. nematophila has been studied for xenematide activity, the function of xenotetrapeptide in this strain has not been explored [48].
Another NRPS with potential biosynthetic novelty shows low homology (38%) with the holomycin gene cluster in Streptomyces clavuligerus ATCC 27064. Holomycin production has been reported in both P. galatheae and P. halotolerans and examined in terms of the biosynthetic gene cluster and physiological role [16,49,50]. Holomycin is a member of the dithiolopyrrolone class that displays antibiotic activity against a broad spectrum of bacteria and inhibits RNA synthesis [51,52]. Further sequence analysis revealed that the CCB-ST2H9 holomycin BGC contains 10 genes homologous to and arranged in the same manner as those in P. galatheae [49]. The cluster is formed by phosphopantothenoylcysteine decarboxylase, HlmF, metallophosphoesterase, HlmY, globin, HlmG, N-acyltransferase, HlmA, acyl-CoA dehydrogenase, HlmB, thioesterase, HlmC, flavin mononucleotide-dependent oxidoreductase, HlmD, MFS transporter, HlmH, transcriptional regulator, HlmX, and core protein NRPS synthetase, HlmE with a characteristic arrangement of cyclization (Cy), adenylation (A), and thiolation (T) domains ( Figure 6). The largest block of conserved genes corresponding to hlmA-hlmE is conserved in all holomycin producers, including S. clauvuligerus, Y. ruckeri, Pseudoalteromonas sp., and Photobacterium [49,[53][54][55]. The holomycin BGC was also found in the P. arenosum and P. salinosoli genomes, even though antibiotic production and antagonism in these species have not been investigated. The presence of nearly identical holomycin BGCs in P. galatheae, P. halotolerans, P. arenosum, P. salinisoli, and CCB-ST2H9 suggests that this cluster may have originated via a horizontal gene transfer (HGT) event from a donor closely related to these species. Overall, the annotation of CCB-ST2H9 genes involved in the biosynthesis of compounds with antibacterial and iron scavenging functions is in agreement with the strain's ability to antagonize the growth of pathogenic bacteria. explored [48].
Another NRPS with potential biosynthetic novelty shows low homology (38%) with the holomycin gene cluster in Streptomyces clavuligerus ATCC 27064. Holomycin production has been reported in both P. galatheae and P. halotolerans and examined in terms of the biosynthetic gene cluster and physiological role [16,49,50]. Holomycin is a member of the dithiolopyrrolone class that displays antibiotic activity against a broad spectrum of bacteria and inhibits RNA synthesis [51,52]. Further sequence analysis revealed that the CCB-ST2H9 holomycin BGC contains 10 genes homologous to and arranged in the same manner as those in P. galatheae [49]. The cluster is formed by phosphopantothenoylcysteine decarboxylase, HlmF, metallophosphoesterase, HlmY, globin, HlmG, N-acyltransferase, HlmA, acyl-CoA dehydrogenase, HlmB, thioesterase, HlmC, flavin mononucleotide-dependent oxidoreductase, HlmD, MFS transporter, HlmH, transcriptional regulator, HlmX, and core protein NRPS synthetase, HlmE with a characteristic arrangement of cyclization (Cy), adenylation (A), and thiolation (T) domains ( Figure 6). The largest block of conserved genes corresponding to hlmA-hlmE is conserved in all holomycin producers, including S. clauvuligerus, Y. ruckeri, Pseudoalteromonas sp., and Photobacterium [49,[53][54][55]. The holomycin BGC was also found in the P. arenosum and P. salinosoli genomes, even though antibiotic production and antagonism in these species have not been investigated. The presence of nearly identical holomycin BGCs in P. galatheae, P. halotolerans, P. arenosum, P. salinisoli, and CCB-ST2H9 suggests that this cluster may have originated via a horizontal gene transfer (HGT) event from a donor closely related to these species. Overall, the annotation of CCB-ST2H9 genes involved in the biosynthesis of compounds with antibacterial and iron scavenging functions is in agreement with the strain's ability to antagonize the growth of pathogenic bacteria.  High variability of the biosynthetic operons in terms of abundance and composition was observed across Photobacterium, with P. galathea carrying the highest number of BGCs (15), while P. carnosum, P. damselae, P. iliopiscarium, and P. toruni have the lowest (2). This is the first study to use clustering based on BGCs profiles to separate the Photobacterium into BGCs-rich and BGCs-low groups. Although no evidence linking environmental sources and BGCs was found, the frequency of BGCs types appears to correlate with genetic proximity. Notably, CCB-ST2H9 displays a similar frequency of BGCs with P. halotolerans, P. arenosum, P. galathea, and P. salinisoli that are phylogenetically close. Except for P. alginatilyticum and P. proteolyticum that are phylogenetically distant, these strains cluster together by the presence of a large number of BGCs in their genomes. Conversely, it could also be claimed that the acquisition of gene clusters leads to ecological diversity and speciation, forming a phylogenetically distinct clade of P. arenosum, P. halotolerans, P. salinisoli, P. galathea, and CCB-ST2H9. The wide distribution of bioactive metabolites in these Photobacterium strains suggests good potential for antibiotic production and as biocontrol agents.
In addition, the variable distribution of BGCs in Photobacterium is indicative of gene loss from the descendants of a cluster-harboring ancestor and HGT or recombination events. Genes encoding traits subjected to weak selection are more likely to be lost, whereas genes that confer positive benefits under certain conditions are likely to be gained. For instance, a polyunsaturated fatty acid (PUFA) biosynthetic cluster that is present in other Photobacterium strains was not detected in CCB-ST2H9. Since the PUFA cluster has been linked to cold temperature adaptation [56], it could have been lost in the evolution of non-psychrophiles Photobacterium. On the other hand, the presence of BGCs for antibiotics in Photobacterium such as P. arenosum, P. halotolerans, P. salinisoli, P. galatheae, and CCB-ST2H9 raises the intriguing question of the ecological significance of the BGCs to the producers. In habitats characterized by dense multispecies communities, bacteria compete with their neighbours for scarce resources [57]. Antibiotic compound productions are proposed as weapons that provide fitness advantage over other occupants of the same ecological niche [58]. Compared to free-living Photobacterium, densely colonized sediment and soil could be the factor driving the evolution of antibiotic-producing traits in strains such as P. arenosum, P. salinisoli, and CCB-ST2H9. This view is in line with results from pelagic ocean samples, which showed increased antagonism of particle-attached bacteria compared to free-living bacteria [59]. Overall, the distinct BGC profiles displayed by Photobacterium from diverse habitats provide clues that metabolite production capacity could be the result of habitat-specific adaptation.

Antimicrobial Activity of EtOAc Extract against Vibrio spp.
As analysed above, CCB-ST2H9 possesses the holomycin biosynthetic gene cluster in its genome. To confirm the production of antagonistic compounds, EtOAc extraction was performed from the culture supernatant. Disc diffusion assay of the crude extract resuspended in methanol was conducted against seven Vibrio spp., including VpAHPND, and two bacterial strains, E. coli DH5α and Bacillus sp. CCB-MMP212, as positive controls. Oliver and co-researchers reported that holomycin exhibited antimicrobial activity against both Gram-positive and negative bacteria such as Staphylococcus and E. coli [51]. Besides that, holomycin isolated from P. galarheae S2753 exhibited the antagonistic activity against several marine bacteria, including the genus Vibrio [49]. As shown in Figure 7, the crude extract also showed antimicrobial activity on Bacillus sp. CCB-MMP212 (Gram-positive) and E. coli DH5α (Gram-negative). All Vibrio spp. tested were susceptible to the crude extract ( Figure 7). It was reported that AHPND could be caused not only by V. parahaemolyticus but also other non-pathogenic Vibrio species because of the introduction of the pVA1 plasmid carrying the pirAB gene encoding homologues of Photorhabdus insect-related (Pir) toxins [60]. This finding suggests that CCB-ST2H9 might have the potential as a novel biocontrol agent due to its broad host range of antagonistic compounds.

Genome Sequencing
Genomic DNA was extracted from mid-logarithmic phase culture following the method of Sokolov et al. [64] with some modifications: sample lysis was performed with a buffer containing 50 mM NaCl, 50 mM Tris-HCl pH8, 50 mM EDTA, 2% SDS, and isopropanol precipitation was replaced by SPRI bead purification. The quality and quantity of isolated DNA were assayed using 1% (w/v) agarose gel electrophoresis and Qubit fluorimetry (Thermo Scientific, Waltham, MA, USA), respectively. The Illumina sequencing library was constructed using the NEBNext Ultra DNA kit (New England Biolabs, Ipswich, MA, USA) according to the manufacturer's instructions. The genome of CCB-ST2H9 was sequenced on a NovaSeq 6000 (Illumina, San Diego, CA, USA) at 150 bp paired-end, generating 10.26 million reads totaling 1.53 Gb. The DNA library for Nanopore sequencing was prepared using the Ligation sequencing kit (SQK-LSK109, Oxford Nanopore, UK). The library was sequenced with a PromethION and yielded 182,476 reads with a total of 1.92 Gb.
Gene clustering was performed with GET_HOMOLOGIES based on the bidirectional best-hit (BDBH), COGtriangles, and OrthoMCL algorithms. BLASTp searches were done with a 75% minimum alignment coverage and E-value set at 1×10 −5 . The core-genome was estimated from the consensus clusters defined by BDBH, COGtriangles, and OrthoMCL, whereas the pan-genome was computed from COGtriangles and OrthoMCL clusters.

Extraction of Antagonistic Compounds from CCB-ST2H9 and Its Disk Diffusion Assay
The extraction of antagonistic compounds from CCB-ST2H9 was conducted according to the method of Mansson [16] with slight modifications. CCB-ST2H9 was cultured in 300 mL of Sigma Sea Salt medium (4% Sigma Sea Salt, 0.4% glucose, and 0.3% casamino acids with 10 mM HEPES, pH 7.6) for 72 h at 30 • C with shaking (200 rpm). After the incubation, the cell suspension was centrifuged at 8000× g for 20 min to remove the cells. An equal volume of ethyl acetate (EtOAc) was added and mixed into the supernatant. The EtOAc fraction was evaporated using a rotary evaporator (Heidolph, Germany). The crude extract was resuspended in methanol (200 mg/200 µL), and the solution was subjected to a test for its antagonistic activity on Vibrio spp.
Double-layer agar plates containing test strains were prepared to test the antagonistic activity of crude extract. Four mL of H-ASWM soft agar (0.3%) containing 200 µL of the exponential phase cells of each Vibrio species were overlayed on H-ASWM agar plates. The disk soaked in 20 µL of the extract was placed on the center of the double-layer agar plate, and the plate was incubated for 1 day at 30 • C. The double-layer agar plates containing E. coli DH5α or Bacillus sp. CCB-MMP212 were prepared by LB agar and MB, respectively. The disk with soaked methanol was used as a negative control.

Conclusions
This study provides insights into the genomic characteristics of Photobacterium sp. CCB-ST2H9 and its antagonistic potential against Vibrio spp. Photobacterium has an open pan-genome encompassing a core genome encoding essential functions and a flexible genome containing environmental genes contributing to adaptation to diverse habitats. While phylogenomic analysis showed that CCB-ST2H9 was closely related to P. galatheae, P. halotolerans, P. salinisoli, and P. arenosum, genomic metrics including 16S rRNA sequence similarity, ANI, and dDDH values between CCB-ST2H9 and these species were below the boundary of a bacterial species. AntiSMASH prediction revealed that CCB-ST2H9 harbours 10 BGCs, including three NRPSs predicted to be involved in the production of antibiotics such as holomycin and xenotetrapeptide. These NRPSs, along with the BGCs for butyrolactone and siderophore, could contribute to the antagonistic potential of CCB-ST2H9. The BGCs profile revealed that CCB-ST2H9, along with the other six Photobacterium spp., formed a BGCs-rich group. In addition, the production of antagonistic compounds against the causative agent of shrimp disease was also experimentally confirmed. Photobacterium spp. are promising resources for the genetic exploration and exploitation of bioactive secondary metabolites.

Conflicts of Interest:
The authors declare no conflict of interest.