2.1.- Motility and adhesion: assets for plant root colonization
Motility of a bacterium is due to the flagellum, enabling it to move towards a vital nutrient source (chemotaxis). In this sense, B. subtilis PTA-271 contains genes (Supplementary Table S1) putatively encoding for flagella maintenance (flh genes) and chemotaxis (che genes). Once reaching a comfortable area, adhesion is due to bacterium pili, allowing the initiation of biofilm formation where both chemotaxis and gene exchanges among microorganisms of microbiota can be amplified.
B. subtilis spp. are also described for their strong swarming motility . The gene swrC putatively encoding for swarming motility protein is predicted in the genome of B. subtilis PTA-271 (Supplementary Table S1). Swarming motility requires the production of functional flagella, pili and surfactant to reduce surface tension.
Motilities and adhesion are considered advantageous characters for a successful host colonization and B. subtilis spp. are already described to grow in biofilm mode involved in root colonization . To this end, the transcription factors (TF) Spo0A and AbrB were described as positive and negative regulators of biofilm formation, respectively . Genes putatively encoding for these 2 TFs are also predicted in the genome of B. subtilis PTA-271 as Spo0A and AbrB (Supplementary Tables S1 and S2).
Beneficial microorganisms that successfully colonize the plant, particularly by the root system, would be advantageous, both for plant growth promotion and for plant biocontrol [38-39].
2.2.- Biofertilizing and morphogenic effects: assets for plant vigor
Plant nutrition depends on soil retention capacity of minerals and nutrient availabilities, thus both on chelating process, mineralization by decomposers and minerals bioavailability towards the plant consumer. Upon nitrogen starvation, some bacteria are described to upregulate the ure gene cluster, since urea is an easy nitrogen source. Such ure genes are predicted in B. subtilis PTA-271 genome (ureA, ureB, ureC). This cluster of genes is known to be controlled by the global nitrogen‐regulatory protein TnrA, also predicted in B. subtilis PTA-271 genome (Supplementary Table S2). Regarding other nutrient access due to phosphate-solubilizing bacteria (PSB) [42-43], genes encoding for proteins involved in the production of gluconic acid and precursor of citric acid are also predicted in the genome of B. subtilis PTA-271 (S19-40_03830, S19-40_03828). Organic acids may lower the soil pH to solubilize phosphate and thus increase its availability to the plant . Bacterial secondary metabolites (PyrroloQuinoline Quinone, PQQ) are also known to control gluconic acid production , and B. subtilis PTA-271 has 3 genes predicted to be related to PQQ production pqqL, pqqF and pqqC . Additionally, B. subtilis PTA-271 contains the phytase gene phy, described in the other Bacillus spp. to encode for phosphatases able to hydrolyze organic complex in order to liberate phosphate and make it available for plants . Iron is another very important nutrient for plant growth and development. B. subtilis PTA-271 possesses the fur gene (Supplementary Table S2) described in the literature to encode for a regulatory protein coordinating the homeostasis of iron uptake depending on its availability in the soil . Regarding soils containing abundant ferric form (Fe3+) which is poorly available to plants, the literature described bacteria producing siderophores with high specificity and affinity for iron, capable of binding, extracting and transporting iron near the plant roots . B. subtilis PTA-271 genome also predicted the production of such siderophores, namely the catecholic siderophore 2,3-dihydroxybenzoate-glycine-threonine trimeric ester bacillibactin encoded by 5 genes (dhbA to dhbF). Surfactants produced by beneficial bacteria also contribute to increase the availability of hydrophobic nutrients. In this sense, B. subtilis PTA-271 is suspected to produce surfactin (with srfAA to srfAD), a powerful biosurfactant due to its amphiphilic nature that strongly anchor with lipid layers, interfering with the structure of biological membranes .
Plant root morphology is also described to impact nutrient uptake and thus plant growth due to the stimulation of lateral root formation and root air formation, while primary root elongation is inhibited [51-52]. Plant hormones are key elements for root morphology changes. Some beneficial bacteria are also described to produce them . Regarding B. subtilis PTA-271 genome, it predicts the trp group, described in literature to produce tryptophan as the main precursor of the auxin IAA (indole-3-acetic acid) . Once synthesized, bacterial IAA is described for two main functions: (i) increase the plant root surface and length, for a deeper soil prospecting capacity and nutrient acquiring capacity, and (ii) release the cell walls of rootlets to facilitate molecule exudations and amplify interactions . The genome of B. subtilis PTA-271 also predicts genes such as yvdD (Supplementary Table S2), linked in the literature to cytokinin synthesis which is known as a plant growth regulator (cell division, organogenesis) in combination with IAA. Gibberellins (GA) produced by some bacteria also affect the plant growth and survival by interfering with the plant signaling pathways through secondary metabolites changes . Regarding the B. subtilis PTA-271 genome, it predicts ispD and GerC3_HepT, described in the literature to be respectively linked to 2-C-methyl-D-erythritol 4-phosphate (MEP) and geranylgeranyl diphosphate (GGPP) production, two successive precursors of GA and abscisic acid (ABA) synthesis in plants. But from GGPP, no additional genes appear encoded by B. subtilis PTA-271 genome to complete the kaurene pathway leading to GA.
Genes described to encode for other plant growth regulators, namely polyamines (PAs), are also predicted in the genome of B. subtilis PTA-271. Among them: speA, speB, speG and speE are respectively described in literature to encode for putative ADC (arginine decarboxylase), agmatinase (leading to putrescine), then spermidine- and spermine- synthases. Additionally, genes encoding for putative S-adenosyl-methionine (SAM) decarboxylase (speH) and SAM-methyltransferase (S19-40_00450) are predicted in B. subtilis PTA-271 genome, and these proteins are mentioned to complete PA synthesis from putrescine . PAs are known to promote flowering and to play important roles in inducing cell division, promoting regeneration of plant tissues and cell cultures , as delaying senescence .
Volatile compounds (VOCs) produced by some beneficial rhizospheric bacteria have also been identified as elicitors promoting plant growth. Regarding B. subtilis PTA-271, its predicted genes encode putatively for (1) acetoin (acuA, acuC…) and (2) 2,3-butanediol (butA and butC) [20,55]. VOCs are especially reported to interact with plant hormones [56-58].
2.3.- Host induced defenses and Microbiota preservation: assets for plant protection
PLANt induced DEFENSES upon biotic stress
Host primed defenses during ISR are regulated by hormones, depending on either JA and ET signaling or SA signaling [7-8,10,59]. Beneficial microorganisms may modulate the plant hormonal balance or directly elicit the plant defenses. Regarding the genome of B. subtilis PTA-271, no gene encodes for ACC deaminase (responsible for the breakdown of the ET precursor: ACC), but the predicted metK gene encodes for the putative S-adenosylmethionine (SAM) synthase (leading to SAM, the ACC precursor). Such ET precursors would appear ISR-useful, for plants which possess the complementary ET metabolic machinery, unless they only feed the PA pathways of the host [54, 60]. SA is another hormone for which several genes encoding its metabolic pathways (from synthesis to hydrolysis) are predicted in B. subtilis PTA-271 genome, among which pchA encoding putatively for the salicylate biosynthesis isochorismate synthase.
Many elicitors also induced host immunity, coming from microorganisms (MAMPs, microbial associated molecular patterns) but also from the plant host (DAMPs, damaged associated molecular patterns). MAMPs can act from the external surface of a beneficial microorganism (flagellin) or result from a secretion outside or inside the host (surfactin, fengycin, VOCs, etc.). Flagellin proteins putatively encoded by the hag gene predicted in B. subtilis PTA-271 (Supplementary Table S1), are described in literature to be recognized by the plant pattern recognition receptors PRRs (FLS2) which are cell surface localized kinase receptors activating host defenses through mitogen-activated protein kinase cascades (MAPK) [40,61]. The lipopeptides surfactin and fengycin are other elicitors of plant ISR, inducing the plant production of antimicrobial phenols , and are putatively encoded by some genes predicted in the genome of B. subtilis PTA-271 (srf and fen genes, respectively). VOCs produced by rhizospheric bacteria, as the 3-hydroxy 2-butanone and acetoin which are putatively encoded by B. subtilis PTA-271 genome, are also well known to induce ISR through SA-independent pathway, but merely through the ET one . Among VICs, the ubiquitous nitric oxide (NO) is another signal molecule both scavenging reactive oxygen species (ROS) and regulating the level of PAs and the hormonal balance (ABA versus SA) in order to reprogram or switch plant development upon stress . Different genes related to NO metabolic pathways are predicted in B. subtilis PTA-271 genome, among which the gene nos putatively encoding for a NO synthase oxygenase. Exopolysaccharides (EPS) and lipopolysaccharides (LPS) are other elicitors reported in several Bacillus genera [9-14,59]. Regarding the genome of B. subtilis PTA-271, it predicts several genes putatively encoding for EPS (S19-40_00800, S19-40_00870, S19-40_00999, S19-40_01009, S19-40_01427) and LPS (lptB, lapA, lapB), additionnally to the other elicitors predicted to be encoded by B. subtilis PTA-271 genome (siderophores, flagella, N-acyl-L-homoserine lactone, etc.).
DAMPs are alternative elicitors produced by lytic enzymes (chitosan, glucans, etc.) of microorganisms (either beneficial or pathogenic) or plants . Genes encoding for lytic enzymes are predicted in B. subtilis PTA-271 genome, such as those encoding for putative chitosanase and ß-glucanase (Supplementary Table S3). Many other genes are also predicted to encode for lytic enzymes in the B. subtilis PTA-271 spore cortex (Supplementary Table S4) for which the roles remain unclear.
PLANt induced DEFENSES upon abiotic stress
Some previously cited hormones are also useful for plant defense against abiotic stress, such as ABA and GA , which precursors are predicted to be encoded by genes identified in the genome of B. subtilis PTA-271 (GerC3_HepT, ispD). From GGPP, the kaurene pathway may lead to GA, while the phytoene path may lead to ABA , and in the genome of B. subtilis PTA-271, yisP (a crtb KEGG gene) encodes for a putative 15-cis-phytoene/all-trans-phytoene synthase. ET is another useful hormone for plant defense against abiotic stress , and B. subtilis PTA-271 genome has genes identified to putatively produce SAM (metK), a precursor of ACC which is required for ET synthesis in plants. Altogether these data predict that B. subtilis PTA-271 genome may putatively encode for key precursors of phytohormones that may influence actively ABA and ET contents in plants. In plants, ABA, GA and ET signaling pathways interfere altogether through different transcription factors (TF) or small proteins (GiD, DELLA, EIN, etc.) that physically interact [65-66]. In the genome of B. subtilis PTA-271, many genes are predicted to encode for sigma factors and many TF (Supplementary Table S2). It is noteworthy to understand that useful TF upon abiotic stress could also be useful upon biotic stress. The set of genes under common regulatory controls (operons) are also listed in the Supplementary Table S2.
PAs such as those predicted to be encoded by the genome of B. subtilis PTA-271 are also described to protect plant cells upon water deficit , temperature changes  and salinity . They are known to increase the activity of various antioxidant enzymes in plants and may contribute to produce H2O2 as a signaling molecule that can activate plant antioxidant defense responses .
Microbiota quality and preservation
As energy and carbon sources, plant root exudates (sugars, organic acids, amino acids, lipophilic compounds, etc.) would enable to selectively recruit biosurfactant producers . In return, these beneficial bacteria can facilitate the bioavailability of root exudates and biofilm formation, thus the colonization of host-plants by beneficial bacteria [41,70], maybe such as B. subtilis PTA-271 which is suspected to produce surfactin. SA was also shown to mediate changes in the composition of root exudates, then in the qualitative microorganism recruitment by plants . Regarding the B. subtilis PTA-271 genome, some genes are also predicted to produce SA (pchA), highlighting another key lever that putatively influence the composition of plant microbiome.
Beneficial microbial interactions can additionally depend on bacterial auto-inducers (AI) that are low-molecular weight signal molecules activating the interactive competences of a bacterium in a quorum-sensing (QS) dependent manner . Among AI, the furanosyl-borate-diester (AI-2) is described as universal for interspecies communication both in gram-positive and gram-negative bacteria . Regarding B. subtilis PTA-271 genome, the predicted luxS gene putatively encodes for AI-2 production, while the predicted EntF and AM373 putatively encode oligopeptides or auto-inducing peptide (AIP) precursors. AIP is another class of AI consisting of 5-34 amino acids residues and produced by Gram‑positive bacteria for their intercellular communication .
When interacting with the environment, a microorganism had also to remain metabolically active to exert beneficial effects. Upon biotic interactions, Bacillus species are exposed to host defenses that include reactive oxygen species (ROS) . Regarding the system of sensing, protection and regulation of ROS in the genome of B. subtilis PTA-271, genes encoding for resistance to hydroperoxide (ohrA, ohrB, ohrR) are predicted to putatively ensure its survival. Upon abiotic stress, beneficial bacteria must survive dehydration, wounding, cold, heat or salinity that in turn lead to regulation of the water status. For this end, bacterial species can control their intracellular solute pools [75-76]. Regarding the genome of B. subtilis PTA-271, genes predicted to encode for potassium uptake proteins (KtrA, KtrB) putatively enable survival in high salinity environments. Interestingly, the genome of B. subtilis PTA-271 also predicts genes to detoxify or resist compounds accumulating in the environment [77,78], such as arsenite (arsR), organic pesticides or nitroaromatic compounds (sugE, qacC, mhqR, mhqA) among others (Supplementary Tables S2 and S5).
Upon extreme environmental conditions, some beneficial bacteria can sporulate, turning on endospore form [1,79]. Regarding the genome of B. subtilis PTA-271, several genes are predicted to be involved in the sporulation process (Supplementary Table S4): spo (sporulation control), ger (germination control), cot (endospore external layer) and cw (spore cortex lytic enzymes), putatively enabling it to survive long lasting periods while preserving all beneficial strengths for plant profits.
2.4.- Direct confrontation with pathogens or aggressive molecules
Upon direct confrontation, Bacillus species also need to protect themselves against phytopathogens that may compete for resources . In addition to ROS protection, diverse transporters mediate antibiotic extrusion, whether specific to a substance or a group of substances. Regarding the genome of B. subtilis PTA-271, the specific transporters predicted would putatively confer it resistance towards: tetracyclin (tetA, tetR, tetD), fosfomycin (fosB), erythromycin (msrA, msrB), bacillibactin (ymfD), bacitracin (BceA, BceB, BcrC), bleomycin (ble) and riboflavin (ribZ, rfnT) for example. Among the non-specific transporters (or multidrug transporters) predicted in the genome of B. subtilis PTA-271 are: mepA, ebrA and ebrB; ykkD and ykkC; bmrA and bmr3; emrY, among others.
Bacillus species can additionally directly detoxify some pathogen aggressive molecules targeting plants, such as phytotoxins, by the mean of antitoxins or detoxifying enzymes such as transferases and CYP450s [81-82]. In the genome of B. subtilis PTA-271, the main transferases predicted are glutathione-S-transferases GST, malonyl-transferases MT, glucosyl-transferases GT and many others, while the main CYP450s predicted are mono-oxygenases and dioxygenases (Supplementary Table S5). Quenching enzymes constitute another lever for beneficial bacteria to directly target pathogen aggressive molecules, by preventing their QS-dependent production [8,83]. Indeed, Bacillus species share aiiA gene encoding for N-acetyl homoserine lactonase able to hydrolyze the lactone ring of the AHLs (Acyl-homoserine lactones) involved in the QS production of some pathogen virulent factors. The genome analysis of B. subtilis PTA-271 predicts such genes putatively encoding for quenching enzymes such as lactonases, β-lactamases, deaminases, deacetylases and other (de)acylases (Supplementary Table S6).
Polyketide synthases (PKS) are another type of transferases, namely acetyltransferases, described to produce plant beneficial molecules as microbicide for phytopathogens: the polyketides (PK) [84-85]. PK are a large group of natural products built from acyl-coenzyme A. Regarding the genome of B. subtilis PTA-271, 15 genes are predicted to encode for putative PKS, many others for acetyltransferases or for enzymes sharing similar part of the PKS functions (Supplementary Table S7). According to antiSMASH 5.1.0, B. subtilis PTA-271 genome predicts 11 secondary metabolites gene clusters, among which: 1 PKS cluster and 1 hybrid PKS-NRPS cluster (Supplementary Table S8).
An extensive range of plant indirect beneficial molecules are additionally produced by Bacillus spp. to directly alter pathogen fitness and aggressiveness, such as the RP (ribosomally synthesized peptides) and NRP (non-ribosomally synthesized peptides) antimicrobial molecules or effectors [20,86]. Some of them are predicted as encoded by the genome of B. subtilis PTA-271, such as: Baillaene (pksD), subtilosin (sboA, albG, albE, albD, albB, albA) and bacilysin (bacE, bacF, bacG) (Supplementary Table S3). Lipopeptides are other NRP antimicrobial molecules [50,87], which encoding genes are predicted in the genome of B. subtilis PTA-271 to putatively produce the powerful antifungal substances fengycin and surfactin (Supplementary Table S3). Fengycin is described as particularly active against filamentous fungi, causing structural hyphae deformations until membrane disruption, thus suppressing fungal proliferation in plant and their production of phytotoxins [50,87]. Besides antibiotics and surfactants, bacterial siderophores can also directly alter pathogen fitness and aggressiveness, by depriving pathogen growth of iron while providing it for plant growth . Regarding the genome of B. subtilis PTA-271, predicted genes putatively encode for the siderophore Bacillibactin (Supplementary Table S3). Lytic enzymes (CWDE) such as cellulases, proteases, chitinases and glucanases, are other important feature of Bacillus spp. that can both alter pathogen fitness and produce DAMPs. Regarding the genome of B. subtilis PTA-271, several genes are predicted to encode for putative CWDEs: 1 chitosanase (csn), 1 β-glucanase (bglS), 1 β-glucanase / cellulase (eglS) and about 80 proteases (Supplementary Table S3). These enzymes are considered as powerful fungicides since they are responsible for the degradation of key structural components of fungal cell walls .
Besides these NRP and RP antimicrobial molecules, the genome of B. subtilis PTA-271 also predicts the genes hcnC, acu and but, putatively encoding for the volatile antimicrobial compounds: VIC (hydrogen cyanide, HCN) and VOC (acetoin and 2,3-butanediol), respectively. HCN is described as a potent inhibitor of cytochrome C oxidase and several other metalloenzymes, and as extremely toxic to aerobic microorganisms at very low concentrations . Acetoin and 2,3-butanediol are described as weapons for phytopathogens [20,55].
According to COG categories, 2.30% of B. subtilis PTA-271 genome is predicted to be devoted to the production of secondary metabolites, considered as one of the most important features in biocontrol activities. AntiSMASH 5.1.0 predicts 11 secondary metabolites gene clusters in B. subtilis PTA-271 genome, among which 3 NRPS clusters and 2 RiPPs clusters (Supplementary Table S8).
3- B. subtilis PTA-271 GENOME Comparison with other genomes
To understand the magnitude of the differences between B. subtilis PTA-271 and other Bacillus strains, the PTA-271 genome has been compared to the complete genomes of 5 type-strains (B. subtilis NCIB 3610, B. subtilis 168, B. subtilis 9407, B. amyloliquefaciens subsp. plantarum strain FZB42, and B. velezensis KTCT 13012)  and 32 non-type strains, represented in Table 6. Among non-type strains showing ≥99% of the 16S ribosomal gene similarity with PTA-271 are 31 distinct strains of B. subtilis and 1 Bacillus velezensis. For this genomic comparison, was used the GGDC 2.1 web server , the DSMZ phylogenomics pipeline to estimate DNA-DNA hybridization (DDH) , and the JSpecies WS web server to estimate the Average Nucleotide Identity (ANI) through pairwise comparisons . The DDH value was estimated using the recommended formula (formula two) for draft genomes, at the GGDC website . The ANI values were calculated using Ezbiocloud . The whole data analysis enabled to obtain the intergenomic distances between genomes and their probability of belonging to the same species or subspecies. The general comparison of genomes is reported in Table 6, while the intergenomic distances (DDH estimate and ANI) are shown in Table 7.
Among the type strain genomes, the closer strain to B. subtilis PTA-271 was B. subtilis 9407, with a 0.0104 distance, a DDH estimate of 91.60%, and an ANIm of 99.02%. As expected, the most distant strain was B. velezensis KTCT 13012, with a 0.2268 distance, a DDH estimate of 19.40% and a 0% probability of being the same species, corroborated with an ANIm percentage of 77.02%. Concerning the non-type strain genomes, the closer strains to PTA-271 were B. subtilis QB5413, B. subtilis SRCM 104005, and B. subtilis QB61 with distances of 0.0112, 0.0119 and 0.0119 respectively, and DDH estimates of 90.90%, 90.20% and 90.20% respectively. The most distant strain was B. velezensis strain ATR2, with a distance of 0.2144 and a DDH estimate of 20.50% corroborated with an ANIm percentage of 77.1%. The most distant B. subtilis strain to PTA-271 was B. subtilis subsp. subtilis RO-NN-1 with a distance of 0.203 and a DDH of 82.60%.