Introduction

Biosurfactants are amphiphilic molecule produced by microorganism that has the ability to lower the surface and interfacial tension between two liquids. They are environmental friendly due to their low toxicity level, extreme biodegradability, stability to high pH and temperature condition which makes them as promising candidate for bioremediation and enhanced oil recovery applications (Mulligan 2009). They are also used in food industry (Nitschke and Costa 2007), pharmaceutical and cosmetics (Banat et al. 2000). Biosurfactants are alternative to the synthetic surfactants; however, the high cost of production limits their application.

There are many bacteria have been reported to produce biosurfactants such as P. aeruginosa N002 (Das et al. 2015), Achromobacter spanius (Alvarez et al. 2015), and Rhodococcus erythroplis (Cai et al. 2015a). Besides that, many strains from genus Bacillus are biosurfactant producers such as B. licheniformis, B. subtilis, and B. pumilus (Płaza et al. 2015a). Bacillus species are genetically diverse and also grow in various environments (Choudhary and Johri 2009). Bacillus species are one of the main sources for producing industrial enzymes such as amylases and protease (Karataş et al. 2013; Kunst et al. 1997).

In the past decade, microbial research was limited to the need to grow bacteria in culture. The revolution of DNA sequencing technology has changed the way scientists think about genetics and genomics information (Lasken and McLean 2014; Zhang et al. 2011). The first complete genome of Bacillus subtilis was published on 20th November 1997 (Kunst et al. 1997). DNA sequencing applications allowed many researchers and scientists to sequence the whole genome or a region of a microbial genome (Kamada et al. 2014; Nishito et al. 2010). To date, the whole genome sequencing application is widely used to analyse microbial biosurfactant producers (Das et al. 2015; Shaligram et al. 2016). Here we report the whole genome sequencing and analysis of the new strain UMX-103, isolated from Terengganu, Malaysia, with the potential for biosurfactant production.

Materials and methods

Sample isolation and preparation

The strain UMX-103 was isolated from a hydrocarbon contaminated site in Terengganu, Malaysia. The bacterium was cultured in Tryptone Soya Agar (TSA) (Merck KGaA, Germany) and incubated overnight at 30 °C. Optimal colony was selected from the cultured bacteria and inoculated in 50 ml of Tryptone Soya Broth (TSB) (Merck KGaA, Germany) using 200 ml conical flask. The broth was incubated overnight at 30 °C in an orbital shaker at 121 rpm.

Determination of bacteria morphology

Gram staining was used to stain the bacteria using GCC Diagnostics (Gainland Chemical Co, UK). Staining was performed according to the manufacture’s protocol. Field Emission Electron Microscope (FESEM)(Quanta 450 FEG, USA) was used to visualize the morphology of the bacteria.

Screening of UMX-103 for capability of producing biosurfactants

The ability of biosurfactant production by UMX-103 was tested using five different methods; (i) hemolytic assay, (ii) oil spreading test, (iii) drop-collapse assay, (iv) emulsification assay, and (v) surface tension measurements. Hemolysis assay on blood agar plates has been widely used as a method to screen surfactant producing bacteria (Banat 1993; Morán et al. 2002; Mulligan et al. 1984; Yonebayashi et al. 2000). In this study, the isolated strain UMX-103 was streaked onto blood agar plates and incubated at 30 °C for 24 h. The plate was visually inspected for clear zone formation around the colonies, which is an indicative of biosurfactant production.

The oil spreading test is to observe a clear zone formation which is a result of dropping a biosurfactant or surfactant solution on an oil–water interface. Approximately, 15 µl of engine oil (10W-40 Shell®) was added to 40 ml of distilled water in a petri dish (150 mm in diameter) to form a thin layer of oil on the surface. Then, 15 µl of culture supernatant was added to the central of the oil layer (Morikawa et al. 1993, 2000; Youssef et al. 2004). Diameter of the clear zone formation on the oil layer was observed and measured after 30 s (Morikawa et al. 2000). We have used (Triton® X-100, USA) as the positive control, while distilled water as negative control (Shoeb et al. 2015).

Drop-collapse qualitative test was conducted on a polystyrene 96-microwell (12.7 × 8.5) plate. Approximately, 2 µl of engine oil (10-40 Shell®) was dropped into the well and equilibrated for 1 h at the room temperature. Then, 5 µl of the culture supernatant was added on the top of oil surface. Water was used as a negative control (Bodour et al. 2003; Shoeb et al. 2015). The shape of the drop on the oil surface was observed after 1 min. In the presence of biosurfactant, the drop will be collapsed in the oil surface.

Emulsification test was performed to check the ability of biosurfactant to emulsify the oil surface. Initially, 5 ml of 50mM Tris buffer (8.0 pH) was added in 30-ml screw-caped test tube. Then, 5 ml of engine oil (10-40 Shell®) was added to the Tris buffer and vortex at room temprature for 2 min. The screw-caped test tube with the mix solution was stabilized for 24 h. The absorbance of aqueous phase was measured by spectrophotometer (Spectroquant® Pharo100, USA) at the wavelength of 400 nm. Distilled water was used as a negative control, while Triton-X as the positive control (Shoeb et al. 2015). The emulsification activity was calculated (Cai et al. 2015b) as stated below:

$${\text{EAbs}}={\text{Sample Emulsification Abs}}/{\text{Optimum Emulsification Abs}} \times {\text{1}}00\% .{\text{ EAbs}}=\left[ {{\text{Emulsification Absorbance}}} \right]$$

For the surface tension measurement the bacteria culture was centrifuged at 3000 rpm for 25 min to obtain a cell-free supernatant. The surface tension of the culture supernatant was determined by the Du Nouy ring method (Gudina et al. 2010; Pereira et al. 2013) using interfacial tensiometer (Force Tensiometer, Sigma700, Biolin Scientific) at room temperature. The measurements of the surface tension were repeated three times and an average value was obtained (Cai et al. 2015b; Pereira et al. 2013; Vaz et al. 2012).

DNA preparation and whole genome sequencing

The whole genome sequence of Bacillus subtilis UMX-103 was obtained from Illumina HiSeq 2000 sequencing technology (Illumina, USA). The DNA was extracted using phenol–chloroform method and the quality and quantity of the DNA was measured using QIAxpert (QIAGEN, Germany). The sample was run on 1.2% agarose gel to determine the integrity of genomic DNA. Fragmentation of the DNA was performed using Covaris S220 (Covaris Inc, USA). Ligation to NEBNext adapters was conducted using NEBNext Ultra, while the PCR-enrichment using the DNA Library Prep Kit (NEB, USA). The final library was quantified using KAPA kit (KAPA Biosystem, USA). Library size was confirmed using Agilent Bioanalyzer High Sensitive DNA Chip (Agilent, USA). The prepared library was sequenced using an Illumina flow cell, consisting of 2 × 100 cycles.

De novo assembly by Velvet and mapping the reads to reference genome by BWA

Quality control assessment was performed using Trimmomatic 0.35 (Bolger et al. 2014). The generated dataset was assembled using Velvet 1.2.10 (Zerbino and Birney 2008) which is a de novo assembly software that use de Bruijn graph algorithm. SSPACE-Standard v3.0 (Boetzer et al. 2011) was used for scaffolding the generated contigs from Velvet assembler. GapFiller v1.10 (Boetzer and Pirovano 2012) was used to close the gaps and replaced the unknown nucleotide with known nucleotides. The scaffolds were sorted along with the reference genome (Bacillus subtilis strain 168; accession number NC_000964.3) using Mauve 2.3.1 (Darling et al. 2004). BWA (Li 2013) was used for mapping the reads to the reference genomes.

Gene prediction and functional annotation

Gene prediction for protein-coding genes was conducted using Prodigal (Hyatt et al. 2010). The tRNA and rRNA screenings were performed using tRNAscan-SE v1.3.1 (Lowe and Eddy 1997) and RNAmmer v1.2 (Lagesen et al. 2007), respectively. Gene annotation was conducted using Prokka v1.11 (Seemann 2014). The functional annotation was performed using EggNOG-mapper 4.5.1 database (Huerta-Cepas et al. 2016). The annotated genes were submitted to IslandViewer3 (Dhillon et al. 2015) to identify the genomic islands in the genome. Pan genome analysis and comparison between all the prospective reference genomes were conducted using Roary version 3.6.1 (Larsen et al. 2012).

Phylogenetic and genomic similarity analysis

16S ribosomal DNA was used to identify the bacteria. The 16S rRNA was obtained from the whole genome sequence and aligned with other 16S rRNA genes from different Bacillus strains. The 16S genes were extracted from each reference strains using RNAmmer (Lagesen et al. 2007). Molecular Evolutionary Genetics Analysis (MEGA) version 7 (Kumar et al. 2016) was used to align and construct the distanced phylogenetic tree. The phylogenetic tree and distance were constructed using Neigbour-joing method. The Average Nucleotide Identity of UMX-103 was determined using GGDC 2.1 server (Meier-Kolthoff et al. 2013). Multilocus Sequence Typing (MLST) was predicted using MLST 1.8 server (Larsen et al. 2012).

Results and discussion

Morphology of UMX-103

The Gram staining showed that UMX-103 is a Gram positive strain. FESEM result shows the morphology of the colony which is a rod shape, with size of 1.954 µm length and 540.9 nm width (Fig. 1). Based on the Gram staining and FESEM results we confirmed that UMX-103 belongs to the genus Bacillus.

Fig. 1
figure 1

FESEM of Bacillus subtilis UMX-103

Full size image

Biosurfactant activity

The biosurfactant producing capability of UMX-103 was tested in five different assays. The results are shown in Table 1. Bernheimer and Avigad (1970) reported that surfactin produced by Bacillus subtilis lyse the red blood cells. There is an association between hemolysis activity and surfactant production, since then hemolytic assay is recommended as a primary technique to screen biosurfactant producers (Youssef et al. 2004). Therefore, this assay was employed in this research. UMX-103 demonstrated beta lysis as it produced a clear zone around the colony which determines the biosurfactant production by the strain. Biosurfactants are well known to have hemolytic, antibacterial, and antiviral activity, owning a precise mechanism that has impact on the membrane permeability and eventually leading to cell disruption (Heerklotz and Seelig 2007).

Table 1 The five different biosurfactant producing capability tests conducted on UMX103; hemolysis assay, oil-spreading, drop-collapse, emulsification assay and surface tension measurement
Full size table

Oil spreading assay is based on the formation of a clear zone and a displacement area, in the presence of biosurfactant in the culture supernatant. The diameter of this clear zone on the oil surface correlates to the amount of biosurfactant produced. In this study supernatant of UMX-103 culture formed a clear zone and oil displacement region about 2 cm for as indication of biosurfactant production.

The drop-collapse test depends on the destabilization of liquid droplets by the biosurfactants produced by the bacterial isolate. The stability of drops is dependent on biosurfactant concentration and correlates with the surface and interfacial tension (Sari et al. 2014). In this study, the distilled water was used as a negative control and there is no droplet collapsing was observed. Whereas, the biosurfactant produced by UMX-103 was tested positive, where the droplet was collapsed.

The emulsification test was used to evaluate the emulsification ability of the isolate UMX-103. We have observed a positive activity of the strain where it emulsifies the oil surface. In this study, Triton-X was used as positive control due to it emulsification ability and it have been widely used as positive control (Shoeb et al. 2015).

The measurement of surface tension using Du nouy ring method is based on measurement of the force required to detach the ring from the culture supernatant surface. The detachment force is directly proportional to the interfacial tension. Our test results showed that UMX-103 has a higher reduction ability which is 26.4 ± 0.02 mN/m. While, Triton X showed reduction value of 34.3 ± 0.003 mN/m and the hexane with the lowest reduction value of 18.1 ± 0.06 mN/m.. Summary of the biosurfactant assays is presented in Table 1.

Mapping reads to the reference genome

A total of 565,068,437 paired-end reads with a length of 101 bp were generated, with average insertion size of 534 bp. Low quality bases and reads were filtered to get an optimal quality score of 30 or higher at each base. The reads were mapped to Bacillus subtilis strain 168, where results showed that 93.44% of the generated reads are mapped to the reference genome.

De novo assembly and scaffold sorting

Velvet assembler generates a total of 69 contigs with an average length of 61,362 bp and with maximum and minimum length of 869,096 and 137 bp, respectively (Table 2). All contigs generated by Velvet were used to generate the scaffolds using SSPACE-Standard software (Boetzer et al. 2011). Scaffolding process generated a total of 39 scaffolds, with average scaffold size of 108,565 bp and with maximum and minimum scaffold sizes of 1,059,836 and 144 bp, respectively (Table 2). Then, GapFiller software (Boetzer and Pirovano 2012) was used to close the gaps in the generated scaffolds. Total of 34 gaps from 41 gaps were closed. In addition, the result after gap closing shows a total of 39 scaffold with an average size of 108,580 bp (Table 2). The scaffolds were sorted according to the reference genome using MAUVE (Rissman et al. 2009).

Table 2 Summary of de novo assembly of UMX-103
Full size table

Gene prediction and functional annotation

The strain UMX-103 contains a single circular chromosome of 4,234,627 bp with an average G+C content of 43.41% (Table 3). The assembled genome consists of 39 scaffolds with an average scaffold size of 108,580 bp (Table 2). By using a combination of several gene-prediction software and manual inspection, a total of 4301 protein-coding genes and 98 RNA genes were identified in this strain (Fig. 2). The annotated genes which identified using Prokka software were used for functional annotation analysis. The functional annotation was conducted using EggNOG-mapper, the summary of functional categories of annotated genes is shown in Table 4. The result revealed existing of biosynthetic gene cluster of genes which are known for coding surfactin. This gene cluster belongs to Nonribosomal Peptide Synthetase (NRPS) family, particularly to the microbial surfactants group. These genes usually present in secondary metabolites biosynthesis, transport and catabolism (Doroghazi et al. 2014).

Table 3 Key features of UMX-103
Full size table
Fig. 2
figure 2

Genome features of UMX-103: the two outmost concentric circles denote the predicted protein-coding genes represent as forward strand (external blue circle) and the reverse strand (internal grey). The third concentric circle (purple) represents tRNAs while the fourth concentric circle (light brown) shows rRNAs genes. The fifth concentric (green and purple) represents the GC content. The green colour shows GC content more than the average while the purple colour shows the GC content below average. Purple and silver in the last inner concentric represent GC skew

Full size image
Table 4 Functional annotation of the predicted genes of UMX-103
Full size table

Phylogenetic and genomic similarity

The 16S rRNA from UMX-103 was used to carry out the phylogenetic analysis where it was aligned with other 16S rRNA genes of Bacillus strains which including; B. subtilis LM 4-2, B. subtilis BEST7003, B. subtilis KCTC 1028, B. subtilis 168, B. subtilis RO-NN-1, B. amyloliquefaciens DSM7, B. licheniformis, B. pumilus GR-8, and Paenibacillus macerans (Fig. 3). Average Nucleotide Identity (ANI) of UMX-103 was determined by comparing the whole genome with the selected references (Table 5). The highest ANI was detected with KCTC 1028 and 168 strains which is 89%. The seven housekeeping genes used to determine the species of the UMX-103 were detected using MLSTserver 1.8 (Larsen et al. 2012) (Supplementary Table 1). All of the housekeeping genes in UMX-103 are highly identical with the housekeeping genes of Bacillus subtilis.

Fig. 3
figure 3

Phylogenetic analysis based on 16S rRNA. Phylogenetic reconstruction was performed based on the sequence of 16S rRNA gene using MEGA7 (Kumar et al. 2016). The 16S rRNA genes sequence of Bacillus pumilus GR-8 and Paenibacillus macerans was used as outgroup

Full size image
Table 5 Average Nucleotide Identity of UMX-103
Full size table

Genomic Islands and horizontal genes transfer

Genomic islands is widely used to compare bacteria strains and identify essential genes in bacterial genome (Dobrindt et al. 2004; Langille et al. 2010). Basically genomic islands associate with horizontal gene transfer (HGT) which is also known as mobile genetic elements. The strain UMX-103 has witnessed a number of Horizontal Gene Transfer events. There are 15 genomic islands in UMX-103 that was predicted by IslandViewer 3 (Dhillon et al. 2015) and the localization of the predicted genomic islands is shown in (Fig. 4). The 15 predicted genomic islands consist of 331 genes (Supplementary Table 2). These genomic islands are possibly having a significant role in adapting and surviving the bacteria to different abiotic stress and antimicrobial resistance, which may occur after the bacteria exposed to different environment including the hydrocarbon contaminated soil. Features of the genomic islands in the strain UMX-103 are given in (Supplementary Table 3).

Fig. 4
figure 4

Genomic Island of UMX-103: Red colour defines predicted genomic islands using integrated method. The blue color shows genomic islands predicted by IslandPath-DIMOB while yellow colour shows genomic islands predicted by SIGI-HMM method. The broken lines represent scaffolds borders

Full size image

Genomic comparisons with closely related bacteria

Comparative genomics of UMX-103 with six related genomes (Table 6) showed that UMX-103 genome is closely related to Bacillus subtilis KCTC 1028 and Bacillus subtilis 168 with the genome sequence similarity of 93.99 and 93.44%, respectively. Analysis showed that UMX-103 has the largest genome size compared to the other bacteria strains studied. The genome contains the highest number of genes and the lowest GC contents which is 43.41%. Pangenome analysis resulted in the identification of 3434 core genes which present in all the Bacillus strains studied (Fig. 5). Pangenome composed of the essential genes in species and it also used as a method in identification of unknown bacteria (Lasken and McLean 2014). The genomic islands comparison (Table 7) showed that UMX-103 has the same number of genomic islands with Bacillus subtilis LM 4-2; however, the total number of genes in the genomic islands of UMX-103 is 322 genes, where only 108 genes were found in Bacillus subtilis LM 4-2.

Table 6 Genomic comparisons with closely related bacteria strains
Full size table
Fig. 5
figure 5

Pan genome analysis of UMX-103 with other related genomes: showing the core genes shared in the 6 compared genomes which UMX-103, B. subtilis LM 4-2, B. subtilis BEST7003, B. subtilis KCTC1028, B. subtilis 168 and B. subtilis RO-NN-1

Full size image
Table 7 Genomic islands comparison of UMX-103 with other related genomes
Full size table

Biosurfactant production genes

Based on the gene prediction and annotation analyses, we have identified 25 genes in UMX-103 which are involved in biosurfactant production. The list of the identified genes is presented in Supplementary Table 4. These genes involved in the biosynthesis and regulation of surfactin, which is a type of biosurfactant produced by Bacillus subtilis with high industrial value (Płaza et al. 2015a). The genes that involved in the biosynthesis of biosurfactant are including; 4-phosphopantetheinyl transferase (sfp), glucose-1-phosphate thymidyly transferase (rmlA), dTDP-glucose 4,6-dehydratase (rmlB), dTDP-4-dehydrorhamnose 3,5-epimerase (rmlC), dTDP-4-dehydrorhamnose reductase (rmlD) (Das et al. 2015), and non-ribosomal peptide synthetase (dhbF) (May et al. 2001).

In this study, we have also identified two operons srfA (Nakano et al. 1991) and pps (Coutte et al. 2010) which are involved in coding of the non-ribosomal peptide synthetase (NRPS) subunits, that catalyse the incorporation of the seven amino acid form surfactin (Coutte et al. 2010; Peypoux et al. 1999). The srfA operon contains four genes; srfAA, srfAB, srfAC, and srfAD, while the operon pps contains five genes; ppsA, ppsB, ppsC, ppsD, and ppsE. The srfA operon encodes surfactin synthetase subunits (Płaza et al. 2015a). Surfactin is made of seven amino acids which are (Glu–Leu–(D)Leu–Val–Asp–(D)Leu–Leu) (Cosmina et al. 1993). The gene srfAA encodes the peptide synthetize subunit which involved in the makeup of amino acids Glu, Leu and D-leu. Whereas, srfAB encodes the subunit that implicate in catalysis of Val, Asp, and D-leu. The third gene in the operon srfAC functions in the foundation of Leu amino acid. The surfactin synthase thioesterase subunit is produced by srfAD (Marahier et al. 1993).The activating enzyme sfp plays essential role in surfactin biosynthesis, as it tranforms the inactive protein that changes surfactin synthetase into an active form (Nakano et al. 1992; Płaza et al. 2015b).

In addition, we have also identified genes implicated in regulation of surfactin; comA and comP which comprise a signal transduction system that involved in the competence development pathway and is required for the transcription of srfA (Marahier et al. 1993; Nakano et al. 1992). The rest of the genes are involved in DNA-binding response, sporulation, phosphate regulon transcription, carbon storage, and sensory transduction protein (Supplementary Table 4).

Our results suggested that presence of biosurfactant genes in UMX-103 has the ability to produce surfactin which is lipopeptide biosurfactant. These results are in agreement with our earlier screening assays.

Conclusion

We conclude that, the new strain UMX-103 belongs to Bacillus subtilis species. It is a Gram positive bacteria, rod shape and with a length of 1.954 µm and a diameter of 540.9 nm. All the five different biosurfactant producing tests showed that UMX-103 has the capability to produce biosurfactant. In addition, we have successfully assembled the genome of the strain UMX-103 using a combination of both de novo and reference-guided assembly methods. The genome was assembled into 39 scaffolds with a size of 4,234,627 bp. Interestingly, we identified 25 genes which are involved in biosurfactant production, where 14 genes involved in biosynthesis and 11 genes associated with the gene regulation. Genomic analysis revealed that UMX-103 has the genes which promote biosurfactant production. Future work will be conducted to characterize the unknown function genes as well as biosurfactant genes using various Omics approaches.