Metagenomic evaluation of peanut rhizosphere microbiome from the farms of Saurashtra regions of Gujarat, India

The narrow zone of soil around the plant roots with maximum microbial activity termed as rhizosphere. Rhizospheric bacteria promote the plant growth directly or indirectly by providing the nutrients and producing antimicrobial compounds. In this study, the rhizospheric microbiota of peanut plants was characterized from different farms using an Illumina-based partial 16S rRNA gene sequencing to evaluate microbial diversity and identify the core microbiome through culture-independent (CI) approach. Further, all rhizospheric bacteria that could grow on various nutrient media were identified, and the diversity of those microbes through culture-dependent method (CD) was then directly compared with their CI counterparts. The microbial population profiles showed a significant correlation with organic carbon and concentration of phosphate, manganese, and potassium in the rhizospheric soil. Genera like Sphingomicrobium, Actinoplanes, Aureimonas _A, Chryseobacterium, members from Sphingomonadaceae, Burkholderiaceae, Pseudomonadaceae, Enterobacteriaceae family, and Bacilli class were found in the core microbiome of peanut plants. As expected, the current study demonstrated more bacterial diversity in the CI method. However, a higher number of sequence variants were exclusively present in the CD approach compared to the number of sequence variants shared between both approaches. These CD-exclusive variants belonged to organisms that are more typically found in soil. Overall, this study portrayed the changes in the rhizospheric microbiota of peanuts in different rhizospheric soil and environmental conditions and gave an idea about core microbiome of peanut plant and comparative bacterial diversity identified through both approaches.


Culture-independent diversity
In the CI approach run, around 7.5 million sequence reads were generated from 65 samples, with an average count of 116,687 reads per sample.The DADA2 pipeline inferred 17,719 Amplicon Sequence Variants (ASVs) from 4.8 million reads (64.33%).After filtering ASVs, 8,042 ASVs from 59 samples remained which were further analyzed (detailed information in supplementary method).All ASVs were taxonomically classified as bacteria, further belonging to 31 phyla, 67 classes, 177 orders, 315 families, 665 genera and 87 species.The reads distribution across taxonomic levels was highlighted in Table S1.

Changes among rhizospheric samples
The changes in the rhizospheric samples were evaluated with respect to their geographic location.NMDS ordination on Bray-Curtis distance revealed that there were separate and distinct clusters of samples as per their geographical location.For example, samples of Rajkot, Amreli and Porbandar district farms formed clusters near to each other, while Gir-Somnath district samples formed separate clusters (Fig. 3).All the samples of Junagadh district were grouped near to Gir-Somnath district samples except F03 farm samples.The difference among geographic locations was further confirmed through PERMANOVA, where a significant difference was observed (p-value < 0.001).A pairwise-adonis between all pairs of farms was significantly different (Table S4).An environmental fit of all variables also revealed a significant association of organic carbon (OC), concentrations of potassium (K 2 O), phosphate (P 2 O 5 ) and manganese (Mn) (Figs. 3, S7).

Core rhizosphere microbiome
Total 168 genera were identified as part of the core microbiome of rhizosphere samples with a minimum abundance of 0.1% across 40% of samples (Fig. 4).To observe the pattern of co-occurrence, these taxa were further correlated with each other.Seven different clusters of genera could be made out from the significantly (p-value < 0.05) correlating genera (Fig. 5).It was discovered that three clusters were negatively correlated with all the other genera.These contained genera like Sphingomicrobium, Chelativorans, Vitiosangium, Lysobacter, Microvirga, Dyadobacter, unknown members of Sphingomonadaceae, Rhizobiaceae, Xanthomonadaceae and Enterobacteriaceae family.A separate cluster could also be made out containing mostly Bacilli class members like Bacillus_BD, Ectobacillus, Domibacillus, Metabacillus and unknown members of Domibacillaceae family, Bacillales_B, Bacillales order and Bacilli class.Many of these genera, particularly the more prevalent ones, exhibited a negative association with all other genera, which might explain their growth during the nodulation phases.www.nature.com/scientificreports/

Comparative analysis using both approaches
Comparative approach of CD and CI In addition to above mentioned data, around 1,282,054 paired-end reads were generated for samples of CD approach run.DADA2-based pipeline generated 18,765 ASVs using sequences of both approaches, of which 6970 ASVs having more than thirty supporting reads were considered for further comparative analysis of CD and CI approaches (Detailed information in supplementary method).

Discussion
In this study, the peanut rhizospheric microbiota was analyzed using Illumina-based 16S rRNA gene sequencing and characterized the culturable bacterial diversity through traditional microbiological cultivation approach.The 16S rRNA gene sequencing technique is extensively used to characterize diverse microbiomes, including rhizosphere.The extensive use of new molecular methods is due to the limitation of traditional microbiological cultivation approaches which are unable to provide a complete picture of bacterial diversity due to the inability www.nature.com/scientificreports/ to cultivate all the microbes under laboratory conditions, probably because of their specific growth requirement.
To provide more resolution to this analysis, the DADA2 denoising algorithm pipeline was used for data analysis 35 .DADA2 is a denoising algorithm designed particularly for Illumina data that infers ASVs based on single nucleotide changes, thus upholding strain-level information.In the present study, several ASVs were distinguished to species level (Fig. S6), including some ASVs with higher abundances.However, the analytical capacity is also affected by the database used for taxonomy assignment.For that, GTDB version 202 was used for the taxonomy assignment.Our choice of GTDB was influenced by the existing results that suggest more number of sequences annotating at the genus level in GTDB, and also because of the taxonomy lineage assignment approach used in GTDB 29,36,37 .GTDB is a curated database with comprehensive genome-based taxonomy based on monophyly and relative evolutionary divergence of taxa, which is an added advantage while annotating ASVs.The reclassifications by GTDB works well by distributing/reclassifying popular genera into several novel ones 36 .This gives a higher resolution to the observed organisms in this study.For example, the abundance of Pseudomonas_F, Pseudomonas_M, Pseudomonas_R, and Pseudomonas_S genera were observed among all Pseudomonas genera.Similar observations were also made with Bacillus genus where ASVs classified as Bacillus_BD, Bacillus_AG, Bacillus_BN and Bacillu_BU among all Bacillus genera.
In the present metagenomic study through the CI approach, the rhizospheric soils of F-08 farm showed higher Shannon diversity (6.92) than other farms, which may be due to influences of individual physico-chemical and abiotic parameters of respected rhizospheric soil.Previous studies showed that the electric conductivity and concentration of different nutrients including N, P and K may alter the diversity of microbial community present in rhizospheric soil [38][39][40] .Additionally, changes in soil pH and OC are typically linked to modification of rhizospheric microorganisms 20,41 .Our study found a strong significant link between OC, concentrations of P 2 O 5 and Mn of rhizospheric soil with rhizospheric microbiota, different from previous studies where a significant link was found between pH, EC and concentration of K 2 O with rhizospheric microbiota 16,19,42 .In all CI samples, greater abundances (relative abundance ≥ 2%) of Proteobacteria, Acidobacteria (called Acidobacteriota in GTDB), Actinobacteria (named Actinobacteriota in GTDB), Planctomycetes (named Planctomycetota in GTDB), Firmicutes, Bacteroidetes (called Bacteroidota in GTDB), and Verrucomicrobia (named Verrucomicrobiota in GTDB) were found as compared to other phyla.Several previous investigations have found a greater abundance of Firmicutes (approximately 3 to 7%) in the rhizosphere 14,18,25,43 including a study on peanut microbiome 28 .However, this was not observed in other studies on the peanut rhizosphere 27,44 .In the present study, the most abundant genera are Sphingomicrobium, UBA2421, Aureimonas _A, unknown member from Sphingomonadaceae, UBA1161 family and Bacilli class, and genus of unknown bacterium.The genera UBA2421 and UBA1161, observed with > 2% abundance in all samples belong to Planctomycetota phylum and are yet uncharacterized organisms.These representative genera are still not reported by cultivable approach and have only been described in the Metagenome assembled genome (MAG) database 45 .Databases like GTDB, which contains many MAGs can be an added advantage of observing an accurate depiction of diversity and illustrates the fact that there are numerous more abundant microorganisms whose roles in the ecosystem have yet to be determined.
One of the objectives of the present study is to observe the rhizospheric community among all samples by considering natural abiotic stress conditions and to find out changes in microbial community and how it represents the core microbiome of peanut plants by using CI approach.While in previous studies, the rhizosphere community of peanut plants was evaluated under controlled environment (in greenhouse), which gives a limited idea of the rhizospheric community 27,28,44 .In this study, the topmost abundant genera were common in each rhizospheric soil sample (Fig. 2B, Table S3).Moreover, the core microbiome was further studied to investigate the potential plant growth-promoting genera among rhizospheric soil samples.Previous studies shows that, as per different plant growth stages, plants release various exudates, which modify surrounding rhizobacterial populations by selecting the finest organisms that can aid in promotion of plant development in various ways, those are commonly called as Plant growth promoting bacteria (PGPB).PGPB can colonize the rhizosphere and form close relationships with roots of host plant 46,47 .The beneficial effects of PGPB on plant growth are achieved through direct mechanisms such as facilitating nutrient uptake, like primarily nitrogen and phosphorus, and by producing phytohormones.Genera like Sphingomicrobium, Actinoplanes, Aureimonas _A, Chryseobacterium, members from Sphingomonadaceae, Burkholderiaceae, Pseudomonadaceae family and Bacilli class were observed in almost all samples.All of these are reported to show PGP activities.For example, Sphingomonadaceae and Burkholderiaceae family members are well studied for their antifungal activity against Rhizoctonia solani, which is the primary plant pathogen in peanuts [48][49][50] , Actinoplanes reported to possess IAA production, siderophore production, and ACC deaminase activity 51,52 while Bacillus and Pseudomonas are reported to possess several beneficial activities including solubilization of phosphate, nitrogen fixation and siderophore production 53,54 .Aureimonas _A is member of the Rhizobiaceae family known for important nitrogen-fixing symbionts of plants.Based on rhizospheric bacterial diversity of the core microbiome, a potential biofertilizer was formulated to check the effect of biofertilizer to promote the growth of different variety of peanut plants (unpublished data).Some species-level assignments of those genera's representatives were also observed in the CI approach, like Sphingomicrobium sp003097155, Pseudoduganella eburnean, Ectobacillus funiculus, Metabacillus sp002871465, Pseudomonas_M indica and Bacillus_BD endozanthoxylicus (Fig. S6).
Further bacterial diversity was also characterized by comparing the CD approach with the CI approach, by doing so expecting to get a complete idea about the peanut bacterial diversity as although metagenomics is extremely popular, it also fails to reflect the true diversity present in the sample due to some of its limitations.For example, if microorganisms are present in very low abundance it may be left out during DNA extraction or possible bias to amplify the target DNA, data analysis pipeline, and database, all of which affect the final interpretation of the results 55 .As noted in method, in this work, all the colonies were taken away from the medium and relied on NGS-based metagenomic platform for identification and analysis.This should reflect almost the entire cultivable diversity, including several microcolonies.By doing so, we have incorporated sequences from all the organisms grown on plates rather than imposing selection biases based on colony observation/morphology.
Further, using various media, generating a high amount of data, and considering enough reads for the analysis can help to provide a complete picture of bacterial diversity.Furthermore, for comparative study, partial 16S rRNA gene was sequenced, similar to metagenomics, rather than sequencing the entire 16S rRNA gene, through Sanger sequencing as done in all previous researches [32][33][34] .As per our knowledge, a similar approach has also been www.nature.com/scientificreports/successfully applied by Zehavi et al. for the study of ruminal microbiota and in our previous study on rhizosphere microbiome 56,57 .This approach could be helpful to analyze the presence-absence based study of microbial diversity.However, it is not appropriate when attempting to analyze the abundance of cultivable organisms.Also, it would not be possible to separate and purify the colony on the media if needed for further experiments.
According to the majority of research, the CI strategy has greater diversity than the CD technique 33,34 .Similar conclusions can also be drawn from this study as well.Total of 232 ASVs exclusively present in CD samples.This might be due to the very low abundance of those organisms, which is a limitation of CI approach.Comparing all the genera of both the approaches, revealed that total 38 genera were exclusively found in CD samples, that is similar to the studied by Hinsu et al. 57 .Most of these genera are commonly found in soil, forest and water sources including marine water.A few of the genera, like Helicobacter, Heliorestis and Herbaspirillum are also linked to nitrogen fixation ability as per some recent studies 58,59 .Surprisingly, the UBA1067 family and UBA3207 genus from the Kiritimatiellae and Bacilli class respectively were also observed, which are candidate taxa with no cultivated representative as of yet.

Experimental design and sample collection
The rhizospheric soil samples were collected from 13 different farms covering 5 districts of Saurashtra region of Gujarat, INDIA, in 2019 (detailed information in Table 1).All the farms have history of continuously sowing of G-20 variety of Groundnut (Arachis hypogaea L.) during the cropping season.All the rhizospheric soil samples were collected within 10 days to avoid differences in crop stages at the nodulation phase of crops.All the farms had sowed the seed almost at the same time, to take advantage of rains by hurricane, hence being a benefit for our www.nature.com/scientificreports/study.From each farm, 5 plant samples were randomly selected as replicates and uprooted gently by removing nearby soil.The plants were vigorously shaken to remove loosely attached soil.Afterward, the tightly adhering soil on the root surface was collected in sterile container for rhizospheric soil property determination.The roots with tightly adhering rhizospheric soil were washed in sterile normal saline (1% NaCl) in a flask, and the washed soil was then collected in sterile 50 ml falcon tubes for microbiome analysis 24,33 .Total 5 g rhizospheric soil from all 5 replicates of the same farm was pooled separately in new sterile tube to study bacterial diversity by culturedependent approach.Same practice was done for all farms.The rhizospheric soil samples for property estimation were transported at room temperature, and samples for microbiome work were transported to lab at 4 °C and then stored at − 20 °C till further processing.Overall, 65 samples of 13 farms were analyzed for metagenomics study, and comparative analysis of CI and CD approach (Fig. S11).

Sample processing
The rhizospheric soil samples were sent for physicochemical examination to a government-approved soil testing laboratory (Gujarat State Fertiliser Company, GSFC, Vadodara, INDIA).The samples were tested for physical properties (pH and electrical conductivity), macronutrients (% organic carbon, concentrations of phosphate and potassium) and micronutrients (concentrations of iron, sulfur, manganese, zinc, and copper).
For the rhizosphere microbiome (CI approach), the rhizospheric samples were thawed and homogenized.The tubes were then centrifuged at 12,000 rpm for 10 min.At this speed, all microbial cells, along with soil particles, will settle down, leaving behind buffer in supernatant which was discarded.After carefully mixing the soil, it was immediately used for DNA extraction 24 .DNA was extracted from 1 g of soil using Qiagen PowerSoil DNA Extraction kit (Qiagen, Germany) following the manufacturer's instructions.
For the CD approach, samples were serially diluted and plated on eight different media supplemented with cycloheximide (50µg/ml) on the next day of collection (Table S5). 10 -4 and 10 -5 dilutions were used for spreading, and plates were incubated at 27 ± 2 °C and 37 ± 2 °C in triplicates.After incubation for 15 days, all the colonies of the same samples were scrapped from each media, collected in phosphate buffer, and mixed.DNA was extracted from this pool of colonies using QIAGEN QIAamp DNA Mini Kit (Qiagen, Germany) following manufacturer's protocol of bacterial genomic DNA extraction (Fig. S12).Extracted DNA was checked on agarose gel for good quality and quantified using Qubit 3.0 (Invitrogen, CA).

Library preparation and sequencing
The 16S rRNA gene amplicon sequencing libraries were prepared separately for CI and CD from 12.5 ng DNA as starting material following double-pass PCR protocol as given in Illumina 16S library preparation guide (Illumina, USA).The primers 341F and 785R coupled with Illumina adapters were used to target the V3-V4 region of the 16S rRNA gene 60 .Agilent Bioanalyser (Agilent, USA) was used to validate the libraries, and Qubit v3 was used to quantify them (ThermoFisher Scientific, USA).The libraries were sequenced separately for CI (run1) and CD (run2) approaches on Illumina MiSeq using 250 × 2 v2 chemistry.

Data analysis
The raw fastq data was analyzed using the Divisive Amplicon Denoising Algorithm 2 (DADA2) pipeline ("dada2" package version 1.22) in R v4.1.3following the steps given at https:// benjj neb.github.io/ dada2/ tutor ial.html for rhizosphere microbiome 35,61,62 .Further, CI and CD samples were sequenced in different runs.Runs of CI and CD were processed independently until sequence table generation and then merged for further steps as indicated in the "big data" tutorial (https:// benjj neb.github.io/ dada2/ bigda ta.html) for comparative analysis of bacterial diversity.Taxonomy of ASVs was assigned using GTDB v202 databases using the files hosted at zenodo 63 .

Ethics declaration
The study included the use of soil associated with plants.No ethical approval was required for the investigation because no plant components were used.Additionally, the owner or farmer was made aware of the research and the kinds of samples that would be taken.Verbal consent and permission were obtained to collect the soil from his farm for the work.

Conclusion
The findings of the current study indicated that a large number of uncultured and unidentified core bacterial genera representative were present in the peanut rhizospheric, many of which may have interacted with the host plant and other microorganisms.Additionally, key core genera that were known to support plant growth were identified from the peanut rhizosphere; this knowledge helped us develop efficient bio-strategies, such

Figure 1 .
Figure 1.Plot highlighting each farm's alpha diversity.Alpha diversity measures Observed ASVs (top), and Shannon Index (bottom) plotted for each farm.p-value from Kruskal-Wallis test comparing all farms is mentioned on the top.

Figure 2 .
Figure 2. Taxonomic distribution of samples.Taxonomic distribution at (A) phylum level and (B) Genus level.Only the most abundant taxa are plotted for both levels.

Figure 3 .
Figure 3. NMDS plot of Bray-Curtis distance calculated from all Rhizosphere samples.District of each farm is used as shape to denote the sample.Arrows are environment fit vectors that represent physical factors and nutritional concentrations.Vectors with significant associations are shown in red coloured arrow.

Figure 4 .
Figure 4. Plot representing core microbiome from rhizosphere samples.The graphic compares genus occurrence in samples with varied degrees of abundance.Only the genera with minimum prevalence of 0.4 at 0.001 abundance are plotted.

Figure 6 .
Figure 6.Observed ASV count and Shannon diversity distribution plot.The farms are coloured differentially.X-axis represents either CI approach or CD approach.

Figure 7 .
Figure 7. Plots representing count of unique ASVs and relative abundance of phyla.(A) ASVs detected across all phyla in all samples (including CD and CI samples, grey colour) and exclusively in culture-dependent samples (orange colour).(B) Relative abundance of top 15 phyla.

Figure 8 .
Figure 8. Upset plots representing shared and unique taxa.Upset plot displaying the distributions of (A) All detected ASVs and (B) Genera-level taxonomy, among CD and CI sample-groups.