Introduction

Floricoccus penangensis, a lactic acid bacterium, was first described by Chuah et al.1. It has been classified belonging to the family Streptococcaceae originally isolated from flowers of Durio zibethinus. The genus Floricoccus was explicitly separated from the genera Lactococcus and Streptococcus, with which it exhibits 94–99% and 49–76% sequence identity, as based on 16S rRNA and the housekeeping rpoB genes, respectively1,2. Floricoccus members have not only been found in various environmental locations, such as plants, flowers and foods1,2, but have also been associated with potential eukaryotic hosts, such as animals (as a component of the fecal/gut microbiome)3 and humans (respiratory secretions)4. Despite their frequent isolation, the precise ecological roles and functions of members of the genus Floricoccus are still largely unknown.

Various members of the genera Lactococcus and Streptococcus, which are phylogenetically closely related to Floricoccus, are extensively used for the production of fermented foods, such as yoghurt, cheese, kefir and lactic butter5,6. For example, Lactococcus raffinolactis is widely used as a starter culture in fish, meat, vegetables and milk fermentations7, while Streptococcus thermophilus strains are employed in the production of yogurt and certain cheeses and have also been reported to exhibit probiotic potential8. However, there have been reports on lactococcal and streptococcal isolates acting as pathogenic bacteria or opportunistic pathogens in humans and animals. For instance, Streptococcus suis can act as the causative agent of arthritis, meningitis and septicemia9, while Lactococcus lactis has been reported to very occasionally cause endocarditis and liver abscesses in humans. Moreover, Lactococcus garvieae is described as a fish pathogen, whereas Lactococcus petauri is associated with abscesses in animals so also exerts pathogenic potential10.

F. penangensis strain ML061-4 is a Gram-positive, coccus-shaped, non-endospore forming, non-motile and catalase-negative bacterium. This strain has been reported to exhibit properties such as acid-base tolerant, ability to aggregate and adhere to human epithelial cells, which may be desirable in a probiotic exploitation context2. In addition, F. penangensis ML061-4 may act as a biological control agent in Assam tea (Camellia sinensis var. assamica) cultivation and during tea leaf fermentation through its ability to produce organic acids as a result of fermentation of a variety of carbohydrates, such as glucose, fructose, mannose, N-acetyl-glucosamine, arbutin, aesculin, maltose, sucrose, trehalose, starch and glycogen1,11.

Although Floricoccus has been described as a novel genus, including species such as F. penangensis and F. tropicus, and being related to streptococci and lactococci, floricoccal genomes have not been extensively analyzed. The genome sequence of F. penangensis JCM 31735T was previously reported by Chuah et al.1. using an Illumina sequencing platform. Illumina sequencing allows low-cost and rapid genome analysis, however, this sequencing technology typically results in the generation of multiple contigs, thus not leading to a fully assembled, high quality genome sequence since it is based on the assembly of short reads12,13. A recent study by Rungsirivanich et al.14. described a completed genome sequence of F. penangensis ML061-4 using a combination of Illumina and Pacific Biosciences (PacBio) sequencing platforms. PacBio sequencing generates long reads and facilitates assembly of reads into fully circular chromosomes (and plasmids where relevant)15.

In silico analysis has been applied for gene predictions to better understand the genetic and metabolic potential as well as to predict functional characteristics of strains16,17. In the present study, we analyze and characterize the genome of F. penangensis ML061-4 and perform an extensive gene identification and annotation effort, combined with comparative genome analysis including genomes from floricoccal, lactococcal and streptococcal groups in order to validate the functionality and corroborate safety aspects of F. penangensis ML061-4.

Results and discussion

General genome features of F. penangensis ML061-4

In a previous study, the complete genome of F. penangensis ML061-4 was reported to constitute a single contig of 2,159,127 base pairs with a 33.2% GC content, and to encompass 2,134 predicted genes comprising 19 rRNAs, 63 tRNAs, 2 ncRNAs, 16 pseudogenes and 2,049 protein-encoding genes14. The genome size of F. penangensis ML061-4 is within the range of those determined for three other Floricoccus species (2.05–2.16 Mbp), or to those of Lactococcus (1.96–2.42 Mbp) and Streptococcus (2.15–2.47 Mbp) species (Table 1). Genomes belonging to members of the floricoccal group exhibit a lower % G + C content (approximately 33%) when compared to those of lactococci (35–38%) and streptococci (39–42%). To validate genetic relatedness among members of these three genera, the average nucleotide identity (ANI) value of strain ML061-4 was determined and compared to type strains representing members of the genera Floricoccus, Lactococcus and Streptococcus. Strain ML061-4 exhibits ANI values of 97.90, 93.70, 69.98 and 70.16% when compared with F. penangensis JCM 31735T, Floricoccus tropicus JCM 31733T, Lactococcus plantarum DSM 20686T and Streptococcus salivarius NCTC 8618T, respectively (Table 1). Congruently, the 16S rRNA gene sequence of strain ML061-4 displayed 100.0 and 99.8% identity to that of F. penangensis JCM 31735T and F. tropicus JCM 31733T, respectively, as previously described by Rungsirivanich et al.2.

Table 1 Genome features of F. penangensis ML061-4 compared with Floricoccus spp., Lactococcus spp. and Streptococcus spp.
Full size table

Genetic fingerprint and whole-genome phylogenetic analysis

DNA fingerprint (GTG)5 profiles showed that F. penangensis ML061-4 is closely related to the genera Streptococcus and Lactococcus. The DNA fingerprint of strain ML061-4 included a band with a size of approximately 1,000 bp which it has in common with the fingerprint obtained for S. thermophilus Rico66 and S. thermophilus Brie28. The same fingerprinting experiment of F. penangensis ML061-4 also generated bands with sizes of ~ 3,500 and ~ 4,000 bp, which corresponds to results obtained for L. lactis Tempeh6L and L. lactis RMN5F, respectively. The obtained results also support the phylogenetic closeness between members of the genera Streptococcus and Lactococcus as illustrated by the generation of very similarly sized products of 2,100, 2,400 and 2,500 bp. In contrast, and as expected based on its distant phylogenomic relationship, the genetic fingerprinting profiles obtained Bacillus species were clearly distinct from each other and from those obtained for floricoccal, lactococcal and streptococcal members (Fig. 1).

Fig. 1
figure 1

(GTG)5-PCR DNA fingerprints of Floricoccus penangensis ML061-4 compared with Streptococcus spp., Lactococcus spp. and Bacillus spp. Lane M : GeneRuler DNA Ladder Mix, Lane 1 : F. penangensis ML061-4, Lane 2 : S. thermophilus Rico66, Lane 3 : S. thermophilus Brie28, Lane 4 : L. lactis Tempeh6L, Lane 5 : L. lactis RMN5F, Lane 6 : B. velezensis ML122-2, Lane 7 : B. licheniformis ML075-1, Lane 8 : B. cereus TISTR 687, Lane 9 : B. subtilis NCTC 10073.

Full size image

A phylogenomic tree was constructed using complete genome sequences of Lactococcus spp. (8 genomes) and Streptococcus spp. (7 genomes) as well as that of F. penangensis ML061-4. The genome sequence of Weissella koreensis KACC 15510 was used as outgroup. The obtained phylogenomic tree showed that F. penangensis ML061-4 genome is clearly distinguished from genomes belonging to members of the genera Lactococcus and Streptococcus. Average nucleotide identity (ANI) between the genome of F. penangensis ML061-4 and those of selected species of Lactococcus and Streptococcus was determined. Strain ML061-4 was shown to elicit ANI values of 70.09, 69.43 and 69.90% when compared with L. cremoris FM-YL12, L. protaetiae KACC 19320 and S. pneumoniae PJ755/1, respectively. Furthermore, F. penangensis ML061-4 exhibits ANI value of 67.05% when compared with W. koreensis KACC 15510. The phylogenetic tree analysis of the whole genome is presented in Fig. 2.

Fig. 2
figure 2

Whole genome sequencing (WGS) phylogenetic tree of Floricoccus penangensis ML061-4 (red diamond) and the members of the genera Lactococcus and Streptococcus. The phylogenetic tree consisting of 17 strains was constructed using the Maximum Likelihood method. Bootstrap values were generated using 1,000 replications. The WGS of Weissella koreensis KACC 15510 was used as outgroup. Different symbols were used to highlight different species: a purple triangle corresponds to Lactococcus lactis, a green triangle indicates Lactococcus cremoris and a blue triangle reflects Streptococcus pneumoniae.

Full size image

Genome annotation of F. penangensis ML061-4

Functional assignment of the gene content of F. penangensis ML061-4 was possible for 1,195 genes (58.0%) involving 187 pathways as predicted by BlastKOALA; these annotations were shown to relate to genetic information processing (15.5%, 319 genes), environmental information processing (6.8%, 141 genes), signaling and cellular processes (6.6%, 136 genes), carbohydrate metabolism (6.5%, 134 genes), amino acid metabolism (4.0%, 82 genes), nucleotide metabolism (2.6%, 54 genes), metabolism of cofactors and vitamins (1.8%, 38 genes), lipid metabolism (1.6%, 33 genes), energy metabolism (1.4%, 28 genes), glycan biosynthesis and metabolism (1.3%, 26 genes), human diseases (0.1%, 3 genes) and unclassified (9.9%, 201 genes) (Fig. 3; Supplementary Table S1). These metabolic pathways may allow strain ML061-4 to survive under nutrient poor conditions18.

Fig. 3
figure 3

Gene functional categories of Floricoccus penangensis ML061-4 genome predicted by BlastKOALA.

Full size image

A total of 1,235 genes (59.9%) harboured by the F. penangensis ML061-4 genome were classified into six KEGG functional categories (Fig. 4; Supplementary Table S2). KEGG functional annotation identified genes associated with metabolism (34.4%), genetic information processing (8.2%), environmental information processing (8.2%), cellular processes (3.3%), organismal systems (1.9%) and human diseases (3.9%). Among these, 199 genes (9.7%) were predicted to be associated with carbohydrate metabolism. Biosynthesis of substrates and products is necessary for body structure, microbial survival, replication and evolution19. Although gene annotation of F. penangensis ML061-4 revealed a number of genes functionally associated with human diseases, this does not mean that the strain is a pathogen.

Metabolic pathways associated with carbohydrate utilization in F. penangensis ML061-4 are listed in Supplementary Table S3. The ML061-4 genome is predicted to encode a fructose-bisphosphate aldolase (locus tag number KIW23_01985), a key enzyme of the Embden-Meyerhof (glycolysis) pathway, and an essential enzyme to generate pyruvate from glucose prior to conversion into L-lactic acid via L-lactate dehydrogenase (locus tag number KIW23_05220). No D-lactate dehydrogenase-encoding gene, which would be expected to be responsible for D-lactic acid production, was identified in the ML061-4 genome, in agreement with Rungsirivanich et al.2. , who reported that this strain produces a high proportion of L-lactic acid (94.6%).

In vitro carbohydrate fermentation abilities by F. penangensis ML061-4 have previously been reported, showing that this strain is able to ferment various monosaccharides (e.g., glucose, fructose and mannose), disaccharides (e.g., lactose, maltose, sucrose and trehalose), oligosaccharide (raffinose), polysaccharide (starch) and a sugar alcohol (mannitol)1,2. These results are consistent with KEGG functional prediction of the F. penangensis ML061-4 genome. The genome of ML061-4 was also shown to encompass genes required for various phosphoenolpyruvate: sugar phosphotransferase systems (PEP-PTSs), which are known to be involved in the transport and metabolism of carbon sources such as fructose, galactose, glucose, lactose and sucrose20,21. The ptsP and ptsH genes of F. penangensis ML061-4 encode the general PTS proteins EI and Hpr, respectively, while genes encoding putative substrate-specific permeases (EII) were also identified, for example treP (positions 1,580,662–1,582,716) responsible for the production of EIIBCtrehalose, and manA (positions 1,613,369–1,614,310) encoding a putative EIIABCD[mannose22.

Fig. 4
figure 4

KEGG functional annotation of genes encoded by Floricoccus penangensis ML061-4 which is separated in six groups including metabolism (blue), genetic information processing (pink), environmental information processing (green), cellular processes (orange), organismal systems (turquoise) and human diseases (purple).

Full size image

Furthermore, our functional prediction efforts assigned 1,419 genes (68.8%) to a putative function by the COG database. We found that 16% of the genes identified on the ML061-4 genome was classified into translation, ribosomal structure, and biogenesis (J) category, 11.7% was associated with amino acid transport and metabolism (category E), and 8.5% was involved in carbohydrate transport and metabolism (category G) (Fig. 5; Supplementary Table S4). Comparison of COG functional classification of genes encoded by F. penangensis ML061-4, F. penangensis JCM 31735T, F. tropicus JCM 31733T, L. plantarum DSM 20686T, L. lactis subsp. lactis DSM 20481T, S. hillyeri DSM 107591T and S. ovuberis CCUG 69612T revealed the close phylogenetic relationship between the genera Floricoccus, Lactococcus and Streptococcus. Meanwhile, Bacillus subtilis Bbv57, an outgroup, presented higher putative function genes except category J which related to floricoccal, lactococcal and streptococcal members, and disappeared with general function prediction (R) category. Between 80 and 170 genes of the genera Floricoccus, Lactococcus and Streptococcus were assigned as function unknown (category S), a much lower number compared to for example Bacillus subtilis strain Bbv57 which had 1,086 genes assigned to this category (Fig. 6; Supplementary Table S5 – S10). This phenomenon is likely due to differences in genome size (flori-, lacto- and streptococcus members possess genomes of approximately 2 Mbp in size (Table 1), whereas bacilli elicit genome sizes that are larger than 4 Mbp)23,24.

Fig. 5
figure 5

COG functional annotation of genes encoded by Floricoccus penangensis ML061-4. Protein-coding sequences are classified according to COG functional categories. Alphabets A-Z exhibits COG functions. A: RNA processing and modification, B: Chromatin structure and dynamics, C: Energy production and conversion, D: Cell cycle control, cell division, chromosome, E: Amino acid transport and metabolism, F: Nucleotide transport and metabolism, I: Lipid transport and metabolism, J: Translation, ribosomal structure and biogenesis, K: Transcription, L: Replication, recombination, and repair, M: Cell wall/membrane/envelope biogenesis, N: Cell motility, O: Posttranslational modification, protein turnover, chaperones, P: Inorganic ion transport and metabolism, Q: Secondary metabolites biosynthesis, transport, and catabolism, R: General function prediction, S: Function Unknown, T: Signal transduction, U: Intracellular trafficking and secretion, Y: Nuclear structure, Z: Cytoskeleton.

Full size image
Fig. 6
figure 6

COG functional classification of genes encoded by Floricoccus penangensis ML061-4 compared to F. penangensis JCM 31735T, F. tropicus JCM 31733T, L. plantarum DSM 20686T, L. lactis subsp. lactis DSM 20481T, S. hillyeri DSM 107591T and S. ovuberis CCUG 69612T. COG functional classification of genes encoded by Bacillus subtilis Bbv57, as previous described by Thiruvengadam et al.24. , is used as an outgroup. Between 80 and 170 genes of the genera Floricoccus, Lactococcus and Streptococcus were assigned as function unknown (category S), whereas Bacillus subtilis had 1,086 genes assigned to category S.

Full size image

Virulence factors, toxin genes and bacterial pathogenicity in F. penangensis ML061-4

To identify possible virulence factors and toxins that may be encoded by the ML061-4 genome, a similarity search was performed employing the virulence factors of pathogenic bacteria database (VFDB). Gene clusters with nucleotide sequence similarity values below 80% and E-value above 1e− 10 were deemed non-significant25. Using these criteria no homologs corresponding to toxin production were found in the genome of ML061-4. The analysis revealed genes associated with virulence factors including clpC, cpsI, lap, clpP, htpB, glf, tufA, groEL, hasC and clpE with a nucleotide identity level of 79–90% (Table 2). These genes, many of which represent cellular housekeeping functions, are not necessarily linked directly to virulence such as clpC and htpB involved in ATP binding, while cpsI and glf encode UDP-galactopyranose mutase involved in cell wall biosynthesis26. Several genes were considered as conserved genes involved in stress responses (clpC, clpP, clpE, tufA, htpB and groEL). Furthermore, PathogenFinder software was applied in order to estimate the bacterial pathogenicity of F. penangensis ML061-4. Two genes were predicted to belong to pathogenic protein families including 30S ribosomal protein S19 (rpsS gene) and 30S ribosomal protein S21 (rpsU gene) with 100.0 and 93.1% identity (Supplementary Table S11). rpsS and rpsU genes have been described to be associated with rRNA binding, structural constituent of ribosome and tRNA binding (Uniport ID: P0A7U3 and P68679)27.

Table 2 Prediction of genes related to virulence factor in F. penangensis ML061-4 using VFDB database.
Full size table

Molecular mechanisms associated with adhesion

Four protein-encoding sequences in the ML061-4 genome were predicted to be involved in adhesion based on the VFDB database, including homologs of the Listeria adhesion protein (LAP; lap gene), 60 K heat shock protein (Hsp60; htpB gene), elongation factor Tu (tufA gene) and chaperonin (groEL gene) with 79–90% nucleotide identity (Table 2). It should be noted that these genes are widely considered to represent housekeeping genes and their physiological role in adhesion has been controversial28,29,30. Previous studies by Burkholder and Bhunia31 and Drolia et al.32. revealed the roles of LAP and Hsp60 which have been associated with bacterial adhesion ability of Escherichia coli and Listeria monocytogenes to Caco-2 cells33, while chaperonin GroEL and elongation factor Tu were described to participate in mucus adhesion16,34. Obtained data corresponding to in vitro experiments by Rungsirivanich et al.2. showed that F. penangensis ML061-4 possesses adhesion ability to an African green monkey kidney cell line (Vero cell) similar to Lactobacillus acidophilus and Lactiplantibacillus plantarum strains. The adhesion of microbes to host mucus and epithelial cells is the first step of human colonization and infection35,36. Bacterial adhesion ability is considered to be an essential trait to colonize and possible invasion of a host in case of pathogenic bacteria. Conversely, candidate probiotic strains with adhesion features may not only act as competitive exclusion to pathogens, yet may also positively modulate the immune response37.

Bacteriophages, antibiotic and antiviral defenses

To predict prophage elements carried by the F. penangensis ML061-4 genome, in silico analysis using PHASTEST software was performed. One region with homology to Streptococcus phage PH15 were identified in the genome of F. penangensis ML061-4. This 16.1 Kbp DNA region with a 32.85% GC content was predicted to represent an incomplete prophage (Table 3; Supplementary Fig. S1 and S2). Of note, a prophage-encompassing region homologous to the complete Enterococcus phage phiFL1A was predicted to be present in the F. penangensis JCM 31735T genome, while two prophage regions related to Lactococcus phage 50101 and Bacillus phage BCJA1c were predicted to be harbored by the genome of F. tropicus JCM 31733T. Furthermore, various predicted (partial) prophage-encompassing regions were predicted in the L. lactis DSM 20481T, S. salivarius NCTC 8618T and S. hillyeri DSM 107591T genomes, while no prophages or remnants thereof appeared to be present in the genomes of L. plantarum DSM 20686T and S. ovuberis CCUG 69612T (Table 3). None of these predicted prophage regions appeared to be intact; however, the relatedness of these regions to phages of diverse bacterial species provides insights (especially regarding phage origins and evolution) into the co-existence and interactions with other organisms in its environmental niche. Phages can infect bacteria and integrate their genomes into the bacterial genome in order to allow phage multiplication. Such integrated phage genomes or prophages may in turn be activated, resulting in phage particle production and host death38. Prophage regions are commonly present in bacterial genomes and have been reported to have many beneficial distinguishing qualities for bacteria such as bacterial adaption in a new environment as well as promoting bacterial resistance26,39.

Table 3 Gene cluster identification of prophages in F. penangensis ML061-4 genome compared with Floricoccus spp., Lactococcus spp. and Streptococcus spp. using PHASTEST.
Full size table

To identify genes involved in antibiotic resistance, the complete genome of ML061-4 was determined using CARD database based on RGI criteria with perfect and strict hits. The results showed that the ML061-4 genome did not contain any obvious antibiotic resistance genes. Nevertheless, the KEGG ontology of ML061-4 revealed genes involved in drug resistance pathways including beta-lactam, vancomycin and cationic antimicrobial peptide. A previous study by Rungsirivanich et al.2. has shown that F. penangensis ML061-4 is susceptible to various tested antibiotics such as ampicillin, cefotaxime, gentamicin, meropenem and tetracycline. The absence of antibiotic resistance genes is suggested to be a desirable property as it eliminates the risk of horizontal transfer to other microorganisms leading to antibiotic drug resistance40,41. The genome of F. penangensis ML061-4 was shown to lack any CRISPR-Cas systems. CRISPR-Cas systems are found in approximately 40% of assessed bacteria and are known to provide immunity against exogenous nucleic acids, especially bacteriophages42. Antiphage systems of strain ML061-4 were predicted by the PADLOC analysis. The results revealed that the genome of F. penangensis ML061-4 contains 5 antiphage systems (DMS_other, RM_type_IV, RM_type_IIG, retron_III-A and PD-T4-6) and 2 phage defense candidate systems (PDC-S30 and PDC-M01). Antiphage systems are multigene (located in operons) or single genes. DNA modification systems (DMS) and restriction-modification systems (RM), retrons have been found to play an important role in bacterial antiviral defense43,44. The predicted antiviral defence systems of F. penangensis ML061-4 are listed in Table 4, Supplementary Fig. S3 and Supplementary Table S12.

Table 4 Antiviral defence systems detected in F. penangensis ML061-4 genome using PADLOC analysis.
Full size table

Secondary metabolite biosynthesis and xenobiotics biodegradation

Secondary metabolic potential of strain ML061-4 was examined using antiSMASH, yet did not produce any credible candidates. The antiSMASH software is widely used to predict microbial secondary metabolites for biosynthetic gene clusters particularly antimicrobial, non-ribosomally synthesized peptides (NRPs), polyketides (PKs) and bioactive compounds45. Conversely, KEGG ontology of strain ML061-4 revealed 48 genes involved in 20 metabolic pathways for terpenoids and polyketides, and secondary metabolite biosynthesis pathways (e.g., antibiotics, insect hormone and plant secondary metabolites) and 31 genes associated with 11 pathways of xenobiotic biodegradation (e.g., organic compounds and drugs) (Supplementary Table S13). Xenobiotics are chemical substances that are not naturally produced or found in organisms or the environment46. In general, xenobiotic refers to environmental pollutants resulting in industry and agriculture such as pesticides, polycyclic aromatic hydrocarbons, pharmaceutical active compounds and phenolic compounds47. Xenobiotic biodegradation ability may assist a microorganism to survive in particular polluted environments48. KEGG annotation revealed gene clusters involved in xenobiotic biodegradation pathway in ML061-4 genome comprising benzoate, aminobenzoate, chloroalkane, chloroalkene, chlorocyclohexane, chlorobenzene, xylene, styrene, dioxin and naphthalene (Supplementary Table 13). Therefore, secondary metabolites and substances related to xenobiotic biodegradation produced by F. penangensis ML061-4 should be further elucidated.

Materials and methods

Bacterial strain and growth conditions

F. penangensis ML061-4 (GenBank accession no. MH050697) was isolated from the leaf surface of an Assam tea plant (Camellia sinensis var. assamica) in the Sakat sub-district, Pua district, Nan province, Thailand at an altitude of 1,038 m above sea level (19°15’53.62"N, 101°0’30.22"E)2. The F. penangensis ML061-4 was cultivated in M17 (Oxoid™, Basingstoke, England) supplemented with 0.5% (w/v) glucose (GM17) broth and incubated at 30 °C for 24 h, as previously reported by Rungsirivanich et al.23.

Genome sequencing and assembly

Whole genome sequencing and assembly of strain Ml061-4 was accomplished as previously described by Rungsirivanich et al.14. The genome sequence was deposited in GenBank under accession number CP075561 and Sequence Read Archive (SRA) number SRP399917.

PCR fingerprinting using the (GTG)5 primer

(GTG)5-PCR DNA fingerprint analysis of F. penangensis ML061-4 was performed according to the protocol of Parlindungan et al.49. Streptococcus thermophilus Rico66, S. thermophilus Brie28, Lactococcus lactis Tempeh6L and L. lactis RMN5F as previously described by Parlindungan et al.50 , Bacillus velezensis ML122-2, B. licheniformis ML075-1, B. cereus TISTR 687 and B. subtilis NCTC 10073 as previously described by Rungsirivanich et al.23. were used for comparison purposes. PCR amplifications were performed using Taq DNA polymerase mastermix (Qiagen, Manchester, UK) with the single oligonucleotide primer (GTG)5, 5′-GTGGTGGTGGTGGTG-3′. The PCR reaction mixtures were initially denatured at 95 °C for 7 min, followed by 30 cycles of denaturation at 90 °C for 30 s, annealing at 40 °C for 1 min, extension at 65 °C for 8 min and a final extension step at 65 °C for 16 min using an Applied Biosystems™ 2720 Thermal Cycler (Thermo Fisher, CA, USA). Amplicons were subjected to electrophoresis on a 1% (w/v) agarose gel consisting of ethidium bromide for 1 h at 110 V 100 mA in 1X Tris-acetate-EDTA (TAE) buffer (40 mM Tris-acetate, 1 mM EDTA, pH 8.0). GeneRuler DNA Ladder Mix (Thermo Fisher, CA, USA) was used as a marker. The GTG PCR profiles were visualized under ultraviolet light by gel documentation system, and gel images were captured using GeneSnap software (Syngene, MD, USA).

Phylogenomic tree analysis

A phylogenomic tree based on whole genome sequences of strain ML061-4 and various members of the genera Lactococcus and Streptococcus was constructed according to the method of L’Huillier et al.51. A total of 16 complete genome sequences were retrieved from the National Centre for Biotechnology Information (NCBI) including Lactococcus lactis K_LL005 (GenBank accession no. NZ_CP060580), Lactococcus lactis AI06 (GenBank accession no. NZ_CP009472), Lactococcus lactis DSM 20481T (GenBank accession no. NC_020450), Lactococcus cremoris FM-YL12 (GenBank accession no. CP071728), Lactococcus cremoris UC073 (GenBank accession no. CP068698), Lactococcus protaetiae KACC 19320T (GenBank accession no. CP041356), Lactococcus allomyrinae 1JSPR-7 (GenBank accession no. CP032627), Lactococcus paracarnosus TMW 2.1615 (GenBank accession no. CP017195), Streptococcus parasuis 1628469 (GenBank accession no. NZ_CP134491), Streptococcus salivarius NCTC 8618T (GenBank accession no. NZ_CP009913), Streptococcus mutans UA159 (GenBank accession no. AE014133), Streptococcus sanguinis CGMH058 (GenBank accession no. CP040798), Streptococcus toyakuensis TP1632 (GenBank accession no. NZ_AP024523), Streptococcus pneumoniae PJ755/1 (GenBank accession no. CP155535), Streptococcus pneumoniae SCAID PHRX1-2021 (GenBank accession no. NZ_CP082820) and Weissella koreensis KACC 15510 (GenBank accession no. CP002899). The kSNP 4.0 was applied to perform alignment-free whole genome phylogenetic analysis by identifying single nucleotide polymorphisms (SNPs) across the genome sequences52. The obtained SNP-matrix results were imported into MEGA 11 for phylogenetic/phylogenomic tree construction53. The phylogenomic tree was analyzed by a Maximum Likelihood method54 with bootstrap values of 1000 replicates55.

Genome annotation

Automatic annotation of predicted open reading frames (ORFs) was performed using a combination of Prodigal (version 2.6.3) and BLASTp (version 2.2.26) sequence alignments56 to assign annotation, using an E-value cut-off of 0.0001 for hits exhibiting at least 50% similarity across at least 50% of the sequence length) against a non-redundant protein database provided by the NCBI portal. Functional prediction of genes and proteins was integrated using the Clusters of Orthologous Groups (COGs)57 and protein family (Pfam)58, respectively, as previously described by Martín et al.59. Ribosomal RNA (rRNA) and transfer RNA (tRNA) genes were detected using RNAmmer version 1.260 and tRNA-scanSE version 2.061, respectively.

The Kyoto Encyclopedia of Genes and Genomes (KEGG) orthology annotation was undertaken62 by the KAAS online server version 2.163 using the bi-directional best hit (BBH) method (https://www.genome.jp/kegg/kaas/). Genome functional characterization was further assigned by BlastKEGG orthology and links annotation (BlastKOALA; https://www.kegg.jp/blastkoala/)64. The functional analysis of singleton genes was classified into categories of the COG using the Batch Web CD-Search Tool (https://www.ncbi.nlm.nih.gov/Structure/bwrpsb/bwrpsb.cgi)65. The virulence factor search in the ML061-4 genome was conducted using the virulence factors of pathogenic bacteria database (VFDB) (http://www.mgc.ac.cn/VFs/)66, and bacterial pathogenicity elements were predicted using PathogenFinder version 1.1 (https://cge.cbs.dtu.dk/services/PathogenFinder/)67. Moreover, the presence of lysogenic bacteriophages within the genome was predicted using PHASTEST (https://phastest.ca/)68, and putative virulence factors and antibiotic resistance genes were identified by CARD database (https://card.mcmaster.ca/)69. Genes involved in secondary metabolite biosynthesis were predicted using antiSMASH (https://antismash-db.secondarymetabolites.org/)45. The presence of clustered regularly interspaced short palindromic repeats (CRISPR) systems was investigated using CRISPRfinder (https://crispr.i2bc.paris-saclay.fr/Server/)70. Antiviral defence systems were identified using PADLOC analysis (https://padloc.otago.ac.nz)71.

Genbank accession numbers and comparative genome analysis

Genome information of Floricoccus penangensis JCM 31735T (GenBank accession no. NZ_MKIQ00000000), Floricoccus tropicus JCM 31733T (GenBank accession no. NZ_MKIR00000000), Lactococcus plantarum DSM 20686T (GenBank accession no. NZ_JXJX00000000), Lactococcus lactis DSM 20481T (GenBank accession no. NC_020450), Lactococcus taiwanensis NBRC 109049T (GenBank accession no. NZ_BNDT00000000), Streptococcus salivarius NCTC 8618T (GenBank accession no. NZ_CP009913), Streptococcus hillyeri DSM 107591T (GenBank accession no. NZ_RCVM00000000) and Streptococcus ovuberis CCUG 69612T (GenBank accession no. NZ_JAAXPR000000000) was retrieved from the NCBI database (https://www.ncbi.nlm.nih.gov/) for comparative genomic analysis. The genome identity assessments were performed by average nucleotide identity (ANI) determination using EzBioCloud databases (https://www.ezbiocloud.net/tools/ani)72.

Conclusions

This study reports on a functional genomic analysis of a member of the genus Floricoccus, and in particular compares its genome to those belonging to closely related members of the genera Lactococcus and Streptococcus. Comparative genomic analysis showed that the G + C content of the genus Floricoccus is explicitly distinguished from Lactococcus and Streptococcus genera. In silico analysis exhibited that F. penangensis ML061-4 encodes enzymes for the metabolism of various carbohydrates. Genome sequence analysis exposed a high level of genetic variation among members of the Floricoccus, Lactococcus and Streptococcus genera. Strain ML061-4 contains genes associated with both probiotic and pathogenic properties, and future research efforts will assist in elucidating the biological functionality, safety and potential biotechnological applications of strains of this species.