Involvement of Peptidoglycan Receptor Proteins in Mediating the Growth-Promoting Effects of Bacillus pumilus TUAT1 in Arabidopsis thaliana

Abstract Bacillus pumilus TUAT1 acts as plant growth–promoting rhizobacteria for various plants like rice and Arabidopsis. Under stress conditions, B. pumilus TUAT1 forms spores with a thick peptidoglycan (PGN) cell wall. Previous research showed that spores were significantly more effective than vegetative cells in enhancing plant growth. In Arabidopsis, lysin motif proteins, LYM1, LYM3 and CERK1, are required for recognizing bacterial PGNs to mediate immunity. Here, we examined the involvement of PGN receptor proteins in the plant growth promotion (PGP) effects of B. pumilus TUAT1 using Arabidopsis mutants defective in PGN receptors. Root growth of wild-type (WT), cerk1-1, lym1-1 and lym1-2 mutant plants was significantly increased by TUAT1 inoculation, but this was not the case for lym3-1 and lym3-2 mutant plants. RNA-seq analysis revealed that the expression of a number of defense-related genes was upregulated in lym3 mutant plants. These results suggested that B. pumilus TUAT1 may act to reduce the defense response, which is dependent on a functional LYM3. The expression of the defense-responsive gene, WRKY29, was significantly induced by the elicitor flg-22, in both WT and lym3 mutant plants, while this induction was significantly reduced by treatment with B. pumilus TUAT1 and PGNs in WT, but not in lym3 mutant plants. These findings suggest that the PGNs of B. pumilus TUAT1 may be recognized by the LYM3 receptor protein, suppressing the defense response, which results in plant growth promotion in a trade-off between defense and growth.


Introduction
In plants, pattern recognition receptors (PRRs) recognize elicitors, such as microbe-associated molecular patterns (MAMPs), leading to MAMP-triggered immunity.Many bacterial elicitors act as MAMPs, such as flagellin, elongation factor Tu, lipopolysaccharides and peptidoglycans (PGNs), which activate plant defense responses (Newman et al. 2013).PRRs are located on plant cell membranes and have different receptor domains, such as extracellular ligand-binding, transmembrane, and intracellular kinase domains.In Arabidopsis, lysin motif (LysM)-containing receptor proteins, LYM1 and LYM3, act as extracellular ligand-binding domains that can physically interact with PGNs in conjunction with the LysM-containing receptor kinase, CERK1, initiating an intracellular signal transduction cascade to increase plant immunity (Willmann et al. 2011).PGNs compose the bacterial cell wall and consist of a glycan backbone of alternating β-1,4-linked N-acetylglucosamine (GlcNAc) and N-acetylmuramic acid (Glauner et al. 1988).The polymeric carbohydrate backbone of PGNs, GlcNAc, is recognized by plant LysM receptors in plants.Plant LysM domain proteins have been widely implicated in the recognition of GlcNAc-containing glycans such as Nod factors (Limpens et al. 2003, Radutoiu et al. 2003) and chitin (Kaku et al. 2006, Miya et al. 2007, Wan et al. 2008, Shimizu et al. 2010).PGNs are common MAMPs that elicit defense responses through recognition by receptor proteins in Arabidopsis (Gust et al. 2007, Erbs et al. 2008, Millet et al. 2010).In contrast, another MAMP, flagellin, is recognized by FLAGELLIN-SENSING 2 which is a receptor-like kinase (RLK) protein consisting of an extracellular leucine-rich repeat, a single membrane-spanning domain and a cytoplasmic serine/threonine kinase domain.Flagellin releases a conserved 22-amino acid peptide, flg-22, and triggered immune responses such as callose deposition, reactive oxygen species (ROS) production and changes in defense gene expression in Arabidopsis.However, this reaction can be suppressed by Nod factor and chitotetraose.Mutant (lyk3) studies suggested that the AtLYK3, a LysM-containing receptor kinase, is necessary for Nod factor suppression of immunity triggered by flg-22 (Liang et al. 2013).
Bacillus pumilus TUAT1 was originally isolated from cultivated soil taken from the fields adjacent to the Tokyo University of Agriculture and Technology (Djedidi et al. 2014).Subsequently, this strain was shown to act as a plant growth-promoting rhizobacterium (PGPR), contributing to the improvement of plant growth by inoculation on various plants, such as rice, Setaria, soybean (Win et al. 2018, 2019, 2020, Agake et al. 2021, 2022a, 2022b, Hasibuan et al. 2021) and Arabidopsis (Borjigin et al. 2024).It is now commercially available as a biofertilizer 'Yume-bio' (Asahi Agria.Co., Ltd.Tokyo, Japan).Under stress conditions, B. pumilus TUAT1 forms spores, which makes this a convenient strain for long-term storage and packaging for use as a biofertilizer.Previous research compared the PGP effects of spores and vegetative cells and found that the former were significant more effective (Ngo et al. 2019).The autoclaved dead spores of Bacillus spp.accelerated the growth of rice seedlings (Seerat et al. 2019), suggesting that at least one substance responsible for the PGP effects is present in heattreated spores.The cell wall of Gram-positive bacteria, including Bacillus, consists of thick PGN layers, and the amounts of PGN are notably higher in spores.This has led to the hypothesis that PGN is one of the candidates inducing the PGP effects of B. pumilus TUAT1.
However, despite numerous reports on the effects by Bacillus PGPR strains, the mechanisms by which plant growth is promoted by these strains remain an open question.Additionally, it is important to clarify this mechanism to further improve the effectiveness of the B. pumilus TUAT1 strain as a commercial biofertilizer.The first step in this PGP process is perception of B. pumilus TUAT1 strain by plants, which initiates a signaling cascade leading to growth promotion.To gain insights into this, we investigated the involvement of receptor proteins for pathogen PGNs, i.e.LYM1, LYM3 and CERK1, in the perception of B. pumilus TUAT1 using Arabidopsis T-DNA knockout mutant plants.The plant responses to purified PGNs of B. pumilus TUAT1 were also investigated.

Growth promotion effects of B. pumilus TUAT1 spores in the lym1, lym3 and cerk1 mutants
To determine the involvement of PGN receptors in growth promotion by B. pumilus TUAT1 spores in Arabidopsis, knockout mutant plants for lym1-1, lym1-2, lym3-1, lym3-2 and cerk1-2 with T-DNAs in the homozygous state of the LYM1, LYM3 and CERK1 genes, respectively, were established.All these mutants are described in Willmann et al. (2011).First, one mutant per each gene lym1-2, lym3-2 or cerk1-2 was analyzed along with the corresponding wild-type (WT) Col-0 plants.As B. pumilus TUAT1 is known to promote root growth (Ngo et al. 2019, Agake et al. 2021, 2022a, 2022b), we focused on roots.A significant increase in root dry weight by B. pumilus TUAT1 inoculation was observed in WT, lym1-2 and cerk1-2 plants, whereas lym3-2 mutant plants showed no increase in root weight (Fig. 1A).
To confirm the different responses in lym3 mutant from those in WT plants, another allelic mutant, lym3-1, was analyzed together with lym3-2.lym1-1 was also included to confirm the result of lym1-2.Significant increases in root dry weight, lateral and total root lengths and root surface area by B. pumilus TUAT1 inoculation were observed in WT and lym1-1 plants, but similar changes were not seen with lym3-1 and lym3-2 plants (Fig. 1B-E).These results clearly implicate the LYM3 receptor protein in mediating the PGP effects of the B. pumilus TUAT1 spores.

Growth promotion effects of purified PGNs from B. pumilus TUAT1
Since LYM3, one of PGN receptors, was found to be involved in the response to the spores of B. pumilus TUAT1, we next confirmed if PGNs purified from this strain promote plant growth.WT, lym3-1 and lym3-2 plants were treated with purified PGNs from B. pumilus TUAT1 or spores of the strain (Fig. 2).Significant growth promotion was observed in WT by treatments of purified PGNs or spores, but not in either lym3-1 or lym3-2 plant.These results indicate that PGNs themselves have growth promotion activity and LYM3 is involved in the mechanism.

Transcriptome analysis by inoculation of B. pumilus TUAT1 in Arabidopsis roots
To further explore the plant response to inoculation with B. pumilus TUAT1 spores, RNA-seq analysis was used to compare uninoculated and inoculated WT, lym3-1 and lym3-2 plants.Total RNA was extracted from roots treated with B. pumilus TUAT1 spores for 1 h and subjected to transcriptome analysis.This time point, 1 hour, was chosen based on our previous reports (Xiao et al. 2023) in which of the gene expression response at 1 h after B. pumilus TUAT1 inoculation in rice seeds correlated with growth promotion.An average of 40.1 million clean reads mapped to the Arabidopsis genome at an average rate of 95.2% (Supplementary Table S1).As shown in the Venn diagram (Fig. 3) and the volcano plots (Supplementary Fig. S1A-C), the numbers of differentially expressed genes (DEGs) are much higher in lym3-1 and lym3-2, compared to those in WT plants.Principal component analysis (PCA) shows that distribution of DEGs of two allelic mutants is not close (Supplementary Fig. S1D), corresponding to the small overlap between lym3-1 and lym3-2 in the Venn diagram (Fig. 3).Additionally, the dendrogram of transcriptome data indicates clear separation among all genotypes (Supplementary Fig. S1E).Top 1,000 genes in the expression level were clustered into eight clusters in terms of their pathways using the K-means method (Supplementary Fig. S2A).Among them, clusters 2 and 7 showed similar patterns in lym3-1 and lym3-2, which are different from WT plants.Genes in cluster 2 were upregulated by inoculation of B. pumilus TUAT1 in WT, while they were downregulated in lym3-1 and lym3-2 (Supplementary Tables S4 and S6).In contrast, genes in cluster 7 were downregulated by inoculation of B. pumilus TUAT1 in WT, while they were upregulated in lym3-1 and lym3-2 (Supplementary Tables S5  and S7).
For the DEGs relevant to the LYM3 mutation, common DEGs observed in both lym3-1 and lym3-2 were selected.For the DEGs of WT, the DEGs observed in WT and only one allele of two lym3 mutants were included as they are irrelevant of LYM3 mutation.Thus, the numbers of upregulated and downregulated DEGs of WT are 15 and 16, and those observed in both alleles of lym3, which are relevant to LYM3 mutation, are 14 and 4, respectively (Fig. 3).
The locus tags and gene descriptions for the 31 DEGs of WT are presented in Table 1, and those for the 18 DEGs commonly observed in both lym3 alleles are presented in Table 2.These 18 genes are possible candidates related to the different responses to B. pumilus TUAT1 in the lym3 mutant compared to WT plants due to the lack of LYM3 protein.
For Gene Ontology (GO) enrichment analysis, the GO slims annotated at the Arabidopsis Information Resource (TAIR) (https://www.arabidopsis.org/portals/genAnnotation/functional_annotation/go.jsp), which are high-level terms from each GO hierarchy that are useful for grouping genes into broad categories (Berardini et al. 2004), were used.The ratio of annotated GO slims in categories of biological process and molecular function, were compared to those in the whole genome (Tables 3, 4, 5, 6).In the enriched GO slims for upregulated DEGs in WT plants (Table 3), 'response to stress' and 'response to chemical' include 3 and 4 DEGs, respectively.In the enriched GO slims for downregulated DEGs in WT plants (Table 4), 'response to stress' , 'response to chemical' and 'response to external stimulus' include 8, 5 and 8 DEGs, respectively.These results indicated that downregulated DEGs are more enriched in GOs related to stress or stimulus responses compared to the upregulated DEGs in WT plants.While in the enriched GO slims for upregulated DEGs in lym3 (Table 5), 'response to chemical' , 'response to stress' , 'response to external stimulus' and 'response to biotic stimulus' include 6, 8, 5 and 5 DEGs, respectively.In the enriched GO slims for downregulated DEGs in lym3 (Table 6), 'response to chemical' , 'response to stress' and 'response to abiotic stimulus' include 1, 2 and 2 DEGs, respectively.These results indicated that upregulated DEGs are more enriched in GOs related to the responses to stress or stimulus responses compared to the downregulated DEGs in lym3 plants.
Confirmation of LYM3 transcript in lym3-1 and lym3-2 mutants Willmann et al. (2011) demonstrated no LYM3 transcript in both lym3-1 and lym3-2 plants by RT-PCR using primers amplifying the entire coding region of LYM3 gene.As lym3-1 and lym3-2 did not act similarly in RNA-seq analysis, we assumed the lack of LYM3 transcripts in either lym3-1 or lym3-2 plants.We confirmed this by RT-PCR using the same primers as in Willmann et al. (2011) with different PCR cycles such as 25, 30 and 35 cycles.With 25 and 30 cycles, LYM3 transcripts were not amplified even in WT plants.With 35 cycles, WT plants showed clear bands, while lym3-1 mutant showed faint bands (Fig. 3C).There was no amplification in lym3-2 mutant.The quantification of the band intensities showed that lower amounts of LYM3 transcripts exist in lym3-1 mutant compared to WT, while no transcript of LYM3 was detected in lym3-2.As T-DNA is inserted in the first intron of lym3-1 while it is in the exon in lym3-2 (Willmann et al. 2011), it is likely that LYM3 gene was transcribed in lym3-1 at lower levels compared to WT, although this was not detected by Willmann et al. (2011) with their RT-PCR condition.Variations of RNA-seq results in lym3-1 and lym3-2 may be because of the difference in the LYM3 transcript.Even though LYM 3 transcript remained at lower amounts in lym3-1, Willmann et al. (2011) observed the same phenotype as defects in perception of pathogen PGNs in both lym3-1 and lym3-2.Therefore, we decided to select common DEGs in both alleles for genes related to perceive PGNs of B. pumilus TUAT1.

Response of flg-22 in the presence of B. pumilus TUAT1 spores or PGNs on lym3 mutant plants
Results of RNA-seq analysis suggested a possibility that B. pumilus TUAT inoculation alleviated defense responses in WT plants but enhanced them in lym3 mutant plants.To confirm this, changes in mRNA abundance of the defense-responsive gene WRKY29 (Liang et al. 2013) were analyzed in both WT and lym3 mutant plants after treatment with an elicitor, flg-22, with or without B. pumilus TUAT1 viable spore pretreatment as well as purified PGNs.The relative transcriptional level of WRKY29 was significantly increased by flg-22 treatment in both WT and lym3 mutant plants (Fig. 4A-F).In WT plants, WRKY29 expression was unchanged compared to the control by the simultaneous application of B. pumilus TUAT1 spores and flg-22 (Fig. 4A), while in lym3-1 and lym3-2 mutant plants, WRKY29 expression was increased by the simultaneous application of B. pumilus TUAT1 spores and flg-22 compared to the control to almost the same level of single flg-22 application (Fig. 4B, C).By the single application of B. pumilus TUAT1 spores, expression of WRKY29 gene was not changed compared to the control in WT plants (Fig. 4A), while it was increased in lym3 mutant plants (Fig. 4B, C).The similar trends were also observed when purified PGNs from B. pumilus TUAT1 were used in place of the spores.In WT plants, WRKY29 expression was unchanged compared to the control by the simultaneous application of B. pumilus TUAT1 PGNs and flg-22 (Fig. 4D), while in lym3-1 and lym3-2 mutant plants, WRKY29 expression was increased by the simultaneous application of B. pumilus TUAT1 PGNs and flg-22 to almost the same level of single flg-22 application (Fig. 4E, F).

Involvement of PGN receptor for growth promotion in Arabidopsis treated with TUAT1 spores and PGNs
Growth promotion by inoculating B. pumillus TUAT1 has been observed in various plants (Win et al. 2018, 2019, 2020, Agake et al. 2021, 2022a, 2022b, Hasibuan et al. 2021, Borjigin et al. 2024), and spores are more effective than vegetative cells (Ngo et al. 2019).Bacillus spores carry a thick PGN layer in their cell wall, so we hypothesized that perception of B. pumillus TUAT1 PGNs by plant receptors is involved in the mechanism of growth promotion by this strain.Consistent with our hypothesis, purified PGNs, as well as spores of B. pumillus TUAT1, promoted growth of WT Arabidopsis thaliana (Fig. 2).PRRs work as a master switch to initiate the signaling pathway by recognizing MAMPs and pathogen-associated molecular patterns like PGNs and flagellin and then transmitted the signal into the inside of the cell with the help of RLKs (Zhang et al. 2019).In the present study, we investigated whether PGN receptors for pathogens, i.e.LYM1, LYM3 and CERK1 (Willmann et al. 2011), are involved in perception of B. pumillus TUAT1 using A. thaliana mutants of lym1, lym3 and cerk1.
In the first study, root growth promotion by B. pumillus TUAT1 spores was observed in lym1-1, lym1-2 and cerk1-2, but not in lym3-1 and lym3-2 mutant plants, suggesting that LYM3 protein is required for the recognition of B. pumilus TUAT1 spores to promote growth (Fig. 1).Similarly, growth promotion by purified PGNs of B. pumilus TUAT1 was observed in WT, but not in lym3 mutant plants (Fig. 2), suggesting that LYM3 receptor protein is required for growth promotion by recognizing PGNs of B. pumilus TUAT1 strain.Since the LYM3 receptor protein lacks a cytoplasmic kinase domain, additional proteins are likely required for transmitting the signal inside of the cell after the perception of B. pumilus TUAT1.Regarding the perception of pathogen PGNs, signaling perceived by LYM3 with LYM1 is transmitted to CERK1.PGP effects by B. pumillus TUAT1 spores were observed in both allelic lym1 mutant plants, indicating that LYM1 is not involved in the response to this strain.As for CERK1, PGP effects were observed in cerk1-2.We also   tried to test another allelic mutant, cerk1-1 with Nossen background (Miya et al. 2007) but could not evaluate the result because WT of Nossen did not respond to B. pumillus TUAT1 (data not shown).Thus, possibility of involvement of CERK1 on PGP effects remains and has to be confirmed in further study.Whether LYM1 and CERK1 are involved together with LYM3 may be the critical point to discriminate responses by PGPR from those by pathogens.

The enriched slim GOs of genes regulated by B. pumilus TUAT1 inoculation
In DEGs of the lym3 mutant, all enriched GO slims of the biological process category were related to response to stress and stimulus, such as 'response to chemical' , 'response to stress' , 'response to external stimulus' and 'response to biotic stimulus' , and upregulated DEGs are more enriched than downregulated DEGs in these GO slims (Tables 5, 6).In contrast, in WT plants, downregulated DEGs are more enriched than upregulated DEGs in GO slims related to stress or stimulus responses, such as 'response to external stimulus' , 'response to stress' and 'response to chemical' (Tables 3, 4).This means that upregulated DEGs are more enriched in lym3, while downregulated DEGs are more enriched in WT plants in GO slims related to stress or stimulus responses.The numbers of DEGs affected by B. pumilus TUAT1 inoculation were much higher in lym3-1 and lym3-2 compared to WT plants (Fig. 3, Supplementary Fig. S1), which may be because stress responses due to the decrease or lack of the LYM3 transcripts in lym3 mutants (Fig. 3C) affected expression of more genes compared in WT plants.
To pay attention to the broader overview of transcriptional reprogramming, we also performed GO enrichment analysis for genes in cluster 2 and cluster 7 of the heatmap (Supplementary Table S4 and S5) as they showed similar expression patterns in the two allelic lym3-1 and lym3-2 mutant plants and the patterns are opposite to those in WT plants.In cluster 2 where genes were upregulated by inoculation of B. pumilus TUAT1 in WT and downregulated in lym3-1 and lym3-2, no stress-related GO slims were enriched (Supplementary Table S4).In cluster 7 where genes were downregulated by inoculation of B. pumilus TUAT1 in WT and upregulated in lym3-1 and lym3-2, the GO slim of 'response to chemical' was enriched (Supplementary Table S5).This also supports that stress response is alleviated in WT and induced in lym3 mutant plants by B. pumilus TUAT1 spores.

Involvement of LYM3 in alleviation of defense response by B. pumilus TUAT1 and PGNs
The results of RNA-seq suggest that B. pumilus TUAT1 spores act as a regulator to reduce defense-related gene expression and that LYM3 is required for this regulation.Supporting this hypothesis, B. pumilus TUAT1 alleviated the induction of the defense-responsive gene WRKY29 by flg-22 treatment in WT plants, but not in lym3-1 and lym3-2 mutant plants (Fig. 4A-C).The expression of WRKY29 gene was also significantly reduced by flg-22 treatment in the presence of PGNs in WT plants, but this trend was not observed in lym3-1 and lym3-2 mutant plants (Fig. 4D-F).These similar results of flg-22 treatment in the presence of PGNs and B. pumilus TUAT1 spores indicate that LYM3 receptor protein is required for recognizing PGNs of B. pumilus TUAT1 spores to alleviate defense responses.
Previously, Liang et al. (2013) demonstrated that rhizobium Nod factors, the signaling molecules for legume infection structurally similar chitotetraose, have the ability to suppress plant innate immunity in both legumes and non-legumes.The same paper showed that flg-22 triggered ROS production and induction of the WRKY29 gene and other defense responses were significantly reduced by pretreatment with Nod factors in both soybean and Arabidopsis.LYK3, which is a LysM-containing receptor kinase, is required for Nod factor suppression of flg-22-triggered immunity.It is possible that rhizobium Nod factors and PGNs in B. pumilus TUAT1 have similar roles in the suppression of defense responses through LysM-containing receptors or receptor kinases.Indeed, similar to our findings with PGN, it was also shown that treatment of plants with Nod factor stimulated root growth in maize and other plants (e.g.Tanaka et al. 2015, Buendia et al. 2019).
It is known that plant growth is negatively correlated with the induction of defense responses, and interactions between plant hormones and MAMP-triggered immunity are suggested in the mechanisms of this trade-off (Lozano-Durán et al. 2013, Huot et al. 2014).The present results are consistent with a model where PGNs of B. pumilus TUAT1 are recognized by LYM3 receptor protein, alleviating defense responses, which results in growth promotion.

Screening of Arabidopsis T-DNA knockout mutants for LYM1, LYM3 and CERK1
Arabidopsis thaliana plants were grown on rockwool in a plastic container at 22 ∘ C under a 16-h light (150 μmol m −2 s −1 )/8-h dark photoperiod using Molecular Genetics Research Laboratory hydroponic solution as a nutrient source (Fujiwara et al. 1992, Hirai et al. 1995).

Measurement of growth promotion effects by B. pumilus TUAT1 and PGNs on selected mutant lines
Bacillus pumilus TUAT1 cultures were grown as previously described (Agake et al. 2022b).Vegetative cells of B. pumilus TUAT1 were cultured in 300 ml of trypticase soy (TSB) broth in one litter Erlenmeyer flask at 180 rpm, 30 ∘ C for 24 h.A high concentration of spores was obtained by culturing for 72 h in Difco sporulation medium instead of TSB (Nicholson and Setlow 1990).The cells were collected by centrifugation, washed with sterilized water purified with reverse osmosis (RO) membrane several times and finally resuspended in sterilized RO water.Spore preparations were incubated at 65 ∘ C for 1 h to kill vegetative cells.The colony forming units (CFU) of the inoculant cultures were confirmed by plating onto T-soy agar.
Plant culture and inoculation with the B. pumilus TUAT1 strain were performed as follows.Seeds were surface-sterilized by immersion in 70% ethanol for 10 min and 1% NaClO for 10 min, washed several times with sterilized water and then sown on germination medium (Umezawa et al. 2009).The seeds were vernalized at 4 ∘ C in the dark for 2-3 d and then transferred into a growth chamber with a photoperiod of 16-h light (150 μmol m −2 s −1 photon flux density) and 8-h darkness at 22 ∘ C. Two weeks after germination, seedlings were transplanted into plastic trays (15 wells/tray) containing 72 g of autoclaved soil (Shinano Baiyoudo Co., Ltd., Nagano, Japan) per well and then grown in the growth chamber set as described earlier.The Shinano soil contains approximately 375 mg N kg −1 , 750 mg P 2 O 5 kg −1 and 375 mg K 2 O kg −1 .The plants were fertilized initially using 625 ml of liquid fertilizer diluted 2000 times with sterile RO water in each tray (15 plants) (Hanakoujyo, N:P:K = 8:10:5 mg ml −1 , Sumitomo Chemical Garden Products Inc., Tokyo, Japan), and subsequently irrigated, as needed, with sterilized RO water.After transplantation, the trays were covered with plastic films to maintain moisture for 1 d.After that, plastic films were removed, and 4 ml of spore suspension of B. pumilus TUAT1 at a concentration of 10 7 CFU ml −1 was applied to the soil in each well, where control plants were treated with 4 ml of sterilized RO water.
In case of PGN treatments, PGNs were extracted and purified from B. pumilus TUAT1 according to a previous report (Kühner et al. 2014).One milliliter of purified PGNs at a concentration of 300 μg ml −1 was applied to a plant grown in soil in each well, where control plants were treated with 1 ml of sterilized RO water.Plants were sampled at 14 d after spore inoculation as well as PGN treatments.
Fresh weight and dry weight of shoots and roots were determined in 15 replicates per treatment.To obtain dry weights, the plant materials were kept in a dryer at 80 ∘ C for 48 h.To measure the total root length, lateral root length and root surface area, plant roots were scanned using an Epson Perfection V700 Photo (Seiko Epson Corporation, Nagano, Japan) and analyzed using WinRHIZO software ver.2004 (Regent Instruments Inc., Quebec, Canada) as described by Haidari et al. (2017).

RNA-seq and data analysis
Plants were cultured and transplanted onto Shinano soil in plastic trays as described earlier.Three weeks after transplanting, each plant was treated with 4 ml of spore suspension of B. pumilus TUAT1 at 10 7 CFU ml −1 .The roots were sampled 1 h after inoculation.Roots from five plants were combined to generate one sample, and three biological replicates were prepared from 15 plants.Total RNAs were extracted using an RNeasy plant mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions.As previously reported (Ju et al. 2017), six libraries were constructed and sequenced by the Beijing Genomics Institute (www.genomics.org.cnBGI, Shenzhen, China).Clean tags were mapped to the reference genome and genes that were available in the Arabidopsis reference genome GCF_000001735.4_TAIR10.1.For gene expression analysis, the matched reads were calculated and then normalized to reads per kilobase of transcript per million mapped reads using the RESM software.The significance of differential gene expression was confirmed with the BGI bioinformatics service using a Q value of ≤0.05.
GOs were based on the GO annotations at the Arabidopsis Information Resource (TAIR) (https://www.arabidopsis.org/portals/genAnnotation/functional_annotation/go.jsp) (Berardini et al. 2004).The ratios of annotated GO slims, which are high-level terms from each GO hierarchy that are useful for grouping genes into broad categories (Berardini et al. 2004), in categories of biological process and molecular function, were compared to those in the whole genome.
Identification of genes with differential expression in WT and lym3 mutants, analysis of volcano plots, PCA, t-SNE plot, dendrogram and K-means clustering were performed using online based software iDEP1.1 (Ge et al. 2018).For the identification of DEGs with up-and downregulation, DEGs were defined as the log 2 fold change (FC) ≥1 and false discovery rate (FDR) was adjusted to ≤0.05.

Quantification of LYM3 transcripts in lym3-1 and lym3-2 mutant plants
The same amount of RNA extracted from roots of 3-weeks-old plants grown on Shinano soil were reverse transcribed using the Prime Script™ RT Reagent Kit with gDNA Eraser (Perfect Real Time) (Takara Bio, Shiga, Japan).cDNA was amplified by GoTaq® DNA Polymerase (Promega, Madison, WI, USA) using primers Lym3-f and Lym3-r (Willmann et al. 2011) or UBQ10_F and UBQ10_R (Supplementary Table S3) at 25, 30 and 35 cycles.LYM3 and UBQ10 transcript detected at 35 and 30 cycles, respectively, were quantified using GelAnalyzer 23.1 (available at www.gelanalyzer.com)by Istvan Lazar Jr., PhD, and Istvan Lazar Sr., PhD, CSc, and the intensities of LYM3 were normalized by those of UBQ10.

Effect of TUAT1 and PGNs on the induction of WRKY29 expression by flg-22
The synthesized flg-22 peptide (containing the conserved 22 amino acid sequence of Pseudomonas aeruginosa) was purchased from Alpha Diagnostic Intl Inc. (San Antonio, TX, USA).Plants were cultured and transplanted on Shinano soil in plastic trays as described earlier.Three weeks after transplanting, each plant was treated with 4 ml of B. pumilus TUAT1 spores suspension at 10 7 CFU ml −1 or sterilized RO water.In the case of purified PGN treatments, each plant was treated with 1 ml of PGNs at a concentration of 300 μg ml −1 or sterilized RO water.Immediately after treatment with B. pumilus TUAT1 spores and PGNs, the plants were treated with 4 ml of 1 μM flg-22.One hour after treatment with flg-22, the roots were sampled, immediately frozen in liquid N 2 and stored at −80 ∘ C until RNA extraction.Sterilized RO water was used as a control treatment instead of B. pumilus TUAT1 spores as well as PGNs and flg-22.
Total RNAs were extracted from the roots using a RNeasy plant mini Kit (Qiagen, Hilden, Germany), according to the manufacturer's instructions.Reverse transcription was performed using the Prime Script™ RT Reagent Kit with gDNA Eraser (Perfect Real Time) (Takara Bio, Shiga, Japan).Quantitative PCR was performed using the KAPA SYBR FAST qPCR Master Mix (2×) Kit (KAPA Biosystems, Wilmington, MA, USA) and Light Cycler 96 (Roche Diagnostics, Basel, Switzerland) following the manufacturer's protocols.The WRKY29 and UBQ10 genes were used as the target gene and internal control, respectively, using the same primers used by Liang et al. (2013) (Supplementary Table S3).Relative expression levels were calculated relative to the calculated values using the 2 -ΔΔC T method (Livak and Schmittgen 2001).

Fig. 1
Fig. 1 Growth promotion effects of B. pumilus TUAT1 in lym1, lym3 and cerk1 mutants.Root dry weights in the first (A) and second (B) experiments; lateral root length (C), total root length (D) and root surface area (E) in the second experiment.NI, not inoculated; I, inoculated with B. pumilus TUAT1.Boxplots denote the span from 25th to the 75th percentile and are centered on the median data of 15 biological replicates.Asterisks indicate significant differences between I and NI for each genotype by Student's t-test (P ≤ 0.05).n = 15.

Fig. 2
Fig. 2 Growth promotion effects of PGNs purified from B. pumilus TUAT1 in lym3 mutants.Fresh and dry weights of shoots and roots (A-D), total root length (E) and lateral root length (F).Control, treated with water; PGNs, treated with purified PGNs from B. pumilus TUAT1; TUAT1, treated with viable spores of B. pumilus TUAT1.The error bars indicate standard errors.Asterisks * and ** indicate significant difference from the control in each genotype by Student's t-test at P ≤ 0.05 and 0.01 levels, respectively.n = 15.

Fig. 3
Fig. 3 DEGs by B. pumilus TUAT1 treatment in roots of WT, lym3-1 and lym3-2 plants and quantification of LYM3 transcripts.Venn diagram of differentially upregulated genes (A) and downregulated genes (B) in WT, lym3-1 and lym3-2.NI, not inoculated; I, inoculated with B. pumilus TUAT1.Quantification of LYM3 transcripts (C).Data are shown as the mean ± SD from three independent biological replicate as the value of WT as 1.Asterisks indicate a significant difference from the controls by Dunnett's test (P ≤ 0.01).The images of gel of the bands of LYM3 and UBQ10 transcripts by RT-PCR are also shown.

Fig. 4
Fig. 4 Effects of B. pumilus TUAT1 variable spores and PGNs on WRKY 29 gene expression by flg-22.Plants were treated with flg-22 with or without B. pumilus TUAT1 spores as well as PGNs for 1 h.The control plants were treated with neither B. pumilus TUAT 1, flg-22 nor PGNs.Relative expression of WRKY29 normalized to UBQ10 is shown as the value of the control in each genotype as 1.WT (A), lym3-1 (B) and lym3-2 (C) plants were treated with flg-22 with or without TUAT1 spores, and WT (D), lym3-1 (E) and lym3-2 (F) plants were treated with flg-22 with or without PGNs.Data are shown as the mean ± SD from three independent biological replicates.Different letters indicate a significant difference by Tukey's test (P ≤ 0.05).

Table 1
DEGs in WT plants by B. pumilus TUAT1 inoculation (Q ≤ 0.05) Gene descriptions are according to The National Center for Biotechnology Information (https://www.ncbi.nlm.nih.gov/).a Asterisks at locus tag are DEGs detected WT and one of two lym3 allele.