Virulence and antibiotic-resistance genes in Enterococcus faecalis associated with streptococcosis disease in fish

Isolation, phenotypic identification, pathogenicity and antibiogram profiling

Enterococcus faecalis strains BFF1B1, BFFF11 and BFPS6 were cultured in Streptococcus selective agar media (Himedia, India). The culture characteristics such as colony, morphological, physiological and biochemical characteristics of these strains BFF1B1, BFFF11 and BFPS6 were summarized in the Supplementary Table S1. All of these strains were Gram-positive, cocci shaped, non-motile and produce dark red colonies while culturing in agar media. They were catalase, oxidase, urease, Arabinose, Fructose, Inositol, Inulin, Raffinose and Xylose negative. They produced β hemolysis in sheep blood agar media. The antibiogram profiling of these strains were performed using disk diffusion methods against eleven commercial antibiotics (Supplementary Table S2). We observed that all strains were resistant to Amoxicillin, Ampicillin, Cefuroxime, Erythromycin and Penicillin. Moreover, the strain BFPS6 was also resistant to Cefradine. Furthermore, in-vivo challenge test revealed that BFFF11, BFF1B1 and BFPS6 were identified as highly virulent (≥ 80% mortality) strains (Supplementary Table S1).

General features of the genome sequences

The genomic characteristics of the three strains of E. faecalis associated with streptococcosis in fish were studied using a subsystem set of seed viewer. The genome sizes of the strains BFF1B1, BFFF11 and BFPS6 were 2.76, 3.07 and 2.87 Mb, respectively. The GC content (%) of the strains was substantially the same (Table 1). The predicted coding sequences (CDSs) by RAST in the strains BFF1B1, BFFF11 and BFPS6 were 2588, 2870 and 2743, respectively. Only 30–52% CDSs in each strain can be functionally categorized into 250–357 subsystems. On the other hand, PATRIC analysis revealed 2631, 2949 and 2745 CDS in the strains BFF1B1, BFFF11 and BFPS6, respectively. In the subsystems, carbohydrates, amino acids and derivatives and protein metabolism had a higher number of functional genes in all of the studied bacteria. None of the strains carried genes related to photosynthesis and secondary metabolism. An overview of the genome features of the E. faecalis strains and their subsystem statistics were shown in Table 1 and Supplementary Fig. S1.

Table 1 General genomic features of the E. faecalis strains BFFF11, BFF1B1 and BFPS6 obtained from RAST and PATRIC analysis.

According to the Kyoto Encyclopedia of Genes (KEGG) study, the CDSs were classified into six sub-categories including cellular processes (CP), environmental information processing (EIP), genetic information processing (GIP), human disease (HD), organisms system (OS), and metabolism (M) (Fig. 1). The strains BFF1B1, BFFF11 and BFPS6 were found to conserve 1280, 1325 and 1356 CDSs, respectively, that were further divided into 37 functional KEGG sub-categories according to the aforementioned six categories (Fig. 1 and Supplementary Table S3). Notably, all of the strains contained a significantly higher number of genes related to metabolic pathways (CDSs 753–826), followed by environmental information processing (CDSs 199–197) and genetic information processing (CDSs 186–187) (Supplementary Table S3).

Figure 1
figure 1

The number of genes categorized by KEGG functional annotation of E. faecalis strains BFF1B1, BFFF11 and BFPS6.

A comparative genomic analysis was performed among the strains where E. faecalis V583 was used as a reference. The genomic map obtained from the BRIG comparison did not show large scale variation between the bacterial genome sequences, and a significant number of non-homologous regions were found around the reference genome with over 80% identity (Fig. 2). Most of these non-homologous regions might be linked to transposable elements.

Figure 2
figure 2

Blast atlas of three Enterococcus faecalis strain (BFFF1B1, BFFF1 and BFP6S6) mapped against reference sequence of Enterococcus faecalis V583. Blast atlas were generated by BRIG using both alignment length and identify cut off values minimum of 50%. The innermost two rings (first and second) represent GC content (black) whereas, the third ring shows GC skew (purble/ green). The remaining 4 rings (rings 4–7) represent a BLASTN comparison with complete genome of E. faecalis strains V583 (megenda ring which was used as reference genome) BFF1B1 (Deep sky), BFFF11 (Blue) and BFPS6 (Maroon).

Virulence factor and biofilm formation-associated genes

The degree of pathogenicity of microbes is greatly influenced by their virulence gene contents. In this study, a total of 69 virulence genes were identified in E. faecalis strains BFFF11, BFF1B1 and BFPS6 (Fig. 3a). Genes associated with protection against oxidative stress (tpx, perR), bacterial cell wall synthesis (psr) and gelatinase toxin (gelE) were identified in all of the three strains. Two extracellular hyaluronidases genes hylA and hylB that evade the phagocytosis process with macrophage persistence of host were identified only in the genome sequence of BFFF11, whereas hylA was found in both BFF1B1 and BFPS6 strains.

Figure 3
figure 3

Heat maps of virulence genes. (A) Presence and absence of VG. Dark red = presence in all three strains, Light red = Presence in BFPS6, Orange = presence in BFFF11, Dark cream = presence in BFF1B1, Light cream = Presence in BFFF11 and BFPS6, Light blue = Presence in BFF1B1 and BFFF11 and Dark blue = absence of VG. (B) Identification of VG according to database. Dark red = Present in all database used for this study, Light red = PATRIC_Victors, Orange = VFDB and PATRIC_VFDB, Orange = VF and PATRIC_Victors, Dark cream = VF and PATRIC_VFDB, Light cream = VF and VFDB, PATRIC_VFDB, Light Blue = VFDB, Sky blue = VF, Dark blue = absence in database.

Many genes associated with biofilm formation were identified such as two aggregation substances (agg and prgB), endocarditis and biofilm-associated pili genes (ebpA, ebpB, ebpC), collagen adhesion precursor (ace), three proteolytic processing of a quorum-sensing system signal molecule precursors (fsrA, fsrB and fsrC), accessory regulator protein (agrBfs), sugar-sensing transcriptional regulator (bopD), serine protease (sprE) and two genes for sortase assembled pili (srtA and strC). Although strain BFFF11 harbored all of the above biofilm-producing factors; two aggregation substance encoding genes agg and prgB were absent in the genomes of BFF1B1and BFPS6.

Numbers of virulence genes for DNA and protein synthesis were found in the genome sequences of all three strains, namely DNA repair enzymes synthesis gene (recQ1 and phrB), purine metabolism gene (purl), thymidylate synthase (thyA), methionine aminopeptidase (map) and sucrose operon repressors genes (scrB-1 and scrR-1). Among three strains, BFF1B1 and BFFF11 harbor ctrA gene that functions as a negative regulator of DNA replication.

Other virulence factors included sex pheromone associated genes (cad, cCF10, camE, cOB1), the cell wall adhesion expressed in serum gene (efaA), the enterococcal Rgg-like regulator gene, amino acid transport and synthesis regulators (brnQ), and gene associated with macrophage persistence (ElrA). Cytolysin toxin-producing genes cylR2 and cylI was identified in BFF1B1 by VFDB and BFPS6 by PATRIC, however, these were not identified in the BFFF11 strain. Furthermore, 11 capsule producing genes (cpsA to cpsK) associated with anti-phagocytosis were identified in the strain BFFF11; among those only 3 genes (cpsA, cpsB and cpsF) were in the Thai sarpunti originated BFPS6 and only cpsA was found in the BFF1B1. Heat shock regulation protease gene (clpP), translation elongation factor (tufA) and UDP-galactopyranose mutase synthesis gene (glf) which play important roles in cell surface formation and infection cycle of pathogens were found only in the strain BFF1B1. A relatively large number of virulence genes were identified using the VFDB database, and the lowest was identified using the virulence finder (Fig. 3b).

Antibiotic-resistance gene

A total of thirty-nine antibiotic-resistance genes belonging to sixteen different groups were identified among the strains of E. faecalis (Table 2). Except for four genes including tet(M), tet(L), tet(S) and tet(45), all of the genes were conserved by the genome sequences of all three strains of E. faecalis. Although tetracyclines, aminoglycosides, phenicol antibiotics resistant gene YkkCD and tetracyclines, glycylcyclines resistant gene S10p were conserved in all the genome sequences of present study strains, four genes such as tet(M), tet(L), tet(S) and tet(45) conferring resistance to tetracycline were identified only in the genome sequence of BFPS6 with 77.14–100% of identity.

Table 2 Acquired antibiotic resistance genes identified in the strains of E. faecalis obtained by ARG-ANNOT Nt, Resfinder and CARD.

Macrolide-lincosamide-streptogramin (MLS) resistant genes lsa(A), RlmA(II) and mph(D) were found in all of the E. faecalis strains. Similarly, two multidrug-resistant efflux pump conferring genes efrA and efrB were identified, found to be resistant against MLS and rifamycin antibiotics. Eight antibiotic-resistance genes included LiaR, LiaS, LiaF, MprF, GdpD, PgsA, rpoB and rpoC were found where seven of them were resistant against daptomycin and rpoB were resistant against rifamycin, daptomycin, rifabutin and rifampin drugs. All three strains harbored genes kasA, FabK, inhA and fabI were resistant to isoniazid and triclosan group of antibiotics. Furthermore, two genes dfr(E) and folA were identified, resistant to diaminopyrimidine (drug class of Trimethoprim), two Aminoglycosides genes gidB and S12p are resistant to streptomycin, two cycloserine resistant genes (Alr and Ddl) and two fluoroquinolones and quinolones resistant genes (gyrA and gyrB) were identified in the genome sequences of all three studied strains. Other notable antibiotic-resistant genes identified in the genome sequences of the studied strains were rho, EF-G, EF-Tu, MurA, Iso-tRNA, folP, VanG and ampS. These genes were found to be resistant against bicyclomycin, fusidic acid, elfamycins, fosfomycin, mupirocin, sulfonamides, and vancomycin and beta-lactamases antibiotics, respectively.

Secondary metabolites

Secondary metabolites often are considered potent sources of virulence factors for a pathogen. Although no potential gene cluster responsible for the biosynthesis of microbial metabolites were found in the strain BFF1B1 by using the antiSMASH software, two gene cluster were identified in the BFFF11, including NRPS (Nonribosomal peptides synthetases) and bacteriocin. Furthermore, a putative bacteriocin gene cluster linked to the potentially interested metabolites was found in the genome sequence of BFPS6.

Bacteriophages

Prophages are bacteriophage mediated mobile genetic elements generally transferred by the transduction process and enable bacteria to obtain antibiotic-resistance or virulence genes. Bacteriophage helps bacteria to become more pathogenic and adapt to a new environment. Bacteriophage related antibiotic-resistance factors were demonstrated by using a phage search tool PHASTER and summarized in Supplementary Table S4. Genomes of the studied Enterococcus strains harbored one incomplete phage region with length 14.6 Kb and GC content ranging from 38.68 to 38.76%. Although one incomplete prophage was identified from genome sequences of BFFF11 and BFF1B1, BFPS6 conserved two incomplete regions. Interestingly, all strains conserved equal length (14.6 kb) of the incomplete region and hit a similar number (15) of phage proteins. The genome of strains BFFF11 and BFPS6 contained one putative intact phage with sizes 40.3 and 37.4 Kb, respectively.

Insertion sequences

Insertion sequences (IS) are small mobile genetic elements that are widely distributed in the bacterial genome. Several groups of IS are found in the genome sequence of bacteria. In the present study, 25 distinct families of insertion elements/ transposes insertions were identified in the genome of E. faecalis strains (Supplementary Table S3). Among these, 14 families (IS3, IS4, IS5, IS6, IS200/IS605, IS256, IS481, IS607, IS630, IS701, IS1595, ISL3, ISKra4 and ISNCY) of transposases were common to the three strains, 3 families (IS30, IS66, IS91) were in BFF1B1 and BFFF11, two were (IS982 and IS1182) in the BFFF11 and BFPS6 and rest 6 families (IS1, IS110, IS1380, IS1634, ISAs1 and Tn3) were found only in the strain BFFF11. In the case of a frequency distribution of IS families, BFPS6 harbored the relatively highest number of IS (N = 115), followed by BFFF11 (N = 85) and BFF1B1 (N = 46). IS family IS200/IS605 were identified as a higher frequency in the strains BFFF11 and BFF1B1 which were 14 and 12, respectively. Significantly higher frequency of IS6 (n = 37) families were identified in BFPS6, followed by IS3 (n = 21) and IS200/IS605 (n = 16). The distribution of the IS families and their members were summarized in Supplementary Table S5.

Phylogenetic tree

A phylogenetic tree was constructed based on the single nucleotides polymorphisms (SNPs) analysis. To understand the degree of relatedness with other pathogenic E. faecalis, the whole genome sequence of human, mouse and pig were extracted from NCBI database and used for the analysis of the phylogenetic tree. As E. faecalis caused streptococcosis like infection in fish, fish pathogenic Streptococcus spp. from NCBI were also included in this analysis. In the phylogenetic tree, two distinct clades were formed among the three strains of the fish pathogenic E. faecalis. Tilapia originated strains BFF1B1 and BFFF11 formed a common cluster with the reference strain whereas BFPS6 formed a separate branch closely related to a human originated strain (Fig. 4). Although common disease symptoms were found, fish originated Streptococcus spp. showed a distinct out-group from the E. faecalis pathogen in the phylogenetic study. The average percentage of reference genome covered by all strains was 0.0292%, whereas, in the present study, strains BFF1B1, BFFF11 and BFPS6 covered 78.675, 83.747 and 80.419%, respectively.

Figure 4
figure 4

The SNP based phylogenetic tree was obtained from CSIphylogeny v1.4 by using reference genome of E. faecalis strain V583 (Accession No AE016830). The dendogram was modified by Fig.Tree v1.3.

Read more here: Source link