Our environment still contains lots of undiscovered micro-organisms that can support potential drug discovery [5]. In this study, the soil metagenomic library was screened to discover the biosynthetic pathways in uncultured microorganisms. Similarly, several studies have involved in the mining of metagenomic libraries to determine the variety of biosynthetic gene clusters [10]. Screening of large biosynthetic gene libraries detected abundant remarkable clones with biosynthetic potential [11, 12, 28]. In this study, the NRPS PCR assay captured successfully several NRPS gene fragments in E. coli clones that matched corresponding NRPS genes on the GenBank. Those positive hits indicate the existence of NRPS biosynthetic genes within the soil metagenome. A similar PCR screening approach was used by other research groups to capture NRPS clones from soil metagenomes [28]. These findings are in agreement with other study that confirmed capturing NRPS genes in Cuban soil metagenome using NRPS PCR using different primer pair [28].
In the herein study, NRPS PCR assay recovered clones that had similarity to sequences belonging to the Delftia genus, Proteobacteria phyla, suggesting that they are the dominant phyla and the tested soil is abundant in metabolite-producing bacteria. Several reports confirmed the biotechnological potential of the Delftia genus. For instance, a recent study has identified Delftibactin A (NRP) isolated from novel environmental Delftia spp. with potent antimicrobial activity against methicillin-resistant Staphylococcus aureus, vancomycin-resistant Enterococcus, Acinetobacter baumannii, and Klebsiella pneumonia [34]. Furthermore, the biotechnological perspective of Delftia sp. JD2 was previously confirmed using a genomic approach [35]. In the same context, functional NRPS genes were identified in Delftia tsuruhatensis MTQ3 for bacteriocins and siderophore production [36].
The level of diversity in the metagenomic library is low since all positive hits related only to the genus of Delftia. Reversely, another study declared high bacterial biodiversity in Cuban soil with abundant enzymatic activity [28]. Due to the small number of clones selected for the sequencing phase, and a huge sequencing work is needed to identify all the gene clusters in the metagenomic library. Additionally, higher biodiversity would be achieved if a range of primer sets were used. Similarly, Amos et al suggested that using different primers focusing on different groups of microorganisms would increase the biodiversity of the positive hits [28]. Several related NRPS sequences were isolated from different sites suggesting a widespread environmental distribution of bacteria harboring NRPS genes. Similar results were recorded by a research study on the metagenomic library extracted from the soil in Cuba [28].
In the case of natural product science, phylogenetic relationships are highly informative in the design and function of the genes involved in secondary metabolite biosynthesis. Non ribosomal peptide synthetases provide a model in which individual domain phylogenies exhibit various predictive capabilities, determining features of substrate specificity correlated to the final metabolic product [37]. In this study, sequencing of selected NRPS clones showed that they belonged to different regions in NRPS gene clusters, thus they were separated in different phylogenetic trees. It also may evidence for the evolutionary process of NRPS genes [38]. Phylogenetic relationships showed that clones no. 15cd30 and 15cd34 were clustered together as a sister group to the genus Delftia, supported by high blast identity values and the tree was in agreement with a previous phylogenetic analysis of the group [28]. Clone no 15cd30 and 15cd34 shared amino acid similarity with strains of Delftia tsuruhatensis and Delftia acidovorans, suggesting similar secondary metabolites production. However, they even have some mismatches at the terminal region indicating a distinct scope of bioactive molecules production. In case of clone 15cd35 and clone 15cd37, NRPS gene sequences of exhibited noticeable similarity values to AMP (A) domain sequences of belonged to members of genus Delftia spp. Though they still showed several differences in amino acid sequences and low bootstrap value (54%) with the closest phylogenetic neighbors. Thus, the metagenomic NRPS genes are predicted to produce distinct non-ribosomal peptides or suggesting many of the clones recovered came from undiscovered NRP pathways. These changes in amino acid sequences can be due to conservative mutation or radical substitution [39, 40]. Especially, the mutations are detected in the protein interior which may be a sign of structural constraints [41]. A similar study confirmed that multiple alignments and the phylogenic tree of amino acid sequences of NRPS genes of Streptomyces sp. BDUSMP 02 reveals the potential to produce a new type of antibacterial compounds belonging to the NRPS type [42]. Other studies supported using multiple alignments and phylogenetic trees approach in identifying conserved sequence regions and establishing evolutionary relationships [38]. Similarly, a study emphasized that sequence analysis of the 21 most active improved variants revealed that each contained between one and three changes to their primary amino acid sequence, suggesting that minimal sequence variation was able effect dramatic improvements in NRPS domain function [43]. Fischbach et al. (2007) declared that NRPS is considered as enzymes that assemble the key skeletons of natural products, so any mutation of these genes is expected to give the clearest impact on the metabolite pattern and their functions. In agreement with our results, a substitutional mutation in NRPS genes can lead to ≈10-fold improvements in enzyme activity and can trigger the creation of new derivatives of NRP antibiotics [23].
In this study, bioinformatics tools such as NRPS predictor tool, NRPS predictive blast webserver, and PDBsum database were used to identify NRPS domains at both genome/proteome level and their substrate suggesting different structures with relevant biological activities. Similarly, other studies support the use of the NRPS predictor tool for reliable prediction of adenylation domain (NRPS) specificities [44, 45]. None of the predictive methods could infer any substrate specificity for 15cd35 and 15cd37 NRPS sequences, suggesting completely new types of specificity and PDBsum database showed their resemblance to acetyltransferase domains still with low z- score. This may be due to mutations from corresponding sequences recorded in the multiple alignments. Different substrate specificity using bioinformatic tools refers to variation in the NRPS clones, thus different non-ribosomal peptide biosynthesized. Similar results recorded that substrate specificity is important in determining the final bioactive product such as the A domain TycB_m3 activates l-tryptophan and phenylalanine for tyrocidine biosynthesis [46, 47]. Additionally, A domain of the barbamide biosynthetic gene cluster activates 100% specificity for leucine and valine, and 80% for trichloroleucine [48].
Finally, this study emphasizes the presence of biosynthetic NRPS genes in the soil metagenomic library. The molecular, phylogenetic, and bioinformatic analysis confirmed that mainly 15cd35 and 15cd37 are distinct clones harboring biosynthetic potential with undetected substrate specificity and thus can produce improved yield or even new antibiotics. Similar results confirmed that directed mutation in substrate specificity code within the A-domain of NRPS genes can produce a high yield of certain antibiotic or even a novel antimicrobial compound [49–52]. This study supports the methodology of using PCR assay for screening soil metagenomes as a tool for drug discovery. Further genetic manipulation of NRPS clones will provide a positive impact on the pharmaceutical sector and consequently the health sector.