Introduction

Streptococcus intermedius belongs to the S. anginosus group (SAG) that also includes S. constellatus and S. anginosus [1]. It is part of the normal oral cavity and upper respiratory tract floras, as well as those of the gastrointestinal and female urogenital tracts [2,3,4,5]. This bacterium was first described by Guthof in 1956 after being isolated from dental abscesses [6]. S. intermedius may also cause human infections, usually monomicrobial, including purulent abscesses of the liver, lungs, psoas, spine and/or central nervous system, and infective endocarditis [7]. Over the years, the role of S. intermedius in human infections has increasingly been reported. Patients with invasive S. intermedius infections were described to cause significantly longer hospital stays and higher mortality than patients with other S. anginosus group infections, suggesting that identifying this species might be important for the management of patients [8].

Various putative virulence factors have been described for Streptococcus intermedius, among which the ability to form biofilms to protect itself from antibiotics and the host immune system [9], the production of hydrolytic enzymes, including both glycosaminoglycan-degrading enzymes, such as hyaluronidase and chrondroitin sulphate depolymerase, and glycosidases, such as α- N- acetylneuramidase (sialidase), β-D-galactosidase, N-acetyl-β-D-glucosaminidase and N-acetyl-β-D-galactosaminidase, which allow S. intermedius to grow on macromolecules found in host tissue [10]; a cytotoxin, intermedilysin (ILY), that can directly damage host tissues and immune defense cells and participate in bacterial pathogenicity; and the surface protein antigens I/II that are involved in adhesion to fibronectin and laminin, which is an important step in the pathogenesis of endocarditis and abscess formation [11].

The development of high throughput nucleic acid sequencing technologies has enabled observing variations of the genetic repertoire among strains of a given bacterial species. Our present study analysis aimed at describing the genetic diversity and pathogenesis substratum of S. intermedius. Twenty-seven genomic sequences from S. intermedius strains, including 13 newly sequenced from our laboratory and 14 from public databases, were used for pan-genomic analysis. Predicted genes were compared among strains to determine the size of the core and dispensable gene pools, the pangenome, the gain/loss of putative virulence determinants, and to identify genomic islands.

Accession numbers

The 13 genome sequences determined in this study were deposited in GenBank and their accession numbers are listed in Table 1.

Table 1 Genomic characteristics of the 27 studied S. intermedius strains
Full size table

Materials and methods

Extraction and genome sequencing

The genomic DNA (gDNA) of each studied S. intermedius strain was extracted in two steps: a mechanical treatment was first performed using acid-washed glass beads (G4649-500 g Sigma) and a FastPrep BIO 101 instrument (Qbiogene, Strasbourg, France) at maximum speed (6.5) for 90 s. following a 2-hour lysozyme incubation at 37 °C, DNA was extracted using an EZ1 biorobot and the EZ1 DNA Tissue kit (Qiagen, Hilden, Germany). The elution volume was 50µL. Genomic DNA was quantified using the Qubit assay (Life technologies, Carlsbad, CA, USA).

The gDNAs were sequenced using a MiSeq sequencer with the Paired-End strategy and the Nextera XT library kit (Illumina, Inc, San Diego, CA, USA). The Paired-End library was prepared using input solutions of 1 ng gDNAs. The gDNAs were fragmented at the tagmentation step. Then, limited cycle PCR amplification (12 cycles) completed the tag adapters and introduced dual-index barcodes. After purification on AMPure beads (Life technologies, Carlsbad, CA, USA), the libraries were normalized according to the Nextera XT protocol (Illumina). Normalized libraries were pooled for sequencing on a MiSeq sequencer (Illumina). Automated cluster generation and paired-end sequencing with dual index reads was performed in a single 39-hour run in a 2 × 250 bp format. The numbers of paired-end reads were summarized in Table 2. The paired-end reads were trimmed and filtered according to the read qualities.

Table 2 Genome sequencing details of the 13 S. intermedius strains from our study
Full size table

Genome assembly, annotation and comparison

After sequencing, the obtained reads were assembled using the A5 software [12] with default parameters and then contigs were compared to NCBI using BLASTn to remove contaminations. Then, the online tool Fasta dataset joiner (http://users-birc.au.dk/biopv/php/fabox/fasta_joiner.php) was used to merge sequences into a single molecule. The Mauve software was used for multiple genomic sequence alignment [13]. Genes were annotated using the Prokka software with default parameters [14] in which the similarity e-value cut-off is 0.000001 and the minimum contig size is 200 bp. This pipeline also includes several other tools like Aragorn for tmRNA detection, Barnap to count rRNAs and Prodigal to identify coding sequences. To estimate the mean level of sequence similarity at the genome level among studied strains, we used the OrthoANI [15] and Genome-to-Genome Distance Calculator (GGDC) [16] softwares, with the following respective threshold values of 95–96 and 70 %.

Phylogenetic analysis

A 16 S rRNA-based phylogenetic analysis of the 27 studied S. intermedius strains was performed using the MEGA 7 software [17]. For constructing the phylogenetic tree, the following options were used: Maximum Likelihood method; Kimura 2-parameter model for substitution model; uniform rates among sites; partial deletion option for gaps/missing data; 1000 bootstrap replicates.

Using genomic sequences and the Roary program [18], a clustered heatmap of core genes was drawn on the basis of the presence/absence approach [18]. We also detected SNPs with the snp-sites program [19] from the core genome alignment and drew a phylogenetic tree with CGEwebface [20].

Virulence factor analysis

Virulence-associated genes were detected by comparing studied genomic sequences with the virulence factor database (VFDB) [21] and sequences described in recent publications [22]. The BLASTp search was performed using the threshold scores reported by Olson et al.: 35 % identity and highest scoring pair length of 50 % [22]. Additionally, we reviewed the literature to identify the proteins involved in interactions with the host [10, 23, 24]. A principal component analysis was performed using the XLSTAT program (Data Analysis and Statistical Solution for Microsoft Excel, Addinsoft, Paris, France 2017) in which the Fisher’s least significant difference (LSD, α = 0.005) and Pearson’s correlation coefficients were used, to detect any association of virulence-associated genes with specific clinical conditions.

Core and pan-genome analysis

Get_homologue [25] was used to reveal orthologous genes among S. intermedius strains, using the following parameters: minimal coverage (-C) 40 %, minimum identity (-S) 50 %, minimum e-value (-E) 1e-05. Sequence similarity searches and clustering of coding sequence (CDS) from the 27 genomes were performed using pairwise BLASTp and OrthoMCL algorithms [26]. Sequential inclusion of all possible combinations of up to 27 strains were simulated and fitted by regression analysis [27] of the amount of conserved genes and of strain-specific genes. This allowed to estimate and extrapolate the sizes of core- and pan-genomes. Roary [18] was also used, with default parameters, to confirm the reliability of the obtained pan-genome analysis results (identity percent ≥ 70 %, coverage ≥ 70 %) and to generate the core genome alignment.

Functional classification of orthologous cluster analysis

The Clusters of Orthologous Groups (COGs) database was used to identify gene functions [28] using BLASTP (E-value 1e− 03, coverage 0.7 and identity percent 30 %).

A circular comparison of genomes was obtained using the online GView Server (https://server.gview.ca/) with S. intermedius strain ATCC 27,335 as reference genome [29]. ResFinder and the ARG-ANNOT database were used to search antibiotic resistance-related markers [30, 31]. The presence of CRISPR repeats and prophages was predicted using the CRISPRFinder [32] and PHASTER softwares, respectively [33].

Results and discussion

Strain characterization

The 27 studied S. intermedius strains originated from China, Canada, South Korea, US, Japan and France. The patients’ data was not available for some strains. The 13 French strains (G1552-G1557 and G1562-G1568, Tables 1 and 2) were isolated in our laboratory from patients with various infections (Table 2), from August 2014 to November 2016, on 5 % sheep blood-enriched Columbia agar (BioMérieux) at 37 °C in anaerobic atmosphere. Their identification was confirmed by the high scores (> 2) obtained using MALDI-TOF MS. In addition, 14 S. intermedius genome sequences were retrieved from GenBank. The 27 strains were divided into 8 groups according to their isolation source (Table 2). The genome sizes and gene numbers among S. intermedius strains were relatively similar, consisting for each strain in a single chromosome but no plasmid was identified in any strains and ranging in size from 1.85Mbp to 2.05Mbp (Table 2).

A schematic view of all 27 studied genomes is provided in (Fig. 1), showing an overall high degree of conservation. The general features of S. intermedius genomes are summarized in Table 2 The G + C content of S. intermedius ranged from 37.3 to 37.8 % (avg 37.641 %, n = 27). All 13 in-house sequenced S. intermedius contained at least 47 tRNA genes, and the number of rRNAs for all strains ranged from 3 to 6. Streptococcus intermedius exhibited an average 1870 CDs with a mean length of 907 bp, accounting for 87.3 % of the whole genome.

Fig. 1
figure 1

Circular representation of the 27 studied S. intermedius genomes. Genomic sequences were aligned using strain ATCC 27335 as reference. The alignment gaps tend to coincide with the regions of low G + C contents. The rings, from the inside out, display the size in kbp; GC skew; G + C content; followed by genomes as listed in the left legend

Full size image

Phylogenetic analysis

The 16 S rRNA-based phylogenetic analysis (Fig. 2), widely used as a gene marker to differentiate Streptococcus species [34], demonstrated that all S. intermedius strains were grouped in a single cluster that was closely-related to S. anginosus and S. constellatus within the S. anginosus group [22] (Fig. 2). In the topology S. intermedius, S. constellatus and S. anginosus strains clustered together with their sub-species. However, the heatmap obtained using Roary [18], based on the core genome, was more discriminatory within the species than the 16 S rRNA-based analysis and identified 3 clusters that were independent from the strain source (Fig. 3).

Fig. 2
figure 2

16S rRNA-based phylogenetic relationships of S. intermedius strains using the Maximum Likelihood method with Kimura 2-parameter. The scale bar indicates the evolutionary distance between the sequences determined by a 0.005 substitution per nucleotide position. Numbers at the nodes indicate bootstrap values obtained from 1,000 replicates

Full size image
Fig. 3
figure 3

Clustered gene presence/ absence and accessory genome distribution calculated by pangenome analysis among the 27 studied S. intermedius strains. Left: core-genome phylogeny; the three clusters in the dendrogram are delineated by red lines; right: heatmap of core genes

Full size image

The three clusters are as follows: strains G1557, G1556, LC4, G1562, SK54, AJKN01, ATCC27335 and JTH08 constituted the first group, strains 30,309, G1563, G1564, 631SC0N and G1554 clustered in the second group while the remaining strains clustered in a third group. There was neither evidence of correlation between strain clusters and their clinical forms, nor between genomic types and the geographical origin of isolates.

To measure the divergence between all studied strains at a deeper level, we also analyzed their phylogenetic relationships on the basis of core genome SNPs, which demonstrated that strains G1562, G1566 and FO413 diverged from other strains and exhibited a higher tendency of recombination. However, again no disease-specific clustering was observed (Fig. 4).

Fig. 4
figure 4

Phylogenetic tree of S. intermedius strains based upon SNPs extracted from the core genome. Sequences were aligned using ClustalW with default parameters and phylogenetic inferences obtained using the Maximum likelihood method within the MEGA, version 7, software. Nodes indicate bootstrap support from 1000 replicates.

Full size image

Genomic similarity

Digital DNA-DNA hybridization (dDDH) values ranged from 80.5 to 99.3 % between all 27 strains, thus confirming their classification within a single species. This was also cross-validated by the OrthoANI program, which produced pairwise values ranging from 97.78 to 100 % which is well above the consensus 95–96 % threshold for prokaryotic species demarcation [35]. This corresponded to 100 % 16 S rDNA sequence identity across all studied isolates. The above data correlate with a strong degree of genome conservation and synteny.

Functional classification of orthologous cluster

The overall distribution of S. intermedius proteins in COG categories was quite similar in all 27 studied strains (Fig. 5). Previous studies of other Streptococcus species also suggested that, within a given species, the majority of strains had a similar COG profile [36,37,38]. Approximately 79.72 % of all proteins predicted in all strains were identified in COG superfamilies. The proportion of each category fluctuated within a very small range, showing almost similar percentages of distribution in all strains. The most abundant sub-categories were related to carbohydrate transport and metabolism (G) and translation, ribosomal structure and biogenesis (J) like their distribution in core genes.

Fig. 5
figure 5

Differential distribution of COG functional categories in S. intermedius: a proportion of six classes of functional categories in strain-specific and core genes; b functional categories in strain-specific and core genes; c functional categories in the 27 S. intermedius strains. Category abbreviations are as follows: C, energy production and conversion; E, amino acid transport and metabolism; F, nucleotide transport and metabolism; G, carbohydrate transport and metabolism; H, coenzyme transport and metabolism; I, lipid transport and metabolism; P, inorganic ion transport and metabolism; Q, secondary metabolites biosynthesis, transport and catabolism; X, mobilome: prophages, transposons; A, RNA processing and modification; B, chromatin structure and dynamics; J, translation, ribosomal structure and biogenesis; K, transcription; L, replication, recombination and repair; D, cell cycle control, cell division, chromosomal partitioning; M, cell wall/membrane/envelope biogenesis; N, cell motility; O, posttranslational modification, protein turnover, chaperones; T, signal transduction mechanisms; U, intracellular trafficking, secretion, and vesicular transport; V, defense mechanisms; W, extracellular structures; Z, cytoskeleton; R, general function predicted only; S, function unknown

Full size image

Less than half of strain-specific genes, but more than 90 % of core genes, had a match in the COGs database. The most abundant functions in core genes were associated with metabolism (Fig. 5a). The overall proportion of metabolic functions in core genes was 32.47 %, whereas that in strain-specific genes was 9.58 %. More specifically, energy production and conversion (C), amino acid transport and metabolism (E), nucleotide transport and metabolism (F), carbohydrate transport and metabolism (G) and coenzyme transport and metabolism (H) were noticeably more abundant in core genes (p-value < 0.01) (Fig. 5b). No mobilome-related functions were detected in S. intermedius. The functional category of information storage and processing showed highly different proportions in sub-categories (Fig. 5b). The functions of translation, ribosomal structure and biogenesis (J) were significantly enhanced (p-value < 0.0001) in core genes, whereas the functions of replication, recombination and repair (L) were significantly enhanced (p-value < 0.01) in strain-specific genes. This trend was also observed in other bacteria [35]. In the cellular processing and signaling category, the function of defense mechanisms (V) was more abundant in strain-specific (p-value < 0.05) than in core genes (Fig. 5c).

Pan- and core-genome analyses

The average number of new genes added by a novel genome was 40 when the 27th genome was added (Fig. 6). The exponential decay model shown in Fig. 7a suggests that the number of conserved core genes approached an asymptote with the comparison of 27 genomes. A total of 1,355 core genes were identified in S. intermedius. The average proportion and sequence identity of core genes per strain were 72 and 97.79 %, respectively, indicating that core genes in S. intermedius are highly conserved and reflecting a low degree of intraspecies genomic variability too. Examination of the functional annotation of these core genes suggests, as expected, that they encode mostly core metabolic processes.

Fig. 6
figure 6

Plot representing the numbers of new and unique genes found as each isolate of S. intermedius is added

Full size image
Fig. 7
figure 7

Fitted curves indicating the characteristics of core- and pan-genomes from 27 studied S. intermedius strains. a curve showing the relationship between the core genome and the number of genomes, b curve showing the relationship between the pan-genome and the number of genomes. As the number of genomes sequenced increased, the pan-genome size increased, whereas the core-genome size decreased, thus indicating an open pan-genome model. The gradual extension of the pangenome with addition of new genomes describes an open pan-genome model of S. intermedius. The number of genes that each strain contains is documented from comprehensive statistical analysis given earlier in Table 2

Full size image

A total of 1,054 strain-specific genes were identified in S. intermedius and the average number of strain-specific genes was 39 (Fig. 7b). Among strain-specific genes, 148 genes were found in strain G1562, 107 in strain TYG1620, 105 in strain BA1, 96 in strain 32,811, 82 in strain G1557, 73 in strain C270, 69 in strain 631_SCON, 61 in strain G1555, 41 in strain G1554, 38 each in strains G1565 and F0413, 33 each in strains G1564 and LC4, 27 each in strains G1556 and ATCC27335, 20 in strain 30,309, 16 in strain G1553, 13 in strain G1552, 11 in strain B196, 6 in strain KCOM1545, 5 in strain FDAARGOS_233 and 1 in strains G1563, G1566, G1567, JTH08, SK54AJKN01, respectively. The size of the pangenome increased steadily without reaching any plateau. The pangenome trend depicted in (Fig. 7b) shows a gradual expansion by addition of new genomes and thus the pangenome of S. intermedius may be considered as open and indicates a homogenous pattern of genome evolution with similar rates of gene gain/ loss process across the whole population. In addition, a total of 1,611 accessory genes that were shared by two or more strains were identified. Overall, we identified a S. intermedius pangenome of 4,020 genes including 1,355 core genes, 1,054 strain-specific genes and 1,611 accessory genes.

Virulence factors and principal component analysis

In the S. intermedius pangenome, 252 virulence factors were identified in total. Of these, 70 core virulence factors were shared by all strains and 78 unique virulence factors were present in one strain each (Table 3). Virulence-associated genes present in all studied genomes included homologous virulence genes that contribute to bacterial avoidance of the immune system, such as ily which encodes an intermedilysin, the lmb, pspA, pavB/pfbB, fss3 genes coding surface proteins, the genes coding the polysaccharide capsule (cps4A, cps4B, cps4C, cps4D, cps8D), the auto-inducer LuxS (luxS), the binding proteins (pavA, hitC, fbpC, psaA, mntA, clpC, fss3), neuraminidase (nanA), hyaluronidase (hysA), and heat shock protein B (htpB), genes from the sil locus known to play a role in quorum-sensing and virulence in S. pyogenes (silA, silD, silE, salX), genes associated with secretion systems (lem11, lem15, sdeC, ceg32, esxA, essC, lpg2372, lirB), and genes associated with Mg2+ transport proteins (mgtB, mgtC); the response regulator CsrR beta-hemolysin gene (cylG), lamanin-binding surface protein like Pac and invasion protein inlA were also present in all strains.

Table 3 Unique Virulence-associated genes detected in the 27 studied S. intermedius genomes
Full size table

Among these core virulence genes, the surface protein antigen I/II that was demonstrated to play a potential role in S. intermedius pathogenesis [39], and human fibronectin and laminin that are supposed to bind to this antigenic protein induce IL-18 release from monocytes [39]; genes from the streptococcal invasion locus (sil) are related to enhanced virulence in the SAG group and may contribute to the invasive behavior of S. intermedius strains; the internalin (inlA), likely acquired from Listeria monocytogenes, increases the virulence of S. intermedius by playing a key role in attachment to host cells [40]; the hyaluronidase (hysA) acts in the liquification of tissues and is also involved in biofilm formation, which protects bacteria from host defenses and antibiotics, and plays a role in infection [9]; the ily-coded intermedilysin can directly damage host tissues and immune defense cells, causing human cell death by membrane bleb formation [23]. It has been also reported that intermedilysin helps in invasion and adhesion of bacteria to human liver cells, and in cytotoxicity [41]; the galE gene codes galactose which plays a role in biofilm formation and its key residues are essential for epimerase activity [42]; the laminin-binding surface protein, homologous to that in Streptococcus agalactiae is coded by the Pac gene and is essential in binding and invasion of different host surfaces, and is present in almost all group B Streptococcus strains causing pneumonia, septicemia and meningitis [43, 44]; psaA codes a surface lipoprotein that plays a role in Streptococcus pneumoniae systemic infections by interacting with monocytes [45]; we also identified the heat shock protein-coding gene htbB that is known in Legionella pneumophila to act in adhesion to host fibronectin [46]; the clpC gene codes a heat shock protein involved in the invasion of hepatocytes in Listeria monocytogenes and has an ATPase activity [47]; ATPase proteins were shown to play a role in the survival and virulence in Salmonella typhimurium and S. aureus [48] ; clpP codes an ATP-dependent caseinolytic protease that was proven in Streptococcus suis to play a role in colonization and bacterial adaptation to various environmental stresses [49], pavB codes a fibronectin-binding protein that mediates bacterial attachment to human epithelial and endothelial cells and also promotes transfer of bacteria to the bloodstream [50, 51]; and nanA codes a highly conserved neuraminidase that also possesses a sialidase activity to catalyze the cleavage of terminal sialic acid residues from glycoconjugates. In S. pneumoniae, it promotes biofilm formation and contributes significantly to broncho-pulmonary colonization [52].

Although most of the strains exhibited one to eight unique virulence genes, strains G1562 and BA1 possessed 14 and 10 specific virulence genes, respectively. Eight strains (G1563, G1566, G1567, G1568, BA1, KCOM1545, JTH08, SK54AJKN01) had no strain-specific virulence factor (Table 3).

Among unique virulence genes, sdcA, ybtE,lpbA, SalR,salK, VopT are secretory system-associated genes that are involved in iron-mediated transport across cellular membranes. Some of these genes are linked with bacterial growth and act as important anti-inflammatory effectors [42, 53,54,55,56,57,58]. Among other unique genes, the pilC gene is suspected to be essential for secretion and assembly of transcription factor P, important in pilus formation [59] while pilT helps in polymerization and depolymerization of pilin [60]. The brkA gene inhibits bactericidal activity and protects the bacterium from complement activation products [61]. Other unique genes are linked with bacterial adherence and colonizationm such as hopH, toxB, mpn 372 and stcE which contribute significantly to actin organization and bacterial attachment to human surfactant proteins [62,63,64,65]. The iraAB gene utilizes iron-loaded peptides that promote iron assimilation [66] while lepA plays a role in bacterial growth and induces inflammatory response. This gene also plays a key role in pathogenicity in Psudomonas aeruginosa [67]. The fcrA gene codes a protein containing receptor domains for immunoglobulins similar to those M-related proteins [68]. Another immunoglobulin-related gene, aga, plays a barrier function for mucosal antibodies by cleaving IgA1 [69]. IpsA controls transcriptional biogenesis of the cell wall in inositol-derived lipid formation in Corynebacterium and Mycobacterium species [70]. The vasL gene is considered to be component of vas genes, associated with the membrane type VI secretion system [71], and ravL is presumably activated at low oxygen level and regulates virulence gene expression via clp gene [72]. The lpg0365 codes a lypophosphoglycan that together with other membrane polypeptides, is necessary for Leishmania pathogenesis [73]. The pvdJ gene is involved in the production of cyclodipeptides that may regulate the production of biofilm [74]. In addition, pvdL is associated to biosynthesis or uptake of the siderphores pyoverdine and pyochelin that act in the transport of heme and ferrous ions [75], while pvdD is involved in the biosynthesis of pyoverdine in Pseudomonas aeruginosa [76]. IpaJ codes a plasmid antigen involved in demyristoylation of proteins by inducing golgi fragmentation and inhibiting hormone trafficking [77]. AliA is associated with nasopharyngeal colonization in Streptococcus pneumoniae [78]. The espN gene is reported in Mycobacterium tuberculosis to play a role in adding an acetyl group to the N-terminus of the esaT-6 virulence factor [79]. Flagella-related unique genes found in different studied strains include flgG, flgI, flgJ and flgk which play a major role in virulence, adhesion and motility. They are mostly involved in flagellum formation and also act as interface with other flagellar proteins [80,81,82,83]. The lnlK gene was reported in Listeria monocytogenes to help avoid autophagy while virB8 localizes to the inner membrane and is related to the export of alkaline phosphatase to the periplasm [84]. Finally, sigA codes a sigma factor linked with galactosidase activity [85].

Using principal component analysis of differentially distributed virulence genes, three distinct clusters were visualized (Fig. 8). A clear separation of virulence genes associated with brain or broncho-pulmonary abscesses (cps4E, sda and lap) from those associated with liver or abdominal abscesses (cpsB, fbp54 and cap8D) was observed. The first component which has maximum coverage and represents the largest variation showed that brain abscess-causing strains were associated with genes coding ATP-dependent proteolytic enzymes, which indicates their potential role in abscess formation. Other virulence genes clustered independently, excluding any association with the previous two disease categories. Among virulence genes associated to brain and broncho-pulmonary infections, sda codes an histidine kinase that regulates sporulation initiation in Bacillus subtilis and mediates the expression of virulence-associated factors [86]; lap codes the Listeria adhesion protein (LAP) that is a host stress response protein responsible for adhesion and promotion of translocation across monolayers [87]; and cps4E codes the capsular polysaccharide biosynthesis protein that was demonstrated in S. pneumoniae to prevent phagocytosis by forming an inert shield essential for encapsulation [88, 89].

Fig. 8
figure 8

Principal component analysis based upon gene presence/absence showing the distribution of virulence genes which may contribute to the particular type of abscess. The green color represents the various clinical forms while virulence genes are represented in red and studied strains are in blue. BPA is Broncho Pulmonary abscess and Abd abscess denotes abdominal abscess

Full size image

In S. pyogenes, fbp54 codes a fibronectin-binding protein that acts as an immunogen in humans. The amino acid sequence of fbp54 in S. intermedius is similar to that of S. pneumoniae. cap8D codes a dehydratase that is essential for the synthesis of the capsule precursor involved in adhesion. It has also been targeted as component for vaccine development [85, 90]; cpsB code capsular polysaccharide biosynthesis proteins that are essential for encapsulation in S. pneumoniae and are involved in the interaction of bacteria with their environment, notably their host organism [91];

In contrast to the above-mentioned genes, some were not found to be disease-specific. These included glf, cpsE, cpsI, cpsA, cps4C, cps8P and hasC. The glf gene is involved in the biosynthesis of unusual monosaccharide galactofuranose [92]; cpsE codes a glycosyl transferase responsible for the addition of activated sugars to the lipid carriers in the bacterial membrane and are essential for encapsulation in S. pneumoniae [93]; cpsI is essential for the production of high molecular weight capsular polysaccharides [94]; cpsA and cps8P are necessary for normal cell wall integrity and composition [95]; cps4C codes a polysaccharide tyrosine kinase adaptor protein that plays a key role in the regulation of capsule biosynthesis [96]; finally, hasC is involved in biosynthesis of hyaluronic acid capsule biosynthesis encodes glucose-1-phosphate uridylyltransferase [97].

Resistance-related genes and prophages

The tetracycline resistance gene tetM was identified in strains G1552, C270, KCOM1545, G1555, LC4, 30,309 and 32,811 whereas tet32 was identified in strain 631_SCON (Table 4). The macrolide resistance gene ermB was detected in strains G1552, C270, G1555 and 30,309. In other strains, no antibiotic resistance gene was identified.

Table 4 Antimicrobial resistance genes of studied S. intermedius strains
Full size table

A set of prophage elements was identified in all 27 strains (Table 1). In addition, four prophage-like elements were detected in strain BA1, three in strain TYG1620, and two in strains G1562, G1564, G1553, G1557, F0413 and G1555. The major difference in the genome size between all 27 studied strains of S. intermedius resided in the phage numbers and this presence of phages also denotes contribution of horizontal gene transfer in the emergence of this species [98].

CRISPR identification analysis

The search for CRISPR elements showed that 14 of the 27 studied genomes contained CRISPRs. Three of these 14 strains (G1564, G1565, 631_SCON) had more than one CRISPR, for a total of 17 CRISPR modules identified in studied strains. The direct repeat (DR) length in identified CRISPRs ranged from 24 to 36 bp while there was variation in the number of spacers present within each CRISPR. CRISPRs also differed among strains but the DR regions were similar for a given CRISPR element subtype. Based on the type of cas proteins, the CRISPRs of strains G1562, G1563, G1564, G1556, G1554, 631_SCON, 30,309 were subtype I-C CRISPRs; those of strains FDAARGOS_233 and KCOM1545 were subtype II-A CRISPRs; finally, the CRISPRs of strains G1565, G1552, B196, G1555 and 32811were subtype II-C CRISPRs [93] (Table 5).

Table 5 CRISPR elements found in studied S. intermedius strains
Full size table

Conclusions

In the present study, we reported 13 new clinical isolates of S. intermedius and, based upon a combined approach of pangenomics, core-genomics and virulence profiling of 27 strains, attempted identification of disease-specific genetic profiles. The comprehensive analysis revealed a genomic variability across strains within the species, although synteny of the core genome was preserved. Our results highlight the importance of surface proteins like pavB, pspA and cps4 (polysaccharide-coding proteins) and the binding proteins psaA, pavA, which are present in all studied strains, in pathogenesis. PCA results suggests two distinct categories of virulence genes, ATP dependent proteolytic virulence genes cps4E, sda and lap that are associated with brain and broncho pulmonary abscess while capsular polysaccharides protein coding genes cpsB and cps8D are linked with liver and abdominal abscess formation. The fibronectin binding protein coded by fbp54 is also showing its connection with liver and abdominal abscess formation. A recent study also attempted to determine the pangenome of S. intermedius.[99] The SNP-based phylogenetic tree as well as core gene-based tree showed no clustering related to any disease entity in S. intermedius strains. The whole study provides a key genetic framework for assessing and understanding the molecular events contributing to S. intermedius pathogenesis. However, due to the limited number of studied strains, validation of the role of these virulence factors will require experimental confirmations.