Compatible Consortium of Endophytic Bacillus halotolerans Strains Cal.l.30 and Cal.f.4 Promotes Plant Growth and Induces Systemic Resistance against Botrytis cinerea

Simple Summary This study evaluated two Bacillus halotolerans-beneficial strains (Cal.l.30 and Cal.f.4) for the levels of their compatibility and their potential to be effective biofertilizers and in parallel biocontrol agents. We investigated their plant growth-promoting effect and ability to induce the expression of defense-related genes in the leaves of young tomato seedling plants, applying them individually or in combination under in vitro and greenhouse conditions using seed biopriming and soil drenching as inoculum delivery systems. The results showed that B. halotolerans strains (singly or in combination) are promising bioproducts for sustainable horticulture as they combine both direct antifungal activity against plant pathogens and the ability to prime plant immunity and enhance plant growth. Abstract Evaluating microbial-based alternatives to conventional fungicides and biofertilizers enables us to gain a deeper understanding of the biocontrol and plant growth-promoting activities. Two genetically distinct Bacillus halotolerans strains (Cal.l.30, Cal.f.4) were evaluated for the levels of their compatibility. They were applied individually or in combination under in vitro and greenhouse conditions, using seed bio-priming and soil drenching as inoculum delivery systems, for their plant growth-promoting effect. Our data indicate that application of Cal.l.30 and Cal.f.4 as single strains and as a mixture significantly enhanced growth parameters of Arabidopsis and tomato plants. We investigated whether seed and an additional soil treatment with these strains could induce the expression of defense-related genes in leaves of young tomato seedling plants. These treatments mediated a long lasting, bacterial-mediated, systemic-induced resistance as evidenced by the high levels of expression of RP3, ACO1 and ERF1 genes in the leaves of young tomato seedlings. Furthermore, we presented data showing that seed and soil treatment with B. halotolerans strains resulted in an effective inhibition of Botrytis cinerea attack and development on tomato leaves. Our findings highlighted the potential of B. halotolerans strains as they combine both direct antifungal activity against plant pathogens and the ability to prime plant innate immunity and enhance plant growth.


Introduction
To meet the growing demand for humans' food and animals' feed, increasing agronomic inputs such as chemical fertilizers and pesticides raise public concerns on the growth-promoting and plant protection abilities as mono-and co-cultures. Cal.l. 30 and Cal.f.4 were selected ahead of other isolates due to their ability to combine direct plant growth effect and strong antimicrobial activity against several plant pathogens with the secretion of novel secondary metabolites [16,17]. Strains were tested for their compatibility under in vitro conditions and their co-colonization capacity on A. thaliana roots. We used seed bio-priming and soil-drenching methods as inoculum delivery systems to apply both strains individually or in combination on tomato plants and evaluate their plant growth promotion. We further studied the activation of tomato plant defense mechanisms after the application of inoculants through seed priming and soil drenching and evaluated their ability to protect plants against gray mold disease. Finally, we evaluated the direct suppression of B. cinerea under ex vivo application on detached grape berries.

Bacterial Preparation
Both B. halotolerans Cal.l. 30 and B. halotolerans Cal.f.4 are natural endophytes isolated from Calendula officinalis, as recently described by Tsalgatidou et al., 2023 [16]. Routinely, both strains were inoculated in 5 mL liquid nutrient broth (NB) medium and incubated overnight with shaking on an orbital platform (200 rpm) at 30 • C. Bacteria were maintained on solid nutrient agar (NA) plates (1.5% w/v agar) until further use.

Compatibility Assays In Vitro
The in vitro compatibility of Cal.l. 30 and Cal.f.4 was studied by the modified mixed colony spot [18] and pairwise interaction methods [19]. Cal.l. 30 and Cal.f.4 have different colonial morphology and could be differentiated based on this. To further support the results, the rifampicin-resistant strain of Cal.l.30 (Cal.l.30-rif) was used to evaluate coexistence of Cal.l. 30 and Cal.f.4. A mixed colony spot was performed between bacterial strains Cal.l.30-rif and Cal.f.4 onto an NA medium in 5 cm diameter Petri dishes. Both bacterial strains, each containing~10 8 CFU/mL (OD 600 = 0.5)-based on standard curves of each bacterial strain-were equally combined, creating a 1:1 mixture. After mixing thoroughly, an aliquot of 10 µL of the mixture (5 µL of each bacterial strain) was spot inoculated on the agar surface and incubated at 30 • C for five days. Colony growth was captured and measured on a daily basis until the fifth day of observation. The bacterial cell density of mono-and co-culture colonies was calculated daily after pouring 1 mL of sterilized ddH 2 O onto petri plates, excluding the cells from the NA medium with a Pasteur pipette and, finally, serially diluting and plating 100 µL of the resuspended cells on NA plates supplemented with or without rifampicin. Plates were incubated overnight at 30 • C. Experiments were conducted in triplicate and repeated three times.
The pairwise interaction method was performed between B. halotolerans Cal.l. 30 and Cal.f.4 strains onto the NA medium in 5 cm Petri dishes. An aliquot of 5 µL of each bacterial strain of cell suspension of 10 8 CFU/mL (OD 600 = 0.5) was spot inoculated at a 1 cm distance between them. Each strain was singly inoculated as a control treatment. Plates were incubated at 30 • C for 7 days and images were taken on a daily basis. For each treatment, four biological replicates were run in three technical replicates.

Bacterial Interactions on A. thaliana Plantlets
Bacterial strains Cal.l. 30 and Cal.f.4 were studied for their in planta compatibility by the root colonization of A. thaliana seedlings and the plant growth effect, as mono-and co-cultures. A. thaliana seeds were surface sterilized and grown on solid half-strength Murashige and Skoog ( 1 /2 MS) medium (Duchefa Biochemie, Haarlem, The Netherlands) supplemented with 0.5% (w/v) sucrose and 0.8% (w/v) bacteriological agar (Sigma, Burlington, MA, USA) as previously described by Tsalgatidou et al., 2023 [16], until further use.
To evaluate the simultaneous A. thaliana root colonization ability of Cal.l. 30 and Cal.f.4 as a mixture, a modified protocol of Beauregard et al., 2013 [20] was conducted. Eight-dayold seedlings grown on solid 1 /2 MS medium were transferred to Eppendorf's containing Biology 2023, 12, 779 4 of 19 300 mL 1 /2 MS medium, inoculated with bacterial cells of~10 8 CFU/root (OD 600 = 0.5) of mono-(Cal.l. 30, Cal.f.4) and co-cultures (Cal.l.30 + Cal.f.4) and put on an orbital shaker at 100 rpm for 18 h. For this experiment, the rifampicin resistant strain Cal.l.30-rif was used. To determine colonization ability, plants were removed from each tube, washed two times in 1 mL of 1 /2 MS after shaking thoroughly for 30 sec and finally washed in 1 mL of 1 /2 MS with vortex agitation for 1 min to remove all cells slightly attached to the roots. Cell-containing solution was serially diluted with 1 /2 MS before plating on NA medium supplemented with rifampicin, only for Cal.l.30-rif determination. The CFU/root was determined 24 h after incubation at 30 • C. For each treatment, six biological replicates were run in three technical replicates. Single strain bacterial cells' colonization was visualized under an Olympus BX40 microscope.
The growth effect on A. thaliana morphological characteristics was determined after bacterial mono-(Cal.l. 30 and Cal.f.4) and co-culture (Cal.l.30 + Cal.f.4) inoculation at 3 cm distance from root tips, as previously described by Tsalgatidou et al., 2023 [16]. Both bacterial strains, each containing~10 8 CFU/mL (OD 600 = 0.5), were equally combined, creating a 1:1 mixture, and an aliquot of 10 µL cell suspension was inoculated. For each treatment six biological replicates were run in three technical replicates. Seedlings' fresh weight, primary root length, lateral roots and root hairs were determined as previously described by Tsalgatidou 30 and Cal.f.4 as a single-strain formulation and as a mixture under greenhouse conditions with a completely randomized block design with three replicates. Tomato seeds were kindly provided by Dr Costas Delis (Department of Agriculture at the University of the Peloponnese). Sterilization and bio-priming of tomato seeds were conducted, as previously described, by Thomloudi et al., 2021 [21]. Briefly, tomato seeds were surface sterilized by immersion in a 3% (v/v) aqueous solution of commercial bleach (5% sodium hypochlorite solution) for 6 min, rinsed six times in double distilled water and air-dried in a UV-sterilized laminar flow cabinet. Dried seeds were immersed in a solution containing a single strain and a 1:1 mixture of bacterial suspension adjusted to~10 8 CFU/mL (OD 600 = 0.5), supplemented with 1% (w/v) Carboxyl Methyl Cellulose (CMC) (Sigma-Aldrich ® , Merk KGaA, Burlington, MA, USA). Control seeds were immersed in a solution containing only CMC. Seeds were incubated for 1.5 h at 25 • C under agitation and then air-dried in a UV-sterilized laminar flow cabinet, until sowing in 7 × 7 cm 2 pots (one plant/pot) using a substrate of peat and perlite (2:1 v/v).
Additional bacterial application was done by the soil drenching (10 mL/pot) of Cal.l.30, Cal.f.4 and Cal.l.30 + Cal.f.4 (1:1 mixture) suspension adjusted to~10 8 CFU/mL, at 15 d 30 d after sowing. The pot experiment was conducted in a greenhouse (average air temperature 25 • C, relative air humidity 60% and an average 10 1 /2 h photoperiod) from October to December of 2019. Plant growth parameters (shoot fresh and dry weight, root fresh and dry weight and shoot length) were recorded 45 days after sowing (das). Plants were captured with a digital camera and shoot length was measured using ImageJ software v.1.8.0 [22].

Rhizosphere Colonization Assay
Bacterial rhizosphere colonization and soil survival were estimated at two time points: at 15 das, to measure the number of cells of each treatment in the rhizosphere soil with bio-priming being the only inoculation system, and at 45 days after two extra bacterial inoculations were conducted through soil drenching. For this experiment, the rifampicin resistant strain Cal.l.30-rif was used. Plants were carefully uprooted and, after removing most of the soil from the roots, one gram of the remaining soil tightly adhered to the roots was taken. The soil was transferred to a falcon tube containing 10 mL of sterilized ddH 2 O and vortexed for 1 min. The bacterial population of each treatment was enumerated at each sample day by a standard dilution plate count method on NA medium. The colonies of Cal.l.30-rif were counted by supplementing the NA medium with rifampicin. The CFU/g of soil were determined 24 h after incubation at 30 • C. For each treatment, four biological replicates were run in three technical replicates.

B. cinerea Infection Assay
After 45 days of bacterial inoculation through seed bio-priming and additional soil drenching applications at 15 and 30 das, B. cinerea was inoculated on tomato leaves in vivo. A droplet of 10 µL of spore suspension (5 × 10 5 CFU/mL) was inoculated on one side of the midvein of the leaves (one leaf/tomato plant). Inoculated plants were covered with transparent plastic to maintain high humidity and plants were left in the greenhouse for five more days. After 5 days post infection (dpi), the lesion area created by B. cinerea on tomato leaves was measured and disease severity index (DSI %) was calculated as described by Lee et al., 2006 [23]. The lesion area was measured using the ImageJ software v.1.8.0 analysis tool [22]. Disease severity (DS %) was calculated as the percentage of infected leaf area on which a 0-to-4 rating scale was created, with 0 = healthy leaves, 1 = 0.1-5%, 2 = 5.1-20%, 3 = 20.1-40% and 4 = 40.1-100% lesion area [23]. The disease severity index (DSI %) was estimated using the above severity scale and based on the following formula: % Disease Severity Index (DSI) = [∑(number of infected leaves in a specific value of disease rating scale × corresponding value of disease scale)/(maximum rating scale value × total number of leaves rated)] × 100.

Quantitative Analysis of Defense-Related Gene Expression
The induction of defense-related genes after bacterial formulation (Cal.l.30, Cal.f.4 and Cal.l.30 + Cal.f.4) through seed bio-priming and both seed bio-priming and an additional root drenching (15 d), was evaluated by quantitative real-time PCR. Tomato plantlets' leaf samples were collected for RNA extraction 15 das of bio-primed seeds and 24 h after an additional bacterial soil drenching application.
Total RNA was extracted from tomato leaves (100 µg/sample) of four different treatments (non-inoculated plants (control), inoculation with B. halotolerans Cal.l.30, inoculation with B. halotolerans Cal.f.4 and inoculation with the mixture of B. halotolerans strains Cal.l.30 + Cal.f.4) at two different time points (15 das of bacterial bio-primed seeds and 24 h after additional soil drenching) following the protocol from the manufacturer, Macherey-Nagel (NucleoSpin ® RNA Plus, Macherey-Nagel, Loughborough, Leicestershire, UK). DNA was DNAase eliminated and cDNA was synthesized with 3 µg of purified RNA using the PrimeScript 1st strand cDNA, Synthesis Kit (Takara Bio, Kusatsu, Shiga, Japan), according to the manufacturer's instructions. The conditions of RT-qPCR and the relative quantification of specific mRNA levels were performed using PowerUp SYBR Green Master Mix Assay (Applied Biosystems, Waltham, MA, USA)) containing 1 µL cDNA template, 10 µL SYBR GREEN PowerUp mix and 2 µL each of 10 µM tomato gene-specific primer pair (Supplementary Table S1). The cycling conditions were as follows: pre-incubation at 95 • C for 2 min, followed by 40 cycles at 95 • C for 15 sec, at 60 • C for 15 sec and at 72 • C for 1 min. Ubi3 gene coding for ubiquitin [24] was used as a housekeeping gene and for the normalization of gene expression. Primers used against genes coding for pathogenesisrelated proteins PR2; PR2 [25] and PR3; PR3 (Beacon Designer v.7.9), for proteins related to ethylene synthesis and response such as ERF1; ERF1 [26] and ACO1; ACO1 (Beacon Designer v.7.9), and the lipoxygenase A encoding gene TomloxA [27] that catalyzes the first step in the octadecanoid pathway as a response to biotic and abiotic stress, were selected.
The efficiency of the reaction for each target gene derived from LinRegPCR software [28] through linear analysis regression on the logarithm of the fluorescence intensity values in each cycle reaction. Relative levels of target gene transcripts for each sample were calculated based on the formula derived from the Pfaffl equation: Eref Cref /Ectarget cttarget [29].

Detached Fruit Assay
The direct biological control ability of B. halotolerans Cal.l.30 and B. halotolerans Cal.f.4 as individual strains and as a consortium was evaluated on detached grape berries (Vitis vinifera L. cv. Sultanina) against gray mold disease caused by B. cinerea as previously described by Tsalgatidou et al., 2022 [17]. Briefly, an aliquot of 10 µL of single strain (Cal.l. 30 and Cal.f.4) and a 1:1 mixture (Cal.l.30 + Cal.f.4) of bacterial suspension adjusted to~10 8 CFU/mL (OD 600 = 0.5) were inoculated on a 3 mm wound on the surface-sterilized grape berries (20 fruit/replicate). The grapes were incubated for 1 h at room temperature until infection with 10 µL of B. cinerea spore suspension (~1 × 10 5 CFU/mL). The grapes were placed into plastic boxes and transferred to a dark growth chamber at 25 • C for 5 d. After 5 dpi, the lesion area created by B. cinerea was measured and the disease severity index and disease incidence were calculated as previously described by Tsalgatidou et al., 2022 [17]. The lesion area was measured using ImageJ software v.1.8.0. Finally, the colonization ability on grapes of both single strains (Cal.f.4 and Cal.l.30) and their mixture was evaluated at 24 h post inoculation as previously described by Tsalgatidou et al., 2022 [17]. For this experiment, the rifampicin-resistant strain Cal.l.30-rif was used to determine colonization on grapes.

Statistical Analysis
Statistical analysis was performed using IBM SPSS Statistics for Windows, version 25 (IBM Corp., Armonk, NY, USA). Plots were created with Sigma Plot, version 12.0 (Systat Software, San Jose, CA, USA). In bacterial compatibility assays, A. thaliana and tomato pot experiments were performed and in detached grapes assays, statistical analysis was performed with ANOVA followed by Tukey's honestly significant difference (HSD) test (p < 0.05) to allow for comparisons among all means. To evaluate the bacterial cell number, data were transformed to the logarithmic scale followed by Tukey analysis. Plots indicate mean values with standard deviation as error bars. Different letters represent statistically significant differences among treatments. Data expressed as percentages were arcsine transformed prior to Tukey analysis (p < 0.05). For gene expression analysis, statistical analysis was performed on the logarithmic values of the relative expression (on ∆Ct) to ensure the assumptions of parametric statistical analysis where data follow a normal distribution and homogeneity of variances. Graphs indicate the mean of gene expression for each treatment and a 95% confidence interval (Mean ± CI).

in In Vitro Compatibility Assays
Members of a bacterial consortium are considered as compatible when they do not inhibit each other's growth during their in vitro co-culture and/or in rhizosphere colonization competition assays [15]. In this context, the in vitro compatibility of the strain mixture involving Cal.l. 30 and Cal.f.4 was evaluated by determining the level of co-existence during growth in colony and biofilm modes in a series of co-culture experiments. First of all, we dropped, inoculated, each strain either singly or as bacterial mixture (1:1) on an NA agar plate and measured the colony morphology ( Figure 1A) as well as the number of bacterial cells ( Figure 1D) in each of the evolving colonies. The results shown in Figure 1B,D indicate that the number of Cal.l. 30 and Cal.f.4 cells in the spot-inoculated two-strain mixturedeveloping colony remained almost identical when compared to the number of cells found in each monoculture, suggesting that Cal.l. 30 and Cal.f.4 strains grown in close proximity do not inhibit the growth of each other and, thus, can co-exist, occupying the same niche under the given environmental conditions. on an NA agar plate and measured the colony morphology ( Figure 1A) as well as the number of bacterial cells ( Figure 1D) in each of the evolving colonies. The results shown in Figure 1B,D indicate that the number of Cal.l. 30 and Cal.f.4 cells in the spot-inoculated two-strain mixture-developing colony remained almost identical when compared to the number of cells found in each monoculture, suggesting that Cal.l. 30 and Cal.f.4 strains grown in close proximity do not inhibit the growth of each other and, thus, can co-exist, occupying the same niche under the given environmental conditions. We also carried out experiments in which two different colonies (Cal.l. 30 and Cal.f.4) were inoculated on NA agar plates with a separation of 10 mm, and their external (in the direction opposite to the other colony) and internal (in the direction of the other colony) radii were periodically depicted and measured for 7 days. As shown in Figure 2A, when Cal.l. 30 and Cal.f.4 strains are inoculated in close proximity to each other, although there is a slight decrease of the Cal.f.4 internal radius as compared to the external one, the area covered by the interacting macrocolonies was largely unaffected as compared to singly grown macrocolonies ( Figure 2B,C). These data suggested that both strains interacted, and no antagonistic relationship was found, suggesting that Cal.f.4 coexisted well with Cal.30. We also carried out experiments in which two different colonies (Cal.l. 30 and Cal.f.4) were inoculated on NA agar plates with a separation of 10 mm, and their external (in the direction opposite to the other colony) and internal (in the direction of the other colony) radii were periodically depicted and measured for 7 days. As shown in Figure 2A, when Cal.l. 30 and Cal.f.4 strains are inoculated in close proximity to each other, although there is a slight decrease of the Cal.f.4 internal radius as compared to the external one, the area covered by the interacting macrocolonies was largely unaffected as compared to singly grown macrocolonies ( Figure 2B,C). These data suggested that both strains interacted, and no antagonistic relationship was found, suggesting that Cal.f.4 coexisted well with Cal.30. Compatibility of the two B. halotolerans strains was also evaluated by investigating whether antagonism between two B. halotolerans strains occurring on the root may affect population density of each strain, using an in vitro root colonization assay. As shown in Figure 3B,C, both bacterial strains can successfully colonize the root surface of A. thaliana plantlets either as single cells or as aggregates (black arrows). Inoculation of Cal.l. 30 and Cal.f.4 as single cultures on hydroponic grown A. thaliana Col-0 resulted in similar levels (3-4 × 10 4 CFU/root) of colonization by each species, whereas inoculation with a mixture (1:1) of both strains resulted in increased and cumulative colonization of roots by both strains (6-7 × 10 4 CFU/root from each strain) ( Figure 3A). These data suggested that both bacterial strains could co-exist on roots in a synergistic manner.  Compatibility of the two B. halotolerans strains was also evaluated by investigating whether antagonism between two B. halotolerans strains occurring on the root may affect population density of each strain, using an in vitro root colonization assay. As shown in Figure 3B,C, both bacterial strains can successfully colonize the root surface of A. thaliana plantlets either as single cells or as aggregates (black arrows). Inoculation of Cal.l. 30 and Cal.f.4 as single cultures on hydroponic grown A. thaliana Col-0 resulted in similar levels (3-4 × 10 4 CFU/root) of colonization by each species, whereas inoculation with a mixture (1:1) of both strains resulted in increased and cumulative colonization of roots by both strains (6-7 × 10 4 CFU/root from each strain) ( Figure 3A). These data suggested that both bacterial strains could co-exist on roots in a synergistic manner.

Application of B. halotolerans Strains in Combination and Individually Enhanced the Growth Parameters of Arabidopsis Plants Grown under In Vitro Conditions
As a first step towards evaluating the plant growth effect of B. halotolerans strains, we evaluated the capacity of Cal.l.30 + Cal.f.4 consortium to promote plant growth in comparison with single-strain applications, using an in vitro A. thaliana co-cultivation approach ( Figure 4A). Treatment with Cal.l.30, Cal.f.4 and a Cal.l.30 + Cal.f.4 mixture significantly increased the growth of A. thaliana as evidenced by the enhanced values of all growth parameters measured (fresh weight, number of lateral roots and root hairs) ( Figure 4B). Treatment of Arabidopsis by the two-strain mixture resulted in plants with higher total fresh weight and lateral roots compared to the single-strain formulations and the control plantlets ( Figure 4B). The single-strain formulation of Cal.l.30 resulted in the highest root hair number compared to the other two formulations.
plantlets either as single cells or as aggregates (black arrows). Inoculation of Cal.l. 30 and Cal.f.4 as single cultures on hydroponic grown A. thaliana Col-0 resulted in similar levels (3-4 × 10 4 CFU/root) of colonization by each species, whereas inoculation with a mixture (1:1) of both strains resulted in increased and cumulative colonization of roots by both strains (6-7 × 10 4 CFU/root from each strain) ( Figure 3A). These data suggested that both bacterial strains could co-exist on roots in a synergistic manner.

Application of B. halotolerans Strains in Combination and Individually Enhanced the Growth Parameters of Arabidopsis Plants Grown under In Vitro Conditions
As a first step towards evaluating the plant growth effect of B. halotolerans strains, we evaluated the capacity of Cal.l. 30 30 and Cal.f.4 as mono-and co-cultures on different growth parameters (total fresh weight, primary root length, lateral root number and root hair number) of A. thaliana seedlings under in vitro conditions. Statistical analysis was performed using one-way ANOVA followed by Tukey's multiple comparison post hoc test (p < 0.05). Different letters suggest statistical differences among treatments.

Application of B. halotolerans Strains as a Mixture and Individually Enhanced Growth Parameters of Tomato Plants under Greenhouse Conditions
The plant growth promotion capacity of Cal.l. 30 and Cal.f.4 was investigated when single strains and bacterial mixtures were applied as tomato seed treatment followed by two additional soil drenching treatments at 15 and 30 das under greenhouse conditions. Plants were harvested at 45 das and growth parameters were measured (total shoot and root, fresh and dry weight and shoot length). As illustrated in Figure 5, seed and soil treatment with Cal.l. 30 and Cal.f.4, individually or in combination, resulted in a pronounced growth of tomato plants' growth parameters compared to untreated plants (control). The results presented in Figure 5 revealed that compared to the control, maximum shoot and root biomass was recorded in plants treated with Cal.l.30, followed by plants treated with Cal.l.30 + Cal.f.4 ( Figure 5C,D). Examination of plants' rhizospheric bacterial populations revealed that Cal.l. 30 and Cal.f.4 remained viable in interactions with roots at 15 and 45 das, supporting their long lasting effects to plants.  30 and Cal.f.4 as mono-and co-cultures on different growth parameters (total fresh weight, primary root length, lateral root number and root hair number) of A. thaliana seedlings under in vitro conditions. Statistical analysis was performed using one-way ANOVA followed by Tukey's multiple comparison post hoc test (p < 0.05). Different letters suggest statistical differences among treatments.

Application of B. halotolerans Strains as a Mixture and Individually Enhanced Growth Parameters of Tomato Plants under Greenhouse Conditions
The plant growth promotion capacity of Cal.l. 30 and Cal.f.4 was investigated when single strains and bacterial mixtures were applied as tomato seed treatment followed by two additional soil drenching treatments at 15 and 30 das under greenhouse conditions. Plants were harvested at 45 das and growth parameters were measured (total shoot and root, fresh and dry weight and shoot length). As illustrated in Figure 5, seed and soil treatment with Cal.l. 30 and Cal.f.4, individually or in combination, resulted in a pronounced growth of tomato plants' growth parameters compared to untreated plants (control). The results presented in Figure 5 revealed that compared to the control, maximum shoot and root biomass was recorded in plants treated with Cal.l.30, followed by plants treated with Cal.l.30 + Cal.f.4 ( Figure 5C,D). Examination of plants' rhizospheric bacterial populations revealed that Cal.l. 30 and Cal.f.4 remained viable in interactions with roots at 15 and 45 das, supporting their long lasting effects to plants.

Seed Priming and Additional Soil Drenching with Cal.l.30 and Cal.f.4 Strains Singly or in Combination Protect Tomato Plants against Gray Mold Disease
To evaluate whether seed and soil treatment with Cal.l. 30

Activation of Long-Lasting Defense-Related Genes of Tomato Plantlets
Besides the impact of B. halotolerans strains in control efficacy against B. cinerea, to understand whether seed bio-priming and soil drenching with Cal.l.30, Cal.f.4 and Cal.l.30 + Cal.f.4 could induce the expression of the long-lasting defense-related genes, RT-qPCR analysis was conducted. Therefore, we analyzed the expression of the SAresponsive genes PR3 (encoding the PR protein 2 β-1,3-glucanase) and PR2 (encoding the PR protein 2 β-1,3-glucanase), the JA synthesis-related gene TomLoxA (encoding for lipoxygenase A), the ET-responsive gene ERF1 and the ethylene biosynthesis-related gene ACO1 in the leaves of young plantlets after being seed bio-primed with Cal.l. 30 As illustrated in Figure 7A, the expression level of ERF1, ACO1 and PR3 at 15 das were strong induced in plants inoculated with Cal.l.30 through seed bio-priming in comparison to the control (unprimed seeds). Specifically, gene transcript levels of ERF1, ACO1 and PR3 were up-regulated 1.98-, 1.68-and 4.12-fold, respectively, compared to the control ( Figure 7C

Activation of Long-Lasting Defense-Related Genes of Tomato Plantlets
Besides the impact of B. halotolerans strains in control efficacy against B. cinerea, to understand whether seed bio-priming and soil drenching with Cal.l.30, Cal.f.4 and Cal.l.30 + Cal.f.4 could induce the expression of the long-lasting defense-related genes, RT-qPCR analysis was conducted. Therefore, we analyzed the expression of the SA-responsive genes PR3 (encoding the PR protein 2 β-1,3-glucanase) and PR2 (encoding the PR protein 2 β-1,3-glucanase), the JA synthesis-related gene TomLoxA (encoding for lipoxygenase A), the ET-responsive gene ERF1 and the ethylene biosynthesis-related gene ACO1 in the leaves of young plantlets after being seed bio-primed with Cal.l. 30 As illustrated in Figure 7A, the expression level of ERF1, ACO1 and PR3 at 15 das were strong induced in plants inoculated with Cal.l.30 through seed bio-priming in comparison to the control (unprimed seeds). Specifically, gene transcript levels of ERF1, ACO1 and PR3 were up-regulated 1.98-, 1.68-and 4.12-fold, respectively, compared to the control ( Figure 7C). Tomato plants raised from Cal.f.4 bio-primed seeds presented up-regulated gene transcript levels of ACO1 and PR3 genes (1.63-and 3.49-fold, respectively), while ERF1 expression was down-regulated ( Figure 7C). Bio-primed plants with a Cal.l.30 + Cal.f.4 consortium presented a significant up-regulation (2.58-fold) of PR3 transcripts compared to control plants, while ERF1 and ACO1 gene expressions presented a similar level in gene expression in both non-and bio-primed plants ( Figure 7C). Expression of PR2 was only observed in plants inoculated with Cal.l.30 + Cal.f.4 consortium ( Figure 7A). These results indicated that defense mechanisms were less activated in plants with combined bacterial priming treatment compared to the individual treatments.
These results indicated that defense mechanisms were less activated in plants with combined bacterial priming treatment compared to the individual treatments. The gene expression levels of ERF1, ACO1, PR2, PR3 and TomLoxA were significantly differentiated after both seed bio-priming and additional bacterial formulation via soil drenching at 15 das ( Figure 7B). As reported in Figure 7D  The gene expression levels of ERF1, ACO1, PR2, PR3 and TomLoxA were significantly differentiated after both seed bio-priming and additional bacterial formulation via soil drenching at 15 das ( Figure 7B). As reported in Figure 7D halotolerans Cal.f.4 as singlestrain biological control agents against B. cinerea in ex vivo applications has already been tested in previous studies [16,17], the application of these strains as a consortium is studied here for the first time. All three applications were tested on detached grape berries in a small-scale storage experiment against B. cinerea. Both single bacterial applications, as well as their mixture, successfully colonized grapes at the point of application, showing a similar cell number (1-3 × 10 7 CFU/wound).
As presented in Figure 8A,B, Cal.l. 30 and Cal.f.4 single-strain applications, as well as their 1:1 mixture, presented a strong inhibition against B. cinerea by reducing both the number of infected grapes and the brown lesion area. Grapes infected only with the pathogenic fungus developed extended brown lesion and visible mycelia growth around the point of formulation and reached 100% infection rate for B. cinerea ( Figure 8A). Negative control grape berries that were wounded artificially but not infected with the fungal pathogen appeared to be unaffected after 5 days of storage. Both single-strain formulations significantly reduced the disease severity index, with strain Cal.l.30 being the most efficient with 31% of DSI and 81% of DI, in comparison to the 46% DSI and 97% DI of Cal.f.4 ( Figure 8C,D). However, treatment with Cal.l.30 + Cal.f.4 synergistically suppressed disease severity, reaching the lowest DI value of 72% and DSI value of 28% at 5 day of incubation, presenting, along with the Cal.l.30 single-strain formulation, the most impressive results.
Although the efficacy of B. halotolerans Cal.l.30 and B. halotolerans Cal.f.4 as singlestrain biological control agents against B. cinerea in ex vivo applications has already been tested in previous studies [16,17], the application of these strains as a consortium is studied here for the first time. All three applications were tested on detached grape berries in a small-scale storage experiment against B. cinerea. Both single bacterial applications, as well as their mixture, successfully colonized grapes at the point of application, showing a similar cell number (1-3 × 10 7 CFU/wound).
As presented in Figure 8A,B, Cal.l. 30 and Cal.f.4 single-strain applications, as well as their 1:1 mixture, presented a strong inhibition against B. cinerea by reducing both the number of infected grapes and the brown lesion area. Grapes infected only with the pathogenic fungus developed extended brown lesion and visible mycelia growth around the point of formulation and reached 100% infection rate for B. cinerea ( Figure 8A). Negative control grape berries that were wounded artificially but not infected with the fungal pathogen appeared to be unaffected after 5 days of storage. Both single-strain formulations significantly reduced the disease severity index, with strain Cal.l.30 being the most efficient with 31% of DSI and 81% of DI, in comparison to the 46% DSI and 97% DI of Cal.f.4 ( Figure 8C,D). However, treatment with Cal.l.30 + Cal.f.4 synergistically suppressed disease severity, reaching the lowest DI value of 72% and DSI value of 28% at 5 day of incubation, presenting, along with the Cal.l.30 single-strain formulation, the most impressive results.

Discussion
Plant growth-promoting bacteria (PGPB) can provide multiple benefits in agriculture by mediating enhancement in crop productivity and nutrient content and suppressing the growth of pathogens. Although members of the B. subtilis species complex have been extensively studied for their ability to promote vegetative plant growth, reduce disease incidence and promote plant tolerance against fungal and bacterial pathogens by inducing the systemic resistance of host plants, the quest for new Bacillus-based biological control agents (BCAs) along with effective methods for their delivery is an ongoing process [30,31].

Discussion
Plant growth-promoting bacteria (PGPB) can provide multiple benefits in agriculture by mediating enhancement in crop productivity and nutrient content and suppressing the growth of pathogens. Although members of the B. subtilis species complex have been extensively studied for their ability to promote vegetative plant growth, reduce disease incidence and promote plant tolerance against fungal and bacterial pathogens by inducing the systemic resistance of host plants, the quest for new Bacillus-based biological control agents (BCAs) along with effective methods for their delivery is an ongoing process [30,31].
B. halotolerans strains Cal.l. 30 and Cal.f.4 have a wealth of plant growth-promoting traits and exhibit a strong antimicrobial activity against fungal pathogens through the production of multiple antifungal metabolites [16,17]. The present study provides evidence that Cal.l. 30  Recent studies have demonstrated that seed bio-priming and soil drenching are the most promising methods for delivering the beneficial microbes to the host plant when PGPB are evaluated for their plant growth-promoting activity under greenhouse or field conditions [32,33]. Furthermore, it has been demonstrated that tomato, cow pea and mung bean seed bio-priming followed soil drenching induced significant plant growth parameters as compared to single treatment [34,35]. In this regard, in our studies, we evaluated the plant growth effect of B. halotolerans Cal.l.30 and B. halotolerans Cal.f.4 when applied to tomato, singly or in combination, through seed bio-priming and additional soil drenching. Our data demonstrated that tomato seed bio-priming followed by soil drenching with all inoculum formulations (Cal.l. 30, Cal.f.4 and Cal.l.30 + Cal.f.4) gave better results in plant growth-promotion parameters compared to untreated plants. These findings are in line with previous studies showing that tomato plants inoculated through seed treatment and soil drenching methods with Pseudomonas sp. strain S3 led to more improved morphological characteristics (root/shoot length, fresh weight and dry weight) compared to a single-method treatment [36]. It should be pointed out that, in contrast to Arabidopsis studies where the bacterial mixture showed a better performance compared to the individual strains, when applied on tomato plants the effect of the Cal.l.30 + Cal.f.4 mixture was more pronounced when compared to the single-strain Cal.f.4 treatment, and somewhat inferior compared to treatment with Cal.l.30, suggesting that consortia may provide better plant growth-promoting effects than some individual strains. This is in line with another study, where it was observed that the effect of seed bio-priming and soil drenching in biomass-promoting effects on Withania somnifera is more pronounced when B. amyloliquefaciens and Pseudomonas fluorescens were applied as a mixture compared to individual treatments [5].
These findings highlights that the plant growth-promoting effect observed in tomato plants treated with Cal.l. 30 and Cal.f.4 could partially be due to the known plant growthpromoting traits of Cal.l. 30 and Cal.f.4, such as production of IAA and siderophores, phosphate solubilization [16] and/or due to the PGPB-mediated long-lasting induction of the expression of genes related to plant growth and development [37][38][39].
Our data indicated that tomato plants treated with Cal.l. 30 and Cal.f.4 through seed bio-priming and soil drenching showed a defense-priming effect; plants raised from Cal.l. 30 and Cal.f.4-treated seeds showed elevated levels of transcripts of SA and JA/ET defenserelated genes, while plants raised from the Cal.l.30 + Cal.f.4 mixture-treated seeds showed enhanced levels of SA-responsive genes. However, plants raised from Cal.l. 30, Cal.f.4 or Cal.l.30 + Cal.f.4 bio-primed seeds that also received a soil drenching application displayed a relatively high level of expression of SA and JA/ET defense-related marker genes, indicating that Cal.l. 30 and Cal.f.4 strains, singly or in combination, do not only sensitize the plants' SA and JA/ET immune responsive systems but also enable the plant to enter the defense-priming state. In this regard, plant's priming is able to maintain a long-lasting state of readiness that does not confer resistance per se, as the accumulation of defense-related transcripts is below the threshold for effective resistance, but rather allows accelerated activation of inducible defense as soon as a pathogen or pest invasion occurs [40,41].
These findings are consistent with previous studies where it was reported that seed treatment followed by soil drenching with certain beneficial ISR-inducing microorganisms, including Bacillus sp. and Trichoderma spp., reprograms their transcriptional apparatus and up-regulates inducible defenses (enhanced expression of SA-and/or ET-defense genes and/or genes coding antioxidative enzymes and enzymes involved in callose deposition and lignification) [9,37,38,[42][43][44][45][46].
The duration of the priming state in tomato plants raised from bio-primed seeds or root-drenched plants is not known. Priming-mediated defenses can be durable and maintain throughout the plant's life cycle and can even be transmitted to the subsequent generation [9]. However, recent studies demonstrated that the inoculum application mode, and most importantly the number of applications, is important for the duration and efficacy of PGPM (plant growth-promoting microorganism) priming-mediated defense [47]. Seed dressing and soil drenching with species of B. amyloliquefaciens FS6 [48], Bacillus subtilis JW-1 [49], Bacillus thuringiensis SE and Pseudomonas aeruginosa KA19 [50] and Bacillus velezen-sis SZAD2 [51], were found to be effective methods for the control of fungal pathogens. However, when seed treatment was followed by soil drenching, the biocontrol efficacy appeared to be more effective than either treatment alone [48,49]. Suppression of Rhizoctonia solani was found more effective after two soil drenching applications of B. amyloliquefaciens FZB42 at short time intervals, compared to single treatment of tomato plants through soil drenching [52]. Similarly, suppression of potato virus Y was found more in tomato plants subjected to B. amyloliquefaciens MBI600 triple soil drench application compared to a single application through soil drenching prior to germination [53]. In addition, suppression of Fusarium was found more when tomato plants were treated with MBI600 dual drenching (soil drenching prior to germination followed by additional soil drenching), compared to single soil drenching just after sowing [33,46]. Taken together, these data suggested that the back-to-back application of ISR-inducing PGPB (e.g., seed bio-priming followed by soil drenching, seedling treatment), as compared to a single ISR-inducing PGPB application, may prolong the priming defense state of the host plant and, thus, keep the elevated levels of gene expression of defense-related genes. This offers better prospects for an accelerated activation of inducible defenses as soon as an invasion occurs [39,40].
In our studies, the biocontrol efficacy of Cal.l. 30, Cal.f.4 and Cal.l.30 + Cal.f.4 was evaluated on tomato plants raised from bio-primed seeds that received two additional doses of bacteria by soil drenching and then challenged with B. cinerea. Under this scheme of inoculum application and seed bio-priming followed by two soil drenchings at short time intervals, tomato plants may sustain the primed defense state (elevated levels of gene expression of SA and ET/JA defense-related genes), thus possessing the means to develop a fast and effective defense response upon facing the ingressive B. cinerea. Based on their disease-suppression abilities, when plants were treated with Cal.l.30 + Cal.f.4, a significant reduction in the plant disease severity was achieved with respect to the control, but this result was not significantly different than those observed in the individualstrain treatments. However, plants treated with Cal.l.30 + Cal.f.4 exhibited better disease inhibition development around the infection zone compared to Cal.l. 30 or Cal.f.4 treated plants, suggesting a strong synergistic effect against B. cinerea ingression and development in tomato leaves.
The ability of PGPM in sensitizing and priming the plant immune defense mainly relies on extracellular compounds synthesized and secreted by the beneficial microbes [54]. Recent studies have demonstrated that cell-free culture filtrates (CFCF) of ISR-inducing beneficial B. amyloliquefaciens MBI600, when inoculated by leaf spraying or root drenching, activated the expression of SA and/or JA defense-related genes [12]. The ISR-eliciting activity of CFCF was attributed to Bacillus sp.-secreted metabolites with antibiotic functions, such as iturinic-type lipopeptides, iturin A and bacilomycin D, as well as to the lipopeptides fengycin and surfactin [55]. In particular, it has been demonstrated that the spraying of A. thaliana leaves with iturin A or mucosubtilin activated the expression of SA and JA defense-related genes, PR1 and PDF1.2 [56], while A. thaliana seedling treatment with mycosubtilin from B. subtilis J-15 activated the expression of ET and JA defense-related genes [57]. Similarly, root treatment of tomato plants with fengycin stimulated the induction of SA-responsive genes [58]. In addition, treatment of mandarin fruit with each of the lipopeptides Iturin A, fengycin and surfactin secreted from B. subtillis ABS-S14 activated the expression of SA and JA defense-related genes [59], while treatment of tomato fruit with iturin A or bacilomycin D activated the expression of SA defense-related genes [60,61]. Previous integrated genomic and metabolomic analysis demonstrated that B. halotolerans strains Cal.l.30, Cal.f.4 and Hil4 have the ability to constitutively and simultaneously synthesize and secrete the lipopeptides fengycin, surfactin and the iturinic-type lipopeptide mojavensin [17,21], suggesting that colonization of plant tissues with these B. halotolerans strains may activate and sustained the defense-priming state of the host plant.

Conclusions
Beneficial microorganisms have been thought to trigger mild plant responses that do not affect the plant host defense-related transcriptome overall and contribute to sensitization rather than the actual induction of defense-related genes [7]. In the present study, B. halotolerans strains Cal.l. 30 and Cal.f.4 belong to a special group of beneficial microorganisms that are able not only to sensitize the induction of defense-related genes upon pathogen challenge but also enable the host plant to maintain elevated levels of defense components around the threshold for effective resistance. Cal.l. 30 and Cal.f.4 created a compatible consortium with both plant growth-promoting effects and direct biocontrol activity against B. cinerea. Furthermore, our findings highlighted that seed treatment followed by the soil drenching of tomato plants with the novel ISR-inducing B. halotolerans strains Cal.l. 30 and Cal.f.4, as both individual strains and as a compatible consortium, could be an effective and practical approach for controlling gray mold disease under greenhouse crop production conditions. In this regard, the ability of B. halotolerans strains to promote growth and activate plants' innate immune responses, along with their potential to synthesize numerous ISR-inducing and antifungal metabolites, make these beneficial Bacillus strains important for the development of effective biological control agents.