Cells from a single colony of P. chlororaphis O6, grown for 2 d on Luria-Broth (LB, Difco Inc., Detroit, MI, USA) agar plates, were inoculated into 500 ml of LB broth and incubated at room temperature for 4 d with agitation at 250 rpm. The EPS of P. chlororaphis O6 were isolated using a modified ethanol precipitation method as described previously (Hung et al., 2005). The bacterial cells were removed by centrifugation for 10 min, and 30 Anson unit protease (0.5 mg/l, Quiagen Inc., Valencia, CA, USA) was added to the supernatant and incubated at 37°C for 12 h with agitation at 70 rpm. Ethanol was added to the supernatant to produce a 1:4 (v/v) mixture of the supernatant:ethanol and stirred at 4°C for 14 h. The precipitate was removed by filtration through a 0.2 μm Millipore membrane. The precipitate was dissolved in sterile water and mixed with ethanol (1:4 ratio) at 4°C for 14 h. The EPS were pelleted by centrifugation for 15,000×g at 4°C for 30 min. The pellet was dried at room temperature for 5 hand the mass measured before dissolving in sterile distilled water to produce solutions with known mass/ml. The final EPS preparation was streaked on LB agar plate to confirm no contamination by any culturable microbes.

Data were analyzed by ANOVA (p<0.01), using SPSS (version 23, SPSS Inc., Chicago, IL, USA) and if the value of the F was significant, differences between each measurement were further elucidated through Duncan’s multiple range test (p<0.01).

To determine the direct effects of EPS on stomatal opening, plants were grown in well plates for 4 weeks. Arabidopsis thaliana Col-O seeds were surfaced-sterilized with 0.3% sodium hypochlorite, followed by extensive washing with sterile water. Seeds were transferred to well plates (12-well microtiter plates; SPL Inc., Seoul, South Korea) that contained half-strength Murashige and Skoog salt (MS, Sigma-Aldrich, St. Louis, MO, USA) medium, 0.3% Phytagel and 3% sucrose (pH 5.7) (Cho et al., 2008, 2012). The plates were sealed with Parafilm to prevent drying and no additional water was added during their growth. To maximize stomatal opening, the plants were maintained under light for at least 3 h before excising the epidermal peels. The peels were transferred on to glass slides. The cuticle side was treated with water (a negative control), defined dilutions of the EPS solution (10, 100, and 1000 μg/ml, the test treatments), or a suspension of P. chlororaphis O6 cells (a positive control). Another positive control for stomatal closure was treatment with 100 μM ABA (Sigma-Aldrich). Two hours after incubation, the stomatal apertures were observed by light microscopy (Leica M165 FC, Leica Microsystems, Wetzlar, Germany). The size of the openings was assessed using the image acquisition software, Image-Pro (version 2.32 for Windows, Media Cybermetics, Rockville, MD, USA). At least 100 stomata/plants were counted, with five plants per treatment being used in each experimental replicate.

The treatment of leaf epidermal peels with ABA and with the suspension of intact P. chlororaphis O6 cells decreased the size of stomatal aperture relative to stomata treated exclusively with water (Fig. 1A). These findings were expected based on the previous findings (Cho et al., 2008). Stomatal closure also was observed with treatment of 100 μg/ml EPS solution (Fig. 1A-C). The extent of closure was dose-dependent (Fig. 1B); when doses of 10, 100, and 1000 μg/ml were used, the 100 μg/ml treatment was optimal.

To determine whether the treatments that caused the narrowing of stomatal openings corresponded with drought tolerance, the degree of wilting was assessed visually using A. thaliana seedlings grown in the well plates as described previously (Cho et al., 2008). The surface-sterilized seeds were placed on Whatman # 1 paper and placed over the MS medium contained in the microtiter wells. The microtiter plates were sealed with parafilm to prevent drying. After two weeks of growth, the seedlings were treated with 10 μl/plant, water as a control, 10 μl per plant, P. chlororaphis O6 cell suspensions (1×10⁸ cfu/ml), or 10 μl per plant, 100 μg/ml EPS solution. After 3 d, seedlings on the filter papers were transferred to open, empty Petri dishes. The extent of wilting of the plants was assessed visually after 2 h. The experiment was repeated three times with 12 plants per treatment.

When A. thaliana seedlings were exposed to the air, prior treatments for 3 d with EPS, or with live P. chlororaphis O6 cells, resulted in plants that exhibited less visible wilting than the plants treated with water as a control (Fig. 2A). Thus, P. chlororaphis O6 cells and extracted EPS reduced the wilt symptoms in intact plants deprived of water.

The changes in photosynthetic potential with drought were assessed with A. thaliana seedlings grown in sterile soil mix. The solid growth matrix was a sterile nursery soil mixture of Bio-Sangto, vermiculite, perlite (125 g/vessel, 3:1:1; Seminis Korea, Seoul, Korea) contained in sterile Magenta boxes (10 cm × 10 cm × 9 cm, Sigma-Aldrich). Five milliliters of water were added daily to the boxes. The seedlings were grown under a 16 h/8 h light/dark cycle with 40-W fluorescent lights (3,000 lx, 80 μmol photon m^-2s^-1). The temperature was maintained at 23±2°C with a relative humidity of 50-60%.

The seedlings were grown for 3 weeks before treatment with 2 ml of sterile water as a negative control, or with equal volumes of 100 μM ABA as a positive control, or P. chlororaphis O6 EPS (100 μg/ml). A subset of the seedlings was drought-stressed by withholding water. After 12 d of drought stress, the chlorophyll fluorescence in the leaf tissues was measured using a Phyto-PAM (Heinz-Walz GmbH, Effeltrich, Germany) as described previously (Woo et al., 2008). The F_v/F_m was determined by the application of a single saturating pulse to dark-adapted leaves for 30 min and used as a quantitative measure of drought tolerance (Woo et al., 2008). The experiment was repeated twice with 3 plants per treatment.

In plants grown in pot cultures that were maintained with normal hydration, the F_v/F_m value for the leaves increased daily, indicating healthy photosynthesizing leaves (Fig. 2B). In plants that were subjected to drought stress by withholding water, the F_v/F_m value steadily declined with the number of days of stress due to the impaired functioning of photosystem II. However, plants that were subjected to water withholding when treated with either ABA or EPS had leaves that showed no reduction in the F_m/F_v values (Fig. 2B). This response was expected, because the partial closure of the stomata maintained function of photosystem II. Thus, treatments of the intact plants with the EPS from P. chlororaphis was as effective as ABA in protecting the plant from reduced photosynthetic efficacy under drought stress.

The present study indicated that the ethanol-precipitated EPS from P. chlororaphis O6 were effective in inducing protection against drought, as measured by limited wilting and by the maintenance of photosynthetic efficacy upon water withholding in intact Arabidopsis plants. Leaf tissues exposed to the EPS showed stomatal closure eliciting the same plant response as 2R, 3R-butanediol (Cho et al., 2008, 2013). Consequently, we demonstrated that EPS from P. chlororaphis O6 is another plant immune effector activating drought tolerance.

The EPS are part of the matrix that embeds bacterial cells in biofilms (Anderson and Kim, 2018). The composition of EPS likely includes different polysaccharides and extracellular DNA (Hung et al., 2005; Wingender et al., 1999). P. chlororaphis O6 forms patchy but layered and complex biofilms on roots as part of the surface colonization process (Bonebrake et al., 2018). Timmusk et al. (2015), based on their work with root-colonizing Gram-positive bacteria that active IST, concluded that the biofilm formation is required in induction of drought tolerance. Thus, the EPS produced in the root surface biofilms may be part of the microbial “microbe-associated molecular patterns” (MAMPs) that activate plant immunity. Indeed, EPS from Bacillus cereus acts as a MAMP for stimulating plant disease resistance (Jiang et al., 2016). We are currently characterizing the components in the EPS from P. chlororaphis O6. An interesting question is whether alginate is present. Alginate present in the EPS from P. aeruginosa, xanthan from Xanthomonas campestris, and amylovoran from Erwinia amylovora suppress plant immunity by binding calcium, thus, preventing MAMP-induced signaling in the plant host (Aslam et al., 2008).

The EPS from P. chlororaphis O6 functions similarly to the EPS from other plant-associated bacteria, e.g., Pantoea agglomerans, Burkholderia gladioli, and P. putida, in the activation of defense mechanisms (Ortmann et al., 2006; Park et al., 2008; Sandhya et al., 2009; Santaella et al., 2008). These findings suggest that it is feasible to develop formulations containing EPS from certain plant-associated bacteria as “green formulations” to boost plant health during crop production. Such formulations would overcome the problem of requiring live microbial cells to induce stress tolerance. Applications of EPS-based formulations maybe especially valuable to aid in alleviating the stress imposed on crops by water shortage caused by higher temperatures and altered precipitation due to climate change.