Pangenome analysis of

Abstract

Objective

Rummeliibacillus, a genus encompassing three known species, R. stabekisii, R. pycnus, and R. suwonensis, has a wide range of potential applications in biodegradation, probiotics, animal feed, and production of arginine, caproic acid, and other compounds. This study aims to explore the genetic diversity of this genus at the genomic level.

Methods

A comparative pangenome analysis of 12 strains isolated from different sources was conducted. In addition, the phylogenetic analysis, functional annotation, genomic metabolic pathway analysis, and prediction of mobile genetic elements were carried out.

Results

A total of 8 024 gene clusters were identified. The core genome, accessory genome, and strain-specific genes comprised 1 550, 3 941, and 2 533 gene clusters, respectively. In the core genome, the arginine cycle of six strains was complete. Seven strains had the ability to completely biosynthesize acetoin. However, only R. pycnus and R. suwonensis 3B-1 were able to completely biosynthesize caproic acid. The phylogenetic tree, DNA-DNA hybridization, and average nucleotide identity showed that Rummeliibacillus sp. G93 and Rummeliibacillus sp. TYF-LIM-RU47 were strains of R. stabekisii. Rummeliibacillus sp. POC4 and Rummeliibacillus sp. TYF005 may belong to a new species of this genus. In addition, genomic islands were identified in all the 12 strains, with the number ranging from four (R. stabekisii DSM 25578 and R. stabekisii NBRC 104870) to 14 (Rummeliibacillus sp. SL167 and Rummeliibacillus sp. TYF005), and prophage sequences were found in five of the 12 strains.

Conclusion

This study provides a genomic framework for Rummeliibacillus that could assist the further exploration of this genus.

Keywords

average nucleotide identity; Bacterial pan-genome analysis (BPGA); genomic islands; pangenome; Rummeliibacillus

Rummeliibacillus was first described in the United States in 2009^[1]. The genus includes three species, namely, R. stabekisii, R. pycnus^[1] and R. suwonensis^[2]. Although minimal research has been conducted on this genus, it has enormous potential for biotechnological applications^[3]. R. pycnus can be used to produce arginine with higher catalytic efficiency than other arginases reported^[4-5]. R. pycnus is able to convert palm oil plant wastewater into a terpolymer of polyhydroxyalkanoates and biodiesel^[6]. R. stabekisii has the potential for biomineralization^[7]. R. suwonensis can produce caproic acid using carbon sources such as sodium acetate^[8]. Microbial communities consisting of Rummeliibacillus sp., Caproiciproducens, and Clostridium_sensu_stricto_12 have been used to produce caproic acid from food waste by high-temperature fermentation^[9]. Rummeliibacillus sp. can produce acetoin using a variety of carbon sources, including pentose, hexose, and lignocelluloses^[10]. In addition, the genus can also be co-cultured with some Clostridium and Bacillus to produce hydrogen^[11]. Rummeliibacillus sp. also has probiotic properties and can be used for the production of food biological preservatives^[12] and animal feed additives^[13]. Overall, Rummeliibacillus is an important genus with many potential industrial applications.

Now, the availability of next-generation sequencing technologies and decreasing sequencing costs have allowed genome-wide approaches for microbial analysis^[14]. Genome sequencing and annotation and comparative genomics have accelerated the study of industrial microorganisms. Genomic information and bioinformatics software and databases can be used to compare multiple genomes of different species or genera^[15]. The pangenome represents the entire genome of a species or genus or given phylogenetic branch, describes genetic variation, and allows the definition of core genomes, which could help to understand species diversity and metabolic capacity^[16-17].

In this study, we systematically studied the specific taxonomic status of Rummeliibacillus sp. through average nucleotide identity (ANI), DNA-DNA hybridization (DDH), and phylogenetic tree analyses. Meanwhile, we analyzed the characteristics of the pangenome, core genes, and accessory genes of Rummeliibacillus sp. and evaluated the phylogenetic relationship of Rummeliibacillus sp. at the genome level. To study the potential functional characteristics of Rummeliibacillus sp., the metabolic capacity of the core genome and the accessory genome were analyzed. We evaluated genomic plasticity and genome evolution by the analysis of mobile genetic elements (MGEs)^[18].

1 Materials and Methods

1.1 Pangenome analysis

The 12 genomes of Rummeliibacillus sp. used for the pangenome analysis were downloaded from the FTP site of the Reference Sequence (RefSeq) database at NCBI (ftp://ftp.ncbi.nih.gov/genomes/, accessed on March 4, 2023). Once the download was complete, we assessed of the completeness and contamination of these 12 genomes using CheckM (https://github.com/Ecogenomics/CheckM; Table S1, the data has been submitted to the National Microbiology Data Center, with the registration number: NMDCX0001747). The detailed genomic information is shown in Table 1. Bacterial pan-genome analysis (BPGA), an ultra-fast software package that provides detailed comprehensive information about microorganismal genomes, was used. Usearch was chosen as the clustering tool with a 50% sequence identity cutoff^[19]. Gene clusters found in the genomes of all the analyzed strains were classified into the core genome. The accessory genome is composed of genes shared by 2-11 strains. The specific genome includes genes that exist only in one single strain of the species^[20]. The pangenome and core genome profiles were evaluated with PanGP^[21], using the gene presence-absence binary matrix (pan-matrix) obtained from BPGA as an input. The calculation of this matrix is based on similarity or dissimilarity among orthologous gene clusters^[22].

Table 1  Rummeliibacillus sp. isolates were used for pangenome analysis in this study

Strain	Country	Source	Size (Mb)	G+C content (%)	Proteins	Scaffolds	Accession number
DSM 15030	Germany	Soil	3.85	34.60	3 554	1	GCA_002884495.1
G20	Korea	Soil	4.11	35.90	3 638	16	GCA_007896435.1
3B-1	China	Pit mud	4.12	35.90	3 530	83	GCA_017578305.1
G93	China	Soil	3.24	37.70	3 121	1	GCA_023515935.1
POC4	Poland	Sewage sludge	3.69	34.50	3 570	163	GCA_003576525.1
SL167	Korea	Soil	4.17	35.80	3 757	14	GCA_007896505.1
TYF-LIM-RU47	China	Vinegar	3.30	37.40	3 273	14	GCA_008638315.1
TYF005	China	Vinegar	3.70	34.40	3 564	117	GCA_003844195.1
NBRC 104870	USA	Missing	3.26	37.30	3 150	23	GCA_007988965.1
PP9	Antarctica	Soil	3.42	37.69	3 334	2	GCA_001617605.1
DSM 25578	Missing	Missing	3.28	37.30	3 105	12	GCA_014202625.1
MER TA 13	USA	Missing	3.35	37.40	3 257	68	GCA_023713565.1

1.2 Phylogenetic analysis

For the phylogenomic analysis, whole-genome sequences of Rummeliibacillus sp. strains were compared by computing the ANI (two strains are considered to belong to the same species if the ANI is greater than 95%)^[23] and DDH (DDH>70% reflects the consistency of the genome type)^[24]. Origin (https://www.originlab.com) and Adobe Illustrator 2020 (https://www.adobe.com/cn/products/illustrator) were used to draw related graphs. To construct the phylogenetic tree based on the core genome, Molecular evolutionary genetics analysis version 11 (MEGA 11)^[25] software was used to align the concatenated amino acid sequences of the core genome, which were obtained from the BPGA pipeline analysis.

1.3 Functional annotation of the pangenome

Clusters of orthologous groups (COGs) of annotated proteins were generated using eggNOG-mapper to assign genes to COG categories^[26]. Kyoto encyclopedia of genes and genomes (KEGG) orthology (KO) annotation of the pangenome genes was carried out via the KEGG automatic annotation server (KAAS) pipeline^[27]. Metabolic pathways of Rummeliibacillus sp. were constructed via KEGG Mapper based on the assigned KO numbers^[28]. To predict the MGEs in the Rummeliibacillus sp. genomes, the genomic islands (GIs) and prophage sequences were predicted. GIs were predicted using IslandViewer 4, which involved three methods: SIGI-HMM, IslandPath-DIMOB, and IslandPick^[29]. Prophage sequences were annotated using PHASTER^[30], and all instant prophage sequences were considered intact, questionable, or incomplete. The sequences of the core genome, the accessory genome, and the specific genome were compared with the virulence factor database (VFDB) to predict virulence factors of 12 strains of Rummeliibacillus sp.

2 Results and Discussion

2.1 Pangenome and core genome of Rummeliibacillus sp.

A total of 12 Rummeliibacillus sp. genomes were used for the pangenome analysis. Genome sizes ranged from 3.24 to 4.17 Mb. The average number of protein-coding genes was 3 404, and the G+C content ranged from 34.40% to 37.70% (Table 1). All protein-coding genes in the 12 genomes of Rummeliibacillus sp. were grouped into 8 024 gene clusters. Among them, 1 550 gene clusters were found in the genomes of all 12 strains, which constituted the core genome of Rummeliibacillus sp. (Figure 1). These genes may represent the common metabolic and physiological characteristics of Rummeliibacillus sp. The accessory genome includes 3 941 gene clusters, made up of genes present in two or more genomes, but not in all the genomes studied. The number of strain-specific genes in each genome ranged from one to 600 (Figure 1). R. pycnus and Rummeliibacillus sp. SL167 had the largest number of strain-specific genes (600 and 419, respectively). R. stabekisii DSM 25578 and R. stabekisii NBRC 104870 had the lowest number of strain-specific genes (one and six, respectively). R. suwonensis 3B-1 and R. suwonensis G20 contained 183 and 201 specific genes, respectively.

Figure 1  Pangenome of Rummeliibacillus sp. core and accessory genome sizes and the number of specific genes in each Rummeliibacillus sp. strain. The same color for a specific gene represents the same species of bacteria.

Analysis of the existence of open and closed genomes can now be performed in many genera as the result of the burgeoning increase in microbial genome sequences from different strains within the same genus^[31]. First, cumulative curves were generated by PanGP. The mathematical formula for pangenome size fitting is a power-law regression based on Heaps’ law (y=Ax^B+C, where y denotes the number of genes of the pangenome, x denotes the analyzed genome number, and A, B, and C are fitting parameters). When 0<B<1, the number of genes of the pangenome increases when newly analyzed genomes are added, and the pangenome is considered as open. When B>1, the number of genes of the pangenome does not increase when newly analyzed genomes are added and the pangenome can be considered as closed. The mathematical formula for the number of genes of the core genome fitting is an exponential regression model (y=Ae^Bx+C, where y denotes the number of genes of the core genome, x denotes the number of analyzed genomes, and A, B, and C are fitting parameters)^[32]. The fitted curves for the pangenome profile analysis of 12 strains of Rummeliibacillus sp. showed that the fitted exponent of the curve was positive (0≤0.2<1), indicating that the Rummeliibacillus pangenome is open, which suggested that each added genome will contribute new genes and increase the number of gene clusters in the pangenome (Figure 2). Although this pangenome is obviously a mere mathematical extrapolation from the available sequenced strains, it makes clear the fact that some species exhibit extreme versatility in gene content.

Figure 2  Development of pangenome and core genome sizes for Rummeliibacillus sp. strains with genome number varying from 1 to 12. The cumulative curve (in orange) supports an open pangenome.

2.2 Phylogenetic analysis of Rummeliibacillus sp.

A comparison of the ANI between unknown genera and known species was performed using JSpeciesWS^[23] (https://jspecies.ribohost.com/jspeciesws/). JSpeciesWS calculated ANI between the genomes for a pairwise comparison using BLAST, and the results are shown in Figure 3A. The DDH values between the genomes were calculated using genome-to-genome distance calculator 3.0^[24] (https://ggdc.dsmz.de/ggdc.php). The ANI and DDH values of Rummeliibacillus sp. G93 and R. stabekisii NBRC 104870 were 98.75% and 90.2%, respectively. The ANI and DDH values of Rummeliibacillus sp. G93 and R. stabekisii sp. DSM 25578 also were 98.75% and 90.30%, respectively. Therefore, Rummeliibacillus sp. G93 belongs to R. stabekisii. Moreover, to analyze the phylogenetic relationship of the 12 Rummeliibacillus sp. strains, phylogenetic trees were constructed based on the concatenated core gene alignments. In the phylogenetic tree, 12 strains were grouped into two main clades (Figure 3B). The tree is divided into two large branches, with R. stabekisii clustered on one branch and R.suwonensis and R. pycnus clustered on the other branch. Meanwhile, in the phylogenetic tree, R. stabekisii NBRC 104870 and R. stabekisii DSM 25578 were in the same branch together with Rummeliibacillus sp. G93, indicating that Rummeliibacillus sp. G93 belongs to R. stabekisii. The ANI and DDH values of Rummeliibacillus sp. TYF-LIM-RU47 and R. stabekisii MER TA 13 were 98.14% and 87.70%, respectively. Meanwhile, Rummeliibacillus sp. TYF-LIM-RU47 and R. stabekisii MER TA 13 were in the same branch of the evolutionary tree, so it can be concluded that Rummeliibacillus sp. TYF-LIM-RU47 also belongs to R. stabekisii. In the other branch, Rummeliibacillus sp. POC4 and Rummeliibacillus sp. TYF005 had an ANI value of 98.41% and a DDH value of 86.50%, confirming that they are the same species. The ANI values of Rummeliibacillus sp. POC4 with R. pycnus, R. suwonensis 3B-1, R. suwonensis G20, and Rummeliibacillus sp. SL167 were 80.17%, 81.02%, 80.97%, and 81.09%, respectively. The DDH values of Rummeliibacillus sp. POC4 with R. pycnus, R. suwonensis 3B-1, and R. suwonensis G20 were only 25.50%, 22.40%, and 22.60%, respectively. This indicates that Rummeliibacillus sp. POC4 and Rummeliibacillus sp. TYF005 are neither R. pycnus nor R. suwonensis and that they may belong to a new species in this genus.

Figure 3  Comparison of the average nucleotide identity (ANI) values between the genomes of the 12 Rummeliibacillus sp. strains. A: Heatmaps display the ANI values between the 12 Rummeliibacillus sp. strains. The color represents the identity of strains, with purple indicating lower ANI values and red indicating higher ANI values. Color bars above and to the right of the heatmaps correspond to the source of each Rummeliibacillus sp. isolate. B: Phylogenetic trees of Rummeliibacillus sp. strains based on the core genome.

2.3 COG functional annotation of Rummeliibacillus sp. genes

COG analysis of pan-genomic gene clusters was conducted. Unknown function (S) was the largest category of the core genome, accessory genome, and strain-specific genes, accounting for 26.7%, 22.3%, and 29.8%, respectively (Figure 4). In terms of COG categories, most of the genes in the core genome are essential for life activities, such as transcription (K) (6.2%), translation, nucleosome structure, and biogenesis (J) (10.3%), amino acid transport and metabolism (E) (7.6%), energy production and conversion (C) (5.2%), replication, recombination, and repair (L) (6.5%), and cell wall/membrane/envelope biogenesis (M) (4.8%) (Figure 4). For the accessory genome, COG annotation showed that the largest categories were nucleotide transport and metabolism (F) (22.0%), transcription (K) (9.5%), and Inorganic ion transport and metabolism (P) (5.8%) (Figure 4).

Figure 4  Distribution of COG categories between the core genome, accessory genome, and strain-specific genes of Rummeliibacillus sp. strains. A: RNA processing and modification; B: Chromatin structure and dynamics; C: Energy production and conversion; D: Cell cycle control, cell division, chromosome partitioning; 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; 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 metabolite biosynthesis, transport, and catabolism; S: Function unknown; T: Signal transduction mechanisms; U: Intracellular trafficking, secretion, and vesicular transport; V: Defense mechanisms; W: Extracellular structures.

2.4 Genomic metabolic pathway analysis of Rummeliibacillus sp.

The KAAS annotation showed that all the 1 550 core gene clusters (19.3%) were assigned with KO numbers. The most annotated gene families in the core genome belong to carbohydrate metabolism. For substrate transport, ATP-binding cassette (ABC) transporters and phosphotransferase systems (PTSs) were the main transporting systems annotated by KAAS. The number of genes distributed in metabolic pathways is shown in Table S2 (The data has been submitted to the National Microbiology Data Center, with the registration number: NMDCX0001748). In the carbohydrate metabolism pathway, 138 genes are annotated in the core genome, 195 genes are annotated in the accessory genome, and 54 genes are annotated in the specific genome.

2.4.1 Caproic acid and acetoin metabolism

In this study, the pathways involved in the KEGG pathway caproic acid and acetoin metabolism were constructed in Rummeliibacillus sp. Some glycolysis genes can be found in the core genome, and the missing genes ptsG and pgi exist in the accessory genome (R. pycnus, Rummeliibacillus sp. POC4, Rummeliibacillus sp. SL167, Rummeliibacillus sp. TYF005, R. suwonensis 3B-1, and R. suwonensis G20). Figure 5B shows the metabolic process of caproic acid and acetoin of Rummeliibacillus sp. strains. R. pycnus, Rummeliibacillus sp. SL167, Rummeliibacillus sp. TYF005, R. suwonensis 3B-1, and R. suwonensis G20 could not synthesize acetoin due to the lack of the alaS gene, while the other seven strains had complete metabolic pathways. Based on existing reports, Rummeliibacillus sp. TYF-LIM-RU47 could indeed produce acetoin from various carbon sources, such as arabinose, xylose, glucose, xylan, and starch^[10]. Therefore, they can synthesize acetoin.

Figure 5  Overview of caproic acid, arginine, and sulfur metabolism in Rummeliibacillus sp. strains and distribution of accessory genome genes. A: Assimilatory sulfate reduction and main carbon metabolism annotated in the pangenome of Rummeliibacillus sp. strains; B: Genes annotated in caproic acid fermentation and metabolism of acetoin. Genes shared by all the 12 genomes (core genome) are shown in red, and genes in the accessory genome are shown in green; C: Distribution of genes involved in arginine metabolism; D: The distribution of genes in the accessory genome among different strains (blue represents the presence of the gene, while orange represents the absence of the gene).

In addition, the caproic acid biosynthesis pathway showed that most caproic acid synthesis genes were present in the core genome, but the bcd gene is found only in R. pycnus and R. suwonensis 3B-1 (Figure 5B). The other 10 strains did not have the ability to synthesize caproic acid. Currently, only Rummeliibacillus 3B-1 has been reported to synthesize caproic acid in this genus^[8]. Acid production by R. pycnus has not been reported, so further research is needed to determine whether R. pycnus has the ability to synthesize caproic acid.

2.4.2 Arginine metabolism

Amino acid metabolism is another important metabolic pathway, with 126, 229 and 58 genes annotated to the core genome, accessory genome, and specific genome, respectively, of this genus. Figure 5C shows the synthetic route of arginine. The genes in the whole arginine cycle pathway exist in the core genome. Arc is present in strains G93, TYF-LIM-RU47, DSM 25578, MER TA 13, NBRC 104870, and PP9, while aspB is present in strains SL167, G20, 3B-1, etc. (Figure 5D). One study reported that a new and heat-resistant arginase was found in R. pycnus SK31.001^[4]. This shows that Rummeliibacillus sp. may have the ability to synthesize arginine. It is worth noting that the metabolic characteristics of arginine in Rummeliibacillus sp. are similar to those reported in Escherichia coli^[33-34]. Glutamate is converted to ornithine, from which arginine is synthesized through the ornithine cycle.

2.4.3 Assimilatory sulfate reduction

In the accessory genome, the gene set for assimilatory sulfate reduction is complete in strains such as 3B-1, POC4, and G20 (Figure 5A, 5D). Sulfate can be absorbed and assimilated by bacteria and degraded to cysteine, and assimilative sulfur reduction by microorganisms provides a large amount of organic sulfur source for growth^[35]. The related genes (sat, cysC, cysH, etc.) mainly exist in R. pycnus and the strains POC4, SL167, TYF005, 3B-1, and G20, which indicates that the strains may have sulfate reducing effects. This suggests that R. stabekisii is capable of biomineralization^[7]. Therefore, it can be inferred that Rummeliibacillus may play an important role in soil maintenance.

2.5 Virulence factor analysis

The sequences of the core genome, the accessory genome, and the specific genome were compared with the VFDB database. In the Rummeliibacillus sp. pangenome, 38 virulence genes were identified in total. Of these, 13 core virulence genes were shared by all strains and four unique virulence factors were present in one strain each (Table S3, the data has been submitted to the National Microbiology Data Center, with the registration number: NMDCX0001749). Rummeliibacillus sp. SL167 (from soil) had the highest number of virulence genes (32), and R. pycnus (from soil) had the lowest number of virulence genes (19). The genomes of all 12 Rummeliibacillus sp. strains harbor genes encoding virulence factors, which are involved in processes including adhesion (flmH and slrA), secretion (clpB and cdsN), regulation (cheY and lisR), and motility (fliQ). Adhesion-related genes can promote adhesion and biofilm formation, which is an important factor in the pathogenesis of Streptococcus^[36]. The adhesion gene slrA encodes many surface proteins^[37]. These surface proteins have been identified as important virulence factors, involving the adhesion of bacteria to the epithelial cells of host cells, mediated by microbial surface components that recognize adhesion matrix molecules, thus contributing to host cell adhesion and tissue colonization^[38]. However, studies have shown that although SlrA is involved in colonization, it does not contribute significantly to invasive pneumococcal disease^[39]. Moreover, R. pycnus, Rummeliibacillus sp. POC4, R. suwonensis 3B-1, and R. suwonensis G20 carried three toxic genes (cylR2, cysC1, and hlyIII). Rummeliibacillus sp. TYF-LIM-RU47, R. stabekisii DSM 25578, R. stabekisii MER TA 13, and R. stabekisii NBRC 104870 contain only one toxic gene, hlyIII. The rest contain two toxic genes (cylR2 and hlyIII). They all have a virulence gene, hlyIII, which encodes an integral outer membrane protein with hemolytic activity that forms pores^[40-41]. However, the expression of enterococcal hemolysin requires the complete set of CylR2, CylA, CylB, and eight other proteins^[41]. Similarly, enterotoxin is not toxic when it is expressed alone^[42]. Therefore, it can be said that Rummeliibacillus are non-pathogenic bacteria.

2.6 Prediction of MGEs

To study the MGEs in Rummeliibacillus sp., we used IslandViewer 4 (an integrated interface for computational identification and visualization of GIs). MGEs can mediate the acquisition of DNA and promote the expansion of the gene pool of bacterial groups^[43]. The number of GIs in the genome of Rummeliibacillus sp. ranged from four (R. stabekisii DSM 25578 and R. stabekisii NBRC 104870) to 14 (Rummeliibacillus sp. SL167 and Rummeliibacillus sp. TYF005), indicating that MGEs are widespread in Rummeliibacillus sp. (Table S4, the data has been submitted to the National Microbiology Data Center, with the registration number: NMDCX0001750). R. suwonensis sp. G20 had the largest total length of GIs, occupying 8.22% of its genome size (4.11 Mb). Rummeliibacillus sp. G93 had the smallest total length of GIs, occupying 3.62% of its genome size (3.24 Mb).

In addition, the genomes of Rummeliibacillus sp. in this study were scanned using the PHASTER online service to obtain phage sequences^[29]. After searching the phages of 12 Rummeliibacillus sp. strains, a total of eight regions were intact, eight regions were questionable, and 26 regions were incomplete (Table 2). R. stabekisii PP9 was most complete, and three hypothetical phage groups were detected, including PHAGE_Paenib_Vegas (NC_028767) and PHAGE_Aeriba_AP45 (NC_048651) (Table 3). However, all regions in R. suwonensis 3B-1 and R. stabekisii DSM 25578 were incomplete. PHAGE_Aeriba_AP45 (NC_048651) was found in Rummeliibacillus sp. G93, Rummeliibacillus sp. TYF-LIM-RU47, and R. stabekisii PP9, suggesting that the phage played an important role in the evolution and diversity of these strains.

Table 2  Number of prophage regions (considered as intact, questionable or incomplete)

Strain	Intact	Questionable	Incomplete
Rummeliibacillus pycnus DSM 15030	0	1	0
Rummeliibacillus suwonensis G20	0	1	2
Rummeliibacillus suwonensis 3B-1	0	0	2
Rummeliibacillus sp. G93	1	0	0
Rummeliibacillus sp. POC4	0	2	4
Rummeliibacillus sp. SL167	1	0	2
Rummeliibacillus sp. TYF-LIM-RU47	2	1	4
Rummeliibacillus sp. TYF005	1	0	1
Rummeliibacillus stabekisii NBRC 104870	0	0	4
Rummeliibacillus stabekisii PP9	3	1	1
Rummeliibacillus stabekisii DSM 25578	0	0	4
Rummeliibacillus stabekisii MER TA 13	0	2	2

Table 3  Phage information for all Rummeliibacillus sp. strains

Strain	Region length (kb)	Total proteins	Most common phage (NCBI accession)	G+C content (%)
R. pycnus DSM 15030	37.4	23	PHAGE_Staphy_6ec (NC_024355)	31.75
R. suwonensis G20	50.3	43	PHAGE_Lister_LP_101 (NC_024387)	31.07
Rummeliibacillus sp. G93	43.9	78	PHAGE_Aeriba_AP45 (NC_048651)	36.52
Rummeliibacillus sp. POC4	35.1	58	PHAGE_Bacill_phBC6A52 (NC_004821)	33.71
Rummeliibacillus sp. POC4	18.6	25	PHAGE_Bacill_SPP1 (NC_004166)	36.28
Rummeliibacillus sp. SL167	55.5	37	PHAGE_Geobac_GBK2 (NC_023612)	42.43
Rummeliibacillus sp. TYF-LIM-RU47	24.7	44	PHAGE_Aeriba_AP45 (NC_048651)	36.96
Rummeliibacillus sp. TYF-LIM-RU47	34.3	52	PHAGE_Lister_B054 (NC_009813)	37.10
Rummeliibacillus sp. TYF005	49.0	72	PHAGE_Bacill_1 (NC_009737)	33.43
R. stabekisii PP9	54.2	78	PHAGE_Paenib_Vegas (NC_028767)	36.81
R. stabekisii PP9	47.0	63	PHAGE_Paenib_Vegas (NC_028767)	36.85
R. stabekisii PP9	44.9	75	PHAGE_Aeriba_AP45 (NC_048651)	37.01
R. stabekisii MER TA 13	17.4	20	PHAGE_Bacill_SPP1 (NC_004166)	38.77
R. stabekisii MER TA 13	31.9	23	PHAGE_Paenib_Vegas (NC_028767)	34.24

3 Conclusions

This work represents the first characterization of Rummeliibacillus species using pan-genomic analysis. The pangenome of Rummeliibacillus sp. strains is open, and the addition of newly sequenced genomes could increase the number of genes and the size of the pangenome. The pathway for arginine metabolism was discovered in all 12 Rummeliibacillus sp. strains, and only two strains had the ability to completely metabolize caproic acid, while seven strains had the ability to completely biosynthesize acetoin. Additionally, a complete assimilation sulfate reduction process was found in six strains. MGEs, including bacteriophages and GIs, were detected in the pangenome of Rummeliibacillus sp., which might give rise to horizontal gene transfer for environmental adaptation and increased genome diversity of Rummeliibacillus sp. This study provides insights for future research on the genetics and practical application of Rummeliibacillus sp.

Contributions Statement

ZOU Wei: Conceptualization, formal analysis, supervision, methodology, writing–review & editing. YANG Lingling: Formal analysis, visualization, writing–review & editing. LIU Chaojie: Methodology, formal analysis, visualization, writing–original draft preparation, data curation. ZHENG Jia: validation. ZHANG Kaizheng: Investigation. QIAO Zongwei: Investigation.

Acknowledgements

We thank International Science Editing (http://www.internationalscienceediting.com) for editing this manuscript.

Conflict of Interest

The authors declare that there is no conflict of interest.