The Arabidopsis BAP1 gene encodes a small protein with a C2-like domain. Here we show that the BAP1 protein is capable of binding to phospholipids in a calcium-dependent manner and is associated with membranes in vivo. We identify multiple roles of BAP1 in negatively regulating defense responses and cell death in Arabidopsis thaliana. The loss of BAP1 function confers an enhanced disease resistance to virulent bacterial and oomycete pathogens. The enhanced resistance is mediated by salicylic acid, PAD4 and a disease resistance gene SNC1. BAP1 is also involved in the control of cell death, which is suggested by an altered hypersensitive response to an avirulent bacterial pathogen in the bap1 loss-of-function mutant. BAP1 overexpression leads to an enhanced susceptibility to a virulent oomycete, suggesting a role for BAP1 in basal defense response. Furthermore, the BAP1 protein probably functions together with an evolutionarily conserved C2 domain protein BON1/CPN1 to negatively regulate defense responses in plants.
Plants utilize two forms of defense to detect and ward off pathogens. One defense is to use disease resistance (R) genes to specifically recognize pathogen avirulence (Avr) genes in an allele-specific manner (Flor, 1971). The specific interaction between R and its cognate Avr induces the hypersensitive response (HR), characterized by rapid calcium and ion fluxes, an extracellular oxidative burst and transcriptional reprogramming (Scheel, 1998). The resulting programmed cell death (PCD) effectively controls the spread of the pathogen. Additionally, this R-mediated defense induces systemic acquired resistance (SAR) in the rest of the plant, conferring resistance to a broad spectrum of pathogens (Ryals et al., 1996). The other form of defense is the basal defense response that provides plants with protection against various pathogens. Without it, plants are more susceptible to invasions of virulent pathogens. This response is triggered by general pathogen-derived signals named pathogen-associated molecular patterns (PAMPs; Nurnberger et al., 2004). These two forms of defense are genetically related. The R-mediated response is probably an accelerated and amplified version of the basal defense response, and is thus more effective in limiting the spread of pathogen growth (Glazebrook, 2001).
A number of R genes have been molecularly identified, and the majority of them encode proteins containing nucleotide binding sites (NBS) and leucine rich repeats (LRRs; Martin et al., 2003). There are about 150 NBS–LRR proteins in Arabidopsis thaliana and they are of either the coiled-coil (CC) type or the Toll/interleukin-1-receptor (TIR) type (Meyers et al., 2003). Genetic studies have identified genes required for R gene signaling (Dangl and Jones, 2001; Glazebrook, 2001). EDS1 (ENHANCED DISEASE SUSCEPTIBILITY 1) and PAD4 (PHYTOALEXIN DEFICIENT 4) are required for the function of the TIR–NBS–LRR proteins while NDR1 (NONRACE-SPECIFIC DISEASE RESISTANCE 1) is required for the CC–NBS–LRR proteins, although there are exceptions (Wiermer et al., 2005). The activation of cell death during a HR is under tight control, as indicated by the isolation of a number of lesion mimic mutants resembling cell death in a HR (Lorrain et al., 2003). Some of the mutants have misregulation of the initiation of cell death and form small, localized, necrotic spots. Others are unable to control the rate and extent of the lesions and form chlorosis in a large area. EDS1, PAD4 and NDR1 are implicated in the amplification of cell death, and this function appears to be independent of their roles in R-gene mediated defense responses (Clarke et al., 2001; Rusterucci et al., 2001).
The Arabidopsis BON1 (BONZAI1)/CPN1 gene is a regulator of defense responses apparently through repressing activity of an R gene. bon1 were isolated as mutants with a temperature-dependent growth defect and an enhanced disease resistance phenotype (Hua et al., 2001; Jambunathan and McNellis, 2003; Jambunathan et al., 2001; Liu et al., 2005). Study of a natural modifier of bon1 revealed that BON1 acts as a repressor of a haplotype-specific R gene SNC1 (SUPPRESSOR OF NPR1 CONSTITUTIVE 1)/BAL present in the Columbia (Col) accession (Yang and Hua, 2004). SNC1 encodes a TIR–NBS–LRR type of R protein and its activation induces constitutive defense responses (Li et al., 2001; Stokes et al., 2002). In bon1-1, the derepression of SNC1 leads to a constitutive activation of defense responses and consequent growth defects. BON1 belongs to an evolutionarily conserved copine gene family found in protozoa, plants, nematodes and mammals (Creutz et al., 1998). The copine proteins are characterized by two C2 domains at the N-terminus and an A domain at the C-terminus. C2 domains possess calcium-dependent phospholipid-binding activities and could confer calcium or lipid regulation on the proteins in which they reside (Rizo and Sudhof, 1998). The A domain refers to the von Willebrand A domain found in the extracellular ligand-binding portion of integrins; A domains are likely to be involved in protein–protein interactions (Whittaker and Hynes, 2002). Roles for copines in membrane trafficking and signal transduction have been suggested based on their structures and biochemical activities. Genetic and biochemical analysis have implicated their functions in signaling of a cation channel (Church and Lambie, 2003), a tumor necrosis factor-alpha receptor (Tomsig et al., 2004) and a nicotinic receptor (Gottschalk et al., 2005). The exact modes of action of these copine proteins and the specific biological processes they are involved in have yet to be revealed.
The identification of BON1 as a negative regulator of the defense response through an R gene provides a great entry point for further study of the molecular mechanism of regulation of the defense response. BON1 ASSOCIATED PROTEIN 1 (BAP1) was subsequently isolated as a BON1 interacting protein from a yeast two-hybrid screen (Hua et al., 2001). Here, we present a genetic and biochemical study suggesting that BAP1 is a functional partner of BON1 and that BON1 and BAP1 represent a new complex and pathway in negatively regulating both basal and R-mediated defenses in plants.
BAP1 is a small lipid-binding protein
The BAP1 protein is a small protein of 192 amino acids with one predicted C2 domain at the N-terminus. The C2 domain is characterized by a compact beta-sandwich composed of two four-stranded beta-sheets as well as three loops at the top of the domain and four loops at the bottom (Rizo and Sudhof, 1998). Calcium ion binding occurs at the top three loops involving four to five aspartates, and calcium binding may facilitate lipid binding through electrostatic changes (Rizo and Sudhof, 1998). Sequence analysis of the predicted C2 domain in BAP1 using the Pfam protein families database (Bateman et al., 2004) revealed that the BAP1 C2 domain has only one of the five aspartates defined as essential for calcium binding (Figure 1a). To determine whether the C2-like domain in BAP1 behaves like a canonical C2, we tested whether BAP1 could bind to phospholipids in a calcium-dependent manner. A maltose-binding protein (MBP):BAP1 fusion protein was expressed in Escherichia coli and the purified protein was assayed for its interaction with phosphatydylserine (PS). The MBP:BAP1 protein was incubated in a binding buffer with or without PS before it was centrifuged. MBP:BAP1, but not MBP, was found in the pellet with PS while no protein was found in the pellet without PS. Furthermore, an increased amount of MBP:BAP1 was precipitated when calcium ions were present in the binding buffer than in their absence (Figure 1b). This demonstrates that BAP1, despite lacking most of the conserved aspartates, is capable of binding to phospholipids in vitro and this lipid-binding ability is enhanced by the presence of calcium ions.
The BAP1 protein is membrane associated
To determine the subcellular localization of the BAP1 protein in Arabidopsis, we fused the BAP1 protein with several small epitope tags either at the N-terminus or at the C-terminus. However, we failed to detect by Western blot analysis any epitope-tagged BAP1 proteins expressed under the control of either its native promoter or the strong 35S promoter of cauliflower mosaic virus (CaMV) in transgenic Arabidopsis plants. Some of these epitope-tagged BAP1 genes when transformed into the bap1 loss-of-function mutant (see below) were capable of rescuing the mutant phenotype, indicating that these fusion proteins are functional. Thus the level of BAP1 protein is likely to be very low in Arabidopsis.
The extremely low abundance of the BAP1 protein but the moderate level of BAP1 transcript (Hua et al., 2001) suggests that the BAP1 protein could be rapidly turned over in Arabidopsis. We therefore transiently expressed a MYC-tagged BAP1 gene under the 35S promoter (Pro35S::MYC:BAP1) in Nicotiana benthamiana. The MYC:BAP1 gene was delivered into leaves of N. benthamiana through Agrobacterium infiltration, and proteins were extracted from infiltrated leaves 2 days later. The total protein extracts were separated into membrane and soluble fractions by centrifugation. We used the BON1:HA protein as a marker for membrane protein in this fractionation because BON1:HA was shown to be localized to the plasma membrane in Arabidopsis (Hua et al., 2001). BON1:HA expressed under the 35S promoter, when co-delivered with MYC:BAP1, was detected in the membrane fraction but not in the soluble fraction (Figure 1c). The MYC:BAP1 protein was detected in both the soluble and membrane fractions by Western blot analysis using the anti-MYC antibody (Figure 1c). These data suggest that BAP1 could be distributed between both membranes and cytosols when overexpressed in N. bethamiana.
We later detected the BAP1 protein in A. thaliana when we fused it with a large reporter protein, GUS (beta-glucuronidase). The BAP1:GUS chimeric gene under the control of the BAP1 promoter (ProBAP1::BAP1:GUS) was capable of rescuing the loss-of-function bap1 mutant (Figure S1), indicating that this is a functional fusion gene. Total proteins were isolated and separated into membrane and soluble fractions from wild-type Col plants carrying this transgene. The activities of GUS were assayed by adding the GUS substrate X-Gluc to the protein extracts. Positive GUS activities, indicated by the blue color, were detected in the membrane fraction but not in the soluble fraction of all three independent lines tested (a representative line is shown in Figure 1d). In contrast, transgenic plants with the GUS gene under the control of the BAP1 promoter (ProBAP1::GUS) exhibited GUS activity only in the soluble fraction in a similar way to transgenics with the GUS gene over-expressed under the 35S promoter (Pro35S::GUS; Figure 1d). The localization of BAP1:GUS to the membrane fraction indicates that BAP1 is associated with membranes in Arabidopsis.
The BAP1 protein interacts with the BON1 protein
BAP1 was isolated as the strongest BON1-interacting protein from a yeast two-hybrid screen (Hua et al., 2001). In a separate yeast two-hybrid screen, we identified BON1 as a protein interacting with BAP1. In this GAL4-based system, the full-length BAP1 was used as a bait to screen several Arabidopsis cDNA libraries. Only one positive clone was isolated and this clone turned out to encode the C-terminal portion (starting from amino acid 308) of the BON1 protein consisting mostly of the A domain. Thus, at least in the yeast two-hybrid systems, BON1 is the strongest interactor with BAP1 and vice versa.
To further verify the physical interaction between BON1 and BAP1 proteins, we carried out an in vitro pull-down experiment. The A domain of BON1 was tagged with a MYC epitope, and the resulting BON1A:MYC fusion was expressed in a rabbit reticulocyte lysate system by in vitro transcription coupled with translation. BON1A:MYC was incubated with either MBP:BAP1 or MBP. BON1A:MYC was co-purified with MBP:BAP1 but not with MBP alone (Figure 2), indicating that BON1 and BAP1 proteins interact with each other in vitro.
The BAP1 loss-of-function mutants have an enhanced disease resistance
To determine the biological function of BAP1, we screened for BAP1 loss-of-function mutants and obtained two T-DNA insertion lines which we named as bap1-1 and bap1-3 (Figure 3a). Both are loss-of-function mutants because the T-DNAs were located in sequences coding for Ile74 and Gln10 of the BAP1 protein, respectively. Furthermore, no wild-type transcripts were detected in the mutants by RNA blot analysis (Figure 3a and data not shown). We used bap1-1 for further analysis because bap1-1 and bap1-3 had very similar phenotypes.
bap1-1 exhibited an enhanced disease resistance to virulent pathogens of oomycetes and bacteria when compared with wild-type Col plants. We grew both the wild type and the bap1-1 mutant under the day/night cycle of 12-h light and 12-h darkness where no morphological difference was observed between them (see below). One week after the seedlings were inoculated with virulent Hyalopernonspora parasitica Noco2 strain, we counted sporangiophores that had formed on leaves. Over 50% of the wild-type leaves had >30 sporangiophores/leaf, while none of the bap1-1 leaves (like those of bon1-1) had more than five sporangiophores/leaf (Figure 3b). bap1-1 also exhibited enhanced resistance to the virulent bacterial pathogen Pseudomonas syringae pv. tomato (Pst) DC3000 compared with the wild-type Col. Three days after dip inoculation, Pst DC3000 grew to approximately 106 colony-forming units (CFU) mg−1 fresh weight in Col while its growth was only about 2 × 105 CFU mg−1 fresh weight in bap1-1 (Figure 3c). The reduced growth of DC3000 was consistently seen in bap1-1, though it was not reduced as much as in bon1-1 (Figure 3c).
The enhanced disease resistance to bacterial and oomycete pathogens suggests an upregulation of defense responses in bap1-1. Consistent with this idea, defense-related genes were upregulated in bap1-1. Pathogenesis Related 1 (PR1), a molecular marker for defense response, was not detected in the Col wild type under normal growth conditions, but was highly expressed in bap1-1 (Figure 3d). Similarly to bon1-1, SNC1 expression was also upregulated in bap1-1 (Figure 3d). The transcript levels of PR1 and SNC1 were lower in bap1-1 compared with bon1-1, suggesting a weaker activation of defense in bap1 than in bon1.
We found that the heightened defense response in bap1-1 was mediated by salicylic acid (SA). When the NahG gene coding for a SA degradation enzyme was introduced into bap1-1, the enhanced resistance to both oomycete and bacterium was greatly compromised. In contrast to almost no growth of the Noco2 strain of H. parasitica in bap1-1, ample growth was found on bap1-1NahG, with more than 60% of leaves having more than 30 sporangiophores/leaf (Figure 3b). Similarly, resistance to Pst DC3000 was compromised by NahG in bap1-1. Three days after inoculation, Pst DC3000 grew to 2 × 105 CFU mg−1 fresh weight in bap1-1 whereas its growth was amplified to 3 × 107 CFU mg−1 fresh weight in bap1-1NahG (Figure 3c). Consistently, PR1 upregulation in bap1-1 was eliminated by NahG (Figure 3d). Thus bap1-1 appears to have constitutive defense responses mediated by SA and consequently has a broad resistance to virulent pathogens.
SNC1 mediates the enhanced disease resistance in the bap1 mutant
Because the regulation of defense responses by BON1 is mediated by the R gene SNC1, we asked whether the mutant phenotype of BAP1 is mediated through SNC1 as well. A double mutant between bap1-1 and the loss-of-function mutant snc1-11 (Yang and Hua, 2004) was constructed. We found that snc1-11 suppressed the resistance to H. parasitica Noco2 in bap1-1. Approximately 40% of the leaves had >30 sporangiophores/leaf in bap1-1snc1-11 compared with 0% in bap1-1 (Figure 3b). The resistance to Pst DC3000 was also reduced in bap1-1snc1-11. Pst DC3000 grew to about 3.5 × 106 CFU mg−1 fresh weight in bap1-1snc1-11 3 days after inoculation, which was at a comparable level to the growth in wild-type Col (Figure 3c). In addition, the upregulation of PR1 expression in bap1-1 was greatly reduced by snc1-11 (Figure 3d). Thus, the SNC1 activity is required for the enhanced disease resistance in bap1-1.
bap1-1 has a growth defect resulting from an activated defense response
bap1-1 exhibits a morphological defect similar to but weaker than that of bon1-1 (Figure 4a). The leaves of bap1-1 were slightly smaller and curlier than those of the wild type. To verify that the mutant phenotype observed in bap1-1 was due to the loss of BAP1 function, we introduced a genomic fragment of the wild-type BAP1 into bap1-1. This 4.9 kb fragment contains 2.6 kb of the promoter region, the coding region and 1.7 kb after the translation stop site. Seven out of the eight transgenic lines generated had a wild-type morphology (Figure 4b), confirming that the loss of BAP1 function is the cause of the mutant phenotype.
The growth phenotype in bap1-1 appears to be a consequence of activated defense responses. When the defense response was blocked by NahG or snc1-11 in bap1-1, the morphological defects were inhibited. Both bap1-1NahG and bap1-1snc1-11 were wild type in appearance (Figure 4c), indicating that the growth defect in bap1-1 was due to an activation of defense responses mediated by SA and SNC1.
The bap1-1 phenotype is fully suppressed by pad4 and partially suppressed by npr1
To further characterize the defense response phenotype in bap1, we generated double mutants between bap1 and defense response mutants pad4 and npr1 (nonexpressers of PR genes 1). pad4-1 suppressed both the growth and resistance phenotypes of the bap1-1 mutant. bap1-1pad4-1 had the appearance of the wild type throughout all developmental stages (Figure 4d). When challenged with the virulent pathogen Pst DC3000, the double mutant behaves like the pad4 single mutant (Figure 4e). Thus, pad4 completely suppressed the enhanced disease resistance as well as the growth defect of bap1-1.
npr1-1 did not rescue the dwarf phenotype of bap1-1. The bap1-1npr1-1 double mutant exhibited reduced growth similar to bap1-1 (Figure 4d) and was slightly yellowish compared with bap1-1. Intriguingly, the enhanced disease resistance in bap1-1 is largely suppressed by npr1-1 (Figure 4e). Pst DC3000 grew in bap1-1npr1-1 by nearly as much as in npr1-1, and much more than in bap1-1. Thus, the enhanced disease resistance but not the growth inhibition in bap1-1 is mediated by NPR1 and the effect of bap1 on regulation of growth and defense can be uncoupled.
The bap1-1 growth phenotype is modified by environmental conditions
We found that the bap1 mutant phenotype is modulated by the environment. Like bon1-1, bap1-1 showed a temperature-dependent growth defect. Under constant light, it had a reduced stature and slightly twisted leaves at 22°C, but was essentially like the wild type at the higher temperature of 28°C (Figure 5). The growth phenotype of bap1-1 was also influenced by light. In contrast to constant light, bap1 grown under a condition of 12 h of light and 12 h of dark did not exhibit any growth defects at 22°C (Figure 5). The severity of the bap1 mutant phenotype varied even under controlled light, humidity and temperature conditions, suggesting that other environmental factors may modulate its phenotype as well.
The bap1-1 mutant has an altered hypersensitive response
To determine whether bap1 mutations affect PCD in plants, we assayed HR to avirulent pathogens in bap1-1. The wild-type Col and the bap1-1 mutant plants were inoculated with a high concentration of Pst DC3000 carrying avrRpm1, avrRpt2 or an empty vector, respectively. Wild-type Col plants underwent a HR in response to Pst DC3000 carrying avrRpm1 or avrRpt2, indicated by a collapse of tissues after infiltration. No HR was observed in Col or bap1-1 infiltrated with Pst DC3000 containing an empty vector (data not shown). An accelerated HR was observed in bap1-1 compared with Col for Pst DC3000 expressing AvrRpt2 (Figure 6a). Thirty per cent of the Col leaves showed HR at 12 h post-inoculation (h.p.i.), while 30% of the bap1-1 leaves had HR at 8 h.p.i. (Figure 6a). A similar effect was observed in bon1-4 (cpn1-1) (Jambunathan et al., 2001) and bon1-1 (Figure 6a). However, the onset of HR to Pst DC3000 with AvrRpm1 was similar in bap1-1 and bon1-1 to that in Col (Figure 6b).
To determine whether the alteration in HR to AvrRpt2 is dependent on the function of SNC1, we assayed HR to Pst DC3000 avrRpt2 in bap1-1snc1-11. bap1-1snc1-11 had an accelerated HR similar to that of bap1-1, while snc1-11 alone did not alter HR (Figure 6c). These data suggest that the accelerated HR phenotype in bap1-1 is not mediated by SNC1. NahG inhibits the early HR phenotype in bap1-1 but it does not abolish HR in bap1-1 as it does in the wild type, suggesting that the bap1 mutation perturbs HR through both SA-dependent and SA-independent pathways.
Plants overexpressing BAP1 are more susceptible to a virulent pathogen
We asked whether BAP1 overexpression would confer more susceptibility to wild-type plants. The BAP1 gene was expressed under the strong 35S promoter of CaMV (Pro35S::BAP1) in wild-type Arabidopsis, and transgenic lines with a higher BAP1 transcript level compared with wild-type Col were selected by RNA blot analysis (Figure 7a). Three such BAP1 overexpressing lines (35S::BAP1-1, -2 and -4) were assayed for disease resistance to the virulent pathogen H. parasitica Noco2. We found that all three lines were more susceptible to Noco2. A week after infection, about 50% of leaves had >30 sporangiophores/leaf in wild type, while the percentage increased to 63%, 81% and 81%, respectively, for each of the three BAP1 overexpression lines (Figure 7b). This increase was consistently seen in three biological repeats and is therefore significant (Figure S2). Thus, BAP1 appears to be a negative regulator of basal defense response. This role of BAP1 does not involve SNC1, because the snc1-11 mutant is not more susceptible to Noco2 than the wild type (Figure 3b).
BAP1 expression is responsive to many stimuli
The BAP1 transcript was shown to be induced by a decrease in temperature (Hua et al., 2001). Expression analysis using a link of the Genevestigator Response Viewer (Zimmermann et al., 2004) from the TAIR site (http://arabidopsis.org/) revealed that the expression of BAP1 at the RNA level is modulated by a variety of stimuli. It was induced by biotic stresses from Botrytis cinerea, Festuca occidentalis, Phytophthora infestans and Pseudomonas syringae, chemical stresses from chitin, cycloheximide, furyl acrylate ester, ozone and syringolin, potassium nutrient deficiency, senescence, cold and salt. The expression of BAP1 was not significantly altered by treatment with light or plant hormones.
We further analyzed the expression of BAP1 during pathogen infection in more detail. When plants were infiltrated with Pst DC3000 (resuspended in 10 mm of MgCl2), BAP1 was induced at 1 h.p.i. and stayed on until 24 h.p.i. (Figure 8a). There was a consistently lower expression of BAP1 around 8 h.p.i. and it was correlated with the timing of reduction of BAP1 expression in MgCl2 infiltration. Infiltrating with MgCl2 increased the BAP1 transcript level from 1 h.p.i. to 6 h.p.i. (Figure 8a). Even infiltrating leaves with water induced a brief upregulation of BAP1 at around 1 h.p.i. (Figure 8b). It thus appears that the earlier induction of BAP1 by bacterial infiltration was due to the infiltration process, and the later induction was due to a response to pathogen.
Other physical stimuli also caused a rapid but transient induction of BAP1. For instance, the BAP1 transcript level went up 30 min after the leaves were cut by scissors and came down by 6 h (Figure 8c). The BAP1 transcript could be induced by SA. One day after being sprayed with 2 mm of SA, the plants had an increased BAP1 expression (Figure 8d).
Temperature modulates the expression of BAP1 as well. The BAP1 transcript was barely detectable at 28°C, while it had a moderate level at 22°C (Hua et al., 2001). A downshift from 28°C to 22°C upregulated the BAP1 transcript as early as 3 h and caused a maximum induction by 12 h (Figure 8e). Conversely, an upshift from 22°C to 28°C downregulated the BAP1 transcript level, and by 6 h no transcripts could be easily detected (Figure 8e). Furthermore, a 4°C treatment greatly induced the BAP1 transcript level (Figure 8e). Thus BAP1 is modulated by various biotic and abiotic stimuli.
Here, we have identified the BAP1 gene as a negative regulator of defense responses and as a potential functional partner of BON1/CPN1. BON1 is a member of the copine family which is evolutionarily conserved across kingdoms. The exact mode of function of these copine proteins is not well understood, but they are probably involved in signaling and membrane trafficking. In Arabidopsis, BON1 appears to be a negative regulator of defense response and cell death (Jambunathan et al., 2001; Yang and Hua, 2004; Yang et al., 2006). This regulation is mediated through an R gene SNC1 because defense responses are induced constitutively in the bon1 mutants in a SNC1-dependent manner. In this study, we gained a further understanding of the BON1 gene and its regulation of defense responses by identifying BAP1 as its potential functional partner.
BAP1 is a small protein with a predicted C2 domain at its N-terminus. We show in this study that it binds to phospholipids in a calcium-dependent manner in vitro although it does not contain all the aspartates coordinating calcium binding found in many other C2 domains. No significant homology is identified for its C-terminal part, although it contributes to the binding to BON1 in an yeast two-hybrid assay (JH, unpublished result). The BAP1 protein appears to be present at an extremely low level in Arabidopsis because no small epitope-tagged BAP1 protein could be detected by Western blot in transgenic Arabidopsis plants. However, the BAP1 protein fused with GUS was detectable and was found to be localized to membranes in Arabidopsis. This could be due to the stability of the GUS protein (Jefferson et al., 1987) or the sensitivity of the GUS activity assay. In N. benthamiana, BAP1 is localized to both the soluble and membrane fractions when transiently overexpressed. The presence of BAP1 in the soluble fraction is either due to an overexpression artifact or due to the dynamic nature of the BAP1 localization captured in an exogenous plant system. Because its binding to lipids could be regulated by calcium ions, BAP1 may shuttle between membranes and cytosols according to its cellular environment. This study has not been able to resolve which membrane the BAP1 protein is associated with due to its low abundance. However, it is possible that BAP1 is at least partially co-localized with BON1 at the plasma membrane and the two proteins could have a direct physical interaction in vivo. This hypothesis is supported by fishing out one protein using the other as a bait in independent yeast two-hybrid screens as well as by their in vitro interaction in a protein pull-down assay.
Our genetic analysis further supports that the BAP1 gene is functionally related to the BON1 gene. The bap1 loss-of-function mutants have similar phenotypes to the bon1 loss-of-function mutants. The relatively weaker phenotype of bap1 is probably due to the presence of a BAP1 homolog in Arabidopsis (HY, S. Yang and JH, Cornell University, Ithaca, NY, USA, unpublished results). Similarly to the bon1 mutants, the BAP1 loss-of-function mutation confers enhanced disease resistance to virulent pathogens largely through an R gene SNC1 and accelerates HR to some avirulent pathogens. On the other hand, overexpression of BAP1 confers more susceptibility to a virulent pathogen, and it has not been feasible to test the effect of BON1 overexpression on virulent pathogens because of the difficulties in overexpression (JH, unpublished data). Thus this study reveals a role for BAP1 in R-mediated resistance similar to BON1 as well as a role in basal resistance not yet observed for BON1.
It remains to be determined how SNC1 is regulated by BAP1 and BON1. R genes were thought to function as receptors for pathogen effector Avr proteins for specific recognition. Emerging evidence indicates that R-gene activities can be regulated at different levels including transcription, protein stability and protein activity. For instance, the expression of a rice R gene Xa27 is induced by an effector protein and is responsible for specific disease resistance (Gu et al., 2005). Signaling of many R genes requires the activities of HSP90 (HEATSHOCK PROTEIN 90), RAR1 (RACE SPECIFIC RESISTANCE 1) and SGT1, indicating that the stability and the conformation of R proteins are highly regulated (Belkhadir et al., 2004; Schulze-Lefert, 2004). R proteins may also ‘guard’ or monitor the status of the host plant proteins that are targets of pathogen Avr effector proteins (Belkhadir et al., 2004; Dangl and Jones, 2001; Martin et al., 2003; Schneider, 2002; Van der Biezen and Jones, 1998). Several mechanisms can account for the apparent negative regulation of SNC1 by BAP1. One possibility is that BAP1 and BON1 repress the protein activity of SNC1 as proposed in the ‘guard’ hypothesis. The subcellular localization of SNC1 is not known and it could be membrane associated similar to what has been found for the R protein RPM1 (Boyes et al., 1998). Thus, the interaction could occur at the plasma membrane where BON1 and possibly BAP1 are localized. However, no direct interaction of BON1 or BAP1 with SNC1 was detected in the yeast two-hybrid system (YL and JH, unpublished results), suggesting that BON1 and BAP1 may indirectly inhibit the protein activity of SNC1 if they exist in a protein complex. Another possibility is that the repression of SNC1 activity occurs at the RNA transcript level. Indeed, the SNC1 transcript is upregulated in the bap1 and bon1 mutants. Because neither BON1 nor BAP1 is likely to possess transcriptional activities based on their sequence homologies, the regulation of SNC1 transcript level might be indirect. As suggested in the bon1 mutant study, this regulation might be mediated by SA through a positive feedback regulation in the mutants (Yang and Hua, 2004). Because not all SA-accumulating mutants have a bap1 like phenotype, induction of SNC1 by SA cannot account for all the phenotypes observed in bap1 or bon1 mutants. It is equally possible that SNC1 is involved in the amplification of defense response that is regulated by BAP1 or BON1. Cellular stresses or basal defense induced by the lack of BAP1 or BON1 could upregulate SA signaling which subsequently upregulates SNC1 and its signaling in the bap1 or bon1 mutants.
Genetic study reveals a SNC1-independent role for BAP1 in addition to a SNC1-dependent role. BAP1 regulates cell death and basal defense apparently in a SNC1-independent manner. Overexpression of BAP1 confers enhanced susceptibility to a virulent oomycete strain indicating a function of BAP1 in basal defense, and this function is not through the repression of SNC1. The bap1-1 mutant had an accelerated HR to Pst DC3000 with avrRpt2 compared to the wild type and this effect is also SNC1-independent. The HR is usually triggered by R-gene activation and involves calcium influx and the accumulation of reactive oxygen species. The acceleration of HR is seen with AvrRpt2 but not with AvrRpm1, suggesting that BAP1 may regulate specific R gene(s) to modulate HR to specific avirulent pathogens. Alternatively, BAP1 could be involved in maintaining the homeostasis of calcium ions and reactive oxygen species and therefore have a general role in modulation of the HR. The absence of acceleration of the HR to AvrRpm1 in bap1 could be due to a fast HR to AvrRpm1 in the wild type and thus the acceleration in bap1 will not be easily detected. Tests of the HR on more avirulent strains and with more sensitive assays will help to clarify this point.
It is tempting to speculate that the ancestral function of BAP1 is to maintain membrane and calcium homeostasis and thus its expression is highly responsive to environmental variations including both biotic and abiotic stresses. The BON1 gene is also responsive to biotic and abiotic stresses, but to a lesser extent than BAP1 (Hua et al., 2001; Jambunathan and McNellis, 2003). BAP1 and BON1 could be involved in regulating basal defense, because membrane and calcium homeostasis are probably critical in the basal defense response. Pathogens may subsequently target BAP1 and BON1 to suppress basal defense mediated by these genes. Plants may have then evolved to counteract this suppression by utilizing R genes to recognize the effects of the modification of the BAP1 and BON1 proteins to mount a HR.
In sum, BON1 and BAP1 are probably components of a complex and pathway that negatively regulate defense responses and their activities could potentially be modulated by environmental conditions. Further study will shed light on the molecular mechanisms of regulation of the defense response and the connection between basal and R-mediated defense responses.
Plant material and growth conditions
Arabidopsis thaliana plants were grown at 22°C or 28°C with 50–70% relative humidity under continuous fluorescent light (100 μmol m−2 sec−1) except for disease resistance tests where they were grown under the day/night cycle of 12 h of light and 12 h of darkness.
For the MBP:BAP1 fusion, the BAP1 cDNA was cloned into the expression vector pMALc2X (New England Biolab Inc., Ipswich, MA, USA) and expressed in E. coli BL21 (DE3). Fusion protein was purified with amylose resin following the manufacturer's protocol. For MYC:BON1A fusion, the coding sequence of the BON1 A domain was digested with SalI and NotI, Klenow-blunted and inserted into the SmaI site of the yeast vector pGBKT7 (Clontech, Mountain View, CA, USA). The MYC–BON1A fusion protein was expressed by the TNT® Quick Coupled Transcription/Translation system (Promega, Madison, WI, USA) following the manufacturer's protocol. For the ProBAP1::BAP1:GUS fusion, a genomic fragment containing the BAP1 promoter and the full-length coding region was translationally fused to the 5′ of the GUS gene in the vector PZP212 (Diener et al., 2000). For Pro35S::MYC:BAP1, six copies of MYC tag were fused to the N-terminus of BAP1 and the fusion protein was expressed in the pGreen0229 vector (http://www.pgreen.ac.uk/) with a 35S promoter and a NOS 3′ terminal sequences.
The full-length BAP1 protein with the MBP tag and the MBP control were each expressed in E. coli and the resulting proteins were purified according to the manufacturer's instructions. These proteins were incubated with phosphatydylserine (PS) in a binding buffer with or without calcium ions (Hua et al., 2001). After being incubated at room temperature for 30 min, the mixture was precipitated by spinning at 13 000 g in a microcentrifuge. Pellets were dissolved in 1× sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) loading buffer (50 mm Tris-Cl pH 6.8, 2% SDS, 100 mm DTT, 1% bromophenol blue, 10% glycerol) and resolved in 12% SDS-PAGE. Proteins were visualized by staining with Coomassie Blue G-250.
Plant tissues were collected, ground in the protein extraction buffer (50 mm Tris-Cl, pH 8.0, 1 mm EDTA, 150 mm NaCl, 10% glycerol), and spun at 8000 g for 20 min to remove the cell debris. The clear extract was transferred to a new tube and centrifuged at 45 000 g for 30 min. The supernatant after the spin is the soluble protein fraction and the pellet is the membrane fraction.
GUS activity analysis
Prepared protein samples were incubated overnight at 37°C in staining solution (50 mm sodium phosphate, pH 7.0, 10 mm EDTA, 2 mm 5-bromo-4-chloro-3-indoyl glucuronide, 1 mm potassium ferricyanide and 1 mm potassium ferrocyanide).
In vitro pull-down assay
The MBP:BAP1 fusion protein and the MBP protein were each expressed and purified from E. coli strain BL21 (DE3). Without eluting them from the amylose resin, the amylose-bound proteins were used for the following analysis. Five microliters of in-vitro transcribed and translated MYC:BON1A and 5 μg of MBP or MBP:BAP1 were mixed in the binding buffer (20 mm HEPES, pH 7.4, 150 mm KCl, 0.1% gelatin, 0.1% Triton X-100, 0.1% NP40, 2 mm MgCl2, 2 mm DTT) and kept at room temperature for 30 min. Amylose resin-bound MBP or MBP:BAP1 were collected by short spin, resuspended in 1× SDS-PAGE loading buffer, and resolved on 12% SDS-PAGE. The existence of MYC:BON1A fusion was detected by anti-Myc monoclonal antibody (Covance, Denver, PA, USA).
Yeast two-hybrid analysis
The BAP1 protein was fused with the DNA-binding domain of the GAL4 transcription factor in the yeast vector pGBD-C2 and was transformed to the yeast strain PJ69-4α (James et al., 1996). This bait strain containing pGBD:BAP1 was used to screen several Arabidopsis cDNA libraries made from young seedlings and inflorescences as described (Hua et al., 2001). Transformants were selected on synthetic complete medium without histidine, tryptophan or leucine.
RNA blot analysis
Total RNAs were extracted from 3-week-old plants using TRI Reagents (Molecular Research Inc., Cincinnati, OH, USA) according to the manufacturer's protocol. Twenty micrograms of RNAs was resolved on a 1.2% agarose gel containing 1.8% formaldehyde. Ethidium bromide was used to visualize the rRNA transcripts to ensure equal loading. The RNA gel blots were hybridized with gene-specific, 32P labeled, single-stranded DNA probes. For BAP1 and PR1, the probes consist of the full-length coding region. For SNC1, the probe contains the first exon of the gene.
Pathogen resistance assay
The Pst DC3000 was grown on KB medium at 30°C overnight. Bacterial growth inside plant leaves was monitored as described (Tornero and Dangl, 2001). Two-week-old plants were dip inoculated with bacteria resuspended at 10−8 CFU ml−1 in 10 mm MgCl2 and 0.02% Silwet L-77 and were kept covered for 24 h. At 1 h after dipping (t = 0) and day 3, the number of CFUs per plant was determined. Three whole seedlings (aerial tissue) were collected in 1 ml of 10 mm MgCl2 with 0.02% Silwet L-77 and shaken at 30°C for 1 h. Serial dilutions of the solution were used to titer the bacterial growth.
Hyalopernonspora parasitica Noco2 was propagated on Arabidopsis ecotype Col. Spores (40 000 ml−1 in water) were spray inoculated onto 2-week-old plants that were subsequently maintained in 100% humidity at 16°C. The number of sporangiophores formed on leaves was counted 1 week later. Approximately 100 leaves were counted for each genotype.
We thank Q. Sun for sequence analysis, S. Yang for technical assistance, the Arabidopsis Bioresource Research Center and Syngenta for T-DNA insertion lines, J. Glazebrook, J. Dangl, J. Parker and X. Dong for mutant Arabidopsis seeds, W. Crosby for the yeast two-hybrid library, A. Collmer for bacterial strains and T. Delaney for oomycete strains. We thank G. Martin, S. Yang and anonymous reviewers for critical reading of the manuscript. This material is based upon work supported by the National Science Foundation under grant no. 0415597.