Computational insights and insecticidal activity of black gram defensin proteins against bruchids (Callosobruchus maculatus)

Abstract

Defensins are small, cysteine-rich peptides involved in plant defense, though their insecticidal properties remain largely unexplored. Previously, based on transcriptome we identified a defensin gene in black gram in response to bruchid (Callosobruchus maculatus) infestation. In the present study, we cloned and sequenced full-length cDNAs of defensin genes from multiple legumes and conducted phylogenetic analyses. Two sequence variants were identified, exhibiting 95–98% homology with a previously reported insecticidal defensin gene (Accession no. AF326687). Variant 1 (DefV1) was present in black gram, pea, cowpea, and common bean, whereas variant 2 (DefV2) was identified in mung bean, chickpea, and pigeon pea. Computational analysis, including molecular docking, visualization, and molecular dynamics (MD) simulations, demonstrated enhanced interactions between DefV1 and bruchid α-amylase, suggesting a “Cork in the Bottle” inhibitory mechanism. Additionally, insect bioassays using artificial seeds supplemented with DefV1 showed no adult emergence. These findings highlight black gram defensin as a promising insecticidal agent and a potential candidate for genetic improvement of bruchid resistance in legumes.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-19446-0.

Introduction

Bruchids (Callosobruchus spp.) are among the most significant insect pests of legumes, recognized as highly destructive storage pests of pulses. These cosmopolitan insects exhibit a broad geographical distribution, spanning tropical Asia, Africa, and Central and South America. The Bruchidae family comprises over 1,700 species across 62 genera, although only 20 species are considered economically important¹. The initial infestation occurs in crop fields, where adult females oviposit on legume pods². Larvae hatch within 6–7 days post-oviposition³, penetrate the eggshell and pod wall, and reach the first available seed⁴. During the harvesting period, larvae feed on the superficial seed coat linings before entering the pupation phase inside the seed. This developmental strategy enables the emergence of young adults, which bore exit holes in stored seeds, facilitating secondary infestation in storage conditions¹.

Due to their multivoltine nature, bruchids contribute significantly to post-harvest losses. Seeds stored in warm and humid environments are particularly susceptible to infestation, underscoring the crucial role of temperature and humidity in regulating the bruchid life cycle⁵. In India, bruchids have gained considerable economic significance, as the country’s diverse climatic conditions are highly conducive to their proliferation. These storage pests cause substantial quantitative and qualitative losses in seeds through feeding and excretion, further predisposing infested seed lots to fungal contamination due to elevated humidity and temperature⁶.

The application of insecticides, particularly fumigants, is a commonly employed strategy for managing bruchid infestations in seed storage facilities⁷. However, this approach entails high costs, environmental pollution, and food safety concerns, posing risks to human health. To date, no cultivated legume has been reported to exhibit complete resistance against bruchids⁸. Nevertheless, certain wild-type accessions, such as Vigna mungo var. silvestris and Vigna radiata var. sublobata, as well as wild Cicer species (C. bijugum, C. judaicum, C. cuneatum, and C. microphyllum), have demonstrated partial resistance⁹. This resistance is primarily associated with factors such as high phenolic content, seed weight, a tough seed coat, and elevated α-amylase inhibitor activity within seeds. However, the precise molecular and biochemical mechanisms underlying bruchid resistance remain to be elucidated.

The resistance or susceptibility of plants to insect pests is primarily determined by the rapidity and efficacy of their defense responses against insect herbivory. Legume seeds are inherently rich in proteins, which can be broadly categorized into storage proteins and defense-related proteins¹⁰. The repertoire of defense-related proteins includes proteinase inhibitors, α-amylase inhibitors, glucanases, arcelins, vicilins, lectins, ribosome-inactivating proteins, allergens, lipid transfer proteins, chitinases, and cysteine-rich proteins, such as plant defensins¹¹.

Plant defensins are small, positively charged peptides comprising approximately 45–54 amino acids (∼5 kDa), characterized by the presence of eight conserved cysteine residues forming a distinct cysteine-stabilized αβ-motif¹². These peptides are encoded by a small multi-gene family and are widely distributed across the plant kingdom^13–16, with their presence reported in various tissues, although they are predominantly studied in seeds^13,17. One such plant defensin, VrCRP, identified in mung bean (V. radiata), has demonstrated 100% insecticidal efficacy against bruchid beetles under in vitro conditions¹⁸. The insecticidal activity of plant defensins is hypothesized to be associated with their proteinase inhibitor activity, translation inhibition, and α-amylase inhibition¹⁷; however, the precise molecular mechanism remains to be elucidated.

Several studies have reported that plant defensins, such as VrD1 and VuD1, exhibit α-amylase inhibitory activity against Coleopteran α-amylase enzymes located in the insect midgut. To date, reports on insecticidal plant defensins have been largely confined to mung bean. However, our ongoing research on black gram (V. mungo) has identified key defense-related genes, including defensins, pathogenesis-related (PR) proteins, and lipoxygenase genes, using suppression subtraction hybridization (SSH) techniques. Our findings revealed a significant upregulation of these genes in bruchid-infested black gram plants compared to uninfested controls¹⁹.

Furthermore, our investigations have expanded to examine the plant’s defensive response to bruchid oviposition. Previously, we reported that bruchid egg-laying triggers distinct transcriptional changes in black gram, priming the plant for subsequent larval attack²⁰. Transcriptomic analyses of immature black gram pods from two cultivars—IC-8219 (moderately resistant) and T-9 (susceptible)—identified differentially expressed genes (DEGs) associated with defense and signaling pathways, including defensins. Based on our transcriptomic data, we prioritized the upregulated defensin genes for further functional characterization²⁰.

Sequence analysis of these black gram defensin genes encountered challenges related to the signal peptide region, which exhibited 97% sequence similarity to the previously reported V. radiata PDF1 gene (Accession No. LN901492.1; Hoang et al., unpublished). Notably, in the present study, we identified a novel black gram defensin gene sharing 95% sequence similarity with the insecticidal defensin gene from mung bean (Accession No. AF326687)¹⁸. Expression profiling of this defensin gene was conducted across various legume species, and full-length cDNA sequences were obtained. Phylogenetic analysis and multiple sequence alignment revealed two distinct variants of the defensin gene: variant 1 (DefV1) and variant 2 (DefV2), with black gram serving as the representative legume for DefV1 and mung bean for DefV2.

Additionally, the interaction mechanism between black gram defensin and the α-amylase enzyme of C. maculatus was investigated using computational approaches, including three-dimensional (3D) modeling, molecular docking, and molecular dynamics simulations. Based on these analyses, we propose a “cork in the bottle” hypothesis, which describes the mode of inhibition of insect α-amylase by plant defensins.

Materials and methods

Plant material and insect rearing

For the present study, seeds of diverse legume germplasms were utilized. These included V. mungo (Black gram: IC-8219, SBC-40, and SBC-47), V. radiata (Mung bean: SGC-16 and SGC-20), Pisum sativum (Pea: AMAN), Cajanus cajan (Pigeon pea: ICPL-332 and ICPL-87), and Cicer arietinum (Chickpea: ILWC-46 and DCP). Black gram and chickpea germplasms were obtained from the Indian Institute of Pulse Research (IIPR), Kanpur, whereas mung bean germplasms were supplied by the Regional Research Station for Pulses, Assam Agricultural University (AAU), Shillongani. Pigeon pea seeds were procured from the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Hyderabad. In addition, seeds of cowpea (Vigna unguiculata) and common bean (Phaseolus vulgaris) were collected from the local market in Jorhat, Assam. All seeds were sown in the greenhouse facility of the DBT-North East Centre for Agricultural Biotechnology (DBT-NECAB), AAU, Jorhat, and plants were maintained under controlled environmental conditions.

The primary culture of C. maculatus was sourced from the Department of Entomology, AAU, Jorhat. Beetles were reared on mung bean seeds (V. radiata) in a growth chamber maintained at 28 ± 2 °C, with a 14:10 h light-dark photoperiod and relative humidity of 65–70%. Growth chambers were sealed to ensure a controlled environment, allowing mating and oviposition to occur. After oviposition, parental insects were removed, and the mung bean seeds containing eggs were transferred to fresh seeds in two-litre plastic containers. These containers were covered with cloth secured by rubber bands to prevent contamination and escape of beetles^21–23.

For infestation assays, ten pairs of male and female adult bruchids were released onto developing pods of each legume species at 15 days after flowering. Sexual dimorphism in C. maculatus was distinguished under a stereomicroscope (40×) based on pygidial morphology: females exhibited a larger, darkly pigmented pygidium with two black stripes and a white transverse line, whereas males possessed a smaller pygidium lacking such markings²³. To prevent insect dispersal, infested pods were enclosed in muslin cloth bags. Immature pod samples were collected at seven days after oviposition (DAO) from both infested and uninfested plants. All harvested tissues were immediately frozen in liquid nitrogen and stored at − 80 °C until further analyses.

cDNA synthesis, PCR amplification and DNA sequencing

Total RNA was extracted from developing seed samples and from pod wall tissues (after seed removal) of both oviposited (treated) and non-oviposited (control) pods using the Plant RNA Purification Reagent (Invitrogen), following the manufacturer’s protocol, with three biological replicates. The corresponding cDNA was synthesized using the PrimeScript™ RT reagent Kit with gDNA Eraser (Clontech, USA). Primers were designed from the defensin sequence (Accession No: AF326687.2) using “PrimerQuest™ Tool” (IDT) software keeping all the parameters as default¹⁸. The primer sequences used for PCR amplification were: 5’ATGGAGAGAAAAACTTTCAGC3’ (Forward primer) and 5’TCAACAGTTGACGAGGCAATAGC3’ (Reverse primer). The PCR amplification was performed with 100 ng of cDNA in a 20 µL volume of reaction mixture having 5 pmol/µL of each primer pair, following the reaction conditions with denaturation at 95 °C for 30 s, annealing at 57 °C for 20 s and extension at 72 °C for 1:30 min for 35 cycles. PCR fragments were cloned in pGEM-T easy vector System I (Promega). The white recombinant clones were selected, and recombinant plasmid DNA was isolated using a mini-prep plasmid isolation kit (QIAGEN). The ligation of the insert was verified by restriction digestion with EcoRI enzyme and Sanger di-deoxy sequencing was done using M13 universal primers at Bioserve Company, Hyderabad, Telangana, India.

Quantitative PCR analysis for defensin gene expression

Quantitative amplification of 1 µg of RNA of seed and pod wall samples was performed by using the “PowerUp™ SYBR™ Green Master Mix” (Applied Biosystems™) following the manufacturer’s protocol on QuantStudio5 (Applied Biosystems™) platform. The temperature profile used was UDG activation at 50 °C for 2 min, an initial denaturation at 95 °C for 2 min, and 40 cycles were performed by three steps at 95 °C for 15 s, 57 °C for 15 s and 72 °C for 1 min followed by a melt curve stage at 60 °C for 1 min. The 2^−ΔΔCT method was used to deduce the relative quantification (RQ) value of each sample based on normalization with the reference genes of black gram (18 S), cowpea (Actin), mung bean (18 S), common bean (bZip), pea (Tubulin) and pigeon pea (Tubulin) and chickpea (GAPDH). The mean RQ values were used for qPCR data analysis. Quantitative PCR (qPCR) analysis was performed using three biological replicates, each consisting of three technical replicates. Statistical significance of the differences in defensin gene expression among the various legume species and tissue types examined in this study was evaluated using Student’s t-test, with a significance threshold set at a 95% confidence level (p < 0.05). The primers utilized for quantitative PCR (qPCR) analysis in this study are detailed in Supplementary Table S1.

Analysis of the sequences and phylogeny using maximum-likelihood (ML) algorithm

Sequence data was analyzed using Technelysium Chromas software Version 2.6.6. Contigs were prepared by aligning the forward and reverse sequences using CLUSTAL OMEGA (EMBL-EBI web-based software). The contigs of each sample were then locally aligned using Nucleotide BLAST (NCBI) against the insecticidal plant defensin sequence, VrCRP (Accession No: AF326687.2). Multiple sequence alignment of the nucleotide sequences was performed accordingly using MEGA 11.0²⁴ and the corresponding amino acid sequences were retrieved and accordingly aligned. The phylogenetic analysis was executed using MEGA 11.0 and using the Maximum-Likelihood (ML) Algorithm, a tree was prepared with 1000 replicates for bootstrapping and the evolutionary distances were calculated by the p-distance method.

Functional annotation of amino acid sequences, disulfide bond prediction and homology modelling of defensin protein

The primary functional annotation of amino acid sequences of DefV1 and DefV2 was carried out in InterPro-EMBL-EBI and subsequent disulfide-bond formations were predicted using the DISULFIND program at the Predict Protein server (https://predictprotein.org/). Furthermore, the 3-D structure of mature defensin peptides DefV1 and DefV2 (without signal) was prepared using the software SWISS-MODEL from Biozentrum^25,26.

Computational analysis of the interaction between α-amylase from C. maculatus and DefV1 and DefV2 proteins

The interaction between the α-amylase enzyme from C. maculatus and the mature defensin protein (DefV1 and DefV2) were investigated through bioinformatics analyses.

Sequence retrieval and 3-D homology modeling of α-amylase

The α-amylase protein of C. maculatus was retrieved from the RCSB Protein Data Bank and UniProtKB database (Entry No. A0A653D1J8). The retrieved PDB file has been inputted into the GASS (Genetic Active Site Search) tool using the CSA (Catalytic Site Atlas) database along with INTERPRO EMBL-EBI to predict the probable active sites of the protein. The 3-D structure was prepared using homology modeling in the SWISS-MODEL software. The 3-D model was downloaded in PDB format for further structural analysis using UCSF Chimera software version 1.16²⁷.

Molecular Docking and validation of the model

The model of the α-amylase and the mature DefV1, DefV2 and VrCRP protein was prepared for protein-protein blind docking using the ClusPro^28–31. In the user interface of the ClusPro, the PDB files of the receptor (α-amylase protein of C. maculatus) and ligand-protein (DefV1/DefV2/VrCRP) were uploaded, and the blind-dock was performed keeping all the parameters as default. The docked structure was validated via web-based software, including the ProSA-Web^32,33 for assessing Z-score plots based on NMR and X-ray Crystallography data points. Further structural validation was conducted using SAVES v6.0, a tool developed by the UCLA-DOE lab, involving Ramachandran Plot via PROCHECK^34,35 overall quality factor and plot analysis via ERRAT³⁶ and visualization of 3D-1D scores for amino acid residues provided by VERIFY 3D^37,38. The docked structures were further visualized in UCSF Chimera v1.16 for the superimposition of the models along with the analysis of the Hydrogen bonding between the proteins and conformational changes of the models.

Molecular dynamic simulation of the docked structures

The docked model with the optimum score was taken into consideration for MD simulation using the WebGRO for Macromolecular Simulations (https://simlab.uams.edu/) using the GROMACS tool³⁹. The structures were submitted in PDB format, and the simulation was carried out with the TIP4P water molecule. The forcefield used was GROMOS96 43a1 using Triclinic box type. Energy minimization was done using the Steepest Descent integrator over 5000 steps. Subsequently, a run time of 50 ns was carried out at 300 K temperature with 5000 frames per simulation conducted using the processor Intel Xeon. The default parameters were retained, and RMSF and RMSD plots were retrieved and analyzed.

α-amylase enzyme Inhibition activity of DefV1 and DefV2 protein

The inhibitory efficacy of α-amylase was assessed by following the method described by Bernfield and co-workers⁴⁰. An amylase inhibitor assay was performed by combining 250 µL of total protein extract from E. coli (strain ER2256) bacterial culture, E. coli culture carrying the pTXB1 vector, and purified DefV1 and DefV2 peptides expressed as a recombinant protein in pTXB1 vector (at concentrations of 0.3 mg/mL) with 250 µL of α-amylase solution, prepared in 20 mM phosphate buffer at pH 6.9. Following a pre-incubation period of 30 min at 37 °C, 250 µL of substrate solution (1% soluble starch) was added, and the reaction was allowed to proceed for 5 min. Termination of the reaction was achieved by the addition of 500 µL of 3,5-dinitrosalicylic acid reagent, followed by a 5-minute boiling step in a water bath. Subsequently, 5 mL of water was introduced, and the mixture was allowed to cool down at room temperature for 15 min. The absorbance was measured at 545 nm. A control sample was also made by replacing the total protein extract with water. The inhibitory activity of α-amylase was calculated as a percentage based on the following formula: Inhibition Percentage = ((Absorbance of Control - Absorbance of Sample)/Absorbance of Control)) x 100. The experiments were carried out using ten biological replicates and five technical replicates. The statistical significance of the inhibition of α-amylase activity by DefV1 and DefV2 was evaluated using Student’s t-test at a 95% confidence level (p < 0.05).

Insect bioassay

Artificial seeds were prepared from chickpea flour following a previous report⁴¹ and subjected to bioassays with C. maculatus. Artificial seeds containing varying concentrations of purified DefV1 and DefV2 (0.1–0.3%, w/w) were prepared for the assay. A total of 45 artificial seeds were used for bruchid infestation assay across all treatments, which included the control (chickpea flour [CF] with sterile distilled water and CF with column buffer), CF + DefV1, and CF + DefV2. The seeds were divided into three replicates, with each replicate containing 15 seeds. These were placed in glass jars that measure 7.5 cm in diameter and 10 cm in height. Ten pairs of C. maculatus, containing an equal number of male and female bruchids, were introduced into each jar to facilitate oviposition. The bruchids were allowed to Lay eggs on the seeds for 24 h. The infested seeds were then kept at 25 °C and 60% relative humidity (RH), and their within-seed development time (WSDT) and percentage adult emergence were periodically monitored.

Results

Differential expression analysis of defensin gene from different legumes

Quantitative PCR (qPCR) analysis was conducted to assess the expression of the defensin gene across various legume species. The results from qPCR analysis of treated seed samples revealed an upregulation of defensin expression in common bean (P. vulgaris), cowpea (V. unguiculata), and black gram (V. mungo), whereas downregulation was observed in pea (P. sativum), mung bean (V. radiata), pigeon pea (C. cajan), and chickpea (C. arietinum) (Fig. 1A). Among the analyzed legumes, common bean exhibited the highest upregulation (5.5-fold), followed by black gram with 2.2-fold (SBC 40) and 1.2-fold (IC-8219), and cowpea with a 1.1-fold increase (Log₂ of mean relative quantification, RQ). Conversely, downregulation was observed in mung bean, with a 1.0-fold (SGC 16) and 1.5-fold (SGC 20) decrease, followed by pigeon pea with a reduction of 1.9-fold (ICPL-87) and 2.5-fold (ICPL-332), and chickpea with 3.2-fold (ILWC 46) and 0.8-fold (DCP) downregulation.

Fig. 1.

Quantitative real-time PCR (qRT-PCR) analysis of defensin gene expression in (A) seed samples and (B) pod wall samples of common bean, pea, cowpea, black gram (Varieties: IC-8219 and SBC-40), mung bean (Varieties: SGC-16 and SGC-20), pigeon pea (Varieties: ICPL-332 and ICPL-87), and chickpea (Varieties: ILWC-46 and DCP). The y-axis represents the log₂ fold change in gene expression, while the x-axis denotes the treated (T) conditions. Gene expression levels in the treated group were normalized against the respective control group. Bars represent the variation in mean expression levels across three biological replicates. Statistical significance was determined using Student’s t-test at a 95% confidence level.

Notably, the extent of defensin gene expression varied across different legumes. The highest upregulation was observed in common bean (5.5-fold increase compared to control), while the greatest downregulation was detected in chickpea (ILWC 46), which exhibited a 3.2-fold reduction in expression.

Additionally, defensin gene expression was analyzed in the pod wall samples of the same legume species (Fig. 1B). The results indicated upregulation in common bean (3.0-fold), black gram with 1.1-fold (SBC 40) and 0.6-fold (IC-8219), and pigeon pea with 1.4-fold (ICPL-87) and 1.0-fold (ICPL-332). In contrast, downregulation was observed in pea (0.6-fold), cowpea (0.5-fold), chickpea (DCP: 1.9-fold, ILWC 46: 1.7-fold), and mung bean (SGC 20: 1.4-fold, SGC 16: 2.3-fold). Among the pod wall samples, the highest upregulation was recorded in common bean, whereas the most significant downregulation was observed in mung bean (SGC 16). The raw qPCR data along with the corresponding statistical analyses are presented in Supplementary Table S2.

Cloning of cDNA encoded defensin from different legumes

A full-length cDNA encoding defensin gene was isolated from different legumes. The PCR amplification suggested the presence of the defensin gene in all the legumes. However, a difference in the intensity of the band was observed in different samples (Supplementary Fig. S1A and B). The purified PCR product from infested samples of each legume was cloned in pGEM-T easy vector and Sanger sequenced for further analysis. The sequence was also submitted to NCBI (Accession No: PP334100.1).

Sequence analysis, functional annotation of defensin variants in legumes and phylogenetic relationships

The nucleotide sequence of the defensin gene responsive to bruchid infection was aligned using the NCBI Nucleotide BLAST tool, with the VrCRP sequence (GenBank Accession No: AF326687) as a reference. The sequence analysis revealed high identity with various legumes, including 98% identity with mung bean (Query Coverage: 56%, E-value: 6e-115), chickpea (Query Coverage: 61%, E-value: 5e-115), and pigeon pea (Query Coverage: 61%, E-value: 5e-115). Additionally, the sequence showed 95% identity with common bean (Query Coverage: 61%, E-value: 5e-105) and black gram (Query Coverage: 63%, E-value: 2e-103), while pea (Query Coverage: 63%, E-value: 2e-103) and cowpea (Query Coverage: 54%, E-value: 6e-100) exhibited 94.62% identity. Multiple sequence alignment (MSA) of the nucleotide sequences indicated high homology with VrCRP, with minor nucleotide variations (Fig. 2A).

Fig. 2.

Multiple sequence alignment and disulfide bridge pattern analysis of legume seed samples and defensin proteins. (A) Nucleotide sequence alignment of common bean, pea, cowpea, black gram, mung bean, pigeon pea, chickpea, and VrCRP using the Clustal W algorithm in MEGA 11.0. ^*Represents conserved nucleotide sequences. (B) Amino acid sequence alignment of seed samples from all legumes using Clustal W in MEGA 11.0. ^*Represents conserved amino acid sequences. (C) Amino acid sequence alignment of DefV1, DefV2, and VrCRP using Clustal W in MEGA 11.0. Differences in amino acid residues, along with their respective positions, are indicated within parentheses. The first 27 amino acids correspond to the signal peptide, while residues 28–73 represent the mature defensin protein. The cleavage site between the 27th and 28th residues is marked by an arrow. (D) Disulfide bridge pattern visualization of DefV1, DefV2, and VrCRP sequences using the DISULFIND program at the Predict Protein server. In the mature peptide, 8 cysteine residues (numbered C1–C8) are linked in the format C1–C8, C2–C5, C3–C6, C4–C7.

Translation of these sequences into amino acids revealed seven substitution sites. Three substitutions—L9F, S11L, and R24K—were located in the signal region, whereas four—K39R, I42M, K51R, and D59N—were present in the mature peptide (Fig. 2B). These variations led to the classification of two defensin gene variants: Defensin Variant 1 (DefV1) and Defensin Variant 2 (DefV2). DefV1 was identified in black gram, pea, cowpea, and common bean, whereas mung bean, chickpea, and pigeon pea defensin proteins were classified as DefV2. The amino acid sequences of VrCRP and DefV2 were nearly identical, except for a single amino acid substitution at position 59, where Aspartic acid (D) in VrCRP was replaced by Asparagine (N) in DefV2. In contrast, DefV1 exhibited a total of seven amino acid substitutions compared to VrCRP. Notably, both DefV1 and DefV2 retained Asparagine at the 59th position, whereas VrCRP contained Aspartic acid (Fig. 2C).

Functional annotation of the amino acid sequences using the InterPro-EMBL-EBI database indicated the presence of a conserved defensin domain. All three sequences—DefV1, DefV2, and VrCRP—possessed an N-terminal 21-amino-acid-long transmembrane signal and a 45-amino-acid-long non-cytoplasmic defensin domain (Supplementary Fig. S2). Characteristically, defensins contain eight conserved cysteine residues involved in disulfide bridge formation. The disulfide bonding pattern in both defensin variants closely resembled VrCRP, suggesting that the functional domain of these plant defensins resides primarily within the non-cytoplasmic region (Fig. 2D).

Phylogenetic analysis revealed two distinct clades (Fig. 3). Black gram, pea, cowpea, and common bean formed one group, with pea and cowpea exhibiting the closest relationship (minimum distance: 0.01), while black gram and common bean were their nearest neighbors. Mung bean, chickpea, and pigeon pea clustered into a separate group, with pigeon pea and mung bean displaying the closest relationship, whereas chickpea was their closest neighbor. Notably, VrCRP was phylogenetically distinct from both defensin variants, further supporting its divergence.

Fig. 3.

Phylogenetic analysis of defensin nucleotide sequences from various legumes using the Neighbor-Joining algorithm with the p-distance method to estimate evolutionary distances (1,000 bootstrap replicates). The analysis reveals two distinct clades: one comprising black gram, common bean, pea, and cowpea, and another including mung bean, pigeon pea, and chickpea, indicating their evolutionary divergence from a common ancestor. The sequence of VrCRP is positioned as an outgroup.

To validate these findings, a comparative sequence analysis was conducted using BLASTn, wherein one representative nucleotide sequence from each defensin variant and the VrCRP sequence was aligned against the non-redundant nucleotide database (nr/nt). The sequences of the two defensin variants exhibited the highest similarity to the plant defensin-like protein of V. nakashimae (Accession No.: AY856095.1), with 99.5% identity, 63% query coverage, and an E-value of 3e-108. The most homologous sequences were subsequently used for multiple sequence alignment in MEGA 11.0, followed by phylogenetic analysis using the Neighbor-Joining algorithm. The resulting phylogenetic tree corroborated the formation of the two major groups observed earlier, both closely related to the defensin-like protein of V. nakashimae (Accession No.: AY856095.1), while the plant defensin D1 gene of V. radiata (Accession Nos.: AY437639.1, FJ591131.1) was identified as the next closest neighbor, sharing a common ancestor (Supplementary Fig. S3A and B). Additionally, the phylogenetic tree reaffirmed the distinct clustering of VrCRP, further supported by the outgrouping of ncRNA of V. angularis (Accession No.: XR001833220.2) and defensin-like protein 4 mRNA of V. umbellata (Accession No.: XM047306565.1). These findings confirm that the identified defensin variants are significantly different from the previously characterized defensin protein by Chen and his co-workers¹⁸, highlighting the novelty of the sequences reported in this study.

Computational analysis and molecular Docking of insect α-amylase and plant defensin interaction

The interaction between insect α-amylase and plant defensin was investigated through a comprehensive computational approach, encompassing sequence analysis, homology modeling, active site prediction, and molecular docking. The α-amylase sequence of C. maculatus was characterized by an N-terminal signal peptide (residues 1–16) and a mature peptide (residues 17–491). InterPro database annotation identified two domains within the mature peptide: a glycosyl hydrolase family 13 catalytic domain (residues 27–391) and an α-amylase C-terminal domain (residues 400–486). A 3D model of the α-amylase was generated using homology modeling, exhibiting 98.17% sequence identity and 100% query coverage with the template A0A168VCK5.1, and a reliable Global Model Quality Estimation (GMQE) score of 0.95. Similarly, structures for the mature peptide chains of DefV1 and DefV2 were modeled, showing high sequence similarity (93.48% and 100%, respectively) with their templates (1TI5.1) and GMQE scores of 0.88 and 0.87, respectively (Fig. 4A).

Fig. 4.

Cartoon representation of the 3-D models of DefV1, DefV2, and the α-amylase of Callosobruchus maculatus. (A) The structures of DefV1 (tan) and DefV2 (blue) show distinct separation of β-sheets and an integrated 31,010​-α-helix, interconnected by loops. The α-amylase model, comprising 491 amino acids, features a structural trench, which may serve as the probable site for amylose digestion. (B) The probable active site residues of α-amylase are depicted in ball-and-stick models: Aspartate 302 and 110 (pink, with negative heteroatoms in red), Glutamate 237 (purple, with a negative heteroatomic charge in red), and Histidine 301 (green, with a positively charged residue in blue). These residues are closely positioned within the enzyme’s structural trench, suggesting their role in catalysis. (C) Interaction between α-amylase and the defensin peptide chain, highlighting Arginine 1 and Threonine 2 (blue ball-and-stick model) positioned proximally within the structural trench, reinforcing the likelihood of active interaction between the two molecules.

The active site of C. maculatus α-amylase was predicted using InterPro software, which identified the catalytic domain of 1,4-α-D-glucan-4-glucanohydrolase (residues 26–394, CDD entry: cd11317). Key residues involved in the active site, calcium-binding, and catalytic activity, such as H301, E237, D110, and D302, were annotated and validated using GASS-WEB. UCSF Chimera visualization confirmed the proximity of these residues, suggesting the presence of an active site (Fig. 4B). Blind protein docking was subsequently performed to explore the interaction between α-amylase and defensin peptides (DefV1 and DefV2).

Molecular docking was conducted using ClusPro, which generated clustered and minimized models. The “Best model” selected based on balanced coefficients (E = 0.40Erep + − 0.40Eatt + 600Eelec + 1.00EDARS)³², consisted of 192 interacting members with a weighted score of −715.6 for the center representative and − 825.2 for the lowest energy. Visualization in UCSF Chimera revealed that defensin peptides penetrated a structural trench in the α-amylase, resembling a pathway to the core of the molecule. An amylose molecule, imported from MolView (https://molview.org/) in mol2 format, was found to be fully situated within the central region of the α-amylase core, further validating the interaction. Residues implicated in the putative active site (H301, E237, D110, and D302) were localized within the trench, while defensin peptide residues (R1 and T2) were observed in close proximity to the active site in the docked structure (Fig. 4C). These findings provide insights into the molecular mechanisms underlying the interaction between insect α-amylase and plant defensin.

Validation and structural assessment of docked defensin-α-amylase complexes

The docked structures of DefV1 and DefV2 with α-amylase, designated as D1AA (DefV1-α-amylase) and D2AA (DefV2-α-amylase), were rigorously validated using ProSA-Web, PROCHECK, ERRAT, and VERIFY-3D. The ProSA-Web analysis provided Z-scores and residue-energy plots, where the average energies over 40-residue fragments were represented as a thick Line, with a smaller window size of 10 residues for reference (Fig. 5A, B, C, D and E, and 5F). The docked structures, comprising 537 amino acid residues, yielded Z-scores of −8.7 for D1AA and − 8.88 for D2AA, closely aligning with the native templates. These scores positioned the structures within the X-ray crystallographic (faded blue) and NMR (dark blue) regions of the ProSA-Web plot, indicating satisfactory conformational similarity to native structures.

Fig. 5.

Structural Validation and Energy Profiling of Docked α-Amylase Complexes. (A-C) Z-Score plots of α-amylase (A) and docked structures D1AA (B) and D2AA (C), revealing Z-Scores of −8.48, −8.70, and − 8.88, respectively. The analysis positioned these structures within the X-ray crystallographic (faded blue) and NMR (dark blue) regions, indicating satisfactory conformational similarity to native structures. (D-F) Residue-energy-based plots of α-amylase (D) and docked structures D1AA (E) and D2AA (F), highlighting notable energy curves. At window 40, the dark green Line indicated significant residue-energy patterns, while the faint green Line at window 10 provided additional insights. Comparisons with native structures confirmed favorable energy profiles.

Further validation using the UCLA-DOE SAVESv6.0 server, including ERRAT analysis, demonstrated reliable structural quality. The overall quality factors were 89.981 for D1AA and 88.847 for D2AA, confirming the structural integrity of the docked complexes (Supplementary Fig. S4A and B). VERIFY-3D analysis revealed high compatibility between the atomic models and their respective amino acid sequences, with 89.01% of residues in D1AA and 86.59% in D2AA scoring ≥ 0.1 in the 3D-1D profile (Supplementary Fig. S5A and B). These results indicate that the majority of residues in both complexes are consistent with their expected structural environments.

Ramachandran plot analysis, performed using PROCHECK, provided insights into the φ and ψ angle distributions of the amino acid residues in D1AA and D2AA (Supplementary Table S3). The analysis revealed that 87.4% of residues in D1AA and 87.9% in D2AA occupied the most favorable regions of the Ramachandran plot, with an additional 11.7% and 11.0% in favorable regions, respectively. These results underscore the high structural quality and integrity of the docked models (Supplementary Fig. S6A and B).

In summary, the comprehensive validation analyses, including Z-score evaluation, ERRAT, VERIFY-3D, and Ramachandran plot assessments, collectively affirm the reliability and structural robustness of the D1AA and D2AA complexes. The docked structures exhibit favorable energy profiles, high residue compatibility, and excellent conformational stability, making them suitable for further functional and mechanistic studies.

Molecular dynamics simulation of docked structures using Web-GRO

In this study, the docked complexes were subjected to molecular dynamics simulations using the GROMACS tool via the Web-GRO server developed by UAMS. Analysis of the root mean square deviation (RMSD) and root mean square fluctuation (RMSF) provided insights into the structural stability of the complexes. For the D1AA complex, the RMSD exhibited an initial deviation ranging from 0.16 to 0.42 nm within the first 5 ns, reaching approximately 0.55 nm by 15 ns. Beyond this point, the RMSD plot remained relatively constant, with no significant peaks observed between 30 and 50 ns, maintaining a deviation of approximately 0.5 nm (Fig. 6A). The RMSF analysis of D1AA indicated overall stability, with fluctuations remaining below the 0.5 nm threshold for most residues. However, residues 490–501 demonstrated fluctuations at or slightly above 0.5 nm, suggesting potential conformational changes at the initiation of the DefV1 peptide sequence following the α-amylase chain in the 537-residue complex (Fig. 6B).

Fig. 6.

Root Mean Square Deviation (RMSD) and Root Mean Square Fluctuation (RMSF) plots for D1AA and D2AA, generated using GROMACS via WebGRO, with simulations run for 50 ns in TIP4P water at 300 K. (A) The RMSD plot for D1AA shows initial deviations from 0.16 to 0.42 nm within the first 5 ns, increasing to 0.55 nm by 15 ns, followed by stabilization around 0.5 nm from 30 to 50 ns. (B) The RMSF plot indicates overall stability, with fluctuations remaining below 0.5 nm, except for residues 490–501, which exhibit higher fluctuations, suggesting potential interactions with DefV1 within the 537-residue docked complex. (C) The RMSD plot for D2AA reveals deviations from 0.15 to 0.45 nm in the first 10 ns, stabilization between 15–30 ns, and a significant increase from 0.4 to 0.55 nm at 35–42 ns before decreasing. (D) The RMSF plot for D2AA presents a distinct fluctuation pattern, with variations from 0.1 to 0.4 nm, reaching up to 0.7 nm within the first 200 residues. Additionally, two peaks of 0.7 nm appear between residues 350–400, with a sharp increase at residues 490–501.

In contrast, the D2AA complex showed an initial RMSD increase from 0.15 to approximately 0.45 nm within the first 10 ns. The complex then achieved temporary stabilization with minor deviations between 15 and 30 ns; however, a notable increase from 0.4 nm to 0.55 nm occurred around 35 ns and persisted until approximately 42 ns, after which the deviation decreased (Fig. 6C). The RMSF profile for D2AA was distinctly different from that of D1AA, exhibiting major fluctuations between 0.1 and 0.7 nm. In the first 200 residues, fluctuations varied from 0.1 to 0.7 nm, with two prominent peaks at approximately 0.7 nm observed between residues 350 and 400, and another sharp peak at residues 490–501 (Fig. 6D).

Collectively, the molecular dynamics simulations revealed a notable difference in the stability of the docked complexes, with D1AA demonstrating superior stability compared to D2AA. The RMSD analysis supports the enhanced stability of the D1AA complex, while the RMSF analysis further corroborates the reduced stability of D2AA, potentially indicating a lower inhibitory potential of DefV2 against C. maculatus α-amylase. These results suggest that the stable conformation of D1AA underlies its higher inhibitory efficacy.

Furthermore, superimposed visualizations generated using UCSF Chimera (Supplementary Fig. S7A, B, C, D, and E) revealed conformational alterations in the structural backbone of the defensin peptides, as well as modifications within the residues comprising the putative active site of the enzyme.

Interactions and mechanistic insights between defensin peptides and α-amylase

The analysis of hydrogen bonding interactions between the defensin peptide chain and the α-amylase chain in the docked structures revealed critical molecular details. Using UCSF Chimera, interchain hydrogen bonding was visualized between the putative active site of α-amylase and specific residues (R1 and T2) in the defensin peptide chain. In the D1AA complex, the R1 residue of DefV1 formed two hydrogen bonds (1.997 Å and 2.059 Å) with H301 and three hydrogen bonds (1.878 Å, 1.910 Å, and 2.171 Å) with D110 of α-amylase. Additionally, T2 of DefV1 established a single hydrogen bond (2.053 Å) with D302 of α-amylase (Fig. 7A). In the D2AA complex, the R1 residue of DefV2 formed one hydrogen bond (1.716 Å) with E237, while T2 formed a single hydrogen bond (1.942 Å) with D302 of α-amylase (Fig. 7B). Beyond the active site, additional hydrogen bond interactions were observed in both complexes: R26-N153 and N154; N32 and K34-D350 for D1AA, and similarly for D2AA, R26-N153; N32-D350 and D348; and K34-D350. Notably, the substitution at the 24th position (K24R) in DefV1 resulted in a hydrogen bond (1.684 Å) with S306, an interaction absent in DefV2. These findings suggest a stronger interaction between DefV1 and α-amylase compared to DefV2 (Fig. 7C, D).

Fig. 7.

Cartoon representations illustrating the hydrogen bonding interactions between defensin variants (DefV1 and DefV2) and the probable active site residues of α-amylase in docked 3D structures using UCSF Chimera. (A) Arginine 1 of DefV1 (blue ball-and-stick model) forms two hydrogen bonds (1.997 Å and 2.059 Å) with Histidine 301 and three hydrogen bonds (1.878 Å, 1.910 Å, and 2.171 Å) with Aspartate 110 of α-amylase. Additionally, Threonine 2 forms a single hydrogen bond (2.053 Å) with Aspartate 302. (B) Arginine 1 of DefV2 (yellow ball-and-stick model) forms a single hydrogen bond (1.716 Å) with Glutamate 237, while Threonine 2 forms a single hydrogen bond (1.942 Å) with Aspartate 302. (C) Arginine 24 of DefV1 (blue ball-and-stick model) forms a hydrogen bond with Serine 306 of α-amylase, whereas (D) Lysine 24 of DefV2 (yellow ball-and-stick model) does not form any hydrogen bond with α-amylase. (E) “Cork in the bottle” model depicting the defensin peptide (green ribbon structure) penetrating the structural trench of α-amylase from Callosobruchus maculatus (tan chain), illustrating the hypothesized inhibitory interaction mechanism.

Furthermore, the docked structures provided insights into a potential “cork in the bottle” mechanism, wherein the defensin peptide precisely blocks the probable entry port of α-amylase. Three-dimensional visualizations demonstrated a seamless fit of the defensin peptide into the structural trench of α-amylase, akin to a cork fitting tightly into a bottle’s mouth (Fig. 7E). This observation led to the hypothesis that the defensin chain impedes amylose molecule access by aligning within the enzyme’s entry port, potentially elucidating the inhibitory mechanism of defensin-α-amylase interactions. These findings not only highlight the molecular basis of their interaction but also provide a foundation for future investigations into the functional and mechanistic aspects of this system.

Inhibitory activity of DefV1 and DefV2 proteins against α-amylase enzyme

The inhibitory effects of the defensin proteins, DefV1 and DefV2, on α-amylase activity were evaluated in vitro. The recombinant proteins were purified from E. coli cells expressing pTXB1 constructs with a chitin tag. The inhibitory potential of these defensins was quantified by measuring the percentage reduction in α-amylase enzymatic activity in the presence of DefV1 and DefV2. DefV1 exhibited a significant inhibitory activity of 61.86%, while DefV2 showed a comparatively lower inhibition rate of 47.377% (Fig. 8). The raw data and corresponding statistical analyses for the α-amylase inhibitory activity of DefV1 and DefV2 proteins are presented in Supplementary Table S4.

Fig. 8.

α-Amylase inhibitor (AI) activity assay of defensin peptides expressed in E. coli. The green bar represents the DefV1 peptide, and the orange bar represents the DefV2 peptide, both purified from the chitin tag fusion protein. The blue bar denotes the total protein isolated from E. coli ER2256 cells, while the pink bar represents the total protein from E. coli cells harbouring the pTXB1 vector. DefV1 exhibited a significant inhibition activity of 61.86%, whereas DefV2 showed a comparatively lower inhibition of 47.38%. The asterisks above bars indicate statistical significance (p ≤ 0.001).

Insect bioassay to check the efficacy of DefV1 and DefV2 against bruchids

The results of the artificial seed bioassays are presented in Fig. 9. In each replicate, the number of eggs was recorded 24 h post-mating, with an average of 10 eggs per replicate of 15 seeds (Supplementary Fig. S8).

Fig. 9.

Percent emergence of Insect in Insect Bioassay. The bar diagram illustrates the percentage emergence of insects in control and artificially formulated seeds. The vertical axis represents the percentage emergence of insects, while the horizontal axis represents the various seed types formulated for insect bioassay. The bar diagram demonstrates a substantial reduction in insect emergence in the formulated seeds compared to the seeds containing chickpea flour (CF) and sterile distilled water (SDW), as well as the seeds containing chickpea flour and Tris saline column buffer (CB) suggesting the efficacy of the defensin protein in inhibiting the.

After 56 days of oviposition, the highest insect emergence was observed in Control 1 (chickpea flour mixed with distilled water), with an average of 10 insects across three replicates, corresponding to an emergence rate of 33.3%. In Control 2 (chickpea flour mixed with column buffer), insect emergence was slightly lower, at 30%.

Among the treated seeds, the highest insect emergence was recorded for DefV2 at a concentration of 0.1% (w/w), followed by 0.2% (w/w). In contrast, an increase in DefV2 concentration to 0.3% resulted in a delayed developmental cycle, with only first- or second-instar larvae being observed.

For DefV1, the maximum percentage of insect emergence occurred at a concentration of 0.1%. However, no larvae or adult insects were detected at higher concentrations of 0.2% and 0.3%, indicating a strong inhibitory effect at these concentrations. The raw data and statistical analyses of the insect bioassay conducted using artificial seeds incorporating different concentrations of DefV1 and DefV2 proteins are presented in Supplementary Table S5.

Discussion

Plants have developed a diverse array of defense mechanisms to safeguard themselves against biotic threats, including pests and pathogens. Among these, the production of small cysteine-rich peptides known as defensins plays a crucial role in antimicrobial defense. These peptides exhibit broad-spectrum antimicrobial activity, including efficacy against insect pests. The VrCRP-encoded cDNA was isolated from the mung bean (V. radiata) near-isogenic line (NIL) VC6089A, which was characterized as resistant to bruchid infestation¹⁸. In the present study, the sequences obtained from various legume species exhibited a 95–98% similarity with VrCRP, suggesting that the isolated cDNA represents a putative defensin gene responsive to bruchid attack. A comparable investigation in adzuki bean (V. angularis) revealed a sequence identity of 78.3% with VrCRP, which is comparatively lower than the similarity observed in the present study¹⁸.

Plant defensins are small, cysteine-rich antimicrobial and insecticidal peptides that are synthesized and secreted by various plant tissues, including seeds, roots, leaves, and flowers^14,16,41. However, they have been most extensively studied in seeds due to their significant role in plant defense mechanisms^17,42,43. Defensins exhibit both constitutive and inducible expression patterns, with the latter being triggered by environmental stresses^44,45.

In the present study, the differential expression of insecticidal defensins was analyzed in legume seed and pod wall samples. The results indicated that seed samples of common bean (P. vulgaris), cowpea (V. unguiculata), and black gram (V. mungo) exhibited upregulation of defensin genes. In contrast, pea (P. sativum), mung bean (V. radiata), pigeon pea (C. cajan), and chickpea (C. arietinum) demonstrated downregulation. Previous transcriptomic analyses conducted by our research group²⁰, as well as findings from Subtraction Suppression Hybridization (SSH) techniques¹⁹, also revealed upregulation of defensin genes alongside other defense-responsive genes in response to bruchid oviposition on immature black gram pods. However, in the present study, the observed increase in defensin gene expression in black gram was limited to a 1.2-fold change, which is relatively low compared to the upregulation reported in previous studies^19,20.

Upon sequence comparison, it was determined that the defensin gene identified in the present study differed from those reported in prior studies. This suggests the possible co-existence of two distinct defensin gene sequences within a single plant, each responding to bruchid infestation in a unique manner. The expression levels of plant defensin genes are known to vary significantly depending on the type of stress, tissue, or cell type, reinforcing their role not only in plant defense but also in growth and development. Consequently, varied expression patterns were observed in different legume tissues, particularly seeds and pod walls.

Defensin protein acting through membrane permeabilization, enzyme inhibition, and modulation of host defense signaling pathways^46,47. For example, the Nicotiana attenuata defensin (Nadefensin) gene was induced by 1.6-fold in Nicotiana attenuata due to Manduca sexta⁴⁸. This suggests that plants elicit defensins to protect itself from insect infestations. Plants maintain tight regulation over defense gene expression, including defensin, to minimize fitness trade-offs associated with excessive resource allocation to defense⁴⁹. This suggests that modest expression levels may represent an evolutionarily optimized balance between growth and protection. Moreover, defensins may participate in broader defense networks by triggering secondary signaling cascades or acting synergistically with other antimicrobial compounds or protein inhibitors, thereby amplifying their protective impact^50,51. It is also important to consider that such expression patterns may represent an early or response to herbivory, which could escalate upon continued or repeated insect attack. Thus, moderate levels of defensin gene expression in legumes could have significant biological consequences during its interactions with bruchids.

Among the legumes analyzed, common bean exhibited the highest resistance to bruchid infestation, which may be attributed to its elevated α-amylase inhibition activity in seeds⁵². This study focuses on a defensin gene isolated from different legumes, showing high sequence similarity to VrCRP plant defensins, suggesting its role as a putative defensin gene in response to bruchid infestation. The differential expression of insecticidal defensins in various legume tissues underscores their multifaceted role in plant defense, growth, and development. However, resistance to bruchid infestation is also influenced by the plant’s genotype⁵³, suggesting that defensin gene expression may be governed by genetic factors inherent to each species.

Plant defensins are characterized by the presence of eight conserved cysteine residues, along with two glycine residues and a single glutamate residue¹⁴. The amino acid sequence analysis of VrCRP indicates a similar conserved pattern for cysteine and glycine residues; however, it lacks the conserved glutamate residue. Notably, other plant defensins, such as γ-purothionin from wheat endosperm, γ-hordothionin from barley endosperm, fabatin-2 from Vicia faba, and VaD1 from azuki bean, exhibit an arginine residue in place of glutamate^18,54. This substitution aligns with the conserved amino acid patterns observed in the present study. Furthermore, comparative nucleotide sequence analysis of VrCRP with other legume defensins reveals that single nucleotide variations in the cDNA sequence correspond to specific amino acid changes in the open reading frame (ORF) sequence.

A comparative analysis of the amino acid sequences of DefV1, DefV2, and VrCRP revealed distinct differences in the VrCRP sequence compared to both DefV1 and DefV2. Notably, a single amino acid substitution at position 59, where aspartic acid replaces asparagine, is likely responsible for the divergence of VrCRP from these two variants. Phylogenetic analysis using the Maximum Likelihood (ML) algorithm (Fig. 5) further supports this distinction, grouping DefV1 with black gram, pea, cowpea, and common bean, while DefV2 clusters with mung bean, chickpea, and pigeon pea. The grouping pattern is consistent with their respective amino acid sequences and further validates the phylogenetic out-grouping of VrCRP from both DefV1 and DefV2.

The InterPro-EMBL database results (Supplementary Fig. 6) indicated that the defensin identified in legumes belongs to Class-I defensins¹⁴. Class-I defensins are characterized by the presence of a signal sequence at the N-terminal and defensin domains. The signal sequence facilitates the export of the defensin molecule to the transmembrane, where it serves as a primary defense mechanism against invading plant pathogens. In contrast, Class-II defensins, which are less prevalent, contain an additional C-terminal prodomain (CTTP), which functions to direct the defensin to the vacuolar space, where it is subsequently cleaved⁵⁵.

Plant defensins share a conserved structure, adopting a cysteine-stabilized αβ (CSαβ) fold as individual units. Despite their small size, these molecules exhibit a compact and highly stable structure, maintained by four or more disulfide bonds. Although structurally similar, plant defensins display a wide range of biological activities^56,57. Their amino acid sequence identity varies significantly, ranging from less than 35% to over 90%. In most plant defensins, the number of cysteine residues remains conserved, particularly C1 to C8. The spacing between the C3-C4, C6-C7, and C7-C8 cysteine residues is fixed at three, one, and three amino acids, respectively⁵⁸. These cysteine residues form disulfide bonds in a specific pattern, where C1 bonds with C8, C2 with C5, C3 with C6, and C4 with C7. This conserved disulfide bridge pattern aligns with the defensins identified in legumes in the current study.

The enzyme α-amylase (Amy gene) is a crucial digestive enzyme involved in the hydrolysis of water-soluble polysaccharides. The enzymatic activity of defensins has been extensively studied in insects, particularly in bruchids, where it plays a crucial role in breaking down starch found in legume seeds⁵⁹. In the present study, the interaction between defensin peptides and α-amylase was investigated, revealing that both DefV1 and DefV2 interact with the α-amylases of C. maculatus. Similar research has demonstrated that VuD1, an unconventional cowpea seed defensin, inhibits insect-pest α-amylases⁶⁰. In contrast to the findings of Pelegrini and co-workers⁶¹, the present study found that DefV1 and DefV2 inhibited the α-amylase of C. maculatus, albeit with varying potency. Biochemical inhibition assays demonstrated that DefV1 exhibited 15% greater inhibition than DefV2. Various studies suggest that plant defensins may be involved in plant defense against insects by interfering with insect digestion, thereby depriving them of energy derived from starch degradation⁴².

Previous studies suggested that the presence of positively charged amino acid residues in Loop 3 contributes to the inhibitory activity of the plant defensin VrD1 against the α-amylase of Tenebrio molitor⁶². However, docking and structural visualization in the current study identified the interaction of R1 and T2 residues instead, interacting with active site residues H301, E237, D110, and D302 within the catalytic site of α-amylase. Similarly, previous studies have reported interactions between K1 of VuD1 and D204 of ZSA. Heterologous expression of VuD1 effectively inhibited α-amylases from Acanthoscelides obtectus and Zabrotes subfasciatus, with minimal effects on α-amylases from mammals, C. maculatus, and Aspergillus fumigatus. Comparable results were observed during docking and molecular visualization in the present study.

The human pancreatic amylase (HPA) sequence was retrieved from the NCBI database (Accession No. AAH20861.1), and a 3D homology model was constructed using SWISS-MODEL with full query coverage and 100% sequence identity to the template P19961.1.A. The model yielded a GMQE score of 0.97. Protein-protein docking between HPA and defensin peptides (DefV1 and DefV2) was performed using ClusPro, revealing a distinct interaction mechanism. Contrary to the “cork in the bottle” hypothesis, docking results demonstrated an inverted and improper fitting of defensin peptides within the structural trench of HPA (Supplementary Fig. 7D). Furthermore, inhibition of mammalian amylase requires a positively charged residue at position 40 adjacent to the third loop of the defensin peptide. The present study confirmed the presence of a neutral cysteine residue at position 40 in both DefV1 and DefV2, which negates their inhibition of mammalian amylase.

Further structural analysis identified a key interaction between K34 of the defensin peptides and D350 of the α-amylase of C. maculatus, forming a hydrogen bond. Notably, no such interaction was observed with human amylase (Supplementary Fig. 7E). The differences in amylase function across species may be attributed to evolutionary divergence. To explore this, a phylogenetic analysis was conducted using MEGA 11.0 software, incorporating mammalian amylase sequences from Pan troglodytes (chimpanzee) (Acc. No. XP009424888.1), Homo sapiens (human) (Acc. No. AAH20861.1), Pongo pygmaeus (orangutan) (Acc. No. XP054350862.1), Felis catus (cat) (Acc. No. XP044892237.1), and Rattus norvegicus (Norway rat) (Acc. No. BAB39466.1), along with the α-amylase sequence of C. maculatus (UniProt ID A0A653D1J8) (Supplementary Fig. 2B). Multiple sequence alignment using Clustal W and phylogenetic tree construction via the Maximum Likelihood algorithm revealed strong evolutionary relationships shaped by temporal mutations, with human amylase clustering closely with chimpanzee amylase, while C. maculatus amylase formed an outgroup.

Previous studies have showed that the 310-helix in the VrD1 peptide chain, facilitated by an arginine residue at position 26, influenced the orientation of Tryptophan (W10), enabling interaction with the active site of T. molitor α-amylase⁵⁹. In the present study, instead of the 310-helix, the R1 and T2 residues of defensin peptides interacted with the active site of C. maculatus α-amylase. Furthermore, the interaction analysis indicated that DefV1 formed hydrogen bonds with C. maculatus α-amylase, where K34 and N32 interacted with D350, R26 with D153 and D154, and R24 with S306. In contrast, such interactions were not observed for DefV2, likely due to conformational differences. DefV1 features longer loops encompassing the 310-helix between the β−2 and β−3 sheets, whereas DefV2 has a single short loop between these sheets. These structural distinctions support the hypothesis that DefV1 exhibits stronger inhibitory interactions with the α-amylase of C. maculatus compared to DefV2, while also providing a plausible explanation for the lack of interaction with human pancreatic amylase and other mammalian amylases.

The inhibitory action of DefV1 against bruchids was confirmed by performing insect bioassay. Although the sequence studied in the present study was quite different from the defensin identified by Chen and his co-workers¹⁸, the efficacy of DefV1 was as par with the VrCRP (Supplementary Fig. 8). A truncated version of VrCRP from VC6089A (resistant variety), a mung bean variety reported by Chen and his co-workers¹⁸ completely inhibited bruchid development in in vitro conditions at a concentration of 0.2%. An in vivo quantitative analysis revealed that seeds of variety VC6089A contained 114.7 µg defensin per gram of dry seed, compared to 32.9 µg/g in VC1973A (susceptible variety), equivalent to 0.012% and 0.0033% (w/w) of dry seed, respectively. The approximately 3.5-fold higher defensin content in VC6089A was proposed as the key factor underlying its resistance to C. maculatus, unlike the susceptible VC1973A seeds. Further studies by Chen and coworkers showed that Pichia-expressed rVrD1 exhibited bruchid-resistant activity comparable to bacterially expressed VrCRP-TSP, with artificial seeds containing 0.2% or 0.4% rVrD1 completely arresting bruchid larval development⁶³. Contrarily, the bruchid-resistant activity of adzuki bean defensin (VaD1) was found to be weaker than that of mung bean VrD1⁶⁴. Artificial seeds containing 0.1% VaD1 exhibited significantly lower resistance, with a prolonged development period of 36.6 days and 53.3% adult emergence. Increasing VaD1 content in artificial seeds from 0.1 to 0.8% reduced adult bruchid emergence from 51 to 32.5%, indicating a significant dose-dependent effect. Complete inhibition of bruchid development was achieved with artificial seeds containing 0.8% VrD1 purified from VC6089A. Similarly, in the present study, higher concentrations of DefV2 (0.3%) were associated with delayed bruchid development, suggesting a potential inhibitory effect at greater concentrations. Additionally, a delayed adult emergence (approximately 20 days) in artificial seeds compared to intact seeds was observed, aligning with the observation of previous study by Chen and co-workers¹⁸. A dose-dependent reduction in larval mass of H. armigera (lepidopteran) was observed, with the highest reduction (37%) occurring at 125 µg/ml CanDef-20 compared to control diets. The findings of the present study suggest that the defensin protein has a dose-dependent insecticidal effect on insect larvae.

Conclusion

The present study provides compelling insights into the insecticidal properties of defensin peptides in black gram (V. mungo), highlighting their potential to enhance plant resilience against bruchid infestation. Through comprehensive analysis, distinct amino acid sequence variants of defensin peptides were identified, alongside differential gene expression patterns, which collectively underscore the intricate nature of defensin-mediated defense mechanisms in plants. These findings emphasize the promising role of defensin genes, particularly those derived from black gram, in crop improvement strategies aimed at conferring durable resistance to bruchids, a major pest affecting leguminous crops.

Insect bioassays conducted in this study demonstrated that the defensin protein confers resistance against bruchid beetles. Based on in vitro assays, we observed that low concentrations (0.2–0.3% w/w) of the protein were effective in reducing bruchid infestation. These preliminary findings suggest that defensin-based approaches may be effective for managing bruchid pests. However, the efficacy of the DefV1 defensin protein against bruchid infestations would be evaluated in stable transgenic lines that express the DefV1 gene in seeds.

Additionally, elucidating the molecular mechanisms underlying defensin-bruchid interactions, would be useful. Investigating these mechanisms will provide critical insights into how defensin peptides exert their insecticidal effects and how these interactions can be optimized for pest control. Furthermore, exploring innovative strategies for integrating defensin-based approaches into sustainable agricultural practices will be essential for developing eco-friendly and effective solutions to mitigate bruchid infestations. Such efforts will contribute to the advancement of crop protection technologies, ultimately supporting global food security and sustainable agriculture.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Sumita Acharjee: Supervision, investigation, Conceptualization; Writing – review & editing, Research Plan and design, Funding acquisition. Jyotsna Dayma: Writing – original draft, Methodology, Formal analysis, Data curation. Padmanav Koushik: Writing – original draft, software resources. Mamta Bhattacharjee: Bidyut Kumar Sarmah: Project-related administration. Prakash Jyoti Kalita: Assisted in insect bioassays. Debajit Das: Validation, Writing, review & editing.

Data availability

The datasets generated and/during the current study are available in the NCBI repository and received accession numbers. Accession Numbers, PP334100 (SUB14221196 Seq PP334100).

Declarations

Competing interests

The authors declare no competing interests.

No potent conflict of interest is stated among the authors.