The Rice Endophyte-Derived α-Mannosidase ShAM1 Degrades Host Cell Walls To Activate DAMP-Triggered Immunity against Disease

ABSTRACT Endophytes play an important role in shaping plant growth and immunity. However, the mechanisms for endophyte-induced disease resistance in host plants remain unclear. Here, we screened and isolated the immunity inducer ShAM1 from the endophyte Streptomyces hygroscopicus OsiSh-2, which strongly antagonizes the pathogen Magnaporthe oryzae. Recombinant ShAM1 can trigger rice immune responses and induce hypersensitive responses in various plant species. After infection with M. oryzae, blast resistance was dramatically improved in ShAM1-inoculated rice. In addition, the enhanced disease resistance by ShAM1 was found to occur through a priming strategy and was mainly regulated through the jasmonic acid-ethylene (JA/ET)-dependent signaling pathway. ShAM1 was identified as a novel α-mannosidase, and its induction of immunity is dependent on its enzyme activity. When we incubated ShAM1 with isolated rice cell walls, the release of oligosaccharides was observed. Notably, extracts from the ShAM1-digested cell wall can enhance the disease resistance of the host rice. These results indicated that ShAM1 triggered immune defense against pathogens by damage-associated molecular pattern (DAMP)-related mechanisms. Our work provides a representative example of endophyte-mediated modulation of disease resistance in host plants. The effects of ShAM1 indicate the promise of using active components from endophytes as plant defense elicitors for the management of plant disease. IMPORTANCE The specific biological niche inside host plants allows endophytes to regulate plant disease resistance effectively. However, there have been few reports on the role of active metabolites from endophytes in inducing host disease resistance. In this study, we demonstrated that an identified α-mannosidase protein, ShAM1, secreted by the endophyte S. hygroscopicus OsiSh-2 could activate typical plant immunity responses and induce a timely and cost-efficient priming defense against the pathogen M. oryzae in rice. Importantly, we revealed that ShAM1 enhanced plant disease resistance through its hydrolytic enzyme (HE) activity to digest the rice cell wall and release damage-associated molecular patterns. Taken together, these findings provide an example of the interaction mode of endophyte-plant symbionts and suggest that HEs derived from endophytes can be used as environmentally friendly and safe prevention agent for plant disease control.

In our previous study, we found that the rice endophyte Streptomyces hygroscopicus OsiSh-2 and its cell-free culture filtrate (CFC) had strong antagonistic activity against M. oryzae via the action of nigericin and siderophores (3,29). OsiSh-2 could regulate thiamine biosynthesis and thus assist in blast resistance in the OsiSh-2-rice symbiont (30). Specifically, OsiSh-2 also has ISR activity in rice for resisting M. oryzae, including induction of a priming response (31). Priming is an important strategy that allows hosts to defend against pathogen attacks in a timely and cost-efficient manner because plants in the "priming state" exhibit faster and stronger activation of specific defense responses after pathogen infection than plants that sustain a full-scale defense response (32,33). As a consequence, the saved energy can drive the growth of the host. By these mechanisms, OsiSh-2 can improve host resistance to rice blast while still sustaining a high yield. However, which active compound produced by OsiSh-2 is involved in ISR and the corresponding molecular mechanism are still not entirely clear.
In this study, we screened, isolated, and obtained the ShAM1 protein from the CFC of OsiSh-2 by HR analysis. This protein was identified as a novel a-mannosidase belonging to the glycosyl hydrolase family 92 (GH92). Purified recombinant ShAM1 can activate typical plant defense responses such as H 2 O 2 accumulation, callose deposition, amplification of phosphorylated mitogen-activated protein kinase (MAPK) cascade signals, activation of genes related to the hormone signaling pathway, and HR in rice. After infection with M. oryzae, ShAM1 pretreatment could induce a priming response, trigger the expression of JA/ET biosynthesis-related genes, and increase resistance to M. oryzae in rice. As a glycosyl hydrolase, when ShAM1 was coincubated with isolated rice cell walls, the release of oligosaccharides and monosaccharides was observed. Importantly, the extract from the ShAM1-digested cell wall could activate the immune response and enhance the disease resistance of the host rice. Thus, ShAM1 degrades the cell wall to release oligosaccharides and monosaccharides, which might serve as DAMPs to activate host immunity and trigger defense against M. oryzae.
To identify a protein with immunity-inducing activity from S. hygroscopicus OsiSh-2, a systematic purification approach was adopted. In every purification step, each fraction was infiltrated into tobacco leaves to monitor HR. Initially, crude proteins from different days of CFC of OsiSh-2 were precipitated with 40, 60, and 80% saturated ammonium sulfate (SAS) (see Fig. S1 in the supplemental material). As shown in Fig. 1A, protein precipitation from the 7-day culture using 40% SAS caused the strongest HR on tobacco leaves and was thus fractionated by column chromatography on an anion-exchange DEAE Fast Flow (FF) device. Five fractions (I to V) were separately collected (Fig. 1B), and the fractions of parts I and II were selected for further purification due to the presence of HR activity (Fig. 1C). A concentrated mixture of parts I and II was passed through a Superdex 200 Increase 10/300 column and showed 5 distinct prominent peaks (Fig. 1D). Then, the protein samples in peaks 1 and 4, which presented high and no HR activity (Fig. 1E), respectively, were subjected to nano-liquid chromatography-tandem mass spectrometry (NanoLC-MS/MS). We suspected that the only expressed proteins in peak 1 might contain HR-inducing protein candidates. Based on the results of mass spectrometry and a search against the OsiSh-2-specific genome databases using the Mascot tool, 78 differentially abundant proteins between peaks 1 and 4 were found. Gene Ontology (GO) annotation grouped these proteins into two categories, biological processes and molecular functions (Fig. S2), which demonstrated that a large proportion of these proteins were related to carbohydrate metabolic progress and hydrolase activities. In addition, genome functional annotation of OsiSh-2 indicated that among these 78 proteins, 15 were hydrolases, including 8 glycoside hydrolases (see Table S1 in the supplemental material).
To determine whether these 15 hydrolases were responsible for the development of HR, we used Agrobacterium tumefaciens-mediated transformation (agroinfiltration) to transiently express the 15 corresponding coding genes in tobacco leaves (Fig. S3).  As shown in Fig. 1F, transient expression of the gene Sh58433664 (designated ShAM1) triggered HR in tobacco 7 days after infiltration. We further heterologously expressed the ShAM1 protein in Escherichia coli BL21(DE3), and the purified recombinant protein induced a strong HR in rice leaves after 48 h of treatment ( Fig. 1G; Fig. S5E). In addition, gradual dilution (from 30 nM to 3 mM) of ShAM1 decreased the development of HR, and ShAM1 induced HR at concentrations as low as 150 nM (Fig. S4). To examine the host specificity of ShAM1, we injected ShAM1 (3 mM) into the leaves of various plant species. The results revealed that ShAM1 could induce HR in Arabidopsis, pepper, cucumber, tomato, and tobacco (Fig. S5). All of these data indicated that ShAM1 is an HR-inducing protein secreted by OsiSh-2. ShAM1 activates immune responses in rice. HR is usually associated with the activation of plant innate immunity. Thus, some key components of the plant immune response were determined. With regard to ROS production, rapid and slight accumulation of H 2 O 2 at 0.5 h posttreatment (hpt) in the veins of rice was observed after being spray purified with recombinant ShAM1. The H 2 O 2 content peaked at 48 hpt and then decreased to a low level. However, no H 2 O 2 accumulation was detected in vector pCold TF-treated (control) rice ( Fig. 2A and C). As a result, the downstream immune response of the H 2 O 2 messenger, the deposition of callose, which works as an effective barrier in the plant cell wall for blocking subsequent pathogen invasion, was induced (34). At 24 hpt, callose deposition around the stomata was observed in ShAM1-treated rice leaves, whereas no callose deposition was found in control rice leaves (Fig. 2B, D).
To further verify the above observations, we tested ShAM1-induced activation of phosphorylated MAPK proteins and immune-responsive gene expression. When ShAM1 was exogenously applied to rice cells, we found that the MAPK pathway was significantly activated, while the control treatment did not activate this pathway (Fig. 2E). The expression of genes associated with the SA signaling pathways (OsPBZ1 and OsPR1a), JA/ET signaling pathways (OsAOS2 and OsLOX2), and cross talk between the SA and JA/ET signaling pathways (OsWRKY70 and OsMPK6) was then examined. Similar to ShAM1-induced MAPK activation, all of the genes were significantly upregulated in the early stage of treatment. For example, the expression of OsPR1a and OsAOS2 increased as much as 45.9-or 22.9-fold, respectively, at 6 hpt (Fig. 2F). These results confirmed that ShAM1 effectively activated the rice immune response.
ShAM1 pretreatment induced rice defense priming and enhanced resistance to M. oryzae in rice. Our previous studies indicated that OsiSh-2 could induce a priming response, an adaptive strategy to improve the defensive capacity of host rice to the pathogen. As an immunity-related active protein of OsiSh-2, ShAM1 is likely to play a role in keeping the host rice in the priming state. When M. oryzae was infected, a significant ROS burst was observed in both ShAM1-pretreated and pCold TF-pretreated (control) rice, but this response in the former was faster. H 2 O 2 accumulated in the ShAM1-pretreated rice leaves at 4 h postinfection (hpi) by M. oryzae, while the H 2 O 2 in control rice just began to largely accumulate at 24 hpi and the level further increased thereafter ( Fig. 3A and C). Similarly, callose deposition in ShAM1-pretreated rice occurred at least 4 h earlier than that in control rice, and the amount was also dramatically increased in ShAM1pretreated rice ( Fig. 3B and D). These results, especially the callose data, presented a typical as a mock control. Images were taken 48 h after injection. Scale bars, 1 cm. (B) Fractionation of a 40% ammonium sulfate precipitate by column chromatography on the DEAE FF column yielded five fractions (I to V). Blue curve, protein concentration; green curve, salt concentration. (C) The tobacco leaves were infiltrated with fractions from the DEAE FF column or buffer as a control. (The buffer was the International Streptomyces Program 2 liquid medium, and "purified" in the same way as the secreted proteins.) The experiment was repeated three times with similar results. Scale bars, 1 cm. (D) The above fractions (parts I and II were mixed) that could induce HR in tobacco were further separated using gel permeation chromatography (Superdex 200 Increase 10/300 GL), yielding five main peaks (1 to 5). Blue curve, protein concentration; red curve, salt concentration. (E) The concentrated mixture of parts I and II was passed through a gel permeation chromatography column (Superdex 200 Increase 10/300) and yielded five main peaks. Proteins corresponding to peak 1 could induce HR on tobacco leaves. Buffer was taken as a control and "purified" in the same way as the secreted proteins. Images were taken 48 h after injection. The experiment was repeated three times with similar results. Scale bars, 1 cm. (F) The leaves of 4-week-old tobacco plants were inoculated with Agrobacterium strains carrying the indicated gene in the vector pCAMBIA-1300-FLAG. The vector pCAMBIA-1300-FLAG was used as a control. The gene Sh58433664 was designated ShAM1. Images were taken 5 days after inoculation. Scale bars, 1 cm. (G) The leaves of 2-week-old rice seedlings were treated with recombinant ShAM1 (3 mM). The vector pCold TF was used as a control. Images were taken 48 h after inoculation. Scale bars, 1 cm.

DAMP-Triggered Immunity by a-Mannosidase
Microbiology Spectrum defense priming response pattern in ShAM1-pretreated rice, indicating that ShAM1 could induce an immune response in rice by priming. We then evaluated the resistance effect of ShAM1 against rice blasts. As shown in Fig. 4A and B, the blast lesions of ShAM1-pretreated rice were significantly smaller than those of control rice, with a 45.9% reduction. Correspondingly, the relative fungal biomass, as indicated by the ratio of the MoPot2 gene (an inverted repeat transposon of M. oryzae) to the OsUbq gene (a rice genomic ubiquitin gene) in ShAM1-pretreated rice, showed a 67.0% reduction compared with that in control rice (Fig. 4C). Interestingly, among the defense-related genes, the marker genes in the SA and JA/ET pathways showed different profiles between ShAM1-pretreated and control rice after M. oryzae infection. The pathogenesis-related genes OsPR1a and OsPBZ1 of the SA signaling pathway were strongly activated by M. oryzae in control rice, but ShAM1 pretreatment led to decreased expression of these two genes ( Fig. 4D and E). In contrast, the expression of the JA/ET biosynthesis-related genes OsAOS2 and OsLOX2 was strongly triggered in ShAM1-pretreated rice ( Fig. 4F and G), whereas they were not induced by M. oryzae infection in control rice. For OsWRKY70 and OsMPK6, there were no obvious differences between these two groups of rice ( Fig. 4H and I). These results indicated that ShAM1   The data shown indicate the means 6 SDs (n = 3; n refers to biological replicates). Statistical significance (****, P , 0.0001) was revealed by Student's t test. The percentage changes followed by "1" or "2" above the bars were calculated by using the above formula (D to I) The expression levels of JA/ET pathway marker genes were significantly upregulated in ShAM1-pretreated rice after M. oryzae infection. ShAM1 was inoculated with 2-week-old rice seedlings and then infected with M. oryzae after recovery for 7 days, and samples were collected for qRT-PCR at 5 days postinfection by M. oryzae in rice leaves. The reference gene OsUbq (a rice genomic ubiquitin gene) was used for the normalization of all qRT-PCR data. pCold TF-pretreated rice was used as a control. The data shown indicate the means 6 SDs (n = 3; n refers to biological replicates). might regulate the immune responses in rice to enhance its resistance to M. oryzae mainly through the JA/ET signaling pathway. ShAM1 is a novel a-mannosidase, and its enzymatic activity is related to immune induction. ShAM1 was predicted to be a glycosyl hydrolase with a molecular weight of 113.7 kDa (Table S1). The protein sequences of ShAM1 contained a GH92 domain predicted by the SMART database (Fig. 5B). ShAM1 is categorized as a secreted protein that contains a signal peptide but lacks transmembrane domains. Comparison of the sequence of ShAM1 with those of the reported a-mannosidases showed that ShAM1 shares 31% identity with the a-mannosidase Bt3990 from Bacteroides thetaiotaomicron (35) (Fig. 5A). The predicted three-dimensional (3D) protein structure from AlphaFold also indicated that ShAM1 was highly similar to Bt3990 (Fig. 5C).
In natural environments, a-mannosidase from microbes is often employed in the hydrolyzation of a-mannosidic linkages from both high-mannose-type and plant complex-type N-glycans (36). To identify the enzyme activity of recombinant ShAM1, we tested the activities using different special substrates for different enzymes. That is, 4nitrophenyl-a-D-mannopyranoside (pNP-a-D-man) was used for a-mannosidase activity, 4-nitrophenyl-b-D-mannopyranoside (pNP-b-D-man) and locust bean gum (LBG) were used for b-mannosidase activity, 4-nitrophenyl-a-D-glucopyranoside and 4-nitrophenyl-b-D-glucopyranoside were used for glucosidase activity, microcrystalline cellulose was used for cellulase activity and xylan was used for xylanase activity. The results showed that ShAM1 only exhibits the expected specificity for pNP-a-D-man (Table S2), indicating that ShAM1 is a novel a-mannosidase rather than b-mannosidase, which is usually found in bacteria, plants, and fungi. To further verify the a-mannosidase activity of ShAM1, swainsonine (SW), an alkaloid that specifically inhibits a-mannosidase, was added (37). As shown in Fig. 5D, the activity of ShAM1 incubated with SW was decreased by 89.9%.
The optimal pH for recombinant ShAM1 activity was 5.5 (Fig. S6A), and the stable pH ranged from 5.5 to 6.5 (Fig. S6B). Moreover, the optimal temperature was approximately ;40°C (Fig. S6C), and 80% relative activity was retained at 45°C (Fig. S6D). Various metal ions showed different effects on the activity of the a-mannosidase ShAM1. As shown in Fig. S6E, only Ca 21 significantly promoted the activity of ShAM1 with a 173.63% increase, while other metal ions, such as Zn 21 , Cu 21 , and Pb 21 , strongly inhibited ShAM1 activity.
To determine whether the immune response triggered by ShAM1 is related to its hydrolase activity, high-temperature-treated (preincubated at 100°C for 20 min) or SW inhibitor-pretreated ShAM1 was infiltrated into tobacco leaves to monitor HR activity. As shown in Fig. 7 below, inactivation treatment of ShAM1 could not induce HR in tobacco, indicating that the enzymatic activity of ShAM1 is required for its immune induction.
ShAM1 hydrolyzes the rice cell wall to produce DAMPs to activate the immune response in rice and enhance rice blast resistance. Generally, protein-induced rice immunity may be related to a PAMP that is directly recognized by a receptor or functions as an enzyme to release cell wall fragments as DAMPs. As ShAM1-induced HR is enzyme dependent, the scenario of rice cell-wall-derived DAMPs seems more likely. Thus, we tested whether ShAM1-digested cell wall (SDCW) extracts could activate the plant immune response. pCold TF-digested cell wall extracts were used as controls. As shown in Fig. 6, strong HR was induced in SDCW extract-treated rice ( Fig. 6A and B), and a large amount of H 2 O 2 and callose accumulation was observed in SDCW extracttreated rice at 48 hpt (Fig. 6C to F), while the control rice showed no immune response. Moreover, SA and JA/ET signaling pathway-related genes were significantly upregulated compared with those in control rice in under early treatment (Fig. 6G). We further demonstrated that the ShAM1 hydrolysis of the cell wall induced strong MAPK (Continued on next page) DAMP-Triggered Immunity by a-Mannosidase Microbiology Spectrum activation in rice cells, while the control treatment did not activate this pathway (Fig. 6H). Regarding resistance to rice blast, lesion lengths were reduced by 42.1% in SDCW extract-pretreated rice compared with control rice, which showed typical blast lesions when infected with M. oryzae ( Fig. 6I and J). Accordingly, DNA-based quantitative reverse transcription-PCR (qRT-PCR) determination of M. oryzae biomass in the rice leaves revealed that the colonization amount in SDCW extract-pretreated rice was decreased by 57.3% compared with that in control rice (Fig. 6K). In addition, we also degraded tobacco cell walls with ShAM1 and then injected or sprayed the tobacco with the extracts of SDCW. As shown in Fig. S8, SDCW could induce HR in tobacco, indicating immune activation by ShAM1 in other species via a model similar to that in rice.

DISCUSSION
Currently, many reports have focused on the application of endophytes with the capability to produce valuable bioactive molecules as biocontrol agents, as the specific biological niche inside host plants makes endophytes efficient in disease resistance for the endophyte-plant symbiont (42). However, as a kind of microbe with the capability of producing a broad spectrum of secondary metabolites with diverse biological activities, actinomycetes, and their active metabolites are scarcely reported to play a role in inducing host disease resistance, let alone endophytic actinomycetes. In this study, we isolated a new participant in the plant immune response, ShAM1, from the rice endophytic actinomycete OsiSh-2. ShAM1 was identified as a novel a-mannosidase ( Fig. 5; see Table S2 in the supplemental material) and further heterologously expressed in E. coli BL21(DE3). Recombinant ShAM1 could trigger a typical HR in rice, tobacco, Arabidopsis, pepper, cucumber, and tomato, indicating its broad spectrum of immunity-inducing activities. Spraying ShAM1 led to a cascade of immune responses in rice, including the production of H 2 O 2 , deposition of callose, amplification of phosphorylated MAPK cascade signals, and expression of several immune-responsive genes (Fig. 2). Consequently, when infected by M. oryzae, ShAM1-pretreated rice showed enhanced resistance to the pathogen (Fig. 3 and 4). Interestingly, ShAM1 is a glycosyl hydrolase, and its induced immunity is enzyme activity dependent. Thus, we proposed that ShAM1 activates immunity and enhances disease resistance in rice by degrading rice cell walls to release DAMPs based on the following elaboration (Fig. 7). The filtered supernatants of ShAM1-and pCold TFhydrolyzed rice cell walls were used for the analysis. Oligosaccharides and monosaccharides were prepared by sugar oxime trimethylsilylation derivatization, and the mass spectrum was compared with the NIST2014 and NIST2017 spectrum retrieval databases to determine the structures of the compounds. Six carbohydrates were identified. (M) Trisaccharides and tetrasaccharides released by ShAM1 from the rice cell wall. The oligosaccharides of the ShAM1hydrolyzed cell walls were detected by MALDI-TOF MS. The filtered supernatants of ShAM1-hydrolyzed rice cell walls were used for the analysis, using 2,5dihydroxybenzoic acid (10 mg/mL in 2:1 acetonitrile-water containing 0.1% TFA) as the matrix. All indicated peaks are single-charged ions of oligosaccharides on the reducing end with Na 1 using 3-aminoquinoline (3-AMQ) as the ionic liquid matrix.

DAMP-Triggered Immunity by a-Mannosidase
Microbiology Spectrum As the first physical barrier to defend against microbial invasion, the plant cell wall is involved in sensing external stresses and transferring the corresponding signal to stimulate defense responses, known as PAMP-triggered immunity. The plant cell wall has a dynamic and highly regulated structure consisting mainly of carbohydrate-based polymers, including cellulose, hemicelluloses, pectin polysaccharides, and minor proportions of mannan and glucomannan (43). Many microorganisms have evolved enzymatic systems capable of degrading plant-related macromolecules (44). Thus, many glycosyl hydrolases (GHs) secreted by microbes can be directly recognized by receptors in plant cell walls. For example, the GH12 protein XEG1 from the pathogen P. sojae (21,45) and two GH12 proteins, VdEG1 and VdEG3, with cellulase activity from the pathogen V. dahliae (28) are recognized via receptor-like proteins in host plants and activate the defense response. These triggered immunizations are independent of GH enzymatic activities.
It should be noted that some GHs have been identified to induce plant immune responses via the DAMP mechanism (46), a currently reported important model of FIG 7 Model of the function of ShAM1 in host plant immunity. The a-mannosidase ShAM1 secreted by endophytic S. hygroscopicus OsiSh-2 degrades the cell wall to release oligosaccharides and monosaccharides, which might serve as DAMPs to activate the immune response. A priming state and JA/ET-dependent signaling pathway in the host rice were then activated. Consequently, ShAM1-treated rice showed enhanced disease resistance to M. oryzae.

DAMP-Triggered Immunity by a-Mannosidase
Microbiology Spectrum plant immunity (22). These GHs break down cell wall polymers, and the products act as DAMPs to regulate immune responses (47). That is, these triggered immunizations are dependent on GH enzymatic activities. For example, a GH17 enzyme from the pathogen Cladosporium fulvum was proposed to disrupt tomato cell walls and trigger host immune responses (48). The pathogen M. oryzae secretes two GH12 endoglucanases, MoCel12A and MoCel12B, to degrade rice cell walls and release trisaccharide and tetrasaccharide to activate immune responses and increase blast disease resistance in rice (16). A GH5 protein from the bacterium Bacillus pumilus GBSW19, BpMan5, has b-mannosidase activity to degrade polysaccharides to release oligosaccharides that serve as DAMPs to activate the immune response and protect rice and tobacco against the pathogens Xanthomonas oryzae and Phytophthora nicotianae (26,49). Here, we infiltrated tobacco leaves with high-temperature-inactivated ShAM1 alone or with an SW inhibitor (see Fig. S7 in the supplemental material) and found no HR induction in tobacco. suggesting that immunity activation depends on enzyme activity. In addition, Y. Gui has reported a secreted protein cutinase, VdCUT11, from the pathogen V. dahliae could activate plant immunity by secreting the protein in plant apoplast to degrade plant cell wall polymers and release DAMPs. In this study, the signal peptide of the ShAM1 protein was predicted by the Conserved Domain Database and SignalP 5.0 server, respectively (Fig. 5B). It was a secreted protein with a predicted N-terminal signal peptide (amino acids 1 to 29). Transient expression of the ShAM1 triggered HR in tobacco 5 to 7 days after treatment (Fig. 1F), while spraying it onto the tobacco leaves with the extracts of ShAM1-digested cell walls could trigger HR in 2 to 3 days (Fig. S8). Transiently expressed ShAM1 protein took a longer time in inducing HR in tobacco than ShAM1-digested cell wall products, which may be due to the process required for ShAM1 to be secreted into the apoplast, where ShAM1 degrades tobacco cell walls and release of DAMPs. However, we have not directly tested whether ShAM1 could be secreted into the apoplast to act on tobacco. It is possible that ShAM1 may have some alternative functions in plant cells leading to HR, which deserves further research. MALDI-TOF MS and GC-MS analysis revealed that ShAM1 can degrade the rice cell wall to release oligosaccharides ( Fig. 6L and M). Mannans are intimately associated with cellulose microfibrils (50). Therefore, it is reasonable that although ShAM1 is an a-mannosidase that is highly specific for a-mannose residues, it can release oligosaccharides by degrading plant polysaccharides. To date, research on a-mannosidases has mainly focused on the diseases caused by the abnormal function of a-mannosidases and the effects of a-mannosidases on fruit ripening and storage (51,52). This is the first report about a-mannosidases serving as cell-wall-degrading enzymes.
As a novel a-mannosidase belonging to the GH92 family. ShAM1 shares only 31% amino acid sequence identity with the B. thetaiotaomicron a-mannosidase Bt3990, and its 3D structure is highly similar to that of Bt3990 (35) (Fig. 5A to C). In addition, the optimal pH of 5.5 for recombinant ShAM1 activity against pNP-a-D-man was similar to the optimal pH (slightly acidic) reported for many reported a-mannosidases (53,54). However, ShAM1 has specific enzymatic properties. For example, the enzyme activity of ShAM1 is independent of the presence of Ca 21 and can only be enhanced in the presence of Ca 21 (Fig. S6E), while by far reported a-mannosidases are dependent on Ca 21 for their enzymatic activity. Future studies will focus on biocontrol agent preparation to facilitate widespread application.
Oligosaccharides are typical DAMPs that have been studied and applied as a plant defense elicitor for several years. The cellotriose derived from the endophytic fungus Piriformospora indica induces the elevation of calcium, production of ROS, changes in membrane potential, and the expression of defense-related genes in Arabidopsis (55). Pretreatment of Arabidopsis with cellobiose increases resistance to P. syringae pv. tomato DC3000 via the induction of a series of immune responses, including fast and short-lived intracellular calcium elevation and activation of kinase (MAPK) cascades, but with no ROS production or callose deposition (24). Pretreatment of Arabidopsis with pentasaccharide 3 3 -a-L-arabinofuranosyl-xylotetraose (XA3XX) from arabinoxylan can trigger immune responses, including calcium influx, ROS production, MAPK phosphorylation, and the expression of PTI-related genes, subsequently enhancing the resistance of tomato and pepper to P. syringae pv. tomato DC3000 and Sclerotinia sclerotiorum (40). A recent study showed that mannan oligosaccharides can serve as DAMPs to trigger multiple defense responses against X. oryzae and P. nicotianae (26). In this study, the ShAM1degrading extracts of rice cell walls activated plant immune responses (Fig. 6), suggesting that the oligosaccharides released by ShAM1 served as DAMPs. However, which cell wall oligosaccharides (disaccharides, trisaccharides, tetrasaccharides, or other identified oligosaccharides) act as DAMPs needs to be further verified, and these oligosaccharides need to be purified. To the best of our knowledge, this is the first report about a-mannosidases that function to activate immunity via DAMP-related mechanisms.
One of the hallmark traits of probiotic treatments is low cost. For instance, beneficial bacteria induce mild but effective immune activation in the absence of pathogens (56), which leads plants in a priming stage to accelerate defense responses in the event of a challenger attack (57). A study on the costs and benefits of priming in Arabidopsis demonstrated that the fitness costs of priming are substantially lower than those of the directly induced defense against pathogens (58). In addition, it was shown that the benefits of priming outweigh its costs when disease occurs (59). Thus, the cost-efficient priming-inducing strategy is increasingly considered for application in disease management. To date, beneficial rhizobacteria and mycorrhizal fungi or virulent/avirulent pathogens have been reported as priming elicitors (60,61). Some natural or synthetic compounds, such as certain agrochemicals, lipopolysaccharide, chitosan, and the flagellin flg22, were also revealed as priming-inducing agents (62,63). However, knowledge of endophytes as priming elicitors to stimulate defense responses is still in its infancy (64). Similarly, reports about proteins playing roles in priming induction are scarce (65). This study revealed a new resource for priming-inducing agents from endophyte-derived proteins. The accumulations of H 2 O 2 and callose in ShAM1-pretreated rice were faster and stronger than those in control rice after infection by M. oryzae, indicating that after the immune response induced by ShAM1, rice was permanently in the primed (alarm) state and efficiently protected against pathogen stress (Fig. 3).
Signaling hormones (i.e., SA, JA, and ET) are well known to play important roles in regulating plant immune responses to various microorganisms and herbivores (66). Generally, JA/ET and SA signaling are extensively reported to be associated with probiotic-plant and pathogen-plant interactions, respectively (67). In this study, the resistance to M. oryzae of ShAM1-pretreated rice mainly relied on the JA/ET pathway ( Fig. 4F and G), while the expression of SA signal-related genes remained lower than that in control rice over the entire M. oryzae infection period ( Fig. 4D and E). These results indicate that immune regulation in plants by active components derived from probiotics is also consistent with the mechanism of action probiotics. Similarly, the protein elicitor PemG1 from Magnaporthe grisea inhibited disease development in rice and was shown to be dependent on the SA signaling pathway but not the JA/ET pathway (68). M. grisea is considered a pathogenic fungus; thus, PemG1 acts as a virulence factor to induce the pathogen-related SA pathway to trigger the host immune response. However, increasing evidence demonstrates that virulence proteins related to the SA signaling pathway promote the infection of pathogens and make plants susceptible to disease (69). However, avirulence proteins involved in the JA/ET signaling pathway can interact directly or indirectly with disease resistance proteins in host plants and then induce plant defense against the pathogen (70). This result was in agreement with the observation for ShAM1. Another piece of evidence is about INF1, an elicitin from the pathogen Phytophthora infestans. INF1 is regarded as an avirulence factor because it can activate JA/ET-mediated signaling pathways and induce resistance to bacterial wilt disease in tomato (71). We subsequently verified the regulatory effect of this avirulence protein INF1 on rice blast disease and found that INF1-pretreated rice exhibited enhanced resistance to M. grisea, which was indeed dependent on the JA/ET but not the SA signaling pathway (Fig. S9). Taken together, the results showed that ShAM1 derived from endophytes can be used as a more environmentally friendly and safe prevention strategy for plant disease control. Moreover, the results also provide new evidence for the interaction mode of endophytic actinomycetes-plant symbionts.
Preparation, purification, and identification of secreted proteins from OsiSh-2. To produce large quantities of secreted proteins, OsiSh-2 was grown in ISP2 broth at 30°C and 170 rpm. Six to 8 days of cell-free culture filtrate (CFC) was collected and filtered through a 0.22-mm-pore Millex-GP syringe filter unit (Millipore). Subsequently, 40%, 60%, and 80% saturated proteins were precipitated overnight at 0°C by the addition of 22.6 g, 36.1 g, and 51.6 g (NH4) 2 SO 4 per 100 mL CFC, respectively. The precipitate was collected by centrifugation at 12,000 Â g for 30 min at 4°C and then resuspended in 20 mM Tris-HCl (pH 7.5). The solution was dialyzed (7,000-Da molecular weight cutoff) to remove ammonium sulfate and passed through a 0.22-mm filter. The resuspended proteins were loaded onto anion-exchange chromatography on a DEAE Fast Flow (FF) column (GE Healthcare) equilibrated with Tris-HCl and eluted with a NaCl gradient from 0.0 to 1.0 M in Tris-HCl and then subjected to fast protein liquid chromatography (FPLC) separation. The fractions corresponding to absorbance peaks at 280 nm were collected. The collected fractions were further desalted and concentrated using 3-kDa ultrafiltration membranes (Millipore). The selected fractions were further purified by a Superdex 200 Increase 10/300 GL column (GE Healthcare). The collected fractions were then concentrated by lyophilization. Then, select protein samples with high and no HR activity were analyzed by nano-liquid chromatography-tandem mass spectrometry (NanoLC-MS/MS) as described previously (75).
Plasmid construction and preparation. The tested genes were amplified from S. hygroscopicus OsiSh-2 DNA using Phanta Max Super-Fidelity DNA polymerase (Vazyme Biotech, Nanjing, China). All sequences of the above genes were cloned separately into the vector pCAMBIA-1300-FLAG. Then, the results were verified by DNA sequencing (Sangon Biotech, Shanghai, China). Subsequently, the recombinant plasmids were introduced into A. tumefaciens strain GV3101, and then A. tumefaciens-mediated transformation (agroinfiltration) was used to transiently express the 15 corresponding coding genes in tobacco leaves. The primers used for transient expression are listed in Table S3.
Expression and purification of recombinant ShAM1 protein. The purified fragment of the ShAM1 gene was cloned into the pCold TF vector and transformed into E. coli strain BL21(DE3) for expression. E. coli was grown in LB medium to an optical density at 600 nm (OD 600 ) of 0.6 at 28°C. ShAM1 expression was induced by 0.5 mM isopropyl b-D-1-thiogalactopyranoside (IPTG) at 16°C overnight. Recombinant ShAM1 protein was purified using Ni-nitrilotriacetic acid (NTA) agarose (Thermo Scientific) according to the manufacturer's instructions. The primers used for protein expression are listed in Table S3.
HR activity assays. The purified recombinant ShAM1 (3 mM) was sprayed onto the leaf surfaces of 2week-old rice seedlings. Photographs were taken 48 h after inoculation. To infiltrate the purified recombinant ShAM1 into the rice leaves, the detached leaves of 2-week-old rice were lightly wounded with a mouse ear punch and then 10 mL of purified recombinant ShAM1 (3 mM) was infiltrated through the wounded areas. Rice leaves were placed in petri dishes with a light/dark light cycle of 16 h/8 h for 48 h. Photographs were taken 48 h after infiltration. Both treatments used pCold TF as control. As for the other plant species (Arabidopsis, pepper, tobacco, tomato, and cucumber), their leaves were infiltrated with the purified recombinant ShAM1 (3 mM) by using a syringe without a needle. pCold TF was used as control. Leaves were photographed at 3, 3, 7, 7, and 8 days postinfiltration, respectively. Three replicates were performed for each treatment.
RNA extraction and gene expression analysis. Rice leaves (0.05 g) were collected for total RNA extraction by using a plant total RNA isolation kit according to the manufacturer's instructions (Sangon Biotech, Shanghai, China). Then, reverse transcription reactions were performed using 1 mg of total RNA with HiScript II Q RT SuperMix for quantitative PCR (qPCR) (1gDNA wiper) with genomic DNA (gDNA) eraser (Vazyme Biotech, Nanjing, China). Gene expression was analyzed by qRT-PCR as described previously (31). The transcript data were normalized using b-actin mRNA expression levels as the internal reference. Independent experiments were repeated twice, and all reactions were performed in triplicate. The primers used for qRT-PCR are listed in Table S3.
H 2 O 2 accumulation and callose deposition assays. The rice leaves were stained with 1 mg/mL 3,39-diaminobenzidine (DAB) (Biotopped) solution overnight at 28°C. After incubation with DAB, leaves were fixed and cleared in absolute alcohol with frequent changes of the fresh solution, and then the red-brown precipitate formed by polymerized DAB in the presence of H 2 O 2 was examined under a stereomicroscope (MZ621MSX1; Mshot, Guangzhou, China). Triplicate samples were performed for each treatment. For the callose deposition assay, the rice leaves were fixed and cleared in absolute alcohol with frequent changes of fresh solution, and the chlorophyll was removed. The transparent leaves were washed with 70 mM sodium phosphate buffer three times and then incubated in stain solution (70 mM sodium phosphate buffer; 0.01% aniline blue) (Macklin) for 2 h in the dark. Then, the leaves were observed using an inverted microscope under UV light (340 to 380 nm) (TS2R-FL; Nikon, Tokyo, Japan). The abundance of H 2 O 2 accumulation and callose intensities are represented as the percentages of DAMP-Triggered Immunity by a-Mannosidase Microbiology Spectrum relative DAB and callose intensities, respectively, which were calculated by using Photoshop CS6 and with 1 mL of 2,5-dihydroxy benzoic acid solution (10 mg/mL in 2:1 acetonitrile-water containing 0.1% trifluoroacetic acid [TFA]). The mixture was then spotted on the target plate with 1 mL of 3-aminoquinoline (10 mg/mL) and dried for MALDI-TOF MS analysis. Statistical analysis. Statistical significance was analyzed using appropriate statistical tests, including by one-way repeated-measures analysis of variance (ANOVA) using SPSS 17.0 software (Chicago, IL). The significant differences between the treatments were determined according to Tukey's post hoc test or two-tailed Student's t test for evaluating the difference between two groups. Duncan's multiple-range tests were used to compare the means for multiple groups. A P value of ,0.05 was considered statistically significant.
Data availability. All data supporting the findings of this study are available within the article and the supplemental material files or are available from the corresponding author upon reasonable request.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 0.9 MB.