First published online October 20, 2006; 10.1105/tpc.105.039339
The Plant Cell 18:2664-2679 (2006)
© 2006 American Society of Plant Biologists
BRANCHED-CHAIN AMINOTRANSFERASE4 Is Part of the Chain Elongation Pathway in the Biosynthesis of Methionine-Derived Glucosinolates in Arabidopsis[W]
Joachim Schustera,
Tanja Knilla,
Michael Reicheltb,
Jonathan Gershenzonb and
Stefan Bindera,1
a Molekulare Botanik, Universität Ulm, 89069 Ulm, Germany
b Max Planck Institut für Chemische Ökologie, Abteilung Biochemie, 07745 Jena, Germany
1 To whom correspondence should be addressed. E-mail stefan.binder{at}uni-ulm.de; fax 49-731-502-2626.
 |
ABSTRACT
|
|---|
As part of our analysis of branched-chain amino acid metabolism in plants, we analyzed the function of Arabidopsis thaliana BRANCHED-CHAIN AMINOTRANSFERASE4 (BCAT4). Recombinant BCAT4 showed high efficiency with Met and its derivatives and the corresponding 2-oxo acids, suggesting its participation in the chain elongation pathway of Met-derived glucosinolate biosynthesis. This was substantiated by in vivo analysis of two BCAT4 T-DNA knockout mutants, in which Met-derived aliphatic glucosinolate accumulation is reduced by  50%. The increase in free Met and S-methylmethionine levels in these mutants, together with in vitro substrate specificity, strongly implicate BCAT4 in catalysis of the initial deamination of Met to 4-methylthio-2-oxobutyrate. BCAT4 transcription is induced by wounding and is predominantly observed in the phloem. BCAT4 transcript accumulation also follows a diurnal rhythm, and green fluorescent protein tagging experiments and subcellular protein fractions show that BCAT4 is located in the cytosol. The assignment of BCAT4 to the Met chain elongation pathway documents the close evolutionary relationship of this pathway to Leu biosynthesis. In addition to BCAT4, the enzyme methylthioalkylmalate synthase 1 has been recruited for the Met chain elongation pathway from a gene family involved in Leu formation. This suggests that the two pathways have a common evolutionary origin.
|
INTRODUCTION
|
|---|
Several pathways of amino acid metabolism in plants have been deduced at least partially from those found in bacteria, fungi, and animals. However, the metabolism of these important compounds in plants appears to be much more complex than in other groups of organisms. For instance, the presence of plastids leads to an additional compartmentalization of pathways. Furthermore, amino acids are important substrates for the biosyntheses of plant-specific hormones, vitamins, and secondary metabolites. These additional pathways and levels of complexity in compartmentalization correlate with the presence of a large number of gene families in which the functions of individual members remain as yet poorly known.
In Arabidopsis thaliana, the branched-chain aminotransferases (BCATs) are encoded by a small gene family. The seven genes are highly conserved and six of them (BCAT1 to BCAT6) have been documented to be transcribed (Diebold et al., 2002 ). Extrapolating from their subcellular localization in plastids, BCAT2 (At1g10070), BCAT3 (At3g49680), and BCAT5 (At5g65780) are likely to participate in branched-chain amino acid (BCAA) biosynthesis, while the mitochondrial BCAT1 (At1g10060) has been proposed to initiate the degradation of Leu, Ile, and Val. Two enzymes (BCAT4 [At3g19710] and BCAT6 [At1g50110]) have been suggested to reside in the cytosol where their function is entirely unknown (Diebold et al., 2002; Schuster and Binder, 2005). In addition to their different subcellular locations, the expression patterns of the individual members of this gene family differ considerably. This observation further substantiates the suggestion of distinct functions for the different enzymes in BCAA metabolism and/or related pathways (Schuster and Binder, 2005).
Almost all members of the Arabidopsis BCAT family show the expected catalytic activity. Five of the six expressed proteins have been found to rescue the BCAA auxotrophy of a yeast mutant lacking endogenous BCAT activity. However, BCAT4 is not able to complement this mutant even when a mitochondrial targeting sequence was attached to the enzyme, and the actual function of this protein remains unclear (Diebold et al., 2002). Hints toward the function of BCAT4 came from enzyme assays performed with the mitochondrial BCAT1. These assays confirmed aminotransferase activities with the standard substrates Leu, Ile, and Val and the respective 2-oxo acids but also revealed a broadened substrate specificity by demonstrating substantial activities with 2-aminobutyrate and Met and their corresponding 2-oxo acids (Schuster and Binder, 2005). The conversion of Met and the corresponding 2-oxo acid 4-methylthio-2-oxobutyrate (MTOB) suggests two possible additional functions. First, a transamination of MTOB to Met is required in the final step of the Met salvage pathway, which allows the recovery of this valuable amino acid from methylthioadenosine. The latter is a by-product arising in the biosyntheses of polyamines and ethylene, which both start from Met (Wang et al., 1982; Miyazaki and Yang, 1987; Walters, 2003). In bacteria, a BCAT also catalyzes this step, suggesting by analogy that one of the homologous enzymes in plants might have a similar function (Berger et al., 2003; Sekowska et al., 2004; Venos et al., 2004; Schuster and Binder, 2005). Second, transamination reactions with Met, its derivatives, and the corresponding 2-oxo acids occur in the chain elongation pathway, the first phase of the biosynthesis of Met-derived aliphatic glucosinolates (Kroymann et al., 2001; Mikkelsen et al., 2002; Wittstock and Halkier, 2002; Textor et al., 2004; Grubb and Abel, 2006). This pathway is initiated by the transamination of Met to MTOB, and further transaminations occur subsequently when chain-elongated 2-oxo acids are converted to Met derivatives, which are intermediates in the biosynthesis of the parent glucosinolates (Mikkelsen et al., 2002; Wittstock and Halkier, 2002). Thus, the substrate spectrum observed in previous investigations suggests a function of one or more of the BCATs in glucosinolate biosynthesis and/or in the Met salvage pathway (Schuster and Binder, 2005).
Here, we report a detailed in vitro and in vivo analysis of the function of BCAT4. Substrate specificity assays with the recombinant enzyme and glucosinolate and amino acid profiling of respective T-DNA mutants reveal this protein to be a component of the chain elongation pathway in the biosynthesis of Met-derived glucosinolates. Furthermore, the increased levels of Met and S-methylmethionine (SMM) in the T-DNA mutants assign BCAT4 to the initial deamination of Met to MTOB. The expression of BCAT4 and methylthioalkylmalate synthase 1 (MAM1; At5g23010), which catalyzes the subsequent step in the chain elongation pathway, can both be triggered by wounding, which is consistent with a function of glucosinolates in protection from herbivores.
Interestingly, BCAT4 and MAM1, the latter having been identified as a member of the isopropylmalate synthase gene family, belong to gene families encoding proteins involved in Leu biosynthesis. This suggests that the pathways of glucosinolate chain elongation and Leu formation may have a close evolutionary relationship.
|
RESULTS
|
|---|
BCAT4 (At3g19710) Is More Active with Intermediates of Met Chain Elongation Than with BCAAs
In a previous complementation analysis, BCAT4 failed to rescue the auxotrophy of a yeast double knockout mutant for BCAAs, suggesting a function other than typical BCAA metabolism (Diebold et al., 2002). To get more information about the potential role of this protein, we compared AtGenExpress expression profile data of BCAT4 with those of 60 other genes potentially involved in the metabolism of branched-chain and other amino acids. This revealed an extraordinarily strong correlation of BCAT4 (At3g19710) expression with that of the MAM1 gene (At5g23010), which encodes a methylthioalkylmalate synthase (data not shown) (Kroymann et al., 2001; Textor et al., 2004; http://www.arabidopsis.org/info/expression/ATGenExpress.jsp). This enzyme belongs to the isopropylmalate synthase gene family, which has four members in Arabidopsis. While the function of one of these proteins (At1g18500) is unsolved, one (At1g74040) is most likely engaged in Leu biosynthesis. MAM1 (At5g23010) and MAM3 (formerly MAML [At5g23020]) are enzymes of the chain elongation pathway, the first phase in the biosynthesis of Met-derived glucosinolates (Field et al., 2004; Textor et al., 2004). This pathway includes several transamination steps, and the striking correlation of MAM1 and BCAT4 transcription in Arabidopsis suggests that one or more of these steps might be catalyzed by the BCAT4-encoded aminotransferase (Mikkelsen et al., 2002; Wittstock and Halkier, 2002). To investigate this possibility, enzyme assays were performed with recombinant BCAT4 protein and a variety of substrates. The complete BCAT4 reading frame was C-terminally fused to the S-Tag in the pET32a vector and expressed in Escherichia coli, and the resulting protein preparation was considerably enriched by affinity purification with S-protein agarose (see Supplemental Figure 1 online and Methods). If BCAT4 were to function in the chain elongation pathway, strong activities would be expected with Met and its elongated derivatives and with the corresponding 2-oxo acids.
The activity of recombinant BCAT4 was first assayed with MTOB using a coupled enzyme test established for BCATs from animals (Schadewaldt and Adelmeyer, 1996). This assay has been applied previously to investigate the mitochondrial BCAT1 of Arabidopsis (At1g10060) (Schuster and Binder, 2005). For MTOB, plotting of reaction velocities and substrate concentrations revealed that the enzyme exhibits Michaelis-Menten kinetics (see Supplemental Figure 2 online). Km and Vmax, calculated by fitting the data into a nonlinear regression, were found to be 0.045 ± 0.006 mM (Km) and 2.7 ± 0.09 µmol/min·mg (Vmax), respectively. Based on these results, activities with different 2-oxo acids were measured at two substrate concentrations: 0.1 and 2.0 mM. These are expressed relative to the activity with MTOB, which was arbitrarily set to 100% (Table 1
, top part). Besides the strong activity detected with this substrate, considerable activities were found with 5-methylthio-2-oxopentanoate (MTOP; 48 and 53% of MTOB at 0.1 and 2.0 mM, respectively). No activity has been observed with 6-methylthio-2-oxohexanoate (MTOH), 3-methyl-2-oxopentanoate (3MOP), or 3-methyl-2-oxobutyrate (3MOB), the latter being the 2-oxo acids of Ile and Val, respectively. Distinct activities were seen with 4-methyl-2-oxopentanoate (4MOP), the 2-oxo acid of Leu. While no substrate conversion was detected at a concentration of 0.1 mM, a relative activity of 41% in comparison to MTOB was found at 2 mM substrate concentration. No activities were found with a control lysate obtained from E. coli expressing an empty pET32a vector, confirming that the measured activities originate from overexpressed BCAT4 and not from any copurified E. coli proteins.
To identify optimal concentrations for relative substrate specificity assays with amino acids, the kinetic of the deamination of Met by the recombinant BCAT4 was determined. This was found to follow a Michaelis-Menten behavior (see Supplemental Figure 3 online). For Met, Km and Vmax values were calculated to be 0.93 ± 0.08 mM and 0.089 ± 0.002 µmol/min·mg protein, respectively. According to these data, various substrates were measured at 1 and 5.0 mM with an HPLC-based activity and detection system as described in Methods (Table 1, bottom part). The activity with Met was used as standard and was arbitrarily set to 100%. At both substrate concentrations, BCAT4 indeed displayed its highest relative activities with this amino acid. Substantial activities were also measured with homomethionine (37% at 1 mM and 65% at 5 mM) and with Leu (29% at 1 mM and 62% at 5 mM), while only little or no activity was detected with Ile (8% at 5 mM) and Val (Table 1, bottom part). We also studied the kinetics of BCAT4 with Leu and observed a perfect Michaelis-Menten correlation. Km and Vmax values were found to be 4.86 ± 0.41 mM and 0.11 ± 0.004 µmol/min·mg protein, respectively (see Supplemental Figure 4 online). Thus, the affinity of BCAT4 to Leu is about five times lower than to Met.
Taken together, BCAT4 catalyzes the transamination of typical intermediates of the Met chain elongation pathway, supporting a role for this protein in this pathway. These results are consistent with the inability of BCAT4 to restore the BCAA auxotrophy of a respective yeast double knockout strain, which lacks endogenous BCAT activity (Diebold et al., 2002).
Accumulation of Met-Derived Glucosinolates Is Substantially Reduced in BCAT4 T-DNA Mutants
To examine the function of BCAT4 in vivo, two T-DNA insertion mutants obtained from the SALK and GABI-Kat collections were investigated (Alonso et al., 2003; Rosso et al., 2003). The SALK mutant (013627, designated bcat4-1) is predicted to contain the T-DNA in the last exon (Figure 1A
). An examination of this line by PCR with gene-specific primers and oligonucleotides specific for right and left borders confirmed the T-DNA in the predicted site (data not shown). The left border primer generates PCR fragments with BCAT4-specific primers annealing to both T-DNA flanking sequences, indicating that at least two left border sequences are present (Figure 1A). Direct sequencing of these PCR fragments identified the insertion at cDNA positions 2142 and 2147 with respect to the ATG (+1), indicating a five-nucleotide deletion. The genotype analysis identified plants #24 and #6 to be homozygous for T-DNA allele bcat4-1, while no insertion could be detected in the BCAT4 gene in plant #7, which is regarded as a wild-type plant from the SALK collection (data not shown). These plants and a normal Columbia (Col-0) wild type were selected for RNA gel blot analysis with a BCAT4-specific probe corresponding to exons 1 to 3 amplified with primers Bcat4UE.H and Bcat4-5'.R (Figure 1A). A strong signal corresponding to a 1.25-kb mRNA was detected in plants carrying a wild-type allele (Figure 1B, left panel). An 500-nucleotide-larger transcript is detected in homozygous bcat4-1 plants, indicating accumulation of a bcat4-1 allele-specific transcript. A densitometric quantification of the RNA gel blot results revealed the latter to accumulate to 20% of the wild-type level.
View larger version (49K):
[in this window]
[in a new window]
|
Figure 1. Characterization of BCAT4 T-DNA Insertion Mutants SALK 013627 and GABI-Kat 163D11.
(A) Diagram of the BCAT4 gene with exons given as black boxes. The T-DNA insertions are located within the last exon (SALK 013627; allele bcat4-1) and in the second intron (GABI-Kat 163D11; allele bcat4-2). Both insertions have left border (LB) sequences at both extremities. Positions of primers bcat4UE.H and bcat4-5'.R used for generation of the hybridization probe are indicated by bent arrows.
(B) RNA gel blot hybridization of total RNA isolated from Col-0 wild-type plants from the SALK (plant #7) and GABI-Kat collections (plants #3 and #6) and plants homozygous for the two different T-DNA alleles: bcat4-1 (plants #6 and #24) and bcat4-2 (plants #2 and #4). Hybridizations were performed with a BCAT4-specific probe, and equal loading of the gel was verified by visualization of the rRNA (bottom panel).
|
|
In the GABI-Kat mutant (163D11, named bcat4-2), likewise two T-DNA left border extremities were found in the predicted location in intron 2 (Figure 1A). The RNA gel blot analysis again detected 1.25-kb mRNAs in the wild-type plants (#3 and #6), while very small transcripts with a size of <0.5 kb were found in bcat4-2 homozygous plants #2 and #4 (Figure 1B). Immunodetection analysis with an antibody raised against the C-terminal dodecapeptide of BCAT4 confirmed that a native BCAT4 protein is not detectable in the plants homozygous for alleles bcat4-1 and -2, so that these lines are complete knockouts (see below; data not shown).
Offspring of mutant plants #24 (bcat4-1) and #4 (bcat4-2) and different wild-type plants of ecotype Col-0 from various sources, including plants #7 and #3 from the SALK and GABI-Kat collections, respectively, were selected for glucosinolate profiling. In leaves, total glucosinolate contents were reduced from 27.23 (±2.65) µmol/mg dry weight in the wild type to 17.62 (±0.99) in bcat4-1 and from 24.87 (±0.11) µmol/mg dry weight in the wild type to 16.66 (±1.83) in bcat4-2 (Table 2
). This corresponds to a reduction to 65 and 67%, respectively. The reduction is even stronger when only Met-derived glucosinolates are considered. These secondary compounds were reduced from 21.54 (±1.80) µmol/mg dry weight in the wild type to 11.38 (±0.90) in bcat4-1 and from 21.04 (±0.10) µmol/mg dry weight in the wild type to 11.37 (±1.23) in bcat4-2, equivalent to a reduction to 53 and 54%, respectively (Table 2). By contrast, total indolic glucosinolates were slightly increased in the mutants. In seeds, similar reductions were observed. Here, total Met-derived glucosinolates were reduced from 83.23 (±7.08) µmol/mg in the wild type to 42.22 (±2.42) in bcat4-1 and from 111.78 (±18.12) µmol/mg in the wild type to 63.12 (±9.40) in bcat4-2, which corresponds to a reduction to 51 and 56%, respectively (Table 3
). These reductions are given with respect to the glucosinolate content of the different wild-type plants isolated from the seed pool obtained for each mutant and from normal wild-type plants. The amounts and profiles measured in the plants are in the range found previously in Arabidopsis ecotype Col-0 (Brown et al., 2003). In leaves of the mutants, strong reductions were observed for almost all types of Met-derived glucosinolates measured, including the most abundant 4-methylsulfinylbutylglucosinolate (4MSOB). Only 5-methylsulfinylpentylglucosinolate (5MSOP) remained unaffected in the mutants. Likewise in seeds, most types of these aliphatic glucosinolates were reduced, including the most prominent 4-methylthiobutylglucosinolate (4MTB). However, in this tissue, 4MSOB and 5MSOP are more abundant in the mutants by a factor of 2.2 and 2.4, respectively (Table 3). In plant #4, homozygous for bcat4-2, almost the same alterations were observed, including the increase of 4MSOB and 5MSOP levels in seeds and unchanged amounts of 5MSOP in leaves. A slight discrepancy was observed for 4-hydroxybutylglucosinolate (4OHB) in seeds, which was somewhat reduced in plant #24 homozygous for bcat4-1 (factor of 1.6) but was slightly elevated in plant #4 homozygous for bcat4-2 (factor of 1.2; Table 3).
Taken together, both mutants bcat4-1 and -2 clearly have reduced levels of Met-derived glucosinolates, strongly suggesting a participation of BCAT4 in one or several transamination reactions of the Met chain elongation pathway.
The Levels of Free Met and SMM Are Substantially Increased in BCAT4 T-DNA Mutants
Since the glucosinolates, reduced in the bcat4 T-DNA mutants, are derived from Met, it is possible that a reduction of these secondary metabolites might lead to a buildup of Met. We therefore determined and compared the amounts of free amino acids in bcat4-1 and -2 homozygous plants with wild-type concentrations. Indeed, in leaves, the level of free Met was found increased from 0.12 (±0.05) to 0.59 (±0.09) and from 0.06 (±0.00) to 0.74 (±0.39) nmol/mg dry weight, which corresponds to an approximate fivefold (bcat4-1) and 12-fold (bcat4-2) increase, respectively (Table 4
). All other proteinogenic amino acids measured in this tissue did not show any clear changes, except Thr, which showed an increase of 25%. In seeds, a slightly weaker increase of free Met was detected from 0.08 (±0.03) to 0.28 (±0.05) nmol/mg and 0.09 (±0.02) to 0.43 (±0.09) nmol/mg seed with a threefold and fivefold increase in bcat4-1 and bcat4-2, respectively, compared with the wild type. However, in this organ, the levels of His, Ile, Lys, and Ser showed significant increases in the mutants (Table 5
). Since S-adenosylmethionine (AdoMet) and SMM are the major derivatives of Met, it seemed probable that the level of at least one of these compounds might also be elevated in the mutants. SMM, which is the main transport form of Met, is a major constituent of the phloem sap in wheat (Triticum aestivum) directed to the developing seeds (Bourgis et al., 1999). It is thus present in the phloem, the main tissue in which BCAT4 is expressed (see below) and where an excess of Met could lead directly to an increase of this transport compound. To investigate this, we measured SMM in leaves of wild-type plants and found SMM at levels of 0.3 nmol/mg dry weight, which is in the range reported previously (Table 4; Bourgis et al., 1999). In the same tissue of the mutants, SMM was found to be increased by a factor of three (0.84 [bcat4-1] and 0.96 [bcat4-2] nmol/mg dry weight, respectively). By contrast, a totally different situation was found in seeds. While SMM could not be detected in this organ from wild-type plants, extremely high levels of this compound were observed in bcat4-1 (4.29 nmol/mg seeds) and in bcat4-2 (8.65 nmol/mg seeds) (Table 5). These levels exceed almost all other free amino acids measured in seeds of both wild-type and mutant plants. This sharp increase of free Met and its transport derivative SMM further support the conclusion that BCAT4 catalyzes the initial transamination of Met in glucosinolate formation generating MTOB.
The BCAT4 Promoter Is Active in the Phloem
In a previous analysis, the BCAT4 promoter activity was mainly found in the vasculature of cotyledons and leaves, but it remained unclear whether the transcriptional activity resides directly in the veins or in the neighboring tissue (Schuster and Binder, 2005). To investigate BCAT4 promoter activity in more detail, segments of a flowering stalk of a transgenic plant carrying a BCAT4 promoter:ß-glucuronidase (GUS) construct were histochemically stained, and cross sections through the intensively colored ends were inspected by light microscopy (Figures 2A and 2B
). Blue staining is restricted to the phloem, while no staining could be detected in cells directly adjacent to the phloem cells or in any other cells of the stalk. This includes the so-called S-cells, which have been identified between the phloem and the endodermis of flowering stalks. These cells have been suggested to be the predominant storage sites for glucosinolates (Koroleva et al., 2000). Thus, biosynthesis of Met-derived glucosinolates starts in a tissue adjacent to the major storage cells. Likewise, BCAT4 promoter activity is found in the stele of the roots (Figure 2C). The examination of cross sections again identified promoter activity in phloem cells (Figure 2D).
View larger version (118K):
[in this window]
[in a new window]
|
Figure 2. BCAT4 Promoter Activity Is Found in the Phloem.
(A) and (B) Cross sections through a flowering stalk detect promoter activity in phloem cells.
(C) In roots, promoter activity is indicated in the vascular tissue.
(D) A cross section through a root detects promoter activity primarily in cells of the phloem. C, cortex; En, endodermis; Ep, epidermis; P, phloem; Pi, pith; S, S-cells; Sc, sclerenchyma; St, stele; and X, xylem.
|
|
Transcription of BCAT4 and MAM1 Is triggered by Wounding
In leaves, BCAT4 promoter activity was almost exclusively found in veins in distinct areas in the basal part of the leaf blade and in petioles (Schuster and Binder, 2005). This pattern might be induced when the leaf is cut from the plant. Strong staining of the stalk segments is also found at the cut surface. To analyze whether BCAT4 promoter activity is indeed induced by wounding, leaf tips were cut off with scissors, squeezed with forceps, or pierced with a pin while the leaf was still attached to the plant. After 5 min, leaves were separated from the plant and transferred to staining buffer. Besides the expected staining of the cut surface of the petiole, strong staining was seen in veins within or very close to areas of mechanical wounding (Figure 3A
). Staining of untreated control leaves was found only at the site of cutting on the petiole (data not shown; Schuster and Binder, 2005). This strongly suggests that wounding indeed induces BCAT4 promoter activity.
View larger version (60K):
[in this window]
[in a new window]
|
Figure 3. BCAT4 Transcription Is Induced by Wounding.
(A) Promoter activity as determined by histochemical GUS staining in respective reporter gene lines is induced upon mechanical wounding by cutting off the leaf tip with scissors (left), by squeezing with forceps (center), and by piercing with a needle (right panel).
(B) Rapid transient induction of BCAT4 and MAM1 steady state mRNAs after wounding. Total RNA was collected from leaf material at different time points after wounding as given in the top line. Steady state levels for mRNAs of BCAT4 and MAM1 were detected by RNA gel blot hybridization and densitometrically analyzed. Loading of the individual lanes was verified by hybridization with an oligonucleotide complementary to 18S rRNA (bottom row).
|
|
To examine this phenomenon by an independent method and to get more information about the time dependence of the response, leaves wounded by squeezing were harvested at different time points after treatment and analyzed by RNA gel blot analysis with a BCAT4 probe. A 1.9-fold increase of BCAT4 steady state mRNA level can already be observed 5 min after wounding (Figure 3B, top panel). After that, only a small further increase of mRNA accumulation was detected (maximum of 2.6-fold). The transcript level starts to drop after 1 h and seems to return to the original level after 2 h. A similar pattern is found with the steady state level of MAM1 mRNA. After 5 min, the MAM1 transcript level is increased by a factor of 1.5, a maximum is reached after 15 min (1.9-fold increase), and the expression level starts to drop after 1 h (Figure 3B, bottom panel).
BCAT4 mRNA Accumulation Is Light Dependent
The experiments described above demonstrate that BCAT4 might catalyze the initial and/or other early steps in the chain elongation pathway of Met. Thus, it might be one of the regulators controlling the flux from primary to secondary metabolism. To gather more information about regulation of this gene, transcript levels were examined in plants grown under a 12-h-light/12-h-dark regime. Leaves were harvested at 4-h intervals, and the isolated total RNA was inspected by RNA gel blot hybridization. While only low BCAT4 mRNA levels were observed during the night, an up to fivefold increase was found upon illumination (Figure 4
, top panel). The level remains high under continued exposure to light. An analogous pattern is found in a parallel experiment with MAM1 (Figure 4, middle panel), demonstrating the light-dependent transcript accumulation of both genes. This suggests a diurnal expression of both proteins and demonstrates that simultaneous expression of BCAT4 and MAM1 extends beyond tissue specificity and promoter activation by wounding to a temporal correlation.
View larger version (42K):
[in this window]
[in a new window]
|
Figure 4. Diurnal Transcription of BCAT4 and MAM1.
RNA gel blot analysis of plants grown under a 12-h-light/12-h-dark regime followed by an extended day was performed. Steady state levels of BCAT4 and MAM1 mRNAs detected at different time points are low in the absence of light and approximately fivefold higher under illumination. mRNA levels remain high under extended illumination. Equal loading of the agarose gels was verified by hybridization with an oligonucleotide complementary to 18S rRNA (bottom row).
|
|
BCAT4 Is a Cytosolic Protein
The subcellular localization of the Met elongation pathway has been so far unclear. The previously identified MAM proteins are in silico predicted to be located in plastids (S. Binder, unpublished data). This subcellular localization has now been experimentally confirmed for MAM3 (S. Textor, J.-W. de Kraker, B. Knoke, S. Schuster, J. Gershenzon, and J. Tokuhisha, unpublished data). To address this question for BCAT4, the subcellular targeting of this protein was analyzed by green fluorescent protein (GFP) tagging. The cDNA representing the complete BCAT4 reading frame was fused to the 5' terminus of the smGFP4 gene (Davis and Vierstra, 1998). This construct was then transiently transformed into tobacco (Nicotiana tabacum) protoplasts, and targeting of the fusion protein was followed by fluorescence microscopy. The green fluorescence is seen exclusively in the cytosol (Figure 5A
). No GFP fluorescence was found in the nucleus, in which GFP alone is frequently observed. Also, no fluorescence was detectable in chloroplasts or peroxisomes, both of which are visible as negative prints within the cytosol (Figure 5A, left and center panels). Likewise, no green fluorescence is found in mitochondria, which were specifically stained with MitoTracker red, or any other subcellular compartment (Figure 5A, right panel).
View larger version (59K):
[in this window]
[in a new window]
|
Figure 5. BCAT4 Is a Cytosolic Enzyme.
(A) A full-length BCAT4:smGFP fusion protein is transiently expressed in tobacco protoplasts. Three filter sets allow the detection of fluorescence at different wavelengths: one window for GFP and chlorophyll autofluorescence, one for GFP alone, and a third optimized for MitoTracker. The protoplasts were stained with a mitochondria-specific dye (MitoTracker). Bars = 10 µm.
(B) Immunodetection analysis of subcellular protein fractions. sp(s), total soluble protein from wild-type and bcat4-1 seedlings; sp(a), total soluble protein from adult wild-type plants; cy, cytosol; cp, chloroplasts; and mt, mitochondria. Antibodies (@) were directed against BCAT4, the cytosolic marker UDP-glucose pyrophosphorylase (UGPase), the chloroplast protein ferredoxin-NADP+ reductase (FNR), and the mitochondrial porin.
|
|
To investigate the subcellular localization by an independent experimental approach, subcellular protein fractions were analyzed for the presence of BCAT4 (Figure 5B). A BCAT4-specific and affinity-purified antibody binds to this aminotransferase in fractions containing total soluble proteins and in a fraction enriched for cytosolic protein from wild-type plants. No BCAT4 protein can be detected in fractions containing purified chloroplasts or mitochondria (Figure 5B, @ BCAT4). The purity of the different fractions was tested with antibodies against cytosolic UDP-glucose pyrophosphorylase, the plastid-located ferredoxin-NADP+ reductase, and the mitochondrial porin. UGPase, which in Arabidopsis presents as two almost identical proteins (At5g17310 [predicted mass of 51.9 kD] and At3g03250 [51.7 kD]; 93% identical amino acids), is found in the total soluble protein fractions and in the cytosol fraction but not in chloroplasts and mitochondria, documenting that no detectable cytosolic contamination is present in these fractions. The background signal observed in the chloroplast fraction is due to a highly abundant protein of this same size, which is also present in the total protein fractions but almost absent from the cytosolic and mitochondrial fractions. This protein is most likely the large subunit of ribulose bisphosphate carboxylase (data not shown). FNR proteins (At5g66190 [40.3 kD] and At1g20020 [41.2 kD]; 74% identical amino acids) are present in the total soluble protein fractions and in the cytosolic fraction but are predominantly found in chloroplasts. No FNR is present in mitochondria. The mitochondrial protein porin is not seen in the total soluble protein fractions, in the cytosol, or in the chloroplasts but is clearly detectable in the mitochondrial fraction. These experiments show that mitochondrial and chloroplast fractions are free or almost free of detectable amounts of contaminant proteins from the respective other subcellular fractions, while the cytosol seems to be slightly contaminated by plastids. However, considering the strong enrichment of the cytosol and chloroplast fractions and the exclusive detection of BCAT4 in the cytosol, these results clearly show that BCAT4 is a cytosolic protein. The relatively weak signal of BCAT4 in the cytosol is most likely due to the use of protoplasts as a source for the cytosol preparation. Cells of the phloem, where BCAT4 is mainly expressed (Figure 2), are most likely underrepresented in the protoplast preparation since their release from the leaf tissue is less efficient in comparison to mesophyll cells.
Taken together, the two independent experimental approaches consistently show that BCAT4 is a cytosolic enzyme and is thus located in a compartment different from the MAM enzymes. No BCAT4 protein is found in the total soluble protein fractions of the bcat4-1 and bcat4-2 plants in line with the above characterization of these mutants as complete expressional knockouts (Figure 5B; data not shown).
|
DISCUSSION
|
|---|
BCAT4 Participates in the Met Chain Elongation Pathway
Although BCAT4 (At3g19710) is a member of the BCAT family in Arabidopsis, several lines of evidence strongly suggest a function of this protein in the Met chain elongation pathway of aliphatic glucosinolate biosynthesis. First, in vitro enzyme assays demonstrate that this protein has a clear substrate preference for intermediates of the Met chain elongation pathway (Table 1). The standard substrates for BCATs are, with the exception of Leu and its 2-oxo acid, only poorly converted if at all. Second, knockout mutants of BCAT4 show reductions in their Met-derived glucosinolate content of 50% (Tables 2 and 3). Third, in these T-DNA insertion mutants, the reduction of the glucosinolate content correlates with a substantial increase of Met, the initial substrate for the biosynthesis of these compounds, and a striking accumulation of SMM, the transport form of Met in the phloem (Tables 4 and 5). Fourth, BCAT4 transcription is closely correlated with that of MAM1, a gene encoding a methylthioalkylmalate synthase catalyzing the condensation of MTOB and other 2-oxo acids with acetyl-CoA in the Met chain elongation pathway. These genes are transcribed in the same tissues (http://www.arabidopsis.org/info/expression/ATGenExpress.jsp), show identical light-dependent mRNA accumulation (Figure 4), and are both induced by wounding (Figure 3). Fifth, BCAT4 is expressed in the phloem in direct vicinity to the S-cells, which have been suggested to be the major storage sites of glucosinolates in flowering stalks (Figure 2; Koroleva et al., 2000). A strong phloem-specific expression of BCAT4 has recently also been observed in a microarray analysis of root hypocotyls in Arabidopsis. In this study, BCAT4 and MAM1 were found to be almost exclusively expressed in the phloem, again confirming that both genes are transcribed in the same tissues (Zhao et al., 2005). Taken together, these data unambiguously demonstrate a participation of BCAT4 in the chain elongation pathway of Met.
Such a promiscuous character has also been found for bacterial BCATs, which are active in the Met salvage pathway (Berger et al., 2003; Sekowska et al., 2004; Venos et al., 2004). In this pathway, also called the Yang cycle, BCATs catalyze the final step by transaminating MTOB to Met. In Mycobacterium tuberculosis and also in the Bacillus species, BCAAs can serve as amino donors in this reaction. (Berger et al., 2003; Venos et al., 2004). The ultimate step in the Met salvage pathway seems to be often catalyzed by species-specific promiscuous aminotransferases, which have Asp and aromatic amino acids as predominant substrates (Berger et al., 1996, 2001; Heilbronn et al., 1999). In human, Met has also been found to be a substrate in transamination reactions catalyzed by a BCAT; however, the importance of this reaction has not been investigated further (Hutson, 2001).
BCAT4 Catalyzes the Initial Reaction of the Met Chain Elongation Pathway
The chain elongation pathway includes several transamination steps with either Met or the 2-oxo acids of various Met derivatives with extended side chains (Wittstock and Halkier, 2002). Thus, BCAT4 could participate in reactions with a variety of substrates (Figure 7). Among the amino acids, highest activity is found with Met (Table 1). The strong activity with this amino acid suggests that BCAT4 catalyzes the transamination step initiating the chain elongation pathway (Figure 6
, transamination 1). This conclusion is supported by the high content of SMM and the elevated Met levels in the bcat4 knockout mutants (Tables 4 and 5), which might be expected when this amino acid is the direct substrate of BCAT4. In addition, the nearly uniform reduction of the various Met-derived glucosinolates supports a function of BCAT4 in the initial Met deamination (Tables 2 and 3).
View larger version (39K):
[in this window]
[in a new window]
|
Figure 7. Comparison of Leu Biosynthesis with the Chain Elongation Pathway of Met-Derived Glucosinolate Biosynthesis.
The enzymes and numbers of genes identified in silico in the Arabidopsis genome are given in gray. For acetohydroxyacid synthase genes, two regulatory subunits and a single catalytic subunit can be detected, and for isopropylmalate isomerase genes, at least one large and three small subunits can be detected in Arabidopsis. The dashed line indicates that 5-methylthio-2-oxopentanoate (or longer 2-oxo acids) can undergo another condensation step with acetyl-CoA for further chain elongation. Light-gray box, Leu biosynthesis; dark-gray box, chain elongation pathway of Met-derived glucosinolate biosynthesis. AHAS, acetohydroxyacid synthase; KARI, ketolacid reducto-isomerase; DHAD, dihydroxyacid dehydratase; IPMS, isopropylmalate synthase; IPMI, isopropylmalate isomerase (also called dehydratase); and IPMDH, isopropylmalate dehydrogenase.
|
|
View larger version (34K):
[in this window]
[in a new window]
|
Figure 6. Model for the Compartmentalization of the Biosynthesis of Met-Derived Glucosinolates in the Flowering Stalk.
Explanations are given in the Discussion.
|
|
However, it can presently not be excluded that BCAT4 is also involved in other transamination steps of the chain elongation pathway, in which 2-oxo acids are transaminated to Met derivatives. Since the general decrease of glucosinolates might alter the composition of these secondary metabolites in unpredictable ways, the in vivo profiling data do not allow conclusions about this function. Thus, further in vitro assays with other 2-oxo acid substrates are required to get more information about the role of BCAT4 in the conversion of these substrates to the respective Met-derivatives.
Interactions of Aliphatic Glucosinolate Biosynthesis with Met and Sulfur Metabolism
As discussed in the previous section, BCAT4 catalyzes the initial transamination of Met to MTOB and acts directly at the interface between primary and secondary metabolism. BCAT4 could thus be an important checkpoint of the metabolite flux into biosynthesis of glucosinolates. This is supported by several observations. First, the localization of BCAT4 in the cytosol (Figures 5 and 6) places this enzyme next to the cytosolic forms of methionine synthase 1 (MS1) (At5g17920) and MS2 (At3g03780), which were suggested to be active in the regeneration of Met in the activated methyl cycle in Arabidopsis (Ravanel et al., 2004). In this pathway, the homocysteinyl moiety is recycled after AdoMet transmethylation. Thus, Met recycled from the AdoMet pool accounting for the major part of the Met metabolism might be an important source for the biosynthesis of the Met-derived glucosinolates. Second, when we compared the diurnal expression of BCAT4 and MAM1 with genes involved in Met biosynthesis based on recently published microarray data (Bläsing et al., 2005), we observed a synchronization of both genes with the above-mentioned cytosolic MS1 and MS1 (see Supplemental Figure 5 online). By contrast, no correlation is found with the other plastid-located Met biosynthesis genes encoding cystathionine  -synthase (CgS; At3g01120), cystathionine ß-lyase (At3g57050), and MS3 (At5g20980). Likewise, no correlation between BCAT4 and Met biosynthetic genes is observed in terms of tissue-specific or developmental expression (http://www.arabidopsis.org/info/expression/ATGenExpress.jsp; Schmid et al., 2005). Thus the spatio-temporal expression patterns do not indicate any correlation between the plastid-located de novo Met biosynthesis and the Met elongation pathway, supporting the conclusion that recycled Met in the cytosol is the preferential substrate for glucosinolates biosynthesis.
Nevertheless, the reduced flux into glucosinolate biosynthesis in the BCAT4 mutants seems to have consequences for Met biosynthesis in green tissues as indicated by the changes in the steady state levels of the respective amino acids. Besides Met and SMM, Thr also is increased in leaves. This can be explained by the well-established mechanism for the regulation of Met and Thr biosynthesis (Hesse et al., 2004). The elevated Met content might have led to an increase of AdoMet, which activates Thr synthase. In Arabidopsis at the same time, the high contents of Met or its metabolites and AdoMet inhibit CgS expression by different mechanisms, including mRNA stability (Chiba et al., 1999, 2003; Gakiere et al., 2000; Hacham et al., 2002, 2006; Kim et al., 2002; Ominato et al., 2002; Hesse et al., 2004). Since both enzymes compete for the common substrate phosphohomoserine, Thr synthase activation and CgS inhibition result in the observed elevated Thr levels (Table 4). Manipulations at this branch point often lead to increased Met and SMM levels, as observed in our studies under reduced glucosinolate biosynthesis.
These control mechanisms seem to be less effective in seeds, where enormous amounts of SMM accumulate in the BCAT4 mutants (Table 5). Here, the flux into Met metabolism has not or has only inefficiently been shut down, resulting in an enhanced generation of SMM, which is undetectable in seeds of wild-type plants (Bourgis et al., 1999). This might also keep AdoMet levels low (Kocsis et al., 2003), which is also required for SMM formation explaining the almost unaffected Thr level. The latter might also be kept low by the increased Ile formation. Difficult to explain is the increase of His, while a simultaneous increase of Lys and Met has been previously observed upon manipulation of Lys degradation in seeds (Zhu and Galili, 2003).
To unravel the flux in more detail, more metabolites have to be followed up. Particularly sulfur-containing compounds should be measured since global transcriptome analyses and integrated transciptome/metabolome approaches revealed the downregulation of MAM1 and other genes involved in glucosinolate biosynthesis and the reduced levels of glucosinolates under sulfur starvation (Hirai et al., 2003, 2004; Nikiforova et al., 2003). In addition, plants in which BCAT4 expression can be shut down by inducible RNA interference would allow further insights into the metabolic flux. Metabolic profiling in such plants at different time points after induction would document resulting changes in metabolite flux. This could be supported by parallel administration of labeled compounds allowing the tracing of metabolites even at low concentrations.
Wounding Induces Expression of BCAT4
Wounding triggers BCAT4 promoter activity in the phloem (Figure 3A). Consistent with the wound-induced twofold accumulation of the corresponding mRNA (Figure 3B), an approximate twofold induction of BCAT4 has also been observed in a recent microarray analysis of the transcriptional response to specialist and generalist herbivores (Reymond et al., 2004). This expression pattern is in line with a function of BCAT4 in Met-derived glucosinolate biosynthesis and the induction of these secondary metabolites in defense against herbivore attack. It has long been known that indole glucosinolates are induced by methyl jasmonate (MeJA) and by wounding. Also, a desulfoglucosinolate sulfotransferase that most likely operates in the biosynthesis of Trp-derived glucosinolates shows the same behavior (Doughty et al., 1995; Mikkelsen et al., 2002; Piotrowski et al., 2004; Hirai et al., 2005; Klein et al., 2006). The latter analysis also revealed a moderate increase in the accumulation of the major leaf glucosinolates 4MSOB after treatment with MeJA and the phytotoxin coronatine, a structural analog of the octadecanoid signaling molecule OPDA (Piotrowski et al., 2004). Likewise, a twofold increase of 5MSOP and an almost fivefold increase of 8-methylsulfinyloctylglucosinolate (8MSOO) and 8-methylthiooctylglucosinolate were observed after treatment with MeJA and/or other signaling molecules (Mikkelsen et al., 2003). Thus, Met-derived glucosinolate contents are also influenced by wounding, which is consistent with the wounding induction of BCAT4 and its function in the biosynthesis of these secondary metabolites.
A Second Aminotransferase Must Be Active in the Met Chain Elongation Pathway
In both BCAT4 knockout mutants, Met-derived glucosinolate accumulation is reduced to 50% of the wild-type level but not completely abolished (Tables 2 and 3). Thus, another aminotransferase is present, which can at least partially maintain the biosynthesis of these secondary metabolites. The identity of this transaminase is unclear at present, but it seems plausible that another member of the BCAT protein family can substitute this function. This enzyme would normally be responsible for the amination of the 2-oxo acids to the Met derivatives, but due to the similarity of the substrates, it might also be able to catalyze the formation of MTOB from Met. Such a side entrance into the chain elongation pathway might be of minor importance under normal conditions but becomes predominant in the bcat4 knockout.
Alternatively, another unrelated aminotransferase might be involved. A candidate protein is the previously purified Met:glyoxylate aminotransferase, with a molecular mass of 50 kD (Chapple et al., 1990). This protein, which was purified from various Brassica species, is in vitro able to catalyze the transamination of Met to MTOB. However, further in vivo studies are required to corroborate a function of this enzyme in the Met elongation pathway.
Compartmentalization of the Chain Elongation Pathway of Met
The GFP tagging experiment and the immunodetection analysis of subcellular protein fractions revealed that BCAT4 is a cytosolic enzyme (Figure 5). These experimental results are supported by the fact that this protein does not carry an N-terminal extension as do its chloroplast and mitochondrial counterparts (Diebold et al., 2002). In addition, there is no indication of the presence of candidate N- or the highly conserved C-terminal peroxisomal targeting signals, so that an additional organellar localization can be excluded. By contrast, MAM activities have only been detectable in extracts obtained from enriched chloroplasts but not in total cellular extracts of Eruca sativa (Falk et al., 2004). In line with this observation, MAM3 has been found to cofractionate with chloroplasts (S. Textor, J.-W. de Kraker, B. Knoke, S. Schuster, J. Gershenzon, and J. Tokuhisha, unpublished data). Thus, two consecutive steps in the Met chain elongation pathway are likely distributed over two different subcellular compartments and require transport of MTOB into the chloroplast (Figure 6). There MTOB is further metabolized by the condensation reaction catalyzed, for instance, by MAM1, and after two further reaction steps, the chain-elongated 2-oxo acid is also most likely synthesized in these organelles (Figure 6; see discussion below). The 2-oxo acid can now either remain in the cycle and undergo further chain elongations or leave the cycle by one of two scenarios. First, the 2-oxo acid is transaminated in the chloroplast and then exported. This plastid-hosted transamination reaction could be catalyzed by another aminotransferase, possibly another member from the BCAT protein family (Figure 6, transamination 2). Second, the 2-oxo acid is exported into the cytosol and transaminated to its corresponding Met derivative (Figure 6, transamination 3). This could be catalyzed by BCAT4 at least for MTOP, as indicated by the in vitro assays. This scenario would be consistent with the observation that the labeled amino group of Met is retained in the pathway, since the deamination of Met could be directly coupled with the amination of chain-elongated 2-oxo acids after one or several cycles (Graser et al., 2000). However, in vitro MTOH is not a substrate for this enzyme, which makes it unlikely that substrates with longer methylene chains are converted by this pathway. In this case, another cytosolic aminotransferase would be required. Regardless of the localization of the transamination reaction, both scenarios require participation of another aminotransferase whose identity so far remains elusive.
In summary, it seems clear that the biosynthesis of Met-derived glucosinolates is distributed over several subcellular compartments. BCAT4 is found in the cytosol, the MAM enzymes and probably other enzymes of the chain elongation pathway are most likely located in plastids, while the further metabolization of the Met derivatives occurs at the endoplasmatic reticulum/cytosol interface (Figure 6; Mikkelsen et al., 2002; Grubb and Abel, 2006). In addition, in flowering stalks, biosynthesis and storage of glucosinolates is accomplished in different cell types (Figure 6). While storage is suggested in S-cells (Koroleva et al., 2000), biosynthesis goes on in the neighboring phloem. This is supported by our studies of BCAT4 (Figure 2) and by the microarray analysis of respective tissues in Arabidopsis root hypocotyls (Zhao et al., 2005). This approach confirmed the phloem-specific expression of BCAT4 and also revealed a phloem-predominant mRNA accumulation of MAM1 and cytochromes P450, CYP79F1, and F2, proteins operating in the biosynthesis of this class of glucosinolates (Chen et al., 2003; Tantikanjana et al., 2004).
The Met Chain Elongation and Leu Biosynthetic Pathways Have a Common Ancestor
MAM1 (At5g23010) and MAM3 (At5g23020) have so far been the only enzymes unambiguously assigned to the Met chain elongation pathway. These proteins are encoded by a small gene family of isopropylmalate synthase–like proteins, one of which (At1g74040) is most likely involved in Leu biosynthesis (Figure 7
; Field et al., 2004). In this work, BCAT4 is found to be another important enzyme of the Met chain elongation pathway. This gene is like the MAM genes, a member of a gene family that also encodes proteins participating in Leu biosynthesis. In addition, the types of reactions occurring in both pathways are identical, including an isomerization and an oxidative decarboxylation, a similarity that was noticed several decades ago (Strassman and Ceci, 1963; Underhill et al., 1973). Now we know that at least three enzymes of the Met chain elongation pathway are encoded by gene families that also code for enzymes of the Leu biosynthesis pathway. These findings suggest that the proteins catalyzing the isomerization and the oxidative decarboxylation steps are members of analogous families as well. This inference is supported by the observation that multiple gene families only exist in the branch toward Leu, while the enzymes operating in the early common steps in BCAA biosynthesis are encoded by single genes (Figure 7). Moreover, these observations strongly support the hypothesis that the complete chain elongation pathway and the complete Leu biosynthesis pathway have a common ancestor. Considering that all plant species have the potential to synthesize Leu, which is thus the evolutionary older reaction chain, it is tempting to argue that in those plant species possessing Met-derived aliphatic glucosinolates the chain elongation pathway has evolved from the Leu biosynthesis pathway. To obtain more data supporting this hypothesis, different isopropylmalate isomerases and isopropylmalate dehydrogenases have to be investigated to determine whether they are part of the chain elongation pathway, whether they are involved in Leu biosynthesis, or possibly in both.
|
METHODS
|
|---|
Plant Cultivation
Arabidopsis thaliana plants ecotype Col-0, including the mutants, were grown on standard soil supplemented with Osmocote Exact Mini (Scotts) and Triabon in a growth chamber under a 16-h-light (160 to 200 µmol/m2s)/8-h-dark regime at 23°C. Plants used for the investigation of diurnal expression were cultivated under a 12-h-light/12-h-dark rhythm under otherwise identical conditions.
Enzyme Activity and Substrate Specificity Assays
For overexpression of BCAT4, the corresponding cDNA was amplified with primers bcat4UE.H (5'-AGGAATTCATGGCTCCTTCTGCGC-3') and bcat4UE.R (5'-ACCTCGAGTCAGCCCTGGCGGTC-3') on a cDNA clone described previously (Diebold et al., 2002). PCR was performed with the BD Advantage 2 PCR enzyme system (BD Bioscience) with 35 cycles, and the resulting cDNA fragment of 1063 nucleotides was digested with BamHI and XhoI, cloned into the respective sites in pET32a (Novagen), and completely sequenced using an ALF Express sequencer following standard procedures given by the manufacturer (Amersham Biosciences). Overexpression and enrichment on S-protein agarose, including the release of the recombinant BCAT4 protein from the N-terminal S-Tag, was performed according to the instructions given in the manual (Novagen). The amount of recombinant BCAT4 protein, which contains only a few additional amino acids at the N terminus, in the enriched fractions was determined in several independent enzyme preparations (see Supplemental Figure 2 online). These were separated by SDS-PAGE and stained with Coomassie Brilliant Blue. The amount of recombinant BCAT4 was then measured and calculated with the Fluor-S MultiImager and Multi-Analyst software (Bio-Rad). The percentage of recombinant BCAT4 in the enriched protein fraction was found to be 18.7% (±4.0) of the total protein in the enriched fraction.
Kinetic studies with MTOB and activity assays with other 2-oxo acids were performed as described previously (Schadewaldt and Adelmeyer, 1996; Schuster and Binder, 2005). Kinetic studies of recombinant BCAT4 with Met and Leu and the analysis of the relative substrate specificities for different amino acids were performed with an HPLC-based assay (Jones and Gilligan, 1983; Schuster and Binder, 2005) with the modification that the amount of the coproduct Glu was measured. In the transamination reaction, Glu is generated in a 1:1 stoichiometry with the conversion of the amino acid to the respective 2-oxo acid (e.g., Met to MTOB). The correlation of the reaction velocity with the substrate concentration followed a typical Michaelis-Menten kinetic after the data were fitted into a nonlinear regression curve using Origin 7.0 software (OriginLab). All substrates if not otherwise mentioned were purchased from Sigma-Aldrich. MTOP and MTOH were synthesized by Applichem.
Nucleic Acid Analyses
Total RNA was isolated from 12-d-old plants using Plant RNeasy kits according to the manual supplied by the manufacturer (Qiagen). For RNA gel blot analysis, 10 µg/per lane of total RNA isolated as described above was separated by glyoxal agarose gel electrophoresis following standard protocols (Sambrook et al., 1989). Size fractionated RNAs were then transferred to Duralon UV membranes (Stratagene) and hybridized with a cDNA probe labeled with the Rediprime II random prime labeling system (Amersham Biosciences) according to the instructions given by the manufacturer. The cDNA probe representing the first three exons was amplified with primers bcat4UE.H (5'-AGGAATTCATGGCTCCTTCTGCGC-3') and bcat4-5'.R (5'-TGATCGACCGAAGGATAAGGC-3').
Wounding Experiments
Transgenic Arabidopsis plants containing promoter:GUS reporter gene constructs were established and histochemically analyzed as described before (Hull and Devic, 1995; Schuster and Binder, 2005). For wounding experiments, leaves of 20-d-old plants grown as outlined above were treated by squeezing with forceps, piercing with a needle, or by cutting off the tip with scissors. Leaves were separated from the plants 5 min after treatment and subsequently histochemically analyzed. For the RNA gel blot analysis, the apical half of the leaves was squeezed with forceps along the center vasculature, and the complete leaves were cut off from the plant at the respective time points. RNA was isolated as described above.
Glucosinolate and Amino Acid Profiling
Glucosinolate content and composition were analyzed in seeds and leaves as described previously (Brown et al., 2003). Profiling of free amino acids in seeds and leaves was done by HPLC-based precolumn derivatization with o-phthalaldehyde/mercaptoethanol following a protocol described previously (Roth, 1971; Sarwar and Botting, 1993). The leaves were harvested from about 3-week-old plants prior to the development of an inflorescence. SMM was identified by comparing the retention time of a corresponding peak in the total free amino acid analysis with the peak observed in a control run with SMM obtained from Sigma-Aldrich. In addition, nonderivatized SMM was identified in the total free amino acid pool from seeds by HPLC tandem mass spectrometry according to a method described previously (van Dam et al., 2003).
Immunodetection Analysis and Subcellular Fractionation
Total soluble protein was isolated from 14- to 21-d-old plants by the following protocol. Approximately 1 to 2 g of leaf material was ground in liquid nitrogen, and the tissue powder was dissolved in 100 mM Tris-HCl, pH 8.0. The solution was centrifuged for 10 min at 12,000g, and the volume of the supernatant was reduced in Amicon Ultra centrifugal filter devices (10 kD; Millipore). Cytosol protein fractions were purified from protoplasts obtained from leaves of 13- to 21-d-old Arabidopsis plants according to methods described previously (Sakamoto et al., 2000; Dovzhenko et al., 2003). Chloroplasts were purified by differential centrifugation and purification on Percoll gradients as outlined previously (Aseeva et al., 2004). Essentially the same procedure with parameters optimized for mitochondria was applied to isolate these organelles from Arabidopsis cell suspension culture (Klein et al., 1998). The BCAT4 antibody was raised against the final dodecapeptide at the C terminus of BCAT4. The antibody was affinity purified using 100 µg ovalbumin-peptide conjugate, which was bound to nitrocellulose (1 cm2) for 30 min at room temperature. The membrane was blocked five times with blocking buffer (1x PBS, 0.1% [v/v] Tween 20, and 3% [w/v] BSA) for 2 min each at room temperature and then incubated in 2 mL antiserum diluted 1:10 with blocking buffer at 4°C overnight. The membrane was again washed three times in blocking buffer, and antibodies were released by an incubation in 1 mL elution buffer (0.2 M glycine-HCl, pH 3.0, 0.5 M NaCl, 1 mg/mL BSA, and 0.02% [w/v] NaN3) for 1 min at room temperature. The solution containing the antibody was neutralized by adding 100 µL 2 M Tris-HCl, pH 8.0, and used in a 1:10 dilution in the protein gel blot analysis. Antibodies against FNR and porin were used in 1:5000 and 1:2000 solution. The anti-UDP-glucose pyrophosphorylase, which is raised against the respective protein from poplar, was purchased from Agrisera and used in a 1:3000 dilution. For immunodetection analysis, 25 µg of total soluble protein and 150 µg from the subcellular fractions were analyzed with anti-BCAT4, anti-porin, and anti-UDP-glucose pyrophosphorylase antibodies, and 10 µg of total and 60 µg of the subcellular fractions were analyzed with the anti-FNR antibody. All immunodetection analyses were done following standard electrophoresis and blotting techniques (Sambrook et al., 1989) and the ECL detection system following the instructions given by the manufacturer (Amersham Biosciences).
Miscellaneuos Methods
For the analysis of the subcellular localization of BCAT4, the cDNA fragment containing the complete BCAT4 reading frame (354 amino acids without stop codon) was amplified by PCR using primers BCAT4GFP.H and BCAT4GFP.R on a respective cDNA clone. The resulting fragment was directly cloned into pGEM-T and confirmed by sequencing. The insert was removed from this plasmid by digestion with XbaI and SmaI and cloned in frame upstream of the GFP gene in the vector psmGFP4 (Davis and Vierstra, 1998). Transformation of tobacco (Nicotiana tabacum) protoplasts and fluorescence microscopy were done as described before (Koop et al., 1996; Däschner et al., 2001; Diebold et al., 2002). All standard methods used were performed according to established protocols (Sambrook et al., 1989).
Accession Numbers
Sequence data from this article can be found in the GenBank/EMBL data libraries under accession numbers AJ271732 (BCAT4) and AY070471 (MAM1).
Supplemental Data
The following materials are available in the online version of this article. - Supplemental Figure 1. Purification of Recombinant BCAT4.
- Supplemental Figure 2. Kinetic of BCAT4 with MTOB.
- Supplemental Figure 3. Kinetic of BCAT4 with Met.
- Supplemental Figure 4. Kinetic of BCAT4 with Leu.
- Supplemental Figure 5. Time-Resolved Expression of BCAT4, MAM1, and Genes Involved in Met Biosynthesis.
|
Acknowledgments
|
|---|
We thank C. Guha and B. Weber for excellent technical assistance, O. Bläsing for his help with in silico data mining, F. Ossenbühl for his help in chloroplast isolation, and A. Svato for the liquid chromatography–mass spectrometry analysis. We also thank A. Sokolenko and H.P. Braun for the kind gift of antibodies against FNR and porin. This work was supported by a fellowship of the Studienstiftung des Deutschen Volkes to J.S., a fellowship according to the Landesgraduiertenförderungsgesetz Baden-Württemberg to T.K., a start-up grant from the Rudolph und Clothilde Eberhardt-Stiftung, the Deutsche Forschungsgemeinschaft (Ge 1126/1-3 and Bi 590/9-1), and the Max Planck Society.
|
Footnotes
|
|---|
The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantcell.org) is: Stefan Binder (stefan.binder{at}uni-ulm.de).
[W] Online version contains Web-only data. 
www.plantcell.org/cgi/doi/10.1105/tpc.105.039339
Received November 8, 2005;
Revision received August 16, 2006.
accepted September 27, 2006.
|
REFERENCES
|
|---|
|
TOP
ABSTRACT
INTRODUCTION