Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
Arabidopsis mutants in the C–S lyase of glucosinolate biosynthesis establish a critical role for indole-3-acetaldoxime in auxin homeostasis
Michael Dalgaard Mikkelsen,
Plant Biochemistry Laboratory, Department of Plant Biology, The Royal Veterinary and Agricultural University, Thorvaldsensvej 40, DK-1871 Frederiksberg C, Copenhagen, Denmark, and Center for Molecular Plant Physiology (PlaCe), The Royal Veterinary and Agricultural University, Thorvaldsensvej 40, DK-1871 Frederiksberg C, Copenhagen, Denmark
For correspondence (fax +45 35 28 33 33; e-mail firstname.lastname@example.org).
Present address: Department of Medicinal Chemistry, The Danish University of Pharmaceutical Sciences, 2 Universitetsparken, DK-2100 Copenhagen Ø, Denmark.
We report characterization of SUPERROOT1 (SUR1) as the C–S lyase in glucosinolate biosynthesis. This is evidenced by selective metabolite profiling of sur1, which is completely devoid of aliphatic and indole glucosinolates. Furthermore, following in vivo feeding with radiolabeled p-hydroxyphenylacetaldoxime to the sur1 mutant, the corresponding C–S lyase substrate accumulated. C–S lyase activity of recombinant SUR1 heterologously expressed in Escherichia coli was demonstrated using the C–S lyase substrate djenkolic acid. The abolishment of glucosinolates in sur1 indicates that the SUR1 function is not redundant and thus SUR1 constitutes a single gene family. This suggests that the ‘high-auxin’ phenotype of sur1 is caused by accumulation of endogenous C–S lyase substrates as well as aldoximes, including indole-3-acetaldoxime (IAOx) that is channeled into the main auxin indole-3-acetic acid (IAA). Thereby, the cause of the ‘high-auxin’ phenotype of sur1 mutant resembles that of two other ‘high-auxin’ mutants, superroot2 (sur2) and yucca1. Our findings provide important insight to the critical role IAOx plays in auxin homeostasis as a key branching point between primary and secondary metabolism, and define a framework for further dissection of auxin biosynthesis.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
The phytohormone auxin is important in virtually all aspects of plant growth and development, including phototropism, cell expansion, vascular differentiation, lateral root development, root gravitropism, and apical dominance. Although auxin was the first phytohormone discovered over 100 years ago, our understanding of its biosynthesis, an essential part of auxin homeostasis, remains limited (for review, see Cohen et al., 2003; Ljung et al., 2002). The classical genetic approach of screening for auxin-deficient mutants has been unsuccessful. This is most likely because multiple redundant pathways are involved in auxin biosynthesis (for review, see Bartel et al., 2001; Normanly and Bartel, 1999). However, three ‘high-auxin’ mutants have been isolated in genetic screens using the model plant Arabidopsis thaliana. The first of these mutants to be isolated was the recessive superroot1 (sur1; Boerjan et al., 1995), allelic to aberrant lateral root formation1 (alf1; Celenza et al., 1995), rooty (rty; King et al., 1995), and hookless3 (hls3; Lehman et al., 1996). Later, the recessive mutant superroot2 (sur2; Barlier et al., 2000; Delarue et al., 1998), allelic to rnt1 (Bak et al., 2001; Winkler et al., 1998), was identified. In the following, these mutants will be referred to as sur1 and sur2. Despite substantial efforts, the function of SUR1 has remained elusive. SUR2 encodes the cytochrome P450, CYP83B1, which is responsible for oxidizing indole-3-acetaldoxime (IAOx) in the biosynthesis of indole glucosinolates (see below). The third ‘high-auxin’ mutant is the dominant activation-tagged yucca1, which overexpresses the flavin monooxygenase-like YUCCA proposed to be involved in tryptophan-dependent indole-3-acetic acid (IAA) biosynthesis (Zhao et al., 2001). The very well-characterized sur1 mutant has been isolated in several independent ‘root morphology’ screens designed to identify mutants phenocopying the wild-type grown on 2,4-D (Boerjan et al., 1995), mutants with altered root formation (Celenza et al., 1995), and mutants with variations in number of adventitious roots on etiolated hypocotyls (King et al., 1995). In addition, a sur1 mutant was isolated in a screen aimed at finding ethylene response mutants based on lack of the apical hook (Lehman et al., 1996). The characteristic ‘high-auxin’ phenotype of sur1 includes excessive adventitious and lateral root formation, epinastic cotyledons, as well as increased endogenous levels of IAA, the main auxin. Several functions have been proposed for SUR1. Based on the increased lateral root formation of sur1 mutant, it was suggested that SUR1 could function as a negative regulator of lateral root formation (Celenza et al., 1995). Identification of the SUR1 gene showed that the deduced protein sequence shared 70% identity with a tyrosine transaminase-like protein (Gopalraj et al., 1996). Furthermore, based on measurements of increased indole-3-acetaldehyde oxidase activity in the sur1 mutant, SUR1 was suggested to transaminate tryptophan to indole-3-pyruvic acid with resulting accretion of indole-3-acetaldehyde to IAA, thereby gaining its ‘high-auxin’ phenotype (Seo et al., 1998).
Glucosinolates are amino acid-derived natural plant products found in the Capparales order, which includes the model plant Arabidopsis. Glucosinolates are implicated in plant defense and have attracted much attention as cancer-preventing agents, biopesticides, and flavor compounds (Halkier, 1999). Recently, major advances have been made with respect to identification of genes in the biosynthetic pathway of glucosinolates (for review, see Wittstock and Halkier, 2002). The first step in biosynthesis of the core glucosinolate structure is conversion of amino acids to the corresponding aldoximes by cytochromes P450 belonging to the CYP79 family (for review, see Mikkelsen et al., 2002). Subsequently, CYP83B1 and CYP83A1 oxidize, respectively, aromatic aldoximes (including IAOx) and aliphatic aldoximes, which are subsequently conjugated to a sulfur donor, most likely cysteine (Wetter and Chisholm, 1968). The C–S bond is then cleaved by a C–S lyase yielding the thiohydroximic acid that is glycosylated and sulfonated to produce the core glucosinolate structure (for review, see Halkier, 1999). Previous efforts directed at identifying the C–S lyase have been unsuccessful. Identification of this enzyme is important for completion of the glucosinolate biosynthetic pathway as well as for metabolic engineering.
In this paper, we identify SUR1 as the C–S lyase in glucosinolate biosynthesis as evidenced by selective metabolite profiling, in vivo labeling experiments, and biochemical characterization of recombinant SUR1. The role of SUR1 in auxin homeostasis is discussed.
Recently, major breakthroughs have resulted in identification of most of the genes involved in biosynthesis of the core glucosinolate structure. However, the C–S lyase responsible for converting S-(alkylacetohydroximoyl)-l-cysteines, in the following referred to as C–S lyase substrates, to the corresponding thiohydroximic acids has remained unidentified. We have taken a bioinformatics approach to identify putative C–S lyases involved in glucosinolate biosynthesis. Three proteins with confirmed C–S lyase activities from rat, human, and zebrafish were aligned and a consensus sequence was constructed. The consensus sequence was used as probe in a blast search against total Arabidopsis proteins. One of the lower scoring hits (E-value: 2E−13) was SUR1 (At2g20610) annotated as a putative aminotransferase. The sequence similarity between aminotransferases and C–S lyases is reflected by the similar reaction mechanism for these enzymes. Both types of enzymes convert amino acids to α-keto acids upon release of the amino group either as free ammonia with concomitant pyruvate release (C–S lyase) or onto a receiving α-keto acid (aminotransferase). Furthermore, both enzymes require pyridoxal-5′-phosphate (PAPS) as co-factor. Although the homology to other eukaryotic C–S lyases was not very high, the ‘high-auxin’ phenotype of the sur1 mutant made it a likely candidate for the C–S lyase in glucosinolate biosynthesis (see Discussion).
In vivo feeding studies were performed to investigate the metabolism of radiolabeled p-hydroxyphenylacetaldoxime, the tyrosine-derived aldoxime, in seedlings of sur1 and the wild-type (Figure 1). We chose to use p-hydroxyphenylacetaldoxime as we have previously characterized its metabolism by CYP83B1 in detail (Hansen et al., 2001) and as Arabidopsis efficiently converts it to the corresponding glucosinolate, p-hydroxybenzyl glucosinolate (p-OHBG; Bak et al., 1999). Following feeding and incubation overnight, radiolabeled metabolites in the plant extracts were analyzed by TLC. In the wild-type, p-hydroxyphenylacetaldoxime was metabolized to p-OHBG as expected. In sur1, radiolabeled p-OHBG was not detected, but a compound co-migrating with an authentic standard of the tyrosine-derived C–S lyase substrate S-(p-hydroxyphenylacetohydroximoyl)-l-cysteine accumulated. When radiolabeled tryptophan or IAOx was administered to the sur1 mutant, a compound co-migrating with the tryptophan-derived C–S lyase substrate accumulated and no radiolabeled indole glucosinolates was detected, whereas in wild-type indole glucosinolates and not the compound co-migrating with the tryptophan-derived C–S lyase substrate standard accumulated (data not shown).
HPLC analyses were performed to investigate how production of glucosinolates was affected by the sur1 mutation. As reference material, we analyzed wild-type plants that were grown on 1 µm of the artificial auxin 2,4-D to phenocopy the ‘high-auxin’sur1 phenotype. In 5-week-old sur1 mutant plants, neither indole, aromatic nor short or long chain aliphatic glucosinolates were detected. However, faint traces of indol-3-ylmethyl glucosinolate (i-3ym) could be detected occasionally (Figure 2). The lack of glucosinolate accumulation in sur1 suggests that it is defective in glucosinolate biosynthesis.
The specific effects of high auxin levels on glucosinolate accumulation were investigated by growing the wild-type on 0.2 µm 2,4-D, in which case plant morphology was not affected to the extent same as that when grown on 1 µm 2,4-D. Some differences in accumulation of individual glucosinolates were observed (Figure 3a). Specifically, i-3ym accumulation was increased almost twofold from 2.9 ± 0.2 to 5.3 ± 0.8 nmol mg−1 in leaves of wild-type plants grown on 2,4-D, as was 8-methylsulfinyloctyl glucosinolate (8-mso) accumulation in both roots and leaves (0.5 ± 0.1 to 1.7 ± 0.3 nmol mg−1 and 0.7 ± 0.1 to 2.5 ± 0.2 nmol mg−1 for roots and leaves, respectively). Conversely, the level of 2-phenylethyl glucosinolate (2-phee) in roots was reduced in plants grown on 2 µm 2,4-D from 3.4 ± 0.7 to 1.9 ± 0.2 nmol mg−1. This shows that although auxin specifically modulates the level of individual glucosinolates, the overall level of glucosinolates is largely unaffected. Furthermore, the data demonstrate that high auxin levels per se do not inhibit glucosinolate biosynthesis and that the abolishment of glucosinolates in the sur1 mutant is not a direct consequence of high auxin levels.
A time course study of the glucosinolate content in sur1 was performed. The sur1 mutation is lethal, and therefore sur1 plants used for analysis were derived from heterozygous parent plants. In 1-week-old sur1 seedlings, significant amounts of all major glucosinolates were detected (Figure 3b). The levels found were on average 26% compared to the wild-type, although 3-hydroxypropyl glucosinolate was present in levels comparable to the wild-type. In 2-week-old sur1 mutants, the levels of 4-hydroxyindol-3-ylmethyl glucosinolate, 8-mso, 4-methoxyindol-3-ylmethyl glucosinolate, 2-phee, and N-methoxyindole glucosinolate were reduced to trace levels. In 5-week-old sur1 mutants, only faint traces of i-3ym could be detected. The time-dependent reduction in levels of all glucosinolates shows that no de novo biosynthesis of these compounds takes place in the sur1 mutant. As the sur1 mutation is lethal, the glucosinolates present in young sur1 seedlings most likely originate from the seed of the heterozygous parent plant. This suggests that time-dependent reduction of glucosinolates in sur1 reflects a combination of turn-over and dilution of the inherent seed glucosinolates as plant biomass increases.
The catalytic activity of SUR1 was investigated by heterologous expression of SUR1 in Escherichia coli. Analysis of cell lysates by SDS–PAGE showed a strong band of 50 kDa corresponding to SUR1 (Figure 4). To characterize the SUR1 catalytic activity, we set up a coupled assay with the commonly used C–S lyase substrate djenkolic acid. In the assay, C–S lyase activity releases pyruvate from djenkolic acid. Subsequently, pyruvate can be reduced by lactate dehydrogenase in an NADH-dependent manner. This can be quantified as oxidation of NADH at 340 nm. Thus, C–S lyase activity can be measured in a coupled assay by djenkolic acid-dependent oxidation of NADH (Figure 5). When desalted lysates from E. coli expressing SUR1 were assayed for C–S lyase activity toward the substrate djenkolic acid, oxidation of NADH was observed (Figure 5). Substrate independent oxidation of NADH was eliminated by desalting the lysates on G-25 gel filtration columns. This demonstrates that SUR1 possesses C–S lyase activity. No djenkolic acid-dependent activity was detected in lysates from control cells. The recombinant SUR1 protein was very unstable as evidenced by loss of measurable activity in the cell-free lysates after a few hours. We could not detect any metabolism of the tyrosine- or the tryptophan-derived C–S lyase substrates. These natural substrates for the C–S lyase have to be generated in situ in the reaction mixtures by recombinant CYP83B1 and cysteine as they rapidly undergo internal cyclization (see Discussion).
In the present report, we demonstrate that SUR1 is the C–S lyase in glucosinolate biosynthesis as evidenced by complete absence of glucosinolates concurrent with accumulation of C–S lyase substrates in the sur1 mutant. Furthermore, C–S lyase activity of recombinant SUR1 was demonstrated.
The complete abolishment of both aliphatic and indole glucosinolates as observed in sur1 has not previously been reported for any mutant. This indicates that the SUR1 function is not redundant and that SUR1 constitutes a single gene family. Although increased levels of 2,4-D in the media modulates the glucosinolate composition with respect to individual glucosinolates, the overall glucosinolate concentration is largely unaffected. Previously, it has been indicated that expression of the genes encoding the tryptophan-metabolizing CYP79B2 and CYP79B3 is induced in Arabidopsis grown on 1 µm 2,4-D, leading to increased accumulation of indole glucosinolates (Mikkelsen et al., 2003). In sur1 where auxin levels are high, CYP79B2 and CYP79B3 are most likely induced, thereby increasing IAOx production and enhancing the superroot phenotype. Conversely, Arabidopsis overexpressing SUR1 displays no apparent phenotype and contains wild-type levels of all glucosinolates (M. Nafisi, B.A. Halkier, M. Gopalraj, N. Olszewski, unpublished data). This is most likely because SUR1 is not rate-limiting under physiological conditions and because SUR1 is unlikely to increase the flow toward glucosinolate biosynthesis as a result of its downstream position in the pathway (Figure 6).
C–S lyase activity of recombinant SUR1 was demonstrated by employing the commonly used substrate djenkolic acid. The endogenous substrates, S-(alkylacetohydroximoyl)-l-cysteines, are derived from oxidized aldoximes produced by CYP83B1 or CYP83A1 that are conjugated to a sulfur donor, most likely cysteine (Bak et al., 2001; Hansen et al., 2001; Naur et al., 2003). The endogenous substrates are unstable as they undergo spontaneous internal cyclization to produce the corresponding R-2-(alkyl)thiazoline-4-carboxylic acids. Accordingly, they would have to be synthesized in situ to function as substrates for the C–S lyase. The products of the CYP83 reaction are highly reactive and excess cysteine has to be present to out-compete other nucleophiles. Unfortunately, cysteine spontaneously reacts with PAPS, thereby inhibiting C–S lyase activity (Nock and Mazelis, 1986). In vitro attempts to demonstrate C–S lyase activity using the endogenous substrates were unsuccessful. In vivo, the CYP83 enzymes and the C–S lyase have to form a tightly coupled complex to prevent cyclization of the endogenous C–S lyase substrates. This complex may possibly also contain a glutathione-S-transferase type of enzyme to ensure proper conjugation with the sulphur donor.
Based on homology, aminotransferases may be evolutionarily related to groups of C–S lyases. C–S lyases may therefore occasionally be incorrectly characterized as aminotransferases, as evidenced for SUR1 (Gopalraj et al., 1996) and for the cysteine lyase CORI3 (Jones et al., 2003). The similar reaction mechanisms for aminotransferases and C–S lyases support an evolutionary relationship. Both reaction mechanisms involve binding of the α-amino group to the PAPS co-factor, which for C–S lyases results in rearrangement of bonds in the substrate molecule, leading to cleavage of the C–S bond and subsequent release of pyruvate. The presence of the α-hydrogen atom as well as the free amino group is crucial for C–S lyase activity, whereas the substituent on the S atom is of little importance for the catalytic activity (Schwimmer and Kjær, 1960). This is in agreement with the general observation that the post-aldoxime enzymes in the glucosinolate pathway have high specificity for the functional group and little specificity for the side chain (for review, see Halkier, 1999).
It is generally believed that glucosinolates are evolutionarily related to the cyanogenic glucosides as both groups of natural products are derived from amino acids, which are first converted to the corresponding aldoximes by CYP79s and then metabolized by the evolutionarily related CYP71s and CYP83s for, respectively, cyanogenic glucoside and glucosinolate biosynthesis. It has been suggested that the enzymes catalyzing the subsequent steps in glucosinolate biosynthesis have been recruited from the detoxification processes such as, e.g. the sulpho- and glucosyltransferases (for review, see Mikkelsen et al., 2002). However, the homology to aminotransferases indicates that SUR1 (Gopalraj et al., 1996), as well as other C–S lyases (Jones et al., 2003), has evolved from those.
Our data indicate that the ‘high-auxin’ phenotype of sur1 is caused by accumulation of endogenous C–S lyase substrates as well as aldoximes, including IAOx that is channeled into IAA. Thereby, sur1 partly resembles the sur2 mutants, which are blocked in the IAOx-metabolizing step in biosynthesis of indole glucosinolates. Although sur2 and sur1 have similar phenotypes at the seedling stage, sur2 is able to overcome its phenotype and produce viable seeds, whereas sur1 is not. This difference is likely because of the presence of CYP83A1 that is able to metabolize IAOx, thereby alleviating the ‘high-auxin’ phenotype (Bak and Feyereisen, 2001; Naur et al., 2003; Figure 6). The presence of CYP83A1 explains why the sur2 mutation results in an approximately 50% decrease of indole glucosinolates (Bak et al., 2001), whereas sur1 does not produce detectable levels of any glucosinolates. Recent expression data confirm that CYP83A1 is upregulated in 5-week-old sur2 knockout mutants (Naur et al., 2003).
The dominant ‘high-auxin’ mutant yucca is mutated in a gene encoding a flavin-monooxygenase-like enzyme. Biochemical data indicate that YUCCA is able to catalyze N-hydroxylation of tryptamine (Zhao et al., 2001) that may subsequently be converted to IAOx. Consequently, all three ‘high-auxin’ mutants gain their phenotypes through IAOx accumulation by either post-aldoxime blockage of the biosynthetic pathway of glucosinolates or activating an upstream biosynthetic gene. The IAOx intermediate in indole glucosinolate biosynthesis is produced from tryptophan exclusively by the functionally redundant CYP79B2 and CYP79B3 (Hull et al., 2000; Mikkelsen et al., 2000). This is evidenced by the lack of tryptophan-derived indole glucosinolates in the cyp79B2::cyp79B3 double knockout (Zhao et al., 2002), which shows that YUCCA-dependent IAOx formation does not contribute significantly to indole glucosinolate biosynthesis. Outside the five indole glucosinolate-producing families in the Capparales order (Griffiths et al., 2001), no CYP79B homologs have been identified. This indicates that a possible role of CYP79B homologs in IAA biosynthesis is limited to a subgroup of plants, and suggests that future genetic screens aimed at identifying genes or regulators of IAA biosynthesis should be carried out in species or mutants that do not produce indole glucosinolates, e.g. in the cyp79B2::cyp79B3 double knockout. In conclusion, identification of SUR1 as the C–S lyase in glucosinolate biosynthesis establishes a critical role for IAOx in IAA homeostasis as a key branching point between primary and secondary metabolism and defines a framework for further dissecting auxin biosynthesis.
A consensus sequence was constructed using the program vectornti (Informax Inc., MD, USA) from three protein sequences with known C–S lyase activity from Homo sapiens, Rattus norwegicus, and Takifugu rubripes, with Accession numbers NP_004050, AAB26845, and CAB44334, respectively. The consensus sequence was used as a probe in a blast search against the A. thaliana sub-base at the National Center for Biotechnology Information (NCBI) homepage (http://www.ncbi.nlm.nih.gov/BLAST/).
Plants were grown on 0.5× MS agar plates containing 3% (w/v) sucrose in a controlled environment Arabidopsis chamber (Percival AR-60 I, Boone, IA, USA) at a photosynthetic flux of 100–120 nmol photons m−2 sec−1 at 20°C and 70% relative humidity at a 16-h photoperiod. Wild-type plants were grown on 1 µm 2,4-D to phenocopy sur1. Plants were analyzed at 1–5 weeks of age. Glucosinolate extraction, analysis, and quantification were performed as described previously by Petersen et al. (2002).
In vivo feeding experiments
Three-week-old seedlings of the wild-type and sur1 were immersed in 200 µl H2O containing 100 nCi of, respectively, [U-14C]p-hydroxyphenylacetaldoxime (specific activity 440 mCi mmol−1, NEN, Boston, MA, USA), and [side chain-3C-14C]-IAOx and [side chain-3C-14C]-tryptophan (specific activity 49 mCi mmol−1; NEN), followed by incubation overnight under saturating light. Metabolites were extracted by boiling in 70% methanol and separated by TLC in 7 : 1 : 2 isopropanol:ethylacetate:H2O. Authentic radiolabeled C–S lyase substrates and glucosinolate standards were applied. Radiolabeled bands were visualized on a storm 840 phosphoimager (Amersham Biosciences AB, Uppsala, Sweden) and quantified with imagequant analysis software (Amersham Biosciences AB). Radiolabeled p-hydroxyphenylacetaldoxime and IAOx were synthesized from 14C-labeled tyrosine and tryptophan by recombinant CYP79A1 (Halkier et al., 1995) and CYP79B2 (Mikkelsen et al., 2000), respectively. The radiolabeled tyrosine-derived C–S lyase substrate S-(p-hydroxyphenylacetohydroximoyl)-l-cysteine was synthesized from p-hydroxyphenylacetaldoxime by recombinant CYP83B1 and cysteine as described previously by Hansen et al. (2001).
Heterologous expression of SUR1
The SUR1 coding sequence (At2g20610) was amplified from plasmid cDNA (kindly provided by Prof. N. Olszewski) by primers S1F: AATAACACCATGGCTAGCGAAGAACAACCACACGCC and S1R: AATAACAGGATCCATTACATTTCGAGATTATTATCACTCAG, thereby incorporating 5′-NcoI and 3′-BamHI restriction sites. The PCR product was purified, digested with NcoI and BamHI, ligated into a similarly digested pET9D vector, and sequenced to exclude PCR errors.
The expression construct was transformed into E. coli strains BL-21 (RIL) and PR745. Cultures were grown at 28°C, 250 r.p.m. for 16 h in Luria Bertani (LB) medium containing 200 µm PAPS, with and without 1 mm isopropyl-beta-D-thiogalactopyranoside (IPTG) for PR745 and BL-21 (RIL) cells, respectively.
The cells were harvested and re-suspended in 0.2 volume 500 mm NaCl, 10% glycerol, 150 mm phosphate buffer, pH 7.8, containing 200 µm PAPS. Following addition of 100 µg ml−1 lysozyme and 15 min incubation on ice, the samples were lyzed by sonication. Lysates were collected after centrifugation at 10 000 g, 15 min, 4°C, and desalted using a Sephadex G-25 gel filtration spin column.
Measurement of C–S lyase activity
C–S lyase activity was measured according to Giovanelli and Mudd (1971) with minor adjustments. Briefly, a typical reaction mixture contained 250 mm NaCl, 75 mm phosphate buffer, pH 7.8, 100 µm PAPS, 5 U lactate dehydrogenase, 250 µm NADH, and 100 µm of the substrate djenkolic acid, in a total volume of 1 ml. NADH oxidation was monitored at 340 nm.
Dr D. Kessler is acknowledged for providing the PR745 strain as well as purified C-DES and expression plasmid pSA16 (Lang and Kessler, 1999), and Prof. N. Olszewski for kindly providing a No-0 sur1 knockout line and SUR1 cDNA.
Relevant Accession numbers: NP_004050, AAB26845, CAB44334, and At2g20610.