Integrated Metabarcoding and Culturomic-Based Microbiome Profiling of Rice Phyllosphere Reveal Diverse and Functional Bacterial Communities for Blast Disease Suppression

Phyllosphere—the harsh foliar plant part exposed to vagaries of environmental and climatic variables is a unique habitat for microbial communities. In the present work, we profiled the phyllosphere microbiome of the rice plants using 16S rRNA gene amplicon sequencing (hereafter termed metabarcoding) and the conventional microbiological methods (culturomics) to decipher the microbiome assemblage, composition, and their functions such as antibiosis and defense induction against rice blast disease. The blast susceptible rice genotype (PRR78) harbored far more diverse bacterial species (294 species) than the resistant genotype (Pusa1602) that showed 193 species. Our metabarcoding of bacterial communities in phyllomicrobiome revealed the predominance of the phylum, Proteobacteria, and its members Pantoea, Enterobacter, Pseudomonas, and Erwinia on the phyllosphere of both rice genotypes. The microbiological culturomic validation of metabarcoding-taxonomic annotation further confirmed the prevalence of 31 bacterial isolates representing 11 genera and 16 species with the maximum abundance of Pantoea. The phyllomicrobiome-associated bacterial members displayed antifungal activity on rice blast fungus, Magnaporthe oryzae, by volatile and non-volatile metabolites. Upon phyllobacterization of rice cultivar PB1, the bacterial species such as Enterobacter sacchari, Microbacterium testaceum, Pantoea ananatis, Pantoea dispersa, Pantoea vagans, Pseudomonas oryzihabitans, Rhizobium sp., and Sphingomonas sp. elicited a defense response and contributed to the suppression of blast disease. qRT-PCR-based gene expression analysis indicated over expression of defense-associated genes such as OsCEBiP, OsCERK1, and phytohormone-associated genes such as OsPAD4, OsEDS1, OsPR1.1, OsNPR1, OsPDF2.2, and OsFMO in phyllobacterized rice seedlings. The phyllosphere bacterial species showing blast suppressive activity on rice were found non-plant pathogenic in tobacco infiltration assay. Our comparative microbiome interrogation of the rice phyllosphere culminated in the isolation and identification of agriculturally significant bacterial communities for blast disease management in rice farming through phyllomicrobiome engineering in the future.

Phyllosphere-the harsh foliar plant part exposed to vagaries of environmental and climatic variables is a unique habitat for microbial communities. In the present work, we profiled the phyllosphere microbiome of the rice plants using 16S rRNA gene amplicon sequencing (hereafter termed metabarcoding) and the conventional microbiological methods (culturomics) to decipher the microbiome assemblage, composition, and their functions such as antibiosis and defense induction against rice blast disease. The blast susceptible rice genotype (PRR78) harbored far more diverse bacterial species (294 species) than the resistant genotype (Pusa1602) that showed 193 species. Our metabarcoding of bacterial communities in phyllomicrobiome revealed the predominance of the phylum, Proteobacteria, and its members Pantoea, Enterobacter, Pseudomonas, and Erwinia on the phyllosphere of both rice genotypes. The microbiological culturomic validation of metabarcoding-taxonomic annotation further confirmed the prevalence of 31 bacterial isolates representing 11 genera and 16 species with the maximum abundance of Pantoea. The phyllomicrobiome-associated bacterial members displayed antifungal activity on rice blast fungus, Magnaporthe oryzae, by volatile and non-volatile metabolites. Upon phyllobacterization of rice cultivar PB1, the bacterial species such as Enterobacter sacchari, Microbacterium testaceum, Pantoea ananatis, Pantoea dispersa, Pantoea vagans, Pseudomonas oryzihabitans, Rhizobium sp., and Sphingomonas sp. elicited a defense response and contributed to the suppression of blast disease. qRT-PCR-based gene expression analysis indicated over expression of defense-associated genes such as OsCEBiP, OsCERK1, and phytohormone-associated genes such as OsPAD4, OsEDS1, OsPR1.1, OsNPR1, OsPDF2.2, and OsFMO in phyllobacterized rice seedlings. The phyllosphere bacterial species showing blast suppressive activity on rice were found non-plant

INTRODUCTION
Microbial communities have an evolutionary association with plant populations where they function as metaorganisms in the natural environment. Here, the microbial activities in total termed microbiomes play a pivotal role in plant development and survival (Hartmann et al., 2008;Bulgarelli et al., 2013;Berg et al., 2016). The microbial communities associated with the plants are called plant microbiome and microbial metagenome that often confer functional flexibility to the plant genome (Sessitsch et al., 2012). The plant microbiomes are presumed to modulate a variety of plant functions. However, the ecological role of the phyllosphere microbial communities on plant functional ecology is among the most understudied and underrated aspects in plant biology.
The microbial life on the foliar niche, the phyllosphere microbiome, is constantly exposed to vagaries of weather events, and other agronomic practices in crop husbandry (Lindow and Leveau, 2002). Currently, the epiphytic phyllosphere microbiomes and their natural functions are increasingly investigated in crops like rice, wheat, maize, and soybean (Andrews and Harris, 2000;Bertani et al., 2016;Compant et al., 2019;Sahu et al., 2020). It is further reported that plantassociated bacteria are prolific for the secretion of primary and secondary metabolites and volatiles for plant growth, developmental regulation, and defense against stresses Rascovan et al., 2016;Sheoran et al., 2016;Gómez Expósito et al., 2017;Eke et al., 2019;Ashajyothi et al., 2020;Vandana et al., 2021).
Rice is the primary staple for the nearly three billion world population and contributes to global food security. Rice production is affected by several biotic and abiotic stresses; among them, blast disease caused by ascomycetous fungus Magnaporthe oryzae (anamorph Pyricularia oryzae Sacc.) is responsible for nearly 30.0% of losses, which can feed 60 million population if prevented preemptively (Dean et al., 2005;Scheuermann et al., 2012;Yasuda et al., 2015;Hashim et al., 2018;Mehta et al., 2019;Prakash et al., 2021). Deployment of fungicides and blast-resistant cultivars are among the blastcombating strategies widely practiced. However, both strategies are under scanner as the chemicals are no longer encouraged due to safety considerations, and the host resistance is not durable (Nalley et al., 2016;Asibi et al., 2019;Sella et al., 2021). Recently, the fungicide residues intercepted on the Indian rice imports have prompted many countries to reject consignments from international trade (Al-Antary et al., 2020). Under this scenario, there is a growing demand among the various stakeholders of the rice production system for an alternative blast mitigation strategy. One promising yet unexplored strategy is the biological control of blast disease by deploying leaf microbiota that shares the same microecological niche with the blast pathogen, M. oryzae. So far, the potential of phyllosphere microbial communities sharing the leaf microhabitat with Magnaporthe-the incitant of blast disease has not yet been harnessed to mitigate the disease in any crop. Hence, the present investigation was carried out to explore the phyllomicrobiome of rice and exploit them for blast disease suppression by microbiome reengineering.

Study Location and Sampling for Microbiome Analysis
The 16S rRNA gene amplicon sequencing by NGS (hereafter cited as metabarcoding; Berg et al., 2020), combined with conventional culturomic investigation of phyllomicrobiome, was conducted. For this, we planted rice genotypes in a blastendemic mountain ecosystem in Palampur, Himachal Pradesh, India (32 • 6 4.7 N and 76 • 32 39.79 E) located at an altitude of 1,275 m above mean sea level [weather conditions: mean temperature 22-23 • C; precipitation 700-1,000 mm; RH 60.0%; sunshine hours 300-350; source: https://en.climate-data.org; www.worldweatheronline.com]. The rice genotypes were grown during the rice-growing season in August to September 2014. Briefly, rice genotypes PRR78-a blast susceptible variety, and Pusa1602-a near-isogenic line of PRR78 introgressed with Pi2 gene conferring complete resistance to blast disease (Singh et al., 2012), were planted in parallel rows with a spacing of 20 cm by adopting all crop husbandry practices. Leaf samples were excised at 15 and 30 days post-sowing in sterilized falcon tubes and brought to the laboratory in an insulated cool container maintained at a temperature of 4.0 ±1.0 • C and processed for microbiome analysis by metabarcoding and culturomic analysis. PBS-T by agitating for 30 min at 250 rpm and vortexing for 10 s. Thus, the collected epiphytic-microbial suspension (300 ml) was collected aseptically in a presterilized container and centrifuged at 12,000 × g for 60 min at 4.0 • C to collect the epiphytic microbial cells. Thus, the obtained pellet was processed to isolate genomic DNA by the cetyltrimethyl ammonium bromide (CTAB) method reported by Moore et al. (2004) with slight modifications like avoidance of phenol in the extraction steps. The quality and quantity of microbial community genomic DNA was determined electrophoretically, spectrophotometrically (Nanodrop 2000, Thermo Scientific, United States), and fluorometrically using Qubit dsDNA BR Assay (Thermo Fisher Scientific Inc., United States).

Preparation of Sequencing Libraries for 2 × 300-bp Run Chemistry
The amplicon libraries were prepared using Nextera XT Index Kit (Illumina Inc.) as per the 16S rRNA Gene Amplicon Sequencing Library Preparation Protocol (Part no. 15044223 Rev. B). PCR primers for the amplification of the 490-bp hypervariable region of V3-V4 of 16S rRNA gene of Eubacteria were designed, synthesized, and used. The sequences of the primers are V3F: 5 CCTACGGGNGGCWGCAG3 and V4R: 5 GACTACHVGGGTATCTAATCC3 . The target amplicons were generated using the fusion primer that consists of Illumina adaptors and multiplex index sequence as per the instructions of the manufacturer. The amplicon libraries were purified by 1× AMpure XP beads and checked on Agilent High Sensitivity (HS) chip on Bioanalyzer 2100 and quantified on fluorometer by QubitdsDNA HS-Assay kit (Life Technologies, United States). Quality-passed libraries were equimolar pooled and then sequenced using the IlluminaMiSeq platform with 2 × 300-bp paired-end sequencing chemistry following the protocols of the manufacturer (Illumina, San Diego, CA, United States).

Bioinformatic Analyses
Initially, the sequenced raw forward reads (R1) and reverse reads (R2) were scanned and analyzed using the FastQC version (Andrews, 2018) to assess the quality of 16rRNA amplicon reads. Thus, the obtained raw reads were end trimmed and curated using Trimmomatic v0.35 (Bolger et al., 2014) with the following command and settings: (i) remove adapter sequences, (ii) ambiguous reads (reads with unknown nucleotides "N" larger than 5.0%), and (iii) low-quality sequences [reads with more than 10.0% quality threshold (quality value) <20 Phred score]. The resultant quality-passed read pairs were joined using PEAR (Paired-End reAdmergeR) version 0.9.8  with default parameters. The joined paired reads were further processed for downstream taxonomic classification where unpaired reads were discarded. The taxonomic classification of resultant high-quality reads was performed using MG-RAST v4.0, wherein (i) 16S rRNA gene sequence reads were sorted using Sortme RNA, (ii) sorted reads were clustered at ≥97% similarity using CD-HIT method, and (iii) clustered reads were taxonomically classified using SILVA SSU database. The clustered reads and taxon abundance downloaded >100 bases and 90.0% similarity through the best hit classification. Furthermore, PAST v2.17c (Hammer et al., 2001) was used for the determination of αdiversity.

Isolation and Characterization of the Culturable Epiphytic Microbiome
Another set of leaves (500 mg) excised from the rice genotypes were analyzed using the culturomic method on nutrient agar medium [NA, g L −1 ; peptone 5.0; beef extract 3.0; NaCl 5.0; agar 15.0; pH 7.0 ± 0.2]. Briefly, the leaf was agitated with 50 ml of sterile phosphate buffer saline amended with 0.1% Tween-20 (PBST) for 30 min at 250 rpm followed by vortexing for 10 s; the aliquot, thus, obtained was decimally diluted up to 10 −5 . An aliquot of 1.0 ml at 10 −3 , 10 −4 , and 10 −5 from each sample was pour plated in nutrient agar media supplemented with 2,3,5-tetrazolium chloride (50 mg L −1 ) and incubated at 28 ± 2 • C for 72 h for morphotyping the bacterial colonies. The culturable bacterial population and their diversity were assessed and quantified based on morphological traits, such as size, shape, color, texture, and margin. The pure culture of the representative isolates was preserved in −80 and −20 • C as glycerol stock (30% V/V) for downstream work.

Identification of Epiphytic Bacterial Species by 16S rRNA Gene Sequencing
Genomic DNA was isolated by the CTAB method described by Moore et al. (2004) with minor modifications as mentioned previously. Isolated and purified genomic DNA was quantitated and quality analyzed as described above. Finally, the genomic DNA reconstituted at 100 ng µl −1 was used as a template in PCR amplification. Box PCR-based DNA fingerprinting was performed for diversity analysis as well as to eliminate the duplicate isolates from the collection (Versalovic et al., 1994); this PCR-based DNA profiling technique specifically amplifies the non-coding conserved sequences in the bacterial genome and is considered a highly discriminatory DNA-fingerprinting technique (Kumar et al., 2004;Eke et al., 2019). Amplicon profiles were resolved in 1.0% agarose gel at 30 V for 10-12 h and imaged (QuantityOne, BioRad, United States). Isolates showing identical amplicon profiles were presumed to be duplicates. PCR amplification of the 16S rRNA gene was performed using primer sets, 27F (27F: 5 -AGAGTTTGATCCTGGCTCAG-3 ) and 1492R (1492R: 5 -GGTTACCTTGTTACGACTT-3 ), to amplify the 1,465-bp region (Sheoran et al., 2015;Munjal et al., 2016). Then the PCR amplicons resolved in the agarose gel (1.0 %) were purified and eluted using an elution kit according to the instructions of the manufacturer' (Promega Corporation, United States). The amplicons were sequenced bidirectionally to achieve maximum coverage of the sequences and analyzed using the nucleotide-Basic Local Alignment Search Tool (BLAST) algorithm at the National Center for Biological Information (NCBI); the bacterial species identity was confirmed by closest match. The diversity analysis of culturable phyllosphere bacterial species was calculated using Shannon diversity indices.

Activity Screening of Epiphytic Phyllospheric Bacteria on Magnaporthe oryzae
Antifungal Activity in vitro Airborne volatile organic compounds emitted, and diffusible metabolites secreted, during bacterial growth were tested for antifungal activities on M. oryzae. Here, we conducted dualculture confrontation assays, and mycelial inhibition over mock was calculated as described (Sheoran et al., 2015;Munjal et al., 2016). Additionally, the fungicidal or fungistatic nature of the antifungal activity of volatiles on M. oryzae was also determined. Here, the volatile exposed mycelia of M. oryzae showing complete inhibition were further incubated after replacing bacterial volatile with lid. Depending on the mycelial growth, the bacterial volatile was either categorized as fungicidal or fungistatic on M. oryzae . The radial mycelial growth of M. oryzae was measured, and mycelial inhibition (%) over mock was calculated using the following formula where I = percent inhibition C = colony diameter in control T = colony diameter in treatment

Blast-Suppressive Activity in planta
The bacterial isolates showing inhibition of mycelial growth of M. oryzae were selected for in planta blast control assay.
Disease severity = (Scale × Number of plants infected) × 100 Total number of plants × Maximum disease scale Furthermore, the percent reduction in disease severity compared with control was estimated using the following formula: Reduction in Blast Severity = C−T C × 100 C = disease severity in control T = disease severity in treatment

Phenotyping for Phyllomicrobiome Conferred Immunocompetence
The bacterial isolates showing antifungal activity on M. oryzae were selected for the immunocompetence assay. Here, the germination and phenotypic alterations induced on rice seedlings by bacterial isolates were monitored and scored. In this assay, the blast susceptible Pusa Basmati 1 was subjected to germination for 5 days in the presence of bacterial cells at varying densities such as 10 6 , 10 7 , 10 8 , and 10 9 (CFU mL −1 ). Similarly, seeds germinated in sterile double distilled water served as control. The experiment was performed in three replications with 50 seeds in each replication and repeated twice. The seed germination (%) was calculated using the following formula to evaluate the effect of bacterial interaction on germination.
Seed germination (%) = Number of germinated seeds Total number of seeds × 100 To test the effect of bacterial colonization on the shoot and root growth, five randomly selected seedlings were scored. The percent deviation in shoot and root growth was calculated against the untreated mock seedlings using the following formula. Here, while the negative value indicated growth inhibition, the positive score indicated growth promotion.
where G = percent growth of shoot/root C = length of shoot/root in control T = length of shoot/root in the treatment

Assessment of Immunocompetence by qPCR
Having confirmed the blast-suppressive potential of the phyllosphere bacterial species, we performed qPCR experiments to decipher the effect of bacterial supplementation on the expression of genes involved in defense pathways in rice. A total of nine phyllosphere bacterial isolates such as Pantoea vagans OsEp-Plm-30B3, Pantoea ananatis OsEp-Plm-15B6, Enterobacter sacchari OsEp-Plm-15B10, P. ananatis OsEp-Plm-30B17, Pantoea dispersa OsEp-Plm-15B14, Rhizobium sp. OsEp-Plm-30B4, Microbacterium testaceum OsEp-Plm-30B1, Pseudomonas oryzihabitans OsEp-Plm-15B16, and Sphingomonas sp. OsEp-Plm-15B2 that showed blast-suppressive activity on rice was chosen for the study. Briefly, whole seedlings of Pusa Basmati 1, bacterized with 2 × 10 7 CFU m −1 and sampled at 24-h interval for three consecutive days were immediately snap frozen in liquid nitrogen to stop all the cellular metabolic activity and then stored instantly at −80 • C until further use. The total RNA was isolated using the SV Tool RNA isolation system according to the instructions of the manufacturer (Promega, Madison, WI, United States). The quality and quantity of RNA were assessed spectrophotometrically (NanoDrop 2000, Thermo Scientific, United States) and electrophoretically.

Transcriptional Analysis of Genes Associated With Immunocompetence
Eight rice genes, such as OsCEBiP (Akamatsu et al., 2013), OsCERK1 (Kouzai et al., 2014), OsPAD4 (Ke et al., 2014), OsEDS1 (Ke et al., 2019), OsNPR1 (Sugano et al., 2010), OsPDF2.2 (Thomma et al., 2002), OsFMO1 (Koch et al., 2006;Mishina and Zeier, 2006), and OsPR1.1 (Breen et al., 2017), which were reported to play a role in rice defense, were selected; PCR primers targeting the above defense genes are furnished in Supplementary Tables 1, 2. qPCR was performed in Real-Time Thermal Cycler (LightCycler 96, Roche Life Science, Switzerland) using GoTaq R 1-Step RT-qPCR System (Promega Corporation, United States); qPCR reaction conditions were as follows: one cycle of reverse transcription at 37 • C for 15 min followed by reverse transcriptase inactivation step at 95 • C for 10 min followed by 30 cycles at 95 • C for 10 s, annealing at 58 • C for 30 s and extension at 72 • C for 30 s followed by three-step melting at 95 • C for 10 s, 63 • C for 60 s, and 97 • C for 1.0 s, and then a final cooling at 37 • C for 30 s. Later, cyclic threshold data points were analyzed for the determination of gene expression relative to the reference housekeeping OsActin gene using the software LightCycler R 96 Roche. The mean Ct values were considered for the calculation of 2 − CT to estimate the fold changes in gene expression.

Hypersensitive Reaction on Tobacco
Upon bacterial infiltration on tobacco leaves, potential plant pathogenic bacteria are known to induce hypersensitive reactions (HR) (Klement, 1963); this is considered as a test to ascertain the plant pathogenic nature of bacterial isolates. The best performing nine bacterial isolates for suppression of blast disease were selected for this assay. Tobacco (Nicotiana tabacum) plants were grown under greenhouse conditions at 20 • C, 50-60% relative humidity, and 12/12-h light/dark per day. Fully expanded leaves of 2 to 3-month-old plantlets were used in all experiments. Bacterial inoculum (1.0 × 10 8 CFU ml −1 , absorbance at 600 nm = 1.0 OD) was infiltrated onto the leaves using a sterile hypodermal syringe. Thus, treated plants were incubated at 25-30 • C under greenhouse conditions (12/12 h of dark/light photoperiods). Similarly, leaves infiltrated with sterile distilled water alone served as a negative control, and a well-known bacterial pathogen, Ralstonia solanacearum, served as a positive control. Plant responses to the bacterial infiltration on tobacco leaves were recorded after 24-h post-inoculation.

Statistical Analysis
All datasets were analyzed using the data analytical tool available in MS Office Excel 2013. The analyzed data obtained were subjected to significance testing by analysis of variance (ANOVA) at a p ≤ 0.05 level of significance. Furthermore, various parameters like the standard error of the mean (SEm), standard error of the difference between two means (SEd), critical difference (CD), and coefficient of variation (CV) were estimated. For figures and tables, the values are represented as the mean of all biological and technical replicates. For the qPCR data analysis, the fold change values calculated for the defense genes were imported into the GraphPad Prism program (https://www. graphpad.com/scientific-software/prism), and two-way ANOVA was conducted using Bonferroni post-hoc test for determining the statistical significance at * p ≤ 0.05, * * p = 0.001, and * * * p = 0.0001.

16S rRNA Barcode Sequence Read Statistics and Diversity Indices
Phyllomicrobiome profiling of blast-susceptible (PRR78) and blast-resistant rice (Pusa1602) genotypes planted in blastendemic locations was conducted using integrated 16S rRNA gene amplicon sequencing and microbiological methods. The total curated sequence generated is in the range of 4, 31,222 for Pusa1602 and 1, 81,250 for PRR78 reads ( Table 1). The αdiversity indices (Shannon diversity) were 1.25 for Pusa1602 and 1.85 for PRR78. Other diversity indices were also marginally higher for PRR78 than Pusa1602 revealing that the blastsusceptible genotype harbored more diverse microbial species than the resistant type ( Table 1).

Enumeration of Cultivable Microbiome and Identification of Bacterial Communities
Both the genotypes recorded the nearly identical epiphytic bacterial population (1.55-5.6 log CFU g −1 ) (Supplementary Table 4). A total of 37 distinct morphotypes of bacterial isolates were enumerated with a diversity index ranging from 1.48 to 1.82 for all the cultured phyllosphere microbiome. The 30-day-old phyllosphere showed more bacterial population (21 morphotypes) and diversity compared with the 15-day seedlings (16 morphotypes). The results of diversity indices further revealed more bacterial diversity in the susceptible genotypes than in the resistant genotypes. The diversity indices of epiphytic bacteria on the rice phyllosphere are presented in Supplementary Table 5. BOX-PCR DNA fingerprinting of all 37 morphotypes culminated in 31 distinct BOX Amplicon Groups (Supplementary Figure 2). Isolates with identical amplicon profiles were considered duplicates, and a representative isolate for each of the BOX groups was retained and subjected to downstream work. The bacterial species identity was established through a 16S rRNA gene sequence analysis. The 31 distinct BOX amplicon groups represented 11 genera and 16 species. We also observed high-frequency occurrence of bacterial species like Acinetobacter (2), Curtobacterium (3), Enterobacter (4), Microbacterium (4), Pantoea (9), Pseudomonas (2), Rhizobium (2), and Sphingomonas (2) on the rice phyllosphere (Table 2 and Supplementary Figures 3, 4a-k). All cultured bacterial genera were also found among the mapped reads in the metabarcoding analysis (Figure 2 and Table 3).

Activity Screening Against Rice Blast Fungus Magnaporthe oryzae
Screening for Antifungal Activity Dual-plate confrontation assay showed inhibition of the mycelial growth of M. oryzae by both volatiles and secreted metabolites produced by bacterial species. Among the 31 bacteria evaluated, 11 phyllosphere-associated bacterial isolates displayed over 40.0% inhibition of mycelial growth by their secreted metabolites ( Table 4 and Supplementary Figure 5). The antagonistic bacterial isolates represented species, such as Acinetobacter calcoaceticus, Enterobacter cloacae, E. sacchari, P. ananatis, P. vagans, P. oryzihabitans, and Sphingomonas sp. Similarly, a total of 12 of them completely inhibited the growth of M. oryzae (100% inhibition) by bacterial volatile organic compounds (Table 4 and Supplementary Figure 6). The antifungal volatile emitting bacterial isolates represented the species, such as E. sacchari, M. testaceum, P. ananatis, P. dispersa, P. vagans, P. oryzihabitans, and Rhizobium sp. Furthermore, while the volatile of seven bacterial isolates displayed fungicidal activity, the other five bacteria released fungistatic volatiles against M. oryzae as mycelial growth re-emerged upon removal of the volatile exposure (Supplementary Figure 7 and Supplementary Table 6).

Suppressive Effect of Phyllomicrobiome on Blast Disease
Blast-susceptible rice cultivar, Pusa Basmati 1, was used for evaluating the blast-suppressive effects of rice phyllomicrobiome. A total of 13 bacterial isolates representing Pantoea (eight strains), Enterobacter (one), Microbacterium (one), Pseudomonas (one), Rhizobium (one), and Sphingomonas (one) were evaluated at three different cell densities. Blast incidence and severity were scored as per the blast score chart recommended by *Genus with less than 10 reads were not considered. Mackill and Bonman (1992 (Figure 3 and Table 5). Nine of the 13 bacterial isolates showed over 50% reduction of blast severity in all bacterial titers. Interestingly, the reduction in blast severity could not be correlated with the bacterial cell densities used for phyllobacterization.

Phenotyping for Phyllomicrobiome Conferred Immunocompetence
An experiment was conducted to study the potential of phyllosphere bacteria to confer immunocompetence in rice. Seedling growth inhibition due to microbial interactions is touted as a phenotypic marker for the induced immune response in plants. Here, the seed germination and the seedling emergence were monitored during bacterial interaction that showed germination in the range of 40.0-86.7% across various bacterial inoculations. Maximum germination was observed in E. sacchari OsEp-Plm-15B10, while the minimum germination was observed in P. ananatis OsEp-Plm-30B2 (Supplementary Figure 8 and Supplementary Table 7). However, the germinability of rice seeds reduced upon increasing bacterial density with poor germination observed at a titer of 10 9 CFU ml −1 . The bacterial inoculation on seeds and seedlings appeared to trigger the plant phenotypic alteration especially on the shoot and root growth. All nine bacterial isolates representing six genera induced a density-dependent alteration on the shoot and root emergence and growth (Supplementary Tables 8, 9). In particular, the shoot and root growth of rice was found inhibited at high bacterial titer (10 9 CFU ml −1 ) (Figure 4).

qPCR Assay on Phyllobacterization-Mediated Immunocompetence
The rice seedlings displaying the altered growth pattern were subjected to transcriptional analysis by qPCR. Defense genes

Hypersensitive Reaction on Tobacco
Rice phyllospheric bacteria upon infiltration in tobacco leaf did not trigger any necrotic reactions even after 48 h. The expression of quick necrosis due to the hypersensitive reaction FIGURE 4 | Assay for phyllomicrobiome-conferred immunocompetence. Seedling growth inhibition as a phenotypic marker of microbiome-conferred immunocompetence, observed with nine blast-suppressive bacterial isolates, represented six genera such as Enterobacter, Microbacterium, Pantoea, Pseudomonas, Rhizobium, and Sphingomonas. The rice phyllosphere bacteria showed various shades of yellow pigmentation; the pink color appearance of the bacterial colony is due to the reduction of tetrazolium dye into insoluble formazan; the inhibition of shoot and root growth-an indicator of innate immunity, can be seen in plantlets interacting with high bacterial titer.
was observed in the pathogen, R. solanacearum, which infiltrated a leaf indicating that the rice phyllosphere-associated bacterial isolates are non-pathogenic on plants (Supplementary Figure 9).

DISCUSSION
The foliar plant niches termed phyllosphere is one of the habitats for a diverse microbiota where the dynamic plant-microbe interactions are believed to impact plant performance in the ecosystem. With assistance from the microbial genome and their metabolic capabilities, the plants are exposed to a plethora of unpredictable, yet competitive or cooperative, microbial interactions (Vorholt, 2012;Hardoim et al., 2015;Reinhold-Hurek et al., 2015;Brader et al., 2017;Lemanceau et al., 2017).
In recent years, the impact of microbe-microbe interactions on the host-microbial pathogen interaction outcomes is gaining the attention of researchers. Studies have shown that the microbiome structure, assemblage, and compositions are directly influenced by both micro and macro abiotic FIGURE 5 | (Continued) FIGURE 5 | qPCR-based transcriptional analysis of defense gene expression in rice seedlings upon phyllobacterization. The fold change values obtained for the defense genes were imported into the GraphPad Prism program (https://www.graphpad.com/scientific-software/prism), and two-way ANOVA was conducted using Bonferroni post-hoc test for determining the statistical significance at *p ≤ 0.05, **p = 0.001, and ***p = 0.0001. Refer to Supplementary Table 10 for data pertaining to fold changes of gene expression. and biotic factors (Jacobs et al., 2005). Most of the previous microbiome studies have focused primarily on metagenomic or amplicon sequence surveys, and a few attempts have been made to validate the microbial mNGS datasets using classical microbiological culturomic tools. Strikingly, we have compared the phyllomicrobiome of blast disease-resistant and -susceptible rice genotypes planted in a blast endemic location in India. We recorded marginally high bacterial diversity and species richness on PRR78 that could be attributed to the innate susceptibility of the genotypes owing to the absence of functional NBS-LRR type of receptors termed as R-genes. Besides, the susceptible leaf showing early necrotic symptoms might have paved way for the inevitable microbial succession on the phyllosphere. In the present work, we not only microbiologically validated the bacterial community structure in the phyllomicrobiome but also functionally characterized them for harnessing their antifungal and defense-inducing potential against rice blast disease. The NGS metabarcoding-based microbiome profiling revealed the predominance of phylum Proteobacteria on the rice phyllosphere. The predominance of Proteobacteria consisting of bacterial communities belonging to Enterobacteriaceae and Pseudomonadaceae is reported on the phyllosphere by many workers (Knief et al., 2012;Ren et al., 2014;Roman-Reyna et al., 2019;Yasmin et al., 2020). At the lower taxonomic genus hierarchy, Pantoea followed by Pseudomonas and Enterobacter were overrepresented on both the rice genotypes. Our findings are in agreement with many other reports that highlighted the ubiquitous occurrence of Pantoea as the most abundant bacterial genera on the phyllosphere (Cottyn et al., 2009;Cother et al., 2010;Kim et al., 2020;Stone and Jackson, 2020;Dhankhar et al., 2021). The microbiological investigation culminated in the isolation of 37 distinct morphotypes on the rice phyllosphere; here, the 4-week-old seedling harbored more bacterial morphotypes compared with the 15-day-old seedlings implying that microbial biomass on plant niches expands with the age of the plants. The morphotypes could be identified as belonging to 31 distinct amplicon groups in BOX-AIR-PCR fingerprinting based on shared amplicon profiles of the isolates. The BOX PCR DNA fingerprinting is one of the widely used molecular tools in bacterial typing and biogeography studies of microbial isolates (Versalovic et al., 1994;Brusetti et al., 2008).
The 16S rRNA gene sequence database search confirmed the identity of cultured bacterial isolates as belonging to Acinetobacter, Enterobacter, Pantoea, Pseudomonas, and Sphingomonas on blast-resistant and -susceptible genotypes. Pantoea, Enterobacter, Microbacterium, and Curtobacterium are recently reported as a member of the core microbiome of the rice leaf endosphere . The genus Microbacterium too is frequently observed on the rice phyllosphere and spermosphere by several researchers in the past (Kaku et al., 2000;Leveau and Lindow, 2001;Midha et al., 2016). Characteristically, most of the phyllosphere bacterial genera produced shades of pigmentation as observed in Acinetobacter (pale brown), Curtobacterium (dark yellow), Enterobacter (pale brown to yellow), Microbacterium (yellow), Pantoea (yellow), Pseudomonas (pale brown to yellow), and Sphingomonas (dark yellow) (Supplementary Figures 4a-k). Production of dark pigmentation is one of the adaptive traits of bacteria and other microbes that encounter harsh environmental abiotic stresses like solar radiation and light on the phyllosphere (Green, 1992;Jacobs et al., 2005). Pigmentation on phyllospheric bacteria is believed to protect them from harmful ultraviolet radiation (Sundin and Jacobs, 1999). The rice phyllosphere is considered as the preferred habitat for yellow-pigmented Pantoea and pink-pigmented methylotrophs, which can survive under nutritional and moisture stress as well as can withstand harmful γ-ray radiation (Green, 1992). The effect of solar radiation on the composition and activities of the phyllosphere microbial community is reported by Carvalho and Castillo (2018).
Currently, growing evidence for microbe-induced seedling growth alteration as a phenotypic marker of activated innate immunity is published (Wang et al., 2021). We observed that the phyllosphere bacterial species in a density-dependent manner impacted the seed germination with consequent seedling growth alterations upon seed inoculation. The results are in agreement with the studies of Damodaran et al. (2013) and Zhu et al. (2017) who observed varying effects of endophytic and rhizosphere bacteria on seed germination. The seedling growth assay enabled us to identify the bacterial species that conferred immune competence in rice seedlings. Whereas the seedling growth was found inhibited at a higher dose (10 9 cells per ml), the lower doses (10 6−8 cells per ml) showed characteristic non-lethal seedling inhibition presumably owing to a tradeoff between growth and immunity. To confirm this, we performed the qPCR-based temporal transcriptional analysis of defense genes involved in innate immunity on phyllobacterized rice seedlings.
Phyllosphere bacterial species evaluated against rice blast under artificial epiphytotic trial in greenhouse showed a reduction in blast disease (50 % over mock) at all tested bacterial titers. Upon prophylactic foliar application (or phyllobacterization), the species belonging to Pantoea, Enterobacter, Microbacterium, Pseudomonas, Sphingomonas, and Rhizobium showed significant blast suppression at all tested doses. However, we could not observe any dosage response for enhanced blast suppression revealing the sufficiency of bacterial augmentation at 10 6−7 cells per ml for reducing blast disease. Suppression of blast disease by Bacillus, Streptomyces, Pseudomonas, Pantoea, Paenibacillus, Burkholderia, Enterobacter, Paraburkholderia, and Actinomycetes was earlier reported in the literature (Gómez Expósito et al., 2017;Harsonowati et al., 2017;Schlatter et al., 2017). Rice blast suppression by Microbacterium, Pseudomonas, and Stenotrophomonas are attributed to the direct antifungal antibiosis and the indirect defense activation as evident from the expression of rice defense genes (Ashajyothi et al., 2020;Sahu et al., 2020). Recently, bacterial volatile belonging to pyrazines is reported to modulate defense against blast disease . Taken together, it is concluded that the enrichment of phyllosphere bacterial communities on the leaf can inflict antifungal antibiosis on M. oryzae and defense elicitation in rice to reduce the incidence and severity of blast disease.
Having confirmed the blast-suppressive potential of phyllosphere bacterial communities, we conducted a tobacco infiltration HR assay to ascertain the biosafety of the phyllosphere bacterial isolate (Klement, 1963). The tobacco HR assay has been recognized to test the pathogenic nature of plant-associated bacterial species (Kucheryava et al., 1999;Medina-Salazar et al., 2020;Sadeghi and Khodakaramian, 2020). None of the phyllosphere bacterial isolates showed any necrotic lesions, while comparing with R. solanacearum served as a positive check. Shades of faded yellowing observed with few phyllosphere bacterial isolates are indicative of activated defense.
In conclusion, the phyllosphere bacterial communities suppressed the blast disease by the dual action of antifungal secreted and volatile metabolites as well as by microbe conferred immunocompetence (Figure 6). The present investigation on phyllosphere microbiome analysis of rice culminated in several potential hitherto unexplored bacterial communities for microbiome-assisted crop protection, especially against rice blast disease.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/ Supplementary Material.

AUTHOR CONTRIBUTIONS
KS conceptualized the study, developed the methodology, investigated and validated the study, performed the formal analysis, and wrote the original draft. AK conceptualized and investigated the study, provided the resources, wrote, reviewed, and edited the manuscript, and was in charge of the visualization, supervision, and project administration. AP, MK, and NS developed the methodology and validated the study. SM and BR performed the formal analysis. PE and NP validated the study. All authors read and approved the final manuscript.