Sample collection and isolation of Bacillus from rhizosphere soil
Rhizosphere soils were collected from wheat crop during the months of February–March 2017 from the four different sites of agricultural field in Roopsi and Sam villages of Jaisalmer district, Rajasthan India (26.9107°N & 70.9144°E and 26.9798° N, 70.7525° E), respectively. Crop plants of wheat were selected and uprooted, with the soil adhered to the roots, from each site and packed in zip locked polybags and transported to the laboratory. The rhizosphere soil was collected by removing the adhered soil from each plant. The putative Bacillus isolates were recovered from rhizosphere soil on nutrient agar (NA) medium supplemented with 5% PEG-6000 using serial dilution method. Briefly, 0.1 ml aliquot of each dilution was placed on Petri plates containing nutrient agar followed by incubation at 37°C for 48 h. The visually distinct colonies were picked up and grew in nutrient broth in order to get pure colony of bacteria. Glycerol stocks (20%) of each bacterium was prepared and stored at −80 °C in a deep freezer. All the subsequent experiments were carried out using fresh cultures.

In vitro evaluation of drought tolerance potential of Bacillus isolates
On solid medium using PEG
Bacterial isolates were grown on NA medium supplemented with varied concentration of PEG-6000 (5–30%) in order to select the most tolerant stain to drought conditions, simulated by osmotic pressure created by addition of PEG- 6000 followed by incubation of plates at 28 ± 2 °C for 48 h. The growth of bacteria on PEG-6000 supplemented media was recorded. Isolates that survived at highest concentration of PEG-6000 concentration were selected for further studies.^18,19

In broth medium using PEG
Bacterial isolates were grown on nutrient broth (NB) supplemented with 10%, 20%, 30% PEG 6000 equivalents to -1.37 MPa (Megapascal), -4.64 MPa, -9.80 MPa osmotic potential, respectively. The nutrient medium supplemented with PEG-6000 was inoculated with 1 ml of 10⁸ cfu ml^-1 freshly grown culture followed by incubation at 28 ± 2 °C up to 72 h. Optical density (O.D.) of the culture was taken at 600nm. Lowry et al²⁰ method was used to determine the protein concentration Uninoculated medium was used as negative control.

Identification of putative Bacillus isolates
Two putative bacilli were identified at species level by morphological, biochemical,²¹ and 16S rRNA gene sequencing methods. The sequencing of 16S rRNA gene of two bacilli was performed using universal primers 27F1 (5- AGAGTTTGATCMTGGCTCAG-3) and 1494R
(5-CTACGGCTACCTTGTTACGAC-3) at National Center for Microbial Resources (NCMR), Pune, India. The 16S rRNA gene sequences of isolates DT-89 and DT-113 were matched with type strains employing EzTaxon server²² and aligned by CLUSTAL W in MEGA version 7²³ software. The topology of evolutionary tree was evaluated by a bootstrap analysis²⁴ of the N-J method based on 1000 replicates by MEGA 7 software. The processed nucleotide sequences data were submitted in the NCBI GenBank to acquire accession number. Finally, the cultures of the Bacillus altitudinis DT-89 (NAIMCC-B-02229) and Bacillus paramycoides DT-113 (NAIMCC-B-02943) were deposited in NAIMCC (National Agriculturally Important Microbial Culture Collection), ICAR-NBAIM, Mau, Uttar Pradesh, India.

In vitro assessment of plant growth-promoting traits of Bacillus spp. under stress conditions
Plant growth promoting attributes of selected strains were studied at three levels of drought stress (10%, 20%, and 30% PEG-6000) and amended in all the media of the following tests. The bacterial strains were functionally characterized based on production of IAA, GA3, exopolysacharide, siderophore, hydrogen cyanide, ammonium production and ACC deaminase; solubilization of zinc, phosphorus, and antagonisms against phytopathogens by using standard procedures. Bacterial strains were tested for IAA production by procedure given by Ahmad et al.²⁵ Bacterial stains were inoculated separately in 10 ml Luria Bertani (LB) broth in 30 ml tubes. Ten (10) ml LB broth containing supplemented with 200 µg ml^-1 L-tryptophan and each tube was inoculated with 48 h grown 100 µl (10⁸ ml^-1) of culture followed by incubation at 28±2 °C for 4 days with shaking at 125 revolutions per minute (rpm) under normal and osmotic stress conditions. After incubation, a 5 ml culture of each bacterial strain was centrifuged at 9,000 rpm for 15 min. Four (4) ml of Salkowski’s reagent (2% 0.5M FeCl₃ in 35% perchloric acid) along with100 μl of orthophosphoric acid were added to 2ml of supernatant followed by incubation at 28±2°C in darkness for 1h. The development of pinkish to red colour indicated indole -3-acetic acid (IAA) production. Gibberellic acid (GA3) production by two strains was carried out under normal and osmotic stress conditions as per method of Burrows et al.²⁶ Briefly, strains DT-89 and DT-113 were grown in LB broth and GA3 was estimated by extracting it with ethyl acetate and measuring absorbance at 680 nm spectrophotometrically. Exopolysaccharide production by two strains was estimated by gravimetric method.²⁷ For siderophore production assay, bacterial culture was spot inoculated on Chrome Azurol S (CAS) agar and incubated at 28±2 °C for 3 days. The appearance of yellow-orange halo around bacterial spot indicates of siderophore production.²⁸ Quantitative estimation of siderophore was done according to the method of Payne et al.²⁹ Zinc solubilization was assayed by streaking bacteria on Tris minimal agar medium supplemented separately with 0.1% Zn as zinc oxide, zinc phosphate and zinc carbonate sources. Clear zone around bacterial colony was indicative of solubilization. For quantification of soluble Zn, quantification of soluble Zn from three different insoluble sources solubilized by bacteria in Tri-minimal broth was assayed by using AAS using method described by Fasim et al.³⁰ The bacterial strains were evaluated for phosphate solubilization on Pikovskaya agar using tricalcium phosphate as an insoluble source of phosphorus according to the standard method of Pikovskaya.^31,32 The qualitative method of Kremer and Souissi³³ was used to determine HCN production. Ammonia production was assayed by addition of 1 ml Nesseler’s reagent to 72h grown cultures in peptone broth. The positive samples showed yellowish brown colour is an indicative for ammonium production. The ACC deaminase activity was studied as per the method described by Govindaswamy et al.³⁴ The DT-89 and DT-113 strains were cultured in 5 ml tryptone soya broth separately and incubated at 28±2 °C for 24h at 120 rpm. The cell pellet was harvested by centrifugation at 6000 rpm for 6 min followed by washing two times with sterile distilled water to remove traces of the medium. The washed pellet was suspended in 1 ml sterile distilled water. The bacterium was spotted on Dworkin Foster (DF) minimal medium plates amended with 3mM of ACC, DF medium alone, a negative control, and DF medium containing ammonium sulphate, as a positive control. All the plates were observed after incubation at 28±2 °C for 96 hrs. The bacterium with prominent growth on ACC-DF was indicative of ACC deaminase production. Antagonistic potential of two strains was performed on PDA against Rhizoctonia solani, Macrophomina phaseolina, Fusarium oxysporum f.sp. cumini, F. oxysporum f. sp. carthami and Sclerotium rolfsii employing dual confrontation plate technique.

Plant growth promotion and drought tolerance potential of Bacillus spp. in pot culture
Two potential drought-adaptive bacteria strains namely DT-89 and DT-113 were taken for evaluation for their growth-promoting and drought tolerance inducing ability on wheat crop (Var. HD-2967) in non-sterile soil under net house conditions. Inoculum of strains DT-89 and DT-113 was added in half-strength NB and incubated at 28°C for 36 h and pellet of bacteria was obtained after centrifugation at 6000 rpm for 12 min. Bacterial suspension was prepared in 0.85% aqueous saline solution of 0.5 OD equivalent to population of 2.3 x 10⁸ cfu ml^-1. Subsequently, seeds of wheat were surface sterilized using 1.0% sodium hypochlorite by exposing it for 30 sec and kept for some time to evaporate trace of sodium hypochlorite. Approximately, 50 g surface-sterilized seeds were soaked in 50 ml of 0.5 OD Bacillus species strains suspensions for one hour and then treated seeds were inoculated in the pots and kept in nethouse conditions in the winter of 2019 in Farm of ICAR-National Bureau of Agriculturally Important Microorganisms, Mau. Each pot was containing 10 Kg of sieved (2 mm sieve) alluvial soil and was not fertilized with chemical fertilizers. In total, there were three treatments, (i) Absolute control without any bacterization, (ii) bacterization with DT-89 strain and (iii) bacterization with DT-113 strain with two conditions watered and stressed (50% – that is half dose of water). All the treatments were replicated three times and total 18 pots were used for two stages of sampling. Initially, each pot was sown with 6 seeds for germination and finally 3 seedlings were maintained for each pot throughout experimentation. After germination, plants were maintained underwater and water stress conditions (50%) throughout experimentation. Plants were harvested after 45 and 75 days, and the soil particles adhered on the roots was removed by washing with tap water. Phenotypic (root and shoot length fresh and dry weight) parameters of the harvested plants were recorded. The leaf samples were preserved at 4 °C for various biochemical estimations like ROS scavenging enzyme activities, proline and chlorophyll content. Chlorophyll content was determined in 80% acetone extract. The absorbance was read spectrophotometrically at 663 and 645 nm. Total chlorophyll concentrations were calculated as per method of Arnon.³⁵ Other parameters were also analysed as mentioned below.

Estimation of ROS scavenging stress marker enzymes and proline content
Assays of superoxide dismutase (SOD)³⁷ and peroxidase (POD)^38,39 activity were performed utilizing supernatant obtained by centrifuging 500mg of leaves in 1.5 ml of 0.1 M phosphate buffer (pH 7) at 18,000 g, 5°C, 15 minutes.

SOD activity was assayed by measuring the inhibition of nitroblue tetrazolium (NBT) by SOD enzyme. Three (3ml) reaction mixture contained 100 µL of crude extract, 50 mM sodium phosphate buffer (pH 7.6), 0.1 mM EDTA, 50 mM sodium carbonate, 12 mM L-methionine, 50 µM NBT and 10 µM riboflavin. The reaction mixture without crude extract and irradiation served as blank, whereas SOD reaction was carried out by exposing the reaction mixture to white light for 15 min at room temperature. After 15 min incubation, absorbance was recorded at 560 nm using a spectrophotometer. One unit (U) of SOD activity was defined as the amount of enzyme causing 50% inhibition of photochemical reduction of NBT.

POD activity was determined by measuring the absorbance changes at 436 nm at 25°C. The rate of formation of GDHP (guaiacol dehydrogenation product) in the guaiacol assay is a measure of the POD activity. The extinction coefficient for GDHP at 436 nm is 6.39 cm² per /µmole guaiacol oxidized.³⁸ Reaction mixture contained 3 ml of the phosphate buffer solution (0.1M), 0.05 ml guaiacol solution, 0.1 ml enzyme extract and 0.03 ml hydrogen peroxide solution in a cuvette. The mixture was well shaken and placed in the spectrophotometer. The time required for the mixture optical density to be increased by 0.1 (Δt) at 436 nm was recorded. In this contribution, the resulting end product is called GDHP (guaiacol dehydrogenation product). Free proline was determined as per Bates et al.⁴⁰ Briefly, 200 mg of leaf sample was homogenized with 3% (w/v) aqueous sulfosalicylic acid and the supernatant obtained was treated with glacial acetic and ninhydrin at 100°C followed by cooling. Now, the chromophore was extracted with toluene by addition in the reaction mixture and finally reading was taken at 520 nm. Proline content was expressed as mg g^-1 fresh weight basis.

Statistical analysis
All the experiments were performed in triplicates. Statistical analysis was done using Prism GraphPad (Version 9, CA USA) for obtaining P values. Analysis of variance (ANOVA) was performed and the means were compared with Tukey’s test at P<0.05, P<0.001, P<0.0001. Microsoft Office Excel 2007 was used for graphics and determination of standard deviation.