Introduction

In silico prediction and characterization of a microbial secondary metabolite of biosynthetic gene clusters has successfully contributed to the development of new medicines (Loureiro et al. 2018). The majority of these metabolites belong to a wide variety of chemical classes and have shown antibiotic and antitumor activities (Law et al. 2019). Recently, the cost of genome sequencing has dramatically decreased and allows for the discovery of thousands of gene clusters encoding the biosynthetic machinery for these compounds (Palazzotto and Weber 2018). The experimental description of each gene cluster is still very difficult and cannot catch up with the rapidity of genomic discovery. Therefore, effective in silico prediction of the most promising targets within the genomes is essential for the successful genomic mining (Horn et al. 2015; Foulston 2019; Sun et al. 2019).

Rare actinobacteria have been recognized as one of the great sources for bioactive compounds and antibiotic (Bredholdt et al. 2007; Khanna et al. 2011; Hacene et al. 2000; Laidi et al. 2006; Balagurunathan and Radhakrishnan 2010). Uncommon and novel types of actinobacteria may contain unexpected genes that are phylogenetically distant from related strains and, subsequently, generate alternative novel antimicrobial agents (Donadio et al. 2005).

Illumina genome sequencing has proven to be an effective approach for genomics study and sequence analysis of individual genes, clusters of genes or operons, full chromosomes, or entire genomes of any organism (Bentley 2006; Castro et al. 2018). Complete sequences of the Streptomyces coelicolor and Streptomyces avermitilis genomes and many Streptomyces genome sequences showed a lot of silent gene clusters that encode bioactive molecules (Bentley et al. 2002; Choi et al. 2015). Several years ago, the traditional fermentation process revealed that Streptomyces ambofaciens produces spiramycin and congocidine. The advancement of genome mining approach has explored many other gene clusters within rare actinobacteria of potential bioactive molecules such as kanamycin derivatives and stambomycin, yet many bioactive molecules remain hidden within silent biosynthetic gene clusters (Aigle et al. 2014; Genilloud 2018).

The advancements in bioinformatics analysis of biosynthetic gene cluster identification have facilitated the processing of enormous genomic data of actinobacteria (Alam et al. 2011; Doroghazi et al. 2014; Abdelmohsen et al. 2015; Hug et al. 2018). Manual annotation found to be difficult, inefficient, and leading to inadequate annotations, while automated annotation of secondary metabolite clusters would lead to more accurate and complete annotations. In silico genomic prediction has successfully provided great analysis of secondary metabolism in bacterial genomes, to list few examples: ClustScan (Cullum et al. 2011), SBSPKS toolbox (Anand et al. 2010), and NP.searcher web server (Li et al. 2009a). Antibiotics and Secondary Metabolite Analysis SHell (antiSMASH) has generated rapid genome annotation of a varied range of bacterial and fungal strains (Medema et al. 2011; Blin et al. 2013; Villebro et al. 2019; Hu et al. 2019).

In the present study, Illumina whole genome sequencing was conducted along with software pipeline antiSMASH.3 for the analysis and annotation of secondary metabolite gene cluster. The antiSMASH database is considered a broad resource for the secondary metabolite biosynthetic gene clusters and encompasses gene clusters for more than 3000 finished bacterial genomes (Blin et al. 2017). This eventually allowed the identification and detection of all recognized classes of the secondary metabolite biosynthetic gene clusters in our strains. The detailed functional annotation was obtained and therefore opening the door for combinatorial biosynthesis for designing novel scaffolds of antibiotics.

Material and methods

Microorganism and their maintenance

Rare actinomycete isolates were cultivated using both starch casein broth and soya bean broth as described by Abbas and Edwards (1990) with slight modification as follows: Spore suspension of the isolates was cultivated into 35 ml of each of the broth media for 7 days at 30 °C. The incubation was conducted by shaking at 150 RPM. The final spore suspension was lypholized at the Mycological center, Assiut University, Assiut, Egypt, and then kept at − 20 °C.

The antimicrobial potentialities of the 50 rare actinobacteria isolates were tested against some bacterial pathogens (methicillin-resistant Staphylococcus aureus “ATCC 6538,” Streptococcus mutans “ATCC 25175,” Escherichia coli “O157:H7 ATCC 51659,” Klebsiella pneumonia “CCM 4415,” Salmonella enterica “ATCC 25566,” and Peudomonas aeuroginosa “ATCC 10145”). Tested pathogens were provided from Ain Shams Specialized Hospital and the Microbial Resources Center (MIRCEN) at the Faculty of Agriculture, Ain Shams University, Cairo, Egypt. Cultivation of these strains was conducted overnight in nutrient broth at 37 °C.

Agar well diffusion method

Agar well diffusion method was used for preliminary screening of the antimicrobial activity of the rare actinobacteria against tested bacterial pathogens. Actinobacteria spore suspensions were prepared. Cell-free supernatant was obtained by centrifugation for 5 min at 37 °C with a speed of 100 RPM. An amount of 250 μl of cell-free supernatant was added to each well in the nutrient agar Petri dishes containing 150 μl of 0.5 McFarland of tested bacterial spores (McFarland 1907). Petri dishes were then incubated for 24 h at 37 °C. Results were recorded by measuring the inhibition zone of bacterial pathogen around the well (Cooper 1972). All tests and experiments were made in duplicates. The most potent actinobacteria Micromonospora sp. Rc5 and Streptomyces sp. Ru87 were selected for further studies (Amin et al. 2017a; Amin et al. 2017b).

Extraction of genomic DNA from actinomycete strains

Extraction of genomic DNA of Micromonospora Rc5 and Streptomyces Ru87 was conducted using Promega Wizard® Genomic DNA Purification Kit as follows: 1 ml of each actinomycete spore suspension (3 × 105 CFU/ml) was aseptically added to 35 ml of sterile starch casein broth in 50-ml Erlenmeyer flasks. The flasks were incubated at 30 °C in a shaking incubator (Spectronics, USA) with an agitation rate (150 RPM) for 7 days. One milliliter of spore suspension was added to a 1.5-ml microcentrifuge tube. Cells were pelleted by centrifugation at 13,000×g for 2 min, and the supernatant was discarded. Cells were resuspended thoroughly in a mixture of 480 μl of 50 mM EDTA (Tris-acetic acid EDTA buffer) and 120 μl of Lysozyme. Incubation of the samples was conducted at 37 °C for 60 min followed by centrifugation for 2 min at 13,000×g. The supernatant was discarded, and 600 μl of Nuclei Lysis Solution was added followed by incubation at 80 °C for 5 min with an immediate cooling down to 37 °C. Three microliters of (4 mg/ml) RNase Solution was added to the lysate. The Eppendorf tubes were then inverted five times to allow complete mixing followed by incubating at 37 °C for 60 min. An aliquot of 200 μl of protein precipitation solution was added to the RNase-treated cell lysate and vigorously vortexed at high speed for 20 s. The mixture was incubated in ice for 5 min and then centrifuged at 13,000×g for 3 min. The supernatant containing the target DNA was transferred to a clean 0.5-ml Eppendorf tube containing 600 μl of isopropanol. Eppendorf tubes were gently inverted until the thread-like strands of DNA formed a visible mass. Samples were centrifuged at 13,000×g for 2 min, and the supernatant was poured off carefully followed by draining the tubes on a clean absorbent paper. An amount of 600 μl of 70% ethanol was added to each Eppendorf tube, and tubes were then inverted several times in order to wash the DNA pellet followed by centrifugation at 13,000×g for 2 min. Careful aspiration of the ethanol was performed, and then, the tubes were poured on a clean absorbent paper to allow the pellet to air-dry for 15 min. One hundred microliters of DNA rehydration solution was added to each tube and incubated at 65 °C for 1 h. DNA concentration was determined using a Nanodrop spectrophotometer (ND-2. 1000, Nanodrop Technologies) and stored at − 20 °C.

Agar gel electrophoresis

To check the integrity and quality of the extracted DNA, an aliquot of 5 μl of each sample was loaded on 1% Agarose Gel Electrophoresis in Tris-acetic acid EDTA buffer for 30 min at 90 V. The gel was stained with 50 mg/ml of ethidium bromide, and digital images were obtained for the DNA bands of expected size using UV transilluminator (Bio-Rad Laboratories, Hercules, CA).

Illumina whole genome sequencing and contig assembly

High-quality DNA extracted from actinobacteria mycelium was sequenced using Illumina MiSeq and HiSeq 2500 platforms using 2 × 250 bp paired-end technology by Microbes NG (http://www.microbesng.uk), which is supported by the BBSRC (grant number BB/L024209/1). The bioinformatics analysis was provided by the sequencing company using Trimmomatic to trim raw reads (Bolger et al. 2014) and other software such as Samtools (Li et al. 2009b) and bwa-mem (Li and Durbin 2009) to quality filter the reads and assemble the genome. The assembly metrics provided by MicrobesNG were calculated using QUAST (Quality Assessment Tool for Genome Assemblies). The taxonomic distribution of both strains was calculated using the Kraken software (Wood and Salzberg 2014). Phylogenetic tree of complete 16S rRNA genes of isolate Rc5 and Ru87 was constructed using the neighbor-joining method (Saitou and Nei 1987). Sequence similarity was conducted using Clustal W within the Mega 7 program. Contigs obtained from the sequenced genomes of each strain were assembled on the basis of reference genomes (Micromonospora carbonaceae in case of Rc5 and Streptomyces coelicolor in case of Ru87), and gaps were filled using Mauve Aligner 2.4. software (Rissman et al. 2009). Artemis software 16.0.11 (http://www.sanger.ac.uk/Software/Artemis) was used in order to allow visualization of sequence features and assembling of contigs and plotting of whole genomes of both isolates.

Annotation of rare actinobacteria genome

Gene annotation was performed via the NCBI (National Center for Biotechnology Information) Prokaryotic Genome Annotation Pipeline (PGAAP) as published online (Tatusova et al. 2016), and an additional annotation was done using Prokka version 1.1 (Seemann 2014) to assist identifying gene clusters. In this study, we used gene annotation via an antiSMASH web server version 3.0.5 for automatic whole genomic identification and analysis of biosynthetic gene clusters in each isolate (Medema et al. 2011).

Results

Antimicrobial potential of rare actinomycete isolates

Fifty rare actinomycete isolates were selectively isolated by physical and chemical means on humic acid vitamin agar media and starch casein agar media from different Egyptian governorates. The isolates were previously identified via morphological, chemotaxonomy, and biochemical methods (Abd-allah et al. 2012). The antimicrobial activity of these isolates was screened against tested bacterial pathogens (Staphylococcus aureus ATCC 6538, Pseudomonas aeruginosa ATCC 10145, Klebsiella pneumonia CCM 4415, Streptococcus mutans ATCC 25175, Escherichia coli ATCC 51659, and Salmonella enterica ATCC 25566).

Results obtained under shaking conditions using soya bean meal broth indicated that 32 (64%) rare actinobacteria showed pronounced antimicrobial activities against the selected bacterial pathogens while it was 29 (58%) on starch casein broth medium (Table 1). In the case of using soya bean broth, 27 isolates (84%) were active against Pseudomonas aeruginosa ATCC 10145, followed by 9 (28%) against Escherichia coli ATCC 51659 and 7 (21%) in case of both Staphylococcus aureus ATCC 6538 and Salmonella enterica ATCC 25566. Four isolates (12%) were active against Streptococcus mutans 25175, while only 3 isolates (9%) were active against Klebsiella pneumonia CCM 4415. The antimicrobial activity of rare actinobacteria growing on starch casein broth reported 23 (79%), 8 (27%), and 7 (24%) active against Pseudomonas aeruginosa ATCC 10145, Staphylococcus aureus ATCC 6538, and Klebsiella pneumonia CCM 4415, respectively. Four isolates (13%) were active against Streptococcus mutans 25175, and 3 isolates (1%) showed activity against Salmonella enterica ATCC 25566. No antimicrobial activity was recorded against Escherichia coli ATCC 51659. Our data indicated that isolate number Rc5 and Ru87 had the highest antimicrobial activity. They also produced a broad-spectrum antimicrobial compound(s) against both Gram-positive and Gram-negative tested pathogenic microorganisms.

Table 1 Screening assay of antimicrobial activity of potent selected rare actinobacteria isolates against food- and blood-borne pathogens
Full size table

Illumina sequencing genome notification

Digital images of agarose gel captured by UV trans illuminator (Bio-Rad Laboratories, Hercules, CA) confirmed the high quality of DNA extracted from Micromonospora Rc5 and Streptomyces Ru87. The whole genome of Micromonospora sp. Rc5 contains 2252-Mb raw reads with 128.284 coverage. The assembly consists of 513 contigs. The draft genome was 7,702,789 bp, with an average GC content of 73.64%. A total of 6792 coding sequences (CDS) with 6504 coding genes were identified by NCBI prokaryote pipeline. In the case of Streptomyces sp. Ru87, the draft genome was 7,662,503 bp, with an average GC content of 73.12% was assembled in 629 contigs. NCBI prokaryote pipeline annotation identified 6527 coding sequences (CDS) with 6051 coding genes, with an average GC content of 73.12%. The taxonomic distribution of both strains calculated using the software Kraken emphasized that Micromonospora sp. Rc5 belongs to genus Micromonospora, while Streptomyces sp. Ru87 belongs to genus Streptomyces. Micromonospora sp. Rc5 Whole Genome Sequencing Bio project has been deposited at EMBL (European Molecular Biology Laboratory)/GenBank under no. PRJNA354176 (BioSample SAMN06041774, Accession MQMK00000000). Streptomyces sp. Ru87 Whole Genome Sequencing Bio project has been deposited at EMBL/GenBank under no. PRJNA413750 (BioSample SAMN07765385, Accession PDIX00000000). Micromonospora sp. Rc5 16S rRNA gene sequence was deposited in Genbank under the accession number KY818317.1 and KY818662.1 for Streptomyces sp. Ru87.

Sequence analysis and phylogenetic tree construction

The whole genomes were sequenced via Illumina and then subjected to genome assembly. Genome annotation using NCBI prokaryote pipeline revealed 16S rRNA genes for each strain. Gene sequences were successfully deposited in Genbank under the accession number KY818317.1 for isolate Rc5 and KY818662.1 for isolate Ru87. The 16S rRNA gene sequence of isolate Rc5 was compared with other Micromonospora gene sequences in the NCBI GenBank database. The phylogenetic tree was generated against closely related Micromonospora strains using the neighbor-joining method. Neighbor-joining (NJ) phylogenetic tree consisted of two main clades. Moreover, isolate Rc5 was gathered with other Micromonospora strains in the same clade, which ensures that isolate Rc5 belonged to this genus. The most similarity of 16S rRNA gene sequence belonged to Micromonospora oryzae and Micromonospora harpali with 73% bootstrap value (Fig. 1).

Fig. 1
figure 1

Phylogenetic tree of complete 16S rRNA genes of isolate Rc5. Sequence similarity was conducted using Clustal W within the Mega 7 program. Phylogenetic trees were constructed using the neighbor-joining method

Full size image

A comprehensive analysis of complete 16S rRNA Streptomyces gene tree was conducted in order to clarify the relationship between Ru87 isolate and closely related Streptomyces species. The 16S rRNA gene sequence of strain Ru87 was compared with the nucleotide sequences of other Streptomyces strains in the NCBI GenBank database. The phylogenetic tree was generated based on the comparison between the 16S rRNA gene sequence of the strain Ru87 and other nucleotide sequences from closely related Streptomyces strains. Isolate Ru87 was grouped with Streptomyces gingianensis and Streptomyces morales partial sequence with 89% bootstrap value (Fig. 2). Sequencing analysis of the 16S rRNA gene sequence confirmed that Rc5 and Ru87 were identified as Micromonospora sp. Rc5 and Streptomyces sp. Ru87.

Fig. 2
figure 2

Phylogenetic tree of complete 16S rRNA genes of isolate Ru87. Sequence similarity was conducted using Clustal W within the Mega 7 program. Phylogenetic trees were constructed using the Neighbor-Joining method

Full size image

Genome mining of whole genome sequence of selected rare actinobacteria using antiSMASH analysis

Biosynthetic pathways are very important in novel antibiotic discovery. Whole genome sequences of Micromonospora sp. Rc5 and Streptomyces sp. Ru87 obtained from the Illumina sequencing were assembled on the basis of reference genomes Micromonospora carbonaceae for Rc5 and Streptomyces coelicolour for Ru87. The gaps were filled using mauve Aligner 2.4. software (Rissman et al. 2009). Genomes were mined using the antiSMASH server for further prediction of secondary metabolite biosynthetic gene clusters. The genomes sequenced of Micromonospora sp. Rc5 displayed more diverse antiSMASH readout than Streptomyces sp. Ru87. A total of 33 potential secondary metabolite gene clusters were predicted by Micromonospora sp. Rc5, 5 polyketide synthase (PKS), 4 non-ribosomal polyketide synthase (NRPS), 10 hybrid polyketide synthases, 4 terpenes, 3 lantipeptides, 2 saccharides, 1 siderophore, 1 bacteriocin, 1 arylpolyene, and 2 unidentified clusters (Table 2). Micromonospora sp. Rc5 whole genome recorded the highest similarity hits with biosynthetic gene clusters coding for the following compounds: sioxanthin (terpene) (100%), SapB (lantipeptide) (100%), desferrioxamie B (siderophore) (66%), methoxyhydroquinones (PKS) (57%), and tetrocarcin A (PKS-hybrid) (53%) (Table 3).

Table 2 Predicted secondary metabolite biosynthetic gene clusters recorded by Micromonospora sp. Rc5 and Streptomyces sp. Ru87 whole genomes analyzed using the antiSMASH 3.0.5 database
Full size table
Table 3 Gene annotation of Micromonospora sp. Rc5 and Streptomyces sp. Ru87 whole genomes using the antiSMASH 3.0.5 database showing the highest similar biosynthetic cluster hits
Full size table

In the case of Streptomyces sp. Ru87, antiSMASH generated 19 potential secondary metabolite gene clusters encoding for 4 NRPSs, 4 terpenes, 2 PKS, 2 bacteriocins, 2 lasso peptide, 1 siderophore, 1 lantipeptide, 1 hserlactone, 1 ectoine, and 1 linaridin (Table 2). The whole genome of Streptomyces sp. Ru87 showed the highest similarity hits with the biosynthetic clusters coding for ectoine (100%), paenibactin (NRPS) (66%), albachelin (NRPS) (60%), erythrochellin (NRPS) (42%), and labrinthopeptin (lantipeptide) (40%) (Table 3).

Discussion

Actinobacteria generate extensive compounds with a variety of biological activities (Bredholdt et al. 2007; Khanna et al. 2011), and rare actinobacteria are known as a great potential source of antibiotic production (Hacene et al. 2000; Laidi et al. 2006; Balagurunathan and Radhakrishnan 2010). We believe that uncommon and rare types of actinobacteria may contain unexpected genes. These genes are phylogenetically far off related strains and, subsequently, generate alternative novel bioactive molecules (Donadio et al. 2005).

In this study, isolates Rc5 and Ru87 showed great inhibition activity against selected Gram-positive and Gram-negative pathogens. Illumina genome sequencing in combination with bioinformatics analysis using antiSMASH software proved to be an effective approach for genomics study and sequence analysis for the selected strains (Bentley 2006; Jakubiec-Krzesniak et al. 2018). The genome size of Micromonospora sp. Rc5 and Streptomyces sp. Ru87 was approximately similar, and it resembles the normal genome size variation among other actinobacteria which is from 4 to 12 Mb (Komaki et al. 2016; Jiang et al. 2015). Our results confirmed that Rc5 and Ru87 isolates are distinct from comparable strains on the database. Our previous phylogenetic analysis via Sanger sequencing of partial 16S rRNA genes (Amin et al. 2017a; Amin et al. 2017b) is in complete agreement with other complete 16S rRNA genes generated with Illumina whole genome sequencing. This will definitely confirm that using Sanger sequencing for characterization of partial 16S rRNA gene (V3 region) of actinobacteria is effective in genera identification. Moreover, whole genome sequencing in taxa identification is expensive and needs more data analysis (Sims et al. 2014; Luo et al. 2014). Further investigations including DNA-DNA hybridization, additional chemotaxonomic, and biochemical tests are required to identify their species level.

The antiSMASH readout of our actinobacteria is in agreement with other publications (Horn et al. 2015) and indicated a direct relation between genome size and biosynthetic potential. However, Micromonospora sp. Rc5 with larger genome size displayed more diverse (antiSMASH) read-out than Streptomyces sp. Ru87. These results showed the higher antimicrobial potential of Micromonospora sp. Rc5 than Streptomyces sp. Ru87. GC content is the main measure of the relatedness of microorganisms; it varies with different organisms due to variation in selection, mutational bias, and biased recombination-associated DNA repair (Birdsell 2002). So, we clarified that fluctuation in genome size and GC content is species dependent.

Both strains have shown great potential to produce post-translationally modified peptides such as terpenes, lantipeptides, saccharides, siderophores, bacteriocin, arylpolyene, lasso peptide, hserlactone, ectoine, and linaridins. Illumina whole genome sequencing results confirmed that both strains possess polyketides and non-ribosomal peptide gene clusters which are the major classes of pharmacologically active natural products. This ensures that the rare actinobacteria produce different metabolites synthesized by conserved enzymes (Komaki et al. 2016; Gomez-Escribano and Bibb 2012). This obviously shed the light on the high metabolic potential of the two actinomycete species to generate diverse bioactive molecules. It is of interest to mention that the combination of both genomics-metabolics sketchings of rare actinobacteria led to the characterization of cryptic or undiscovered biosynthetic clusters with a new mode of action to inhibit resistant bacteria (Foulston 2019; Xu and Wright 2019).

Actinobacteria strains Rc5 and Ru87 were tested in vitro and showed significant antibacterial activity. This antagonistic activity found to be in agreement with antiSMASH pipeline prediction of several biosynthetic gene clusters. Physicochemical analysis of the purified antimicrobial compounds was performed in order to determine the possible chemical group for each strain (Amin et al. 2017a; Amin et al. 2017b). The physicochemical analysis and bioinformatics genome mining confirmed the ability of Micromonospora sp. Rc5 to produce tetrocarcin antibiotic harboring phthalate core and caused inhibitory effect against S. aureus ATCC 6238. We have found that the labyrinthopeptin structure is in agreement with the physicochemical analysis of the active fraction produced by Streptomyces sp. Ru87, which is cyclic or aromatic peptide structure. Our data indicated that the mutual understanding from in silico and in vitro approaches leads to the identification of the closest possible antimicrobial compound produced by actinobacteria strains. This is a unique and infrequent approach of Illumina whole genome sequencing for observing the biosynthetic clusters of antibiotic-producing actinobacteria in Egypt.

Conclusion

Our findings illustrated that the Egyptian soil is very rich of high potential biosynthetic of rare actinobacteria. The genetic potential of secondary bioactive molecule producers was successfully determined via genome mining. Sequencing actinomycete genomes provide useful information for inventing novel antimicrobial agents. We ensure that rare actinobacteria genome sequencing guided with bioinformatics analysis will open the door for scientists to explore more about the biochemical pathways and consequently the discovery of novel bioactive molecules. This approach would contribute to more discovery of natural antibiotics and therefore enhance pharmaceutical industry. The current study helps to control the problem of antimicrobial drug resistance and improve the health care in Egypt, the UK, and worldwide. In addition, it introduces potential bioactive agents that would support the drug discovery in Egypt.