Phosphorylation of BIK1 is critical for interaction of downstream signaling components

Botrytis-induced Kinase 1 (BIK1) is a receptor-like cytoplasmic kinase (RLCK) involved in the defense, growth, and development of higher plants. It interacts with various receptor-like kinases (RLKs) such as Brassinosteroid Insensitive 1 (BRI1), Flagellin Sensitive 2 (FLS2), and Perception of the Arabidopsis Danger Signal Peptide 1 (PEPR1), but little is known about signaling downstream of BIK1. Interestingly, Arabidopsis thaliana BIK1 (AtBIK1) displays strong autophosphorylation kinase activity on tyrosine and threonine residues, whereas Brassica rapa BIK1 (BrBIK1) does not exhibit autophosphorylation kinase activity in vitro. Herein, we demonstrated that four proteins (RGP2, PATL2, PP7, and SULTR4.1) interact with BrBIK1 but not AtBIK1 in a yeast two-hybrid (Y2H) system. We subsequently employed bimolecular uorescence complementation (BiFC) to conrm interactions between BIK1 and candidates in Nicotiana benthamiana, and found that only BrBIK1 bound the three proteins tested. We selected three phosphosites, T90, T362, and T368, based on amino acid sequence alignment between AtBIK1 and BrBIK1, and performed site-directed mutagenesis (SDM) on AtBIK1 and BrBIK. S90T, P362T, and A368T mutations in BrBIK1 restored autophosphorylation kinase activity on threonine residues comparable with AtBIK1. However, T90A, T362P, and T368A mutations in AtBIK1 did not alter autophosphorylation kinase activity on threonine residues compared with wild-type AtBIK1. Interestingly, BiFC results showed that BIK1 mutations restored kinase activity but not binding to RGP2, PATL2, or PP7 proteins. Our results suggest that phospho-BIK1 might be involved in plant innate immunity, while non-phospho plant development through interactions with RGP2, PATL2, and marker, the Amp r marker for Escherichia coli selection, and the replication origin of the 2µ plasmid for replication in yeast. PBN204 contains ADE2, URA3, and lacZ genes as reporters based on different gal promoters. Yeast transformants were selected on synthetic dextrose (SD) minimal medium lacking leucine and tryptophan (SD-LW). Transformants were replica plated onto various SD selection media, including medium lacking leucine, tryptophan, and uracil (SD-LWU), and medium lacking leucine, tryptophan, and adenine (SD-LWA). Yeast transformants containing bait did not grow on SD selection media and they did not show beta-galactosidase activity in X-gal lter assays. This indicates that the BIK1 bait does not have transcriptional activator function, making it suitable for YTH selection in the PBN204 yeast strain. We also carried out BIK1 bait self-activity tests in the AH109 strain, which has a different genetic background, and the results were the same as those in PBN204. As a positive control, yeast were transformed with the Tag bait plasmid and the p53 prey plasmid; pGBKT7-53 encodes the Gal4 DNA-binding domain fused with murine p53, and pGADT7-Tag encodes the Gal4 AD fused with the SV40 large T-antigen. Negative controls were transformed with parental bait vector (pGBKT7) and prey vector (pGADT7).


Introduction
Plants must adapt to changing environmental factors in order to grow and develop. Receptor-like kinases (RLKs) are extremely important in transmembrane signaling-mediated regulation of plant growth, development, and adaptation to diverse environmental conditions, including pathogens, and they are believed to play critical roles in the perception and transmission of external signals 1 . Therefore, transcriptional and post-translational regulation of their activity is pivotal for receptor signaling in plants.
Recognition of microbe-associated molecular patterns (MAMPs) by pattern recognition receptors (PRRs) triggers the induction of various proteins such as FLS2 ( agellin) and Elongation Factor Tu (EFR), which is the rst line of inducible defense against invading pathogens 2 .
Among RLKs, Receptor-like Cytoplasmic Kinases (RLCKs) share homology with RLKs in their kinase domains, but lack a transmembrane domain, and they have emerged as a major class of signaling proteins that regulate plant cellular activities in response to biotic/abiotic stresses and endogenous extracellular signaling molecules as downstream components of RLKs 3 . Some functionally characterized RLCKs from plants play roles in development and stress responses. For example, 149 and 187 RLCK-encoding genes have been identi ed in Arabidopsis thaliana and Oryza sativa, respectively 4 . However, to date, little is known about the precise functions of RLCKs in higher plants.
As we know, the AtBIK1 gene is regulated by Botrytis cinerea infection, and inactivation of BIK1 causes severe susceptibility to necrotrophic fungal pathogens but enhances resistance to a virulent strain of the bacterial pathogen Pseudomonas syringae pv tomato 5 . Interestingly, AtBIK1 associates with the agellin receptor complex to initiate plant innate immunity 6 , and BIK1 is rapidly phosphorylated by g22 perception in an FLS2-and BAK1-dependent manner. As another aspect, BIK1 is monoubiquitinated following phosphorylation, then released from the FLS2-BAK1 complex, and internalized dynamically into endocytic compartments 7 .
BIK1 associates with multiple RLKs to regulate pathogen defense responses and brassinosteroid (BR) signaling. Since BR signaling regulates plant growth and development, BIK1 might be involved in both defense and growth through different post-translational modi cations such as phosphorylation status. BIK1 and ERECTA (ER) play opposing roles in leaf morphogenesis and in orescence architecture, and BIK1 is required to maintain appropriate auxin responses during leaf margin morphogenesis 8 . Based on genetic analysis, g22-induced PTI against Botrytis cinerea requires BIK1, EIN2, and HUB1, and BIK1mediated PTI against P. syringae is modulated by salicylic acid (SA), ethylene (ET), and jasmonate signaling 9 . An Asp residue (D 287 ) of BAK1 is crucial for its proteolytic cleavage, and plays an essential role in BAK1-regulated plant immunity, growth hormone BR-mediated responses, and cell death containment. BAK1 D287A mutation impairs BAK1 phosphorylation of its substrate BIK1, and its plasma membrane localization 10 .
As a signaling molecule, Pep1 is an endogenous elicitor that potentiates PAMP-inducible plant responses and association with PEPR1, and acts as a negative regulator in BR signaling through direct interaction with the BR receptor BRI1 cytosolic kinase domain 11 . The small peptide Pep1 trigger immunity through Perception of the Arabidopsis Danger Signal Peptide 1 and 2 (PEPR1 and PEPR2) in A. thaliana. Interestingly, PEPR1 speci cally interacts with BIK1 to mediate defenses in the presence of Pep1 and PEPR1, and PEPR2 phosphorylates BIK1 with Pep1 treatment in vitro and in vivo 12 . Generally, phosphorylation of proteins by protein kinases controls many cellular and physiological processes, including intracellular signal transduction. BIK1 also interacts with PBL1 in PTI defenses 12 and acts as a positive regulator of the PAMP g22-induced increase in cytosolic calcium.
Although much is known about BIK1, the relationships between BIK1 and downstream components related to plant growth, development, and defense signaling remain poorly understood in higher plants. In the present study, we found that AtBIK1 displayed strong autophosphorylation kinase activity on tyrosine and threonine residues, unlike BrBIK1. We employed yeast two-hybrid (Y2H) assays to identify downstream components and explore BIK1 downstream signaling interactors with AtBIK1 and BrBIK1 as bait. We identi ed four proteins: UDP-arabinopyranose mutase 2 (RGP2), which is essential for cell wall and pollen development 13 ; PATL2, a carrier protein that may be involved in membrane tra cking events associated with cell plate formation during cytokinesis 14 ; serine/threonine-protein phosphatase 7 (PP7), which acts as a positive regulator of cryptochrome signaling involved in hypocotyl growth inhibition and cotyledon expansion under white and blue light conditions 15 , control of stomatal aperture 16 , and thermotolerance in Arabidopsis 17 ; and SULTR4.1, a sulfate transporter that interacts with BrBIK1 but not AtBIK1, and displays strong autophosphorylation kinase activity. The results suggest that interactions between BIK1 and downstream components are dependent on the BIK1 phosphorylation status.

Material And Methods
Yeast Two-hybrid Assay To construct the BIK1 bait, we ampli ed full-length BIK1 (392 residues) using primers 5'-GC GAA TTC GGT TCT TGC TTC AGC TCT TG-3' (forward) and 5'-CG GGA TCC TTA ATG CTT CCC AAA AGG TCT T-3' (reverse). The BIK1 bait (1182 nt) was cloned into EcoRI/BamHI restriction sites of the pGBKT vector, which contains the DNA-binding domain of GAL4 (GAL4DB), the TRP1 yeast selection marker, the Kan r marker, and the replication origin sequence of the yeast 2µ plasmid. It contains a fully functional pADH1 promoter in front of the bait gene. The junction between the GAL4 DNA-binding domain and the BrBIK1 gene was con rmed by DNA sequencing. For Y2H screening, a Brassica rapa cDNA activation domain (AD) library and an A. thaliana cDNA library were used. The cDNA inserts were integrated into the pGADT7 vector in three different frames. The cDNA inserts were introduced into yeast strain PBN204 using a SmaI-linearized pGADT7-Rec vector in three different frames by yeast homologous recombination. For self-activity testing of the BIK1 bait, we used a Y2H system based on transcriptional activity of the reconstituted GAL4 transcriptional activator. Therefore, the bait by itself should not function as a transcriptional activator.
To test the transcriptional activity of the BIK1 bait, the bait plasmid was introduced alongside an empty prey vector into yeast strain PBN204. The negative control prey vector was pACT2, which contains a GAL4 transcriptional activation domain, the LEU2 yeast selection marker, the Amp r marker for Escherichia coli selection, and the replication origin of the 2µ plasmid for replication in yeast. PBN204 contains ADE2, URA3, and lacZ genes as reporters based on different gal promoters. Yeast transformants were selected on synthetic dextrose (SD) minimal medium lacking leucine and tryptophan (SD-LW). Transformants were replica plated onto various SD selection media, including medium lacking leucine, tryptophan, and uracil (SD-LWU), and medium lacking leucine, tryptophan, and adenine (SD-LWA). Yeast transformants containing bait did not grow on SD selection media and they did not show beta-galactosidase activity in X-gal lter assays. This indicates that the BIK1 bait does not have transcriptional activator function, making it suitable for YTH selection in the PBN204 yeast strain. We also carried out BIK1 bait self-activity tests in the AH109 strain, which has a different genetic background, and the results were the same as those in PBN204. As a positive control, yeast were transformed with the Tag bait plasmid and the p53 prey plasmid; pGBKT7-53 encodes the Gal4 DNA-binding domain fused with murine p53, and pGADT7-Tag encodes the Gal4 AD fused with the SV40 large T-antigen. Negative controls were transformed with parental bait vector (pGBKT7) and prey vector (pGADT7).

Site-directed Mutagenesis
The described GST-AtBIK1, Flag-AtBIK1, and Flag-BrBIK1 constructs were used as templates for sitedirected mutagenesis with a QuikChange XL Site-Directed Mutagenesis Kit (Stratagene). Six individual constructs were generated with the substitutions T90S, T362P, and T368A for AtBIK1, and S90T, P362T, and A368T for BrBIK1. All constructs were sequenced in both directions to verify speci c mutations and the absence of additional mutations.

Results And Discussion
Higher plants have evolved numerous RLKs and RLCKs that modulate many different biological processes, including innate immunity, growth and development, abiotic stresses, and stomatal aperture. Phosphorylation of the RLK/RLCK complex is an essential step in the initiation of diverse biological and physiological processes. Among RLCKs in Arabidopsis, BIK1 is a serine/threonine kinase that possesses tyrosine kinase activity, and mass spectrometry, immunoblot, and genetic analyses have shown that it is autophosphorylated at multiple Ser/Thr/Tyr residues. BAK1, a co-receptor of BRI1, is able to phosphorylate BIK1 at both tyrosine and serine/threonine residues 22 . BIK1 (Y150) blocks both tyrosine and serine/threonine kinase activity, whereas BIK1 (Y243 and Y250) are more speci cally involved in tyrosine phosphorylation 22 . In the present study, AtBIK1 showed strong autophosphorylation kinase activity on tyrosine and threonine residues, in contrast to BrBIK1, despite high amino acid similarity between these two proteins (Fig. 1). To identify downstream signaling components of BIK1 related to growth and defense signaling in A. thaliana, we used an Arabidopsis library for Y2H screening, and identi ed four proteins that interact with BrBIK1 but not AtBIK1 (Fig. 2A); UDP-arabinopyranose mutase 2 (RGP2), Patellin-2 (PATL2), serine/threonine-protein phosphatase 7 (PP7), and sulfate transporter 4.1 (SULTR4.1). RGP2 is essential for proper cell wall and pollen development 13 . PATL2 may be involved in membrane tra cking events associated with cell plate formation during cytokinesis 14 . PP7 acts as a positive regulator of cryptochrome signaling involved in hypocotyl growth inhibition, cotyledon expansion under white and blue light conditions 15 , and control of Arabidopsis stomatal aperture 16 . SULTR4.1 is an H + /sulfate cotransporter that may play a role in the regulation of sulfate assimilation. Y2H results suggested that BrBIK1 interactors might be involved in plant growth and development processes, including sulfate assimilation, which indicates that non-phosphorylated BIK1 might not be involved in defense mechanisms, including PTI.
We also employed BiFC experiments to con rm interactions between BIK1 and RGP2, PP7, and PATL2 in N. benthamiana leaves. As a positive control we used SOS2, which encodes a novel Ser/Thr protein kinase that also functions in salt tolerance in Arabidopsis 22 , and SOS3 that encodes a novel EF-hand Ca 2+ sensor 23 . We con rmed that AtBIK1 does not interact with RGP2, PP7, and PATL2 (Fig. 2C), whereas BrBIK1 strongly interacts with these proteins (Fig. 2B). The results indicate that phosphorylation of BIK1 is not necessary for RGP2, PP7, PATL2, and BrBIK1 interactions, suggesting that BIK1 phosphorylation is essential for plant defense signaling, but not necessary for plant growth and development signaling. Therefore, we performed amino acid sequence alignments and investigated known phosphorylation sites of AtBIK1 in A. thaliana. Protein phosphorylation is a key event in signal transduction pathways. When upstream signals are stimulated through receptor kinases, protein kinases are activated and phosphorylate their substrates on Ser/Thr and tyrosine residues, modulating their localization, conformation, and activity. In some cases, phosphorylated substrates become recognizable to other proteins, and such interactions transduce and propel the signal onward. Certain domains speci cally recognize phosphorylated residues of proteins, regulating cell growth and differentiation 24 . Analysis of recombinant RLCKs revealed distinct autophosphorylation activity at tyrosine and threonine residues in E.
coli 25 . Given that phosphorylation of RLCKs generally initiates and mediates signal transduction pathways, these two RLCKs might also be involved in sensing and responding to cellular needs in Arabidopsis 25 .
In the present work, we focused on the status of BIK1 phosphorylation, which may be essential for interactions between BIK1 and downstream components to regulate diverse physiological processes. Therefore, we cloned BrBIK1 and AtBIK1 into the pDEST15B and pFlag recombinant expression vectors. The results of western blotting analysis with anti-pThr and anti-pY antibodies showed that GST-BrBIK1 and Flag-BrBIK1 did not undergo autophosphorylation at threonine and tyrosine residues. However, in contrast with BrBIK1, AtBIK1 exhibited strong autophosphorylation kinase activity at threonine and tyrosine residues ( Fig. 3A and B). Thus, we performed site-directed mutagenesis (SDM) to manipulate autophosphorylation of AtBIK1 and BrBIK1 based on amino acid sequence alignment (Fig. 1), and generated three threonine phosphorylation site mutants in GST-AtBIK1 (T90S, T362P, and T368A). The three sites differ from those in BrBIK1. The results showed that GST-AtBIK1(T90S), GST-AtBIK1(T362P), GST-AtBIK1(T368A), and wild-type GST-AtBIK1 exhibited almost identical autophosphorylation kinase activities (Fig. 3B). We also used SDM to generate Flag-BrBIK1 mutants P362T and A368T. Interestingly, both mutations rescued autophosphorylation kinase activity at threonine residues to levels comparable with AtBIK1 in vitro (Fig. 3C). Thus, we tested protein interactions using BiFC experiments with BrBIK1 (P362T, A369T) and RGP2, PATL2, and PP7. Surprisingly, BrBIK1 (P362T, A369T) did not bind downstream components RGP2, PATL2, and PP7 ( Fig. 4A and B). These results suggest that the phosphorylation status of BIK1 may play an important role in regulating between growth and defense in higher plants. The ndings indicate that autophosphorylation of RLCKs such as BIK1 provides a convenient and powerful system for elucidating kinase speci city in vitro and in vivo.