PPFM bacteria were isolated from the phylloplane of crop plants listed in Table 1 through leaf imprinting technique on solid ammonium mineral salts (AMS) medium (pH 6.8) supplemented with 0.5% methanol and cyclohexamide (30 μg ml⁻¹) and incubated at 28°C for 3–5 days (Whittenbury et al., 1970). The single colonies with reddish pink pigmentation were picked and purified. The PPFM isolates were grown for 72 h on AMS medium. Gram reaction and other biochemical tests were carried out as described previously (Gerhardt et al., 1981). Plant pathogens (Fusarium oxysporum f. sp. lycopersici MTCC 4356, S. rolfsii KACC 43068, R. solani MTCC 4633, and P. ultimum KACC 40705) were obtained from Microbial Type Culture Collection, Chandigarh, India and Korean Agricultural Culture Collection, Jeonju, Republic of Korea. Fungal pathogens were grown and maintained in potato dextrose agar (PDA) before further use. Tomato cultivar CO₄, which is susceptible to both seedling wilt and root rot, was obtained from the Department of vegetables at Horticultural College and Research Institute, Tamil Nadu Agricultural University, Coimbatore, India.

All PPFM bacterial isolates were screened for their ability to inhibit the growth of fungal pathogens on PDA plates by dual culturing technique with three replications (Yoshida et al., 2001). After 120 h incubation at 30 ± 2°C, the distance between the bacterial growth and fungal mycelial growth was measured to find out the antagonistic effect of PPFM bacterial isolates on test pathogenic fungi. The antagonistic effect of volatile antifungal compounds produced by PPFM bacterial isolates against test fungal pathogens was evaluated on PDA medium with three replications after 120 h of incubation at 30 ± 2°C as described by Trivedi et al. (2008). Briefly, PPFM isolates antagonism due to volatile compounds was evaluated by preparing a bacterial lawn on agar plates. After incubation for 24 h, the lid of the plate was replaced by a plate containing an agar block of the test fungus grown on PDA. The two plates were sealed together with parafilm. Control sets were prepared in a similar manner, without PPFM in the bottom plate. Then the Petri dishes were incubated at 28°C, and the observations were recorded after 5 days. Growth inhibition of the test fungus was calculated in % using the formula: (r1 − r2/r1) ^* 100, where r1 (a control value) represents the radial growth of the fungus in control sets without, and r2 with bacteria.

A volume of 1 ml of bacterial culture grown in potato dextrose broth for 72 h (10⁹ cfu ml⁻¹) and a disc of test fungus (10 mm) from a well-grown fungal colony on PDA plates were inoculated into 50 ml potato dextrose broth in 250 ml conical flasks and incubated at 30 ± 2°C on a rotary shaker for 120 h. Potato dextrose broth inoculated only with fungi served as positive control. The broth with fungal mat in various treatments was passed through a pre-weighed Whatman No. 1 filter paper. The filter papers were dried for 24 h at 70 ± 1°C to obtain a constant weight. The percentage reduction in fungal mycelial weight was calculated as described previously by Trivedi et al. (2008).

Siderophore, hydrocyanic acid (HCN), and salicylic acid production, β-1,3-glucanase activity and chitinase activity of the isolates were measured as mentioned in Supporting Information File Materials and Method I.

Bacterial isolates were grown in AMS medium for 72 h and DNA was extracted according to Sambrook et al. (1989). The gene encoding bacterial 16S rRNA was amplified through PCR with forward primer 27f: 5′-AGAGTTTGATCCTGGCTCAG-3′ and reverse primer 1492r: 5′-GGTTACCTTGTTACGACTT-3′. Nearly, complete 16S rRNA gene sequences of PPFM bacterial isolates from the automatic sequencer were aligned using the integrated SINA alignment tool from the ARB-SILVA website (Pruesse et al., 2007) and bacterial identities deduced by using the EzTaxon-e server (http://eztaxon-e.ezbiocloud.net/) to ascertain their closest relatives (Kim et al., 2012). Bacterial gyrA gene is a potential chromosomal marker for the phylogenetic identification of Bacillus genera. PCR amplification and detection of gyr A gene was given in the Supporting Information File Materials and Method II.

Methanol utilization by the isolates was checked according to Sy et al. (2001) in M72 medium. The bacterial suspensions were diluted in M72 medium [optical density (OD) = 0.05], and added with one of the following compounds; methanol (MeOH) (10, 50, 100, or 500 mM), pyruvate (10 mM), or succinate (10 mM). Growth was monitored by measuring OD at 620 nm. Detection of mxaF gene was given in Supporting Information File Materials and Method III.

This experiment was conducted to test the production of exotoxins called hemolysin able to destroy the Red Blood Cells (RBC) and hemoglobin by the methylotrophic isolates Delftia lacustris PPO-1, Bacillus subtilis PPT-1, Bacillus cereus PPB-1 in sheep blood agar plates (catalog no. MP1301 Himedia, India). The plates were incubated for 48–72 h at 30°C and observed for presence of clear zone, greenish brown, and no zone which indicates the complete (β hemolysis), partial (α hemolysis), and no hemolytic (ɤ) activity, respectively. Susceptibility to the antibiotics rifampicin, ampicillin, oleandomycin, chloramphenicol, tetracycline, novobiocin, streptomycin, spectinomycin, kanamycin, trimethoprim, hygromycin, ceftriaxone, gentamycin, cefepime, and amikacin (each at 50 μg ml⁻¹) was tested for D. lacutris PPO-1 on R2A agar plates as per Bauer et al. (1966).

Plant growth chamber experiment was conducted to assess the biocontrol potential of PPFM bacterial isolates on wilt and rot diseases of tomato. Bacterial cultures grown for 72 h were harvested by centrifugation at 10,000 × g for 10 min at 4°C, washed twice with sterile distilled water and suspended in 0.03 M MgSO₄ solution. Tomato seeds were surface sterilized with 70% ethanol for 1 min, 0.5% NaOCl for 2 min, and washed four times with sterilized distilled water. The surface-sterilized seeds were imbibed in the 5 ml of bacterial suspension (10⁹ cfu ml⁻¹) for 2 h, and single seed was sown in a 400 ml plastic pot filled with mixture of sterilized soil and farm yard manure in 3:1 ratio. The pots were subjected to following treatments; T1—B. subtilis PPT-1 + F. oxysporum f. sp. lycopersici, T2—B. cereus PPB-1 + F. oxysporum f. sp. lycopersici, T3—D. lacustris PPO-1 + F. oxysporum f. sp. lycopersici, T4—B. subtilis PPT-1 + S. rolfsii, T5—B. cereus PPB-1 + S. rolfsii, T6—D. lacustris PPO-1 + S. rolfsii, T7—B. subtilis PPT-1 + R. solani, T8—B. cereus PPB-1 + R. solani, T9—D. lacustris PPO-1 + R. solani, T10—B. subtilis PPT-1 + P. ultimum, T11—B. cereus PPB-1 + P. ultimum, T12—D. lacustris PPO-1 + P. ultimum, T13—F. oxysporum f. sp. Lycopersici, T14—S. rolfsii, T15—R. solani, T16—P. ultimum (Figure S1). One milliliter of fungal spore suspension (10⁶ spore ml⁻¹) was applied to the soil on 30th day after sowing. Ten potted plants were maintained for each treatment. The experiment was conducted in a completely randomized block design with three replications. Pots were placed in growth chambers at 20 ± 1°C, photoperiod starting with 12 h dark followed by 12 h lights (18 μ mol m⁻² s⁻²). Plants were watered with 1 X Hoagland solution. The disease incidence was observed by recording the wilting of leaves leading to plant drying. Randomly 15 plants were selected from each treatment and the number of wilted plants was recorded. The mean wilt disease incidence was expressed in percentage. Disease incidence was recorded up to 30 d after challenge inoculation with fungal pathogens. The percent disease incidence (PDI) was calculated, by using the formula. PDI = Number of plants infected/Total number of plants observed × 100.

Tomato seeds were surface sterilized and imbibed in the PPFM bacterial suspension as described earlier. Four seeds were sown in a 20 L earthen pot containing 10 Kg of unsterilized soil. Control seeds were imbibed with 0.03 M MgSO₄ for 2 h. All treatments followed in the plant growth chamber assay were imposed. The experiment was conducted in a completely randomized block design with three replications. Twenty-one days after sowing, PPFM bacterial cells suspended in 0.03 M MgSO₄ (10⁹ cfu ml⁻¹) were sprayed over tomato plants with handheld pneumatic sprayer till the plants were completely wetted. Thirty days after sowing, 20 ml of fungal spore suspension (10⁶ spore ml⁻¹) was applied into the soil. Initially, F. oxysporum f. sp. lycopersici (1.2 × 10² cfu), S. rolfsii (1.5 × 10² cfu), R. solani (2.1 × 10² cfu), and P. ultimum (1 × 10² cfu) were present in 1 g of experimental dry soil. Germination percentage was recorded on 7 days after sowing (DAS). Seedling vigor index was calculated as % germination × seedling length (shoot length + root length) in cm on 15 DAS (Baki and Anderson, 1973). Leaf samples were collected on 0, 3, 5, 7, and 9 days after challenge inoculation of pathogen and frozen immediately at −20°C for analyses of PR proteins and defense enzymes. The plant height was recorded on 30, 60, and 90 DAS. The crop was harvested at physiological maturity and yield was recorded.

Frozen leaves were ground at 4°C in an ice-chilled mortar with liquid nitrogen, and the resulting powder was suspended in 100 mM potassium phosphate buffer, pH 7.0 (2:2.5, w/v). Crude homogenates were centrifuged at 8000 × g for 10 min at 4°C, and the supernatants were kept frozen at −20°C until use. β-1,3-glucanase was assayed based on the release of reducing sugars from laminarin, as described by Liang et al. (1995). One unit of β-1,3-glucanase activity was determined as the amount of enzyme required to liberate 1 ng of glucose equivalent at 37°C in 1 h. Chitinase activity was assayed by measuring the release of N-acetyl-glucosamine from prepared colloidal chitin, as described by Singh et al. (1999). One unit of chitinase activity was determined as the amount of enzyme required to liberate 1 n mol of N-acetyl-glucosamine equivalent at 37°C in 1 h. PAL activity was measured using the procedure described by El-shora (2002). One unit represented the conversion of 1 μ mol of L-phenylalanine to cinnamic acid per min. Oxidative enzymes like PO and PPO were estimated as described previously (Compant et al., 2005). The activity of PO and PPO were expressed as the change in unit of absorbance at 420 nm/g of fresh weight per min. Catalase activity was assayed spectrophotometrically as described by Stern (1937). The activity was measured by monitoring the degradation of H₂O₂ using a spectrophotometer at 240 nm over 1 min. expressed in mmol min g⁻¹ of leaf tissue⁻¹.

The data were analyzed by an analysis of variance (ANOVA) using the general linear model version 9.1; SAS Institute Inc., Cary, NC, USA. Means were compared using the least significant difference (LSD). The significance levels were within confidence limits of 0.05 or less.

Twenty PPFM bacteria were isolated from phylloplane of various crop plants (Table S1). Among the PPFM bacterial isolates, 12 isolates showed direct antagonism toward tested pathogens. The isolates PPO-1, PPT-1, and PPB-1 exhibited mycelial growth inhibition of Fusarium oxysporum f. sp. lycopersici (2.15, 2.05, and 1.95 cm), S. rolfsii (2.14, 2.04, and 1.94 cm), P. ultimum (2.12, 2.02, and 1.92 cm), R. solani (2.18, 2.08, and 1.98 cm), respectively, in a dual plate assay on day 5 (Table 1). All the tested methylotrophic isolates were able to produce siderophores and salicylic acid and none of the isolates produced hydrocyanic acid. The isolates PPO-1, PPT-1, and PPB-1 produced the maximum siderophore (27.94, 27.45, and 25.44 μg ml⁻¹) and salicylic acid (89, 88, and 87 μg ml⁻¹), respectively (Table 1). Chitinase and β-1,3-glucanase activities were observed higher in PPO-1 isolate followed by PPT-1 and PPB-1 isolates (Table 1).

All PPFM bacterial isolates produced volatile antifungal compound(s), and inhibited the mycelial growth of the tested fungal pathogens in sealed Petridishes. The higher mycelial growth inhibitions were observed in F. oxysporum f. sp. lycopersici (52.84, 51.73, and 50.65%), S. rolfsii (43.34, 42.23, and 41.12%), P. ultimum (57.78, 56.67, and 55.56%), and R. solani (44.45, 43.34, and 42.23%) due to the antagonistic effect of respective PPO-1, PPT-1, and PPB-1 isolates (Table 2). All PPFM bacterial isolates showed the reduction in pathogenic fungal biomass in the co-cultured liquid medium. The isolates PPO-1, PPT-1, and PPB-1 showed respective biomass reduction of F. oxysporum f. sp. lycopersici (77.14, 76.57, and 76.00%), S. rolfsii (79.26, 78.45, and 77.64%), P. ultimum (88.08, 87.65, and 85.95%), R. solani (89.81, 86.03, and 85.66%) (Figure 1).

PPFM bacterial isolates PPO-1 (JN 088183), PPT-1 (JN 088185), and PPB-1 (JN 088184) showed the highest 16S rRNA gene sequence similarities to type strains of D. lacustris DSM 21246^T (100%), B. subtilis KCTC 13429^T (99.87%), and B. cereus ATCC 14579^T (99.67%), respectively. The colony morphology of pink pigmented methylotrophic isolates PPO-1, PPT-1, and PPB-1 are shown in Figure S2. Bacterial growth was noted on either pyruvate or succinate as the sole carbon source. Also, growth was obtained at a methanol concentration up to 500 mM. Presence of methanol dehydrogenase gene (mxaF) was examined using Methylobacterium extorquens AM1 as reference strain. An amplification product of the expected size, 560 bp, for all the isolates similar to reference strain M. extorquens AM1 were noted which indicated the presence of mxaF. The isolate PPO-1 was Gram negative and isolates PPT-1 and PPB-1 were Gram positive. Isolates PPO-1, PPT-1, and PPB-1 did not exhibit hemolytic activity and isolate PPB-1 grew on mannitol egg yolk polymyxin agar MYP agar forming white precipitation which is the characteristic feature for B. cereus. Presence of housekeeping gene (gyr A) in two isolates of Bacillus sp. (PPT-1 and PPB-1) was examined, PPT-1 showed positive (Figure S3). Methylotrophic isolates PPO-1, PPT-1, and PPB-1 were susceptible to the antibiotics such as rifampicin, ampicillin, oleandomycin, chloramphenicol, tetracycline, novobiocin, streptomycin, spectinomycin, kanamycin, trimethoprim, hygromycin, ceftriaxone, gentamycin, cefepime, and amikacin at the concentration of 50 μg ml⁻¹. For hemolytic activity the PPFM isolates did not produce any clear or greenish zone around the colonies in sheep blood agar medium, which may indicates that the isolates are non-pathogenic to human.

Under plant growth chamber condition, soil inoculated with fungal pathogens showed 65% disease incidence. Higher diseases incidence was observed in the plants treated with R. solani followed by P. ultimum, F. oxysporum f. sp. Lycopersici, and S. rolfsii. However, the PPFM bacterial isolates inoculated tomato plants when challenged with the fungal pathogens recorded only 15% disease incidence (Figure 2). In greenhouse experiment, bacterization of tomato seeds with PPFM isolates D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 significantly increased the germination percentage and the seedling vigor compared to control (Table 3). However, pathogen alone treatment has negative impact on seed germination, seedling vigor and plant height. No much variation was observed in plant height among the plants treated with different PPFM isolates and challenge inoculated with pathogens. However, significant variation in plant height was observed between PPFM challenge-inoculated with pathogen treatment and pathogen alone treated plants (Figure S4). D. lacustris PPO-1 inoculated tomato plants when challenged with R. solani, S. rolfsi, and F. oxysporum f. sp. lycopersici recorded the higher fruit yield of 9.35, 9.08, and 8.90 fruits plant⁻¹, respectively (Table 3).

In this study, D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 inoculations induced significant protection in tomato plants against the tested pathogens. PPFM isolates D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 inoculated tomato plants when challenge-inoculated with F. oxysporum f. sp. lycopersici, S. rolfsii, P. ultimum, and R. solani, increased PR proteins over uninoculated controls. The higher activities of β-1,3 glucanase, chitinase, PO, PPO, PAL, and catalase enzymes were recorded on 5th day after challenge inoculation with pathogens (Table 4). Increase in PO enzyme activity was observed in tomato plants inoculated with PPFM isolates D. lacustris PPO-1, B. subtilis PPT-1, and B. cereus PPB-1 and challenged with F. oxysporum f. sp. lycopersici, S. rolfsii, P. ultimum, and R. solani. The activities of PPO, PAL, chitinase, and catalase enzymes also had a similar trend (Table 4). In particular, D. lacustris PPO-1 inoculated tomato plants when challenge inoculated with R. solani, P. ultimum, S. rolfsii, and F. oxysporum f. sp. lycopersici exhibited significant increase in β-1,3-glucanase activity (Table 4). Delftia lacustris PPO-1 challenge inoculated with pathogen exhibited higher chitinase and catalase activity compared to other treatments.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.