Deciphering metabolomic signatures of drought adaptation in groundnut triggered by encapsulated microbial biostimulant

Microbial biostimulants enhance crop resilience to drought and promote sustainable agricultural productivity. The present research aimed to elucidate the metabolomic variations induced by encapsulated microbial biostimulants for drought tolerance in groundnut. Two rhizobacterial strains, Acinetobacter calcoaceticus AC12 and Bacillus amyloliquefaciens BA13, were functionally tested for plant growth promoting attributes and osmolyte production under osmotic stress. AC12 and BA13, as consortium, were encapsulated using sodium alginate and chitosan to formulate a plant-growth-promoting rhizobacteria (PGPR)-based microbial biostimulant. The encapsulated microcapsules were characterized by SEM, FTIR, and other physiochemical properties. The results of a field trial demonstrated that the encapsulated microbial biostimulant had a significant (p < 0.05) impact on the physiological, biochemical, and yield attributes of groundnut under drought. Groundnut seed metabolomic analysis identified 52 altered metabolites, including sugars, fatty acids, and amino acids. Partial least square discriminant analysis and heat map identified potential biomarker metabolites. A pathway enrichment analysis revealed that inoculation of encapsulated biostimulants significantly altered 25 metabolic pathways under water deficit conditions. By deciphering key metabolomic shifts, the study provides insights on biostimulant-based drought endurance and lays a foundation for developing new biostimulant products with encapsulation technology to enhance global food production and nutrition.

Microbial biostimulant is an environment-friendly strategy in agriculture to secure agricultural production, yield stability, and promote soil quality (Naqve et al., 2023; Garg et al., 2024). It refers to a single or a combination of microbes designed to enhance plant growth, induce systemic and acquired resistance in plants, and improve tolerance to stress factors (Jiang et al., 2024). These biostimulants facilitate the plants’ defense mechanism by hormonal adjustment, transcription regulation, osmolyte accumulation, and modification of root architecture (Kumar et al., 2024).

Plant growth promoting rhizobacteria (PGPR) as biostimulants are natural habitants of the root rhizospheric region and has ideal power to enhance agronomic sustainability and reduce drought (Zeng et al., 2025). Drought stress is a predominant constraint responsible for multifaceted consequences in plants (Qiao et al., 2024). Arachis hypogaea L. (groundnut), a valuable leguminous crop which is an excellent source of protein, dietary minerals, fatty acids, and bioactive compounds, faces an annual loss of 6 million tons and 85% pod yield reduction due to drought (Prodic et al., 2023). It also hinders pod formation and negatively impact the nutritional quality of groundnut, resulting in poor seed grade, vigor, and viability (Pokhrel et al., 2025).

Discovering microbial consortia from extreme drought hides countless potential to enact systemic tolerance in groundnut (Krishnan et al., 2025). According to Chieb and Gachomo (2023), stress-resilient rhizobacteria produce biochemicals with significant potential as biostimulants and essential for plant stress responses. Over the years, several research revealed the positive stimulation impact of PGPR on plant growth and its survival in drought through mechanisms such as (i) hormone regulation, (ii) phosphate solubilization, (iii) siderophore and exopolysaccharide production, (iv) nutrient acquisition, (v) antioxidant activity, and (vi) production of osmolytes such as trehalose, alginate, glycine betaine, salicylic acid, and ectoine (Tanvere et al., 2023; Azri et al., 2024).

The encapsulation of PGPR in a biopolymer matrix is one of the efficient techniques to conserve the functional strains and improve its stability, viability, and plant root colonization during fluctuating climatic factors, thus favoring the valuable commercialization of biostimulant formulations (Balla et al., 2022; Tariq et al., 2025). Alginate and chitosan are commonly utilized as potential carrier materials for microbial encapsulation due to their benefits like being non-toxic, biocompatible, biodegradable, abundant, and sustainable (Riseh et al., 2023).

In the presence of sodium alginate, chitosan, a polyvalent cation, undergoes instant electrostatic polymerization through cross-linking, which decreases the porosity of alginate and forms gel microcapsules with increased surface hardness and internal stability (Parsana et al., 2023; Waqar et al., 2024). Starch is utilized as filler material to stabilize the formulation as well as preserve and shield microbial cells during transportation and storage (Rani et al., 2023). The chitosan-coated alginate microcapsules can endure in soil with stability, guaranteeing the controlled release of PGPR and successful root colonization.

Metabolomics is currently utilized to gain insights into the mechanisms of action of microbial biostimulants on drought-stressed crops (Kashyap et al., 2024). The omics sciences, especially untargeted metabolomics, are employed to assess the levels of plant metabolites within a specific molecular pathway of both transcriptional and post-transcriptional gene regulation, influenced by drought and biostimulants (Roychowdhury et al., 2023; Li et al., 2024).

Analyzing the metabolic snapshots of plants through correlation analysis will lay the groundwork for the construction of metabolic networks. These statistical metrics offer important insights into the alerted behavior of different metabolites (Kisiel et al., 2023).

Despite ongoing efforts that are emerging in understanding the effects of biostimulants on plants, metabolomic studies on the tripartite interaction (biostimulant–plant–abiotic stress) remain enigmatic. This knowledge gap due to the lack of fundamental understanding of the modes of action of encapsulated biostimulants hampers the novel formulation of biostimulants and the implementation of these products into agronomic practices. Therefore, to effectively establish novel biostimulant-based agricultural strategies, in the present study, through metabolomics, we reported an elucidation of metabolic alterations and differentially expressed biomarkers induced by an encapsulated microbial biostimulant in groundnut under drought. Such insights and findings herein contribute toward understanding the molecular knowledge base of biostimulants and to leverage the values of such formulations in agronomic avenues.

Source of drought-tolerant rhizobacterial strains

The rhizobacterial strains Acinetobacter calcoaceticus (AC12) and Bacillus amyloliquefaciens (BA13) were retrieved from the Soil Biology and PGPR laboratory, Department of Botany, Periyar University, Tamil Nadu, India. The strains were originally isolated from the rhizospheric soil of groundnut. They were genotypically identified by 16S rRNA gene sequence and deposited in NCBI with the accession number ON495939 for AC12 and ON495964 for BA13 (Supplementary Figure 1).

The bacterial strains were assessed for their drought tolerance under PEG concentrations (up to 30% PEG-6000) in nutrient broth, followed by incubation for 72 h at 30 °C. The optical density at 600 nm was measured to record the growth of AC12 and BA13 (Ansari and Ahmad, 2018). The stock cultures of the strains were maintained for long-term storage at −70°C in nutrient broth (NB) with 30% glycerol and used for further analysis.

In vitro bioassay for multifarious plant-growth-promoting traits

Drought-tolerant AC12 and BA13 strains were characterized in vitro for direct and indirect multifunctional plant-growth-promoting properties such as indole acetic acid (IAA), phosphate, solubilization, siderophore, gibberellic acid, and exopolysaccharide production.

Indole production

Bacterial cultures were inoculated in NB medium with L-tryptophan (100 µg/mL). After incubation, the cultures were centrifuged (10,000 rpm, 10 min, 4°C), and 1.5 mL bacterial supernatant with 4 mL of Salkowski’s reagent (12 g FeCl 3 per liter in 7.9 M H 2 SO 4) was mixed and incubated for 15 min. The intensity of indole (IAA, μg mL -1) was quantified at a wavelength of 520 nm using a spectrophotometer (Gusmiaty et al., 2019).

Phosphate solubilization

The bioassay for the phosphate solubilization efficiency was performed on Pikovskaya agar plates as defined by Pikovskaya (1948). The streaked plates were incubated at 28°C for 6–12 days. A halo zone around the colonies was counted as positive, and the phosphate solubilization index (PSI) was formulated as:

The quantitative analysis of phosphate solubilization was performed using the method of Nautiyal (1999). Bacterial cultures (500 μL) were inoculated into 50 mL National Botanical Research Institute’s phosphate (NBRIP) broth media with Ca 3 (PO 4) as insoluble phosphate source and incubated at 28°C for 72 h. After incubation, the cultures were harvested by centrifugation at 10,000 × g for 10 min, and the phosphate in the supernatant was estimated using the method of Fiske and Subbarow (1925).

Siderophore production

For the qualitative analysis, blue agar plates containing chrome azurol sulphate (CAS) dye were used as a universal method to identify siderophore production with the emergence of an orange halo encircling the rhizobacterial colony (Schwyn and Neilands, 1987). CAS shuttle assay was executed for siderophore quantification as described by Alexander and Zuberer (1991). The percentage of siderophore was measured as the proportion of CAS color using the following relation:

Ar = absorbance of reference (CAS solution and uninoculated medium) and As = absorbance of sample.

The catecholate type of siderophore was determined by using Arnow’s test (Arnow, 1937), and the detection of hydroxamate siderophore was carried out by using tetrazolium test (Snow, 1970).

Gibberellic acid production

Gibberellic acid was evaluated following Abou-Aly et al. (2019) by mixing an equal ratio of HCl, Folin–Ciocalteu reagent, and the bacterial supernatant (1:1:1) with distilled water to make a total of 6 mL. The mixture was boiled in a water bath for 5 min and cooled. The intensity of the generated bluish-green color was calculated by using a spectrophotometer at 760 nm.

Exopolysaccharide production

Exopolysaccharide (EPS) quantification was performed as described (Breedveld et al., 1990). The strains were inoculated in 50 mL of nutrient broth and grown at 28°C ± 2°C for 5 days at 120 rpm. The culture was centrifuged at 11,500 × g for 20 min at 4°C. The supernatant was filtered through a 0.45-μm nitrocellulose filter. EPS was precipitated from the filtrate by using chilled 96% ethanol (1:3). The solution was refrigerated at 4°C for 24 h and centrifuged again. The precipitated EPS was quantified (mg mL -1) after drying at 80°C for 48 h.

Osmolyte production of drought-tolerant bacterial strains (AC12 and BA13)

The bacterial strains were characterized for osmolyte production such as glycine betaine, trehalose, salicylic acid, alginate, and ectoine with supplementation of PEG 6000 to lower the water potential and mimic drought stress.

Glycine betaine

The protocol of Vasconcellos et al. (2021) was followed for glycine betaine quantification by incubating the drought-tolerant strains in NB with distinct PEG concentrations (0% and 30%) at 27°C for 24 h. The bacterial cells were centrifuged (5,000 × g), weighted, and resuspended in ethanol (1 mL) and vigorously shaken for 30 min. The extract was then filtered through a 0.45-µm filter and verified for glycine betaine production (mg g -1) using HPLC on an RP-18 column (Merck, Germany) and detected at 200 nm.

Trehalose

The concentration of trehalose was determined by using the reformed protocol of García et al. (2017). Concisely, strains were grown in NB supplemented with a range of PEG concentrations (0% and 30%). The bacterial cultures were centrifuged, rinsed with water, and resuspended in 80% ethanol abide by incubation at 85°C for 15 min. The mixture was centrifuged at 10,000 × g for 5 min, and the supernatant was retrieved. The samples were air-dried and resuspended in sterile deionized water. The trehalose concentration was examined by HPLC using the column Aminex HPX-87C (Bio-rad Labs, CA, USA) with acetonitrile and water (80:20, v/v). The values are referred to as microgram per milligram (μg mg -1) of cells.

Salicylic acid

The protocol of Lukkani and Reddy (2014) was followed to estimate the salicylic acid (SA) production of rhizobacterial stains. A total of 4 mL of rhizobacterial supernatant was acidified with 1 N HCl and adjusted to pH 2. Salicylic acid was precipitated with 1:1 (v/v) of chloroform (CHCl 3). Then, 5 mL of 2M FeCl 3 and 4 mL of distilled water were added to pool the chloroform phase. The absorbance of the purple iron–SA complex was measured spectrophotometrically at 527 nm and quantified as microgram per milliliter (μg mL -1) of bacterial supernatant.

Alginate

The cell-free supernatant of the drought-tolerant rhizobacterial cultures was utilized for alginate extraction. A proportional amount of isopropanol was added to the supernatants (left under static conditions for 24 h) to ensure the precipitation of deacetylated alginate. The precipitate was recovered by centrifugation at 10,000 rpm for 10 min. The resultant pellet was successively washed with 1 mL of ethanol (70% and 96%), dehydrated at 37°C for 15 min, and dispersed in distilled water. Then, 1 mL of borate sulfuric acid (10 mM) and 30 μL of carbazole reagent were added. The mixture was allowed to stand for 15 min at room temperature, and absorbance at 500 nm was recorded. Alginate quantity was expressed in terms of microgram per milliliter (μg mL -1) (Wozniak et al., 2003).

Ectoine

Ectoine content was quantified by using the protocol of Nagata and Wang (2001) with minor alterations. The cell pellets of the strains obtained under distinct concentrations of PEG (0% and 30%) were concentrated by freeze-drying and then extracted with 80% ethanol for 1 h. The supernatant was dried (40°C) and extracted again for 20 min with absolute ethanol/chloroform/water (1:1:1). The extraction mixture was blended with equal quantities of chloroform, and the mixture was agitated for 10 min. The supernatant was subjected to lyophilization and dissolved in ultrapure water. Using HPLC and UV detection at 210 nm, the extracted ectoine was measured as milligram per gram (mg g -1) of cells (Shimadzu, Tokyo, Japan).

Preparation of PGPR inoculum for encapsulation

Bacterial strains AC12 and BA13 were cultured in 100 mL of nutrient broth, each in a 250-mL flask, and incubated at 28°C for 48 h under constant agitation. The culture was harvested at stationary phase, centrifuged (4,000 × g, 10 min), washed with 0.85% sterile physiological saline (NaCl) solution, and resuspended in 200 mL sterile saline (1.5 × 10 9 CFU/mL) (Ma et al., 2019). The resulting suspension of each strain was mixed in equal ratio, and the PGPR consortium was stored at 4°C until subjected to encapsulation.

Biopolymeric encapsulation of rhizobacterial consortium for the production of PGPR-based biostimulant

The encapsulation of PGPR consortium (Figure 1) was carried out accordance with the procedure outlined by Zago et al. (2019) with slight amendments. The encapsulation process was executed in a sterile laminar airflow hood. A solution of sodium alginate (HiMedia) (2 g/100 mL) was made with distilled water and 2% starch. It was autoclaved at 121°C for 15 min. After cooling down to ambient temperature (38°C–40°C), the solution was uniformly incorporated with rhizobacterial consortium at a ratio of 2:1.

Aliquots of the biostimulant (BS)/alginate solution was vigorously agitated for 1 h for homogeneous alginate dissolution formerly being extruded drop-wise over a 2-mL syringe (needle size, 0.2 mm) into sterile 150 mL of 0.1 M CaCl 2 as a hardening solution under constant agitation for cross-linking at room temperature to form encapsulated microbial beads by ionic gelation. The syringe was sustained at a distance of 5 cm to the surface of the CaCl 2 solution to create spherical capsules. They were held in the solution overnight. The encapsulated microcapsules were rinsed thrice with sterile water and further subjected to chitosan coating.

Chitosan (0.4%) was dissolved in 90 mL of distilled water and acidified with 0.4 mL glacial acetic acid. The pH was increased to 6.0 using 1 N NaOH, and the volume was adjusted to 100 mL, followed by autoclaving at 121°C for 15 min. Subsequently, 15 g of the prepared alginate microcapsules was suspended in the chitosan solution while swirled at 100 rpm for 40 min. The chitosan-coated alginate microcapsules were washed twice with 0.85% NaCl solution and were stored in peptone water for further analysis.

Characterization of biostimulant-encapsulated microcapsules

Visualization by SEM

The morphology of the encapsulated microcapsules was visualized using a scanning electron microscope (ZEISS EVO 19, Germany). Before being loaded onto the microscope, the chitosan-coated alginate microcapsules were freeze-dried and sputter-coated with gold (10 nm). The images of the microcapsules were documented in the representative fields and performed at 15 kV under low vacuum with an angle of 45° (de-Bashan et al., 2016).

Chemical structure characterization by FTIR

Fourier transform infrared spectroscopy (FTIR) analysis was performed to analyze the chemical structure of the microcapsules and to determine the presence of the various bonds in the polymer matrix. The microcapsules were vacuum-dried, ground into a fine powder, which was subsequently blended with potassium bromide powder (1:200). This mixture was then pressed to form a disk for FTIR characterization. The spectra were collected over the range from 4,000 to 500 cm −1 at a resolution of 4 cm -1 by using an Agilent FTIR spectrophotometer (Shimadzu, Kyoto, Japan) (Agnes et al., 2023).

Moisture content of encapsulated microcapsules

The weight of the wet microcapsules put on filter paper was calculated and recorded as Ww. They were then dried for 48 h at room temperature (approximately 27°C ± 2°C) to achieve a consistent weight in order to assess the moisture content. The dried microcapsules bearing the AC12 strain were examined to get the final weight (Wd). The following formula was used to calculate the moisture percentage (Chen et al., 2012) of the microcapsules.

Ww—wet weight (g) of the microcapsules.

Wd—dry weight (g) of the microcapsules.

Swelling analysis of encapsulated microcapsules

The swelling analysis was conducted using weight expansion technique. Dried microcapsules (1 g) were submerged in 10 mL of sterile saline solution (0.9%) for 24 h. The surplus water was removed by tissue paper and promptly quantified. The swelling ratio was obtained as

W1—weight of dried microcapsules.

W0—weight of the swelling microcapsules.

Embedding efficacy of encapsulated microcapsules

The methodology of Petraitytė and Šipailienė (2019) was employed to evaluate the encapsulation efficiency in terms of the total number of PGPR enclosed in microcapsules. This study was performed via standard plate count method of live bacteria on nutrient agar (NA) from various dilutions made with phosphate-buffered saline. The agar plates were incubated at a constant temperature of 28°C for 48 h. The embedding efficiency was computed using the following equation:

N 0 specifies the initial number of rhizobacteria introduced to the biopolymer mixture, and N e denotes the number of free bacteria in the CaCl 2 solution (Wu et al., 2011).

Survival viability and release behavior

The major goals of our investigation were to maintain bacterial viability and analyze its release behavior. Survival was assessed using the depolymerization approach described by Gonzalez et al. (2018). In brief, 1 g of microcapsules was resuspended in a 10% sodium citrate dehydrate solution (20 mL, pH = 8) and homogenized by vortexing for 30 min. The encapsulated rhizobacteria within the chitosan-coated alginate matrix were liberated during homogenization. The resulting mixture was centrifuged at 4,000 × g, and the supernatant was discarded, leaving the pellet suspended in 0.85% NaCl. The suspension was serially diluted and plated on LB agar, incubated for 40 h at 28°C. The technique was repeated for a 90-day storage period, and surviving viability was recorded as viable log CFU per gram of microcapsules.

In vivo experiment in groundnut under drought

The efficacy of the prepared PGPR-based encapsulated biostimulant on alleviating the behavior of groundnut under drought stress was determined by conducting a field study. Details about the materials and techniques adopted during experimentation, plant sampling, and analyses are described below.

Plant material and inoculum preparation

The groundnut cultivar used for the field assay was TMV 7 (Arachis hypogaea L.). Groundnut seeds were procured from Tamil Nadu Agricultural University (TNAU), Coimbatore, Tamil Nadu, India. The seeds were surface-sterilized by soaking them in 1% sodium hypochlorite solution for 20 min and rinsed thrice with distilled water.

There were two levels of inoculation in the field experiment compared with the uninoculated control: (i) encapsulated PGPR biostimulant and (ii) non-encapsulated PGPR biostimulant. The chitosan-coated alginate microcapsules were used as inoculum for encapsulated biostimulant at the rate of 20 g/L of water. AC12 and BA13 rhizobacterial strains from the preserved stock were cultured in nutrient broth for 48 h at 28°C ± 2°C. The bacterial suspension was adjusted to a final concentration of 10 8 CFU/mL. The bacterial suspension was taken in equal ratio and used as inoculum for non-encapsulated biostimulant at the rate of 20 mL/L of water.

Experimental strategy and drought imposition

The factorial components comprise two levels of inoculation and two levels of drought stress. Hence, the subsequent nine experimental trials were executed:

T1—uninoculated control

T2—encapsulated BS

T3—non-encapsulated BS

T4—mild drought (MD)

T5—MD + encapsulated BS

T6—MD + non-encapsulated BS

T7—severe drought (SD)

T8—SD + encapsulated BS

T9—SD + non-encapsulated BS

The plants were subjected to two levels of drought regime extended for the entire growth period. Drought stress was induced by withholding water, and the seedlings were exposed to water stress by limiting the irrigation levels to corresponding field capacities of 100% (no drought), 60% (mild drought), and 40% (severe drought) (Chiappero et al., 2019).

Execution of field trial

The open-field trial was carried out in a farmer’s farmland at Poosaripatty (11°86′ N, 78°05′ E), Salem, Tamil Nadu, India. The average temperature of the site was 26.4°C, the average rainfall was 232 mm, and the relative humidity was 56%. The farmland soil physicochemical parameters were analyzed. The seeds were sown in the field, and irrigation was given as per the aforementioned treatments in a completely randomized block design of 5 × 3 m 2 (15 m 2) with three replicate blocks. Plant-to-plant and row-to-row spacing was arranged at 10 and 15 cm, respectively. Weeding was executed manually. Throughout the entire growth phase, no agrochemicals were applied with the exception of biostimulants. After 90 days of the experiment, the plants were gently uprooted for further analysis.

Determination of relative water content

Relative water content (RWC) was measured by using the protocol of Barrs and Weatherly (1962) with slight modifications. The fresh weight (FW) of five leaves was noted immediately on an electronic scale; the leaves were kept immersed in distilled water for 24 h, and the turgor weight (TW) of the saturated leaves was measured. Dehydration of the leaves in an oven at 70°C for 48 h was done to measure the dry weight (DW). RCW was computed as follows:

Determination of lipid peroxidation

Lipid peroxidation, quantified as malondialdehyde (MDA), was measured following the protocol of Heath and Packer (1968). In total, 200 mg of fresh leaves was homogenized in 600 µL of 0.1% trichloroacetic acid (TCA). Centrifugation (15,000 × g) for 20 min at 4°C was done to obtain the supernatant. Then, 1.5 mL of 20% TCA containing 0.5% TBA was mixed with 0.5 mL of the supernatant. The mixture was heated (95°C, 25 min) and cooled, followed by recentrifugation (15,000 × g, 4°C, 5 min). The absorbance of the supernatant was recorded at 532 nm.

Estimation of total soluble sugars

Soluble sugar content was determined by using the protocol of Dubois et al. (1956).

For the extraction, 100 mg of leaf sample was ground and treated with 5 mL of 2.5 N HCl. The resulting solution was incubated for 3 h in a water bath at 80°C. The hydrolyzed sample was neutralized with sodium carbonate, made up to 100 mL with distilled water and centrifuged at 3000 × g for 5 min. The resulting supernatant was preserved.

For the estimation, 0.1 mL of the prepared supernatant was diluted to 1 mL with distilled water. Subsequently, 1 mL of 80% phenol and 5 mL of 96% concentrated sulfuric acid were added, and the mixture was shaken (10 min) and incubated at 37°C for 30 min. The absorbance of the sample was recorded at 490 nm. The concentration of the soluble sugar (mg g -1 FW) of the samples was calculated with a standard curve of glucose solution.

Determination of protein content

The procedure of Lowry et al. (1951) was used to estimate the total protein content.

For the extraction, 0.1 g of the leaf sample was ground with 20 mL of 20% trichloroacetic acid (TCA) using a pestle and mortar. The homogenate was centrifuged at 3,000 × g for 15 min. The supernatant was discarded, and the pellet was added with 5 mL of 0.1 N sodium hydroxide (NaOH). Then, the aliquot was centrifuged again at 3,000 × g for 15 min. The supernatant was collected, and the total volume was made up to 10 mL with 0.1 N NaOH and used for the estimation of protein content.

For the estimation, 0.5 mL of the extract was added to 4.5 mL of alkaline copper sulphate, and the mixture was incubated at room temperature for 10 min. The resultant solution was subjected to incubation (30 min) at 20°C ± 2°C after the addition of 0.5 mL Folin’s reagent. The absorbance was noted at 640 nm, and total protein was measured (mg g -1 FW) with the reference curve of bovine serum albumen (BSA).

Determination of yield-contributing characters

The groundnut plants were harvested after the maturity period, and the pods were dried to attain constant weight. The agronomic yield parameters included the number of pods per plant, 100-seed weight, and yield per plant (Vu et al., 2022).

Untargeted metabolomic profiling

Extraction and derivatization of metabolites

Untargeted metabolomic (whole metabolomic) profiling was done in accordance with the modified protocol established by Lisec et al. (2009). Seed samples from the biostimulant-treated plants under stressed and unstressed conditions were collected for metabolic profiling. The collected seed samples were quickly frozen in liquid nitrogen, stored at -80°C, and lyophilized for 72 h. One gram of lyophilized samples (seed) was homogenized using a prechilled mortar–pestle in 10 mL of 80% precooled (−20°C) HPLC-grade methanol. Next, 20 μL of ribitol (0.2 mg/mL of distilled water) was subsequently incorporated as an internal quantitative standard and vortexed for 1 min at room temperature. The samples were ultrasonicated for 20 min, followed by centrifugation at 11,000 × g for 10 min. The supernatant was transferred into a separate test tube.

The extracted metabolites were derivatized as outlined by Sanimah et al. (2013) with slight modifications. For the derivatization process, 80 µL of methoxyamine hydrochloride dissolved in pyridine (20 mg mL −1) was introduced into each tube and allowed to incubate at 37°C for 2 h. Following this, 200 µL of N-methyl-N-(trimethylsilyl)-trifluoroacetamide (MSTFA) was added, and the mixture was incubated again at 37°C for 1 h. The resultant mixture was then transferred into appropriate GC vials.

Metabolomic analysis

The derivatized samples were subjected to a gas chromatography−mass spectrometry (GC−MS; Perkin Elmer, Clarus 680 GC, US) instrument linked to a mass selective detector (Perkin Elmer, clarus 600 (EI) US) and were operated following the manufacturer’s guidelines. The GC used in the analysis employed a fused silica column, packed with Elite-5MS (5% biphenyl, 95% dimethylpolysiloxane, 30 m × 0.25 mm ID × 250 μm df), and the components were separated using helium as carrier gas (99.99% purity) at a constant flow of 1 mL/min. The injector temperature was set at 260°C during the chromatographic run. Then, 1 μL of the sample extract was injected into the instrument. The oven temperature was set as follows: 60°C (2 min), followed by 300°C at the rate of 10°C min −1, where it was held for 6 min.

The mass detector conditions were as follows: transfer line temperature of 240°C, ion source temperature of 240°C, and ionization mode electron impact at 70 eV, scan time of 0.2 s, and scan interval of 0.1 s with the scan fragments from 40 to 600 Da. The spectra of the components were compared with the database of mass spectrum of known components stored in the GC-MS NIST (2008) library (National Institute of Standards and Technology, Gaithersburg, MD, USA). For chromatogram acquisition, baseline correction, and peak deconvolution, GC−MS real-time analysis software (Turbo Mass version 5.4.2) was used.

The pathway analysis was executed using MetaboAnalyst, and the metabolites were annotated using the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database. The data were normalized by log2 transformation and Pareto scaling, and a heat map with hierarchical cluster analysis was performed to investigate relatedness across treatments. The supervised method, partial least square discriminant analysis (PLS-DA), was performed using centering and Pareto scaling. The parameters (R 2 Y, Q 2 Y, degree of correlation) were recorded, and the model was validated through a permutation test (n = 100) and CV-ANOVA (p< 0.05) to visualize differential metabolites selected by variable importance in projection (VIP).

Statistical data processing and analysis

The experimental data were performed in triplicate and presented as mean and standard deviation (SD). The values were statistically analyzed using analysis of variance (ANOVA) by using IBM SPSS software v.20, Tukey’s post-hoc test (p < 0.05) for multiple comparisons, and homogeneity of variance using Levene’s test. Graphs were drawn using GraphPad Prism 9 software.

The rhizobacterial strains Acinetobacter calcoaceticus AC12 and Bacillus amyloliquefaciens BA13 were assessed for drought tolerance, plant-growth-promoting activities, and osmolyte production, with the sole target of designing encapsulated microbial biostimulants with effective bacterial strains. The strains AC12 and BA13 were resistant to a maximum of 30% PEG and with a concentration of 10 6 CFU/mL.

Bioassay for multifarious plant-growth-promoting traits

The in vitro results of rhizobacterial characterization involved in plant growth promotion such as IAA, siderophore, phosphate solubilization, gibberellic acid, and EPS are depicted in Table 1. The maximum value of IAA was recorded by AC12, followed by strain BA13. Both strains were positive for siderophore production with an orange-colored zone in the CAS agar medium. The results of the chemical characterization showed that the siderophore of both strains were of the catecholate type. The strain AC12 was the most potent with the most significant siderophore unit. Qualitatively, AC12 had a higher phosphate solubilization index than BA13 with the visualization of a sharp, clear halo around the colony. The quantitative analysis showed that AC12 had solubilized the maximum amount of phosphate, followed by BA13. Regarding gibberellic production, maximum GA was noted in AC12. Similarly, the test for EPS was recorded as positive for both strains.

Osmolyte production of drought-tolerant strains (AC12 and BA13)

The rhizobacterial strains were analyzed for osmolyte production. Osmolytes associated with drought tolerance, such as glycine betaine, trehalose, salicylic acid, alginate, and ectoine, were assessed using 30% PEG 6000-simulated drought (-1.03 MPa) as shown in Figure 2.

Regarding glycine betaine accumulation (Figure 2A), both strains produced maximum quantities at 30% PEG compared to the unstressed condition. Meanwhile, the maximum amount of glycine betaine was recorded in AC12, being 83.33% more than BA13 under stress. Trehalose was formally found to be minimum in both strains under 0% PEG (Figure 2B). However, the contents were at peak under stress. The production of trehalose by AC12 at 30% PEG was 86% greater than at 0% PEG and exceeded the rate produced by strain BA13 by 3.2% under drought.

The content of salicylic acid by AC12 and BA13 was significantly increased by 64.08% and 51.91%, respectively, under osmotic stress (30% PEG) relative to the unstressed control. AC12 exhibited 12.17% higher salicylic acid production than BA13 at 30% PEG (Figure 2C). With regard to alginate production under stress, AC12 produced 26.8% and BA13 produced 17.39% alginate compared to their respective production under normal growth (Figure 2D). The ectoine content of AC12 and BA13 was increased by 248% and 175%, respectively, under drought stress (Figure 2E). These data demonstrated a positive association between osmotic stress tolerance and osmolyte production in drought-tolerant rhizobacteria.

Analytical characterization of encapsulated PGPR microcapsules

The microscopic morphology of encapsulated freeze-dried microcapsules was examined under a scanning electron microscope. The SEM analysis exhibited successful entrapment of the bacterial strains in biopolymeric formulation. The freeze-dried microcapsule was slightly wrinkled due to dehydration on lyophilization and showed a particle size of approximately 1–1.5 mm. Figure 3 shows the SEM micrographs of microcapsules carrying the bacterial consortium of AC12 and BA13 in biopolymeric matrix at varied magnifications.

Figure 4 shows the FTIR spectrograph of encapsulated microcapsules. Specifically, sodium alginate has a characteristic absorption peak at 3,298 cm -1 formed by the bending vibration of -OH and a peak at 1,415 cm -1 which belongs to the asymmetric stretching of the COO - group. The peaks of chitosan at 1,598 cm -1 were attributed to the deformation vibration of N–H and those at the 1,077-cm -1 region correspond to C–N stretching vibrations. The starch spectral peaks at 1,148 cm -1 were due to the asymmetric stretching of C–O–C and that at 1,014 cm -1 was caused by C–O–H stretching vibrations. Thus, the spectrum of encapsulated microcapsules depicts the characteristic peaks of almost all components and had no new peaks, confirming that no reaction occurred between chitosan, starch, and sodium alginate.

The drying of encapsulated microcapsules (28°C ± 2°C) resulted in a moisture content of 5.20%. The swelling rate of the microcapsule was denoted as an expansion rate of approximately 82.05%. In the study, the encapsulation efficiency of microcapsules was found to reach 99.5%. This indicated the efficient entrapment of rhizobacterial consortium into the chitosan-coated alginate matrix by internal gelation and electrostatic adsorption methods.

The bacterial cells from the microcapsules were released in a slow-release effect, and the number of live bacteria increased significantly in the first 4 days, reaching 10 6 CFU/g of microcapsules. There was a slow-release rate of bacteria up to 10 days, and it reached saturation after 20 days (Figure 5A). The survival viability of the strains in the microcapsules was also calculated for 90 days. Figure 5B reveals the viable cell count up to 30 days, and it remains with 10 7 CFU/g of microcapsules.

Field experiment data

The encapsulated PGPR biostimulants were analyzed for their impact on groundnut in a field experiment (Supplementary Figure 2) under a range of induced drought stress.

Effects of encapsulated PGPR biostimulant on relative water content of groundnut under drought

For relative water content (RWC), a substantial (p< 0.05) decrease was recorded in stressed plants compared to the well-irrigated plants (17.64% reduction in MD and 64.70% reduction in SD). In plants treated with encapsulated and non-encapsulated BS, RWC was detected significantly (p< 0.05) with 10% and 2.85% increments, respectively, under MD. Meanwhile, in SD, the encapsulated PGPR BS notably enhanced the RWC by 22.21% in contrast to stressed plants compared to the non-encapsulated BS, as shown in Figure 6A.

Effects of encapsulated PGPR biostimulant on lipid peroxidation, TSS, and protein under drought in groundnut

As shown in Figures 6B, MD and SD induced the overproduction of ROS, resulting in lipid peroxidation (denoted by MDA accumulation). MDA was not significantly affected (p > 0.05) among treatments in the well-watered plants. MDA was remarkably amplified in the stressed plants (MA and SD) by 112.56% and 196.22%, respectively. Under MD, the biostimulant application (encapsulated and non-encapsulated) induced 23.11% and 8.6% reduction in MDA compared to the stressed plants. On the other hand, bacterial inoculation reduced (p< 0.05) the leaves’ MDA content by 24.03% in the case of encapsulated BS and by 4.4% in the non-encapsulated BS in SD stress.

The tremendous increase in TSS content of groundnut was induced by drought stress (Figure 6C). The drought-stressed plants (MD and SD) showed 113.17% and 142.41% increments in TSS compared to the well-watered control. The PGPR biostimulant significantly (p< 0.05) enhanced the TSS concentration compared to the uninoculated control under drought. TSS was maximally induced by the inoculation of encapsulated BS, representing 10.18% in MD stress, than the non-encapsulated BS. Under SD, both encapsulated and non-encapsulated BS significantly (p< 0.05) manifested the TSS concentration by 8% than the respective control.

The protein content in groundnut leaves increased in the inoculated treatment. The data (Figure 6D) showed the impact of drought stress in uninoculated plants and a significant effect (p < 0.05) on the protein with the biostimulant treatment. The treatment with the highest significant effect was achieved by the encapsulated BS microcapsules with increases of 29% and 25%, followed by non-encapsulated BS representing an increase of 8.6% and 2.1% under MD and SD, respectively.

Effects of encapsulated PGPR biostimulant on the yield-contributing characteristics of groundnut under drought

The pod number per plant was significantly improved (p< 0.05) by the implementation of the microbial biostimulant in both MD and SD plants (Figure 7A). Higher values of 37.97% in MD and 32.44% in SD were observed in the encapsulated-BS-treated plants, while the non-encapsulated BS exhibited 10.34% and 18.98% increment in the number of pods under both drought stress compared to the control.

The inoculation of encapsulated BS had a significant (p< 0.05) effect on 100-seed weight. In MD, the seed weight increased by 2.73% compared to the non-encapsulated BS. It also exhibited a higher seed weight in SD (Figure 7B).

The biostimulant-treated plants achieved a statistically superior yield over the uninoculated plants (p< 0.05). In MD, the encapsulated BS caused a maximum increase in the per hectare yield of groundnut, improving it by 17.98% compared to the control, followed by the non-encapsulated BS with 8.07% (Figure 7C). In SD plants, a substantial yield improvement was produced by both encapsulated and encapsulated BS. There was no significant variation among these two inoculations, but a considerable difference was shown with the uninoculated control.

Metabolic reprogramming upon application of microbial biostimulant bound by drought stress

The effect of microbial biostimulant on the groundnut seed was visualized via hierarchical clustering with a heat map. The heat map was created for the experimental groups, signifying a clear variance between the metabolites under well-watered condition and drought stress (MD and SD) with the inoculation of encapsulated and non-encapsulated biostimulants. (Figure 8). A profiling analysis identified 52 metabolites spanning sugars, amino acids, fatty acids, ester, ketone, organic acids, and aromatic compounds. The maximum differential metabolites expressed among MD and SD stress were significantly dominated by encapsulated BS than non-encapsulated BS compared to the respective control.

Plants treated with non-encapsulated BS exhibited elevated levels of hexadecadiene, methyl lino, tribromo-6- methylpyridine, propanoic acid, and l-azetidine carboxy under MD and increased levels of metabolites such as pyrrolopyrazine, pyridazinone, and mannosamine, and moderate levels of dibutoxybutane, oxalic acid, mannoic lactone, and formic acid were enriched in SD compared to the unstressed control plants treated with non-encapsulated BS alone that primarily contained oxobutyric acid and 10-undecenal. The treatment with encapsulated BS resulted in a maximum number of expressed metabolites (21 metabolites) with significantly higher concentrations of glycerin, hexenylpyruvate, D-glucose, homoserine, allantoic acid, cystathionine, butyric acid, D-ribopyranoside, mannoic lactone, formic acid, glycine, L-mannose, oxalic acid, propanoic acid, asparagine, 4-butoxybutanol, beta-lactic acid, pyruvic acid, hexadecanoic acid, 4-piperidinecarboxamide, and L-arabinose in MD stress.

Imposing SD, plants treated with encapsulated BS showed enriched levels of dibutoxybutane, isopropylphosphonic acid, benzisoxazole, propanediamine, 2-propanone, homoserine lactone, propanamine, aminobutyric acid, linoleic acid, and heptanoic acid compared to those treated with non-encapsulated BS. A higher accumulation of metabolites including pyridazinone, 10-undecanal, hexadecadiene, pyrrolopyrazine, methyl linolelaidate, and oxobutyric acid were noted in groundnut with the inoculation of non-encapsulated BS. Significant levels of certain metabolites, such as farnesyl acetate, 2-thiophenecarboxylicacid, allantoic acid, cystathionine, butyric acid, 4-butoxybutanol, lactic acid, benzisoxazole, linoleic acid, and heptanoic acid, were identified, and they were recognized as encapsulated-BS-induced drought-responsive metabolites.

Multivariate statistical analysis

For a comprehensive analysis, PLS-DA, a more powerful pattern recognition method by multivariate analysis, was performed. The 2D score plot of PLS-DA (Figure 9) between PC1 and PC2 revealed differential changes in metabolites among three distinct groups (uninoculated, encapsulated, and non-encapsulated-BS-treated samples) associated with drought and control. The PLS-DA results further illustrated that the application of encapsulated and non-encapsulated BS leads to the accumulation of distinct metabolites and discrimination between samples under varied stress conditions.

The VIP (variable of importance in prediction) scores obtained from the PLS-DA identified distinct stress-specific metabolites (VIP > 1) across treatments (uninoculated, encapsulated, and non-encapsulated BS) (Figures 10A, B). These results specified that the metabolomes across treatments in groundnut seed were evidently distinguishable.

KEGG analysis of differentially expressed metabolites

The metabolic pathways and networks impacted by the application of microbial biostimulant under drought were identified using data from the KEGG pathways database, as shown in Figure 11. According to KEGG (Kyoto Encyclopedia of Genes and Genomes), the enrichment and pathway analysis of 52 matched metabolites mapped them to 25 upregulated and enriched metabolisms (Figure 11A). The top 10 most influenced pathways (Figure 11B) of differential metabolites having high statistical significance scores were annotated and enriched in alanine, aspartate, and glutamate metabolism, glycine, serine, and threonine metabolism, butanoate metabolism, propanoate metabolism, cysteine and methionine metabolism, glyoxylate and dicarboxylate metabolism, beta-alanine metabolism, valine, leucine, and isoleucine biosynthesis, galactose metabolism, and linoleic acid metabolism.

The total number of metabolite hits, log p -value, false discovery rate, and pathway impact value of each metabolism are summarized in Table 2. Among these metabolites, a maximum of four metabolite hits was acquired by alanine, aspartate, and glutamate metabolism, glycine, serine, and threonine metabolism, and cysteine and methionine metabolism. In linoleic acid metabolism, only one metabolite was enriched, but the impact value (=1) was higher than those of the other metabolisms. The pathway analysis revealed that differentially expressed metabolites were mainly involved in the biosynthesis of amino acid and the biosynthesis of galactose and linoleic acid. It has been reported that these pathways are involved in plant stress responses.

Microbial biostimulants offer a promising strategy for the mitigation of detrimental effects of drought on agricultural productivity. It enhances plant defense response as well as improves nutrient uptake and metabolic pathways to thrive under adverse stress (Ali et al., 2024). The present research focuses on designing a PGPR-based microbial biostimulant by characterizing the osmo-adaptive features of rhizobacterial strains, its encapsulation with biopolymers, and assessing growth promotion, yield, and seed metabolomic profile in groundnut under induced drought. The rhizobacterial strains Acinetobacter calcoaceticus AC12 and Bacillus amyloliquefaciens BA13 exhibited mechanisms like drought tolerance, phosphate solubilization, and synthesized auxin, siderophore, gibberellin, and EPS. The study was consistent with earlier studies (Upadhyay et al., 2022 and Srivastava et al., 2022), with Acinetobacter and Bacillus strains acting as plant microbial biostimulants. IAA synthesized by rhizobacterial strains is a direct mechanism to influence plant growth and regulate gene expression (Lata et al., 2024). Feng et al. (2024) reported the stimulation of plant cell division and lateral root growth by IAA-producing strains (Bacillus and Pseudomonas).

PGPR synthesizes organic acids to facilitate the solubilization of inorganic phosphate in soil (Zhu et al., 2024). A report suggested that phosphate-solubilizing PGPR (Pseudomonas, Serratia, and Enterobacter) persisted its viability and increased the resilience of peanut under drought (Lopez et al., 2025). Bacterial siderophore chelate Fe 3+ improves its solubility and uptake by the plant (He et al., 2024), showing indirect biostimulation effect. A previous study had reported the role of siderophore-mediated plant growth in Sesuvium sp (Cen et al., 2024). Gibberellic acid is a phytohormone that assists in seed germination and plant growth. Bacillus sp. PG-8 have been reported for the production of gibberellic acid and confer stress tolerance in A. hypogaea (Gohil et al., 2022). A study by Woo et al. (2023) also highlighted the growth-promoting effect of gibberellin-producing bacteria KW05 and KW07 in salty environments. EPS production was triggered in bacteria as a defense response, and previous studies demonstrated that a high EPS secretion of Pseudomonas aeruginosa EPB9 and Bacillus cereus was responsible for bacterial attachment, water retention, and root colonization as well as enabled adaptation against desiccation (Naseem et al., 2024).

The AC12 and BA13 strains produced compatible osmolytes (glycine betaine, trehalose, salicylic acid, alginate, and ectoine), demonstrating their survival under osmotic stress. Glycine betaine synthesis acted as an osmoprotectant, preserving the cell turgor pressure for bacterial viability (Vasconcellos et al., 2021). In accordance with García et al. (2024), Azosprillum strain Az19 exhibited remarkable trehalose production and transcended the stress tolerance in maize. Salicylic acid by PGPR can be allied with the degree of drought tolerance and its endurance in soil (Injamum-Ul-Hoque et al., 2024). Ashry et al. (2022) stated the maximum production of SA by Bacillus strains under the PEG stress. Pseudomonas and Azotobacter sp. are known to produce large quantities of alginate (Núñez et al., 2022), which are hygroscopic and make the bacterial cell hydrated. Alginate production was reported to afford a competitive advantage for bacteria under water scarcity, providing ecological success and sustaining its membrane integrity (Eswaran and Sundaram, 2024). Another compatible osmolyte synthesized by bacteria is ectoine, which has superior osmoprotectant property. In a study, OsLif-2 (Streptomyces) produced a high amount of ectoine in response to osmotic stress (Niu et al., 2022).

The consortium of AC12 and BA13 was used to develop encapsulated microbial biostimulants. In this research, alginate and chitosan were chosen as appropriate carriers for the PGPR biostimulant as they are natural polymers, inexpensive, and biodegradable and maintain bacterial survival in field conditions (Brahmaprakash et al., 2020; Riseh et al., 2023). The results of scanning electron microscopy revealed the morphology of encapsulated microcapsules as evidenced in several investigations (Souza-Alonso et al., 2021), with a diameter of approximately 1–1.5 mm (Zhang et al., 2023). By FTIR spectroscopy, the composition and the functional groups of microcapsules confirmed the presence of sodium alginate, chitosan, and starch with the characteristic spectral bands observed at 3,298, 1,598, 1,415, 1,148, 1,077, and 1,014 cm -1, similar to that cited in the literature (Rani et al., 2023). The study showed that the microbead has a moisture content of 2.87%, ensuring strain stability inside the capsules.

The cross-link between sodium alginate and calcium (CaCl 2) ions resulted in a higher encapsulation efficiency of Bacillus (Mendoza-Labrador et al., 2021), and chitosan coating reduced the porosity in the alginate matrix and minimized the leakage of encapsulated bacteria (Parsana et al., 2023). Furthermore, the prior work of Rani et al. (2023), with the microcapsules of E. ludwigii PS10, showed effective viable bacteria (10 7 CFU) under 90 days of storage, which was analogous to the present study. Consequently, the encapsulated PGPR biostimulant has extreme stability, efficiency, and longer shelf-life, which are major priorities in the development of biostimulant formulations.

The proposed research demonstrated that the encapsulated PGPR consortium stimulated the drought tolerance and biochemical responses of groundnut under field trials. Relative water content is a substantial criterion for observing plant water status, and its decline triggers stomatal closure and causes electrolyte leakage (Manandhar et al., 2024). Plants treated with encapsulated biostimulants showed an increase in RWC than plants without inoculation under MD and SD. Comparable results were reported in Brassica treated with PGPR consortium (Sheoran et al., 2025) under drought. A reduced level of MDA was detected in inoculated plants over non-inoculated plants. These results were aligned with prior findings (Tayyab et al., 2020), showing a favorable correlation between reduced MDA levels and priming against drought by microbial biostimulants. The present study observed a significant enhancement in the protein and soluble sugar content by the inoculation of the encapsulated biostimulant than the inoculated broth solution. Jing et al. (2024) reported an increase in soluble sugar and protein to mitigate drought stress in poplar seedlings with response to PGPR. Prior research documented that groundnut pod yield was boosted by PGPR consortium in both pot and field trials (Bigatton et al., 2024; Alam et al., 2024), while Azospirillum and Bradyrhizobium maximized the yield by 18.52% over the control (Frezarin et al., 2023). These findings corroborated the present work with a substantial rise in groundnut production and seed weight by using encapsulated biostimulants.

In order to understand the underlying molecular defense mechanisms driven by the encapsulated microbial biostimulant, an untargeted metabolomic approach with pathway enrichment analysis was exploited. It revealed the most significantly reconfigured groundnut seed metabolic pathways under drought. They are alanine, aspartate, and glutamate metabolism, glycine, serine, and threonine metabolism, propanoate metabolism, glyoxylate and dicarboxylate metabolism, cysteine and methionine metabolism, valine, leucine, and isoleucine biosynthesis, beta-alanine metabolism, butanoate metabolism, and galactose and linoleic acid metabolism. Amino acids play essential roles in various physiological and biochemical processes to enhance stress resilience as well as the growth and nutritional characteristics (Sun et al., 2024). Metabolic reprogramming in Cicer arietinum L. was noted following the treatment with Bacillus sp (El-Ballat et al., 2023). The research indicated upregulation in amino acid metabolism and biosynthesis, along with other metabolic pathways, after treatments with PGPR.

In alanine, aspartate, and glutamate metabolism, the aspartate family, which includes aspartic acid and asparagine, is essential for plants in combating drought stress. Studies have shown that the build-up of the aspartate family, particularly asparagine, in plants primed with endophytic PGPR improves metabolism in grass during water deficit (Errickson and Huang, 2023). It is noteworthy that glutamate, a precursor for ornithine, arginine, and proline, accumulates significantly in inoculated plants under drought, thereby facilitating protein synthesis and stress recovery (Kasemsap and Bloom, 2022). The present work confirmed that the concentrations of alanine, aspartate, and glutamate were amplified under the treatment of PGPR biostimulants, which aligns with observations reported in other plant species (Mun et al., 2024). Amino acids of glycine, serine, and threonine metabolism are remarkable allies in drought resilience, stomatal governance, osmoprotection, photorespiration, and antioxidant defense systems (Sun et al., 2024; Timm et al., 2025). The vital role of these amino acids against oxidative damage in drought was underscored by recent studies (Ingrisano et al., 2023; Chaturvedi et al., 2024), making them crucial for enhancing seed protein quality and boosting seed and grain yields in crops like chickpea. Our study illustrated that the microbial biostimulant intricately restructured the groundnut seed metabolism through the differential regulation of glycine, serine, and threonine metabolism under stress.

In the context of cystine and methionine metabolism, it is highly susceptible to various reactive oxygen species, highlighting their antioxidant properties (Mehari et al., 2025). Inoculation of the consortium of PGPR influenced the abundance of several amino acid metabolisms, including cysteine and methionine metabolism, under reduced irrigation in Salvia officinalis (Jaoudé et al., 2025). In valine, leucine, and isoleucine metabolism, the amino acids participate in the drought-resistant pathways of groundnut, and the inoculation of encapsulated biostimulants further influenced its accumulation. These amino acids lower osmotic pressure and sustain cell turgor pressure and integrity (Ma et al., 2020). In previous research, the strain DGL-1 (Bacillus amyloliquefaciens) showed an upregulation trend in terms of the content of valine, leucine, and isoleucine under drought (Yang et al., 2024) in Medicago sativa and M. truncatula (Shen et al., 2025).

In butanoate metabolism, the inoculation of biostimulants upregulated the abundances of 4-aminobutyric acid, succinic acid, malic acid, and maleic acid. These accumulated metabolites of butanoate metabolism and TCA cycle maintain ATP production and metabolic regulation under stress (MacLean et al., 2023). In a former study, PGPR Pantoea eucalyptus (B13) inoculation in Leymus chinensis upregulated the gene expression of butanoate metabolism (Yuan et al., 2025). The other pathways with significant contribution in drought tolerance include glyoxylate and dicarboxylate metabolism and propanoate metabolism. These altered metabolic pathways, along with their intermediates and products, provide a framework of drought tolerance in groundnut under biostimulant inoculation. Acetyl-CoAs from glyoxylate cycle are converted into malate and succinate, providing energy for life activities and for amino acid synthesis (Wang et al., 2023). Bacillus amyloliquefaciens -strain-SN13-inoculated seedlings exhibited increased levels of butanoic acid and propanoic acid as well as glyoxylate and dicarboxylate metabolism when simultaneously exposed to nutrient deficiency and drought stress (Bisht et al., 2020).

Galactose metabolism links with energy conversion activities, supplies the energy substrate (glucose) for TCA metabolism (Zhi et al., 2024; Wang et al., 2022), supports membrane stability, and preserves osmotic equilibrium in drought. In this scientific study, the encapsulated-biostimulant-inoculated plants expressed elevated seed sugar concentrations, including glucose, galactose, fructose, ribose, and mannose, confirming that the biostimulant regulates and improves plant nutrition under stress. It was found that the plants inoculated with the B. amyloliquefaciens SN13 strain have enhanced xylose, mannose, and trehalose under salt and drought stress (Joshi et al., 2023). Our research work demonstrated that in fatty acid metabolism, different expression levels of unsaturated fats such as dodecanoic acid, linolenic acid, methyl linolelaidate, heptanoic acid, cyclohexanedione, and octadecadienoic acid in groundnut inoculated with encapsulated biostimulant under stress were noted. Plants preserve their integrity by fluctuating the cell membrane lipids during dehydration. Increased unsaturated fat levels correlated with drought resistance in plants (Dutta et al., 2025). Metabolic fingerprinting of olive plant showed an accumulation of 3,4-dimethyphenol, octadecatrienoic acid, benzoic acid, and other fatty acids by bacterial consortium under drought (Azri et al., 2024).

To develop an effective microbial biostimulant product, we analyzed all of the required parameters such as rhizobacterial drought-tolerant ability, PGP traits, and osmolyte production. To further extend their shelf-life, viability, and stability, the microbial biostimulants were encapsulated and characterized. A metabolomic analysis on the tripartite interaction (biostimulant–groundnut–drought) revealed that the encapsulated microbial biostimulant reprogrammed the groundnut seed metabolites to cope with drought stress. Specifically, 25 upregulated metabolic pathways participate in the regulation of plant nutrition and stress responses. These metabolomic findings attempt to decipher the precise action of encapsulated microbial biostimulants and contribute a scientific background for the manufacturing of biostimulant products for plants. This work deserves further attention and in-depth functional study through integrated multi-omics approach.