|Home | About | Journals | Submit | Contact Us | Français|
Brassinosteroids (BRs) are important plant hormones that act synergistically with auxin to regulate a variety of plant developmental and physiological processes. In the past decade, genetic and biochemical studies have revealed a linear signaling pathway that relies on protein phosphorylation to transmit the BR signal into the nucleus, altering expression of hundreds of genes to promote plant growth. We conducted an activation-tagging based suppressor screen to look for Arabidopsis genes that, when overexpressed by inserted 35S enhancer elements, could suppress the dwarf phenotype of a weak BR receptor mutant bri1-301. This screen identified a total of six dominant activation-tagged bri1 suppressors (atbs-Ds). Using a plasmid rescue approach, we discovered that the bri1-301 suppression effect in four atbs-D mutants (atbs3-D to atbs6-D) was caused by overexpression of a YUCCA gene thought to be involved in tryptophan-dependent auxin biosynthesis. Interestingly, the three activation-tagged YUCCA genes belong to the YUCCA IIA subfamily that includes two other members out of 11 known Arabidopsis YUCCA genes. In addition, our molecular studies revealed a T-DNA insertion near a basic helix-loop-helix gene in atbs1-D and a T-DNA insertion in a region carrying a BR biosynthetic gene in atbs2-D. Further studies of these atbs-D mutants could lead to better understanding of the BR signaling process and the BR–auxin interaction.
Brassinosteroids (BRs) are a class of plant-specific polyhydroxylated steroids that play important roles in plant growth and development (Clouse and Sasse, 1998). Mutants defective in either BR biosynthesis or signaling share similar morphological/developmental phenotypes including dwarfed stature, round and dark green rosette leaves, short petioles, delayed flowering and senescence, reduced male fertility, and de-etiolation in the dark (Clouse and Feldmann, 1999). Unlike the animal steroid hormone receptors that function as nuclear transcription factors, the major BR receptor BRASSINOSTEROID INSENSITIVE 1 (BRI1) is a cell surface-localized leucine-rich-repeat receptor-like kinase (Kinoshita et al., 2005; Li and Chory, 1997) that initiates a phosphorylation-mediated signaling cascade to regulate stability, nuclear localization, and/or DNA binding activity of several plant-specific transcription factors (Li, 2005; Li and Jin, 2007). BR binding to BRI1 leads to rapid dissociation of BRI1 Kinase Inhibitor 1 (BKI1) (Wang and Chory, 2006) to allow heterotetramerization and transphosphorylation of BRI1 and its co-receptor BRI1-Associated Receptor Kinase 1 (BAK1) (Li et al., 2002; Nam and Li, 2002; Wang et al., 2008). The activation of two receptor kinases leads to activation of a family of BR SIGNALING KINASES (BSKs) (Tang et al., 2008) and a protein phosphatase bri1 SUPPRESSOR 1 (BSU1) (Mora-Garcia et al., 2004) and inhibition of a GSK3-like kinase, BRASSINOSTEROID INSENSITIVE 2 (BIN2) (Choe et al., 2002; Li and Nam, 2002), likely via proteasome-mediated protein degradation or tyrosine dephosphorylation (Kim et al., 2009; Peng et al., 2008). As a result, two BIN2 substrates bri1 EMS SUPPRESSOR 1 (BES1) (Yin et al., 2002; Zhao et al., 2002) and BRASSINAZOLE RESISTANT 1 (BZR1) (Wang et al., 2002) become dephosphorylated and more stable (He et al., 2002), accumulate in the nucleus (Gampala et al., 2007; Ryu et al., 2007), and bind to their corresponding regulatory sequences (He et al., 2005; Yin et al., 2005), to influence expression of many known BR-responsive genes (Vert et al., 2005), thus promoting cellular elongation and plant growth.
Many strategies have been employed in the past decade to identify the aforementioned BR signaling components and proteins that regulate BR biosynthesis and signaling. These strategies include forward genetic screens for BR-insensitive dwarf mutants, which has so far identified over 30 loss-of-function alleles of BRI1 (Bouquin et al., 2001; Clouse et al., 1996; Li and Chory, 1997; Noguchi et al., 1999; Tanaka et al., 2005; Xu et al., 2008) and eight gain-of-function alleles of BIN2 (Choe et al., 2002; Li and Nam, 2002; Perez-Perez et al., 2002), suppressor screens for mutations that revert the dwarf phenotype or resistant to a BR biosynthesis inhibitor (BAK1, BES1, BZR1, and BSU1) (Li et al., 2002; Mora-Garcia et al., 2004; Wang et al., 2002; Yin et al., 2002), and yeast two-hybrid screen coupled with reverse genetics (BAK1, BKI1, BES1, and BZR1) (Nam and Li, 2002, 2004; Wang and Chory, 2006; Zhao et al., 2002). Among them, the most successful approach is the suppressor screen using the activation-tagging approach (Weigel et al., 2000). The insertion of a T-DNA that carries four copies of an enhancer element derived from the cauliflower mosaic virus 35S promoter into an Arabidopsis genome could lead to significant overexpression of a T-DNA flanking gene, resulting in a desirable morphological/developmental/physiological phenotype and easy identification of the responsible gene. The activation-tagging-based suppressor screens were responsible for the identification of BAK1 (Li et al., 2002), BSU1 (Mora-Garcia et al., 2004), bri1 SUPPRESSOR 1 (BRS1, a secreted carboxypeptidase) (Li et al., 2001a), BRI1-LIKE 1 (BRL1, a BRI1 homolog specifically expressed in vascular tissue) (Zhou et al., 2004), and bri1-5 ENHANCED 1 (BEN1, a dihydroflavonol 4-reductase-like protein regulating the levels of active BR) (Yuan et al., 2007).
One requirement for a successful activation tagging-based suppressor screen is that the primary mutant should produce many seeds from which to select T1 transgenic plants. The majority of known bri1 mutants are male sterile, explaining why all the activation-tagging screens for potential BR signaling were carried out in bri1-5, a weak BR receptor mutant with bri1-5 retained in the endoplasmic reticulum (ER) (Hong et al., 2008). Interestingly, the other known weak allele of BRI1, bri1-9, is also caused by ER retention of a mutated BR receptor (Jin et al., 2009, 2007). Suppressor screens using ethylmethanesulfonate (EMS) mutagenized bri1-9 seeds led to identification of three identical mutations in BES1 (Zhao et al., 2002) and mutations in key components of the bri1-9 ER retention mechanism (Jin et al., 2009, 2007). The weakest, possibly the most interesting, known bri1 mutant so far is bri1-301 that carries a two-nucleotide change (GG-AT) in the BRI1 gene, causing conversion of Gly989 to Ile in the cytoplasmic kinase domain of BRI1 (Xu et al., 2008). Interestingly, while the Gly989Ile mutation completely inhibits the in vitro kinase activity of a BRI1 fusion kinase, bri1-301 can grow as tall and produce as many seeds as the wild-type control yet exhibits many characteristic phenotypes of known BR mutants, including a compact rosette with rounder leaves and short petioles (Xu et al., 2008). We expected that an activation-tagging-based suppressor screen in this mutant background might be able to uncover additional components/regulators of the BR signaling pathway.
It is well known that BRs act synergistically with another plant hormone auxin to promote plant growth and regulate plant development (Hardtke, 2007). Auxin is synthesized from tryptophan or its precursor via multiple biosynthetic routes (Chandler, 2009). One of the major breakthroughs in studying auxin biosynthesis was the discovery of the Arabidopsis YUCCA gene by activation tagging, which encodes a flavin monooxygenase-like protein catalyzing the hydroxylation of tryptamine, a key step in the tryptamine route (Zhao et al., 2001). The Arabidopsis genome encodes 11 YUCCA genes that can be grouped into three subfamilies: YUCCA I, YUCCA IIA, and YUCCA IIB (Cheng et al., 2006; Xia et al., 2009), and overexpression of YUCCA3, 4, 5, or 6 results in similar morphological phenotypes and elevated auxin levels (Kim et al., 2007; Marsch-Martinez et al., 2002; Woodward et al., 2005; Zhao et al., 2001). A previous study suggested that the auxin-overproduction yucca phenotype requires a functional BR receptor, since a yucca bri1 double mutant still exhibits a dwarf phenotype (Nemhauser et al., 2004). Interestingly, neither auxin nor BR mutant was identified by genetic screens for mutants defective in the biosynthesis or response of the other hormone.
Here, we report isolation of six dominant activation-tagged bri1 suppressors (atbs-Ds) from a collection of 25,000 activation-tagged transgenic bri1-301 lines. DNA blot analysis revealed that four of them carry a single T-DNA, with the other two containing two T-DNA insertions. Using a plasmid rescue approach, we recovered seven T-DNA/genome junction fragments and discovered that the bri1-301 suppression phenotype in four atbs-D mutants was caused by overproduction of one of the three closely related members of the YUCCA family. Our discoveries thus provided additional genetic support for the interaction between the plant steroid hormone and auxin. These atbs-D mutants will be excellent tools in studying the BR signaling process and the biochemical mechanisms of the BR/auxin interaction.
Among all studied alleles of the BR receptor gene BRI1, bri1-301 is one of the weakest alleles (Figure 1A). A recent study, using an E. coli-expressed fusion protein between glutathione-S-transferase and the cytoplasmic kinase domain of BRI1, showed that the bri1-301 mutation completely inactivates the kinase activity of the BR receptor in vitro (Xu et al., 2008). To test whether this was just an artifact of the E. coli-expressed fusion protein, we decided to use our previously established yeast expression system, which was used to show that activation of the full-length receptor kinases of BRI1 and its partner BAK1 requires their heterodimerization and transphosphorylation (Nam and Li, 2002). If one of the kinases was inactivated by single amino acid changes, no phosphorylation activity could be detected on either protein. As shown in Figure 1B, co-expression of wild-type BRI1 and BAK1 led to strong phosphorylation of the two full-length proteins. However, co-expression of the wild-type BAK1 with bri1-101 or bri1-301 resulted in no detectable phosphorylation on either receptor-like kinase. Our results thus confirmed that activation of BRI1 or BAK1 requires transphosphorylation by their partners and that the bri1-301 mutation inhibits the kinase activity of BRI1. Although it remains to be tested whether bri1-301 is a kinase-dead BR receptor in Arabidopsis, bri1-301 is certainly an interesting BR receptor mutant that can be used for an activation-tagging-based genetic screen to look for additional regulators of BR signaling.
We thus transformed a total of 10 trays (~150plants/tray) of bri1-301 mutants via a vacuum infiltration method with Agrobacterium strain GV3101 carrying the activation-tagging vector pSKI015 that contains the BASTA-resistance gene and the pUC19 plasmid sequence (Weigel et al., 2000). Seeds collected from infiltrated bri1-301 plants were directly sown into soil and 2–4-week-old growing seedlings were sprayed with BASTA to select herbicide-resistant lines. A total of ~25,000 BASTA-resistant bri1-301 plants were generated and screened for plants exhibiting wild-type-like rosette, leading to identification of six activation-tagged bri1 suppressor-Dominant (atbs-D) mutants (Figure 2A–2F).
To determine how many copies of the activation-tagging T-DNA were inserted into each atbs-D mutant, total genomic DNAs isolated from these six atbs-D mutants and bri1-301 were digested with either EcoRI or HindIII and hybridized with a 32P-labeled pUC19 plasmid-derived DNA probe in several DNA blot experiments. Because both restriction enzymes cut the inserted T-DNA once near the left border while the pUC19 plasmid-derived fragment is near the T-DNA right border (Figure 3A), the hybridizing DNA fragments detected in each mutant must contain flanking genomic DNA of the bri1-301. As shown in Figure 3B–3E, four atbs-D mutants, atbs1-D, 2-D, 5-D, and 6-D, contain a single hybridizing band, while both atbs3-D and atbs4-D contain at least two copies of the pUC19-carrying T-DNA inserts. Consistent with the DNA blot results, atbs3-D segregated out BASTA-resistant bri1-301-looking plants in its T2 generation. However, no BASTA-resistant bri1-301-like seedlings were found in the T2 generation of atbs4-D (data not shown), suggesting that one of the inserted T-DNAs might be truncated and lack the intact BASTA-resistance gene that is located near the left border, since integration of a T-DNA into plant chromosomes is a polar process, with the right border integrating first.
To understand the underlying biochemical mechanisms by which these atbs-D mutations suppress bri1-301, one has to know where the T-DNAs are inserted and which flanking gene(s) are overexpressed. Since each pUC19-hybridizing EcoRI or HindIII-digested fragment is smaller than 16kb in size (Figure 3B–3E) and carries the pUC19 plasmid DNA with a bacterial replication origin and an ampicillin resistance gene (Weigel et al., 2000), we decided to use the plasmid rescue approach (Weigel et al., 2000) to recover the T-DNA flanking genomic DNA fragments. Consistent with the DNA blot result, two flanking genomic DNA fragments were rescued from atbs3-D (Figure 4A and 4B). Sequencing analysis of a rescued ~6.3-kb EcoRI fragment revealed that one T-DNA was inserted ~270bp upstream of the predicted transcriptional initiation site of At5g55050 that encodes a GDSL-motif lipase/hydrolase-like protein (Figure 4A). Based on the orientation of the inserted T-DNA, the 1.4-kb-long 35S enhancer sequence was located ~80bp upstream of the predicted TATA-box of At5g55050 but was ~12kb away from the predicted initiation site of its upstream neighbor At5g55040 encoding a 145-amino-acid polypeptide of no known function. RNA blot analysis indicated that At5g55050 was overexpressed in the atbs3-D mutant (Figure 4C); however, PCR-based genotyping analysis showed that this insertion was not co-segregated with the atbs3-D phenotype (Figure 4D), since the T-DNA-specific junction fragment could be amplified from both atbs3-D-like and bri1-301-like T2 segregants of atbs3-D.
It is thus likely that the atbs3-D phenotype is caused by the other T-DNA insertion. Rescue and sequencing analysis of a ~8.0-kb HindIII fragment revealed that the second T-DNA was inserted ~3.5kb upstream of the predicted initiation site of At4g28720 and ~7.5kb downstream of the predicted 3′-end of At4g28710 (Figure 4B). The sequence data also indicated that the 35S enhancer was located ~3.5kb upstream of At4g28720 but was ~22.6kb downstream of the predicted initiation site of At4g28710. At4g28710 is annotated as the Arabidopsis MYOSIN XIH involved in actin-based movement (Avisar et al., 2009), while At4g28720 is known as a member (YUCCA8) of the Arabidopsis YUCCA family involved in tryptophan-dependent auxin biosynthesis (Cheng et al., 2006; Zhao et al., 2001). RNA blot analysis revealed that the YUCCA8 gene was overexpressed in the atbs3-D mutant (Figure 4E), explaining its narrow and epinastic leaves also observed with other known YUCCA-overexpressing mutants (Kim et al., 2007; Marsch-Martinez et al., 2002; Woodward et al., 2005).
The effect of YUCCA8 overexpression on bri1-301 was confirmed by discovery of the T-DNA insertion site in atbs4-D. Sequencing analysis of a rescued ~12.5-kb HindIII fragment from atbs4-D revealed the presence of a T-DNA insertion between At4g28710 and At4g28720 with the 35S enhancer sequence located ~8.5kb upstream and ~16.6kb downstream of the initiation sites of YUCCA8 and MYOSIN XIH genes, respectively (Figure 4B). Consistent with our hypothesis of the second T-DNA being truncated, the two ~6-kb pUC19-hybridizing EcoRI/HindIII fragments seen on the DNA blot (Figure 3D) were not recovered in our repeated plasmid-rescue experiments (likely caused by truncation of the pUC19 plasmid) and both the ~12.5-kb HindIII and ~13-kb EcoRI fragments were derived from the first T-DNA insertion. As expected, RNA blot analysis showed that the YUCCA8 gene was indeed overexpressed in atbs4-D (Figure 4F). The identification of two T-DNA insertions upstream of the YUCCA8 gene in two independent atbs-D mutants argued strongly for the causal relationship between YUCCA8 overexpression and the suppressed bri1-301 phenotype.
The bri1-301-rescuing effect of an overexpressed YUCCA gene was further confirmed by recovery of flanking genomic DNA fragments from two additional atbs-D mutants, atbs5-D and atbs6-D. As shown in Figure 5A and 5B, sequencing analysis of a rescued ~7.7-kb EcoRI fragment from atbs5-D revealed that the 35S enhancers were located ~3.3kb upstream and ~18.6kb downstream of the predicted transcription sites of At5g43890 and At5g43900 genes, respectively, while the sequence information of a rescued ~7.2-kb EcoRI fragment from atbs6-D predicted that the 35S enhancers were located ~5.5kb upstream and ~17kb downstream of initiation sites of At1g04610 and At1g04600 genes, respectively. Interestingly, the genes closer to the 35S sequence of the inserted T-DNAs encode close homologs of YUCCA8, YUCCA5 [also known as SUPER1 (Woodward et al., 2005)] in atbs5-D, and YUCCA3 in atbs6-D, while the other flanking genes far away from the 35S enhancers encode homologs of MYOSIN XIH, MYOSIN XI 6 in atbs5-D, and MYOSIN XIA in atbs6-D. Consistent with locations of the 35S enhancers, RNA blot analysis showed that YUCCA5 and YUCCA3 genes were overexpressed in atbs5-D and atbs6-D, respectively. Taken together, our study demonstrated that overexpression of a YUCCA gene was able to suppress the compact rosette phenotype of the weak BR receptor bri1-301 and provided additional genetic support for the well known interaction between BR and auxin. Our discoveries coupled with a previous finding that YUCCA overexpression was not able to suppress a severe bri1 dwarf mutant (Nemhauser et al., 2004) strongly support an earlier hypothesis that auxin response requires a functional BR signaling pathway.
It is interesting to note that the three YUCCA genes that were tagged in the four atbs-D mutants are clustered together with two other YUCCAs, YUCCA7 (At2g33230) and YUCCA9 (At1g04180), in phylogenetic analysis, forming the YUCCA IIA subfamily out of 11 Arabidopsis YUCCA genes that were previously classified into three groups: YUCCA I, YUCCA IIA, and YUCCA IIB (Cheng et al., 2006; Xia et al., 2009) (Figure 5E). Besides sequence similarity, the five YUCCA IIA genes all have a MYOSIN XI-type gene as their neighbors on chromosomes. One of the differences between the three tagged YUCCA IIA genes and two other YUCCA IIA genes is the presence of additional genes between YUCCA and MYOSIN: At2g33233 between YUCCA7 and MYOSIN XID and At1g04170 plus At1g04171 between YUCCA9 and MYOSIN XIB. It is thus quite possible that an insertion of the activation-tagging T-DNA could produce other morphological phenotypes that would prevent their discovery as activation-tagged bri1 suppressors. This likely explanation might also be responsible for our failure to tag other YUCCA genes in our genetic screen, although the activation-tagged YUCCA6 (also known as HYPERTALL1) was shown to produce long, narrow leaves with elongated petiole (Kim et al., 2007)—a phenotype often associated with enhanced BR signaling (Nam and Li, 2002; Yan et al., 2009).
Our plasmid rescue experiments also identified the T-DNA insertions in both atbs1-D and atbs2-D mutants. As shown in Figure 6A, the 35S enhancer-carrying T-DNA was inserted 5.9 and 1.5kb upstream of At1g74500 and At1g74510, respectively, with four copies of the 35S enhancer pointing towards the predicted start codon of At1g74500. RNA blot analysis using probes derived from the two flanking genes indicated that At1g74500 was overexpressed in atbs1-D (Figure 6C and 6D), suggesting that the suppressor phenotype is likely caused by At1g74500 overexpression. Interestingly, At1g74500 encodes a 93-amino-acid basic helix-loop-helix (bHLH) protein that is similar in sequence to the rice BRASSINOSTEROID UPREGULATED 1 that was recently implicated in BR signaling (Tanaka et al., 2009). Further experiments are needed to confirm whether At1g74500 is indeed the ATBS1 (activation-tagged bri1 suppressor 1) gene and, if so, to investigate how this small bHLH protein regulates BR signaling in Arabidopsis.
Figure 6B shows that the T-DNA in atbs2-D was inserted into a region on chromosome 2 that contains DET2 encoding a steroid 5α-reductase catalyzing a critical step of BR biosynthesis (Li et al., 1997, 1996). However, the clustered 35S enhancers are located very close to the TATA-box of the annotated At2g38070 gene encoding a member of the DUF740 family, but is ~11.8kb away from the DET2 transcriptional initiation site. Besides, there is another coding sequence between the T-DNA and DET2. Therefore, it is unlikely that the atbs2-D phenotype was caused by DET2 overexpression. Indeed, RNA blot analysis indicated that the DET2 expression level was not affected by the T-DNA insertion (Figure 6E). Further experiments are underway to investigate the underlying mechanism by which the atbs2-D mutation suppresses the weak BR receptor mutant.
Arabidopsis thaliana ecotype Columbia (Col-0) was used as the wild-type control for phenotypic comparison, while bri1-301 (Xu et al., 2008) was used as the starting genetic materials for the activation-tagging suppressor screen. Seed germination and plant growth conditions were described previously (Li et al., 2001b).
The activation-tagging plasmid pSKI015 was previously described (Weigel et al., 2000) and transformed into GV3101 Agrobacterium tumefaciens cells. Due to instability of the clustered 35S enhancers in Agrobacterial cells when kept at 4°C, the pSKI015 plasmid was freshly transformed into competent GV3101 cells for each plant transformation experiment. Agrobacterial cells of stationary phase were collected by centrifugation at 5,000g at 4°C, re-suspended in 3vol. of liquid half-strength Murashige and Skoog (MS) medium containing 2% sucrose, 10μgL−1 benzylaminopurine, and 0.15% (v/v) Silwet L-77 (Lehle Seeds, Round Rock, TX), and used to transform 6–7-week-old flowering bri1-301 mutants via the vacuum infiltration method (Bechtold and Pelletier, 1998). Seeds collected from infiltrated bri1-301 plants were sterilized as previously described (Li et al., 2001b), mixed with 0.08% (w/v) Phytagar (Invitrogen), stored at 4°C for 48h, and sown directly into soil. Transgenic bri1-301 mutants were selected by spraying Finale (AgrEvo, Montvale, NJ) at 1:1000 dilution once every 3d for half a month and visually inspected for plants exhibiting wild-type-like morphology. A potential atbs-D mutant was PCR-genotyped to eliminate any pollen or seed contaminant using the following CAPS primers: 5′-CTGGCGATAGAACTGCTAAC-3’ and 5′-GCTGTTTCACCCATCCAAC-3’ (the restriction enzyme DpnII cuts the PCR fragment from wild-type twice but cuts that from bri1-301 only once).
Approximately 1μg total genomic DNA was digested overnight at 37°C in a 40-μL reaction with EcoRI or HindIII. After denaturing enzymes by 10min heating at 65°C and phenol/chloroform extraction, the digested DNAs were ethanol precipitated and ligated overnight in a 200-μL reaction volume with 80U T4 DNA ligase (New England Biolabs) at 16°C. Ligated DNAs were extracted with phenol/chloroform, precipitated with ethanol, re-suspended in water, and transformed into E. coli competent cells by electroporation with a MicroPulser Electroporator (BIO-RAD) following the manufacturer's suggested protocol. Electroporated cells were selected on ampicillin-containing medium; and the rescued plasmids were purified using the Qiagen's Plasmid Mini Kit, sequenced at the University of Michigan's sequencing core facility, and compared with the published Arabidopsis genome to determine T-DNA/genome junctions.
Total genomic DNAs were isolated either by the Qiagen's DNeasy Plant Mini Kit according to the manufacturer's recommended protocol (for DNA blot analysis and plasmid rescue) or by a previously described ‘miniprep’ method (Li and Chory, 1998) for genotyping analysis. Approximately 0.5-μg genomic DNAs of bri1-301 or an atbs-D mutant was digested with EcoRI or HindIII overnight at 37°C and separated by 0.8% agarose gel. A previously described protocol (Lincoln et al., 1990) was used to transfer DNAs to Hybond-N membrane (Pharmacia) and hybridized with a 32P-labed probe derived from the pUC19 plasmid. Total RNAs were isolated from shoots of 4-week-old soil-grown plants using the Qiagen's RNeasy Plant Mini Kit according to the manufacturer's recommended procedure and analyzed by Northern blot following a previously described protocol (Li et al., 2001b) and a 32P-labeled DNA probe amplified from a gene of interest.
The plasmids pYES2-BRI1:HA and pESC-BAK1:MYC and detailed methods for expressing the full-length BRI1 and its co-receptor BAK1 in yeast cells, immunoprecipitating the BRI1/BAK1 complex, extraction of total yeast proteins, and immunoblot analysis of BRI1 or BAK1 were previously described (Nam and Li, 2002). PCR-based site-directed mutagenesis was conducted to introduce bri1-101 or bri1-301 mutation into pYES2-BRI1:HA using the QuickChange XL Site-Directed Mutagenesis Kit (Stratagene). The in vitro phosphorylation assay using immunoprecipitated BRI1/BAK1 proteins was performed as previously described (Nam and Li, 2002).
Sequence data from this article can be found in the EMBL/GenBank data libraries under the following accession numbers: NP_194980 (YUCCA1, At4g32540), NP_193062 (YUCCA2, At4g13260), NP_171955 (YUCCA3, At1g04610), NP_850808 (YUCCA4, At5g11320), NP_199202 (YUCCA5, At5g43890), NP_197944 (YUCCA6, At5g25620), NP_180881 (YUCCA7, At2g33230), NP_194601 (YUCCA8, At4g28720), NP_171914 (YUCCA9, At1g04180), NP_175321 (YUCCA10, At1g48910), and NP_173564 (YUCCA11, At1g21430).
This work was supported by a grant from the National Institutes of Health (GM060519 to J.L.).
We would like to thank members of the Li lab for helpful discussion. No conflict of interest declared.