Comprehensive analysis of orthologous genes reveals functional dynamics and energy metabolism in the rhizospheric microbiome of Moringa oleifera

Abstract

Moringa oleifera, known for its nutritional and therapeutic properties, exhibits a complex relationship with its rhizospheric soil microbiome. This study aimed to elucidate the microbiome's structural composition, molecular functions, and its role in plant growth by integrating Clusters of Orthologous Genes (COG) analysis with enzymatic functions previously identified through KEGG, CAZy, and CARD databases. Metagenomic sequencing and bioinformatics analysis were performed from the rhizospheric soil microbiome of M. oleifera collected from the Mecca district in Saudi Arabia. The analysis revealed a role for the rhizospheric microbiome in energy production, storage, and regulation, with glucose serving as a crucial precursor for NADH synthesis and subsequent ATP production via oxidative phosphorylation. Key orthologous genes (OGs) implicated in this process include NuoD, NuoH, NuoM, NuoN, NuoL, atpA, QcrB/PetB, and AccC. Additionally, OGs involved in ATP hydrolysis, such as ClpP, EntF, YopO, and AtoC, were identified. Taxonomic analysis highlighted Actinobacteria and Proteobacteria as the predominant phyla, with enriched genera including Blastococcus, Nocardioides, Streptomyces, Microvirga, Sphingomonas, and Massilia, correlating with specific OGs involved in ATP hydrolysis. This study provides insights into the molecular mechanisms underpinning plant–microbe interactions and highlights the multifaceted roles of ATP-dependent processes in the rhizosphere. Further research is recommended to explore the potential applications of these findings in sustainable agriculture and ecosystem management.

Supplementary information

The online version contains supplementary material available at 10.1007/s10142-025-01580-7.

Introduction

Moringa oleifera, colloquially referred to as the drumstick, horseradish, or miracle tree, is a member of the Moringaceae family indigenous to various Asian regions, including Saudi Arabia (Gupta and Ahmed 2020). This autochthonous plant species exhibits a plethora of potential salutary effects attributable to its rich nutritional profile and bioactive constituents (Gopalakrishnan et al. 2016). The plant's seeds and foliar structures are replete with essential nutrients, encompassing indispensable amino acids, ascorbic acid, retinol, and minerals such as calcium, potassium, and iron, as well as an array of antioxidants including flavonoids, polyphenols, and ascorbic acid (Bibi et al. 2024; Kashyap et al. 2022; Milla et al. 2021). These components confer potent antioxidant and anti-inflammatory properties, potentially conferring cytoprotective effects, enhancing immunological function, and mitigating inflammatory processes associated with pathological conditions such as rheumatoid arthritis (Pareek et al. 2023). Furthermore, crude extracts have demonstrated antimicrobial efficacy and putative benefits for cognitive function and gastrointestinal health (Pareek et al. 2023). The pharmaceutical relevance of this autochthonous plant is underscored by its biosynthesis of key phytochemicals, including glycosides, β-sitosterol, and N-α-rhamnophyranosyl vincosamide, which contribute to the regulation of blood pressure and management of cholesterol homeostasis (Yeshi et al. 2022). M. oleifera is also recognized for its dermatological benefits, efficacy in ameliorating chronic maladies, and its hepatoprotective properties against oxidative stress and toxicity, as well as its capacity to promote tissue regeneration (Islam et al. 2021; Zouboulis et al. 2023). Beyond its therapeutic applications, this plant demonstrates potential in diverse domains such as aqueous purification, cosmetic formulations, soil amendment, and biofuel production (Takase et al. 2022). While these potential benefits are promising, it is imperative to note that a significant proportion of studies have been conducted in animal models or in vitro systems, necessitating further human clinical trials to elucidate fully the effects of this plant on human physiology and pathophysiology (Su et al. 2023).

Recent investigations have revealed that the rhizospheric soil microbiome of M. oleifera demonstrated an intricate and fluid interrelationship with its host plant, distinguished by enhanced microbial populations and taxonomic diversity in comparison to surrounding bulk soil (Ashy et al. 2023; Tashkandi et al. 2022). Recent metagenomic studies have revealed several key aspects of this interaction, including increased abundance of genes encoding Carbohydrate-Active Enzymes (CAZymes) and other KEGG enzymes in the rhizosphere (Ashy et al. 2023; Alshareef 2024; Tashkandi and Baz 2023). The rhizospheric microbiome contains a higher abundance of antibiotic resistance genes (ARGs) compared to bulk soil, which may have implications for the spread of antibiotic resistance (Shami et al. 2022; Jalal and Sonbol 2024). Beneficial rhizobacteria in the M. oleifera rhizosphere contribute to plant growth promotion through various mechanisms, including the production of indole acetic acid and other growth-promoting compounds (Lalarukh et al. 2022; Tashkandi et al. 2022). The composition and quantity of root exudates from M. oleifera significantly influence the density, composition, and biological activity of microbial communities in the rhizosphere (Tashkandi et al. 2022). Additionally, the rhizospheric phageome of M. oleifera plays a role in horizontal gene transfer and may impact the distribution of ARGs (Tashkandi and Baz 2023). These findings highlight the mutually beneficial relationship between M. oleifera and its rhizospheric microbiome, emphasizing its importance for plant health and soil ecology. Notwithstanding the extant corpus of research on M. oleifera (and other plant species), there remains a conspicuous lacuna in the utilization of the Clusters of Orthologous Genes (COG) database (Tatusov et al. 2003) for the elucidation of enzymes and proteins encoded by orthologous genes potentially extant in the rhizospheric soil microbiome of this economically and ecologically significant plant species. The database of Clusters of Orthologous Genes (COG) is a comprehensive resource used for the classification and analysis of orthologous genes (OGs) across a diverse array of eukaryotic or prokaryotic species based on their evolutionary relationships (Tatusov et al. 2003). The implementation of such a comprehensive analysis would provide a novel dimension to our understanding, bridging the existing knowledge gap and contributing to a more holistic comprehension of the role of rhizospheric soil microbiome in the overall health and vitality of their associated wild plant species.

This study aimed to address the knowledge gap regarding the functional roles of the rhizospheric soil microbiome associated with Moringa oleifera by employing a comprehensive Clusters of Orthologous Genes (COG) analysis. Our specific objectives were to: (1) characterize the taxonomic composition of the rhizospheric soil microbiome of M. oleifera; (2) identify the prevalent functional categories and specific orthologous genes (OGs) enriched in the rhizosphere compared to bulk soil; (3) investigate the roles of these OGs in key metabolic processes, particularly those related to energy production, ATP hydrolysis, and plant–microbe interactions; (4) integrate COG data with existing KEGG, CAZy, and CARD database information to provide a holistic understanding of the microbiome's functional potential; and (5) identify specific microbial taxa associated with highly abundant OGs to elucidate potential keystone species driving these processes.

Materials and methods

DNA extraction and whole genome shotgun sequencing

Environmental specimens were procured in biological triplicates from the rhizospheric soil of indigenous Moringa oleifera specimens situated in the Mecca district, Saudi Arabia (21°12′17.8"N, 39°31′26.4"E) (Al-Eisawi and Al-Ruzayza 2015). Control specimens consisting of bulk soil were systematically collected at a one-meter radius from each selected plant specimen. All samples underwent immediate cryopreservation in liquid nitrogen followed by maintenance at −20°C. Genomic DNA isolation was executed utilizing the cetyltrimethylammonium bromide (CTAB) methodology as previously delineated (Hurt et al. 2001). Post-extraction, RNA contamination was eliminated through RNase A treatment (10 μM) at 37°C. DNA structural integrity was validated via electrophoretic separation on 1% agarose gel. Quantification and standardization of DNA concentration to 10 ng/μL was achieved utilizing a double-stranded DNA Assay kit. Six DNA specimens (1 μg per sample) were subsequently submitted to Novogene Co., Ltd. for next-generation sequencing and computational analyses.

DNA sequencing and library preparation

DNA samples underwent physical fragmentation utilizing a Covaris Sonicator to achieve a target fragment size of 350 bp. The resultant fragments were subjected to a series of enzymatic modifications, including end-polishing and A-tailing, followed by ligation with full-length adaptors compatible with PCR amplification using the i7/i5 dual index primer set 1 (E7600). Sequencing libraries were constructed employing the Ultra DNA Library Prep kit for Illumina (NEB, USA), adhering strictly to the manufacturer's protocol.

High-throughput sequencing, data deposition, and quality control

High-throughput sequencing was performed on the Illumina HiSeq 2500 platform. The resulting high-quality sequencing data were deposited under Bioproject no. PRJEB55112 in the European Nucleotide Archive (ENA; https://www.ebi.ac.uk/ena) and the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov/), the two partners in the International Nucleotide Sequence Database Collaboration (INSDC). These sequencing data have been assigned unique identifiers within the repository, where the rhizosphere soil biosamples were allocated the run accession numbers ERR12764799, ERR10100771, and ERR10100772. Correspondingly, the bulk soil biosamples were designated the run accession numbers ERR12764800, ERR10100774, and ERR10100781. Rigorous quality control measures were implemented to ensure data integrity. These encompassed the retrieval of filtered clean data, excision of low-quality bases (Q-value ≤ 38) surpassing a 40-bp threshold, and elimination of reads containing N nucleotides exceeding a 10-bp threshold or exhibiting adapter overlap beyond a 15-bp threshold.

Metagenomic assembly and gene prediction

De novo assembly was executed utilizing MEGAHIT (K-mer = 55), subsequent to the elimination of chimeric sequences as previously elucidated (Mende et al. 2012; Karlsson et al. 2012; Oh et al. 2014). Unassembled reads from all samples underwent a secondary assembly process to generate NOVO_MIX scaffolds, adhering to established protocols (Mende et al. 2012; Nielsen et al. 2014). Sequence alignment was performed using Soap 2.21, with effective open reading frames (ORFs) and scaftigs (termed NOVO_MIX) serving as the reference. Gene prediction was conducted employing MetaGeneMark (Nielsen et al. 2014), followed by a dereplication step utilizing the Cluster Database at High Identity with Tolerance (CD-HIT) algorithm (Li and Godzik 2006; Fu et al. 2012).

Construction of non-redundant gene catalogues and functional annotation

Subsequently, non-redundant gene catalogues (nrGC) were constructed through greedy pairwise comparison methodologies (Li et al. 2014). The resultant gene set underwent annotation via MEGAN (MEta Genome ANalyzer), succeeded by the quantification of functional abundance based on the eggNOG database (version 4.0) (Huson et al. 2016, 2011; Powell et al. 2014; Huerta-Cepas et al. 2019). Subsequent analyses, encompassing table clustering and Principal Coordinate Analysis (PCoA), were executed (Lozupone and Knight 2005; Lozupone et al. 2007, 2011). Deduced amino acid sequences of diverse annotated genes were further mapped to the eggNOG database employing DIAMOND (https://github.com/bbuchfink/diamond) and eggNOG-mapper (v2.0.1) (Maranga et al. 2023; Huerta-Cepas et al. 2019). The annotated genes were categorized into various ontologies, including Gene Ontology for enzyme detection consortium (Consortium 2019), Kyoto Encyclopedia of Genes and Genomes (KEGG) (https://www.genome.jp/kegg/pathway.html) for metabolic pathway elucidation (Kanehisa et al. 2024), and Clusters of Orthologous Genes (COG) (https://www.ncbi.nlm.nih.gov/research/COG) for orthologous gene (OG) clustering (Tatusov et al. 2003). Furthermore, the generated unique OGs were subjected to a rigorous homology search utilizing the Blastp algorithm against Comprehensive Antibiotic Resistance Database (CARD, https://card.mcmaster.ca/ontology/), employing a stringent e-value threshold of ≤ 1^e−5 (Martínez et al. 2015). To facilitate a more nuanced and comprehensive understanding resistome of the soil rhizospheric microbiome of M. oleifera, the identified ARGs were systematically categorized into antimicrobial resistance (AMR) families (Liu and Pop 2009).

To generate a robust taxonomic classification, HMMER (http://hmmer.janelia.org) was employed to identify single-copy marker genes from the Genome Taxonomy Database reference genomes (Eddy 2011). This approach facilitated a comprehensive and precise taxonomic assignment of the metagenomic sequences.

Results

Fidelity of assembled and annotated orthologous gene sequences

The results in Table S1 provides a comprehensive characterization of the orthologous genes (OG) queries subsequent to their alignment with homologous sequences (subject) in the National Center for Biotechnology Information (NCBI) database. The assembly process of OG or gene fragments yielded an impressive array of approximately 280,000 individual open reading frames (ORFs) and 720,000 scaftigs (NOVO_MIX) (Table S1). These queries correspond to aligned sequences exhibiting a remarkable range of nucleotide lengths, spanning from approximately 30 to 5000, with a stringent minimum identity threshold of 50%. The alignment process revealed mismatches ranging from 0 to approximately 750 nucleotides. Notably, the aligned sequence lengths of both query and corresponding subject demonstrated minimal interference from gaps, which were observed to range from 0 to a mere 32 nucleotides. The robustness of these alignment parameters, including the substantial number of assembled sequences, the wide range of sequence lengths, the high identity threshold, and the minimal gap interference, collectively provide strong evidence for the high fidelity and reliability of the sequence alignment and annotation results. This comprehensive dataset forms a solid foundation for subsequent analyses and interpretations in our study.

The fidelity of eggNOG datasets was further corroborated through the implementation of principle coordinate analysis (PCoA), a multivariate statistical technique employed to elucidate the spatial relationships among samples within each of the two soil types surrounding Moringa oleifera (Fig. 1). The results of this analysis revealed a distinct and unambiguous separation between the microbiomes of the two soil types under investigation. Specifically, the bulk soil microbiome samples consistently clustered on the positive axis of PCoA 1 (or PC1), while the rhizobiome samples clustered on the negative axis of PC1. This clear demarcation was observed at both the categorical gene level (Fig. 1a) and the orthological level (Fig. 1b). The consistent positioning of samples across these two hierarchical levels of analysis provides robust evidence for the divergence in microbial community composition between the bulk and rhizospheric soil environments.

Fig. 1.

Principal Coordinate Analysis (PCoA) derived from the Clusters of Orthologous Genes (COGs) database at the categorical (a) and orthological (b) tiers, depicting the microbial consortia of prokaryotic and eukaryotic organisms in specimens collected from the rhizospheric (R) and surrounding bulk (S) soils of Moringa oleifera

Functional profiling of soil microbiomes via eggNOG analysis

The eggNOG-based annotation process has enabled the assignment of Clusters of Orthologous Genes (COG) identifiers to the non-redundant orthologous gene queries. These COG IDs are categorized into distinct functional groups, designated by alphabetical letters A through Z. Each COG ID is associated with a specific enzyme or protein, thus providing a comprehensive functional classification of the annotated genes (Table S2). The robustness of the eggNOG-based results is further corroborated by the data presented in Fig. S1, which demonstrates a clear delineation between the rhizospheric soil microbiomes and those of the surrounding bulk soil, based on the abundance of functional groups (Fig. S2). Analysis of the quantitative distribution of orthologous group (OG) queries across the microbiomes of both soil types revealed that categories C (Energy production and conversion), E (Amino acid transport and metabolism), G (Carbohydrate transport and metabolism), L (Replication, recombination and repair), and S (Function unknown or uncertain) exhibited the highest prevalence (Fig. S2 and Table S3). Notably, when examining the abundance profiles within the microbiomes of the two soil types, all functional categories, with the exception of N (Cell motility), W (Extracellular structures), and Z (Cytoskeleton), displayed elevated levels of OG queries in rhizospheric soil microbiomes compared to their bulk soil counterparts (Fig. 2 and Table S4). This observation underscores the distinct functional enrichment patterns characteristic of the rhizosphere environment.

Fig. 2.

Abundance (a) and relative abundance (b) of orthologous genes at the categorical level of Clusters of Orthologous Groups (COGs) within the rhizospheric (R) and bulk (S) soil microbiomes of Moringa oleifera. Further information is available in Table S4

The results presented in Fig. S3 and Table S5 elucidate the top 200 orthologous genes (OGs) as determined by the frequency of gene queries (≥ 660) across the rhizospheric and bulk soil microbiomes associated with Moringa oleifera. Among this subset, 33 OGs exhibited significantly elevated relative abundance in the rhizospheric microbiome compared to their bulk soil counterparts (Fig. 3 and Table S6). These OGs are distributed across diverse functional categories, including C, CH, G, I, J, K, L, O, P, Q, S, T, U, and V (Table S7). The heat map visualizations presented in Figs. S4 and S5 provide an inconclusive representation of the abundance profiles for selected OGs. Comprehensive information pertaining to the 33 selected OGs is delineated in Table S7.

Fig. 3.

Abundance (a) and relative abundance (b) of orthologous genes predominantly present in the rhizospheric (R) soil microbiome relative to the bulk (S) soil microbiome of Moringa oleifera. These genes are among the top 200 most frequently queried orthologs, with query number of ≥ 660. Further information is available in Table S6. Letters preceding COG IDs refers to their functional categories

Notably, nine of these OGs are integrated within eggNOG pathways, with four of them—specifically AtpA, UbiH, AccC, and CypX— also being represented in KEGG pathways. These four genes encode for FoF1-type ATP synthase (alpha subunit), 2-polyprenyl-6-methoxyphenol hydroxylase, acetyl-CoA carboxylase, and cytochrome P450, respectively. Of particular interest, the OG atpA is implicated in two distinct KEGG pathways, e.g., "Oxidative phosphorylation" and "Photosynthesis". The remaining three OGs are associated with the KEGG pathways "Ubiquinone and other terpenoid-quinone biosynthesis", "Fatty acid biosynthesis", and "Biotin metabolism", respectively (Table S7).

Microbial taxa associated with highly abundant orthologous groups

The results delineated in Table S8 encompass the comprehensive set of orthologous genes (OGs) present in the microbiomes of both soil types associated with Moringa oleifera. Among the six most prevalent phyla, as determined by the number of OG queries, Actinobacteria and Proteobacteria exhibited the highest representation, while the eukaryotic phylum Streptophyta demonstrated the lowest (Fig. 4a). At the genus level, the six most prominent taxa were identified as Blastococcus, followed by Nocardioides, and Streptomyces within the phylum Actinobacteria, and Microvirga, followed by Sphingomonas, and Massilia within the phylum Proteobacteria (Fig. 4b). The pattern of the 33 selected OGs at the phylum level largely corroborated that observed in the entire OG dataset, with a notable exception being the phylum Streptophyta, which exhibited a higher number of OG queries compared to Acidobacteria, Bacteroidetes, and Gemmatimonadetes (Fig. 4c). This observation suggests that one or more of the 33 OGs are predominantly associated with eukaryotic microorganisms. At the genus level, the relative abundance of the six aforementioned genera based on the 33 selected OGs closely mirrored that of the entire OG dataset, with the exception of Sphingomonas, which displayed a comparatively lower number of OG queries (Fig. 4d). The abundance and relative abundance profiles of microbial phyla illustrated in Fig. 5 and Table S9 generally align with the OG query numbers presented in Fig. 4c. However, a notable divergence was observed for the phylum Streptophyta, which demonstrated higher representation for the COG IDs L-COG2801, T-COG0515, and U-COG4886 compared to Actinobacteria and Proteobacteria. The absence of abundance and relative abundance data at the genus level for COG IDs P-ENOG410XNNV and U-COG4886 (Fig. 6) is attributed to their exclusion from the bacterial phyla Actinobacteria and Proteobacteria. Instead, P-ENOG410XNNV is predominantly associated with the phylum Bacteroidetes, while U-COG4886 is primarily found in the phylum Streptophyta (Fig. 5 and Table S8). At the genus level, the COG IDs Q-COG1020 and Q-COG3321 exhibit an almost exclusive association with the genus Nocardioides (Fig. 6 and Table S10).

Fig. 4.

The six most dominant phyla based on the total number of orthologous gene (OG) queries (a), and the six most prominent genera within the two predominant phyla, Actinobacteria and Proteobacteria (b), across the rhizospheric and bulk soil microbiomes of Moringa oleifera. Panels (c) and (d) depict the quantitative distribution of queries across the selected 35 orthologous groups (OGs) at the taxonomic levels of phylum and genus, respectively. Red columns denote the phylum Actinobacteria and its descending genera, while blue columns represent the phylum Proteobacteria and its constituent genera. The orange column corresponds to the eukaryotic phylum Streptophyta. Further information is available in Tables S8, S9, and S10

Fig. 5.

The abundance (a) and relative abundance (b) of the most dominant gene orthologs at the phylum taxonomic level across the rhizospheric and bulk soil microbiomes of Moringa oleifera. Further information is available in Table S9

Fig. 6.

The abundance (a) and relative abundance (b) of the most predominant gene orthologs at the genus level across the rhizospheric and bulk soil microbiomes of Moringa oleifera. The initial trio of genera are sourced from the phylum Actinobacteria, while the subsequent quartet of genera is derived from the phylum Proteobacteria. Further information is available in Table S10

Integrative functional analysis: eggNOG complementing KEGG, CAZy, and CARD databases

To comprehensively analyze the functional profile of the selected Orthologous Group (OG), we employed an integrative approach that leverages the complementary strengths of multiple databases. In this analysis, eggNOG (evolutionary gene genealogy Non-supervised Orthologous Groups) datasets play a crucial role in complementing and filling gaps in microbial functions identified by other specialized databases such as KEGG, CAZy, and CARD, ranging from broad metabolic roles to specific enzymatic activities and possible resistance mechanisms that might be overlooked when relying on a single database.

Biosynthesis of ATP

The results generated from eggNOG datasets regarding functional merits of the rhizospheric soil microbiome of M. oleifera complement those generated from the three databases KEGG, CAZy, and CARD in prior research. The prior results of KEGG and CAZy databases for this wild plant emphasized the enrichment of 17 enzymes contributing to the hydrolysis of bi- or polysaccharides in the "Sucrose and Starch metabolism" pathway towards the biosynthesis of glucose (Fig. 7). This action takes place when cells require energy in the form of ATP for its metabolic processes, where glucose serves as the fundamental precursor for the generation of NADH, which subsequently facilitates ATP production through the Krebs cycle. An extra avenue towards the biosynthesis of glucose is mediated by the orthologous gene BglX that encodes periplasmic beta-glucosidase, where cellobiose is the substrate (Fig. 8).

Fig. 7.

Enzymes exhibiting heightened abundance in the rhizospheric soil microbiome of Moringa oleifera, compared to the bulk soil microbiome, that were identified through KEGG and CAZy database analyses in two recent publications (Tashkandi et al. 2022; Tashkandi and Baz 2023). These enzymes collectively facilitate glucose biosynthesis via diverse routes of the "Sucrose and Starch metabolism" pathway, orchestrating the synthesis of glucose from the three complex carbohydrates starch, sucrose, and cellulose (in orange boxes). Notably, the cellulose-based route for glucose synthesis, with cellulose as a primary polysaccharide substrate, was detected in the KEGG database analysis, although this pathway was not explicitly mentioned or investigated in the corresponding article (Tashkandi et al. 2022). Note that glucose serves as the fundamental precursor for the generation of NADH, which subsequently facilitates ATP production through the Krebs cycle. EC:3.2.1.54 = cyclomaltodextrinase, EC:5.4.99.15 = (1- > 4)-alpha-D-glucan 1-alpha-D-glucosylmutase, EC:2.4.1.19 = cyclomaltodextrin glucanotransferase, EC:3.2.1.68 = isoamylase, EC:3.2.1.1 = alpha-amylase, EC:2.4.1.18 = 1,4-alpha-glucan branching enzyme, EC:2.4.1.25 = 4-alpha-glucanotransferase, EC:5.4.99.16 = maltose alpha-D-glucosyltransferase, EC:3.2.1.141 = 4-alpha-D-{(1- > 4)-alpha-D-glucano}trehalose trehalohydrolase, EC:3.2.1.20 = alpha-glucosidase, EC:3.2.1.10 = oligo-1,6-glucosidase, EC:2.4.1.64 = alpha,alpha-trehalose phosphorylase, EC:2.4.1.4 = amylosucrase, EC:2.7.1.2 = glucokinase, EC:3.2.1.4 = cellulase, EC:2.7.1.4 = fructokinase, EC:5.3.1.9 = glucose-6-phosphate isomerase

Fig. 8.

Highly abundant enzymes and proteins encoded by nine orthologous genes (OG) in the Moringa oleifera rhizospheric soil microbiome that facilitate glucose metabolism to promote electron transport and ATP biosynthesis of “Oxidative phosphorylation” pathway. The BglX gene-encoded periplasmic β-glucosidase provides an additional glucose synthesis route, complementing KEGG and CAZy enzyme pathways by hydrolyzing beta-glucosidic bonds in glucose-rich compounds like cellobiose. Other OG-encoded enzymes participate in electron transport chain complexes I, III, and V of this pathway. Complex I included five enriched NADH:ubiquinone oxidoreductase subunits (chains D, H, N, M, L) encoded by NuoD, NuoH, NuoM, NuoN, and NuoL OGs, with chain L functioning as Na + /H + antiporter (MnhA) and Na + /bicarbonate transporter (MpsA). Complex III included cytochrome b subunit that is encoded by QcrB/PetB OG, while Complex V included α subunit of FoF1-type ATP synthase that is encoded by AtpA OG. Generated ATP molecules can either be consumed by ATPase or stored via "Fatty acid biosynthesis" pathway. The latter pathway is initiated by Acetyl-CoA carboxylase (ACC) encoded by AccC OG, which catalyzes the reaction between excess ATP and acetyl-CoA with CO₂ to produce ADP, phosphorus, and malonyl-CoA

The analysis of highly abundant OGs in the rhizospheric soil of M. oleifera uncovered the role of enzymes and proteins of Complexes I, II, and V of “Oxidative phosphorylation” pathway in electron transport towards the biosynthesis of ATP (Fig. 8). Complex I included the enrichment of NADH:ubiquinone oxidoreductase subunits (chains D, H, N, M, L) encoded by five OGs, namely NuoD, NuoH, NuoM, NuoN, and NuoL, with chain L functioning as Na + /H + antiporter (MnhA) and Na + /bicarbonate transporter (MpsA). Complex III included cytochrome b subunit that is encoded by QcrB/PetB OG, while Complex V included α subunit of FoF1-type ATP synthase that is encoded by AtpA OG. Generated ATP molecules can either be consumed by ATPase or stored via "Fatty acid biosynthesis" pathway. The latter pathway is initiated by acetyl-CoA carboxylase (ACC) encoded by AccC OG, which catalyzes the reaction between excess ATP and acetyl-CoA with CO₂ to produce ADP, phosphorus, and malonyl-CoA.

ATP hydrolysis and the hierarchy of downstream consequences

The rhizospheric soil microbiome of Moringa oleifera exhibits a notable abundance of enzymes and proteins encoded by four orthologous genes (OGs), all of which are associated with ATP hydrolysis (Fig. 9). These four OGs encode for enzymes with diverse functional roles in cellular processes. The first of which is ClpP that encodes an ATP-dependent protease that plays a pivotal role in cellular proteostasis by facilitating the degradation of aberrant, damaged, or transient proteins. The second OG, e.g., EntF, encodes a non-ribosomal peptide synthetase that is instrumental in the biosynthesis of the enterobactin, a siderophore used by bacteria to acquire iron from the environment. The third OG, e.g., YopO, encodes a serine/threonine protein kinase functioning as an effector protein that confers advantages to bacterial pathogens. While the AtoC OG encodes a response regulator that is an integral component of a two-component system, partnering with the sensor kinase AtoS. These highly abundant ATP-hydrolyzing enzymes collectively contribute to the complex functional landscape of the M. oleifera rhizosphere microbiome, underscoring the importance of ATP-dependent processes in this unique ecological niche.

Fig. 9.

Highly abundant enzymes and proteins encoded by orthologous genes (OG) and associated with ATP hydrolysis in the rhizospheric soil microbiome of Moringa oleifera. The ClpP OG-encoded ATP-dependent protease plays a crucial role in maintaining cellular protein quality by degrading misfolded, damaged, or short-lived proteins. The enzyme collaborates with ClpX and ClpA ATPases to unfold, tag, and translocate proteins into its protease chamber for degradation. The EntF OG-encoded non-ribosomal peptide synthetase contributes to the siderophore enterobactin biosynthesis, and iron acquisition by catalyzing ATP-dependent serine activation. The resulting serine-AMP is transferred to the enzyme's peptidyl carrier protein (PCP) domain for incorporation into the growing peptide chain, facilitating the synthesis of natural products such as siderophores and antibiotics. The YopO OG-encoded serine/threonine protein kinase, an effector protein beneficial to bacterial pathogens, transfers phosphate groups from ATP to serine or threonine residues in host target proteins, modulating host cell signaling and promoting bacterial virulence. The AtoC OG-encoded response regulator, part of a two-component system with the sensor kinase AtoS, undergoes conformational changes upon phosphorylation of its REC Domain in response to specific environmental signals. Subsequently, the AAA-type ATPase domain of AtoC mediates cellular responses through sigma54-specific transcriptional regulation

Mode of action of other highly abundant OGs

Functional analysis of the other 20 highly abundant Orthologous Groups (OGs) revealed their involvement in eight distinct biological processes. Two orthologous antibiotic resistance genes, AcrB and MdlB, were identified as integral components of efflux pump systems, contributing to multidrug resistance mechanisms. SufI was found to be implicated in cell division processes, while RpsG, RplB, RpoB, and RpoC were elucidated to play crucial roles in transcriptional and translational machinery. PurR, BaeS, and AcrB were determined to be involved in various stress response pathways and regulatory cascades. CirA and CusA were characterized as mediators of iron acquisition and metal ion efflux, respectively. UbiH, PksD, and CypX were ascertained to be pivotal in the biosynthetic pathways of ubiquinone (coenzyme Q), polyketides, and pulcherriminic acid/pulcherrimin, respectively. YopM, SalY, and SPS1 were identified as modulators of host–pathogen interactions, with YopM exhibiting broad immunosuppressive properties, SalY loss-of-function enhancing host immunity, and SPS1 regulating key innate immune pathways in insects. Furthermore, CheY and BaeS genes were elucidated to be integral components of the two-component signal transduction systems CheA/CheY and BaeS/R, respectively. These sophisticated signaling networks of the latter two signal transduction systems enable bacteria to perceive and respond to environmental stimuli by transducing external cues into alterations in gene expression or protein activity.

Discussion

Comparative analysis of orthology-based and specialized functional databases

The Clusters of Orthologous Genes (COG) database represents a significant advancement in phylogenetic classification of proteins encoded in complete genomes (Tatusov et al. 2003). The eggNOG (evolutionary genealogy of genes: Non-supervised Orthologous Groups) is another orthology-based database that provides orthology relationships, gene evolutionary histories, and functional annotations across all domains of life. Both COG and eggNOG databases epitomize successive generations of orthology classification frameworks, each exhibiting distinctive characteristics and analytical capabilities (Huerta-Cepas et al. 2019; Galperin et al. 2024; Zdobnov et al. 2021; Trgovec-Greif et al. 2024). In the present investigation, we elected to employ the COG database due to its established reliability in prokaryotic genome analysis and its consistent historical framework for taxonomic classification (Carhuaricra-Huaman and Setubal 2024). The eggNOG-based annotation process has significantly enhanced our understanding of the functional diversity within the Moringa oleifera rhizosphere microbiome. By assigning COG identifiers to orthologous gene queries, this approach has revealed distinct functional enrichment patterns characteristic of the rhizosphere environment, and identified several highly prevalent functional categories, including energy production and conversion, amino acid and carbohydrate transport and metabolism, and replication, recombination and repair (López et al. 2023).

As COG database provides broad evolutionary context and functional classification across diverse organisms, other specialized databases such as KEGG (Kyoto Encyclopedia of Genes and Genomes), CAZy (Carbohydrate-Active enZYmes), and CARD (Comprehensive Antibiotic Resistance Database) offer in-depth, curated information on specific biological processes, enzyme families, and resistance mechanisms, respectively, complementing orthology-based databases in comprehensive functional annotation (Abulfaraj et al. 2024; Alshareef 2024; Huerta-Cepas et al. 2019; Zeller and Huson 2022). KEGG (Kyoto Encyclopedia of Genes and Genomes) serves as a comprehensive, pathway-oriented repository that establishes intricate connections between discrete genetic elements and their corresponding biological functions (Kanehisa et al. 2016). Unlike KEGG, COG is functionally oriented, but does not necessarily emphasize pathways, rather evolutionary conservation and function. COG and CAZy databases represent two synergistic bioinformatics platforms, each offering unique analytical capabilities (Miller 2024; Hobbs et al. 2023). However, the integration of these two databases provides researchers with robust frameworks for biological interpretation of genomic data, particularly in the context of complex carbohydrate metabolism and energy production (Galperin et al. 2024; Hobbs et al. 2023). Although they are two distinct functional annotation databases, COG and CARD can be integrated to provide a more comprehensive understanding of microbiome function in the context of antibiotic resistance (Zeller and Huson 2022; Alcock et al. 2020).

Orthology-driven insights into host-microbe interactions

The increased prevalence of a particular orthologous gene (OG) in the rhizospheric soil signifies that microbes within this zone are dynamically engaged in essential processes such as nutrient cycling, plant–microbe symbioses, energy production, and biodegradation, all of which are integral to plant growth and physiological well-being (Ling et al. 2022). The presence of these conserved genes in both host and microbial genomes facilitates a molecular dialogue that can foster mutually advantageous relationships (Starr et al. 2018). Symbiosis between hosts and their associated microbiomes is exemplified in diverse ecological contexts. In the human gastrointestinal tract, resident microorganisms play a crucial role in the catabolism of complex carbohydrates, yielding simpler molecular entities that are readily assimilable by the host for energy acquisition. Reciprocally, these microbial communities often harness host-derived metabolic byproducts or nutritional compounds to sustain their own bioenergetic processes (Tashkandi and Baz 2023). An analogous paradigm is observed in the rhizosphere, where the soil microbiome exhibits a sophisticated response to plant-derived signaling molecules. This microbial consortium engages in the degradation of complex carbohydrate structures, facilitating a mutually beneficial exchange of resources between the microorganisms and their plant host. This bidirectional metabolic interplay underscores the complex and interdependent nature of host-microbe relationships across different biological systems (Jamil et al. 2022).

Orthologous genes bridging core functions in rhizospheric microbiome of M. oleifera

Orthologous genes (OGs) serve as the evolutionary scaffolding for life's complex molecular networks, facilitating the integration of ancient metabolic pathways with newly evolved functions (Stamboulian et al. 2020; Glover et al. 2019). These conserved genetic elements act as crucial connectors in the intricate biochemical circuitry governing cellular processes across diverse organisms. By bridging primordial metabolic cascades with recent functional adaptations, OGs illuminate the dynamic interplay between evolutionary conservation and functional diversification in living systems.

The primary objective of the current investigation was to integrate the invaluable insights derived from the Clusters of Orthologous Genes (COG) database with previously published findings utilizing other databases, thereby providing a more comprehensive understanding of the functional capabilities and contributions of the rhizospheric soil microbiome in enhancing the growth of the wild plant Moringa oleifera. The initial cluster of Orthologous Groups (OGs) that aligns with the aforementioned assertion operates via the "Oxidative phosphorylation" pathway. This pathway embodies an intricate and refined process for energy generation and modulation within living cells, adapting to immediate metabolic demands (Wilson 2017). The process begins with glucose generation through two distinct routes: the hydrolysis of cellobiose by periplasmic β-glucosidase (Ngo et al. 2023), and through several KEGG/CAZy enzymes (Fig. 7) (Tashkandi et al. 2022; Tashkandi and Baz 2023). The periplasmic β-glucosidase, encoded by the OG BglX, serves as the initial catalyst in glucose generation through the hydrolytic cleavage of β-glucosidic linkages within cellobiose molecules. The assimilated glucose subsequently enters a series of catabolic pathways, including glycolysis and the tricarboxylic acid (TCA) cycle, culminating in the production of reduced nicotinamide adenine dinucleotide (NADH). This high-energy electron carrier serves as the primary substrate for the electron transport chain (ETC). The ETC, a sophisticated multi-component system embedded in the inner mitochondrial membrane, comprises five major protein complexes. Of particular interest are complexes I, III, and V, which incorporate multiple enzyme subunits encoded by distinct Orthologous Groups (OGs). These OG-encoded subunits exhibit a high degree of functional synergy, orchestrating a series of redox reactions that ultimately drive the synthesis of adenosine triphosphate (ATP) through chemiosmotic coupling (Wilson 2017). Complex I (NADH-ubiquinone oxidoreductase) contains chains D, H, M, and N, and includes specialized transporters—the Na + /H + antiporter (MnhA) and Na + /bicarbonate transporter (MpsA)—which maintain pH and acid–base balance (Demaurex and Grinstein 1994; Fan 2021). The enzymatic constituents of this complex are encoded by a specific set of Orthologous Groups (OGs), comprising NuoD, NuoH, NuoM, NuoN, and NuoL genes, respectively. Complex III features the cytochrome b subunit, encoded by QcrB/PetB, which facilitates electron transfer and helps establish the crucial proton gradient. Complex V (ATP synthase) utilizes the proton gradient established by previous complexes to synthesize ATP. Its α subunit, encoded by AtpA, is part of the F1 portion that catalyzes ATP synthesis from ADP and inorganic phosphate. When mitochondrial respiration is compromised, this complex can reverse its function, hydrolyzing ATP to maintain the proton gradient (Long et al. 2015). The AccC-encoded Acetyl-CoA carboxylase provides an alternative pathway for ATP utilization, converting excess ATP and acetyl-CoA to malonyl-CoA in the presence of CO₂. This reaction represents a key step in fatty acid biosynthesis, allowing energy storage when ATP production exceeds immediate cellular needs (Long et al. 2015; Wilson 2017). The efficiency of oxidative phosphorylation significantly surpasses that of glycolysis alone, producing approximately 30–32 ATP molecules compared to glycolysis's mere 2 ATP. This high-efficiency bioenergetic system has demonstrated its indispensability in the phylogenetic progression and perpetuation of complex multicellular organisms. The process represents a cornerstone of cellular metabolism, its significance underscored by the conservation of OGs encoding its key components across diverse taxa (Ghifari et al. 2023; Wilson 2017).

The rhizospheric soil microbiome of Moringa oleifera exhibits a notable abundance of enzymes and proteins encoded by four orthologous genes (OGs), all of which are associated with ATP hydrolysis (Fig. 9). These four OGs encode enzymes with diverse functional roles in cellular processes, where ClpP encodes an ATP-dependent protease that plays a pivotal role in cellular proteostasis by facilitating the degradation of aberrant, damaged, or transient proteins (Mabanglo and Houry 2022). This protease operates in concert with two ATPases, ClpX and ClpA, which are responsible for substrate recognition, unfolding, and translocation into the proteolytic chamber of ClpP for subsequent proteolysis (Xu et al. 2024). The second OG, designated as EntF, encodes a non-ribosomal peptide synthetase that is instrumental in the biosynthesis of the siderophore enterobactin (Miller et al. 2016; Miller and Gulick 2016; Ehmann et al. 2000). This enzyme catalyzes the ATP-dependent activation of serine, resulting in the formation of a serine-AMP intermediate (Miller and Gulick 2016). The third OG, identified as YopO, encodes a serine/threonine protein kinase functioning as an effector protein that confers advantages to bacterial pathogens (Singaravelu et al. 2017). This kinase catalyzes the transfer of phosphate groups from ATP to serine or threonine residues on host target proteins, thereby modulating host cell signaling cascades and promoting bacterial virulence (Lee et al. 2017). The forth OG, designated as AtoC, encodes a response regulator that is an integral component of the AtoS/AtoC two-component system (Kawamura et al. 2023). Upon detection of specific environmental stimuli, the receiver (REC) domain of AtoC undergoes phosphorylation-induced conformational changes5. Subsequently, the AAA-type ATPase domain of AtoC mediates cellular responses through sigma54-specific transcriptional regulation (Kawamura et al. 2023). These highly abundant ATP-hydrolyzing enzymes collectively contribute to the complex functional landscape of the M. oleifera rhizosphere microbiome, underscoring the importance of ATP-dependent processes in this unique ecological niche.

The functional interrogation of the 20 predominant OGs unveiled their involvement in eight discrete biological processes, underscoring the multifaceted and intricate functional architecture of the microbial consortium under scrutiny. Among these, two antibiotic resistance genes (ARGs), AcrB and MdlB, were discerned as pivotal constituents of efflux pump systems, playing a central role in the mechanisms underpinning multidrug resistance. AcrB is part of the resistance-nodulation-division (RND) antimicrobial resistance family (ARM), while MdlB belongs to the ATP-binding cassette (ABC) transporter family (Nishino et al. 2021). Previous investigations into the resistome of the rhizospheric soil microbiome associated with M. oleifera have revealed the presence of antibiotic resistance genes (ARGs) such as mtrA, soxR, and golS, which are involved in the metabolic processes of resistance-nodulation-division (RND) antibiotic efflux. In contrast, the oleC and novA genes are implicated in the metabolic pathways of ATP-binding cassette (ABC) antibiotic efflux. The identification of AcrB and MdlB as highly prevalent OGs within this microbiome signifies a substantial enhancement of our comprehension regarding its resistome. These two newly characterized ARGs offer potential functional redundancy and resilience to the antimicrobial resistance mechanisms within this microbial community. Their presence may serve as a safeguard against potential mutations in previously identified ARGs associated with RND (e.g., mtrA, soxR, and golS) and ABC (e.g., oleC and novA) efflux systems. The evolutionary conservation of these OGs suggests a higher degree of stability and a reduced susceptibility to mutational events, potentially conferring a more robust and adaptable resistance profile to the rhizospheric soil microbiome of M. oleifera.

Other OGs involve SufI, which was implicated in cellular division, specifically in the formation and regulation of the septal ring during bacterial cytokinesis (Gong et al. 2024). In parallel, the RpsG, RplB, RpoB, and RpoC genes were identified as critical players in both transcriptional and translational machineries, driving essential processes of protein biosynthesis and gene expression regulation (Roymondal et al. 2009). CirA and CusA genes were delineated as central mediators of iron sequestration and metal ion efflux, respectively (Chacón et al. 2014). CirA gene is essential for the uptake of iron under conditions of limited availability, while CusA gene is implicated in the expulsion of excess copper and silver ions, contributing critically to the maintenance of metal ion equilibrium. UbiH, PksD, and CypX genes were characterized as essential enzymes within the biosynthetic cascades governing the synthesis of ubiquinone (coenzyme Q), polyketides, and pulcherriminic acid/pulcherrimin, respectively (Lin and Qu 2022). These catalytic entities facilitate pivotal reactions in the generation of vital cellular constituents and secondary metabolites, some of which are endowed with potential antimicrobial properties. YopM, SalY, and SPS1 genes were identified as modulators of host–pathogen interactions, with YopM exhibiting potent immunosuppressive activities, SalY's loss-of-function enhancing host immune responses, and SPS1 modulating key innate immune pathways (Chung et al. 2016; Yoo et al. 2022). These genes are instrumental in bacterial virulence and the fine-tuning of host immune responses. Moreover, CheY and BaeS were delineated as integral components of the two-component signal transduction systems CheA/CheY (Porter et al. 2011) and BaeS/BaeR (Lin et al. 2014), respectively. These elaborate signaling networks empower bacteria to perceive and transduce environmental signals into modulations of gene expression or protein activity, thereby enabling rapid and dynamic adaptation to fluctuating ecological conditions.

Microbial taxa as key players in the rhizospheric microbiome of M. oleifera

The analysis of orthologous genes (OGs) in the microbiomes associated with M. oleifera reveals important insights into the functional potential of key microbial taxa. The predominance of Actinobacteria and Proteobacteria aligns with their known abundance and diverse metabolic capabilities in soil environments (Henke et al. 2020). At the genus level, several taxa stand out, where the Actinobacterial genus Blastococcus is known for its ability to thrive in oligotrophic environments (Hezbri et al. 2024). The high representation of OGs suggests it may play important roles in nutrient cycling and stress tolerance in the rhizosphere. The other Actinobacterial genus Nocardioides is notable for its diverse metabolic capabilities, including the degradation of complex organic compounds (Yan et al. 2022). The near-exclusive association with COG1020 and COG3321 (both involved in secondary metabolite biosynthesis) suggests Nocardioides may be a key producer of bioactive compounds in this ecosystem (Shtratnikova et al. 2021). The third Actinobacterial genus Streptomyces is a well-known for producing a wide array of secondary metabolites (Otani et al. 2022). The high representation of Streptomyces OGs indicates potential importance in plant–microbe interactions and soil health. Microvirga is a Proteobacterial genus that includes nitrogen-fixing symbionts (Shi et al. 2024). Its prevalence suggests a potential role in nitrogen cycling, even in association with non-leguminous plants. Though showing lower representation in the 33 selected OGs, Sphingomonas genus is known for its diverse metabolic capabilities and potential plant growth-promoting effects (Sorouri et al. 2023). The Proteobacterial genus Massilia has also been associated with plant growth promotion and disease suppression (Han et al. 2024). The higher representation of Streptophyta OGs for certain functions (e.g., L-COG2801 for DNA repair; T-COG0515 as a protein kinase; U-COG4886 as a vesicle transport) likely reflects the capture of plant root cells or endophytes in the samples, and highlights the importance of these processes in plant–microbe interactions (Guo et al. 2022). The exclusive association of P-ENOG410XNNV with Bacteroidetes and U-COG4886 with Streptophyta underscores the unique functional contributions of different taxonomic groups to the overall microbiome. This analysis provides valuable insights into the potential functional roles of key microbial taxa in the rhizosphere of Moringa oleifera, suggesting a complex network of interactions involving nutrient cycling, secondary metabolite production, and plant growth promotion.

Although this study might provide valuable insights into the rhizospheric microbiome of Moringa oleifera, particularly regarding energy production and regulation, several limitations exist. The metagenomic approaches may not capture the full microbial diversity due to biases in DNA extraction and sequencing (Dias et al. 2020). Furthermore, the study likely represents a temporal snapshot, potentially missing seasonal or developmental variations in the microbiome (Shami et al. 2022; Refai et al. 2023). While associations between microbial taxa and specific functions are identified, establishing causal relationships requires further experimental validation (Dias et al. 2020; Ma et al. 2023).

Conclusion

The exhaustive examination of orthologous genes (OGs) within the rhizospheric soil microbiome of Moringa oleifera yielded pivotal insights germane to sustainable agricultural practices. This meticulous investigation, employing a synergistic approach that integrated the COG database with KEGG, CAZy, and CARD, afforded a more nuanced and holistic perspective on the functional repertoire of the rhizosphere microbiome. The study elucidated the intricate interplay among diverse microbial taxa and their multifaceted roles in energy metabolism, nutrient cycling, and plant–microbe symbioses, thereby underscoring the paramount importance of fostering and maintaining heterogeneous, robust soil ecosystems. These findings serve as a cornerstone for the development of innovative agronomic strategies that harness the potential of beneficial microorganisms, and the implementation of precision agriculture techniques.

Supplementary information

Below is the link to the electronic supplementary material. 

Author contributions

Conceptualization, L.B., A.A.B., A.M.A., S.A.A., S.N.A., H.W.A., R.S.J.; methodology, F.O.S., A.A. Abulfaraj, F.M.A., A. Al-Andal, A.A.Alnahari, M.T., software, F.O.S., A.A. Abulfaraj, F.M.A., A. Al-Andal, A.A.Alnahari, M.T.; validation, L.B., A.A.B., A.M.A., S.A.A., S.N.A., H.W.A., R.S.J.; formal analysis, L.B., A.A.B., A.M.A., S.A.A., S.N.A., H.W.A., R.S.J., A.S.A; writing—original draft preparation, F.O.S., A.A. Abulfaraj, F.M.A., A. Al-Andal, A.A.Alnahari, M.T., L.B., A.A.B., A.M.A., S.A.A., S.N.A., H.W.A., R.S.J., A.S.A; writing—review and editing, F.O.S., A.A. Abulfaraj, F.M.A., A. Al-Andal, A.A.Alnahari, M.T., L.B., A.A.B., A.M.A., S.A.A., S.N.A., H.W.A., R.S.J., A.S.; visualization, L.B., A.A.B., A.M.A., S.A.A., S.N.A., H.W.A., R.S.J., A.S.A; project administration, A.S.A.

Funding

Princess Nourah bint Abdulrahman University Researchers Supporting Project number (PNURSP2025R357), Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia.

Data availability

The clean sequencing data were archived under Bioproject number PRJEB55112 in two preeminent genomic repositories: the European Nucleotide Archive (ENA) and the National Center for Biotechnology Information (NCBI). Within these repositories, the sequencing data have been assigned distinctive identifiers. The rhizosphere soil biosamples were allocated the run accession numbers ERR12764799, ERR10100771, and ERR10100772. In a similar vein, the bulk soil biosamples were designated the run accession numbers ERR12764800, ERR10100774, and ERR10100781.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.