All wild-type A. thaliana (Col-0) and mutant (sid2, npr1, jar1, and myc2) seeds were obtained from the Arabidopsis Biological Resource Center (ABRC) in Columbus, Ohio. The homozygous lines were identified based on the information from a published report (Jarret and McCluskey, 2019) and salk website (http://signal.salk.edu/). A. thaliana plants were grown in a potting mix (Lambert LM-111) in a walk-in growth chamber (Winnipeg, MB, Canada) at 22 °C with 12-hour light, 12-hour darkness and ~60% relative humidity (Yoo et al., 2022). The light intensity was adjusted as ~120 μmol/m² s (Zhao et al., 2020). Tobacco and tomato seeds were purchased from Park Seed Company at U.S.A. Plants were grown in a professional potting mix (Lambert LM-111) in a greenhouse compartment with air temperatures set at 28 °C, day/night cycles with 16-hour light and 8-hour darkness, and relative humidity at ~45%. Plants were fertilized with Osmocote’s general fertilizer (14-14-4) (Calhoun et al., 2021) to support the needed nutrients.

OSUB18 was cultured on Tryptic Soy Agar (TSA) plates at 28 °C for 2 days before the bacteria were transferred tosterile 15 mL tubes. 2 mL sterile water was used to wash out all bacterial residues from each TSA plate. TSA plates without OSUB18 were also washed with 2ml sterile water and used as the controls. The 15 mL tubes with OSUB18 cells or controls were homogenized with pipette tips and then transferred to new sterile 2 mL tubes and centrifuged at 12000 rpm for 10 minutes. The above clear cell-free supernatant (CFS) was defined as 1x BEE and stored at -80°C until use. 10x BEE was obtained by concentrating the 1xBEE from 2 mL to 200 µL with a lyophilizer machine. The 5x BEE was diluted two times from the 10x BEE with sterile water.

OSUB18 was cultured on Tryptic Soy Agar (TSA) plates at 28 °C. Phytopathogenic P. syringae pv. tomato DC3000 (Pst DC3000) was cultured on King’s B Agar (KBA) plates containing 50 mg/L rifampicin at 28 °C. The grey mold fungus B. cinerea was cultured on Potato Dextrose Agar (PDA) plates at 22 °C. Antagonism of OSUB18 against B. cinerea was conducted on Potato Soy Agar (PSA, made of PDA and TSA at 1:1 ratio) plates by the dual-culture test. In brief, one agar disc of B. cinerea was first placed on one side of the PSA plate. Then, OSUB18 cells (10ul per drop at 10^8 CFU/ml) were spotted on the opposite side away from the fungal plugs in the plates. The plates were then incubated for five days at 22 °C before the fungal growth was measured. The control set was conducted simultaneously by replacing the OSUB18 cells with sterile water. A similar dual-culture assay was carried out to test the direct inhibition effect of OSUB18 against Pst DC3000. In brief, Pst DC3000 cells (100ul, 10^8 CFU/ml) were evenly distributed on KBA by sterile glass beads (Eyler, 2013). OSUB18 cells (20ul, 10^8 CFU/ml) were spotted to the same plate ~3 cm apart from the plate center. The plates were then incubated for two days at 28 °C before measuring the inhibition zone. The OSUB18/Pst DC3000 cells were collected freshly and washed 3 times with sterile water to remove the nutrient residues before use. The above methods were modified from previous reports (Hu et al., 2014; Chen et al., 2018). All experiments were repeated three times with four plates as the replicates for each treatment.

The bacterial 16S ribosomal DNA (16S rDNA) sequence of OSUB18 was amplified by polymerase chain reaction (PCR) from its genomic DNA extracted by the Quick-DNA Fungal/Bacterial Micro prep Kit (Zymo Research, Irvine, CA, USA) with the primer pair 799F/1193R (Chen et al., 2020). The purified 16S rDNA sequence was subjected to the Basic Local Alignment Search Tool (BLAST) provided by the National Center for Biotechnology Information (NCBI). We used the Molecular Evolutionary Genetics Analysis (MEGA) software (Tamura et al., 2021) to construct the phylogenetic tree with the Gram positive bacterium Deinococcus radiophilus DSM 20551T as an out-cluster control.

The leaves of the 4~6-week-old Arabidopsis plants were syringe-injected with Pst DC3000 at 1 x 10⁶ CFU/mL, dipping inoculated, or spray inoculated with Pst DC3000 at 5 x 10⁸ CFU/mL containing 0.05% Silwet L-77 based on the experimental needs. Three days after inoculation, the bacterial growth was quantified by measuring Pst DC3000 growth in infected leaves using the serial dilution technique (Katagiri et al., 2002; Yao et al., 2013).

The leaves of the 4~6-week-old Arabidopsis plants were inoculated with B. cinerea spores (10ul-drop, 5 x 10⁵ spores/mL in half-strength V8 juice) and covered with a plastic dome to maintain high humidity to facilitate disease development at 22 °C. Three days later, the fungal disease symptom was assessed by measuring the lesion size using the ImageJ software (Weiberg et al., 2013).

The A. thaliana seedlings were transplanted into pots (3.5” x 3.5” x 2”). They were then subjected to a weekly treatment of OSUB18 (10⁷ CFU/mL, 50ml per pot) or water (Ctrl) as the root drench for three consecutive weeks. A week after the last root drench, the treated A. thalianaplants were used for bacterial or fungal pathogen inoculation to check ISR responses.

Superoxide anion (•O₂ ⁻) detection by nitroblue tetrazolium (NBT) staining was performed as described by Jambunathan (Jambunathan, 2010). Hydrogen peroxide (H₂O₂) detection by diaminobenzidine tetrahydrochloride (DAB) staining was performed based on a previously published method (Nie et al., 2017). Callose deposition was measured following a previous method (Wang et al., 2017). Because the bacterial pathogen was infiltrated into the plant leaves for the infection assay and the fungal pathogen was drop-inoculated on the plant leaves, different patterns of callose deposition were induced. Therefore, they were quantified as callose/mm2 (Jin and Mackey, 2017) in the bacterial infection assay and relative callose (Nie et al., 2017) in the fungal infection assay with Image J software (https://imagej.nih.gov/ij/ ). ROS burst assay was performed following a published protocol (Sang and Macho, 2017).

A. thaliana (Col-0) plants were root-drenched with OSUB18 or water (Ctrl) for bacterial or fungal phytopathogen infection assay. At 0 or 24 hours post-infection, the infected leaves were sampled for plant defense-related hormone extraction. SA, SA glucoside (SAG), JA, and JA isoleucine (JA-IIe) were extracted and quantified from ~0.1g fresh weight leaf tissues, as described previously (Zhao et al., 2022b). In brief, phytohormones were extracted from leaf samples with a methanol buffer with the addition of acetic acid and isotope-labeled internal standards (²H₄-SA and ²H₂-JA, CDN Isotopes) on ice. The supernatant was collected after being centrifued at 4 °C and analyzed by UPLC-MS on the Thermo Fisher Ultimate 3000 system (Thermo Fisher Scientific) equipped with a C18 (100 mm × 2.0 mm) column (Waters company, Milford, MA, USA). The solvent gradient used was 100%A (99.9% H₂O: 0.1% CHOOH) to 100%B (99.9% CH3CN: 0.1% CHOOH) at 200 ul/min. The endogenous and isotope-labeled SA and JA were detected using the following mass transitions: SA 137 > 93, 2H4 SA 141 > 97, SAG 299 > 93, JA 209 > 59, 2H2-JA 211 > 61, JA-IIe 247 > 97.

The total RNA of A. thaliana (Col-0) leaves with 0 or 24 hours of the bacterial or fungal pathogen infection was extracted with TRIzol Reagent (Invitrogen, Carlsbad, MA, USA). First-strand cDNA was synthesized from 1 µg total RNA with a reverse transcription kit (Applied Biosystem, Waltham, MA, USA). The qRT-PCR was performed on the CFX96 real-time PCR machine (Bio-Rad, Hercules, CA, USA) with UBIQUITIN10 (UBQ10; At4g05320) as the internal reference gene. Table S1 lists the gene-specific primers. A standard curve was made by determining the threshold cycle (Ct) values for each primer pair’s dilution series of the cDNA product. The gene expression was calculated using the ΔC_T method using the UBQ10 as the reference gene, following the Bio-Rad Real-Time Application Guide. In summary, the relative quantity of each gene is expressed in comparison to UBQ10 using the formula 2^ (Ct ^UBQ10 -Ct ^GENE ), where 2 represents the perfect PCR efficiency. For each qRT-PCR reaction, the Ct was determined by setting the threshold within the logarithmic amplification phase. The ΔC_T method is a variation of the Livak (2^-ΔΔCT) method. The ΔC_T method is simpler to perform and gives essentially the same results as the Livak (2^-ΔΔCT) method (BIO-RAD Real-Time PCR Applications Guide).

The O-CAS assay was conducted to detect the bacterial siderophore (Pérez-Miranda et al., 2007). The bacterial exopolysaccharide (EPS) was detected as previously described (Mu’minah et al., 2015). The bacterial acetoin and diacetyl were detected according to the previous report (Khalaf and Raizada, 2018). The bacterial hydrogen cyanide (HCN) was detected according to the previous study (Castric and Castric, 1983). The ammonia production was tested according to the previous report (Vyas et al., 2018).The production of indole acetic acid (IAA) and phosphate solubilization assays were performed according to the previous studies (Maheshwari et al., 2015; Batista et al., 2021).The organic acid was detected on Bromocresol purple Agar (BCPA) with 5′,5″-dibromo-o-cresolsulfophthalein as a pH indicator. The catalase activity test was performed according to the previous report (Odds, 1981).All the related experiments had at least 4 replicates and had been repeated 3 times with consistent results.

All statistical analyses were performed using the SPSS software. Data were summarized as mean ± s.e.m unless otherwise stated. Graph bars with different letters or * indicate the p-value < 0.05 by a Student t-test (to compare two groups) or analysis of variance (ANOVA, to compare three or more groups). Data from at least three biological replicates were presented. All experiments were repeated three times to ensure consistency.

OSUB18 was initially isolated from switchgrass plants (Panicum virgatum L.) in Ohio by the previous method (Xia et al., 2013). We investigated the potential capability of OSUB18 in plant pathogen inhibition and disease control. We noticed that one-day-old OSUB18 cultured on Tryptic Soy Agar (TSA) plates exhibited colonies with undulate margins ( Figure 1A , upper). The representative crinkle colony was confirmed under a light microscope ( Figure 1A , bottom). OSUB18 in vitro plate assay showed its ability to inhibit the growth of Pst DC3000 ( Figures 1B, C ), which is the causing agent of bacterial speck disease and the top first plant pathogenic bacterium with a hemibiotrophic lifestyle in molecular plant pathology study (Mansfield et al., 2012). Additionally, we examined the capacity of OSUB18 to control B. cinerea, a prevalent fungal pathogen with a necrotrophic lifestyle that causes the destructive grey mold disease on more than 200 plant species (Dean et al., 2012). We found that in vitro, OSUB18 significantly inhibited the growth of B. cinerea by releasing diffusible functional compounds to the agar media ( Figures 1D, E ). We also found that OSUB18 could increase Arabidopsis plants’ shoot biomass and seed yield by the root-drench treatment ( Figure S1 ). To characterize OSUB18 at the molecular level, we amplified a PCR fragment from its genome DNA with the internal transcribed sequences (ITS) primers 799F/1193R (Chen et al., 2020) and conducted the Sanger sequencing for the purified PCR fragment. Based on the 16S ribosomal RNA sequences of OSUB18 and the other various sequenced Bacillus spp. from the NCBI database, we built a phylogenetic tree for OSUB18. The phylogenetic analysis indicated that OSUB18 belongs most closely to Bacillus proteolyticus at the species level ( Figure S2 ).

To further study the efficacy of OSUB18 in antagonizing phytopathogens in plants, we mixed OSUB18 cells or sterile water (control) into the pathogen inoculum and conducted the pathogen infection assay in plants, which is a much more complicated and natural system than the in vitro co-culture agar plate assay. As expected, the mixing of OSUB18 in the pathogen inoculum significantly inhibited the growth of Pst DC3000 in A. thaliana Col-0 leaves ( Figures 2A–C ). In addition to the model plant A. thaliana, tobacco leaves co-syringe injected with Pst DC3000 and OSUB18 showed much weaker disease symptoms ( Figure 2D ) than the water control. In addition, similar results were obtained from 4 different cultivars of tomato plants ( Figures 2E, F ), validating the roles of OSUB18 against the bacterial phytopathogen in different plant species. Likewise, we mingled OSUB18 into the fungal pathogen inoculum of B. cinerea for its infection assay. Consistently, the presence of OSUB18 significantly reduced the virulence of B. cinerea on Arabidopsis Col-0 plants ( Figures 3A, B ), tobacco plants ( Figures 3C, D ), and tomato plants ( Figures 3E, F ). Our results indicate that OSUB18 has great potential in controlling bacterial and fungal diseases with a broad spectrum in different plant species.

To further investigate the mechanisms of OSUB18 in activating the ISR of host plants against pathogens, we root-drenched A. thaliana Col-0 plants with OSUB18 and then challenged their above-ground leaves with the pathogenic bacterium Pst DC3000. Both syringe injection and dipping inoculation assays exhibited significantly higher ISR response in Arabidopsis against Pst DC3000 infection ( Figure 4A ). The leaves were collected 24 hours after the Pst DC3000 infection to proceed with assays assessing the callose deposition and ISR-responsive gene expression. Compared with control plants (Ctrl, pre-drenched with water), Col-0 plants pre-drenched with OSUB18 accumulated much more callose ( Figures 4B, C ) and significantly higher expression levels of the ISR-responsive gene MYC2 ( Figure 4D ) upon the Pst DC3000 inoculation. Unlike the hemibiotrophic phytopathogens such as Pst DC3000, B. cinerea has a necrotrophic lifestyle. As a result that OSUB18 antagonized both Pst DC3000 and B. cinerea ( Figures 1 – 3 ), we were interested in the ISR efficacy of OSUB18 against phytopathogens with different lifestyles. We thus infected the pre-drenched OSUB18 and water control A. thaliana Col-0 plants with B. cinerea. We found that Col-0 plants pre-drenched with OSUB18 showed significantly higher resistance against B. cinerea infection ( Figures 5A, B ). The Col-0 plants pre-drenched with OSUB18 also exhibited a more robust callose deposition ( Figures 5C, D ) and significantly higher expression of the ISR-responsive gene MYC2 ( Figure 5E ) against B. cinerea. Consistent with previous reports that MYC2 was a transcription factor constitutively expressed in rhizobacteria-induced systemic resistance (Pozo et al., 2008), our data showed that OSUB18-drenched plants had a remarkably higher expression level of MYC2 compared to the control plants ( Figures 5E , 6E ). In agreement with the above results, Arabidopsis myc2 mutant plants were defective in OSUB18-triggered ISR response against both the bacterial pathogen Pst DC3000 ( Figures S5 ) and the fungal pathogen B. cinerea ( Figure S6 ). The result suggested that OSUB18 could successfully activate ISR induction against phytopathogens with different lifestyles.

Upon the Pst DC3000 infection, Arabidopsis leaves were collected to proceed with assays assessing the superoxide anion production, hydrogen peroxide production, and ROS-responsive gene expression. Compared with the control plants (Ctrl, pre-drenched with water), Col-0 plants pre-drenched with OSUB18 accumulated much more superoxide anion ( Figure 6A ), hydrogen peroxide ( Figure 6B ), and significantly higher expression of the ROS-responsive gene RBOHD ( Figure 6C ), which is essential for the accumulation of ROS in the apoplast (Torres et al., 2002; Zhao et al., 2022a). To verify whether a similar mechanism pattern is present in OSUB18-induced ISR against fungal pathogens, we also infected the pre-drenched OSUB18 A. thaliana Col-0 plants with B. cinerea. After B. cinerea infection, Arabidopsis leaves were collected to proceed with assays to assess the superoxide anion production, hydrogen peroxide production, and ROS-responsive gene expression. Indeed, the above-ground leaves of Col-0 plants pre-drenched with OSUB18 also exhibited more robust superoxide anion ( Figure 7A ), hydrogen peroxide ( Figure 7B ), and significantly higher expression level of the ROS-responsive gene RBOHD ( Figure 7C ) against the B. cinerea infection. To further validate the positive role of ROS production in OSUB18-activated ISR against pathogens in A. thaliana, we performed the PAMP-triggered ROS burst assay with leaf discs collected from OSUB18-drenched and water-drenched Col-0 plants. As expected, A. thaliana Col-0 plants pre-drenched with OSUB18 showed a significantly stronger ROS burst elicited by the bacterial PAMP flg22 ( Figures 6D, E ) and the fungal PAMP chitin ( Figures 7D, E ) than control plants.

To further decipher the underlying mechanism of OSUB18-triggered ISR against the hemibiotrophic phytopathogen Pst DC3000, we quantified the levels of SA and SAG in the water- and OSUB18-drenched A. thaliana Col-0 plants at 0 hpi and 24 hpi of the Pst DC3000 inoculation. We found that the free SA and conjugated SA (SAG) levels were significantly increased in OSUB18-drenched plants 24 hours after the Pst DC3000 inoculation compared with the control plants ( Figures 8A, B ). In line with the SA phytohormone production, the representative genes in the SA-signaling pathway (Wang et al., 2019) (PR1, PR2, PR5, EDS5, NPR1 and SID2) were significantly upregulated in OSUB18-drenched plants at 24 hours after the Pst DC3000 inoculation, compared with the control plants ( Figures 8 , S3 ). There were no significant differences in the JA or JA-lle levels in OSUB18-drenched plants at 24 hours after the Pst DC3000 inoculation, compared with the control plants ( Figures 8E, F ), as supported by the normal gene expression levels of LOX3 and JAR1 ( Figures S3 ). Interestly, the plant JA responsive gene PDF 1.2 was significantly upregulated after the Pst DC3000 inoculation. One of the reasons could be that the PDF1.2 gene is also regulated by the other upstream genes, such as NPR1 (Nie et al., 2017). Indeed, we found that the NPR1 gene was significantly upregulated in the process of OSUB18-induced priming against Pst DC3000 ( Figure S3D ), which may lead to the increased expression of PDF 1.2.In agreement with the above results, Arabidopsis sid2 and npr1 mutant plants were defective in OSUB18-triggered ISR response against the bacterial pathogen Pst DC3000 but not jar1 mutant plants ( Figures S5 ).

To further examine the underlying mechanism of OSUB18-triggered ISR against the necrotrophic phytopathogen, we quantified the levels of JA and JA-lle in the water- and OSUB18-drenched A. thaliana Col-0 plants at 0hpi and 24hpi of the B. cinerea inoculation. Higher JA levels ( Figures 9E ) and significantly higher levels of JA-lle (the bioactive form of jasmonate) were detected in OSUB18-drenched plants at 24 hours after the B. cinerea inoculation, compared with the control plants ( Figures 9F ). Consistent with the phytohormone quantification results, the representative genes in the JA-signaling pathway (PDF1.2, COI1, LOX3, and JAR1) and the main ISR regulator NPR1 were significantly upregulated in OSUB18-drenched plants at 24 hours after the B. cinerea inoculation, compared with the control plants ( Figures 9 , S4 ). In agreement with the above results, Arabidopsis jar1 and npr1 mutant plants were defective in OSUB18-triggered ISR response against the fungal pathogen B. cinerea but not sid2 mutant plants ( Figures S6 ). Though the representative genes in the SA-pathway (PR1, PR2, PR5, SID2) were also upregulated ( Figures 9C, D , S3 ), there were no significant differences in the SA or SAG levels in OSUB18-drenched plants 24 hours after the B. cinerea inoculation, compared with the control plants ( Figures 9A, B ). Possibly, it is due to the downregulated expression of EDS5 ( Figure S4 ), which is responsible for the transport of the SA precursor isochorismate (IC) from chloroplast to cytosol for the SA biosynthesis in cytosol.

Some chemical compounds, such as siderophores, polysaccharides, and volatile acetoin/diacetyl, are critical in ISR elicitation (Berendsen et al., 2015; Jiang et al., 2016). For our study, we found that OSUB18 was positive in producing siderophores ( Figure 10A ), exopolysaccharides ( Figure 10B ), and acetoin/diacetyl ( Figure 10C ). These data aligned with OSUB18’s efficacy in activating ISR against both bacterial and fungal phytopathogens ( Figures 4 , 5 ). OSUB18 was positive in producing acetoin/diacetyl ( Figure 10C ) but not HCN ( Figure 10D ), suggesting acetoin/diacetyl-related compounds can be the functional VOCs, in contrast to Pseudomonas fluorescens strain Pf5, one well-known bacterial strain involved in ISR. The production of ammonia by OSUB18 ( Figure 8E ) may also provide host plants with an additional nitrogen source for growth and defense. The phytohormone auxin could promote plant shoots’ and lateral roots’ growth (Woodward & Bartel, 2005). The enhanced shoot growth of OSUB18-inoculated plants ( Figures S1C, D ) is consistent with the fact that OSUB18 could produce ample amounts of IAA ( Figure 10F ). However, OSUB18 did not solubilize phosphate ( Figure 10G ) nor produce organic acid ( Figure 10H ), suggesting its plant growth-promoting role might be independent of these two factors. The catalase activity of OSUB18 ( Figure 10I ) may also play a positive role in protecting host plants by increasing oxygen availability in potting mix/soil and facilitating the gas exchange of plants with the environment. In agreement with the above beneficial traits, OSUB18 metabolites (cell-free crude extracts) inhibited the growth of Pst DC3000 ( Figure S7A ) and prevented the spore germination and hypha development of B. cinerea ( Figure S7B ).