Bioconversion of poly-γ-glutamic acid (γ-PGA) from fulvic acid powder produced from the wastewater of yeast molasses fermentation

Background: Molasses is a wildly used feedstock for fermentation, but it poses a severe wastewater-disposal problem worldwide. Recently, the wastewater produced by yeast during molasses fermentation is being processed into fulvic acid (FA) powder as a fertilizer for crops, but it consequently induces a problem of soil acidication after being directly applied in soil. In this study, the low-cost FA powder was bioconverted into a value-added product, γ-PGA, by a glutamate independent producer, Bacillus velezensis GJ11. Results: With FA powder, the substrates of sodium glutamate and citrate sodium used in medium were decreased around one third. Moreover, FA powder could completely substitute Mg 2+ , Mn 2+ , Ca 2+ and Fe 3+ in the fermentation medium. In the optimized FA powder fermentation medium, the γ-PGA was produced with its maximum concentration at 42.55 g/L and a productivity of 1.15 g/(L·h), while only 2.87 g/L was produced in the medium without FA powder. Hydrolyzed γ-PGA could trigger induced systemic resistance (ISR), e.g. H 2 O 2 accumulation and callose deposition, against the pathogen infection in plants. Further investigations found that the ISR triggered by γ-PGA hydrolyzates was dependent on the ethylene signalling and NPR1. Conclusions: To our knowledge, this is the rst report of using the industry waste, FA powder, as a sustainable substrate for the microbial synthesis of γ-PGA. This bioprocess can not only develop a new way of FA powder as a cheap feedstock for producing γ-PGA, but also help to reduce pollution from the wastewater of yeast molasses fermentation.

Without sodium glutamate added in the medium, GJ11 produced γ-PGA at about 25.0 g/L. With the increase of sodium glutamate added in the medium, the γ-PGA production was also increased in the broth. However, the γ-PGA production was not increased continuously when the sodium glutamate concentration was more than 40.0 g/L (Fig. 1a). This was consistent with the result of SDS-PAGE (Fig. 1b). Thereby, GJ11 is a glutamate-independent γ-PGA producer, and an addition of glutamate can further improve its γ-PGA production. We further used gel permeation chromatography to determine the molecular weight of γ-PGA produced by GJ11. The retention time of standard γ-PGA was about 8.39 min, while the retention time of γ-PGA produced by GJ11 was about 6.92 min (Fig. 1c). Thereby, the molecular weight of γ-PGA produced by GJ11 is higher than that of the γ-PGA standard with a molecular weight of 580 kD.
Compatibility of GJ11 to FA powder FA powder was dissolved in water as the medium for culturing GJ11. The results showed that the biomass of GJ11 was gradually increased with the increase of FA powder (< 40 g/L), indicating that FA powder could provide nutrients to support the bacterial growth. After FA powder was increased to more than 40 g/L, the biomass of GJ11 was decreased (Fig. S1a). We further found that, with an increase of FA powder used for culturing GJ11, the pH value of the medium was decreased due to the presence of organic acids (Fig. S1b). The low pH value suppressed the bacterial growth. Thereby, we adjusted the pH value of the medium to 7.0, and fount it was later decreased to ~ 6.5 after sterilization (Fig. S1c).
The original fermentation medium with FA powder was adjust to pH 7.0 for culturing GJ11. As shown in Fig. S1d, the biomass of the culture reached its maximum when FA powder was used at 40 g/L. When the concentration of FA powder was more than 40 g/L, the biomass decreased dramatically when FA powder added in the fermentation medium increased. The γ-PGA production reached the highest value, ~ 42 g/L, when no FA powder was added into the fermentation medium. Then, the production decreased as FA powder added in the fermentation medium increased. Thereby, the excessive FA powder is unfavorable for the bacterial growth and γ-PGA production.
Effects of carbon and nitrogen sources on γ-PGA production We found that FA powder was unfavorable for GJ11 to produce γ-PGA in the original fermentation medium. Thereby, we further detected whether FA powder (40 g/L) could substitute some nutrients in order to reduce the cost of γ-PGA production. Glucose is known as the most e cient carbon source for producing γ-PGA currently. Thus, we studied whether FA powder could substitute glucose in the fermentation medium. As shown in Fig. 2a, the biomass and γ-PGA production in the broth were both decreased without an addition of glucose. Under the glucose concentration of 70 g/L, the biomass and γ-PGA production were both increased when the concentration of glucose added in the medium increased. When the glucose concentration was more than 70 g/L, the biomass and γ-PGA production was no longer increased with the increase of glucose. Besides glucose, citrate acid is regarded as a common carbon source for γ-PGA production. As shown in Fig. 2b, the increase of citrate sodium could improve γ-PGA production rather than biomass. This was probably due to the fact that citrate acid was mainly contributed to biosynthesizing glutamate, a monomer for biosynthesizing γ-PGA. When the concentration of citrate sodium was more than 10 g/L, the production of γ-PGA was no longer increased with the increase of citrate sodium added in the medium.
Although GJ11 is a glutamate-independent γ-PGA producer, addition with glutamate could improve its γ-PGA production. As shown in Fig. 2c, the biomass was not signi cantly in uenced by addition with sodium glutamate, but the γ-PGA production was increased with the increase of sodium glutamate added in the medium. The highest γ-PGA production was obtained when sodium glutamate was added at a nal concentration of 10 g/L, and higher concentrations of sodium glutamate could not improve but reduce γ-PGA production in the broth.
As shown in Fig. 2d, an addition of NaNO 3 could improve both biomass and γ-PGA production in the broth. When the concentration of NaNO 3 was more than 2 g/L, the biomass was no longer increased with the increase of NaNO 3 , but the γ-PGA production was still increased with the increase of NaNO 3 (< 12 g/L). However, another inorganic nitrogen source, NH 4 Cl, could neither improve γ-PGA production nor increase biomass in the broth (Fig. 2e). Moreover, the bacterial growth was inhibited by NH 4 Cl when it was added into the medium at a nal concentration more than 2 g/L. Additionally, the biosynthesis of γ-PGA was reduced when NH 4 Cl was added into the medium at a nal concentration more than 4 g/L.
Effects of inorganic salts on γ-PGA production Inorganic salts have been reported to be important for γ-PGA production [26,27]. As shown in Fig. 3a, KH 2 PO 4 could improve both biomass and γ-PGA production. However, when the concentration of KH 2 PO 4 was more than 0.3 g/L, the biomass was no longer increased with the increase of KH 2 PO 4 added in the medium. Moreover, the biomass was reduced when the concentration of KH 2 PO 4 was more than 0.7 g/L.
It has been reported that Mn 2+ can improve cell growth, prolong cell viability, and assist the utilization of different carbon sources and increase γ-PGA production [26]. As shown in Fig. 3b, an addition of MnSO 4 could not signi cantly improve γ-PGA production, indicating that Mn 2+ in the FA powder was enough for GJ11 to produce γ-PGA. Moreover, with the increase of MnSO 4 added in the medium, γ-PGA production in the broth was gradually decreased. Mg 2+ has been reported to be necessary for the activity of PgsBCA in biosynthesizing γ-PGA [27]. Our results showed that an addition of MgSO 4 could not improve biomass and γ-PGA production. When the concentration of MgSO 4 increased, γ-PGA production in the broth was gradually reduced (Fig. 3c).
Thereby, an addition of Mg 2+ is not necessary for γ-PGA production in the medium with FA powder.
We further investigated whether an addition of Ca 2+ could improve γ-PGA production. We found that an addition of CaCl 2 could not signi cantly improve biomass and γ-PGA production in the medium with FA powder. Moreover, the excessive Ca 2+ inhibited γ-PGA production (Fig. 3d). The results indicated that Ca 2+ was already enough for producing γ-PGA in the medium with FA powder.
As shown in Fig. 3E, an addition of FeCl 3 in the medium with FA powder was favorable for improving biomass rather than γ-PGA production. However, excessive Fe 3+ (> 0.02 g/L) could not further improve biomass in the broth.
Orthogonal test for optimizing fermentation medium On the basis of our results, we further optimized the fermentation medium with orthogonal test. Glucose, citrate sodium, sodium glutamate, NaNO 3 , KH 2 PO 4 , and FA powder were selected for further optimization according to the orthogonal experiments (L18(3 7 )). As shown in Table 2, glucose, NaNO 3 , sodium glutamate, and citrate sodium improved γ-PGA production, while FA powder was negative for γ-PGA production. According to the R-value obtained from the orthogonal tests, we found that γ-PGA production was successively affected by sodium glutamate, NaNO 3 , citrate sodium, FA powder, glucose, and KH 2 PO 4 . The optimal combination of the medium was A3B3C2D1E3F3, corresponding to 80 g/L glucose, 20 g/L NaNO 3 , 0.7 g/L KH 2 PO 4 , 20 g/L FA powder, 20 g/L sodium glutamate, and 20 g/L citrate sodium, which resulted in the highest γ-PGA production of 35.54 g/L. On the other hand, glucose and FA powder improved the bacterial biomass, while NaNO 3 , sodium glutamate, citrate sodium, and KH 2 PO 4 were unfavorable for the bacterial growth. On the basis of R-value, we found that the biomass was successively affected by glucose, NaNO 3 , citrate sodium, FA powder, sodium glutamate, and KH 2 PO 4 . The bacterial growth was negatively related with γ-PGA production. In this study, we mainly focused on γ-PGA production. Thus, our optimized medium contained 80 g/L glucose, 20 g/L NaNO 3 , 0.5 g/L KH 2 PO 4 , 20 g/L FA powder, 20 g/L sodium glutamate, and 20 g/L citrate sodium. Compared to original fermentation medium, the cost of sodium glutamate and citrate sodium were both decreased around one third due to an addition of FA powder in the medium. Moreover, FA powder could be a substitute for NH 4 Cl, MgSO 4 , MnSO 4 , CaCl 2 , and FeCl 3 in the original fermentation medium. Optimization of fermentation conditions for γ-PGA production On the basis of optimized fermentation medium, we detected the effect of medium pH on γ-PGA production. As shown in Fig. 4a, with the increase of original pH value of medium, the γ-PGA production was gradually decreased. At pH 7.0, the biomass achieved its highest value. This result was consistent with a previous literature [21]. We also investigated the effect of liquid volume on γ-PGA production, and found the production was decreased when the liquid volume increased (Fig. 4b). The biomass reached its highest amount when the liquid volume was 50 mL which was loaded in a 250 mL ask. However, excessive liquid volume was unfavorable for the bacterial growth. As shown in Fig. 4c, the γ-PGA production and biomass were both in uenced by the inoculation amount. When the inoculation amount was more than 3%, the γ-PGA production was gradually decreased with the increase of inoculation amount. Similarly, the biomass was decreased with the increase of inoculation amount (> 5%).
Veri cation of optimized fermentation medium and conditions for γ-PGA production We cultured GJ11 in the optimized fermentation medium and conditions, and found glucose in the medium was not signi cantly consumed by GJ11 during the rst 12 h, corresponding to our result that no γ-PGA was accumulated in the broth. During 12-36 h, the biomass was dramatically increased, corresponding to a rapid decrease of residual glucose in the broth. Consistently, γ-PGA was rapidly biosynthesized in this period, with a maximum production of 41.47 g/L and a high productivity of 1.15 g/(L·h). During 36-48 h, the biomass, residual glucose, and γ-PGA production had no signi cant change. During 48-96 h, the biomass was increased again, accompanied by a decrease of residual glucose in the broth. In this period, γ-PGA production was gradually decreased (Fig. 5a), suggesting that some γ-PGA was consumed as carbon and nitrogen source for the bacterial growth. FA in the medium could not be used by GJ11, which was consistent with the previous report [18]. However, other organic acids could be gradually consumed by GJ11, accompanied by a gradually increased pH value of broth (Fig. 5b).
We further veri ed the in uence of FA powder on biomass and γ-PGA production in the optimized formula. We found that the optimized fermentation medium with 20 g/L FA powder could produce γ-PGA at 42.55 g/L, while the medium without FA powder only produced 2.87 g/L γ-PGA (Fig. 5c). These results indicated that FA powder was important for γ-PGA production in our optimized formula because it probably contained some nutrients for the bacterial cells to produce γ-PGA.
HR induced by γ-PGA and its hydrolyzates Lei et al. have reported that γ-PGA could protect plants against abiotic stress, such as high and low temperature [28]. In our study, we used γ-PGA puri ed from the fermentation broth of GJ11 to treat plants, then invested whether HR could be triggered in the leaves. The results showed that γ-PGA could not induce neither HR nor ISR to protect plants form the pathogen (Pst DC3000) infection ( Fig. 6a and b). We deduced that this might be due to the high molecular weight of γ-PGA. Thus, we digested γ-PGA into γ-PGA hydrolysates with smaller molecular weights. We found that, as hydrolysis time increased, more and more γ-PGA was digested into the hydrolysates with smaller molecular weights (Fig. 6c).
After hydrolysis, the solution was adjusted to pH 7.0, then used for injecting the tobacco leaves. As shown in Fig. 6d, the 5 h hydrolysates could trigger HR in the leaves signi cantly. Consistently, the plants with ISR triggered by irrigating roots with 5 h hydrolysates showed a signi cant resistance against the pathogen (Pst DC3000) infection (Fig. 6e). We further found that 5 g/L γ-PGA hydrolysates was more effective for inducing the resistance of plants against Pst DC3000 infection (Fig. 6f) (Fig. 7a) and callose deposition (Fig. 7b) in the leaves.
We used different defense-compromised lines of Arabidopsis, including NahG, jar1-1, ein2, and npr1, to detect the possible signals induced by γ-PGA hydrolyzates. Compared to the control (water), after inoculation with Pst DC3000 for 12 h, the lines, including Col-0, NahG, and jar1-1, were all observed to have H 2 O 2 accumulation (Fig. 7c) and callose deposition (Fig. 7d), while the lines, such as ein2 and npr1, were not. After 24 h, pre-treatment with γ-PGA hydrolyzates signi cantly enhance H 2 O 2 accumulation ( Fig. 7c) and callose deposition (Fig. 7d) in the lines Col-0, NahG, and jar1-1, rather than in the lines ein2 and npr1. These results suggested that the ISR induced by γ-PGA hydrolyzates is dependent on NPR1, and the ET signal, rather than the SA and JA signals in plants.
We further recovered the pathogen in different lines, and found that pre-treatment with γ-PGA hydrolyzates could signi cantly reduce the amount of Pst DC3000 in Col-0, NahG, and jar1-1 when compared to that in CK (Fig. 7e). However, the pathogen amounts recovered in ein2 and npr1 was similar between the group pre-treated with γ-PGA hydrolyzates and the control (CK). The results further veri ed that the ISR induced by γ-PGA hydrolyzates is dependent on the ET signaling and NPR1 in plants.

Discussion
In the industrial bioprocess, molasses is a cheap and widely used carbon source for microbial fermentation, which can be used to produce various kinds of highly valuable metabolites [29]. However, the wastewater produced during molasses fermentation can lead to an environmental pollution if it is directly discharged into the environment without appropriate pre-treatment. Since molasses contains a large amount of fulvic acid (FA), the wastewater from molasses fermentation can be made into powder by dry-spraying, which can be used as a fertilizer for crops [15,16]. However, fertilizing with FA powder can result in soil acidi cation. In this study, we managed to use FA powder in microbial fermentation. After fermentation, some organic acids in the FA powder were consumed as carbon sources by microorganism. Thus, microbial fermentation can reduce the negative in uence of FA powder on soil acidi cation. On the other hand, FA powder has been potentially used as a low-cost feedstock for fermentation to produce highly valuable metabolites, such as γ-PGA [22]. In agriculture, γ-PGA is regarded as a new environmental-friendly fertilizer synergist, as well as an activator to induce resistance against plant diseases [23]. However, due to low yield and high cost of production, γ-PGA is di cult to be wildly used in agriculture. Our study found that FA powder could be used as an alternative feedstock to produce γ-PGA at a low cost. After fermentation, most of organic acids were consumed to produce γ-PGA, but FA was di cult to be used by GJ11 because of its high amount of oxygen-rich and carbon-poor functional groups [18]. Thereby, FA, together with γ-PGA, remained in the broth to promote plant growth and protect crops from stresses. These results indicated that the fermentation broth containing both FA and γ-PGA could be a better and more effective fertilizer for crops.
Previously, we isolated a strain, B. velezensis GJ11, which could produce acetoin to trigger ISR in plants [24]. Here, we found it could also produce γ-PGA e ciently by using glucose and ammonium chloride as substrates in a glutamate-independent manner. However, most of these producers have low yield of γ-PGA. Thus, people began to use glutamate to enhance the yield of this polymer, resulting in a signi cant increase of the cost. In order to produce γ-PGA at a low cost, we tested whether FA powder could substitute some nutrients in the original formula. We rstly determined whether GJ11 was compatible to FA powder. Due to the low pH of FA powder medium, increasing the concentration of FA powder resulted in a gradual decrease in pH value of the medium, which was unfavorable for the bacterial growth. However, the decrease could be reversed by increasing the original pH value of the medium with NaOH. pH is an important environmental factor for γ-PGA fermentation. It has been reported that pH 6.5 can support a high γ-PGA production, whereas pH 7.0 is favorable for cell growth [26]. Thereby, in order to enhance γ-PGA production, the medium pH should be maintained at 7.0 for the rst 24 h of culturing to obtain a maximum biomass, then shifted to 6.5 to maximize the γ-PGA production.
Although FA powder could promote the growth of GJ11 at a certain concentration, it reduced γ-PGA production in the broth. Thereby, we further optimized the medium formula with FA powder to improve γ-PGA production and reduce the production cost. Generally, glucose is used as a preferred carbon source for γ-PGA production [21]. In our study, an addition of glucose could improve both biomass and γ-PGA production in the medium with FA powder, indicating that FA powder alone is not enough to provide carbon sources for producing γ-PGA. Thereby, FA powder could not substitute glucose in the original fermentation medium. Production of γ-PGA often requires the supplementation of glutamate, resulting in an increase in the overall cost of production [15]. Although GJ11 is a glutamate -independent producer, our study showed that an addition of glutamate could signi cantly increase γ-PGA production. Interestingly, FA powder could partially substitute glutamate in the original fermentation medium. Glutamate -independent producer can de novo synthesize glutamate monomer for biosynthesis of γ-PGA via the TCA cycle [30]. Thereby, citrate acid is able to enhance γ-PGA production by joining the TCA cycle directly to elevate the level of α-ketoglutarate. In this way, more glutamate are generated to produce γ-PGA [21,31]. In this study, we tested whether FA powder could substitute citric acid. According to our results, FA powder with organic acids could partially substitute citric acid in the medium for producing γ-PGA.
GJ11 can use inorganic nitrogen sources to synthesize glutamate for γ-PGA production. In this study, we found that FA powder could substitute NH 4 Cl rather than NaNO 3 , indicating that FA powder mainly contains NH 4 + rather than nitrate. Mg 2+ is necessary for γ-PGA production because the activity of PgsBCA that is responsible for biosynthesis of γ-PGA is dependent on it [27]. Mn 2+ is important for the stereochemical and enantiomeric composition of γ-PGA [26]. Ca 2+ and Fe 3+ are also needed for high production of γ-PGA. In this study, we found that FA powder could substitute all of these ions listed above. As a result, an addition of 20 g/L FA powder could decrease the cost of glutamate and citrate acid around one third, and substitute NH 4 + , Mg 2+ , Mn 2+ , and Fe 3+ when compared to the original fermentation medium. Thereby, the formula of medium with FA powder is low-cost for producing γ-PGA via fermentation.
Although B. subtilis and B. licheniformis have been reported to be promising native bacteria for commercial production of γ-PGA most frequently [21], our study proved that B. velezensis GJ11 could be potential for producing γ-PGA. With the optimized fermentation medium and conditions, the γ-PGA production reached a maximum concentration of 42.55 g/L, and a high productivity of 1.15 g/(L·h). In this formula, FA powder is essential because the medium without FA powder had a much lower production of γ-PGA which was only 2.87 g/L.
After fermentation, the pH value of the broth was signi cantly increased and FA was slightly consumed by GJ11. Thereby, the broth containing both γ-PGA and FA with a high pH value could be a better and more effective fertilizer than that containing FA alone. Although γ-PGA is generally used as a fertilizer synergist in agriculture, it has been reported to have the ability to protect crops from plant diseases (e.g. Fusarium root rot) [32]. We hypothesized that γ-PGA might act as an activator to induce resistance against the pathogen infection in plants. Plant activators that can induce defense response have attracted increasing attentions due to their potentials in controlling plant diseases whilst reducing the environmental burdens. Their action mechanisms can activate a complex signaling network, including the pathways regulated by SA, ET, JA, etc [19]. In our study, we further investigated whether γ-PGA could trigger resistance against the pathogens infection in plants. We found that γ-PGA with high molecular weight could not effectively trigger the resistance against Pst DC3000 infection, but the γ-PGA hydrolyzates were effective to induce the defense response (e.g. HR, H 2 O 2 accumulation and callose deposition) against pathogen infection. Our results also revealed that γ-PGA hydrolyzates mainly triggered the induced systemic resistance (ISR) via ET signaling and NPR1. Thereby, in addition to acting as a fertilizer synergist, γ-PGA is potential to be used as a new activator to trigger the defense response against plant diseases.

Conclusions
We used the industry waste, FA powder, as a sustainable substrate for microbial synthesis of some biotechnological products, e.g. γ-PGA. This technology can not only alleviate the soil acidi cation induced by directly returning FA powder into soil, but also develop a new application of FA powder as a cheap raw material for producing γ-PGA at a low cost. Moreover, the novel use of FA powder as a raw material is favorable for reducing the possible pollution induced by wastewater from yeast fermentation with molasses. Thereby, the bioprocess of converting FA powder to highly valuable products, such as γ-PGA, is circular economic.

Strains, mediums, and chemicals
The strain used for fermentation was B. velezensis GJ11 [24]. Assaying in uences of FA powder on γ-PGA production and biomass FA powder was added into the original fermentation medium at different concentrations (0, 10, 20, 40, 60, 100 g/L, respectively). Then, the pH value of the medium was adjusted to 7.0 using 6 M NaOH solution.
GJ11 was cultured in LB medium overnight, and transferred into 50 mL of the fermentation mediums with FA powder in a 250 mL ask at a ratio of 3%, for further culture at 37 o C and 180 rpm for 36 h. After the incubation, the samples were collected for determining γ-PGA production and biomass of the broth.

Optimization of fermentation medium
In the fermentation medium, the concentration of FA powder was set at 40 g/L, then the impacts of glucose, citrate sodium, sodium glutamate, NaNO 3 , KH 2 PO 4 , NH 4  On the basis of above optimization, the orthogonal test was set up for further optimizing the fermentation medium. Glucose, citrate sodium, sodium glutamate, NaNO 3 , KH 2 PO 4 and FA powder were selected for the orthogonal experiment design (L18(3 7 )) ( Table 1). Table 1 Factorial Level Optimization of culture conditions On the basis of the orthogonal experiments, we further studied the effect of culture conditions on γ-PGA production. To study the impact of pH on γ-PGA production, the initial pH value of medium was set at 6.0, 6.5, 7.0, 7.5 and 8.0, respectively. To study the impact of ventilation on γ-PGA production, several 250 mL asks containing 25, 50, 75, 100 and 125 mL medium, respectively, were prepared for culturing GJ11. The amounts of inoculation were set at 1%, 3%, 5%, 7% and 10% (v/v), respectively, for producing γ-PGA. All other factors were held constantly.
Detecting γ-PGA production and molecular weight, biomass, residual glucose, pH and FA of broth The biomass of GJ11 was assayed by detecting the OD 600 value of broth. γ-PGA production of the broth was determined by the cetyltrimethylammonium bromide (CTAB) turbidimetry method, and the SDS-PAGE with methylene blue staining [30,32,34]. Residual glucose in the broth was determined using SBA-40D Bio-analyzer (Shandong Academy of Sciences, China) [35]. The pH value of broth was detected using a pH meter. The content of fulvic acids was determined by the KMnO 4 oxidation method. After dissolution, fulvic acids were valued by titration with ferrous ammonium sulphate and N-phenyl anthranilic acid, which could indicate the end point [17,36]. The γ-PGA molecular weight was measured by using gel permeation chromatography with a RI-10 refractive-index detector and a SuperposeTM 6 column (Shimadzu Corp) [37].
Analysis of hypersensitive reaction (HR) induced by γ-PGA and its hydrolyzates γ-PGA was recovered from the GJ11 broth [20]. 20 µL of γ-PGA in gradient concentration (1, 5, 10 and 15 g/L, respectively) was used for injecting the tobacco leaves with an 1 mL syringe without needles.
After 24 h, the hypersensitive response (HR) was detected via trypan blue staining [24]. After irrigating tobacco roots with γ-PGA (1, 5, 10 and 15 g/L, respectively) at 5 ml per seedling in a pot (one plant per pot, 10 pots per group) for 3 days, the pathogen of Pseudomonas syringae pv. tomato DC3000 (termed as Pst DC3000, 1 × 10 8 cfu mL − 1 ) was used for infecting tobacco plants by spraying the leaves evenly. The leaves were collected three days after the infection, then sterilized and homogenized for spreading plates. After incubation, the bacterial colony was counted for calculating the Pst DC3000 content per gram of fresh leaf (cfu g − 1 ) [24].
The γ-PGA solution (5 g/L) was adjusted to pH 2.0, then incubated at 80 o C for 0, 1, 2, 3, 4, 5, 6, 7 and 8 h, respectively. After that, the pH value of solution was adjusted to 7.0 with 6 M NaOH, then the samples were collected for analysis of hydrolyzates using SDS-PAGE [32]. After hydrolysis, the γ-PGA hydrolyzates were detected for the activity to trigger the hypersensitive response, and the ISR against Pst DC3000 infection as above.
Detecting cellular defensive responses induced by γ-PGA hydrolyzates Different lines of A. thaliana seedlings (6-week old, 10 seedlings per group), including Col-0, NahG, npr1, jar1-1, and ein2, were irrigated with 5 mL of γ-PGA hydrolyzates (5 g/L of γ-PGA was hydrolyzed for 5 h), then infected with Pst DC3000 by spraying the leaves as above. After that, the leaves were collected and     Optimization of fermentation conditions for γ-PGA production. a: Effect of pH on cell biomass and γ-PGA production. b: Effect of loading liquid volume on cell growth and γ-PGA production. c: Effect of inoculation amount on cell growth and γ-PGA production.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.