Mutation of GT-1 binding sites in the Pr-1A promoter ... - Springer Link

0 downloads 0 Views 116KB Size Report
of binding of nuclear protein GT-1 to far-upstream regions (−902 to −656) of the PR-1a ... A possible model for GT-1's mode of action in PR-1a gene expression.
Plant Molecular Biology 40: 387–396, 1999. © 1999 Kluwer Academic Publishers. Printed in the Netherlands.

387

Mutation of GT-1 binding sites in the Pr-1A promoter influences the level of inducible gene expression in vivo Annemarie S. Buchel, Frans Th. Brederode, John F. Bol and Huub J.M. Linthorst∗ Institute of Molecular Plant Sciences, Gorlaeus Laboratories, Leiden University, P.O. Box 9502, 2300 RA Leiden, Netherlands (∗ author for correspondence) Received 5 June 1998; accepted in revised form 12 January 1999

Key words: gene expression, GT-1, PR-1a, PR proteins, salicylic acid-induced, transcription factors

Abstract Infection of Nicotiana tabacum Samsun NN with tobacco mosaic virus (TMV) results in a hypersensitive plant response and leads to systemic acquired resistance (SAR). The induction of SAR is mediated by the plant hormone salicylic acid (SA) and is accompanied by the induced expression of a number of genes including the pathogenesisrelated (PR) gene 1a. Previously, it has been found that TMV infection and SA treatment resulted in a reduction of binding of nuclear protein GT-1 to far-upstream regions (−902 to −656) of the PR-1a gene. To test if GT-1 is a negative regulator of PR-1a gene expression, the effects of mutations in the seven putative GT-1 binding sites in this region were studied in vitro using dimethyl sulfate interference footprinting and band shift assays. This showed that at least one of the seven sites is indeed a GT-1 binding site. However, when tested in transgenic plants, the mutations did not result in constitutive expression of the chimeric PR-1a/GUS transgene, while inducible expression after SA treatment was decreased. The results suggest that binding of GT-1-like proteins to far-upstream PR-1a promoter regions indeed influences gene expression. A possible model for GT-1’s mode of action in PR-1a gene expression is discussed.

Introduction Upon infection with tobacco mosaic virus (TMV), necrotic lesions appear on the leaves of Nicotiana tabacum cv. Samsun NN. This hypersensitive response of the plant inhibits virus spread and induces systemic acquired resistance (SAR) against further pathogen infection (Ross, 1986). The onset of SAR is accompagnied by the synthesis of a number of host factors including pathogenesis-related (PR) proteins (van Loon and Antoniw, 1988). In addition to pathogen infection, accumulation of PR proteins is differentially induced during plant development and by external stimuli such as UV light, wounding or treatment with the hormones salicylic acid (SA) and ethylene (Memelink et al., 1990; Brederode et al., 1991; Pierpoint, 1997). Tobacco PR proteins have been classified into different families represented amongst others by PR-1 proteins, β-1,3-glucanases (PR-2), chitinases (PR-3), thaumatin-like proteins (PR-5) and

proteinase inhibitors (PR-6) (van Loon et al., 1994). The tobacco PR-1 protein family encompasses three acidic (PR-1a, -1b and -1c) and two basic (PR-1g and PRB-1b) proteins (Gianinazzi et al., 1970; Ahl et al. 1982; Cornelissen et al., 1986; Pfitzner et al., 1987; Eyal et al., 1992; Niderman et al., 1995). PR1 proteins are thought to play a rolein resistance to oomycete fungal infection (Alexander et al., 1993; Niderman et al., 1995). Of all tobacco PR genes, expression of the genes encoding acidic PR-1 proteins is most highly induced by TMV infection and SA treatment (Ward et al., 1991). Studies by van de Rhee et al. (1990) with PR1a promoter fragments fused to the β-glucuronidase (GUS) reporter gene revealed that 0.9 kb of the PR1a upstream region is necessary for high levels of inducible gene expression. Stepwise deletion of sequences between −902 and −689 gradually reduced the level of inducible gene expression. Since no induction by TMV or SA was found for constructs with

388

Figure 1. Schematic representation of the PR-1a promoter region from −906 to −571. The position of fragments A, B and B0 -C, relative to the PR-1a promoter (solid line), is indicated. The seven putative GT-1 binding sites are indicated by the open circles.

PR-1a upstream sequences of −643 bp or shorter, a TMV- and SA-responsive element is thought to be present between the −689 and −643. Several studies have been performed to identify the cis-acting sequences in the upstream region of the PR1a gene and the trans-acting factors interacting with these sites (Strompen et al., 1988; Ward et al., 1991; Hagiwara et al., 1993; Buchel et al., 1996). Results of Hagiwara et al. (1993) suggested that negative regulators bound to two regions of the promoter (−184 to −172 and −68 to −51). On the other hand, evidence for positive regulation of PR-1a gene expression was obtained by Yang and Klessig (1996), who found that a TMV-inducible myb homologue specifically interacted with the promoter from −520 to −514. Another positive regulator was recently identified by Strompen et al. (1998) who showed that transcription factor TGA1a can interact with an as-1-like element in the PR-1a promoter (−592 to −577). Interestingly, Buchel et al. (1996) found that the binding activity in nuclear extracts of tobacco of as-1-like proteins increases after TMV infection and SA treatment (Figure 6 in Buchel et al., 1996). In previous studies we divided the 0.9 kb PR-1a promoter region into eight fragments (fragments A to H) and analysed the binding of tobacco nuclear proteins to the fragments in gel retardation assays (Buchel et al., 1996). The experiments showed that the PR1a promoter contains a number of elements that bind the same factors with different affinity. The factors involved in these interactions are GT-1-like proteins (Buchel et al., 1996). GT-1 is a plant nuclear activity originally identified as a factor that binds to the conserved elements box II and box III present in the light-responsive element (LRE, −166 to −55) of the pea Rbc-3A promoter (Green et al., 1987; Jefferson et al., 1987; Kuhlemeier et al., 1989; Davis et al., 1990). GT-1 can exist as a homotetramer in solution (Lam, 1995). Possibly,

a GT-1 complex simultanously interacts with multiple promoter regions. Boxes II and III are essential for the phytochrome-responsiveness conferred by the LRE. However, since box III appears to be a much weaker binding site for GT-1 than box II the interaction between box III and GT-1 may have only a minor effect on transcriptional regulation directed by the LRE (Gilmartin and Chua, 1990; Gilmartin et al., 1992). The ability of a tetramer of box II to confer light-responsive gene expression on a heterologous promoter in vivo suggests that GT-1 binding is involved in Rbc-3A gene induction by light (Kuhlemeier et al., 1987; Lam and Chua, 1990). Green et al. (1988) identified a box II core sequence of six residues (GGTTAA) that is critical for GT-1 binding in vitro. Sequences homologous to the GT-1 binding box II (core) sequence have also been found in promoter regions of other genes, for example in defence-related tobacco PR-1a, Catharanthus roseus Tdc, soybean chs15, circadian clockcontrolled Arabidopsis cab2, light-repressed rice phya and tissue-specific spinach rps1 (Dehesh et al., 1990; Lawton et al., 1990; Anderson et al., 1994; Villain et al., 1994; Buchel et al., 1996; Ouwerkerk, 1997). Analysis of the effect of mutations that prevent binding of (factors related to) GT-1 to the upstream regions of these differentially regulated genes may shed light on additional functions of GT-1 in control of gene expression. Villain et al. (1996) studied the cell-specific transcription of the GUS reporter gene controlled by tetrameric wild-type and mutant GT-1 binding box II elements in transgenic tobacco. Mutation of the GT-1 binding sites resulted in a decrease in promoter activity in leaves. These data suggest that binding of GT-1 has a positive function on transcription in leaves. The box II mutation provoked an increase of transcriptional activity in tobacco roots. A similar effect on root-specific transcription was reported earlier for the mutated S1F binding site, a box II-related sequence present in the rps1 promoter (Villain et al., 1994). Apparently, binding of GT-1 or the GT-1related factor S1F to box II (-related) sequences acts as a root-specific repressor of transcription. More recently, Ouwerkerk et al. (1997) showed that mutation in GT-1 binding sites present in the tryptophan decarboxylase (Tdc) promoter did not affect the expression of the GUS reporter gene in transgenic tobacco leaves. However, binding of GT-1 appeared to activate cotyledon-specific gene expression, both in the presence and the absence of light. Together, these data

389 indicate that GT-1 can serve as either a repressor or an activator of transcription. The binding of GT-1-like factors to the far upstream region (−902 to −656) of the tobacco PR-1a gene is reduced after TMV infection and SA treatment (Buchel et al., 1996). If the removal of GT-1-like proteins contributes to the induction of PR-1a gene expression, mutations that prevent GT-1-like protein binding may result in a constitutive expression of the gene. Seven putative GT-1 binding sites, resembling the box II core sequence, can be distinguished in the PR-1a upstream region between −902 and −643 (Buchel et al., 1996). By dimethyl sulfate interference footprint experiments we showed that at least one putative GT-1 binding site in PR-1a promoter fragment A (−902 to −787) is involved in the interaction with GT-1-like proteins. Other residues within fragment A appear to be important for binding of proteins distinct from GT-1. To investigate the possible role of GT-1-like factors in regulation of the PR-1a gene the seven putative GT-1 binding sites were mutated. The effect of the mutations on protein binding in vitro and chimeric PR-1a/GUS gene expression in vivo was analysed. Although point mutations in three of the seven putative GT-1 binding sites reduced the binding of GT-1-like factors in vitro, the mutations did not result in constitutive reporter gene expression in transgenic plants. The SA-inducible expression of the chimeric PR-1a/GUS gene in tobacco was decreased as a result of mutation of GT-1 binding sites. No additive effect of the seven mutations could be detected. Materials and methods Preparation of the wild-type and mutant PR-1a promoter fragments The PR-1a promoter fragments A (−906 to −764, Figure 1), B (−787 to −656) and B0 -C (−787 to −571) were generated by PCR. Point mutations were introduced by sequential PCR steps. The position of mutations mi and m1 to m7 within the PR-1a promoter is shown in Figure 2A. The numbering of the nucleotides in the PR-1a upstream sequence is according to Payne et al. (1988). For the introduction of a mutation one upstream and one downstream primer carrying the same mutation were used. Each mutant primer set had at least 12 bases of overlap. In the first PCR step two overlapping subfragments were synthesized with a mutant primer set and the wild-type

primers corresponding to the 50 and the 30 end of the promoter fragment. The first subfragment was generated using the wild-type upstream primer and a mutant downstream primer. The second subfragment was synthesized with the wild-type downstream primer and a mutant upstream primer. The two subfragments were denatured and annealed and the resulting fragment was amplified by a second PCR step with the 50 and 30 primers (Ausubel et al., 1995). The PCR products were gel-purified and cloned. The identity of each cloned fragment was confirmed by sequencing. Mutations were introduced one at the time, for example for the synthesis of a fragment with point mutations m1 and m2 the cloned fragment with mutation m1 was used as a template. Since fragment B0 -C (Figure 1) was synthesized with the upstream primer M5-r and the downstream C2 primer, this fragment contained PR-1a promoter sequences from −720 to −571, including mutation m5. For the synthesis of the B0 -C fragment with mutations m6 and m7 the cloned B-C fragment carrying m6 and m7 was used as a template. Gel retardation assays Nuclear protein extracts were prepared from leaves of uninduced tobacco (Nicotiana tabacum cv. Samsun NN) as described by Green et al. (1989). DNA fragments for gel retardation assays obtained after digestion with the proper enzymes, were gel-purified and end-labelled with α-32 P-dCTP, using Klenow polymerase. DNA-binding reactions typically contained 0.5 ng labelled DNA fragment, 3 µg poly(dIdC)·poly(dI-dC) (Pharmacia), and 5 µg nuclear protein extract, in 10 µl of binding buffer (20 mM HEPES-KOH pH 6.7, 40 mM KCl, 0.1 mM EDTA, 10% glycerol, 1 mM DTT, 0.8 mM PMSF). To the reaction mixture with fragment B0 -C a 20-fold molar excess of an internal 231 bp EcoRV fragment of the βglucuronidase (GUS) gene was added as a non-specific competitor (Jefferson et al., 1987). The binding reactions were incubated for 30 min at room temperature and subsequently loaded on a 5% (w/v) polyacrylamide gel in 44.5 mM Tris-borate, 44.5 mM boric acid, 1 mM EDTA (0.5× TBE). The gels were dried on chromatography paper and autoradiographed. Dimethyl sulfate interference footprinting Methylation interference assays were performed essentially as described by Green et al. (1987). Endlabelled probes were partially methylated by treatment

390 Figure 2. The PR-1a promoter sequence from −907 to −638. A. The extra nucleotides introduced into the promoter sequence to create HindIII sites for cloning are given in bold. The putative GT-1 binding sites are indicated (I to VII). Nucleotides that replace PR-1a promoter residues in mutations mi and m1 to m7 are shown beneath the wild-type sequence. B. Schematic representation of the constructs used to transform tobacco. PR-1a promoter sequences are represented by the black bars. The GUS box depicts the GUS reporter gene and the nopaline synthase terminator region. The position of the HindIII restriction sites is indicated. PR-1a promoter mutations m1 to m7 are shown as black circles. The position of each mutation relative to the transcription start site is given.

with dimethyl sulfate (DMS). Band shift reaction mixtures contained 32 µg nuclear protein extract derived from uninduced tobacco, 24 µg poly(dI-dC)·poly(dIdC) and 4 ng probe in binding buffer. The reactions were electrophoresed in 5% (w/v) acrylamide gels in 0.5× TBE. The wet gel was autoradiographed and the DNA was isolated from the complexes by overnight elution in 0.5 M ammonium acetate, 0.1% SDS, 1 mM EDTA. After phenol extraction and ethanol precipation the DNA samples were chemically cleaved with piperidine (G reaction) or HCl (A>G reaction; Maxam and Gilbert, 1977), precipitated with butanol and/or ethanol and resuspended in formamide loading buffer. Equal amounts of bound and unbound DNA fractions were loaded on a 6% sequencing gel. After electrophoresis the gel was dried under vacuum and autoradiographed. Vector construction All constructs used for transformation of tobacco are presented in Figure 2B. Cloning was started with PRB643 (van de Rhee et al., 1990), a binary plant vector containing nucleotides −643 to +29 of the PR-1a gene fused to the GUS reporter gene and the T-Nos terminator in pBI101 (Jefferson et al., 1987). In PRB643 a HindIII site was present at position −643 in the PR-1a promoter. To generate a HindIII site in the wild-type PR-1a promoter sequence at position −643, a subfragment was synthesized by PCR with primers containing a HindIII site. The wild-type PR1a promoter sequence from −906 to +29, cloned in pIC19H (Marsch et al., 1984), was used as a template. The subfragment was gel-purified and cloned into the HindIII site of PRB643. Mutant PR-1a promoter/GUS constructs were synthesized in the same way with the same mutation primers as described for the PR-1a promoter band shift probes. The mutant constructs M1 to M16 were obtained using the mutation primers in combination with the HindIII primers. The orientation

391 and identity of each cloned fragment were confirmed by sequencing. Plant transformation and preparation of leaf extracts The PR-1a promoter/GUS constructs were transformed via Agrobacterium tumefaciens LBA4404 to N. tabacum cv. Samsun NN using the leaf disc transformation procedure (Horsch et al., 1985; Jefferson et al., 1987). Transgenic plants were selected with 100 µg/ml kanamycin and 200 µg/ml carbenicillin. Leaf discs of 2.5 cm were floated for 24 h at 26 ◦ C on distilled water or on 1 mM sodium salicylate, pH 7. Leaf discs were homogenized in 1.5 ml GUS lysis buffer (50 mM NaH2 PO4 pH 7, 10 mM EDTA, 0.1% Triton X-100, 0.1% sarcosyl, 10 mM 2-mercaptoethanol; Jefferson et al., 1987), using a Potter homogenizer (Braun). The extracts were directly frozen in liquid nitrogen and stored at −80 ◦ C until measurement of GUS activity. β-Glucuronidase assays Frozen homogenates were quickly melted at 37 ◦ C and placed on ice. After 5 min of centrifugation the GUS enzyme activity of the supernatant was analysed according to Jefferson et al. (1987). Samples of 20 µl of the supernatant were incubated in a final volume of 100 µl of GUS lysis buffer containing 1 mM 4-methylumbelliferyl glucuronide at 37 ◦ C in black microtitre plates (Flow Laboratories). After 18 h the reactions were stopped by the addition of 25 µl 1 mM Na2 CO3 . The GUS activity of the samples was measured with a Titertek Fluoroskan II apparatus (Flow Laboratories). Protein concentrations were determined according to Bradford (1979). GUS activity is given in pmol 4-methylumbelliferone (MU) per hour per gram soluble protein.

Results Nucleotides involved in protein binding to PR-1a promoter fragment A The PR-1a promoter region between −902 to −643 includes seven potential GT-1 binding sites (Figures 1 and 2A; Buchel et al., 1996), three of which are present in fragment A (−906 to −764). To analyse if residues involved in protein binding to fragment A coincide with one or more of these potential GT1 binding sites, a DMS interference footprint assay

was performed. For this experiment the bottom strand of fragment A was radioactively labelled and partially methylated with dimethyl sulfate before it was used in gel retardation reactions. Upon incubation with the nuclear extract, fragment A was retarded in two protein-DNA complexes (the third retarded band immediately above the position of the unbound probe, as previously described (Buchel et al., 1996), is not always clearly visible). Of these complexes the lowmobility complex has been shown to contain GT-1-like factors (Buchel et al., 1996). Recent supershift experiments with antiserum against GT-1 support this finding (J. Memelink, results not shown). Besides unbound fragment A, probe A present in the two protein-DNA complexes was eluted from the acrylamide gel. The DNA was chemically cleaved at the methylated guanine residues (Figure 3A) or at both the adenine and guanine residues (Figure 3B) as described by Maxam and Gilbert (1977). As shown in Figure 3A, the G nucleotide at position −816 was partially depleted from the bound fraction derived from the low-mobility GT-1-like complex (B) as compared to the free fraction (F). Apparently, this G residue is involved in the binding of GT-1-like proteins to fragment A. This G residue is present in a putative GT-1 binding site. The G residues at −854 and −855 play no part in binding to the GT-1-like proteins within the low-mobility protein-DNA complex since molecules methylated at these residues are equally distributed in both the bound and the unbound fractions (Figure 3A). Because only the lower strand of fragment A was end-labelled, methylation interference at the G-residue present in the putative GT-1 binding site at −844 could not be analysed. To detect residues in fragment A involved in the formation of the fast-migrating protein-DNA complex, the DMS interference assay shown in Figure 3B was performed. In this experiment four residues (CTTT, −855 to −858) of the putative GT-1 binding site at −854 were found to be involved in protein binding. Methylation of the G residue at −816 did not interfere with the formation of the fast-migrating complex with PR-1a promoter fragment A. In addition the residues CTT (−860 to −862) and the T residues at −869, −827 and −818 appeared to be involved in protein-DNA interaction.

392

Figure 4. Effect of mutations in putative GT-1 binding sites in PR-1a promoter fragment A on binding of tobacco nuclear proteins. Nuclear protein extract from uninduced tobacco was incubated with labelled wild-type fragment A or with fragment A carrying mutations mi, m1, m2 and/or m3, as indicated. Plus and minus signs indicate the presence or absence of nuclear proteins in the reaction mixtures, respectively. Arrowheads at the left indicate the position of the low-, intermediate- and high-mobility complexes, respectively.

Figure 3. DMS interference footprint assay using nuclear proteins from non-induced tobacco and PR-1a promoter fragment A end-labelled at the bottom strand. Lanes F and B show the reaction products after chemical cleavage, at the methylated G residues (Panel A, left) or at both the A and G residues (Panel B right), of the free (f) and the bound (B) gel retardation fractions. Methylation of the G residue at −816 (designated by the arrow) interferes with the protein binding that results in the formation of the low-mobility GT-1-like protein-DNA complex with fragment A (Panel A). In Panel B the A and G residues that are involved in the formation of the fast-migrating GT-1-like complex with fragment A are marked.

Point mutation decreases binding of GT-1-like factors to upstream PR-1a promoter fragment To further define their role in complex formation, the seven putative GT-1 binding sites in the PR-1a promoter upstream of −643 were mutated by PCR. The protein binding to mutant and wild-type promoter fragments A, B, and B0 -C was studied by gel retardation analysis. A schematic representation of mutations m1 to m7 is shown in Figure 2A. In addition to mutating the putative GT-1 binding site at −857, mutation mi replaces a T residue at −862. Figure 4 shows the

binding of wild-type fragment A and fragments A containing the separate mutations mi, m1, m2, m3 or m2+m3 to nuclear proteins from uninduced tobacco. Although methylation interference analysis revealed the importance of the T residues at −857 and −862 for formation of the fast-migrating protein-DNA complex, replacement of these residues by two A nucleotides did not change the binding pattern (Figure 4, lane mi+). The C-residues at position −854 and −855 were changed into AT by mutation m1. As shown by DMS interference footprinting the latter C residue appears to be involved in the formation of the fast-migrating complex with fragment A. However, introduction of m1 into fragment A did not result in a change of the fast-mobility protein-DNA complex. In contrast, an additional complex is detected that migrates in between the slow- and the fast-migrating complexes (Figure 4, lane m1+). An intermediate complex was also formed with fragment A containing mutation m2 (Figure 4, lane m2+). This mutation resulted in a slight reduction of the low-mobility GT-1 complex, while the fast-migrating complex was nearly absent. Mutation m3 replaces the C residue at −816 by a T. This residue was previously shown to be involved in formation of the slow-migrating GT-1 complex. As expected, the introduction of point mutation m3 resulted in a strongly reduced formation of the GT-1 complex (Figure 4, lane m3+). The fast-migrating protein-DNA complex with fragment A was not affected by mutation m3 (Figure 4, lane m3+). As for fragments carrying m1 or m2, fragment A containing

393 able to form GT-1 complexes with nuclear proteins from uninduced tobacco (data not shown). GUS competitor DNA partly interfered with the formation of a GT-1 complex using the overlapping wild-type promoter fragment B0 -C, containing putative binding sites VI and VII, and the corresponding fragment with mutations m6 and m7. As a result, characterization of these sites as GT-1 binding sites was inconclusive. Point mutations in the PR-1a promoter reduce SA-inducible gene expression in vivo

Figure 5. Effect of mutations in putative GT-1 binding sites in the PR-1a −902/−643 promoter region on gene expression. N. tabacum Samsun NN plants were transformed with the indicated constructs and leaf discs were assayed for GUS activity after floating on SA or H2 O. For each transformant the SA-inducible GUS activity (GUS activity on SA minus GUS activity on H2 O) was determined. The data are graphically displayed in the corresponding block diagram (GUS activity in 0.1 pmol MU per hour per g soluble protein, fixed at one decimal). The number of individual transformants per construct is indicated by ‘n’.

mutation m3 formed a complex of intermediate mobility. Fragment A including both mutations m2 and m3 showed a clearly decreased formation of the low and the fast complexes with the tobacco nuclear extract. Mutations m4 and m5 replace residues in two potential GT-1 binding sites in fragment B (position −787 to −656, Figure 1). However, in spite of the presence of mutation m4 or m5 fragment B was still

To analyse the effect of mutations m1 to m7 in vivo, Samsun NN tobacco was transformed with wild-type and mutant 0.9 kb PR-1a promoter/GUS constructs (Figure 2B). A HindIII restriction site was introduced in all constructs at position −643 to enable more efficient cloning. Transgenic plants carrying the wildtype construct containing this HindIII site showed an average SA-inducible GUS activity similar to plants containing construct PRB902 without the HindIII site (van de Rhee et al., 1990). As shown in Figure 2B, mutant constructs M1, M2 and M3 carry the singlepoint mutations m1, m2 and m3, respectively. For significant interaction nuclear GT-1 requires at least two box II binding sites positioned in tandem. A trimer or tetramer of box II has a higher affinity for GT-1 than a dimer (Green et al., 1988). To study the additive effect of mutations in potential GT-1 binding sites within the PR-1a promoter upstream of −643, mutations m1 to m7 were introduced into the PR-1a promoter/GUS constructs one by one (mutants M12 to M17, Figure 2B). The effects of each construct were analysed by measuring the GUS activity of 14 to 20 independent primary transformants, after incubation of leaf discs on water or 1 mM SA. Figure 5 shows the data obtained for the different constructs in ten graphs. Each graph contains the results of the plants transformed with a particular construct (e.g. wild type, M1, M2). The number of analysed plants carrying the same construct is represented by ‘n’. The level of GUS activity is divided into thirteen categories, as indicated in the legend to Figure 5. For each construct the number of plants per category is shown. The level of GUS activity in each plant after SA treatment was corrected for the background level of GUS activity measured after floatation of the leaf material on H2 O. In most cases H2 O treatment resulted in GUS expression levels that could be

394 classified as category 1 (0–0.1 pmol MU per hour per µg soluble protein). A first glance at the diagrams of Figure 5 shows that the GUS activity of plants carrying the same construct shows a high variability. However, the expression levels of the tobacco plants containing the wild-type construct are equally distributed over all categories of GUS activity. Due to the introduction of point mutation m1, m2 or m3 this random distribution changes in such a way that the majority of the plants falls in the categories of lower GUS activity (mutants M1, M2 and M3, Figure 5). Together, the data indicate that the residues at −854/−855, −844 and −816, with mutations m1, m2 and m3, respectively, each are part of a cis-acting PR-1a promoter element. Mutation m3 affects GT-1 binding (Figure 4) but mutations m1 and m2 do not. Possibly, the residues at −854/−855 and −844 are involved in binding of proteins other than GT-1. Sequential addition of mutations m2 to m7 to construct M1, resulting in constructs M12 to M17, did not further change the distribution pattern of GUS activity of the M1 plants.

Discussion The 0.9 kb PR-1a upstream region, previously shown to be required for high levels of inducible gene expression, contains a number of elements that are bound by tobacco GT-1-like nuclear factors (van de Rhee et al., 1993; Buchel et al., 1996). While sequences between −902 and −689 have been shown to be critical for the level of PR-1a gene expression, sequences within the region −689 to −643 determine the inducibility of the gene by TMV infection and SA treatment (van de Rhee and Bol, 1993; van de Rhee et al., 1993). Seven potential GT-1 binding sites are located between −902 and −643. The reduced binding of GT-1-like factors to PR-1a promoter fragments A (−902 to −764) and B (−787 to −656) after induction of the gene by TMV infection or SA treatment suggested that GT-1 may function as a repressor of gene expression in uninduced plants. In the present study we have investigated the effects of mutations in the putative GT-1 binding sites on GT-1 binding in vitro and expression of PR-1a promoter/GUS genes in vivo. Introduction of point mutations that affect the critical core sequence of the putative GT-1 binding sites in fragments A, B and B0 -C (−720 to −571) revealed that one of the seven sites is indeed involved in the

formation of a GT-1-DNA complex. The binding site is located at −816 (site III). The importance of the G residue at −816 (site III) for binding GT-1 was confirmed by both DMS interference footprinting (Figure 3A) and by mutational analysis (Figure 4, lane m3+). The methylation interference of the G residue at −816 did not result in the complete absence of the methylated form of this residue within the GT-1-DNA complex. This suggests that more than one binding site for GT-1 is present in PR-1a promoter fragment A. However, the position of additional GT-1 binding sites in this fragment remains unclear. Although sites I, II, IV and V match the core GT-1 binding consensus sequence, mutations in these sites (m1, m2, m4, m5) were not able to abolish the binding of GT-1 (Figure 4 and unpublished results). This suggests that, in spite of mutation of the conserved box II-like site, GT-1 can still bind to this position. Alternatively, it binds to other regions of fragments A and B. In both cases, conservation of the box IIlike sequence seems not to be the sole determinant for GT-1 binding in this region of the promoter. DMS interference footprint results indicate that with the fast-mobility complex formed with fragment A, nuclear proteins interact with sequences located at −860 to −862, −855 to −858 and some single residues at −869, −827 and −818 (Figure 3B). Although the residues at −855 to −858 coincide with the potential GT-1 binding site I, gel shift assays using competitor box II (Buchel et al., 1996) exclude the involvement of these residues in binding GT-1 in vitro. To analyse the effect of mutations m1 to m7 on gene expression, tobacco plants were transformed with constructs containing the 0.9 kb PR-1a promoter fused to the GUS reporter gene. Leaf discs from the transformed plants were assayed for GUS activity after floatation on water or 1 mM SA. SA treatment of leaf discs from plants carrying the wild-type construct showed induced GUS activities varying between 0 and 1.4 pmol MU per hour per µg soluble protein, with a random distribution of the fifteen individual transformants over the thirteen arbitrary categories of GUS activity (Figure 5, wt). We expected that, if a decreased binding of GT1-like proteins to the PR-1a promoter would induce gene expression, mutation m3, which abolishes this interaction, would result in a consitutive expression of PR-1a in vivo. However, this did not appear to be the case (results not shown). Moreover, introduction

395 of point mutation m3 resulted in a decreased level of SA-inducible gene expression (Figure 5, M3). Also mutations that have no effect on GT-1 binding in vitro (m1 and, to a lesser extent, m2) have a negative effect on SA-inducible gene expression (Figure 5, M1 and M2). The reduced SA inducibility was found for all transformants, irrespective of the number of mutated potential GT-1 binding sites. This finding implies that mutations m1 to m7 do not have an additive effect on reporter gene expression. However, since mutations m4 to m7 were not analysed separately, it cannot be excluded that the influence of these mutations is masked by the effect of mutations m1 to m3. Each of the mutations m1, m2 and/or m3 in fragment A results in a new complex with intermediatemobility when assayed in vitro. Since mutations m1, m2 and m3 affect different residues in fragment A, the proteins in the new complex probably interact with DNA in a non-sequence-specific manner. The finding that mutation m1 did not result in the formation of the intermediate-mobility complex suggests that the protein-DNA interaction in the latter complex is restricted to sequences between positions −855 and −816. Our working hypothesis is that removal of GT-1 from the far-upstream low-affinity sites of the PR-1a gene enables positive regulators to bind, which results in gene expression. Although the finding of the new complex supports this hypothesis, we cannot exclude that in the mutant promoter plants additional effects occur. It could be argued that in these plants similar interactions have a negative effect on inducible gene expression. The protein(s) involved in such interactions might be histones or other DNA scaffolding proteins. In the transgenic plants, where GT-1 can no longer efficiently bind to the upstream promoter region, the presumed DNA scaffolding proteins could wrap up this region, which would negatively influence the inducibility of gene expression. In wild-type plants GT-1 secures the region for irreversible wrapping in chromatin structures, while its low-affinity interaction with GT-1 enables it to be available for positive transcription factors after induction. In conclusion, our experiments suggest that the binding of GT-1-like factors to the PR-1a promoter influences the level of SA-inducible gene expression. In addition, protein factors distinct from GT-1 appear to be involved in this process. Further characterization of the proteins that interact with wild-type PR1a upstream sequences and the intermediate complex formed with mutant PR-1a promoter fragment A will

be required to elucidate their role in regulation of PR-1a gene expression.

References Ahl, P., Cornu, A. and Gianinazzi, H.M. 1982. Soluble proteins as genetic markers in studies of resistance and phylogeny in Nicotiana. Phytopathology 72: 80–85. Alexander, D., Goodman, R.M., Gut-Rella, M., Glascock, C., Weymann, K., Friedrich, L., Maddox, D., Ahl-Goy, P., Luntz, T., Ward, E. and Ryals, J. 1993. Increased tolerance to 2 oomycete pathogens in transgenic tobacco expressing pathogenesis-related protein-1a. Proc. Natl. Acad. Sci. USA 90: 7327–7331. Anderson, S.L., Teakle, G.R., Martino-Gatt, S.J. and Kay, S.A. 1994. Circadian-clock- and phytochrome-regulated transcription is conferred by a 78 bp cis-acting domain of the Arabidopsis CAB2 promoter. Plant J. 6: 457–470. Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K., Albright, M., Coen, D.M., Varki, A. and Janssen, K. (Eds.) 1995. Introduction of a point mutation by sequential PCR steps. In: Current Protocols in Molecular Biology, Vol. 1, Suppl. 15, Wiley, New York. Bradford, M.M. 1979. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248–254. Brederode, F.T., Linthorst, H.J.M. and Bol, J.F. 1991. Differential induction of acquired resistance and PR gene expression in a tobacco by virus infection, ethephon treatment, UV light and wounding. Plant Mol. Biol. 17: 1117–1125. Buchel, A.S., Molenkamp, R., Bol, J.F. and Linthorst, H.J.M. 1996. The PR-1a promoter contains a number of elements that bind GT-1-like nuclear factors with different affinity. Plant Mol. Biol. 30: 493–504. Cornelissen, B.J.C., Hooft van Huijsduijnen, R.A.M., van Loon, L.C. and Bol, J.F. 1986. Molecular characterization of messenger RNAs for ‘pathogenesis-related’ proteins 1a, 1b and 1c, induced by TMV infection of tobacco. EMBO J. 5: 37–40. Davis, M.C., Yong, M.-H., Gilmartin, P.M., Goyvaerts, E., Kuhlemeier, C., Sarokin, L. and Chua, N.-H. 1989. Minimal sequence requirements for the regulated expression of rbcS-3A from Pisum sativum in transgenic tobacco plants. Photochem. Photobiol. 52: 43–50. Dehesh, K., Bruce, B. and Quail, P.H. 1990. A trans-acting factor that binds a GT-motif in a phytochrome gene promoter. Science 250: 1397–1399. Eyal, Y., Sagee, O. and Fluhr, R. 1972. Dark-induced accumulation of a basic pathogenesis-related (PR-1) transcript and a light requirement for its induction by ethylene. Plant Mol. Biol. 19: 589–599. Gianinazzi, S., Martic, C. and Vallee, J.C. 1970. Hypersensibilité aux virus, température et protéines solubles chez le Nicotiana Xanthi n.c. Apparition de novelles macromolécules lors de la répression de la synthèse virale. C.R. Acad. Sci. (Ser. D) 270: 2383–2386. Gilmartin, P.M. and Chua, N.-H. 1990. Localizaition of a phytochrome responsive element within the upstream region of pea rbcS-3A. Mol. Cell. Biol. 10: 5565–5568. Gilmartin, P.M., Memelink, J., Hiratsuka, K., Kay, S.A. and Chua, N.-H. 1992. Characterization of a gene encoding a DNA binding protein with specificity for a light-responsive element. Plant Cell 4: 839–849.

396 Green, P.J., Kay, S.A. and Chua, N.-H. 1987. Sequence specific interactions of a pea nuclear factor with light-responsive elements upstream of the rbcS-3A gene. EMBO J. 6: 2543–2549. Green, P.J., Yong, M.-H., Cuozzo, Kano-Murakami, Y., Silverstein, P. and Chua, N.-H. 1988. Binding site requirements for pea nuclear factor GT-1 correlate with sequences required for lightdependent transcriptional activation of the rbcS-3A gene. EMBO J. 7: 4035–4044. Green, P.J., Kay, S.A., Lam, E. and Chua, N.-H. 1989. In vitro DNA footprinting. In: S.B. Gelvin and R.A. Schilperoort (Eds.) Plant Molecular Biology Manual, Kluwer Academic Publishers, Dordrecht, Netherlands, B11, pp. 1–22. Hagiwara, H., Matsuoka, M., Ohshima M., Watanabe, M., Hosokawa, D. and Ohashi, Y. 1993. Sequence-specific binding of protein factors to two independent promoter regions of the acidic tobacco pathogenesis-related-1 protein gene (PR-1). Mol. Gen. Genet. 240: 197–205. Horsch, R.B., Fry, J.E., Hoffmann, N.L., Eichholtz, D., Rogers, S.G. and Fraley, R.T. 1985. A simple and general method for transferring genes into plants. Science 227: 1229–1231. Jefferson, R.A. 1997. Assaying chimeric genes in plants: the GUS gene fusion system. Plant Mol. Biol. Rep. 5: 387–405. Jefferson, R.A., Kavanagh, T.A. and Bevan, M.W. 1987. GUS fusions: β-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J. 6: 3901–3907. Kuhlemeier, C., Green, P.J. and Chua, N.-H. 1987. Regulation of gene expression in higher plants. Annu. Rev. Plant. Physiol. 38: 221–257. Kuhlemeier, C., Strittmatter, G., Ward, K. and Chua, N.-H. 1989. The pea rbcS-3A promoter mediates light responsiveness but not organ specificity. Plant Cell 1: 471–478. Lam, E. 1995. Domain analysis of the plant DNA-binding protein GT-1a: requirement of four putative α-helices for DNA-binding and identification of a novel oligomerization region. Mol. Cell. Biol. 15: 1014–1020. Lam, E. and Chua, N.-H. 1990. GT-1 binding sites confer lightresponsive expression in transgenic tobacco. Science 248: 471– 474. Lawton, M.A., Dean, S.M., Dron, M., Kooter, J.M., Kragh, K.M., Harrison, M.J., Tanguary, L., Dixon, R.A. and Lamb, C.J. 1991. Silencer region of a chalcone synthase promoter contains multiple binding sites for a factor SBF-1, closely related to GT-1. Plant Mol. Biol. 16: 235–249. Marsch, J.L., Erfle, M. and Wykes, E.J. 1984. The pIC plasmid and phage vectors with versatile cloning sites for recombinant selection by insertional inactivation. Gene 32: 481–485. Maxam, A.M. and Gilbert, W. 1977. A new method for sequencing DNA. Proc. Natl. Acad. Sci. USA 74: 560–564. Memelink, J., Linthorst, H.J.M., Schilperoort, R.A. and Hoge, J.H.C. 1990. Tobacco genes encoding acidic and basic isoforms of pathogenesis-related proteins display different expression patterns. Plant Mol. Biol. 14: 119–126. Niderman, T., Genetet, I., Bruyère, T., Gees, R., Stinzi, A., Legrand, M., Fritig, B. and Mösinger, E. 1995. Pathogenesis-related PR-1

proteins are antifungal: Isolation and characterization of three 14 kilodalton proteins of tomato and of a basic PR-1 of tobacco with inhibitory activity against Phytophthora infestans. Plant Physiol. 108: 17–27. Ouwerkerk, P.B.F. 1997. Cis-acting elements and trans-acting factors involved in regulation of terpenoid indole alkaloid genes from Catharanthus roseus. Ph.D. thesis, University of Leiden, Netherlands. Payne, G., Parks, T.D., Burkhart, W., Dincher, S., Ahl, P., Métraux, J.P. and Ryals, J. 1988. Isolation of the genomic clone for pathogenesis-related protein 1a from Nicotiana tabacum cv. Xanthi nc. Plant Mol. Biol. 11: 89–94. Pfitzner, U.M. and Goodman, H.M. 1987. Isolation and characterization of cDNA clones encoding pathogenesis-related proteins from tobacco mosaic virus infected tobacco plants. Nucl. Acids Res. 15: 4449–4465. Pierpoint, W.S. 1997. The natural history of salicylic acid. Interdisciplin. Sci. Rev. 22(1): 45–52. Ross, A.F. 1986. Systemic acquired resistance induced by localized virus infection in plants. Plant Mol. Biol. 9: 389–410. Strompen, G., Grüner, R. and Pfitzner, U.M. 1988. An as-1like motif controls the level of expression of the gene for the pathogenesis-related protein 1a from tobacco. Plant Mol. Biol. 37: 871–883. 0 van de Rhee, M.D. and Bol, J.F. 1989. Induction of the tobacco PR1a gene by virus infection and salicylate treatment involves an interaction between regulatory elements. Plant J. 3: 71–82. van de Rhee, M.D., van Kan, J.A.L., González-Jaén, M.T. and Bol, J.F.: Analysis of regulatory elements involved in the induction of two tobacco genes by salicylate treatment and virus infection. Plant Cell 2: 357–366. van Loon, L.C. and Antoniw, J.F. 1988. Comparison of the effects of salicylic acid and ethephon treatment with virus-induced hypersensitivity and acquired resistance in tobacco. Neth. J. Plant Path. 88: 237–256. van Loon, L.C., Pierpoint, W.S., Boller, Th. and Conejero, V. 1994. Recommendations for naming plant pathogenesis-related proteins. Plant Mol. Biol. Rep. 12: 245–264. Villain, P., Clabault, G., Mache, R. and Zhou, D.X. 1994. SIF binding site is related but different from the light-responsive GT-1 binding site and differentially represses the spinach rps1 promoter in transgenic tobacco. J. Biol. Chem. 269: 16626–16630. Villain, P., Mache, R. and Zhou, D.X. 1996. The mechanism of GT-element-mediated cell type specific transcriptional control. J. Biol. Chem. 51: 32593–32598. Ward, E.R., Uknes, S.J., Williams, S.C., Dincher, S.S., Wiederhold, D.L., Alexander, D.C., Ahl-Goy, P., Métraux, J.-P. and Ryals, J.A. 1991. Coordinate gene activity in response to agents that induce systemic acquired resistance. Plant Cell 3: 1085–1094. Yang, Y. and Klessig, D.F. 1996. Isolation and characterization of a tobacco mosaic virus-inducible myb oncogene homolog from tobacco. Proc. Natl. Acad. Sci. USA 93: 14972–14977.