The Plant Cell, Vol. 4, 721-733, June 1992 O 1992 American Society of Plant Physiologists

Two Anthranilate Synthase Genes in Arabidopsis: Defense-Related Regulation of the Tryptophan Pathway Krishna K. Niyogi and Gerald R. Finkl Department of Biology and Whitehead lnstitute for Biomedical Research, Massachusetts lnstitute of Technology, Nine Cambridge Center, Cambridge, Massachusetts 02142

Arabidopsis thaliana has two genes, ASA7 and ASA2, encoding the a subunit of anthranilate synthase, the enzyme catalyzing the first reaction i n the tryptophan biosynthetic pathway. As a branchpoint enzyme i n aromatic amino acid biosynthesis, anthranilate synthase has an important regulatory role. The sequences of the plant genes are homologous to their microbial counterparts. Both predicted proteins have putative chloroplast transit peptides at their amino termini and conserved amino acids involved in feedback inhibition by tryptophan. ASA7 and ASA2 cDNAs complement anthranilate synthase a subunit mutations i n the yeast Saccharomyces cerevisiae and i n Escherichia coli, confirming that both genes encode functional anthranilate synthase proteins. The distributions of ASA7 and ASA2 mRNAs i n various parts of Arabidopsis plants are overlapping but nonidentical, and ASA7 mRNA is approximately 10times more abundant in whole plants. Whereas ASA2 i s expressed at a constitutive basal level, ASA7 is induced by wounding and bacterial pathogen infiltration, suggesting a nove1 role for ASA7 i n the production of tryptophan pathway metabolites as part of an Arabidopsis defense response. Regulation of key steps i n aromatic amino acid biosynthesis i n Arabidopsis appears to involve differential expression of duplicated genes.

INTRODUCTION

The enzyme anthranilatesynthase (AS; EC 4.1.3.27) catalyzes the first reaction branching from the aromatic amino acid pathway toward the biosynthesisof tryptophan in plants, fungi, and bacteria. Tryptophan is required primarily for protein synthesis in bacteria and fungi, whereas in plants the tryptophan pathway also provides precursorsfor the synthesis of key “secondary” metabolites such as the major endogenous auxin, indole-3-acetic acid (IAA), and other molecules that may help protect plants against pathogens and herbivores. Because AS is a branchpoint enzyme in aromatic amino acid biosynthesis, regulation of AS is critical for controlling the flux of intermediates in the pathway. In plants, fungi, and bacteria, AS enzyme activity is feedback inhibited by tryptophan. In microorganisms, the genes encoding AS and other tryptophan biosynthetic enzymes are regulated in a manner that reflects the functional importance of the amino acid as the end product of the pathway. For example, in Escherichia coli the synthesis of tryptophan pathway enzymes is regulated by tryptophan via transcription repression and attenuation (for review, see Yanofsky and Crawford, 1987). As a result, the enzymes are synthesized only under conditions of tryptophan starvation. In the yeast Saccharomyces cerevisiae and other fungi,

transcription of the genes encoding tryptophan biosynthetic enzymes is under general amino acid control rather than pathway-specific control (for review, see Hinnebusch, 1988). The genes are expressed ata constitutive basal level, and increased transcription occurs not only in response to tryptophan starvation, but also upon starvation for a number of other amino acids. The regulation of tryptophan biosynthetic genes in plants is undefined. Microbial AS is composed of two nonidentical subunits, an a subunit (also called Component I), which binds the substrate chorismate and carries out its aromatization, and a subunit (also called Component II), which transfers an amino group from the other substrate glutamine (for reviews, see Zalkin, 1980; Hütter et al., 1986; Yanofsky and Crawford, 1987; Crawford, 1989). This glutamine-dependentAS reaction (1) requires both a and p subunits. The a subunit alone can synthesize anthranilate from chorismate using ammonia as the amino donor rather than glutamine, but only if the ammonia is present at a sufficiently high concentration. This reaction is termed the ammonia-dependent AS reaction (2). Both AS reactions require Mg2+ as a cofactor. Chorismate

1

To whom correspondence should be addressed.

+ glutamine -. anthranilate +

Chorismate

+

glutamate NH3 -. anthranilate

+

pyruvate pyruvate

+ (11 (2)

722

The Plant Cell

The a subunit is usually a monofunctional polypeptide,whereas the p subunit is often part of a multifunctional polypeptide in which one domain contains the glutamine amidotransferase activity, and the others catalyze subsequent reactions in the tryptophan biosynthetic pathway. For example, in Neurospora crassa the TRPl gene encodes a single polypeptide containing the AS p subunit, phosphoribosylanthranilateisomerase, and indole-3-glycerolphosphate synthase (Hütter et al., 1986). The a and p subunits associate to form a heteromeric AS enzyme, i.e., an a$;! heterotetramer in E. coli (Ito and Yanofsky, 1969) and N. crassa (Hulett and DeMoss, 1975) and an ap heterodimer in S. cerevisiae (Prantl et al., 1985). Recent studies of AS from Salmonella typhimurium (Caligiuri and Bauerle, 1991) and Brevibacferium lacfofermenfum (Matsui et al., 1987) have pinpointed particular amino acids in the a subunit that are important for feedback inhibition by tryptophan. Microbial AS genes have homology to genes encoding at least two other chorismate-utilizingenzymes, para-aminobenzoicacid (PABA) synthase (Goncharoff and Nichols, 1984; Crawford, 1989) and isochorismate synthase (Ozenberger et al., 1989). Relatively little is known about the biochemistry and regulation of AS from plants (for reviews, see Gilchrist and Kosuge, 1980; Poulsen and Verpoorte, 1991; Singh et al., 1991). Crude fractionation studies have shown that AS is separable from other enzymes in the pathway (Hankins et al., 1976), but the subunit composition of plant AS has not been determined. The role of feedback inhibition of AS enzyme activity by tryptophan (Belser et al., 1971) has been investigated using plant cell cultures resistant to 5-methyltryptophan, a false feedback inhibitor of AS. The mutant cells contain a feedback-resistant AS, and as a result the cultures accumulate tryptophan (Widholm, 1972a. 1972b; Carlson and Widholm, 1978; Ranch et al., 1983) and, in some cases, no longer require exogenous auxin for growth (Widholm, 1977; Sung, 1979). Plant cell cultures fed with anthranilate or indole show apparently unregulated accumulation of tryptophan (Widholm, 1974). Together, these results imply that tryptophan levels in cultured cells are controlled mainly by regulation of AS. However, besides feedback inhibition of AS, control of tryptophan biosynthesis in plants is largely uncharacterized.The synthesis of pathway enzymes,

including AS, is not repressed by tryptophan in plant tissue cultures (Widholm, 197l). Subcellular fractionation experiments (Mousdale and Coggins, 1986) have led to the conclusion that AS, like the rest of the aromatic amino acid pathway (SchulzeSiebert and Schultz, 1989), is sequestered in plastids, although there has been a report of a cytosolic AS isozyme (Brotherton et al., 1986). We are studying AS in Arabidopsis fhaliana to gain further insight into the structure and regulation of the tryptophan biosynthetic pathway in plants. We find that Arabidopsis has two genes encoding the a subunit of AS. The two genes are regulated differently in response to wounding and bacterial pathogen infiltration, providing molecular genetic evidence that the tryptophan branch of the aromatic pathway has a role in plant defense responses.

RESULTS

lsolation of Two Genes Encoding Arabidopsis Anthranilate Synthase A fragment of the yeast TRP2 gene (Zalkin et al., 1984) containing the region most highly conserved between yeast and bacterial AS a subunit genes was used as a DNA hybridization probe to isolate homologous cDNAs from Arabidopsis. A cDNA with 500/0 nucleotide and 39% amino acid identity to the yeast probe was subsequently used to identify two classes of additional Arabidopsis cDNA clones. Restriction mapping and further sequence analysis revealed that one class was identicalto the initial Arabidopsis clone, whereas the other class represented a second gene that also had significant similarity to known AS a subunit genes. Genomic clones corresponding to each cDNA were isolated and sequenced. These two Arabidopsis AS genes, designated ASAl (GenBank accession no. M92353) and ASA2 (GenBank accession no. M92354), share 63% nucleotide identity within proteincoding exons. The structures of the ASAl and ASA2 genes are summarized schematically in Figure 1.

ASAl: XSHd PHd SHd P

Hc

Hc

HdHdP

EXHc

PHc

Hd C

HcHd

ASA2:

,

t o

Ikb

Figure 1. Schematic Structures of ASAl and ASA2 Genes. The boxes represent exons, and dashed lines connect homologous exons between ASAl and ASAP. The 5' to 3' transcriptional orientation of both genes is from left to right. The restriction sites shown are A, Accl; C, Clal; E, EcoRI; Hc, Hincll; Hd, Hindlll; P, Pstl; S, Sacl; and X, Xbal.

Arabidopsis Anthranilate Syntheses

B RNA

RNA

!-*-*

GATC - +

723

two genes cross-hybridized weakly, but there was no evidence of additional hybridizing sequences (data not shown). Restriction fragment length polymorphism (RFLP) linkage analysis (S. Hanley and H. Goodman, personal communication) placed

kb -9.5-

-7.5-4.4H*

«. -2.4-

it

UJ X

-1.4-

X

zi _

o

^

S 8 U UJ

—23130—

—9416 — — 6557 —

-0.24-

—4361 —

ASA1

ASA2

ASA1

ASA2

Figure 2. Analysis of /AS/47 and ASA2 mRNAs. (A) RNA gel blot analysis. Ten micrograms of Arabidopsis total RNA and 1 ng of poly(A)+ RNA were fractionated by electrophoresis on a 1% agarose gel containing formaldehyde (Ausubel et al., 1989), transferred to nitrocellulose, and hybridized with a 32P-labeled 2-kb Xhol fragment of pKN41 (ASA1) and a 1.8-kb BamHI fragment of pKN108A (ASA2). The sizes of the RNA markers (Bethesda Research Laboratories) are shown in kilobases. (B) Primer extension analysis. For ASA1, an antisense oligonucleotide complementary to nucleotides 2391 to 2418 in the ASA1 sequence (GenBank accession no. M92353) was hybridized to 25 ng of Arabidopsis total RNA at 55°C, and cDNA was synthesized using reverse transcriptase (Ausubel et al., 1989). Products were analyzed by denaturing polyacrylamide gel electrophoresis. The ASA1 sequencing ladder was generated using the same primer and pKN211A template DNA. For ASA2, an antisense oligonucleotide complementary to nucleotides 2846 to 2873 in the ASA2 sequence (GenBank accession no. M92354) was hybridized to 50 ng of Arabidopsis total RNA at 55°C. The ASA2 sequencing ladder was generated using the same primer and pKN143C template DNA.

— 564—

ASA1

ASA2

B 2 4553 - - 0.0 Cp-2

24.2

er

39.2

ASA2 Both genes are transcribed, resulting in mRNAs approximately 2200 nucleotides in length, as shown by RNA gel blot analysis in Figure 2A. The 5' ends of the mRNAs encoded by the two genes were determined by primer extension (Figure 2B) and confirmed by S1 nuclease protection (data not shown). For ASA1 mRNA, there are two major 5' ends, resulting in 5' untranslated leaders of 90 and 93 nucleotides. The mRNA corresponding to ASA2 begins at a single site 41 nucleotides upstream of the first AUG codon. As determined by cDNA sequencing, the ASA1 and ASA2 genes have 3' transcribed but untranslated regions of 156 nucleotides and 162 nucleotides, respectively. As is the case for many plant genes (Joshi, 1987), no sequences were found in these 3' regions that exactly match the AAUAAA consensus sequence for animal poly(A) addition signals. Arabidopsis ASA1 and ASA2 are duplicated, but unlinked, nuclear genes. The cloned genes hybridized to different sets of fragments on a genomic DNA gel blot, as shown in Figure 3A. Under conditions of reduced hybridization stringency, the

TSB1

125.3

AB5-13

160.6

-

47.2

21502

49.9

cer-8

71.9

Figure 3. Analysis of ASA1 and ASA2 Genomic Loci. (A) Genomic DNA gel blot analysis. Two micrograms of Arabidopsis genomic DNA was digested with the indicated restriction enzymes, fractionated by electrophoresis on a 0.8% agarose gel, transferred to a Zeta-Probe membrane (Bio-Rad), and hybridized with a ^P-labeled 1-kb EcoRI fragment of pKNSA (ASA1) and a 1.2-kb EcoRI fragment of pKN1A (ASA2). Sizes of DNA markers (Hindlll-digested X DNA) are shown in base pairs. (B) Schematic representation of the locations of ASA1 and ASA2 on Arabidopsis chromosomes 2 and 5. Selected visible and RFLP genetic markers are shown to the left of each vertical line, and map distances in centimorgans (S. Hanley and H. Goodman, personal communication) are to the right.

724

The Plant Cell

Plurality ASA1 ASA2 TRP2 (S.C.) TrpE (E.c. ) TrpE ( B . s . ) PabB (E.c )

.

......... I . ........ I ......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .I . 10

20

30

50

40

60

70

80

90

100

170

180

190

200

*** Plural ity ASAl ASA2 TRP2(S.c.) TrpE (E.c. ) TrpE ( B . s . ) PabB (E.c.)

......... I .........l.........I.........I.........I.........I.........l............l.........l......... 110

Plurality ASAl ASA2 TRP2(S.c.) TrpE (E.c. ) TrpE(E3.s.) PabB (E.c.)

120

140

160

150

I

......... I ......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210

Plurality ASAl ASA2 TRP2 ( S . c. ) TrpE (E.c. ) TrpE (E3.s.) PabB (E.c.)

130

220

230

240

250

.........I . . ....... l ......... I..... .... I.........I. 310

320 tt

330

340

2 60

270

280

2 90

I

300

........ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .I 350

**

360

370

3 90

380

400

t

t

Plurality ASAl

ASA2 TRPZ(5.c.) TrpE (E.c.) TrpE(B.s.) PabB (E.c. )

......... I ......... I . . . . . . . . . I . . 410

420

.......I.... ..... I..... .... I ......... I . ........ (.........I......... 430

440

450

t

Plurality

4 60

470

480

t

I

4 90

500

t

...GSV.V..I ... I..FSHVMH..S.V.G.L...L...L)ALRA..P.GT.SGAPKVW\MELI.ELE..RRG.\l.G..G..SF.G.-MD..I..RT...--

ASAl

ASA2 TRPZ(S.c.) TrpE (E.c. ) TrpE ( B . s . ) PabB (E.c.)

......... I ......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 510

520

..

530

t

-_-----_----____ Plurality G.A..QAGAGIVRDS.P ASA1 ASA2 TRP2 ( S .c. ) TrpE(E.c.1 TrpE (F3.s.) PabB (E.c ) I . . . . . . . . . . . . . . .

.

......... ......... 610

620

630

540

550

. .- . . .

-%

--.

570

580

5 90

I

600

t

..E ..E..NKA......I..A.......

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I. . . 640

650

Figure 4. Alignment of Amino Acid Sequences of AS a Subunit Proteins.

, ..E... -

5 60

660

670

680

6 90

700

Arabidopsis Anthranilate Synthases

ASAl and ASA2 on different chromosomes on an Arabidopsis RFLP genetic map (Nam et al., 1989). The ASAl gene is located on chromosome 5 at 14.8 centimorgans, whereasASA2 is located on chromosome 2 at 47.2 centimorgans (Figure 38). Sequence analysis of 5908 bp of ASAl and 6661 bp of ASA2 genomic DNA did not reveal any sequences related to microbial AS p subunit genes or any other tryptophan biosyntheticgenes. Based on available mapping data, there does not appear to be significant clustering of the cloned Arabidopsis tryptophan biosynthetic genes in any particular region of the genome (Figure 38; Last et al., 1991).

Comparison of the ASA1 and ASA2 Deduced Proteins The amino acid sequences of deduced proteins encoded by

ASAl and ASA2 are shown in Figure 4. For each gene, protein translation was assumed to begin at the first AUG codon downstream of the transcription start site(s). The ASAl gene is capable of encoding a protein of 595 amino acids with calculated molecular mass of 66212 D; ASA2 can encode a 69711-D protein composed of 621 amino acids. The ASAl and ASA2 predicted amino acid sequences were aligned with the AS a subunits from yeast, E. coli, and Bacillus subtilis, along with the PABA synthase a subunit from E. coli (Figure 4). As seen in Table 1, ASAl and ASA2 are 67% identical to each other and 30 to 36% identical to microbial AS a subunit sequences. The ASAl and ASA2 predicted proteins contain conserved amino acids involved in feedback inhibition of AS by tryptophan (Figure 4). 60th ASA1 and ASA2 proteins are predicted to have aminoterminal extensions not present in the corresponding microbial proteins (Figure 4). The amino acid compositions and secondary structures of these extensions are characteristic of plant chloroplast transit peptides (von Heijne et al., 1989). The aminoterminal amino acid sequences of the mature, processed Arabidopsis ASA1 and ASA2 proteins are not known, but the sequences IKCV in ASAl and IKCS in ASA2 are near matches to a proposed cleavage site consensus [(V/I)X(A/C).1A] (Gavel and von Heijne, 1990). The putative chloroplast transit peptides are the most dissimilar regions of the aligned proteins, as evidenced by the fact that there is only 20% amino acid

725

identity between the peptides encoded by the first exons of

ASAl and ASA2. There is no apparent homology between the ASA1 and ASA2 transit peptides and the transit peptides of other Arabidopsis aromatic amino acid biosynthetic enzymes (Klee et al., 1987; Berlyn et al., 1989; Keith et al., 1991; Last et al., 1991).

ASA7 and ASA2 Encode Functional AS a Subunits Arabidopsis ASAl and ASA2 complemented AS mutations in yeast and E. coli, verifying the assertion, based on sequence analysis, that ASAl and ASA2 are AS a subunit genes. To test complementation in yeast, we made a deletionlinsertion mutation of the TRP2 gene encoding the AS a subunit in yeast. A strain (KNY1) containing this mutation (frp2A9O::HlS3)requires tryptophan supplementation of minimal medium for growth. As shown in Figure 5, ASAl and ASA2 cDNAs under the control of the yeast GALl promoter complemented the auxotrophic phenotype of the frp2 mutation, confirming that both ASAl and ASA2 encode functional anthranilatesynthases. Consistent with the control of expression by the inducible GALl promoter, the KNYl strain containing the ASAl cDNA (pKN41) or the ASA2 cDNA (pKNlO9A-1) grew on galactose medium

Table 1. Pairwise Amino Acid ldentity between Plant AS a Subunit Sequences and Microbial AS a and PABA Synthase a

Subunit Sequences

ASAl ASA2 TRP2a

ASA2

TRP2a

TrpEb

TrpEC

PabBb

(04

(O4

(04

( 0 4

(O4

67

33

30 31 31

36 35 35 31

32

33

TrpEb TrpEC

30 27 26 29

Percent identity was calculated for the sequences as aligned in Figure 4. a

Derived from S. cerevisiae. Derived from E. coli. Derived from 6. subfilis.

Figure 4. (continued).

The predicted amino acid sequences of ASAl and ASA2 proteins were fit to the alignment of bacterial and yeast AS a subunit sequences (Crawford, 1989), with minor modifications. The sequences shown are Arabidopsis ASAl and ASA2, S. cerevisiae TRP2 [TRPP S.C.)](Zalkin et al., 1984), E. coli TrpE [TrpE (E.c.)] (Nichols et ai., 1981), 6. subtilis TrpE [TrpE (B.s.)](Band et al., 1984), and E. coli PabB [PabB (E.c.)] (Goncharoff and Nichols, 1984), the a subunit of PABA synthase. The TRP2 sequence was modified by the introduction of frameshifts at amino acid positions 622 and 638, as described by Crawford (1989). The plurality shows amino acids that are conserved in two-thirds of the aligned sequences at each position. Hyphens indicate introduced gaps. Amino acid positions showing identity to the ASA1 sequence are shaded. Conserved amino acids that are affected in inactivating missense mutants of S. typhimurium WpE (Bauerle et al., 1987) are designated with daggers. Five of six of these amino acids are invariant in the AS a subunit sequences shown. Amino acid positions affected in strongly feedback-resistant mutants of S.typhimurium trp€ (Caligiuri and Bauerle, 1991) are noted with asterisks. Five of nine of these amino acids are identically conserved in ASAl and ASA2, and two of nine are replaced by similar amino acids.

726

The Plant Cell

Regulation of ASA1 and ASA2 mRNA Levels

ra

! 1

a s I K l

undiluted

glc - ura

glc - ura - trp gal - ura - trp

Figure 5. Complementation of a Yeast trp2 Deletion Mutation by Arabidopsis ASA1 and ASA2. The yeast strain KNY1 (MATa trp2A90::HIS3 ura3-52 his3A200 leu2A2 Gal+) was transformed with the plasmids pKN10 (TRP2), pSE936* (vector), pKN41 (ASA1), pKN109C (antisense ASA2), and pKN109A-1 (ASA2), and Ura+ transformants were grown to saturation at 30°C in liquid SC medium containing 2% glucose lacking uracil. Serial dilutions were made with water, and 2 nL of each were spotted on the indicated agar plates: SC medium with 2% glucose lacking uracil (glcura), SC medium with 2% glucose lacking uracil and tryptophan (glcura-trp), and SC medium with 2% galactose lacking uracil and tryptophan (gal-ura-trp). Growth after 4 days at 30°C is shown.

lacking tryptophan but not on glucose medium without tryptophan. The URA3 ARS1 CEN4 vector alone (pSE936*) or antisense ASA2 in the URA3, 2-um vector (pKN109C) did not allow growth of KNY1 on either galactose or glucose medium lacking tryptophan. As a positive control, the yeast TRP2 gene under the control of its own promoter (pKN10) complemented the trp2 mutation in KNY1 for growth on both glucose and galactose medium without tryptophan. We also tested the ability of ASA1 cDNAs to complement trpE mutations in £ coli. As shown in Figure 6, expression of an ASA1 cDNA allowed growth of E. coli strains containing either a deletion mutation in trpE (trpAES) or a nonsense mutation in trpE (trpE5972) on M9 glycerol minimal medium lacking tryptophan. The vector alone was unable to complement either trpAES or trpE5972. Complementation of trpAES by ASA1 (pKN37) was also observed on modified M9 glycerol minimal medium without tryptophan containing only 1 mM NH4CI (data not shown), conditions under which E. coli requires glutamine-dependent AS activity (1) in order to grow (Paluh et al., 1985). Expression of ASA1 in an £ coli strain lacking both endogenous AS subunits (trpAED27; Jackson and Yanofsky, 1974) allowed growth on M9 medium lacking tryptophan (data not shown), conditions under which the ammoniadependent AS activity (2) provided by the AS a subunit is sufficient. However, this strain did not grow on modified M9 (1 mM NH4CI) lacking tryptophan (data not shown), suggesting that ASA1 encodes an AS a subunit that can interact with the E. coli AS p subunil in vivo and that such interaction is necessary for catalyzing the glutamine-dependent AS reaction (1).

We observed a quantitative difference in the steady state levels of ASA1 and ASA2 mRNAs in Arabidopsis plants (Figure 2A). The ASA1 mRNA was the major AS a subunit transcript, present at a level about 10-fold greater than that of ASA2, as determined by RNA gel blot analysis using probes of approximately equal specific activity. RNA gel blot analysis of RNAs isolated from different parts of Arabidopsis plants revealed a qualitative difference in the patterns of ASA1 and ASA2 expression, as shown in Figure 7A. Roots and rosette leaves of 5-week-old plants contained approximately equal amounts of ASA1 mRNA. The level of ASA1 mRNA was lower in stem tissue and undetectable in flowers, developing (green) siliques, and mature dry seeds. In contrast, ASA2 mRNA was present at low, but detectable, levels in roots, leaves, and developing siliques and at slightly lower levels in stem tissue. There was no detectable ASA2 mRNA in flowers or dry seeds. The absence of detectable ASA1 orASA2 mRNA in certain tissues was not due to any general problem with the RNA from those tissues, because the mRNA corresponding to the ALS gene encoding acetolactate synthase (acetohydroxyacid synthase) (Mazur et al., 1987), the first enzyme of the isoleucine-valine biosynthetic pathway, was detected in all Arabidopsis tissues examined (Figure 7A). Addition of tryptophan or auxin did not affect the expression of ASA1 or ASA2 RNAs in wild-type plants. Arabidopsis plants grown on synthetic medium without tryptophan had steady state levels of ASA1 and ASA2 mRNAs equivalent to plants grown with 50 uM tryptophan (data not shown). Because

trpAES

trpE5972

vector

ASA1

+trp Figure 6. Complementation of E. coli trpE Mutations by Arabidopsis ASA1.

The E. coli strains trpAES tnaA2 and trpE5972 were transformed with pSE936* (vector) or pKN37 (ASAT), and ampicillin-resistant transformants were grown to saturation at 37°C in liquid Luria-Bertani medium containing 100 ng/mL ampicillm and then streaked onto agar plates containing M9 minimal medium plus 0.2% glycerol, 100 ng/mL ampicillin, 200 uM L-tryptophan (+trp) or M9 minimal medium plus 0.2% glycerol, 100 |ig/mL ampicillin, 100 uM isopropylthiogalactoside (-trp). Growth after 4 days at 37°C is shown.

Arabidopsis Anthranilate Synthases

B

hours after Infection

hours after wounding « £

r

«

8

0

727

1

3

6

controlI065(avlrulent)

9 121518

4326 (virulent)

0 1.53.5 612.524 50 0 1.53.5 612.524 50 0 1.53.5 612.524 50

ASA1

•*• -99•*»*?* f» ASA1

ASA2

ASA2

ALS

ALS

•«•-.«•-* m-1f.-m If ^ » w ^ 1F» » VW ASA1

ASA2

*' "» »*

•-*»«•• ALS

•«*• **> «•• •«• — -i"

Figure 7. RNA Gel Blot Analysis of ASA1 and AS/42 Expression. Total RNA was fractionated by electrophoresis on 1% agarose gels containing formaldehyde (Ausubel et al., 1989), transferred to nitrocellulose, and hybridized with 32P-labeled probes. (A) Tissue specificity. Five micrograms of total RNA per lane from rosette leaves (leaf), roots (root), stems of flowering stalks (stem), immature and fully opened flowers (flower), developing green seed pods (silique), and mature dry seeds (seed) of 5-week-old Arabidopsis pgm plants were hybridized with a 32P-labeled 1-kb EcoRI fragment of pKNSA (ASA1), a 1.2-kb EcoRI fragment of pKN1A (ASA2), and a 3.3-kb Ncol-Xbal fragment of pGH1 (ALS) (kindly provided by G. Haughn, University of Saskatchewan). Equal loading of RNA in each gel lane was confirmed by ethidium bromide staining. (B) Wounding. Ten micrograms of total RNA per lane from wounded Arabidopsis leaves were hybridized with a 32P-labeled 2-kb Xhol fragment of pKN41 (ASA1), a 1.8-kb BamHI fragment of pKN108A (ASA2), a 3.3-kb Ncol-Xbal fragment of pGH1 (ALS), and a 1-kb Kpnl-BamHI fragment of pABT4 (ABT4) (Marks et al., 1987). (C) Pathogen infiltration. Ten micrograms of total RNA per lane from leaves of 3-week-old Arabidopsis plants inoculated as described previously (Dong et al., 1991) with 10 mM MgSO4 (control), P. s. pv tomato MM1065 (1065 [avirulent]), and P. s. pv maculicola ES4326 (4326 [virulent]) were hybridized with the same probes as given in (B). The liter of the inoculated bacterial strains was 1 x 105 colony-forming units per milliliter. Bacterial growth was followed as described by Dong et al. (1991). Disease symptoms caused by ES4326 were readily apparent 48 hr after inoculation.

the tryptophan pathway provides the precursors for IAA biosynthesis in plants, we also examined the effect of exogenous IAA application on the steady state levels of ASA1 and ASA2 mRNAs. Spraying 3-week-old wild-type Arabidopsis plants with 100 uM IAA did not discernibly affect ASA1 and ASA2 mRNAs relative to water-treated control plants (data not shown), whereas levels of mRNAs corresponding to known auxininduced genes were increased (J. Normanly and G. R. Fink, unpublished results). Expression of ASA1 mRNA, but not ASA2 mRNA, was strongly induced by wounding of Arabidopsis leaf tissue. As shown in Figure 7B, the steady state level of ASA1 mRNA increased dramatically 3 hr after wounding and then increased again after 12 hr, reaching a maximum at 18 hr. In striking contrast to ASA1, the level of ASA2 mRNA declined gradually after wounding of Arabidopsis leaves. A similar decrease was observed for ALS and (5-tubulin (ABT4) mRNAs (Figure 7B). Infiltration of Arabidopsis leaves with virulent and avirulent strains of Pseudomonas syringae (Dong et al., 1991) specifically induced ASA1 mRNA. The avirulent strain P. s. pv tomato MM1065 causes a resistance response when inoculated into Arabidopsis ecotype Columbia, whereas infiltration of the virulent strain P. s. pv maculicola ES1065 results in disease (Dong

etal., 1991). Figure 7C shows that the steady state level of ASA1 mRNA increased transiently 6 hr after inoculation with the avirulent pathogen R s. pv tomato MM1065. Infiltration of the virulent strain P. s. pv maculicola ES4326 resulted in a strong induction of ASA1 mRNA, beginning at 12.5 hr and persisting until the end of the time course. Mock inoculation of Arabidopsis leaves caused a slight, transient increase in the level of ASA1 mRNA at 1.5 hr that may be due to wounding of the plants during the inoculation procedure; similar increases were observed 1.5 hr after bacterial inoculations (Figure 7C). The levels of ASA2 and ALS mRNAs were relatively unaffected by bacterial infiltration, whereas (3-tubulin (ABT4) mRNA increased slightly 12 to 48 hr after infiltration of ES4326 (Figure 7C).

DISCUSSION Arabidopsis AS Resembles Microbial AS We have isolated and characterized two Arabidopsis genes, ASA1 and ASA2, encoding the a subunit of AS. The initial clone was obtained using part of the yeast TRP2 gene (Zalkin et al.,

728

The Plant Cell

1984) as a heterologous probe. Two lines of evidence have led to the conclusion that ASAl and ASA2 encode Arabidopsis AS. First, the sequences of both genes are homologous to AS a subunit genes from bacteria and yeast (Figure 4 and Table 1). Second, ASA7 and ASA2 cDNAs complemented mutations in yeast TRP2 and E. coli frpE (Figures 5 and 6). The structure of plant AS appears to resemble that of microbial AS. With the exception of the putative chloroplast transit peptides at their amino termini, the Arabidopsis ASAl and ASA2 proteins align well with AS a subunit proteins from yeast and bacteria (Figure 4). We found no evidence for fused a and p subunits as in Rhizobium meliloti (Bae et al., 1989) and the photosynthetic eukaryote Euglena gracilis (Hankins and Mills, 1976). When expressed in E. coli, the Arabidopsis a subunit encoded by ASA7 was apparently able to interact with the E. coli p subunit (Figure 6 and data not shown). Although there is no information about the subunit structure of plant AS, our data suggest that in Arabidopsis the ASA1 and ASA2 proteins interact with an AS p subunit, presumably encoded by a distinct gene(s).

Two Genes, Two Pathways The presence of two Arabidopsis genes encoding the AS a subunit is consistent with the suggestion that there may be two aromatic amino acid pathways in plants (Jensen, 1986; Keith et al., 1991; Last et al., 1991). Duplication of genes has also been observed for three other Arabidopsis aromatic amino acid biosynthetic enzymes: 3-deoxy-o-arabino-heptulosonate bphosphate (DAHP) synthase (Keith et al., 1991), 5enolpyruvylshikimate-3phosphate(EPSP) synthase (Klee et al., 1987), and tryptophan synthase p subunit (Berlyn et al., 1989; Last et al., 1991). The available evidence suggests that duplication of genes encoding certain enzymes of this pathway may be a general feature of plants (Last et al., 1991). Two aromatic amino acid biosynthetic pathways could be separated within the plant cell due to different compartmentation (Jensen, 1986). Cytosolic and plastidic isozymes of AS have been described in 5-methyltryptophan-resistant tobacco cell cultures (Brothertonet al., 1986). However, both Arabidopsis AS a subunit genes encode proteins with putative amino-terminal chloroplast transit peptides, suggesting that the products of both genes may reside within plastids. Like ASA7 and ASA2, the gene pairs encoding tryptophan synthase p (Berlyn et al., 1989; Last et al., 1991) and DAHP synthase (Keith et al., 1991) and the one published EPSP synthase gene (Klee et al., 1987) encode putative chloroplast transit peptides at their amino termini. Although we have not proven that ASAl and ASA2 proteins are plastid-localized,preliminary subcellular fractionation experiments indicated that there is AS activity in Arabidopsis chloroplasts, whereas none was detectable in a cytosolic fraction (K. K. Niyogi and G. R. Fink, unpublished results). It is possible that ASAl and ASA2 are sequestered in different compartments within the plastid. If our results can be extrapolated to tobacco, the cytosolic AS isozyme, which

was detectable only in tissue culture cells and not in regenerated plants (Brotherton et al., 1986), might be due to aberrant processing or localization of a plastidic AS protein in cultured cells (Singh et al., 1991). Of course, there may be additional, distantly related AS genes in Arabidopsis that encode cytosolic isozymes. 60th aromatic amino acid pathways could be present in plastids, but the two pathways might be expressed in distinct cell types. Our results showing expression of ASA7 and ASA2 in different parts of Arabidopsis plants revealed subtle qualitative differences in ASAl and ASA2 tissue specificity (Figure 7A). Transgenic plants containing fusions of ASA7 and ASA2 promoters to the reporter gene 0-glucuronidase (Jefferson et al., 1987) should be useful for determining precisely in which cell types the two genes are expressed.

Two Genes, Two Functions Expression of the ASAl gene is regulated by increased demand for tryptophan pathway metabolites, whereas theASA2 gene appears to be constitutivelyexpressed. Addition of tryptophan or auxin to Arabidopsis plants had no effect on ASA7 orASA2 mRNA levels (data not shown), suggestingthat repression below the basal levels of expression does not occur in response to exogenous primary and secondary products of the pathway. The dramatic increases in ASAl steady state mRNA levels following wounding (Figure 76) and bacterial pathogen infiltration (Figure 7C) suggest a role for ASAl in production of secondary metabolites as part of an Arabidopsis defense response. Because ALS mRNA was unaffected by these treatments, the induction of ASA7 is probably not due to a general stress-relatedrequirement for amino acids for protein synthesis. Tryptophan pathway metabolites that may be involved in plant defense responses include indole-3-methylglucosinolates and indole phytoalexins. lndole glucosinolates, which are found in Arabidopsis (Hogge et al., 1988) and which have a number of toxic hydrolysis products (for review, see Fenwick et al., 1983), accumulate following insect infestation and wounding of Brassica species (Koritsas et al., 1989). Production of indole phytoalexins by cruciferous plants has been observed following elicitation with bacterial and funga1 pathogens and UV light (Takasugi et al., 1986; Devys et al., 1988; Browne et al., 1991). In Arabidopsis, the phytoalexin produced in response to ? I s. pv syringae infection is 3-thiazol-2'-yl-indole(Tsuji et al., 1992). lnductionof ASA1 mRNA may therefore be necessary for increased synthesis of indole phytoalexins. The second AS gene, ASA2, behaves constitutively under conditions that induce ASA1. The different regulation of ASAl and ASA2 may reflect different functions. Duplication of AS genes to perform distinct functions has a precedent in the bacterium I? aeruginosa, which has one AS gene for tryptophan biosynthesis and another involved in pyocyanin pigment synthesis (Essar et al., 1990).

Arabidopsis Anthranilate Synthases

In Arabidopsis, the constitutiveASA2 gene may function primarily to maintain a basal leve1 of tryptophan pathway metabolites. The regulated ASA7 gene may also have a role in basal biosynthetic activity, so the two genes may have some functional redundancy, but the induction of ASAl in situations requiring increased synthesis of secondary products derived from the tryptophan pathway suggests that ASA1 has additional roles. Arabidopsis mutants affected in AS activity should be useful for testing this hypothesis. The different regulation of ASAl and ASA2 parallels the situation for the duplicated genes encoding Arabidopsis DAHP synthase (Keith et al., 1991), the first enzyme in the aromatic amino acid pathway. The DAHP synthase gene DHSl, like ASAl, respondsto wounding and pathogen infection, whereas the second DAHP synthase gene DHS2, like ASAP, is unaffected. These results suggest that for two canonical control points in the aromatic amino acid pathway, AS and DAHP synthase, regulation in plants involves differential expression of duplicated genes.

METHODS

Plant Material Arabidopsis thaliana plants derived from the Columbia ecotype were grown in soil under continuous illumination at 22OC as described by Last and Fink (1988). Total DNA from rosette leaves of 3-week-oldwildtype plantsand RNAfrom 3-week-oldwhole pgm (Caspar et al., 1985) plants were isolated as described by Ausubel et al. (1989). Total RNA was isolatedfrom small amounts of plant materialbya miniprep method (Nagy et al., 1988). Wounding of plants was performed by cutting rosette leaves of 3-week-oldpgm plants into a damp glass beaker that was then covered and incubated in the dark at room temperature. At various time points, 2 to 3 g of tissue was frozen in liquid nitrogen and stored at -7OOC. lnoculation of 3-week-old wild-type Arabidopsis plants with Pseudomonas syringae pv tomato MM1065 and P s.pv maculicola ES4326 was performed as described by Dong et al. (1991).

DNA and RNA Methods Standard techniques of DNA analysis and cloning were performed as described by Ausubel et al. (1989). DNA fragments were purified from agarose gels using GeneClean(Bio 101, La Jolla, CA). DNA hybridization probes were labeled using a-32P-dATP(Amersham) and random hexamer primers (Prime Time, lnternationalBiotechnologies, New Haven, CT). For DNA gel blot analysis of genomic DNA, total DNA was digestedwith restrictionenzymes(New England Biolabs, Beverly, MA), separated on 0.8% agarose gels, and transferred to Zeta-Probe (Bio-Rad) in 0.4 M NaOH. Genomic DNA blots were hybridized in 1.5 x SSPE (1 x SSPE is 0.15 M NaCI, 10 mM sodium phosphate, 1 mM EWA, pH 7.4), 05% Blotlo, 1%SDS, O5 mg/mL heat-denatured sheared herring sperm DNA, and 10% dextran sulfate at 65OC overnight and washed in 0.1 x SSC (1 x SSC is 0.15 M NaCI, 15 mM sodium citrate), 0.1% SDS at 5OOC.

729

RNA gel blot analysiswas done as describedby Ausubel et al. (1989). except that 0.1 mg/mL heat-denatured sheared herring sperm DNA and 5% dextran sulfate were used in the hybridizationsolution. Nitrocellulose filters were washed in 0.1 x SSC, 0.1% SDS at 65OC. An RNA ladder (Bethesda Research Laboratories) was used to estimate sizes on RNA gels. Densitometric scanning of RNA blot autoradiograms was done using a computing densitometer (Molecular Dynamics, Sunnyvale, CA). Primer extensionanalysisof RNA was done as described by Ausubel et al. (1989)with minor modifications. Yeast tRNA (10 pg) was included in the hybridization. Moloney murine leukemiavirus (M-MLV) reverse transcriptase(400 units; Bethesda Research Laboratories) was used insteadof avian myeloblastosisvirus reversetranscriptase,and alkaline hydrolysis was substituted for RNase A treatment. Screening of cDNA and Genomic Libraries The Saccharomyces cerevisiae TRP2 gene (Zalkin et al., 1984) and the Escherichiacoli trpE gene (Nichols et al., 1981) were compared to identify conservedamino acid sequences, and a 91Sbp EcoRV fragment (nucleotides 1271 to 2189) of the yeast TRP2 gene in pME514 (Braus et al., 1985) was chosen as a DNA hybridization probe. An Arabidopsis cDNA library in IgtlO (Clontech, Palo Alto, CA) was grown in E. coli host strain C600, and 50,000 plaqueswere transferredto Biotrans nylon filters (ICN Biomedicals, Irvine, CA) in duplicate and hybridized to the 32P-labeledprobe in 25% formamide, 5 x SSPE, 5 x Denhardt's solution (1 x Denhardfssolution is 0.02% Ficoll,0.02% polyvinylpyrrolidone, 0.02% BSA), 0.1 mg/mL heat-denaturedsheared herringsperm DNA, 5% dextran sulfate at 42OC overnight. Filterswere washed four times for 15 min each in 25% formamide, 5 x SSC, 0.1% SDS at 42OC and then exposedto x-rayfilm (Kodak)with an intensifying screenat 70% for 7 days. Phageclones yieldingduplicated positive hybridization signals were purified by two more rounds of plaque hybridization. The 1.2-kb cDNA insert of I K N l was subcloned in both orientations into the EcoRl site of pUC118 (Vieira and Messing, 1987) to create pKNlA and pKNlC. The EcoRl fragment containing the 1.2-kb cDNA and pKNlA was then used as a hybridization probe to isolate additional cDNAs from a randomhexamer-primedArabidopsis cDNA library in IgtlO (Learned and Fink, 1989). Hybridizationwas done in 5 x SSPE, 5 x Denhardt's solution, 0.2% SDS, 0.1 mg/mL heat-denaturedsheared herringsperm DNA, 5% dextran sulfate at 65OC overnight, and the most stringent wash was in 0.25 x SSC, 0.1% SDS at room temperature. The 1.0-kb cDNA insert from purified phage IKN8 was subcloned in both orientations into the EcoRl site of pUC118, yielding pKN8A and pKN8C. The cDNAs from pKN8A and pKNlA were then used to isolate genomic clones corresponding toASA7 and ASA2, respectively, from an Arabidopsis genomic library in IEMBL3 (Clontech). The ASA7 genomic clone IKN21was subcloned as follows: a 3.2-kb Hindlll partia1digest fragment was subcloned in both orientations into the Hindlll site of pUC119 to make pKN212A and pKN212C; an overlapping 4.2kb EcoRl fragment was subcloned into EcoRlcut pUC118 to make pKN211A; a 3.2-kb EcoRl-Xbalfragmentthat is a subset of the pKN211A insertwas ligated in the opposite orientation into pUC118 cut with EcoRl and Xbal to make pKN214. Two adjacent Hindlll fragments from ASA2 genomic clone IKN12 were each ligated into the Hindlll site of pUCll9 in both orientations; pKN140A and pKN140C contain a 3.1-kb fragment, and pKN143A and pKN143C contain a 3.6-kb fragment. Additional ASA7 and ASA2 cDNAs were isolatedfrom an Arabidopsis cDNA library in IYES-R (Elledge et al., 1991). A 533-bp EcoRl fragment of pKN212Awasusedto isolate kKN37and XKN41, from which

730

The Plant Cell

pKN37 and pKN41were looped out in vivo by site-specific recombination using JM107lhKC as host (Elledge et al., 1991). A 612-bp Sacl fragment of pKN143A was used to isolate IKN34 and IKN35 and thereby pKN34and pKN35.The cDNA in pKN35 lacked 198 bp at the 5'end and 438 bp at the 3'end of the coding region, so a nearly fulllengthASA2 cDNA was constructed as follows. The Xhol fragmentcontaining the cDNA from pKN35was ligated into the Sal1site of pUC119 to make pKN104, and then the 844-bp Bglll-Pstl fragment of pKNlC was ligated into Bglll- and Pstlcut pKN104 to generate pKNlO8A.

DNA Sequencing

DNA sequence was determined using the dideoxy chain termination method (Sanger et al., 1977)with modified T7 DNA polymerase(Tabor and Richardson, 1987) and either single-stranded templates generated using M13KW (Vieira and Messing, 1987) or alkaline-denatured double-stranded templates (Chen and Seeburg, 1985). Overlapping unidirectional deletion series were made using exonuclease 111 (Henikoff, 1984) with exonuclease VI1 instead of S1 nuclease. Deletion series of fragments subcloned in both orientationswere made so that both strands were completely sequenced. Sequencing reactions (Sequenase, United States Biochemical) were analyzed on denaturing gradient polyacrylamide gels (Biggin et al., 1983). The overlapping deletions in combination with synthetic oligonucleotideswere used to sequence the entire genomic or cDNA clones in pKN212A, pKN212C, pKN211A,pKN214, pKNBA, pKNBC, pKN140A, pKN140C, pKN143A, pKN143C,pKNlA, and pKN1C. Partia1sequences were obtained for the inserts in pKN34, pKN35, pKN36, pKN37, and pKN41 to determine the 5' ends of these cDNAs and to confirm the positions of introns in the genomic clones. The 5'ends of the cDNAs in pKN37 and pKN41 are nucleotides2508 and 2386, respectively, in the ASA7 sequence (GenBankaccession no. M92353). The 5'end of the pKN35 cDNA is nucleotide3023 in theASA2 sequence (GenBank accession no. M92354). In addition, the cDNA in pKN35 has a single silent T- to -C transition mutation (nucleotide3753 in theASA2sequence) relative to the sequence of pKN34 and of the genomic clone, and the cDNA in pKN34 has an unspliced fourth intron. Sequence analysis was done using UWGCG programs (Devereux et al., 1984). Data base searcheswere performedusing FastA (Pearson and Lipman, 1988). Protein secondary structure predictionswere done using PREDICT89 (Finer-Moore and Stroud, 1984).

Expression ln Yeast and E. coli

Yeast media were prepared as described by Sherman et al. (1986). The yeaststrainKNYl (MATa ttp2A90::HlS3ura3-52his3A2OOleu2A7 Gal+) was constructed as follows. The ends of the 1.77-kb BamHl fragment of pJH-H1(kindly provided by J. Hill, Carnegie Mellon University, Pittsburgh, PA) wntaining the HlS3gene were made blunt using Klenow fragment of DNA polymerase I,and the resultingfragment was ligated into Stul- and Hpal-cutpME514to create pKNtrpPAB, a constructcontaining a deletionof nucleotides 1106to 1794 in the TRP2 coding region (Zalkinet al., 1984)as well as an insertion of the HlS3 gene. The plasmid pKNtrp2ABwas digestedwith EwRl and Sal1and usedto transform yeast strain L4242 (MATa ura3-52 his3A200leu2A 7 Gal+) (6. Ruskin and G. R. Fink, unpublished results) by the lithium acetate method (Ausubel et al., 1989). His+transformants were selected on synthetic complete (SC) medium lacking histidine, and His+Trp- clones were identifiedby replica plating. Genomic DNA gel blot analysisconfirmed

the replacement of the wild-type TRP2 gene by the deletionlinsertion mutation trp2A 90::HIS3. The 4.2-kb BamHI-Sal1fragment of pME514 was subcloned into BamHI- and Sall-cut YEp24 (Botstein et al., 1979) to create pKN10, which contains the wild-type yeast TRP2 gene under the control of its own promoter in a high-copy 2-pm, URA3 vector. The plasmid pSE936*, the URA3 ARS7 CEN4 empty vector counterpart of pKN37 and pKN41, was generated by site-specificrecombination betweenlox sites on IYES-R in host JMlWlIKC (Elledge et al., 1991); pSE936' differs from pSE936 (Elledge et al., 1991) because it lacks the Notl site and has only a single lox site. The BamHl fragment containing the ASA2 cDNA from pKNlO8A was subcloned in both orientations into the yeast expression vector pDAD2 (URA3, 2 Wm) to generate pKNlO9A (sense orientation) and pKNlO9C (antisense).The plasmid pDAD2(D. Miller, D. Pellman, and G. R. Fink, unpublished results)was created by inserting a polylinker and PH05 terminator downstream of the GAL7 promoter in pCGSlO9 (a kind gift of J. Schaum and J. Mao, Collaborative Research, Bedford, MA). Yeast strain KNYl was transformed with pKNlOSA, grown to saturation in SC containing2% galactose and lacking uracil, and cells were plated on SC containing 2% galactose without uracil and tryptophan. Trp+ colonies arose at a frequency of approximately 5 x 10-8. The plasmid was rescued from one of these Trp+clones (pKNlO9A-I), and DNA sequencing revealed that a single base pair substitution had occurred to generate an in-frameATG in the adaptor region three codons upstream of the ASA2 reading frame. E. coli mediawere prepared as described by Ausubel et al. (1989). E. coli strains were transformed by the CaCI, method (Ausubel et al., 1989).

ACKNOWLEDGMENTS

We gratefully acknowledgethe work of Paula L. Grisafi in sequencing the 5'half of theASA7 gene, the RFLP mappingdata provided by Susan Hanley and HowardGoodman (MassachusettsGeneral Hospital), and the help of Brian Keith with the pathogen infection experiment. We thank Charles Yanofsky for bacterial strains and Stephen Elledge for providing the IYES-R library prior to publication. We thank Bonnie Bartel, Judy Bender, Brian Keith, and Jennifer Normanly for critical reading of the manuscript. K.K.N. thanks past and present members of the lab, especiallyRob Last and Brian Keith, for helpful discussions. This work was supportedby the National Science Foundation. K.K.N. was supported by a National Science Foundation Graduate Fellowship. G.R.F. is an American Cancer Society Professor of Genetics.

Received March 25, 1992; accepted April 22, 1992.

REFERENCES

Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., and Struhl, K., eds (1989). Current Protocolsin

Molecular Biology. (New York: Greene PublishingAssociates and Wiley-lnterscience), Bae, Y.M., Holmgren, E., and Crawford, I.P. (1989). Rhizobiummeliloti anthranilatesynthase gene: Cloning, sequence, and expression in Escherichia coli. J. Bacteriol. ln, 3471-3478.

Arabidopsis Anthranilate Synthases

Band, L., Shimotsu, H., and Henner, D.J. (1984).Nucleotidesequence of the Bacillus subtilis trpE and trpD genes. Gene 27, 55-65. Bauerle, R., Hess, J., and French, S. (1987).Anthranilate synthase-anthranilate phosphoribosyltransferase complex and subunits

of Salmonella typhimurium. Methods Enzymol. 142,366-386. Belser, W.L., Murphy, J.B., Delmer, D.P., and Mills, S.E. (197l).End

product control of tryptophan biosynthesis in extracts and intact cells of the higher plant Nicotiana tabacum var. Wisconsin 38.Biochim. Biophys. Acta 237, 1-10, Berlyn, M.B., Last, R.L., and Fink, G.R. (1989).A gene encoding the tryptophan synthase p subunit of Arabidopis thaliana. Proc. Natl. Acad. Sci. USA 86, 4604-4608.. Biggin, M.D., Gibson, T.J., and,Hong, G.F. (1983).Buffer gradient gels and 35Slabel as an aid to rapid DNA sequence determination. Proc. Natl. Acad. Sci. USA 80::3963-3965. Botstein, D., Falco, S.C., Stewait. S.E., Brennan, M., Scherer, S., Stlnchcomb, D.T., Struhl, K., a.nd Davis, R.W. (1979).Sterile host yeasts (SHY): A eukaryotic system of biological containment for recombinant DNA experiments. Gene . .8, 17-24. Braus, G., Furter, R., Prantl, F., Nlederberger, P., and Hütter, R. (1985).Arrangement of genes cerevisiae strains. Arch. Micro Brotherton, J.E., Hauptmann, R.M.,, and Widholm, J.M. (1986).Anthranilate synthase forms in plants a:nd cultured cells of Nicotiana _' tabacum L. Planta 168, 214-221. Browne, L.M., Conn, K.L., Ayer, W.A.;'andTewarI, J.P. (1991).The camalexins: New phytoalexinsproduced in the leaves of Camelina sativa (Cruciferae). Tetrahedron Lett. 47~3909-3914. Caligiuri, M.G., and Bauerle, R. (1991).ldentification of amino acid residues involved in feedback regulation of the anthranilate synthase cynplex from Salmonella typhimurium. Evidence for an aminoterminal regulatory site. J. Biol. Chem. 266,8328-8335. Carlson, J.E., and Widholm, J.M. (1978).Separation of two forms of anthranilatesynthetasefrom 5-methyltryptophan-susceptibleand -resistant cultured Solanum tuberosum cells. Physiol. Plant. 44, 251-255. Caspar, T., Huber, S.C., and Somerville, C. (1985).Alterations in growth, photosynthesis,and respiration in a starchless mutant of Arabidopsis thaliana (L.) deficient in chloroplast phosphoglucomutase activity. Plant Physiol. 79, 11-17. Chen, E.Y., and Seeburg, P.H. (1985).Supercoil sequencing: A fast and simple method for sequencing plasmid DNA. DNA 4,165-170. Crawdord, I.P. (1989).Evolutionof a biosyntheticpathway: The tryptophan paradigm. Annu. Rev. Microbiol. 43,567-600. Devereux, J., Haeberli, P., and Smithies, O. (1984).A comprehensive set of sequence analysis programs for the VAX. Nucl. Acids Res. 12, 387-395. Devys, M., Barbier, M., Loiselet, I., Rouxel, T., Sarnlguet, A., Kollmann, A., and Bousquet, J.-F. (1988).Brassilexin, a novel sulphur-containingphytoalexin from Brassicajuncea L. (Cruciferae). Tetrahedron Lett. 29, 6447-6448. Dong, X., Mindrlnos, M., Davis, K.R., and Ausubel, F.M. (1991).Induction of Arabidopsis defense genes by virulent and avirulent Pseudomonas syringae strains and by a cloned avirulence gene. Plant Ceil 3, 61-72. Elledge, S.J., Mulllgan, J.T., Ramer, S.W., Spottswood, M., and Davis, R.W. (1991).LYES: A multifunctionalcDNA expression vector for the isolation of genes by complementation of yeast and Escherichiacoli mutations. Proc. Natl. Acad. Sci. USA 88,1731-1735. '

731

Essar, D.W., Eberly, L., Hadero, A., and Crawford, I.P. (1990).Iden-

tification and characterization of genes for a second anthranilate synthase in Pseudomonas aeruginosa: lnterchangeabilityof the two anthranilate synthases and evolutionary implications. J. Bacteriol. 172, 884-900. Fenwick, G.R., Heaney, R.K., and Mullin, W.J. (1983).Glucosinolates and their breakdown products in food and food plants. CRC Crit. Rev. Food Sci. Nutr. 18, 123-201. Finer-Moore,J., and Stroud, R.M. (1984).Amphipathic analysisand possibleformation of the ion channel in an acetylcholine receptor. Proc. Natl. Acad. Sci. USA 81, 155-159. Gavel, Y., and von Heijne, G. (1990). A conservedcleavage-sitemotif inxhloroplast transit peptides. FEBS Lett. 261, 455-458. Gilchrlst, D.G., and Kosuge, T. (1980).Aromatic amino acid biosynthesis and its regulation. In The Biochemistry of Plants, Vol. 5,P.K. Stumpf and E.E. Conn, eds(New York Academic Press),pp. 507-531. Goncharoff,P., and Nichols, B.P. (1984).Nucleotidesequenceof Escherichia coli pabB indicates a common evolutionary origin of p-aminobenzoate synthetase and anthranilate synthetase. J. Bacteriol. 159,57-62. Hankins, C.N., and Mills, S.E. (1976). Anthranilate synthase-amidotransferase (combined). A novel form of anthranilate synthase from Euglena gradis. J. Biol. Chem. 251, 7774-7778. Hankins, C.N., Largen, M.T., and Mills, S.E. (1976).Some physical characteristics of the enzymes of L-tryptophanbiosynthesis in higher plants. Plant Physiol. 57, 101-104. Henikoff, S. (1984).Unidirectionaldigestion with exonuclease 111 cre-

ates targetedbreakpoints for DNA sequencing. Gene 28,351-359. Hinnebusch, A.G. (1988).Mechanisms of gene regulation in the

general control of amino acid biosynthesis in Saccharomyces cefevisiae. Microbiol. Rev. 52, 248-273. Hogge, L.R., Reed, D.W., Underhill, E.W., and Haughn, G.W. (1988).

HPLC separation of glucosinolates from leaves and seeds of Arabidopsis thaliana and their identificationusing thermospray liquid chromatographyhnassspectrometry. J. Chromatogr. Sci. 26, 551-556. Hulett, F.M., and DeMoss, J.A. (1975).Subunit structure of anthranilate synthase from Neurospora crassa. J. Biol. Chem. 250,

6648-6652. Hütter, R., Niederberger, P., and DeMoss, J.A. (1986).Tryptophan

biosynthetic genes in eukaryotic microorganisms. Annu. Rev. Microbiol. 40, 55-77. Ito, J., and Yanofsky, C. (1969).Anthranilate synthase, an enzyme

specified by the tryptophan operon of Escherichia coli: Comparative studies on the complex and the subunits. J. Bacteriol. 97, 734-742. Jackson, E.N., and Yanofsky, C. (1974).Localizationof twofunctions of the phosphoribosyl anthranilate transferase of Escherichia coli to distinct regionsof the polypeptidechain. J. Bacteriol. 117,502-508. Jefferson, R.A., Kavanagh, T.A., and Bevan, M.W. (1987).GUS fu-

sions: P-Glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J. 6, 3901-3907. Jensen, R.A. (1986).The shikimate/arogenatepathway: Link between

carbohydrate metabolism and secondary metabolism. Physiol. Plant. 66, 164-168. Joshi, C.P. (1987).Putativepolyadenylationsignals in nuclear genes

of higher plants: A compilation and analysis. Nucl. Acids Res. 15, 9627-9640.

732

The Plant Cell

Keith, B., Dong, X., Ausubel, F.M., and Fink, G.R. (1991). Differentia1 induction of 3-deoxy-D-arabino-heptulosonate 7-phosphate

Pearson, W.R., and Lipman, D.J. (1988). lmproved tools for biological sequence comparison.Proc. Natl. Acad. Sci. USA 85,2444-2448.

synthasegenes in Arabidopsisthalianaby wounding and pathogenic . . attack. Proc. Natl. Acad. Sci. USA 88, 8821-8825. Klee, H.J., Muskopf, Y.M., and Gasser, C.S. (1987). Cloning of an Arabidopsis thaliana gene encoding 5-enolpyruvylshikimate-3phosphate synthase: Sequence analysis and manipulation to obtain glyphosate-tolerantplants. MOI.Gen. Genet. 210, 437-442. Koritsas, V.M., Lewis, J.A., and Fenwick, G.R. (1989). Accumulation of indoleglucosinolates in PsylliodeschrysocephalaL.-infested, or -damaged tissues of oilseedrape (Brassicanapus L.). Experientia 45,493-495. Last, R.L., and Fink, G.R. (1988). Tryptophan-requiring mutants of the plant Arabidopsis thaliana. Science 240, 305-310. Last, R.L., Blsslnger, P.H., Mahoney, D.J., Radwanski, E.R., and Fink, G.R. (1991). Tryptophan mutants in Arabidopsis: The consequences of duplicatedtryptophan synthase p genes. Plant Cell 3, 345-358. Learned, R.M., and Fink, G.R. (1989). 3-Hydroxy-3-methylglutarylcoenzyme A reductase from Arabidopsis thaliana is structurallydistinct from the yeast and animal enzymes. Proc. Natl. Acad. Sci. USA 86, 27'79-2783. Marks, M.D., West, J., and Weeks, D.P. (1987). The relatively large beta-tubulingene family of Arabidopsis contains a member with an unusual transcribed 5' noncoding sequence. Plant MOI.Biol. 10, 91-104. Matsui, K., Miwa, K., and Sano, K. (1987). Two single-base-pairsubstitutions causing desensitizationto tryptophan feedback inhibition of anthranilate synthase and enhanced expression of tryptophan genes of Brevibacterium lactofermentum. J. Bacteriol. 169, 5330-5332. Mazur, B.J., Chui, C.-F., and Smith, J.K. (1987). lsolation and characterizationof plant genes coding for acetolactatesynthase, the target enzyme for two classes of herbicides. Plant Physiol.85, 1110-1117. Mousdale, D.M., and Coggins, J.R. (1986). Detectionand subcellular localization of a higher plant chorismate synthase. FEBS Lett. 205, 328-332. Nagy, F., Kay, S.A., and Chua, NrH. (1988). Analysis of gene expression in transgenic plants. In Plant Molecular Biology Manual, S.B. Gelvin and R.A. Schilperoort, eds (Dordrecht, The Netherlands: Kluwer Academic Publishers), pp. B4/1-B4/29. Nam, H.-G., Glraudat, J., den Boer, B., Moonan, F., Loos, W.D.B., Hauge, B.M., and Goodman, H.M. (1989). Restriction fragment length polymorphism linkage map of Arabidopsis thali?na.-Plant Cell 1, 699-705. Nichols, B.P., van Cleemput, M., and Yanofsky, C. (1981). Nucleotide sequence of Escherichia coli trpE. Anthranilate synthetase component I contains no tryptophan residues. J. MOI. Biol. 146, 45-54. Ozenberger, B.A., Brickman, T.J., and Mclntosh, M.A. (1989). ., Nucleotidesequence of Escherichia coli isochorismate synthetase gene entC and evolutionary relationship of isochorismate synthetase and other chorismate-utilizing enzymes. J. Bacteriol. 1T1, 775-783. Paluh, J.L., Zalkin, H:, Betsch, D., and Weith, H.L. (1985). Study of anthranilate synthasefunctionby replacementof cysteine84 using site-directed mutagenesis. J. Biol. Chem. 260, 1889-1894.

Poulsen, C., and Verpoorte, R. (1991). Roles of chorismate mutase,

isochorismate synthase and anthranilate synthase in plants. Phytochemistry 30, 377-386. Prantl, F., Strasser, A., Aebi, M., Furter, R., Niederberger, P., Kirschner, K., and Huetter, R. (1985). Purification and characterization of the indole-3-glycerolphosphate synthaselanthranilate synthase complex of Sacchammyces cerevisiae. Eur. J. Biochem. 146, 95-100. Ranch, J.P., Rick, S., Brotherton, J.E., and Widholm, J.M. (1983). Expression of 5-methyltryptophan resistancein plants regenerated from resistant cell lines of Daturiainnoxia. Plant Physiol. 7l, 136140. Sanger, F., Nicklen, S., and Coulsen, A.R. (1977). DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74, 5463-5467. Schulze-Siebert,D., and Schultz, G. (1989). Formation of aromatic

amino acids and valine from 14C02or 3-[U-14C]phosphoglycerate by isolated intact spinach chloroplasts.Evidencefor a chloroplastic Sphosphoglycerate 2-phosphoglycerate phosphoenolpyruvate -,pyruvate pathway. Plant Sci. 59, 167-174.

-

-

Sherman, F., Hlcks, J.B., and Fink, G.R. (1986). Methods in Yeast

Genetics. (ColdSpring Harbor, NY: Cold Spring Harbor Laboratory). Singh, B.K., Siehl, D.L., and Connelly, J.A. (1991). Shikimate pathway: Why does it mean so much to so many? In Oxford Surveys of Plant Molecular Biology, Vol. 7, B.J. Miflin, ed (New York: Oxford University Press), pp. 143-185. Sung, Z.R. (1979). Relationshipof indole-3-aceticacid and tryptophan concentrations in normal and Smethyltryptophan-resistantcell lines of wild carrots. Planta 145, 339-345. Tabor, S., and Rlchardson, C.C. (1987). DNA sequence analysiswith a modified bacteriophage T7 DNA polymerase. Proc. Natl. Acad. Sci. USA 84, 4767-4771. Takasugi, M., Katsui, N., and Shirata, A. (1986). lsolation of three

nove1 sulphur-containing phytoalexinsfrom the Chinese cabbage Brassica campestris L. ssp. pekinensis (Cruciferae).J. Chem. SOC. Chem. Commun. 1986, 1077-1078. Tsujl, J., Jackson, E.P., Gage, D.A., Hammerschmldt, R., and Somerville, S.C. (1992). Phytoalexinaccumulation in Arabidopsis

thaliana during the hypersensitivereaction to Pseudomonas syringae pv. syringae. Plant Physiol. 98, 1304-1309. Vielra, J., and Messing, J. (1987). Productionof single-strandedplasmid DNA. Methods Enzymol. 153, 3-11. von Heijne, G., Steppuhn, J., and Herrmann, R.G. (1989). Domain

structure of mitochondrialand chloroplast targeting peptides. Eur. J. Biochem. 180, 535-545. Widholm, J.M. (197l). Control of tryptophan biosynthesis in plant tis-

sue cultures: Lack of repression of anthranilate and tryptophan synthetases by tryptophan. Physiol. Plant. 25, 75-79. Widholm, J.M. (1972a). Anthranilatesynthetase from B-methyltrypto-

phan-susceptible and -resistant cultured Daucus camta cells. Biochim. Biophys. Acta 279, 48-57. Widholm, J.M. (1972b). Cultured Nicotiana tabacum cells with an altered anthranilate synthetase which is less sensitive to feedback inhibition. Biochim. Biophys. Acta 261, 52-58.

Arabidopsis Anthranilate Synthases

Wldholm, J.M. (1974). Control of aromatic amino acid biosynthesis

in culturedplant tissues: Effect of intermediatesand aromatic amino acids on free levels. Physiol. Plant. 30, 13-18. Widholm, J.M. (1977). Relationbetween auxin autotrophy and tryptophan accumulation in cultured plant cells. Planta 134, 103-108. Yanofsky, C., and Crawford, I.P. (1987). The tryptophan operon. In Escherichiacoli and Salmonella typhimurium: Cellular and Molecular Biology, F.C. Neidhardt,J.L. Ingraham, K.B. Low, B. Magasanik, M. Schaechter, and H.E. Umbarger, eds (Washington, DC: American Society for Microbiology), pp. 1453-1472.

733

Zalkin, H. (1980). Anthranilate synthase: Relationships between bifunctional and monofunctionalenzymes. In MultifunctionalProteins, H. Bisswanger and E. Schmincke-Ott, eds (New York: John Wiley & Sons), pp. 123-149.

Zalkin, H., Paluh,J.L., van Cleemput, M., Moye, W.S., and Yanofsky, C. (1984). Nucleotidesequence of Saccharumycescerevisiae genes TRP2 and TRP3 encoding bifunctionalanthranilate synthase:indole3-glycerol phosphate synthase. J. Biol. Chem. 259, 3985-3992.

Two anthranilate synthase genes in Arabidopsis: defense-related regulation of the tryptophan pathway. K K Niyogi and G R Fink Plant Cell 1992;4;721-733 DOI 10.1105/tpc.4.6.721 This information is current as of April 13, 2015 Permissions

https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298 X

eTOCs

Sign up for eTOCs at: http://www.plantcell.org/cgi/alerts/ctmain

CiteTrack Alerts

Sign up for CiteTrack Alerts at: http://www.plantcell.org/cgi/alerts/ctmain

Subscription Information

Subscription Information for The Plant Cell and Plant Physiology is available at: http://www.aspb.org/publications/subscriptions.cfm

© American Society of Plant Biologists ADVANCING THE SCIENCE OF PLANT BIOLOGY