Comparative genomics insights into Pseudomonas rhodesiae: environmental distribution, resistance determinants, virulence factors, and evolutionary implications

Abstract

Background

Pseudomonas rhodesiae is a member of the Pseudomonas genus that is ecologically adaptable but has received little research. Despite its presence in clinical, environmental, agricultural, and aquaculture contexts, there is just a small amount of genomic information available for this species.

Results

We sequenced the genome of P. rhodesiae strain SK22, which was identified after a mortality outbreak in Dicentrarchus labrax (European seabass) farming in Turkey, and conducted a thorough comparison with 32 other publicly accessible P. rhodesiae genomes sourced from different environments. Whole-genome-based core genome phylogeny and Average Nucleotide Identity (ANI) identified two misclassified strains, 26B3 and CIP104664, that grouped with P. quebecensis (> 99% ANI). After eliminating these genomes, comparative studies revealed significant strain-specific variability in AMR determinants, virulence-associated genes, integrons, prophage content, and secretion systems. Class 1 integron was found only in one mineral-water isolate, whereas environmental isolates—particularly those from aquaculture, agricultural systems, and urban areas—had the largest AMR gene loads, suggesting their potential involvement in the dissemination of resistance and opportunistic pathogenicity. All genomes expressed key virulence characteristics such as T1SS-T6SS secretion systems, type IV pili, flagella, and siderophore-related genes, but supplementary virulence factors differed significantly. Prophage profiling found intact phage regions throughout all genomes, with agricultural and aquaculture isolates having abnormally large prophage loads, indicating phage-mediated adaptation and gene acquisition.

Conclusions

This study provides the first high-resolution genomic framework for P. rhodesiae, identifies incorrectly classified genomes within public databases, and highlights the species’ extensive genomic plasticity. The combination of diverse AMR determinants, mobile genetic elements, and virulence factors suggests that P. rhodesiae has the potential to function not only as an environmental saprophyte but also as an opportunistic pathogen capable of disseminating resistance genes across ecosystems. These findings highlight the importance of increased surveillance and functional validation in understanding the ecological and clinical roles of this emerging species.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-026-12681-3.

Background

The genus Pseudomonas is a diverse and ubiquitous group of Gram-negative, non-spore-forming, motile bacteria that are widely recognized for their remarkable metabolic versatility and environmental adaptability. These bacteria are commonly characterized by their rod-shaped morphology, oxidative metabolism, and ability to produce distinctive pigments. Classified within the Gammaproteobacteria, Pseudomonas species possess a broad range of biochemical and physiological capabilities, including proteolytic, lipolytic, and psychrotrophic activities. This metabolic flexibility allows them to thrive in a wide variety of environments, including soil, water, vegetation, and even within animal hosts [1–5].

Many Pseudomonas species have significant ecological and clinical importance. For instance, P. aeruginosa is a well-known human pathogen, recognized for its role in opportunistic infections, particularly in immunocompromised individuals. In contrast, other species, such as P. fluorescens, are considered effective biofilm producers, which can foster the growth of other pathogens, especially in food-related environments. Additionally, various Pseudomonas species, including P. anguilliseptica, P. brenneri, P. baetica, P. defensor, P. chlororaphis, P. haemolytica, P. koreensis, P. lactis, P. lurida, P. mandelii, P. luteola, P. proteolytica, P. plecoglossicida, P. simiae, P. meridiana, P. weihenstephanensis, P. pseudoalcaligenes, P. fluorescens, and P. putida, are known to cause diseases in aquatic organisms, inducing a range of external lesions including ulcers, fin erosion, and hemorrhages. These bacteria are also significant spoilage organisms in food industries, including dairy, meat, and seafood [6–8].

Bacterial whole-genome sequencing has become a cornerstone in microbiology and related sciences, due to the advances in next-generation sequencing technology, which provides a comprehensive view of an organism’s genetic makeup, including all its genes and regulatory components. It is also helpful in understanding bacterial physiology, metabolism, and pathogenicity, and predicting antimicrobial resistance patterns. The comparison of diverse bacterial strains’ genomes within the same species aids in the construction of phylogenetic links, genetic differences such as virulence factors and antibiotic resistance, the identification of unique genes, and regulatory mechanisms from various environmental or pathogenic perspectives [9–13].

P. rhodesiae strain CIP 104,664^T was initially isolated from natural mineral water and identified in 1996 [14], and it was sequenced in 2022. P. rhodesiae strain FF9 was the first reported genome of the P. rhodesiae species on the National Center for Biotechnology Information (NCBI). P. rhodesiae was classified into the P. fluorescens phylogenetic subgroup based on branch length, grouping, and bootstrap values from both the rpoD and whole genome phylogenies [4]. P. rhodesiae is a relatively underexplored species within the well-studied Pseudomonas genus, despite its known presence in diverse ecological niches such as soil, water, and plant environments. While other Pseudomonas species have been extensively studied, particularly in terms of their genomic and ecological characteristics, P. rhodesiae remains poorly characterized at the genomic level. To address this gap, the present study offers the first comprehensive genomic analysis of a P. rhodesiae strain, accompanied by a detailed comparative genomic analysis with other P. rhodesiae strains. The aim is to identify key genetic features associated with virulence factors, antibiotic resistance mechanisms, biofilm formation, and environmental adaptability. This comparative approach will not only enhance our understanding of P. rhodesiae’s unique genetic makeup and its role within microbial ecosystems but also highlight both conserved and species-specific traits. The findings will contribute to a broader understanding of the Pseudomonas genus and lay the groundwork for future research into the ecological and potential clinical significance of P. rhodesiae.

Materials and methods

Bacterial isolation and whole genome sequencing

In 2012, a disease outbreak in Dicentrarchus labrax (European seabass) aquaculture in the Black Sea caused significant fish mortality, prompting the submission of five infected fish for diagnostic examination to Recep Tayyip Erdogan University’s Fisheries Faculty. The major bacterial agent recovered from these fish was Pseudomonas rhodesiae, which was first identified as Pseudomonas spp. using the API 20NE (bioMérieux) method. The isolate, strain SK22, was then stored at -70 °C in 15% glycerol. The bacterium was moved from − 70 °C to Tryptic Soy Agar (TSA) and cultured at 20 °C for 24 h for whole genome sequencing. Then, a colony was transferred to 10 mL of Tryptic Soy Broth (TSB) and cultured overnight at 20 °C with shaking at 150 rpm. Bacterial genomic DNA was isolated using the Zymo Research “Quick-DNA Fungal/Bacterial Miniprep Kit, Cat. No: D6005” following the manufacturer’s instructions. The quality and quantity of DNA were evaluated using spectrophotometric and fluorometric techniques. The sequencing library was created using the Nextera XT DNA Library Preparation Kit, and sequencing was conducted on the Illumina NovaSeq 6000 platform, producing paired-end (PE) 2 × 150 base reads. FastQC was used to verify the quality of the raw whole genome sequence (WGS) readings. The data showed that the sequencing quality was high, with 97.78% of bases getting a Q20 score and 93.82% getting a Q30 score. After sequencing, adapter sequences, contamination, and low-quality bases were deleted, and reads that didn’t meet quality or length standards were left out. Subsequently, utilizing pristine data, the consensus genome was constructed through de novo assembly employing the CLC Genomics Workbench (version 22.0.1), resulting in 37 contigs and 6,074,418 bp reads at 242X coverage. The final version of the assembled P. rhodesia strain SK22 sequencing data was submitted to the National Center for Biotechnology Information (NCBI) genome database. The GenBank accession number is JBFQEB000000000. All available P. rhodesiae genomes protein and sequence files were acquired from the NCBI genomes database (as of July 30, 2024) (Table 1). Genomes with more than 300 contigs were excluded from the analysis due to concerns about analysis efficiency and genome quality.

Table 1.

Genome features of Pseudomonas rhodesiae

#	Organism Name	Source	Location	Level	Scaffolds	Size (bp)	Release Date	Accession #	References
1	Pseudomonas rhodesiae CTOTU48138	Urban	USA	Scaffold	45	5,916,096	09.26.23	GCA_031993805.1	[89]
2	Pseudomonas rhodesiae 3.8	Brewery	Belgium	Complete	2	6,474,451	06.01.23	GCA_030144705.1	[56]
3	Pseudomonas rhodesiae ERR9968924_bin.9_MetaWRAP_v1.3_MAG	Homo sapiens	Australia	Contig	55	5,923,447	10.01.23	GCA_963521225.1	NA
4	Pseudomonas rhodesiae AAMF24	Farm pasture	USA	Complete	1	5,772,816	05.20.21	GCA_018417555.1	[90]
5	Pseudomonas rhodesiae CTOTU50844	Urban	USA	Scaffold	56	5,765,194	09.27.23	GCA_032087575.1	[89]
6	Pseudomonas rhodesiae SK22	Dicentrarchus labrax	Turkiye	Contig	37	6,074,418	07.27.24	GCA_040931495.1	This study
7	Pseudomonas rhodesiae S5_IA_3a	Tomato	Turkiye	Scaffold	104	6,844,694	12.22.20	GCA_016307315.1	[91]
8	Pseudomonas rhodesiae WS 5107	Raw milk	Germany	Contig	58	5,961,510	05.07.20	GCA_012985855.1	[92]
9	Pseudomonas rhodesiae FSL M9-0606	Packaged baby spinach	USA	Contig	212	5,901,518	12.28.23	GCA_034808945.1	[93]
10	Pseudomonas rhodesiae FSL S12-0141	Packaged baby spinach	USA	Contig	133	5,928,270	12.28.23	GCA_034808985.1	[93]
11	Pseudomonas rhodesiae BML-PP011	Homo sapiens	Japan	Contig	80	5,849,171	11.22.21	GCA_021603345.1	[94]
12	Pseudomonas rhodesiae SPPC 2818	Plant rhizosphere	Canada	Contig	43	5,914,570	07.13.23	GCA_030462455.1	[95]
13	Pseudomonas rhodesiae CTOTU48728	Urban	USA	Scaffold	70	6,245,817	09.27.23	GCA_032056925.1	[89]
14	Pseudomonas rhodesiae S00114	N/A	USA	Scaffold	30	5,771,111	08.14.20	GCA_014204275.1	NA
15	Pseudomonas rhodesiae WS 4669	Cow	Germany	Contig	65	5,983,759	05.07.20	GCA_012985905.1	[92]
16	Pseudomonas rhodesiae B21-046	Soil	Canada	Complete	1	5,753,837	08.26.22	GCA_024748755.1	[96]
17	Pseudomonas rhodesiae CTOTU49380	Urban	USA	Scaffold	36	5,904,190	09.27.23	GCA_032052475.1	[89]
18	Pseudomonas rhodesiae cycles.exp.s105	Lake water	Switzerland	Contig	36	5,915,029	04.21.24	GCA_964032085.1	NA
19	Pseudomonas rhodesiae FF9	N/A	France	Scaffold	13	6,051,076	10.18.14	GCA_000821225.2	NA
20	Pseudomonas rhodesiae UBA2601	Metal	USA	Scaffold	49	5,712,426	09.22.17	GCA_002340035.1	[97]
21	Pseudomonas rhodesiae FP1707	Plant rhizosphere	China	Complete	1	5,934,031	08.08.23	GCA_030687185.1	NA
22	Pseudomonas rhodesiae BML-PP031	Homo sapiens	Japan	Contig	136	6,129,167	11.22.21	GCA_021602225.1	[94]
23	Pseudomonas rhodesiae DLA23	Soil	USA	Complete	1	6,109,631	05.20.21	GCA_018417575.1	NA
24	Pseudomonas rhodesiae NL2019	Soil	China	Complete	1	5,779,157	06.08.20	GCA_013285305.1	NA
25	Pseudomonas rhodesiae CTOTU48783	Urban	USA	Scaffold	116	5,878,822	09.27.23	GCA_032062085.1	[89]
26	Pseudomonas rhodesiae FP69	Plant rhizosphere	China	Complete	1	6,072,001	08.08.23	GCA_030687355.1	NA
27	Pseudomonas rhodesiae 88A6	Soil	USA	Scaffold	18	5,958,491	11.15.18	GCA_003732025.1	NA
28	Pseudomonas rhodesiae 90F12-1	Soil	USA	Scaffold	26	5,959,528	11.15.18	GCA_003732045.1	NA
29	Pseudomonas rhodesiae JCM 11940	Natural mineral water	France	Contig	2	6,416,178	02.23.24	GCA_039522895.1	NA
30	Pseudomonas rhodesiae DSM 14020	Natural mineral water	France	Contig	82	6,333,130	08.06.19	GCA_007858255.1	NA
31	Pseudomonas rhodesiae LMG 17764	Natural mineral water	France	Complete	1	6,417,799	10.21.16	GCA_900105575.1	NA
32	Pseudomonas rhodesiae cycles.exp.s118	Lake water	Switzerland	Contig	37	5,664,902	04.21.24	GCA_964031495.1	NA
33	Pseudomonas rhodesiae 26B3 *	N/A	USA	Contig	2	5,724,985	03.30.23	GCA_029504075.1	NA
34	Pseudomonas rhodesiae CIP104664 *	Natural mineral water	USA	Contig	2	5,725,440	07.06.22	GCA_024169765.1	NA

* indicates that they are Pseudomonas quebecensis not Pseudomonas rhodesiae

Phylogenetic relationship and Average Nucleotide Identity (ANI) value

All strains of P. rhodesiae were analyzed using a phylogenetic tree and an ANI value to demonstrate the relationship between them. A phylogenetic tree was constructed by computing the core genes of 34 genomes (each genome had a core of 3784 genes, for a total of 128656 genes, and the core had a total of 1244143 AA-residues, 42300862 in total) utilizing the EDGAR v3.2 [15] software platform for comparative genomics. EDGAR was also used to calculate the ANI value [16]. Following this, MUSCLE [15] was used to create alignments for each of the key gene sets, and then all the alignments were merged into a super-matrix. The EDGAR tree construction methods take this alignment as input. The FastTree program (http://www.microbesonline.org/fasttree/) [17] is used to create phylogenetic trees that are close to maximum-likelihood trees.

Integron

IntegronFinder 2.0 [18] was used to identify the presence of novel integrons in P. rhodesiae genomes, utilizing default features and genome nucleotide sequence files downloaded from NCBI. The default features were used as follows: thorough local detection: no, default replicons topology: linear, use a topology file: no, search also for promoter and attI sites: yes, genbank output: yes, pdf output: yes.

Antibiotic susceptibility testing and antimicrobial resistance elements predictions

The antibiotic resistance profiles of the confirmed Pseudomonas spp. isolate were determined using the disc diffusion method, which followed Clinical and Laboratory Standards Institute (CLSI) criteria [19]. The following antibiotics were tested: Ampicillin (10 µg), Erythromycin (15 µg), Enrofloxacin (5 µg), Florfenicol (30 µg), Amoxicillin-Clavulanic Acid (10 µg), Gentamicin (10 µg), Sulfamethoxazole (25 µg), Trimethoprim+Sulfamethoxazole (25 µg), and Oxytetracycline (30 µg). Bacterial suspensions were adjusted to the 0.5 McFarland standard and distributed on Mueller-Hinton agar. Antibiotic discs were applied to plates and incubated at 25 °C for 24 h. The susceptibility was categorized as sensitive, moderate, or resistant.

To identify the anticipated antibiotic-resistant components, a complete protein dataset from the Comprehensive Antibiotic Resistance Database (CARD) [20] was downloaded and imported into the BioEdit software [21]. Putative antimicrobial resistance elements were identified using BioEdit software’s local BLAST function, with a cutoff E-value of 10^–20 considered significant. In addition to E-values, we personally evaluated query coverage and subject coverage to reduce false positives from short or incomplete alignments. Specially, only hits that covered for at least 70% of the query length were kept for further analysis.

Alignments with very short local matches were discarded even if their E-values were low.

Bacterial prophage

Bacterial prophage sequences in P. rhodesiae genomes were identified with the web service, PHASTEST [22–24], using default features; choose file: nucleotide files, select bacterial sequence annotation mode: lite, my input consists of multiple separate contigs (fasta format only): yes, use pre-computed results if available (faster): yes.

Bacterial secretion systems

To identify protein secretion systems in P. rhodesiae genomes, MacSyFinder (Version 2) [18, 25, 26] was used with default values (the type of dataset to deal with: unordered, what kind of microorganism do you want to analyze: bacteria, tune or leave default values to Hmmer options: default).

Predicted virulence factors

A full dataset of the virulence factor database (VFDB) [27] protein file was downloaded to define the predicted virulence factors. The CLC Genomics Workbench (22.0.1) was used to perform local searches on all protein files. The potential virulence factors were identified, and a cutoff E-value of 10^–100 was deemed significant. We personally evaluated query coverage and subject coverage to reduce false positives from short or incomplete alignments. Specially, only hits that covered for at least 70% of the query length were kept for further analysis. Alignments with very short local matches were discarded even if their E-values were low. A customized Python script was then used to generate a binary presence-absence matrix based on this cutoff. To highlight the distribution of unique virulence factors across genomes, genes that were encoded by all genomes were excluded from the final visualization. A heatmap of the resulting matrix, containing unique virulence genes, was generated using the heatmap package in R Studio (v2024.12.1 + 563). The heatmap was clustered using hierarchical clustering based on the presence and absence of the genes. The raw heatmap with all virulence genes across genomes is provided in Supplementary file 4.

Results

Bacterial genome sequencing and genome data source

We performed whole genome sequencing of an outbreak-causing P. rhodesiae isolate that infected fish and caused a significant increase in fish mortality in Dicentrarchus labrax (European seabass) aquaculture in the Black Sea. The final product was condensed to 37 contigs, and this was compared to the sequences of 33 P. rhodesiae genomes in NCBI. Table 1 provides a list of the 34 genomes and their distinctive features. The bacteria were isolated from various sources including the urban environment, breweries, human digestive systems, farm pastures, Dicentrarchus labrax (European seabass), tomatoes, raw milk, packaged baby spinach, Homo sapiens, Cannabis sativa, cattle, agricultural soil, lake water metagenomes, metal, plant rhizospheres, soil, and natural mineral water. Their locations include the United States, Belgium, Australia, Germany, Japan, Canada, Switzerland, France, China, and Turkiye. The average genome size ranges from 5.664 Mb to 6.845 Mb. Among the investigated P. rhodesiae strains, eight full genomes were present. Additionally, the brewery isolate strain 3.8 carried one large plasmid totaling 514,920 bp. The other genome sequences were incomplete and presented as contigs and scaffold genomes.

Phylogenetic relationship and Average Nucleotide Identity (ANI) value

The ANI values for 34 P. rhodesiae genomes are presented in Table 2. The phylogenetic tree that was created is displayed in Fig. 1, and the ANI values support it. The average ANI value among 32 P. rhodesiae genomes is greater than 96%. The ANI values between strains 26B3 and CIP104664 are greater than 99%, although they are roughly 91% to other P. rhodesiae strains. The strains 26B3 and CIP104664 are more similar to Pseudomonas quebecensis with a 99% ANI value (Table 2) and are likely miscategorized. Therefore, these two genome sequences were removed from subsequent comparisons.

Table 2.

Average Nucleotide Identity (ANI) value of evaluated Pseudomonas genomes. For the same species, the ANI value should be higher than 95%. As a result of analysis, strains 26B3 and CIP104664 were identified as Pseudomonas quebecensis with higher than 97% ANI value

Fig. 1.

A phylogenetic tree for 34 genomes, derived from a core of 3,784 genes per genome, totaling 128,656 genes (32 genomes belong to Pseudomonas rhodesiae strains. * indicates that they are Pseudomonas quebecensis strains)

Integron

Complete class 1 integron region was encoded by P. rhodesiae strains DSM 14020, JCM 11940, and LMG 17764 (Supplementary file 1). DSM 14020, JCM 11940, and LMG 17764 are three different culture collection numbers assigned to the same strain and have different genome assemblies.

Antibiotic susceptibility testing and predicted antimicrobial resistance elements

Based on the antibiotic susceptibility testing results, P. rhodesiae strain SK22 showed resistance to ampicillin, amoxicillin-clavulanic acid, sulfamethoxazole, and erythromycin antibiotics. On the other hand, it showed moderate resistant to florfenicol, and was sensitive to gentamicin, oxytetracycline, enrofloxacin, and trimethoprim+sulfamethoxazole antibiotics.

In the investigation of 32 P. rhodesiae genomes, predicted antibiotic classes, resistance mechanisms, and resistance-associated elements are given in Supplementary File 2. The predicted classes of antibiotics include Aminocoumarin, Aminoglycoside, Antibacterial free fatty acids, Bicyclomycin-like, Cephalosporin, Diaminopyrimidine, disinfectants and antiseptics, Fluoroquinolone, Glycopeptide, Lincosamide, Macrolide, multidrug antibiotics, Nitroimidazole, Nucleoside, Penicillin beta-lactam, Peptide, Phosphonic acid, polycationic antibiotics, Sulfonamide, Tetracycline, Mupirocin-like, and Rifamycin. Their resistance mechanisms included antibiotic efflux, target replacement, target change, target protection, inactivation, and reduced permeability to antibiotics. 109 predicted antimicrobial resistance elements were common to all genomes, despite the presence of unique proteins in each genome (Table 3). All predicted antimicrobial resistance elements identified in the analyzed P. rhodesiae genomes are listed in Supplementary file 2. Strains FF9 and UBA2601 had the fewest antimicrobial resistance elements, but strains SK22, S5_IA_3a (fish and plant isolate from Turkiye), CTOTU48728, and CTOTU49380 (urban isolate from the USA) displayed the highest number of predicted antimicrobial resistance elements. Furthermore, components of Mupirocin-like antibiotics and the Rifamycin antibiotic class were exclusive to specific genomes (Table 3).

Table 3.

Predicted Unique Antimicrobial Resistance (AMR) Elements of Pseudomonas rhodesiae genomes. The table includes resistance-associated genes, corresponding antibiotic classes, and resistance mechanisms

1 = present proteins, 0 = absent proteins

Bacterial prophage

All evaluated P. rhodesiae genomes contained phage regions and phage-associated proteins. The genomes’ phage regions and the quantity of proteins they contain are shown in Table 4. There are three levels of phage status: intact (score > 90), questionable (score70–90), and incomplete (score < 70). All genomes contained at least one phage region with intact completeness scores; however, among the genomes, strains CTOTU48728, cycles.exp.s105, and DLA23 possessed only one phage region. Strains DLA23 and CTOTU48728 exhibited the lowest phage hit protein (21), whereas strain cycles.exp.s105 had 67 phage hit proteins. Conversely, strain WS 5107 showed the highest number of phage areas (6 regions) and phage hit proteins (204 proteins), whereas strain SK22 included 4 phage regions with 152 phage hit proteins.

Table 4.

Pseudomonas rhodesiae genomes’ bacterial phage regions and elements

Specific Keywords: Specific phage-related keywords in the GenBank protein name field of the input file; Phage-hit, Hypothetical, and Bacterial protein numbers: predicted phage-like hit proteins; ATT-site Show-up: potential phage attachment sites; Phage species: predicted number of phage species

Secretion systems

All P. rhodesiae genomes contained type I secretion system (T1SS), type II secretion system (T2SS), type III secretion system (T3SS), Flagellum, type VI secretion system (T6SS), and type IVa pilin (T4aP). Although all genomes encoded T1SS, the number of proteins varied. One of the subgroups, including strains FP1707, BML-PP031, DLA23, NL2019, CTOTU48783, FP6988A6, 90F12-1, JCM11940, DSM14020, and LMG17764, contained the highest number of tight adherence pili (Tad) elements. All genomes encoded the type V secretion system (T5SS) except the T5cSS_PF03895 proteins (Tables 5 and 6). Additionally, all genomes were lacking secretion-related genes, including gspN (T2SS), pilY and pilU (T4aP), and evpJ (T6SS) (except strains FP1707, DLA23, and NL2019). Seventeen genomes possessed a minimum of one type IV secretion (T4SS) sub-system (T4SS_G, T4SS_F, and T4SS_T), which contained the majority of proteins. Strains SK22, S5_IA_3a, cycles.exp.s105, and UBA2601 had the majority of type IVb pilin (T4bP) proteins (Supplementary file 3). Importantly, we underline that these predictions reflect genetic potential, rather than experimentally verified activity.

Table 5.

Bacterial secretion systems of Pseudomonas rhodesiae genomes. Values indicate the number of proteins associated with each secretion system per genome, as identified using MacSyFinder. Missing components indicate incomplete systems or strain-specific variation

Table 6.

Bacterial secretion systems of Pseudomonas rhodesiae genomes. Values indicate the number of proteins associated with each secretion system per genome, as identified using MacSyFinder. Missing components indicate incomplete systems or strain-specific variation

Predicted virulence factors

A total of 201 shared proteins associated with adherence, antimicrobial activity/competitive advantage, biofilm formation, effector delivery systems, exoenzymes, exotoxins, immune modulation, invasion, motility, nutritional/metabolic factors, regulation, and stress survival virulence factors were identified across all strains. Furthermore, each genome possessed distinct virulence-associated genes. A subgroup of P. rhodesiae genomes, comprising strains FP1707, BML-PP031, DLA23, NL2019, CTOTU48783, FP69, 88A6, 90F12-1, JCM 11,940, DSM 14,020, and LMG 17,764, exhibited adherence-related pili proteins. Strain FF9 has exotoxin phytotoxin proteins known as syringopeptins. Although all genomes possessed immune modulation-related capsules, lipopolysaccharides (LPS), O-antigens, and capsular polysaccharide proteins, each genome or subgroup contained distinct proteins. Furthermore, the nutritional/metabolic component associated with pyochelin proteins is found in all genomes except for strains FP1707, BML-PP031, NL2019, CTOTU48783, FP69, 88A6, 90F12-1, JCM 11,940, DSM 14,020, and LMG 17,764. The predicted unique virulence factors are given in Fig. 2, and all predicted virulence factors were included as supplementary file 4 for the readability.

Fig. 2.

Distribution of unique virulence genes within the P. rhodesiae genomes. Red boxes indicate the presence of unique virulence genes across all Pseudomonas rhodesiae strains, while blue boxes indicate the absence of the corresponding virulence genes. Hierarchical clustering was applied to group the genomes based on the distribution of virulence genes

Discussion

The current investigation involved sequencing the genome of P. rhodesiae strain SK22, which was isolated from disease outbreaks in Dicentrarchus labrax (European seabass). The genome was then compared with other available P. rhodesiae genomes using comparative genomics approaches. A total of ten countries—the USA, Belgium, Australia, Germany, Japan, Canada, Switzerland, France, China, and Turkiye—reported P. rhodesiae genomes. They originated from seventeen distinct hosts and were subsequently isolated (Table 1).

Multi-locus sequence typing (MLST) and 16 S rRNA sequences have been extensively used for the identification and categorization of prokaryotes. However, the similarities in the shared sequence (> 99%) may result in misclassification. Consequently, the use of ANI value has been strongly recommended in conjunction with the development of next-generation sequencing technology [28–30]. It should be emphasized that the ANI value should be greater than 95% for the same species [31]. The ANI value between the genomes was evaluated, and it showed that strains 26B3 and CIP104664 were not members of the P. rhodesiae species (Table 2). The analysis determined that isolates 26B3 and CIP104664 belonged to the P. quebecensis species, with an ANI value more than 99%. Thus, all P. rhodesiae genomes were studied to determine their relationship. A phylogenetic tree was constructed using the core genome, which consists of 3784 genes, rather than MLST and 16 S rRNA sequencing, to further validate the ANI values (Fig. 1). Based on both the ANI calculations and phylogenetic relationships, strains 26B3 and CIP104664 should be reclassified as P. quebecensis. Furthermore, P. rhodesiae genomes are split into clades. P. rhodesiae’s subgroup divergence enhances its ecological flexibility, potentially enabling it to colonize a wide range of environments, interact with a variety of species, and play a crucial role in system dynamics. This variation may result in a greater range of ecological responsibilities, from nutrient cycling and host interactions to antibiotic production and bioremediation, underscoring the complex relationship between microbial diversity and environments.

Integrons are genetic components which enable gene cassettes to be acquired, incorporated, and expressed, with many of them encoding antimicrobial resistance determinants. Integrons are not plasmids or transposons, yet they are commonly found inside these mobile genetic elements. This enables horizontal gene transfer and rapid adaptability of bacteria to antimicrobial pressure [18, 32–35]. Resistance genes that are present in both gram-negative and gram-positive bacteria are primarily derived from the class 1 integron [36]. Class 1 integrons have been shown to confer multidrug resistance, such as beta-lactamase and fluoroquinolones, in clinical isolates of P. aeruginosa [37–39]. In addition, Kiddee et al. examined the antimicrobial susceptibility, presence of class 1 integrons, and arrangement of gene cassettes of 50 P. aeruginosa isolates in northern Thailand. Class 1 integrons were present in 82% of the isolates, and the gene cassettes were associated with resistance to aminoglycosides, chloramphenicol, β-lactams, and rifampicin. Integron-positive isolates exhibited a higher frequency of antimicrobial resistance than those lacking integrons [40]. The class 1 integron is present in the genomes of the mineral water isolate strains DSM 14,020, JCM 11,940, and LMG 17,764 from France, but its function in conferring antimicrobial resistance is unclear (Fig. 2). These strains may have acquired this integron horizontally from other bacterial groups to adapt to environmental changes. Although integron-associated resistance determinants are discussed above based on genomic findings, a general definition of antimicrobial resistance is provided below to place these results within a broader biological and clinical context.

Bacterial antimicrobial resistance (AMR) is a developing global health issue in which bacteria develop antibiotic resistance. Infection susceptibility makes treatment more challenging, extending disease, increasing healthcare expenses, and increasing mortality. Bacteria can develop antimicrobial resistance due to random genetic mutations, horizontal gene transfer from other bacteria, and antibiotic-inactivating enzymes generated by the bacteria. Multiple-drug-resistant Gram-negative bacteria infections cause morbidity and mortality worldwide. These bacteria acquire antibiotic resistance quickly [20, 41–44]. While P. aeruginosa is the best known species of the genus, other species, including P. fluorescens and P. putida, may also demonstrate antimicrobial resistance (AMR). Multidrug-resistant (MDR) Pseudomonas species exhibit resistance to multiple kinds of antibiotics, including beta-lactams, aminoglycosides, fluoroquinolones, and carbapenems [45–52]. P. aeruginosa can gain resistance by overexpressing AmpC β-lactamases due to gene mutations. Transferable aminoglycoside-modifying enzymes (AMEs) lower the bacterial cell’s affinity for their target, making Pseudomonas resistant to them. Multidrug-resistant P. aeruginosa is common in clinics [53–55]. Svet et al. found that competitive interactions between P. rhodesiae and Raoultella terrigena, isolated from a brewery, significantly influenced the development of resistance to the antimicrobial sulfathiazole. While single-strain cultures showed limited resistance acquisition, mixed-strain cultures acquired resistance more readily, highlighting that microbial competition can accelerate antimicrobial resistance [56]. All analyzed P. rhodesiae genomes contained common antimicrobial resistance-related elements, but they also contained unique elements. Horizontal gene transfer may have occurred from different bacterial species in the region where the bacteria were isolated. Moreover, urban isolates and those from agricultural plants and fish possessed a greater number of antimicrobial resistance elements. The spread of antimicrobial resistance elements may have been influenced by the inappropriate or excessive use of antibiotics in production areas, which are frequently employed for disease prevention.

Bacteriophages that are integrated into bacterial chromosomes or plasmids are referred to as prophages. Integration increases bacterial genetic diversity and often brings host bacteria beneficial virulence traits like antibiotic resistance and toxins. Prophages can help bacteria grow in adverse environments or with antibiotics [22, 57, 58]. Johnson et al. conducted a comprehensive investigation of prophage sequences using 5,383 publicly available P. aeruginosa genomes from both environmental and human isolates. Prophages were discovered in a diverse range of populations, including tailed phages, inoviruses, and microviruses [59]. Most clinical isolate P. aeruginosa strains harbor prophages, which are present in numerous circulating strains and exhibit a similar clonal distribution pattern. Several viral defense proteins, including anti-CRISPR, toxin/antitoxin modules, and restriction-modification system proteins, as well as prophage interference with their host’s quorum sensing system and regulatory cascades, have been discovered. However, many open reading frames (ORFs) in prophages have unclear roles. This implies that prophages alter bacterial pathogenesis and anti-phage defense [60, 61]. Interestingly, P. rhodesiae strains isolated from agricultural animals such as fish and cows, as well as animal products such as milk and plants, displayed an unusually high number of phage sites and phage hit proteins (Table 4). Plasmids were found in only one genome (strain 3.8). Plasmids may be present in the strains, which include more phage areas. However, they were not reported due to the lack of a fully completed genome. The absence of plasmids in closed genomes greatly increases the likelihood of phage presence on the chromosome. Phages may have helped these strains form a distinct group within the lineage of evolution (Fig. 1). No obvious phage-borne AMR or virulence genes were discovered; however, prophages are thought to help with genome remodeling, gene transfer between organisms, and ecological fitness. They may also confer a selection advantage during the adaptation process, contributing to the development of antibiotic resistance and pathogenicity.

The protein secretion systems in bacteria are very complicated machines made up of molecules that transfer proteins and toxins within the bacterial cytoplasm to the outside or into other cells. Pathogenicity and disease are promoted in a variety of animals, people, and plants by these systems, through their critical role in numerous physiological processes, such as nutrition acquisition, toxin delivery, and interactions with host organisms. There are various types of bacterial secretion systems (T1SS-T6SS), each possessing a distinct role in pathogenicity [62–69]. There are three subclasses of type IV pili (T4P), known as fimbriae, which are dynamic, thread-like extensions on the surface of many Gram-negative bacteria. These subclasses are type IV class a pilin (T4aP), type IV class b pilin (T4bP), and tight adherence pili (Tad). They play an important role in the pathogenicity of many bacteria and have multiple roles related to virulence, such as in adhesion, motility, biofilm development, immunological evasion, and host-microbe interactions [70–72]. The flagellum is a complex molecular mechanism that helps bacteria’s motility, chemotaxis, adhesion, and biofilm formation. Although the bacterial flagellum and T3SS have diverse roles, including motility and pathogenicity, they share structural commonalities and evolutionary linkages. It affects bacterial survival and behavior [73, 74]. Major virulence factors in Pseudomonas include T1SS, T2SS, T4P, T3SS, flagella, T4SS, T5SS, and T6SS, which function in adhesion, biofilm formation, motility, and the effector delivery system of toxins and enzymes [27, 75, 76]. The genomes of P. rhodesiae provide evidence of the presence of secretion systems that have been identified in the past for Pseudomonas species (Tables 5 and 6). A significant number of the system’s genes are present, even though certain systems are lacking genes (Supplementary file 3). Although these secretion systems may play a role in bacteria’s ability to adapt to their surroundings and compete with one another for survival, they also have the potential to cause disease in the host. It is critical to note that these predictions represent what the genome is capable of, rather than what has been tested in the lab. Thus, the functional inference is not certain and requires confirmation from transcriptomic or proteomic data.

The identification of virulence factors is crucial for understanding pathogenicity and developing effective methods for preventing, diagnosing, and treating bacterial infections [77]. P. aeruginosa represents a significant human pathogen, implicated in a range of opportunistic infections, while P. syringae is recognized as a prominent plant pathogen within the Pseudomonas genus. A multitude of Pseudomonas species serve as opportunistic pathogens affecting animals, humans, and plants alike. Reports indicate that adherence-related factors such as Fap (Functional amyloid in Pseudomonas) and T4P [78, 79], biofilm-related alginate [80, 81] and Quorum sensing [82], along with effector delivery system-related secretion systems and motility-related flagella, immune modulation-related LPS and Rhamnolipid, as well as nutritional/metabolic factors including metal-uptake, iron-uptake, and the siderophore uptake system, are recognized as significant virulence factors of P. aeruginosa [83–86]. Many studies on virulence factors have been conducted to elucidate the pathogenicity of Pseudomonas, with several significant virulence factors identified [27]. P. aeruginosa can produce a significant siderophore known as pyochelin (PCH) during the period of an infection. This siderophore can help P. aeruginosa to survive in an iron-restricted environment and cause infection [87]. Moreover, P. rhodesiae can immobilize toxic lead carbonate by secreting organic acids that dissolve metal compounds, highlighting its potential in bioremediation [88]. P. rhodesiae genomes share most of the virulence factor-related genes (Supplementary file 4) while some genomes contain unique virulence factor-related genes (Fig. 2), such as adherence-related type IV pili, secretion systems, exotoxins, and immune modulation-related LPS genes. They may act as a potent virulence factor by activating and modulating the host immunological response. Similarly, nutritional and metabolic factor-related pyochelin genes may have a considerable impact on bacterial virulence for some strains by facilitating the uptake of essential nutrients, enabling adaptability to harsh environments, promoting the formation of protective biofilms, and producing substances that contribute to disease.

In conclusion, this study provides a high-resolution comparative genomic analysis of P. rhodesiae, including the newly sequenced strain SK22, obtained from Dicentrarchus labrax (European Seabass), as well as publicly available genomes from diverse environmental sources. Using whole-genome-based phylogeny and average nucleotide identity (ANI), we found two genomes that were incorrectly classified in public databases. This demonstrates the significance of genome-scale methodologies for the accurate classification of taxa within the Pseudomonas genus. Comparative studies revealed considerable genomic diversity among P. rhodesiae strains, including variations in putative antimicrobial resistance determinants, virulence-associated genes, secretion system components, integrons, and prophage content. These characteristics highlight the species’ genetic adaptability and its capacity to acquire and retain mobile genetic material across diverse ecological contexts. However, it is essential to emphasize that all resistance- and virulence-related characteristics discussed in this study are based on in-silico predictions derived from genomic sequence data. The presence of genes associated with antibiotic resistance, secretion systems, and other virulence-related processes indicates a hypothesized genetic potential rather than a confirmed biological function. No definitive conclusions regarding pathogenicity, resistance characteristics, or functional expression can be established without experimental validation. Consequently, the findings need to be regarded as a means to formulate novel hypotheses and establish a genetic framework for forthcoming functional investigations. Additionally, even though the genomes that were studied come from different places and environments, there aren’t many P. rhodesiae genomes available, and they aren’t all in the same place. This makes it hard for us to draw broad conclusions about the species’ global population structure, ecological niche breadth, and evolutionary dynamics. This research establishes the fundamental genetic resource for P. rhodesiae and identifies significant areas for future exploration. Experimental validation, including phenotypic characterization, expression analyses, and host interaction studies, as well as enhanced genome sampling, is essential to determine the biological significance of the predicted traits and to clarify the ecological and potential clinical relevance of this underexplored species.

Authors’ contributions

SK: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. FP: Data curation, Formal analysis, Methodology, Visualization, Validation, Writing – review & editing. JB: Data curation, Methodology, Resources, Software. AE: Data curation, Resources, Methodology, Validation, Writing – review & editing. SK: Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Project administration, Resources, Validation, Visualization, Writing – review & editing. LAH: Conceptualization, Data curation, Methodology, Validation, Supervision, Writing – review & editing. HCT: Conceptualization, Data curation, Methodology, Visualization, Supervision, Validation, Writing – review & editing.

Funding

This research has been supported by the Scientific Research Projects Coordination of Recep Tayyip Erdogan University under the project number FBA-2021-1264.

Data availability

The genome sequence data of P. rhodesiae strain SK22 from the current study have been submitted to the National Center for Biotechnology Information (NCBI) genome database, assigned GenBank accession number JBFQEB000000000 and Bioproject PRJNA1138762. The NCBI genomes database was used to get all genomes of P. rhodesiae (https://www.ncbi.nlm.nih.gov/datasets/genome/?taxon=76760). Also, the information supporting the study’s findings is available in the main paper and its supplementary files.

Declarations

Ethics approval and consent to participate

Not applicable.

Conflict of interest

The authors state there are no conflicts of interest.

Competing interests

The authors declare no competing interests.