SlBIR3 Negatively Regulates PAMP Responses and Cell Death in Tomato

Bri1-associated kinase 1 (BAK1)-interacting receptor-like kinase (BIR) proteins have been shown to play important roles in regulating growth and development, pathogen associated molecular pattern (PAMP)-triggered immunity (PTI) responses, and cell death in the model plant, Arabidopsis thaliana. We identified four BIR family members in tomato (Solanum lycopersicum), including SlBIR3, an ortholog of AtBIR3 from A. thaliana. SlBIR3 is predicted to encode a membrane localized non-arginine-aspartate (non-RD) kinase that, based on protein sequence, does not have autophosphorylation activity but that can be phosphorylated in vivo. We established that SlBIR3 interacts with SlBAK1 and AtBAK1 using yeast two-hybrid assays and co-immunoprecipitation and maltose-binding protein pull down assays. We observed that SlBIR3 overexpression in tomato (cv. micro-tom) and A. thaliana has weak effect on growth and development through brassinosteroid (BR) signaling. SlBIR3 overexpression in A. thaliana suppressed flg22-induced defense responses, but did not affect infection with the bacterial pathogen Pseudomonas syringae (PstDC3000). This result was confirmed using virus-induced gene silencing (VIGS) in tomato in conjunction with PstDC3000 infection. Overexpression of SlBIR3 in tomato (cv. micro-tom) and A. thaliana resulted in enhanced susceptibility to the necrotrophic fungus Botrytis cinerea. In addition, co-silencing SlBIR3 with SlSERK3A or SlSERK3B using VIGS and the tobacco rattle virus (TRV)-RNA2 vector containing fragments of both the SlSERK3 and SlBIR3 genes induced spontaneous cell death, indicating a cooperation between the two proteins in this process. In conclusion, our study revealed that SlBIR3 is the ortholog of AtBIR3 and that it participates in BR, PTI, and cell death signaling pathways.


Introduction
Plants are often subjected to various adverse environmental conditions during their entire life cycle, which are detrimental to growth and development [1][2][3][4]. Consequently, they have evolved various strategies to perceive and rapidly adapt to external environmental conditions. Response mechanisms include leucine-rich repeats receptor-like protein kinases (LRR-RLKs), the largest subgroup of the receptor-like protein kinase (RLKs) family in plants, which play vital roles in transmitting extracellular environmental stimuli to intracellular chemical signals [5][6][7]. Typical LRR-RLKs are composed of an extracellular leucine-rich repeat domain that senses a specific external environmental stimulus, a single-pass transmembrane domain that localizes the protein to the plasma membrane and an intracellular domain that relays the downstream intracellular signal via a sequential phosphorylation mechanism [8][9][10]. Based on the presence of an arginine (R) residue preceding the reported to exhibit a spontaneous cell death phenotype and constitutive activation of defense responses, which can be rescued by the functional loss of suppressor of bak1-interacting receptor like kinase 1 (BIR1)-1 (SOBIR1), calreticulin 3 (CRT3), ER-localized DnaJ-like protein 3b (ERdj3b), stromal-derived factor-2 (SDF2), or BAK1 [35,39,41]. BIR2, an atypical non-RD LRR-RLK, constitutively associates with BAK1 and interferes with the interaction between BAK1 and FLS2, resulting in the negative regulation of immune signaling and cell death, but has no effect on BR signaling [38].
Studies to date have focused particularly on the role of BIR1 and BIR2 in PTI signaling, resistance to fungal infection, control of cell death in association with, BAK1, but less is known about the involvement of BIR3 proteins. Here, a tomato ortholog to A. thaliana AtBIR3 (At1g27190), SlBIR3 (Solyc05g006570.1.1), was cloned and characterized. Genetic data suggested that SlBIR3 has weak effect on BR signaling and suppress PAMP responses. We verified that a developed Agrobacterium-mediated tobacco rattle virus (TRV)-based virus-induced gene silencing (VIGS) method [35,42] with TRV-RNA2 vector containing fragments of both the SlSERK3 and SlBIR3 genes was reliable and effective in tomato, and used this approach to show that SlBIR3 together with SlSERK3 suppresses cell death.
To confirm this, a phylogenetic analysis was carried out, in which orthologous genes from other species were identified from a plant membrane protein database (http://aramemnon.unikoeln.de/). Phylogenetic analysis showed Solyc02g067560.1.1 and Solyc02g087460.1.1 clustered together with AtBIR2, and Solyc05g006570.1.1 aligned within the same clade as AtBIR3 and AtBIR4 ( Figure S2), suggesting that the closest ortholog of AtBIR3 in tomato is Solyc05g006570.1.1, which we renamed SlBIR3.
An analysis of the deduced SlBIR3 protein sequence indicated that it contains an N-terminal signal peptide, five LRR regions, an extracellular juxtamembrane domain, a transmembrane domain, an intracellular juxtamembrane domain, a Ser/Thr cytoplasmic domain and a carboxy-terminal domain ( Figure 1A,B). These structures are almost identical to those of AtBIR2 and AtBIR3. A sequence alignment of the predicted kinase domain of SlBIR3, AtBIR3, AtBIR2, AtBAK1 and somatic embryogenesis-related kinase 5 (AtSERK5) revealed that SlBIR3 lacks a highly conserved R and the conserved catalytic D residue in subdomain VIb of the kinase domain ( Figure 1C). This is important for the catalytic function and kinase activity and so we concluded that SlBIR3 is a non-RD kinase, and that SlBIR3 is an atypical membrane localized receptor-like protein kinase.
To further confirm that SlBIR3 is a membrane localized receptor-like kinase, a subcellular localization experiment was performed, using agrobacterium mediated transient transformation of tobacco epidermal cells. The results suggested that SlBIR3 is mainly localized to the plasma membrane (Figure 2A), which is the same localization as that previously reported for BAK1 [18]. An in vitro phosphorylation assay was also performed to test whether SlBIR3 has autophosphorylation and transphosphorylation activities. As shown in Figure 2C, only the recombinant FLAG-BAK1 protein showed strong autophosphorylation of the threonine residues and also had the ability to transphosphorylate E. coli proteins on these. Neither autophosphorylation of recombinant FLAG-SlBIR3 nor the phosphorylation of E. coli proteins transphosphorylated by FLAG-SlBIR3 on the threonine residues was detected by the anti-pThr antibody, similar to the inactive recombinant kinase proteins FLAG-mBAK1(317E) and FLAG-AtSERK5. These results indicated that SlBIR3 is a membrane localized atypical receptor-like protein kinase.
Int. J. Mol. Sci. 2017, 18,1966 4 of 20 threonine residues was detected by the anti-pThr antibody, similar to the inactive recombinant kinase proteins FLAG-mBAK1(317E) and FLAG-AtSERK5. These results indicated that SlBIR3 is a membrane localized atypical receptor-like protein kinase.   In vitro SlBIR3 phosphorylation activity. The phosphorylation activity of the recombinant FLAG-SlBIR3, FLAG-BAK1, FLAG-mBAK1 (L317E) and FLAG-SERK5 proteins was detected using an anti-pThr antibody, and aliquots of the recombinant proteins were detected in different gels using anti-FLAG antibodies in Western blot analysis. CBB staining (bottom) shows equal loading.

SlBIR3 Interacts with SlBAK1 and AtBAK1
Previous studies have indicated that AtBIR3 interacts with AtBAK1 [38]. Since SlBIR3 is the tomato ortholog of AtBIR3, we investigated whether SlBIR3 could interact with SlBAK1 (SlSERK3B) and AtBAK1. An activation domain (AD) fused to the cytoplasmic domain of SlBIR3 (AD-SlBIR3) and a DNA-binding domain (BD) fused to the kinase domain of SlBAK1 and AtBAK1 (BD-SlBAK1, BD-BAK1) were generated for expression in yeast. As shown in Figure 3A, separate co-expression of AD-SlBIR3 with BD-SlBAK1 and BD-BAK1, conferred yeast cell growth on the SD/-Trp-Leu-Ade-His selection medium, as well as X-α-galactosidase activity. This demonstrated that SlBIR3 interacts with SlBAK1 and AtBAK1 in yeast.
Next, a pull-down assay was carried out to verify the interaction between SlBIR3 protein and BAK1. The cytoplasmic domain of SlBIR3 fused to maltose binding protein (MBP) immobilized on amylose resin was used as a bait to bind BAK1-FLAG fusion proteins. As shown in Figure 3B, BAK1-FLAG interacted with MBP-SlBIR3, the BAK1-FLAG and MBP-PUB13 (positive control), but not with MBP alone, indicating that SlBIR3 physically interacts with BAK1 in vitro. The lower panels show that the SlBIR3-GFP fusion protein mainly localizes to the plasma membrane. The top panels show that the non-infected tobacco leaf cells do not have green fluorescence signal and the middle panels show that the GFP is distributed throughout the control cells. Scale bars, 50 µm. (B) Western blot analysis of the expression of SlBIR3-GFP shown in (A) detected using an anti-GFP antibody. Coomassie brilliant blue (CBB) staining shows equal protein loading. (C) In vitro SlBIR3 phosphorylation activity. The phosphorylation activity of the recombinant FLAG-SlBIR3, FLAG-BAK1, FLAG-mBAK1 (L317E) and FLAG-SERK5 proteins was detected using an anti-pThr antibody, and aliquots of the recombinant proteins were detected in different gels using anti-FLAG antibodies in Western blot analysis. CBB staining (bottom) shows equal loading.

SlBIR3 Interacts with SlBAK1 and AtBAK1
Previous studies have indicated that AtBIR3 interacts with AtBAK1 [38]. Since SlBIR3 is the tomato ortholog of AtBIR3, we investigated whether SlBIR3 could interact with SlBAK1 (SlSERK3B) and AtBAK1. An activation domain (AD) fused to the cytoplasmic domain of SlBIR3 (AD-SlBIR3) and a DNA-binding domain (BD) fused to the kinase domain of SlBAK1 and AtBAK1 (BD-SlBAK1, BD-BAK1) were generated for expression in yeast. As shown in Figure 3A, separate co-expression of AD-SlBIR3 with BD-SlBAK1 and BD-BAK1, conferred yeast cell growth on the SD/-Trp-Leu-Ade-His selection medium, as well as X-α-galactosidase activity. This demonstrated that SlBIR3 interacts with SlBAK1 and AtBAK1 in yeast.
Next, a pull-down assay was carried out to verify the interaction between SlBIR3 protein and BAK1. The cytoplasmic domain of SlBIR3 fused to maltose binding protein (MBP) immobilized on amylose resin was used as a bait to bind BAK1-FLAG fusion proteins. As shown in Figure 3B, BAK1-FLAG interacted with MBP-SlBIR3, the BAK1-FLAG and MBP-PUB13 (positive control), but not with MBP alone, indicating that SlBIR3 physically interacts with BAK1 in vitro. Yeast strain AH109 containing the BAK1 cytoplasmic domain in the prey vector pGADT7 and BAK1 or SlBAK1 cytoplasmic domain in bait vector pGBKT7 were tested for growth on SD/-Leu-Trp (DDO), SD/-Leu-Trp-His-Ade (QDO), and SD/-Leu-Trp-His-Ade+X-a-gal (QDO/X-α-gal). The positive control was AH109:pGBKT7-53/pGADT7-RecT, the negative control were AH109:pGBKT7-Lam/pGADT7-RecT, AH109:pGBKT7/pGADT7-SlBIR3, AH109:pGBKT7-BAK1/pGADT7, and AH109:pGBKT7-SlBAK1/pGADT7; (B) SlBIR3 interacts with BAK1 in vitro. A MBP pull down assay was performed by incubating FLAG-tagged BAK1 cytoplasmic domain with MBP-tagged SlBIR3 cytoplasmic domain. The MBP-bound beads were used as a negative control and the MBP-tagged PUB13 cytoplasmic domain was used as a positive control. Bound proteins were subjected to Western blot analysis with an anti-FLAG antibody. The input proteins were analyzed using an anti-MBP antibody. CBB staining shows the protein inputs; (C) SlBIR3 interacts with BAK1 in vivo. The total proteins extracted from 11-day-old A. thaliana expressing SlBIR3-GFP in the Col-0 or bak1-4 background were immunoprecipitated with anti-GFP antibody and protein A agarose beads. The total and immunoprecipitated proteins were analyzed with anti-GFP, anti-pThr, and anti-BAK1 antibodies; (D) SlBIR3 interacts with SlBAK1 in vitro. A MBP pull down assay was performed by incubating FLAG-tagged SlBAK1 cytoplasmic domain with MBP-tagged SlBIR3 cytoplasmic domain. MBP-bound beads and MBP protein were used as a negative control. The bound proteins were subjected to Western blot analysis with an anti-FLAG antibody. The input proteins were analyzed with an anti-MBP antibody.
We then performed a co-immunoprecipitation analysis to test the interaction between SlBIR3 and BAK1 in vivo. First, a construct expressing SlBIR3-GFP protein driven by the CaMV 35S promoter was transformed into Arabidopsis thaliana Col-0 and the A. thaliana bak1-4 mutant. The transgenic plants were grown in liquid culture for 11 days and then protein extracts were made and subjected to immunoprecipitation using anti-GFP antibody and protein A agarose beads. The immunoprecipitated proteins were subjected to Western blot analysis with an anti-BAK1 antibody. As shown in Figure 3C, BAK1 was detected in the immunoprecipitated proteins from total protein extracts of SlBIR3 transgenic plants in Col-0, but not in extracts from wild type (Col-0), bak1-4 or SlBIR3 overexpression lines in the bak1-4 background. We noted that the phosphorylation of SlBIR3-GFP detected with an anti-pThr antibody in wild type was much stronger than that in bak1-4, indicating SlBIR3 may be phosphorylated by BAK1 in planta, and that SlBIR3 is a substrate of BAK1.
To further characterize the association between SlBIR3 and SlBAK1, another MBP pull-down assay was performed using MBP-tagged SlBIR3 and FLAG-tagged SlBAK1 cytoplasmic domain expressed in E. coli, which were purified using FLAG agarose beads and MBP amylose resin, respectively. As shown in Figure 3D, the SlBAK1-FLAG proteins interacted with MBP-SlBIR3 but not Yeast strain AH109 containing the BAK1 cytoplasmic domain in the prey vector pGADT7 and BAK1 or SlBAK1 cytoplasmic domain in bait vector pGBKT7 were tested for growth on SD/-Leu-Trp (DDO), SD/-Leu-Trp-His-Ade (QDO), and SD/-Leu-Trp-His-Ade+X-a-gal (QDO/X-α-gal). The positive control was AH109:pGBKT7-53/pGADT7-RecT, the negative control were AH109:pGBKT7-Lam/pGADT7-RecT, AH109:pGBKT7/pGADT7-SlBIR3, AH109:pGBKT7-BAK1/ pGADT7, and AH109:pGBKT7-SlBAK1/pGADT7; (B) SlBIR3 interacts with BAK1 in vitro. A MBP pull down assay was performed by incubating FLAG-tagged BAK1 cytoplasmic domain with MBP-tagged SlBIR3 cytoplasmic domain. The MBP-bound beads were used as a negative control and the MBP-tagged PUB13 cytoplasmic domain was used as a positive control. Bound proteins were subjected to Western blot analysis with an anti-FLAG antibody. The input proteins were analyzed using an anti-MBP antibody. CBB staining shows the protein inputs; (C) SlBIR3 interacts with BAK1 in vivo. The total proteins extracted from 11-day-old A. thaliana expressing SlBIR3-GFP in the Col-0 or bak1-4 background were immunoprecipitated with anti-GFP antibody and protein A agarose beads. The total and immunoprecipitated proteins were analyzed with anti-GFP, anti-pThr, and anti-BAK1 antibodies; (D) SlBIR3 interacts with SlBAK1 in vitro. A MBP pull down assay was performed by incubating FLAG-tagged SlBAK1 cytoplasmic domain with MBP-tagged SlBIR3 cytoplasmic domain. MBP-bound beads and MBP protein were used as a negative control. The bound proteins were subjected to Western blot analysis with an anti-FLAG antibody. The input proteins were analyzed with an anti-MBP antibody.
We then performed a co-immunoprecipitation analysis to test the interaction between SlBIR3 and BAK1 in vivo. First, a construct expressing SlBIR3-GFP protein driven by the CaMV 35S promoter was transformed into Arabidopsis thaliana Col-0 and the A. thaliana bak1-4 mutant. The transgenic plants were grown in liquid culture for 11 days and then protein extracts were made and subjected to immunoprecipitation using anti-GFP antibody and protein A agarose beads. The immunoprecipitated proteins were subjected to Western blot analysis with an anti-BAK1 antibody. As shown in Figure 3C, BAK1 was detected in the immunoprecipitated proteins from total protein extracts of SlBIR3 transgenic plants in Col-0, but not in extracts from wild type (Col-0), bak1-4 or SlBIR3 overexpression lines in the bak1-4 background. We noted that the phosphorylation of SlBIR3-GFP detected with an anti-pThr antibody in wild type was much stronger than that in bak1-4, indicating SlBIR3 may be phosphorylated by BAK1 in planta, and that SlBIR3 is a substrate of BAK1.
To further characterize the association between SlBIR3 and SlBAK1, another MBP pull-down assay was performed using MBP-tagged SlBIR3 and FLAG-tagged SlBAK1 cytoplasmic domain expressed in E. coli, which were purified using FLAG agarose beads and MBP amylose resin, respectively. As shown in Figure 3D, the SlBAK1-FLAG proteins interacted with MBP-SlBIR3 but not with amylose resin (beads) or MBP. Taken together, these results were consistent with an association of SlBIR3 with BAK1 and SlBAK1.

Overexpressing SlBIR3 Has Weak Effect on BR Signaling
Previous results indicated that BAK1 interacts with BRI1 to initiate early steps in BR signaling, resulting in the promotion of growth and development [9,17,18]. Since SlBIR3 interacts with BAK1, we reasoned that SlBIR3 may be involved in regulating growth and developmental processes through BR signaling. To test this hypothesis, a construct expressing the SlBIR3-GFP fusion protein driven by CaMV35S promoter was introduced into the micro-tom tomato cultivar. As shown in Figure S3, transgenic lines were obtained and expression of the transgene and recombinant protein verified by semi RT-PCR and Western blot analysis, respectively. We observed that overexpression of SlBIR3-GFP had no significant effect on growth or developmental features though appeared a slight reduction on plant height and expansion diameter, and no significant changes were detected in the transcript levels of the BR-responses genes SlCPD and SlDWF4 between SlBIR3 transgenic lines and wild type plants. These results suggested that SlBIR3 overexpression in tomato has weak effect on BR signaling.
Next, we tested whether overexpression of SlBIR3 in A. thaliana affects growth and development through BR signaling. The same transgenic construct used for tomato transformation described above was introduced into A. thaliana Col-0 by agroinfiltration and several transgenic lines with different recombinant protein expression levels were identified and analyzed. As shown in Figure S4, no significant differences were observed in the phenotypes of the transgenic lines and wild type plants, regardless of the level of SlBIR3-GFP protein accumulation. To further determine whether the BR sensitivity of these transgenic lines was abnormal, a root growth inhibition assay and a hypocotyl elongation assay were carried out. As shown in Figures S5A-E, no significant phenotypic differences were observed as a consequence of overexpressing SlBIR3 in wild type, but the growth of roots and hypocotyls of SlBIR3-GFP overexpression lines were less sensitive to exogenous 24-epiBL in a series of concentrations from 1 to 1000 nM compared to wild type plants. We then determined, using semi RT-PCR, that the expression levels of the BR-responsive genes constitutive photomorphogenesis and dwarfism (CPD) and DWARF4 (DWF4) were similar in wild type and in SlBIR3 overexpression seedlings ( Figure S5F). These results suggest that SlBIR3 does not dramatically but weakly affect the BR signaling pathway or BR-mediated growth and development in wild-type A. thaliana.

SlBIR3 Negatively Regulates PAMP-Triggered Responses but Not Responses to PstDC3000 Treatment
According to the results of previous studies indicating that BAK1 plays important functions in PTI signaling [27], we speculated that SlBIR3 may also play a role in PAMP-triggered responses. To test this, and to elucidate the biological significance of SlBIR3 in PTI signaling, we tested for seedling growth inhibition by flg22 (a conserved 22 amino acids peptide of flagellin) [43]. SlBIR3 overexpression lines (Col-0 background) were less sensitive to exogenous flg22 compared to wild type (Col-0) over a range of flg22 concentrations from 10 to 100 nM ( Figure 4A,B). It was reported that FRK1, an LRR-RLK, is involved in defense responses and was used as a marker gene in early PTI signaling activated by flg22 [44]. Accordingly, we measured the transcript levels of FRK1 using real-time qRT-PCR. Compared with wild type (Col-0), the expression level of FRK1 was significantly lower in 14-day-old SlBIR3 overexpression seedlings (Col-0) after treatment with 100 nM flg22 for 3 h. Meanwhile, we observed that the relative expression level of FRK1 in bak1-4 plants was even lower compared to that in wild type ( Figure 4C).
We next investigated the role of SlBIR3 in resistance against plant pathogen Pseudomonas syringae pv. tomato DC3000 (PstDC3000). Functional analyses of SlBIR3 were carried out in SlBIR3 silenced tomato plants (cv. Moneymaker) using a virus-induced gene silencing method. The pTRV2:00 and pTRV2-SlPDS vectors were used as negative (VIGS-vector) and positive (VIGS-SlPDS) controls, respectively. As previously reported [45], approximately four weeks after infiltration, a clear photo-bleaching phenotype was observed in almost all the positive control plants ( Figure S6A). The efficiency of SlBIR3 silencing in the transgenic plants (VIGS-SlBIR3) was assessed by semi-quantitative RT-PCR, which showed that the SlBIR3 transcript levels were lower than that in negative control ( Figure S6B). PstDC3000 infection assays were carried out in VIGS-SlBIR3 plants and the bacterial growth in VIGS-Vector and VIGS-SlBIR3 groups were very similar to each other ( Figure S6C). To further examine the role of SlBIR3 in PstDC3000 resistance, bacterial suspensions were injected into bak1-4, Col-0 and SlBIR3 overexpression lines (all Col-0 background). Bacterial growth in the various genotypes was measured after 4 days post-infection, and no significant differences were observed ( Figure S6D). 18,1966 8 of 20 quantitative RT-PCR, which showed that the SlBIR3 transcript levels were lower than that in negative control ( Figure S6B). PstDC3000 infection assays were carried out in VIGS-SlBIR3 plants and the bacterial growth in VIGS-Vector and VIGS-SlBIR3 groups were very similar to each other ( Figure  S6C). To further examine the role of SlBIR3 in PstDC3000 resistance, bacterial suspensions were injected into bak1-4, Col-0 and SlBIR3 overexpression lines (all Col-0 background). Bacterial growth in the various genotypes was measured after 4 days post-infection, and no significant differences were observed ( Figure S6D). with ACTIN2 used as the reference gene. 14-day-old seedlings were treated with 100 nM flg22 for three hours. Data indicate mean values ± SD and the asterisks indicate significant differences compared to wild-type Col-0, according to a Student's t test (* p < 0.05). Experiments were performed three times with similar results.

SlBIR3 Negatively Regulates the Resistance to Botrytis Cinerea in Tomato and A. thaliana
BAK1 has been shown to promote resistance to Botrytis cinerea [46]. To investigate whether SlBIR3 plays a similar role in tomato, detached leaves from wild type (micro-tom background) and SlBIR3-GFP overexpression lines (micro-tom) were incubated with 10 μL of a B. cinerea suspension (5 × 10 5 spores/mL) for 7 days prior to analysis. The disease symptoms were assessed and the lesion areas in the SlBIR3-GFP overexpression lines were approximately four-fold larger than in wild type ( Figure 5A,B), indicating that SlBIR3 suppresses resistance to B. cinerea infection in tomato. To verify this, detached leaves from a susceptible tomato cultivar, moneymaker, in which SlBIR3 expression was silenced by VIGS, were incubated with 5 mm diameter mycelium plugs of B. cinerea. The silencing efficiency of SlBIR3 gene is shown in Figure S6B and the disease symptoms were evaluated 4 days after the infection. As shown in Figure 5C,D, no significant difference was observed between SlBIR3 silenced plant and the control on lesion areas in leaves. We also examined infections phenotypes in plants of the same genotype using a drop-inoculation method ( Figure 5C,E), which confirmed that loss of SlBIR3 function did not alter the disease resistance to B. cinerea. . At least 12 seedlings were measured; (C) The relative expression level of FRK1 was measured by real time quantitative RT-PCR, with ACTIN2 used as the reference gene. 14-day-old seedlings were treated with 100 nM flg22 for three hours. Data indicate mean values ± SD and the asterisks indicate significant differences compared to wild-type Col-0, according to a Student's t test (* p < 0.05). Experiments were performed three times with similar results.

SlBIR3 Negatively Regulates the Resistance to Botrytis Cinerea in Tomato and A. thaliana
BAK1 has been shown to promote resistance to Botrytis cinerea [46]. To investigate whether SlBIR3 plays a similar role in tomato, detached leaves from wild type (micro-tom background) and SlBIR3-GFP overexpression lines (micro-tom) were incubated with 10 µL of a B. cinerea suspension (5 × 10 5 spores/mL) for 7 days prior to analysis. The disease symptoms were assessed and the lesion areas in the SlBIR3-GFP overexpression lines were approximately four-fold larger than in wild type ( Figure 5A,B), indicating that SlBIR3 suppresses resistance to B. cinerea infection in tomato. To verify this, detached leaves from a susceptible tomato cultivar, moneymaker, in which SlBIR3 expression was silenced by VIGS, were incubated with 5 mm diameter mycelium plugs of B. cinerea. The silencing efficiency of SlBIR3 gene is shown in Figure S6B and the disease symptoms were evaluated 4 days after the infection. As shown in Figure 5C,D, no significant difference was observed between SlBIR3 silenced plant and the control on lesion areas in leaves. We also examined infections phenotypes in plants of the same genotype using a drop-inoculation method ( Figure 5C,E), which confirmed that loss of SlBIR3 function did not alter the disease resistance to B. cinerea. To further confirm the role of SlBIR3 in B. cinerea resistance, a B. cinerea spore suspension (5 × 10 5 spores/mL) was sprayed on A. thaliana wild type, the bak1-4, mutant and SlBIR3 overexpression lines (all Col-0 background) and the plants were placed in high humidity conditions under light for 7 days. We observed necrotic spots over the entire leaf surface of the SlBIR3 overexpression lines and the bak1-4 mutant, while the necrotic symptoms in the wild type leaves were restricted to the infection sites ( Figure 6A). The lesion areas in the SlBIR3 overexpression lines were also much larger than in wild type ( Figure 6B), which correlated with the reduced chlorophyll levels caused by the B. cinerea infection ( Figure 6C). These results suggest that SlBIR3 suppresses resistance to B. cinerea infection in A. thaliana, which is consistent with its action in tomato. Taken together, these results suggest that overexpression of SlBIR3 weakens the basal defense and impairs resistance to B. cinerea in tomato and A. thaliana. To further confirm the role of SlBIR3 in B. cinerea resistance, a B. cinerea spore suspension (5 × 10 5 spores/mL) was sprayed on A. thaliana wild type, the bak1-4, mutant and SlBIR3 overexpression lines (all Col-0 background) and the plants were placed in high humidity conditions under light for 7 days. We observed necrotic spots over the entire leaf surface of the SlBIR3 overexpression lines and the bak1-4 mutant, while the necrotic symptoms in the wild type leaves were restricted to the infection sites ( Figure 6A). The lesion areas in the SlBIR3 overexpression lines were also much larger than in wild type ( Figure 6B), which correlated with the reduced chlorophyll levels caused by the B. cinerea infection ( Figure 6C). These results suggest that SlBIR3 suppresses resistance to B. cinerea infection in A. thaliana, which is consistent with its action in tomato. Taken together, these results suggest that overexpression of SlBIR3 weakens the basal defense and impairs resistance to B. cinerea in tomato and A. thaliana. To further confirm the role of SlBIR3 in B. cinerea resistance, a B. cinerea spore suspension (5 × 10 5 spores/mL) was sprayed on A. thaliana wild type, the bak1-4, mutant and SlBIR3 overexpression lines (all Col-0 background) and the plants were placed in high humidity conditions under light for 7 days. We observed necrotic spots over the entire leaf surface of the SlBIR3 overexpression lines and the bak1-4 mutant, while the necrotic symptoms in the wild type leaves were restricted to the infection sites ( Figure 6A). The lesion areas in the SlBIR3 overexpression lines were also much larger than in wild type ( Figure 6B), which correlated with the reduced chlorophyll levels caused by the B. cinerea infection ( Figure 6C). These results suggest that SlBIR3 suppresses resistance to B. cinerea infection in A. thaliana, which is consistent with its action in tomato. Taken together, these results suggest that overexpression of SlBIR3 weakens the basal defense and impairs resistance to B. cinerea in tomato and A. thaliana.

SlBIR3 and SlSERK3 Control Cell Death in Tomato
It has been reported that co-silencing SlSERK3A and SlSERK3B in tomato leaves using VIGS triggers cell death [47]. To investigate whether SlBIR3 is involved in controlling cell death, we used VIGS to co-silence SlSERK3 and SlBIR3 in the tomato cultivar, Moneymaker. The silencing efficiency of the target genes ( Figure S7A) was determined using real-time qRT-PCR. As shown in Figures 7A  and S7B, the cell death phenotype was observed in both SlSERK3A/SlBIR3 and SlSERK3B/SlBIR3 cosilenced plants, which is similar to that in SlSERK3A/SlSERK3B co-silenced plants. We observed no cell death symptoms in plants in which SlSERK3A, SlSERK3B or SlBIR3 were silenced separately ( Figures 7A and S7B). Cell death was visualized by trypan blue staining ( Figure 7A).
Another study showed that defense related genes are upregulated in SlSERK3A/SlSERK3B cosilenced plants [47] and so we measured the relative transcript levels of two pathogenesis genes, SlPRb1 and SlPR2, in the leaves shown in Figure S7B, using real-time qRT-PCR. The expression levels of these two genes was significantly higher in both SlSERK3A/SlBIR3B and SlSERK3B/SlBIR3 cosilenced leaves, and was similar to that in SlSERK3A/SlSERK3B co-silenced leaves ( Figure 7B,C), which is consistent with the cell death results ( Figure 7A). These results indicated that SlBIR3 suppresses cell death together with SlSERK3A and SlSERK3B in tomato.

SlBIR3 and SlSERK3 Control Cell Death in Tomato
It has been reported that co-silencing SlSERK3A and SlSERK3B in tomato leaves using VIGS triggers cell death [47]. To investigate whether SlBIR3 is involved in controlling cell death, we used VIGS to co-silence SlSERK3 and SlBIR3 in the tomato cultivar, Moneymaker. The silencing efficiency of the target genes ( Figure S7A) was determined using real-time qRT-PCR. As shown in Figure 7A and Figure S7B, the cell death phenotype was observed in both SlSERK3A/SlBIR3 and SlSERK3B/SlBIR3 co-silenced plants, which is similar to that in SlSERK3A/SlSERK3B co-silenced plants. We observed no cell death symptoms in plants in which SlSERK3A, SlSERK3B or SlBIR3 were silenced separately ( Figure 7A and Figure S7B). Cell death was visualized by trypan blue staining ( Figure 7A).
Another study showed that defense related genes are upregulated in SlSERK3A/SlSERK3B co-silenced plants [47] and so we measured the relative transcript levels of two pathogenesis genes, SlPRb1 and SlPR2, in the leaves shown in Figure S7B, using real-time qRT-PCR. The expression levels of these two genes was significantly higher in both SlSERK3A/SlBIR3B and SlSERK3B/SlBIR3 co-silenced leaves, and was similar to that in SlSERK3A/SlSERK3B co-silenced leaves ( Figure 7B,C), which is consistent with the cell death results ( Figure 7A). These results indicated that SlBIR3 suppresses cell death together with SlSERK3A and SlSERK3B in tomato.

SlBIR3 and SlSERK3 Control Cell Death in Tomato
It has been reported that co-silencing SlSERK3A and SlSERK3B in tomato leaves using VIGS triggers cell death [47]. To investigate whether SlBIR3 is involved in controlling cell death, we used VIGS to co-silence SlSERK3 and SlBIR3 in the tomato cultivar, Moneymaker. The silencing efficiency of the target genes ( Figure S7A) was determined using real-time qRT-PCR. As shown in Figures 7A  and S7B, the cell death phenotype was observed in both SlSERK3A/SlBIR3 and SlSERK3B/SlBIR3 cosilenced plants, which is similar to that in SlSERK3A/SlSERK3B co-silenced plants. We observed no cell death symptoms in plants in which SlSERK3A, SlSERK3B or SlBIR3 were silenced separately ( Figures 7A and S7B). Cell death was visualized by trypan blue staining ( Figure 7A).
Another study showed that defense related genes are upregulated in SlSERK3A/SlSERK3B cosilenced plants [47] and so we measured the relative transcript levels of two pathogenesis genes, SlPRb1 and SlPR2, in the leaves shown in Figure S7B, using real-time qRT-PCR. The expression levels of these two genes was significantly higher in both SlSERK3A/SlBIR3B and SlSERK3B/SlBIR3 cosilenced leaves, and was similar to that in SlSERK3A/SlSERK3B co-silenced leaves ( Figure 7B,C), which is consistent with the cell death results ( Figure 7A). These results indicated that SlBIR3 suppresses cell death together with SlSERK3A and SlSERK3B in tomato.

Discussion
Plant LRR-RLK proteins play key roles in the sensing of external biotic and abiotic stimuli, and in their conversion into internal chemical signals that regulate responses to ever changing environmental conditions [48]. According to their activation or catalytic mode of action, LRR-RLK proteins can be grouped into RD kinases and non-RD kinases [15], and the latter have usually been associated with plant innate immunity [15]. Previously, a small subfamily of BIR proteins were identified in A. thaliana and categorized as non-RD kinases [38]. Here, a tomato non-RD kinase gene, named SlBIR3 (Solyc05g006570.1.1), an ortholog of AtBIR3, was shown to play roles in regulating PTI responses and controlling cell death.
A phylogenetic analysis revealed the widespread presence of the BIR proteins family in plants ( Figure S2). Genomic DNA and complementary DNA sequences analyses showed that SlBIR3 has only one exon, which is consistent with AtBIR3 [38] and the closest orthologs from other plant species, indicating conserved exon composition ( Figure S2). A domain analysis indicated that SlBIR3 has five leucine-rich repeats (LRR) regions, a transmembrane domain and a Ser/Thr cytoplasmic domain ( Figure 1). This is also consistent with the domain structures of AtBIR3 and AtBIR2 [38] and is a typical and conserved signature motif of an LRR-RLK [48]. Previous studies demonstrated that LRR-RLK proteins lacking the conserved catalytic subdomain in the cytoplasmic domain usually do not have autophosphorylation and transphosphorylation activity in vitro [11,15,39,49,50]. Our results ( Figures 1C and 2) support the idea that SlBIR3 does not have autophosphorylation and transphosphorylation activity due to the absence of the RD motif in the subunit VIb domain. Previous studies showed that non-RD kinases can also be phosphorylated in vivo [50,51] and some of phosphorylation sites are critical for maintaining their functions [51,52]. In the current study, the phosphorylation of threonine residues in SlBIR3 was detected in vivo ( Figure 3C) by immunoblotting, using a phosphothreonine antibody, which is consistent with previous studies of BIR2 [50].
It has been reported that functional BAK1 in A. thaliana and tomato, and members of the BIR proteins family in A. thaliana are mainly localized in the plasma membrane [16,38,39]. Domain analysis suggested that there is a signal peptide in the N-terminus of SlBIR3 ( Figure 1A,B) and subcellular localization analysis indicated that SlBIR3 is mainly localized on the cell surface ( Figure  2A,B). Taken together, these results suggest that SlBIR3 may interacts with BAK1 in vivo. Indeed, yeast two hybrid, co-IP and MBP pull down assays all confirmed that SlBIR3 can interact with SlBAK1 and AtBAK1 (Figure 3), which is consistent with previous reports that BIR proteins can interact with BAK1 in A. thaliana [38,41]. We noticed that SlBIR3 phosphorylation was greatly weakened in the

Discussion
Plant LRR-RLK proteins play key roles in the sensing of external biotic and abiotic stimuli, and in their conversion into internal chemical signals that regulate responses to ever changing environmental conditions [48]. According to their activation or catalytic mode of action, LRR-RLK proteins can be grouped into RD kinases and non-RD kinases [15], and the latter have usually been associated with plant innate immunity [15]. Previously, a small subfamily of BIR proteins were identified in A. thaliana and categorized as non-RD kinases [38]. Here, a tomato non-RD kinase gene, named SlBIR3 (Solyc05g006570.1.1), an ortholog of AtBIR3, was shown to play roles in regulating PTI responses and controlling cell death.
A phylogenetic analysis revealed the widespread presence of the BIR proteins family in plants ( Figure S2). Genomic DNA and complementary DNA sequences analyses showed that SlBIR3 has only one exon, which is consistent with AtBIR3 [38] and the closest orthologs from other plant species, indicating conserved exon composition ( Figure S2). A domain analysis indicated that SlBIR3 has five leucine-rich repeats (LRR) regions, a transmembrane domain and a Ser/Thr cytoplasmic domain (Figure 1). This is also consistent with the domain structures of AtBIR3 and AtBIR2 [38] and is a typical and conserved signature motif of an LRR-RLK [48]. Previous studies demonstrated that LRR-RLK proteins lacking the conserved catalytic subdomain in the cytoplasmic domain usually do not have autophosphorylation and transphosphorylation activity in vitro [11,15,39,49,50]. Our results ( Figures  1C and 2) support the idea that SlBIR3 does not have autophosphorylation and transphosphorylation activity due to the absence of the RD motif in the subunit VIb domain. Previous studies showed that non-RD kinases can also be phosphorylated in vivo [50,51] and some of phosphorylation sites are critical for maintaining their functions [51,52]. In the current study, the phosphorylation of threonine residues in SlBIR3 was detected in vivo ( Figure 3C) by immunoblotting, using a phosphothreonine antibody, which is consistent with previous studies of BIR2 [50].
It has been reported that functional BAK1 in A. thaliana and tomato, and members of the BIR proteins family in A. thaliana are mainly localized in the plasma membrane [16,38,39]. Domain analysis suggested that there is a signal peptide in the N-terminus of SlBIR3 ( Figure 1A,B) and subcellular localization analysis indicated that SlBIR3 is mainly localized on the cell surface (Figure 2A,B). Taken together, these results suggest that SlBIR3 may interacts with BAK1 in vivo. Indeed, yeast two hybrid, co-IP and MBP pull down assays all confirmed that SlBIR3 can interact with SlBAK1 and AtBAK1 (Figure 3), which is consistent with previous reports that BIR proteins can interact with BAK1 in A. thaliana [38,41]. We noticed that SlBIR3 phosphorylation was greatly weakened in the absence of BAK1 ( Figure 3C); however, it was detectable ( Figure 3C), indicating that SlBIR3 can be targeted in planta by protein kinases other than BAK1. This conclusion is supported by previous studies in which BIR3 was identified in SERK1 immunoprecipitates using tandem mass-spectrometry [53].
Functional analysis of SlBIR3 overexpression or knockdown tomato and A. thaliana plants revealed that SlBIR3 participates in BAK1-regulated PTI and cell death pathways, but has weak effect on BR signaling. We found that overexpressing SlBIR3 in tomato (micro-tom background) caused no significant phenotypic changes or to the expression levels of the BR responsive genes, SlCPD [47] and SlDWARF [54] (Figure S3A-D). This is consistent with a previous study of BIR2 in A. thaliana [38], and indicates that SlBIR3 has weak effect on plant growth and development through BR signaling. Micro-tom is a weak BR biosynthetic mutant and contains a mutation in the D gene (CYP85A1) [55]. Two orthologous genes in A. thaliana are brassinosteroid-6-oxidase 1 (BR6OX1) and brassinosteroid-6-oxidase 2 (BROX2) [56] and a BR-deficient br6ox1 br6ox2 double mutant was used to study the BR biosynthesis mechanism [57]. The growth impairment of another weak BR-deficient A. thaliana mutant, de-etiolated 2 (det2) [58,59], could be partially rescued by loss of suppressor of BRI1 (SBI1) [59], BIK1 [60], BKI1 [61], and protein phosphatase 2A (PP2A) [62]; all components of the BR signaling pathway. Together, these studies suggest that BR-deficient micro-tom can provide a valuable resource for BR signaling research.
It was reported that overexpression of SlBRI1 and GmBRI1, respectively, could partially complement the phenotype of the weak BR-insensitive mutant bri1-5 [63][64][65] and overexpression of SlSERK3 in the BAK1 loss-of-function A. thaliana mutant bak1-4 restored its sensitivity to exogenous BR [47]. We demonstrated that overexpressing SlBIR3 in A. thaliana Col-0 did not result in any significant phenotypic changes ( Figure S4) and that the BR sensitivity of SlBIR3 transgenic lines was only slightly altered ( Figure S5). These results suggest that SlBIR3 has no apparent but weak effect on BR signaling, although it has yet to be determined whether the other SlBIR proteins have functions in BR signaling.
In a microbe rich environment, plants mainly depend on innate immunity to defend themselves against invasion [66,67]. In our study, we observed that PTI signaling induced by flg22 was reduced in SlBIR3 overexpression lines (Figure 4), indicating that SlBIR3 negative regulates PTI signaling. This is contrary to the function of its partner BAK1 [27] and that of the flagellin receptor FLS2 [68], but is consistent with its paralog, BIR2. This suggested that SlBIR3 may be a negative regulator of BAK1 in PAMP responses, but the response of the SlBIR3 transgenic and knockdown lines to PstDC3000 bacteria was unaffected ( Figure S6), which is consistent with the phenotypes of knocking out AtBAK1 in A. thaliana [46] and knocking down SlBAK1 in tomato [47]. We propose that this may be due to the involvement of SlBIR3 in cell death (Figures 5-7).
The necrotrophic fungus, B. cinerea, a pathogen worldwide, has a wide range of hosts and causes severe damage to crops, especially tomato [69]. Previous studies reported that loss of function of BAK1 [46] and BIR2 [38] in A. thaliana, respectively, resulted in decreased resistance to B. cinerea. In our study, we found that role of SlBIR3 in resistance to B. cinerea is opposite to that of BIR2 [38], indicating functional differentiation of BIR during evolution. This result also supports the idea that SlBIR3 is a negative regulator of BAK1 in PAMP responses.
Several studies reported that BAK1 plays important roles in spontaneous cell death. Loss-of-function of BAK1 and BAK1-like 1 (BKK1) together causes a spontaneous cell death phenotype in a salicylic acid (SA) dependent manner, which could be rescued by loss-of-function of suppressor of BAK1 BKK1 (SBB1), and partial loss of endoplasmic reticulum quality control (ERQC) components, pep receptor 1 (PERP1) and PEP RECEPTOR 2 (PEPR2) [35,36,70]. Loss-of-function of BAK1 and BIK1 together was also reported to induce spontaneous cell death, which was rescued by loss-of-function of phytoalexin deficient 4 (PAD4) and salicylic acid induction deficient 2 (SID2), respectively [71]. The role of BAK1 in controlling spontaneous cell death, together with other RLKs, was further verified by our results showing that co-silencing of SlBIR3/SlSERK3A and SlBIR3/SlSERK3B lead to leaf necrosis, which was similar to the co-silencing of SlSERK3A/SlSERK3B by VIGS in tomato [47]. This suggests that SlBIR3 negatively regulates spontaneous cell death and that SlSERK3s is critical for the development of SlBIR3 knockdown mutant (Figure 7 and Figure S7). BIR1 regulates temperature-dependent spontaneous cell death with BON1 [40] and the spontaneous cell death phenotype in loss of BIR1 function mutant, bir1, could be rescued by loss of function of BAK1 [41]. Cell death in the bir2 mutant was only observed after fungal infection [38]. We conclude that the cell death pathway controlled by SlBIR3 is distinct from that regulated by BIR1 and BIR2. It will be interesting to test whether the components that can rescue BAK1-related spontaneous cell death is involved in SlBIR3-controlled cell death and whether SlBIR3 is involved in BAK1-regulated other signaling, such as that associated with abscisic acid (ABA) signaling in guard cells [72].

Tobacco Transient Transformation
The coding sequence of SlBIR3 was subcloned into a CaM35SGFP vector [75] forming a construct encoding SlBIR3 with a C-terminal GFP, driven by the CaMV35S promoter. The primers used are listed in Table S1. An A. tumefaciens (GV3101) culture transformed [76] with the above vector was shaken at 28 • C, 200 rpm for overnight. The OD 600 of the culture was adjusted to 0.2 with infiltration medium (27.8 mM d-glucose, 2.5 mM MES, pH 5.8 (adjusted by addition of 1 M NaOH), 0.1 mM Na 3 PO 4 ·12H 2 O, 5 mM acetosyringone). The culture A was then left at room temperature for at least 3 h before infiltration into tobacco leaves [76]. The GFP signal in the transformed leaves was detected three days after infiltration using a fluorescence microscope (BX51, Olympus, Tokyo, Japan) under blue light. Images were processed by the ImageJ software.

Yeast Two-Hybrid Assay
To verify interactions between SlBIR3 and BAK1, the pGADT7 plasmid (Clontech, Mountain View, CA, USA) containing the cytoplasmic domain of SlBIR3, and pGBKT7 plasmid (Clontech, Mountain View, CA, USA) containing the cytoplasmic domain of AtBAK1 or SlBAK1 were co-transformed into yeast (Saccharomyces cerevisiae) strain AH109 (Clontech, Mountain View, CA, USA). The resulting transformants were grown on SD/-Leu-Trp (DDO) medium at 30 • C for 3 days. Individual colonies were picked from the plates and subcultured in 2 mL of SD/-Leu-Trp liquid medium at 30 • C for 24 h. The cultured cells were spotted on SD/-Leu-Trp, SD/-Leu-Trp-His-Ade (QDO, and SD/-Leu-Trp-His-Ade + X-α-Gal (QDO + X-α-Gal) medium and incubated at 30 • C. Plates were photographed after 4 days of incubation.

Protein Extraction, Co-IP and Western Blot Analyses
Ten micrograms of total protein extracted from the transgenic plants or the infected tobacco leaves of the same age were loaded on SDS-PAGE gels for the western blot analyses with anti-GFP (1:5000) antibody (Transgene, Beijing, China, Cat. #HT801-01) as previously described [78]. A. thaliana seedlings used for co-IP were grown in shaking liquid culture [9] without 1 mM 24-epiBL (Sigma-Aldrich, Saint Louis, MO, USA) treatment. Total protein isolation, immunoprecipitation, and immunoblot analysis were performed as previously described in [78]

In Vitro Pull-Down Assay
The cytoplasmic domains of SlBAK1 and AtBAK1 were amplified and subcloned into the pFLAG-MAC vector (Sigma-Aldrich, Saint Louis, MO, USA). The cytoplasmic domains of SlBIR3 and PUB13 were amplified and subcloned into the pMal-C2 vectors (New England Biolab, Ipswich, MA, USA), to generate MBP tagged proteins. The primers used are listed in Table S1. Each of the constructs was transformed into E. coli BL21(DE3)pLysS. Recombinant SlBAK1-FLAG and AtBAK1-FLAG were purified with Anti-FLAG M2 Affinity Gel beads (Sigma-Aldrich, Saint Louis, MO, USA) according to manufacturer's protocols, and MBP, MBP-SlBIR3, MBP-PUB13 fusion proteins were purified after immobilization with amylose resin (New England Biolab, Ipswich, MA, USA, Cat. #DP301-01) following standard protocols. Three micrograms of FLAG fusion proteins were preincubated with 10 µL prewashed amylose resin in 120 µL incubation buffer (1 mM NaCl, 20 mM MgCl 2 , 0.5% Triton X-100, and 0.1 M HEPES, pH 7.2 (adjusted by addition of 1 M NaOH) for 1 h at 4 • C. After centrifugation at 5000× g for 5 min at room temperature the resin was collected and washed 5 times with incubation buffer. The pull-down proteins were detected by Western blot analysis with an anti-FLAG antibody (1:5000) (Transgene, Beijing, China, Cat. #HT201-01).

Root and Hypocotyl Growth Analyses
After surface sterilization with 15% NaClO containing 0.1% Triton X-100 (Sigma-Aldrich, Saint Louis, MO, USA), seeds were stored at 4 • C for two days and then planted on 1/2 MS medium supplemented with 0.7% (w/v) agar, 1% (w/v) sucrose and different concentrations of 24-epiBL. The plates were then placed vertically at 22 • C and the root length was measured seven days after germination under long-light conditions (16 h/8 h), and the hypocotyl length was measured five days after germination in the dark [33].

Seedling Growth Inhibition Assay
The seedling growth inhibition assays of the mentioned seedlings were performed as previously described [79].

VIGS of SlBIR3 in Tomato
A 400 bp of SlBIR3 corresponding to 1-400 bp of SlBIR3 was amplified from cDNA derived from tomato (cv. moneymaker) and subcloned into EcoRI-BamHI-digested pTRV2 vector [45], generating the construct pTRV2-SlBIIR3, which was used to silence SlBIR3 expression in tomato. A 409 bp of SlPDS, corresponding to 858-1266 bp of SlPDS, was also amplified from the tomato (cv. moneymaker) cDNA and subcloned into EcoRI-BamHI-digested pTRV2 vector forming the construct pTRV2-SlPDS, which was used to silence SlPDS and as a positive control. The empty pTRV2 vector was used as a negative control. VIGS constructs to silence SlSERK3A and SlSERK3B separately, and for both genes SlSERK3A/SlSERK3B together, as well as fragments of SlSERK3A (152 bp), SlSERK3B (174 bp) and SlSERK3A/B (177 bp), were amplified from tomato (cv. Moneymaker) cDNA and subcloned into BamHI-XhoI-digested pTRV2 vector, forming the constructs pTRV2-SlSERK3A, pTRV2-SlSERK3B and pTRV2-SlSERK3A/B [47]. VIGS constructs for both genes SlBIR3/SlSERK3A and SlBIR3/SlSERK3A, fragments of SlBIR3 (400 bp), as mentioned above, were subcloned into EcoRI-BamHI-digested pTRV2-SlSERK3A and pTRV2-SlSERK3B vectors, respectively, forming the constructs pTRV2-SlBIR3/SlSERK3A and pTRV2-SlBIR3/SlSERK3B, respectively [35]. The recombinant constructs and the empty vector were each transformed into A. tumefaciens strain GV3101. The primers used to generate these constructs are listed in Table S1. The tomato and the leaf infiltration method used for the VIGS assay was as previously described [45].

Pathogen Preparation and Inoculation
Pseudomonas syringae pv. tomato DC3000 (PstDC3000) infections were performed as previously described [80]. Botrytis cinerea strain ACCC30091 was obtained from the Agricultural Culture Collection of China (http://www.accc.org.cn/htdocs/epages.asp?id=13) and the strain was rejuvenated in tomato fruit, separated and grown in potato dextrose agar (PDA) medium [81]. B. cinerea spores were washed down from the PDA plate with distilled water. After gentle shaking (100 rpm) at 26 • C for 30 min, the suspension was filtrated with absorbent cotton gauze and then centrifuged at 800× g for 3 min at room temperature. The supernatant was poured off and the collected spores were gently resuspended in distilled water containing 0.025% Tween and 2% glucose. The spores were counted using a hemacytometer under a light microscrope and the spore suspension (5 × 10 5 spores/mL) were prepared. For B. cinerea infections of tomato, each detached leaf from 4-week-old tomato was incubated with 10 µL B. cinerea suspension using a drop-inoculation method or 5 mm B. cinerea diameter mycelium plugs [69,81]. The detached leaves were placed in a chamber at 21-23 • C, 16 h light/8 h dark for 7 days in the case of drops incubation and for 4 days in the case of mycelium plug incubation. For B. cinerea infections of A. thaliana, spore suspension (5 × 10 5 spores/mL) containing 0.025% Tween and 2% glucose were sprayed on 5-week-old plants, and the plants were then grown in 100% relative humidity and at 24 • C under light for 7 days. Leaf areas were calculated using ImageJ software.

Semi-Quantitative and Real-Time Quantitative RT-PCR Analyses
To measure the expression levels of SlBIR3, SlDWARF and SlCPD in the transgenic tomato lines, total RNA was extracted from five-week-old plants. To measure the expression levels of DWF4 and CPD in A. thaliana transgenic lines, total RNA was extracted from 14-day-old seedlings grown on 1/2 MS medium supplemented with 0.7% (w/v) agar and 2% (w/v) sucrose. For the flag22 sensitivity test, total RNA was extracted from 14-day-old seedlings grown on 1/2 MS medium supplemented with 1% (w/v) sucrose with or without a 3 h 100 nM flg22 (Genscript, Wuhan, China) treatment. For VIGS assay, the inoculated leaves were used for extracting total RNA. In each case, total RNA was extracted with RNAiso Plus kit (TaKaRa, Dalian, China) according to the manufacturer's protocol and transcribed to cDNA with Transcriptor First Strand cDNA Synthesis Kit (Roche, Mannheim, Germany) according to the manufacturer's protocol after treatment with DNase I (Thermo Scientific, Waltham, MA, USA). For semi-quantitative RT-PCR, PCR was performed using a 2× Es Taq MasterMix (Cwbio, Beijing, China). For real-time quantitative RT-PCR, PCR was performed using a SYBR Green Master Mix kit (Vazyme, Nanjing, China) with the above cDNA, and the relative quantitative expression of FRK1, SlPR1b1, SlPR2.0, SlSERK3A, SlSERK3B, and SlBIR3 were detected using ABI StepOne plus (ABI, Foster, CA, USA) with ACTIN2 or SlUBI3 as a normalized control and calculated using the delta-delta Ct method [82]. The PCR reaction was performed in a 20-µL volume in 8-well tubes heated for 1 cycle of 5 min at 95 • C, 40 cycles of 10 s at 95 • C and 30 s at 60 • C, and final a dissociation stage. For each primer combination, the specificity and the amplification efficiency was checked by melting-curve analysis and standard-curve analysis, respectively. The primers used are listed in Table S1.

Trypan-Blue Staining and Chlorophyll Measurement
Trypan-blue Staining used for detecting cell death was performed as previously described [83]. Chlorophyll content measurements were performed as previously described [84] with modifications. Briefly, 10 rosette leaves infected by B. cinerea were placed in a vial containing 15 mL 80% acetone and chlorophyll was extracted into the fluid without grinding at room temperature by incubating for 24 h under dark condition. The supernatant from the vial was made up to 20 mL with 80% acetone. The absorbance of the mixture was measured using ultraviolet spectrophotometer (UV-1750, Shimadzu, Tokyo, Japan) and the total chlorophyll (chlorophyll a + b) was calculated as described in [84].

Conclusions
Solyc05g006570.1.1 encodes SlBIR3, which is the A. thaliana ortholog of AtBIR3. Amino acid sequence and kinase activity analyses indicated that SlBIR3 is a membrane localized non-RD kinase, which interacts with, and could be phosphorylated by, BAK1. Overexpression of SlBIR3 in plants suppresses flg22 induced PAMP responses, enhances the susceptibility to B. cinerea infection, but has weak effect on BR signaling. The use of VIGS with TRV-RNA2 containing fragments from SlBIR3 and SlSERK3 revealed that SlBIR3 suppresses cell death in tomato together with SlSERK3s.