Plant Molecular Biology43: 429–438, 2000.© 2000Kluwer Academic Publishers. Printed in the Netherlands.

429

cDNA cloning and characterization of a plant protein that may beassociated with the harpinPSS-mediated hypersensitive response

Cheng-hsien Chen1,2, Hao-jan Lin1, Mang-jye Ger2, David Chow1 and Teng-yung Feng1,2,∗1Institute of Botany, Academia Sinica, Taipei 115, Taiwan (∗author for correspondence; e-mail: bofeng@cc vax.sinica.edu.tw);2Institute of Life Science, National Defense Medical Center, Taipei, Taiwan

Received 27 July 1999; accepted in revised form 8 March 2000

Key words:defense-related protein, harpinPSS, hypersensitive response, hypersensitive response-assisting protein,sweet pepper

Abstract

Hypersensitive response-assisting protein (HRAP) is a novel plant protein that can intensify the harpinPSS-mediatedhypersensitive response (HR) in harpinPSS-insensitive plants, such as the vegetative stage of sweet pepper. In thisreport, we identified a HRAP cDNA clone from sweet pepper (Capsicum annuumcv. ECW). The sequence ofthis cDNA clone showed no appreciable similarity to any other known sequences. However, it contained threepositively charged regions, a typical signal peptide and a cAMP-dependent phosphorylation site. ThehrapmRNAaccumulated preferentially during the incompatible interaction of sweet pepper leaves with a pathogenic bacterium,Pseudomonas syringaepv. syringae. When thehrap gene transcription level was high, the sweet pepper leavesreadily expressed the harpinPSS-mediated HR. Thehrap gene transcription level in sweet pepper was also higherduring the reproductive stage than during the vegetative stage. The HRAP distribution in an individual plant anddifferent plant species was investigated. We found that all the organs of sweet pepper, except fruit, could expresstwo different forms of HRAP. Moreover, thehrap gene was presented in many plant species including tobacco,Arabidopsis, and rice. In conclusion, our results suggest that thehrap gene is widely distributed throughout theplant world and its transcription level correlates with plant sensitivity to harpinPSS. The interaction between HRAPand harpinPSSreveals a novel way to interpret the interaction mechanism between plants and bacterial pathogens.

Introduction

The hypersensitive response (HR), characterized bythe rapid, localized death of plant cells at the siteof pathogen invasion (Klementet al., 1964; Tunerand Novacky, 1974; Goodman and Novacky, 1994),is an important defense response to prevent furthermultiplication and restrict the spread of the pathogenin plant tissue. The ability of many Gram-negativeplant pathogens, such asPseudomonas, Xanthomonas,and Erwinia, to elicit a HR or cause a disease inplants is controlled byhrp genes (Bonas, 1994).Many hrp genes encode components of a type IIIprotein secretion system conserved in both plant and

The nucleotide sequence data reported will appear in theEMBL and GenBank Nucleotide Sequence Databases under theaccession number AF168415 (hrap cDNA).

animal pathogens (Galán and Collmer, 1999). Oneprotein known to be secreted by the Hrp system isharpinPSS, encoded byhrpZ in the hrp gene clus-ter from Pseudomonas syringaepv. syringae (Heet al., 1993). HarpinPSS is capable of eliciting a HRwhen infiltrated into the leaf intercellular spaces oftobacco and several other plants (Heet al., 1993).Two pieces of evidence to support that harpinPSScan interact with molecules on the plant cell surfacesand induce HR through membrane signal transductionpathways were previously reported. First, harpinPSS-mediated HR could be prevented or delayed by in-hibitors of membrane signal transduction in tobacco(Heet al., 1993). Second, Hoyoset al.(1996) reportedthat fluorochrome-tagged harpinPSSantibodies boundthe outer portion of the plant cell surface under theharpinPSS-HR induction of tobacco suspension cells.

430

It is reasonable to suggest that the plant molecules thatinteracted with harpin elicitors were involved in harpinsensing by the plant cells.

In previous studies, we documented that aplant amphipathic protein (hypersensitive response-assisting protein, HRAP) from sweet pepper mayinteract with harpinPSS. HRAP was able to cause thedissociation of harpinPSSmultimeric forms in neutralbuffer systems (Chenet al., 1998). This dissocia-tion resulted in inducing HR during the harpinPSS-insensitive stage of sweet pepper. In this paper, thefull-length cDNA of HRAP was cloned and character-ized by the rapid amplification of 5′ and 3′ cDNA ends(RACE) on the basis of polymerase chain reaction(PCR). The function of the recombinant HRAP pro-tein encoded by the cDNA clone was consistent withthe biochemically purified native HRAP. Based onamino acid sequence analysis, a putative signal pep-tide was located at the N-terminus of HRAP and maylead HRAP into extracellular matrixes. An increasedhrap transcript level was observed when sweet pepperleaves were infiltrated with the native harpinPSSgen-erator,Pseudomonas syringaepv. syringae. Addition-ally, we found that the extent of harpinPSS-mediatedHR expressed in sweet pepper during different devel-opmental stages was not the same. In northern blotanalysis, this phenomenon was shown to correlatewith the transcriptional level of thehrap gene. Ac-cording to these results, we suggest that HRAP hasthe potential to interact with harpinPSSin vivoand mayinfluence the plant’s sensitivity to harpinPSS.

Materials and methods

Plants and harpinPSSpreparation

Sweet pepper (Capsicum annuumcv. ECW) seedswere germinated for 14 days and then planted in plots.Sweet pepper plants were grown in a growth chamberat 25◦C, 60% RH, and a photoperiod of 12 h (100µEm−2 s−1). Plants flowered 10 weeks after germination.In our experiments, we used 8-week old plants as thevegetative stage materials, and 11-week old plants asthe reproductive stage materials. The positions of up-per leaves were located above the first fork, and lowerleaves were below. The lower leaves (ca. 6 weeks oldat vegetative stage and 8 weeks old at reproductivestage) were older than the upper ones (3 weeks oldat each stage) in this experiment.

The harpinPSS clone was provided by Dr H.-C. Huang at the Agricultural Biotechnology Labo-

ratories, National Chung-Hsien University, Taiwan.HarpinPSS protein was extracted by the methods de-scribed by Heet al. (1993).Escherichia coliDH5α(pSYH10) which harbored the harpinPSSgene (hrpZ)was grown in Luria Broth containing ampicillin(50µg/ml) at 37◦C in the dark with shaking overnightin the presence of isopropylthio-β-D-galactoside. Toobtain harpinPSS, the bacteria were washed and soni-cated for 30 s with 0.01 M phosphate buffer pH 6.5 andthen boiled for 10 min. After boiling, the extracts werecentrifuged at 10 000× g for 10 min. Supernatantswere desalted by a Microconcentrator (Amicon) andstored at 4◦C.

Plant hypersensitive response assay

The assay was performed according to Huanget al.(1988). Sweet pepper leaves ca. 4–7 weeks old andfully expanded tobacco leaves were wounded by a 25gauge needle to form tiny holes on the lower surfaceof the leaves. HarpinPSSwith and without HRAP treat-ments were prepared in 50 mM Tris buffer pH 7.5 andinfiltrated by pressing a 1 ml blunt syringe through thehole. The infiltrated plant was incubated in a 25◦C,12 h light/12 h dark incubator.

Infiltration of sweet pepper withPseudomonassyringaepv.syringae

TheP. syringaepv. syringaestrain used in this studywas provided by Dr H.-C. Huang. This strain inducesthe development of an HR on sweet pepper leaves(during any plant developmental stage) within 18–24 h of infiltration. This strain was grown at 30◦Cin Nutrient Broth medium, and then diluted in waterfor infiltration. Sweet pepper leaves 4–7 weeks oldwere infiltrated with theP. syringaesuspension (107

c.f.u./ml) or water. The infiltration procedures were asdescribed for the HR assay.

Total RNA and genomic DNA isolation

Total RNA was isolated from sweet pepper leaveswith a Qiagen Plant RNA kit (Qiagen) and quantitatedby spectrophotometry, assumingA260 = 40 µg/ml(Sambrooket al., 1989). Genomic DNA was extractedfrom sweet pepper, tobacco (Nicotiana tabacum), rice(Oryza sativa), andArabidopsis columbialeaves ac-cording to the procedure provided by Qiagen. In brief,2 to 5 g ofyoung plant leaves were ground in liquidnitrogen, and lysis buffer (20 mM EDTA, 0.5 mg/ml

431

cellulase, 1% Triton X-100, 500 mM guanidine-HCl, 200 mM NaCl, 10 mM Tris-HCl, pH 7.9) wasadded. Supernatant of lysates was transferred to anion-exchange column (Qiagen, Genomic Tip) to iso-late DNA. The eluted DNA was then precipitated andwashed by a standard protocol, and then quantitatedby spectrophotometry, assumingA260 = 50 µg/ml(Sambrooket al., 1989).

Gene cloning

The internal sequencing of HRAP was performedfor gene cloning work because HRAP was blockedat the amino terminus. We obtained a short aminoacid sequence, KIPLQFTLVGFW. Based on 1 to 8amino acids in this sequence, a degenerate primer (5′-AA(A/G)AT(A/C/T)CCI(C/T)TICA(A/G)TT-3′) wasdesigned for the HRAP cloning work.

3′ and 5′ RACE cloning methods with total RNA asthe template were used to clone the HRAP gene fromsweet pepper. First, the 3′-end fragment was ampli-fied with a poly-T primer and the degenerate primerdescribed above. Synthesis of the cDNA was carriedout as recommended by the manufacturer (Tital RT-PCR System, Boehringer Mannheim). The PCR wasconducted for 5 cycles with a regime of 94◦C for30 s, 40◦C for 30 s, 68◦C for 1 min and then 30cycles with a regime of 94◦C for 30 s, 50◦C for 30 s,68 ◦C for 1 min. The PCR products were sequencedand an 80 bp clone (H-C-1) was obtained. The second-round RT-PCR was carried out with total RNA, apoly-T primer and a specific sense oligonucleotide(5′-GATTTTGGGGAACAGATTCCAG-3′) designedfrom the sequence of H-C-1. The sequence of theoligonucleotide corresponded to the nucleotide posi-tions 199 to 220 in the full-length cDNA (Figure 1).The PCR mixture was heated at 94◦C for 2 min andthen subjected to 35 cycles of PCR with a regime of94 ◦C for 30 s, 60◦C for 30 s, 68◦C for 1 min.

The RACE approach was employed to clone the5′ end of the HRAP cDNA clone. For use as primer,two oligonucleotides corresponding to the 3′-end re-gion of the H-C-1 cDNA clone (oligomer 1, 5′-GGCAACCTTGTTGAATTCCT-3′; oligomer 2, 5′-CTGGAATCTGTTCCCCAAAATC-3′) were synthe-sized in the antisense direction. The sequences of theoligonucleotides corresponded to nucleotide positions199–220 and 222–243 in the full-length cDNA clonerespectively (Figure 1). The first-strand cDNA wasprepared using total RNA with oligomer 1. Synthe-sis of the cDNA was carried out as recommended by

Figure 1. Nucleotide and deduced amino acid sequence of the sweetpepper HRAP cDNA clone. The internal amino acid sequence ob-tained for gene cloning work is in bold type, the predicted signalsequence is lightly underlined, and the cAMP-dependent phospho-rylation site is heavily-underlined. The amino acid charge index ofHRAP is shown in the bottom table (analyzed by GCG software).The positively charged regions are marked with black bars.

the manufacturer (5′-RACE kit, Clontech). The first-round PCR was conducted with oligomer 2 and anAbridge Anchor Primer supplied by the manufacturer(Clontech). The first-round PCR was conducted for35 cycles with a regime of 94◦C for 1 min, 55◦Cfor 1 min, 72 ◦C for 2 min. The first-round PCRproduct was diluted and used as a PCR template forsecond-round PCR. Oligomer 2 and the AUAP primersupplied by the manufacturer (Clontech) were used inthe second-round PCR. The PCR mixture was heatedat 94◦C for 2 min and then subjected to 35 cycles ofPCR with a regime of 94◦C for 1 min, 55◦C for 1 min,72 ◦C for 2 min.

After cloning and sequencing both end frag-ments, the HRAP gene coding region was ampli-fied using total RNA and specific primers for ex-pression work. The primers were designed from se-

432

quences located at 27–46 (sense oligomer 5′-CG-CGGATCCATGAAAATGAAGAACCTCTC-3′, witha degenerateBamHI restriction sequence underlined)and 849–878 (antisense oligomer 5′-TACTTAGTAA-ACTTTCTTATAAAGCTTCAC-3′, with a nativeHindIII restriction sequence underlined) from the full-length HRAP cDNA.

The expression and purification of sweet pepperHRAP inEscherichia coli

The expression and purification of HRAP inE. coliwas carried out with a QIAexpressionist kit (Qia-gen). The PCR fragments of the HRAP cDNA codingregion were digested withBamHI and HindIII for24 h and ligated into pQE-30 vectors supplied bythe manufacturer (Qiagen). TheE. coli strain M15was transformed with the pQE-30 plasmids containingHRAP cDNA. Recombinant HRAP was extracted bythe native method for use in HR assays or by the de-naturing method for use in antiserum preparation. Therecombinant HRAP was purified with Ni-NTA agaroseas recommended by the manufacturer (Qiagen). Inaddition, we transformed M15 directly with pQE-30plasmids supplied by the manufacturer (Qiagen), andextracted and purified proteins from this transformantby the same procedures as per recombinant HRAP.This protein solution was treated as a blank control inthe HR assay of recombinant HRAP.

DNA sequencing and analysis

The sequences of HRAP cDNA clones were de-termined by an automated sequencer (model 381A,ABI Applied Biosystems, Foster City, CA) using thedideoxy chain termination method according to theprocedure provided by the manufacturer (ABI). Wecompared the HRAP cDNA clone against the Gen-Bank and the National Center for Biotechnology In-formation non-redundant protein databases with theBLAST search program (Altschulet al., 1990).

Antiserum

Adult New Zealand white female rabbits were usedfor anti-HRAP antiserum preparations. RecombinantHRAP was obtained in denaturing buffer (8 M urea,0.1 M sodium hydroxyphosphate, 0.01 M Tris-HCl,pH 5.9) with the QIAexpress system (Qiagen) andconcentrated to 7.5µg/µl, because a high concentra-tion of HRAP caused precipitation in non-denaturingsolution. We mixed 20µl of recombinant HRAP

solution with 200µl of TiterMax Gold (CytRx Cor-poration), 90µl of 10 mM phosphate buffer and 90µlrabbit pre-immune serum. This mixture was the anti-gen solution for a subcutaneous injection. No boostershot for this antibody preparation was executed andthe rabbit was bled 50 days after the injection.

Southern blot analysis

The sweet pepperhrap gene clone was labeled withdigoxigenin-11-dUTP (Boehringer Mannheim) usingPCR as recommended by the manufacturer. GenomicDNA (15 µg) of examined plants were digested withrestriction enzymes, and the samples were Southern-blotted by standard procedures (Sambrooket al.,1989). Membranes were hybridized at 55◦C overnightwith the hrap probe, washed at room temperature in2× SSC/0.1% w/v SDS twice for 5 min each, and at68 ◦C in 0.5× SSC/0.1% w/v SDS twice for 15 mineach, and then detected with the DIG luminescentdetection kit (Boehringer Mannheim).

Northern blot analysis

Total RNA (15µg) was electrophoresed through 1%agarose/formaldehyde gels and then transferred ontonylon membranes. Membranes were hybridized at55 ◦C overnight with thehrap probe, washed at highstringency (60◦C) in 2× SSC/0.1% w/v SDS twicefor 5 min each, and in 0.5× SSC/0.1% w/v SDS twicefor 15 min each, and then detected with the DIGluminescent detection kit (Boehringer Mannheim).

Western blot analysis

Proteins were extracted by homogenizing 5 g of tissuein 5 ml Tris buffer (50 mM Tris pH 7.2) using a plasticpestle fitted to a 15 ml centrifuge tube. The sam-ples were precipitated with 60% ammonia sulfate anddialyzed in 50 mM Tris buffer. The protein concen-tration of samples were detected by using Coomassiebrilliant blue dye with a microassay method as rec-ommended by the manufacturer (BioRad). A 10µgportion of each protein sample was electrophoresedthrough gels containing 12.5% polyacrylamide plusSDS (SDS-PAGE), then either stained with Coomassieblue or electro-transferred onto mitrocellulose mem-branes with the BioRad blue tank method. HRAPproteins were detected on western blots using anti-HRAP antibodies followed by mouse anti-rabbit IgG-peroxidase conjugate.

433

Figure 2. The protein pattern of HRAP expressed byE. colion SDS-PAGE. Recombinant HRAP andE. coli protein extractswere extracted by a denaturing method and purified with Ni-NTAagarose. Recombinant HRAP (1µg) was loaded onto a 12.5% poly-acrylamide gel plus SDS (SDS-PAGE) and silver-staine. m, proteinmarkers; H, HRAP expressed byE. coli; B, E. coli protein extracts.

Results

Cloning and characterization of the HRAP gene

We used 3′- and 5′-RACE cloning methods with to-tal RNA as templates to clone the HRAP gene fromsweet pepper. Using a poly-T primer and a degenerateprimer designed from an internal amino acid sequenceof the HRAP protein, a 1.1 kb 3′-end fragment of theHRAP cDNA was cloned from 3′-RACE. 5′-RACEwas then used to clone the 5′-end fragment with spe-cific primers resulting in a 210 bp cDNA product. Afull-length cDNA was cloned by flanking PCR withtwo specific primers designed from both the 3′- and5′-end fragments.

The full-length cDNA of sweet pepper HRAP is996 bp, and encodes a 29.9 kDa protein (Figure 1).This encoded protein has an isoelectric point of 8.39,which is consistent with the biochemically purifiednative HRAP (Chenet al., 1998). The sequence com-parison of the HRAP gene showed no similarity withany reported cDNA sequences. The HRAP containsa cAMP-dependent phosphorylation site, a 21 aminoacid NH2-terminal hydrophobic region, and three pos-itively charged regions (Figure 1). Although cAMP-dependent phosphorylation sites are often associatedwith cAMP-dependent regulation, there is no other ev-idence to support that cAMP is related to the functionof HRAP. The hydrophobic region is a putative signalpeptide and its speculative cleavage site is located atthe 22nd amino acid (von Heijine, 1990). This findingsuggests that HRAP can be secreted into extracellularmatrices because signal peptides can direct proteinsinto secretory pathways.

Figure 3. The HR promoting activity of HRAP expressed byE. coli.Each treatment was inoculated into sweet pepper leaves at the plantvegetative stage, and HR areas were observed in 24 h. 1, 15 ngrecombinant HRAP; 2, 10µg harpinPSSmixed with 15 ng recombi-nant HRAP; 3, 10µg harpinPSSmixed withE. coli protein extracts(1.5 ng in total) (see Materials and methods); 4, 10µg harpinPSSonly. HR necrosis was photographed five days after inoculation.

The activity of HRAP expressed byEscherichia coli

We expressed the HRAP gene inE. coli in order to seewhether the protein encoded from the HRAP cDNAclone matched the native HRAP protein from sweetpepper. The expressed HRAP protein purified by His-tag affinity chromatography was used for biochemicaland biological analysis. The protein solution extractedfrom E. coli without HRAP cDNA plasmid was alsopurified by His-tag affinity chromatography as a blankcontrol. The molecular weight (MW) of the HRAPprotein expressed byE. coli was the same as the es-timated MW derived from the cDNA clone (Figure 2).In biological analysis, HRAP purified from sweet pep-per leaves could assist harpinPSSto induce HR duringthe vegetative stage of sweet pepper leaves, whereits activity highly depends on the ratio of HRAP toharpin (Chenet al., 1998). Similarly, the HRAP ex-pressed byE. coli was also able to assist harpinPSStoinduce HR during the vegetative stage of sweet pepperleaves (Figure 3). A 10µg portion of harpinPSSmixedwith an optimal quantity (ca. 15 ng) of recombinantHRAP could cause a most significant HR in sweetpepper leaves (Figure 4). These results suggest thatthe obtained cDNA clone can encode the sweet pepperHRAP protein.

Southern blot analysis

Arabidopsis, rice, sweet pepper, and tobacco genomicDNA was digested with restriction enzymesBamHIandHindIII, respectively, to identify the distributionand copy number of thehrap gene in plant genomes.A probe synthesized from the cDNA coding region

434

Figure 4. The titration effect of HRAP expressed byE. coli in thepromotion of HR. A 10µg portion of harpinPSSmixed recombinantHRAP was infiltrated into sweet pepper leaves at the plant vege-tative stage. 1, 1 ng HRAP; 2, 15 ng HRAP; 3, 30 ng HRAP; 4,harpinPSS only. The degree of HR measured as the percentage ofthe HR area over the infiltration area is given in the table to theright. HR necrosis was photographed four days after inoculation.

of HRAP was used. The result showed that at leastsix bands could be detected in theBamHI-digestedlane of sweet pepper genomic DNA and five bandsin the HindIII-digested DNA (Figure 5). Since thereis no BamHI or HindIII cutting site located withinthe HRAP cDNA coding region in sweet pepper, thepredicted copy number for the HRAP gene shouldbe a minimum of five. The band pattern detected intobacco was similar to the one found in sweet pep-per. In addition, low-molecular-weight bands (2.4 kb,4.2 kb, and 4.9 kb bands in theBamHI digestion and2.4 kb, 4.2 kb, and 8.5 kb bands in theHindIII di-gestion) detected in sweet pepper and tobacco, werealso detected inArabidopsisand rice genomic DNA(Figure 5). These results suggest that thehrap genemay exist in different plant species in a very conservedform.

Development-specific expression of sweet pepperHRAP

The temporal patterns ofhrap gene expression havebeen determined in sweet pepper with northern blotanalysis. We extracted total RNA from upper- andlower-position leaves of sweet pepper at the vegetativeand reproductive stages and hybridized the total RNA

Figure 5. Southern blot analysis of plant genomic DNA using thesweet pepper HRAP cDNA as a probe.Arabidopsis, rice, sweetpepper, and tobacco genomic DNA (10µg per lane) were respec-tively digested byBamHI (lane B) orHindIII (lane H) before beingsubjected to Southern blot analysis.

with a HRAP cDNA probe. As shown in Figure 6,a 1.1 kb band was detected in northern blot analy-sis using RNA from both upper and lower leaves atvegetative and reproductive stages. The HRAP tran-scription level in the upper leaves at the reproductivestage was higher than at the vegetative stage. However,lower leaves at both the vegetative and reproductivestages transcribed very few HRAP mRNAs. This re-sult may be due to the low transcriptional efficiencyof old leaves. Nevertheless, our results suggest thatregulation of HRAP is development-specific with sig-nificant expression associated with young leaves at thereproductive stage.

The potency of harpinPSSto induce a HR in sweetpepper at different developmental stages was alsotested. A distinct HR was shown only on the upper leafat the reproductive stage (Figure 7). Minor HR thatshowed small HR necrosis areas was found on vegeta-tive stage leaves and lower leaves at the reproductivestage. These results suggest that sweet pepper leavesare capable of more potently expressing the harpinPSS-mediated HR when the HRAP transcription level ishigh.

435

Figure 6. Northern blot analysis at different developmental stagesof sweet pepper plant. Total RNA (15µg per lane) of upper leaves(lane U), lower leaves (lane L) at the vegetative stage (lanes markedV) and reproductive stage (lanes marked R) was isolated and probedwith HRAP cDNA clone. The ethidium bromide-stained bands of28S rRNA were used to verify the loading amount of total RNA(bottom row).

Figure 7. The harpinPSS-mediated HR in sweet pepper leaves atdifferent developmental stages. A 10µg portion of harpinPSSwasinoculated into sweet pepper leaves and HR areas were observed.1, the upper leaf at reproductive stage; 2, the lower leaf at repro-ductive stage; 3, the upper leaf at vegetative stage; 4, the lower leafat vegetative stage. HR necrosis was photographed four days afterinoculation.

Figure 8. hraptranscript accumulation in sweet pepper leaves in-filtrated withP. syringaepv. syringaesuspension. Two leaves froma single vegetative stage sweet pepper plant were infiltrated withP. syringaepv. syringaesuspension. About 0.3 g of leaf materialwas harvested from the infiltration areas 0, 1, 4, 8, or 16 h afterinfiltration. HR symptoms became visible 20 h after infiltration onthe test plant. Total RNA (15µg per lane) was hybridized with aHRAP cDNA probe. The ethidium bromide-stained bands of 28SrRNA were used to verify the loading amount of total RNA (bottomrow).

Figure 9. Western blot analysis with anti-HRAP antibody of differ-ent sweet pepper organs. The plant materials used in the analysiswere harvested from the reproductive stage of sweet pepper. Totalproteins from each organ (5µg) were purified (except 10µg infruits) and western blot analysis was performed as described. Theanti-HRAP antibody can detect a minimum of 0.1 ng HRAP.

hrapgene expression during the incompatibleinteraction of sweet pepper withP. syringaepv.syringae

To investigate the relationship betweenhrap geneexpression and infection by the native harpinPSSgen-erator, P. syringaepv. syringae, the hrap transcriptlevel in infiltrated sweet pepper leaves was detected(Figure 8). HR symptoms became visible 20 h afterinfiltration on the test plant. As shown in Figure 8, thehrap gene expression level was very low at 0 and 1 hafter inoculation, and increased progressively duringthe first 4 h. It reached a maximum level at ca. 4 h, andthen decreased to a low level within 16 h. This resultindicates thathrap gene expression is induced earlyduring the incompatible interaction of sweet pepperwith P. syringaepv. syringae.

Distribution of HRAP proteins in various organs ofsweet pepper

The western blot analysis of HRAP expressed in var-ious organs of sweet pepper showed that HRAP wasdetected in all except fruit (Figure 9). Interestingly,two different molecular masses of HRAP were de-tected: one at 30 kDa and another at 29 kDa. Theamino acid sequences of these two proteins are prob-ably very similar because anti-HRAP polyclonal an-tibodies could detect them both. The multiple HRAPgenes observed in the Southern blot analysis suggestthat these two proteins may belong to the same gene

436

family (Figure 4). The 30 kDa peptide was mainly de-tected in stems and roots, but not in leaves and flowers.In contrast, the 29 kDa peptide was mainly detected inleaves, flowers, and roots, but not in stems. This resultsuggests that two isoforms of HRAP are expressed inan organ-specific manner.

Discussion

The sequence of the cDNA clone encoding HRAPfrom sweet pepper showed no appreciable similarityto any other known sequences. This result reveals thatHRAP is a novel plant protein able to intensify theharpinPSS-mediated HR in sweet pepper. Interestingly,our data show that the higher the transcription levelof hrap, the easier it is to express harpinPSS-mediatedHR in sweet pepper leaves. This proportional relation-ship suggests that the intensifying effect of HRAP onharpinPSS-mediated HR observed in our experimentsshould existin vivo and correlate with the plant’s sen-sitivity to harpinPSS. In addition, Southern blot analy-sis indicates thathrap exists in both dicotyledons andmonocotyledons. The mechanism of plant-pathogeninteraction, in which HRAP is involved, may be alsoprevalent within vegetable world.

The mechanism of intensifying the harpinPSS-mediated HR is hypothesized such that HRAP candissociate multiple forms of harpinPSS into the moreactive monomer or dimer form in HR induction (Chenet al., 1988). The mechanism of the HRAP-harpinPSSinteraction is still unknown. We have found a numberof positively charged regions within the HRAP aminoacid sequence. Positively charged residues oftenplay an important role in protein-protein interactions,such as in the binding of protein kinase C-substrate(Chapline et al., 1993), thrombomodulin-thrombin(Liu et al., 1994), nicotinic acetylcholine receptor-curaremimetic toxins (Fulachieret al., 1994) andbutyrylcholinesterase-succinyldithiocholine (Massonet al., 1997). Therefore, these positively charged re-gions may provide the ability for the HRAP proteinto bind to negatively charged molecules. This charac-teristic of HRAP may be involved in intensifying theharpinPSS-mediated HR. Additionally, we find at leasttwo HRAP isoforms expressed in an organ-specificmanner in sweet pepper. There may be a few func-tional differences between these isoforms to fulfilldifferent demands in different organs. Further genecloning work using the different HRAP isoforms willhelp us to further understand the HRAP activity.

hrap gene expression is induced by an incompat-ible pathogenic bacterium infiltration. A number ofplant defense genes have been shown to be activatedduring plant-pathogen interactions, especially duringincompatible interactions (Kombrink and Somssich,1995). Products of these genes include a broad spec-trum of antimicrobial proteins, such as pathogenesis-related (PR) proteins (Fritiget al., 1998), andmetabolites deriving from secondary metabolic path-ways, such as the sesquiterpenoid (Stoesslet al.,1976) and phenylpropanoid (Nicholson and Hammer-schmidt, 1992) pathways. These proteins are thoughtto play a role in resistance during the late stage ofHR (Hammond-Kosack and Johnes, 1996). The ex-pression of these defense-related genes often increased5–8 h after pathogen inoculation and maintained at ahigh level after more than 16 h. However, the tran-script level of inducedhrap increased early and turnedover within 16 h, differing from other defense-relatedgenes. This confirms that the function of HRAP is nec-essary during the initial stage of HR. Moreover, HRAPis known to intensify the harpinPSS-mediated HR inthe intercellular spaces of plant leaves. Therefore,HRAP may have a correlation to pathogen recog-nition or HR triggering during the early stage ofincompatible plant-pathogen interaction.

Presently, the relationship between plant HR anddevelopment has not been elucidated. Some reportssuggest that the resistance of a number of plants topathogens correlate with their development. In to-bacco plants, the levels of salicylic acid (SA) arehigh during the flowering stage (Henselet al., 1993).The pathogen-related proteins (PR proteins) are alsohighly expressed in tobacco flowers (Lotanet al.,1989) and the leaves of flowering tobacco (Fraser,1981; Nealeet al., 1990; Hensel et al., 1993). Asthese molecules accumulate in the foliage of matureplants, there is an increase in resistance to infectionby viral (Fraser, 1972; Henselet al., 1993) and fungal(Reuveniet al., 1986; Wyattet al., 1991) pathogens.Our data show that harpinPSS-mediated HR is ex-pressed more strongly during the reproductive stageof sweet pepper than during the vegetative stage. Wealso find that HRAP is highly expressed in the foliageof flowering sweet pepper and may cause floweringsweet pepper to become sensitive to harpinPSS. Thisphenomenon, in comparison with SA and PR proteins,shows a similar developmental regulatory characteris-tic, leading to increased resistance during flowering. Itis possible that the HRAP, SA and PR proteins share acommon regulatory mechanism.

437

Hoyoset al. (1996) reported that the existence ofthe plant cell wall is essential for harpinPSS-mediatedHR. This result suggests that harpinPSScould interactwith certain plant molecules or be recognized by itsreceptor in the extracellular matrices of plant cells.Furthermore, Dangl (1994) proposed that the elicitorpresenting systems in plant cells are able to modify ex-ternal elicitors and present their active regions to plantcell surface receptors. For example, it was reportedthat the protein elicitor (AVR9) fromCladosporiumfulvum is processed by plant proteinases in HR in-duction (Van den Ackervekenet al., 1993). Therefore,the elicitor presenting system for harpin proteins mayexist in extracellular matrices for harpin proteins tobe delivered to the surface of plant cells (Galán andCollmer, 1999). Whereas HRAP contains a typicalsignal peptide to lead the mature HRAP protein intoextracellular matrices (von Heijne, 1990), HRAP alsohas a chance to interact with harpinPSSand achieve theintensifying effect on harpinPSS-mediated HR. Thus,the dissociation of harpinPSS, promoted by HRAP, isperhaps an elicitor presenting mechanism located onplant cell surfaces. Further HRAP biochemical studiesshould help to reveal the details of the harpin present-ing system in plant cells and the role of harpin proteinsin bacterial pathogen invasion.

Acknowledgements

We would like to thank Dr S.-C. Huang for providingthe harpinPSS clone. Thanks are also due to Dr J.-F.Wang for providing sweet pepper seeds and plant ma-terials. This work was supported by grants to T.-Y.F.from Academia Sinica, Taiwan.

References

Altschul, S.F., Gish, W., Miller, W., Myers, E.W. and Lipman, D.J.1990. Basic local alignment search tool. J. Mol. Biol. 215: 403–410.

Bonas, U. 1994.Hrp genes of phytopathogenic bacteria. In: J.L.Dangl (Ed.) Current Topics in Microbiology and Immunology,Vol 192: Bacterial Pathogenesis of Plant and Animals: Molecularand Cellular Mechanisms, Springer-Verlag, Berlin, pp. 79–98.

Chapline, C., Ramsay, K., Klauck, T. and Jaken, S. 1993. Interactioncloning of protein kinase C substrates. J. Biol. Chem. 268: 6858–6861.

Chen, C.H., Lin, H.J. and Feng, T.Y. 1998. An amphipathic proteinfrom sweet pepper can dissociate harpinPSS multimeric formsand intensify the harpinPSS-mediated hypersensitive response.Physiol. Mol. Plant. Path. 52: 139–149.

Dangl, J.L. 1994. The enigmatic avirulence genes of phytopatho-genic bacteria. In: J.L. Dangl (Ed.) Current Topics in Mi-crobiology and Immunology, Vol 192: Bacterial Pathogenesisof Plants and Animals: Molecular and Cellular Mechanisms,Springer-Verlag, Berlin, pp. 99–118.

Fraser, R.S.S. 1972. Effects of two strains of tobacco mosaic viruson growth and RNA content of tobacco leaves. Virology 47: 261–269.

Fraser, R.S.S. 1981. Evidence for the occurrence of the‘pathogenesis-related’ proteins in the leaves of healthy tobaccoplants during flowering. Physiol. Plant. Path. 19: 69–76.

Fritig, B., Heitz, T. and Legrand, M. 1998. Antimicrobial proteinsin induced plant defense. Curr. Opin. Immunol. 10: 16–22.

Fulachier, M.H., Mourier, G., Cotton, J., Servent, D. and Menez,A. 1994. Interaction of protein ligands with receptor fragments.On the residues of curaremimetic toxins that recognize frag-ments 128–142 and 185–199 of theα-subunit of the nicotinicacetylcholine receptor. FEBS Lett. 338: 331–338.

Galán, J.E. and Collmer, A. 1999. Type III secretion machines: bac-terial devices for protein delivery into host cells. Science 284:1322–1328.

Godiard, L., Grant, M.R., Dietrich, R.A., Kiedrowski, S. and Dangl,J.L. 1994. Perception and response in plant disease resistance.Curr. Opin. Genet. Dev. 4: 662–671.

Goodman, R.N. and Novacky, A. 1994. The hypersensitive defensereaction in plants to pathogens: a resistance phenomenon. Amer-ican Phytopathological Society Press, St Paul, MN, pp. 7–8.

Hammond-Kosack K.E. and Jones, J.D.G. 1996. Resistance gene-dependent plant defense response. Plant Cell 8: 1773–1791.

He, S.Y., Huang, H.C. and Collmer, A. 1993.Pseudomonas sy-ringae pv. syringaeharpinPSS: a protein that is secreted via theHrp pathway and elicits the hypersensitive response in plants.Cell 73: 1255–1266.

Hensel, L.L., Grbic, V., Baumgarten, D.A. and Bleecker, A.B.1993. Developmental and age-related processes that influencethe longevity and senescence of photosynthetic tissues in ara-bidopsis. Plant Cell 5: 553–564.

Hoyos, M.E., Stanley, C.M., He, S.Y., Pike, S., Pu, X.-A. andNovacky, A. 1996. The interaction of harpinPSS with plant cellwalls. Mol. Plant-Microbe Interact. 9: 608–616.

Huang, H.-C., Schuurink, R., Denny, T.P., Atkinson, M.M., Baker,C., Yucel, I., Hutcheson, S.W. and Collmer, A. 1988. Molecularcloning of aPseudomonas syringaepv. syringaegene clusterthat enablesPseudomonas fluorescensto elicit the hypersensitiveresponse in tobacco plants. J. Bact. 170: 4748–4756.

Klement, Z., Farkas, G.L. and Lovrekovich, L. 1964. Hypersensitivereaction induced by phytopathogenic bacteria in the tobacco leaf.Phytopathology 54: 474–477.

Kombrink, E. and Somssich, I.E. 1995. Defense responses of plantsto pathogens. Adv. Bot. Res. 21: 1–33.

Liu, L.W., Rezaie, A.R., Carson, C.W., Esmon, N.L. and Esmon,C.T. 1994. Occupancy of anion binding exosite 2 on thrombindetermines Ca2+ dependence of protein C activation. J. Biol.Chem. 269: 11807–11812.

Lotan, T., Ori, N. and Fluhr, R. 1989. Pathogenesis-related proteinsare developmentally regulated in tobacco flowers. Plant Cell 1:881–887.

Masson, P., Legrand, P., Bartels, C.F., Formant, M.T., Schopfer,L.M. and Lockridge, O. 1997. Role of aspartate 70 and trypto-phan 82 in binding of succinyldithiocholine to human butyryl-cholinesterase. Biochemistry 36: 2266–2277.

Neale, A.D., Wahleithner, J.A., Lund, M., Bonnett, H.T., Kelly,A., Meeks-Wagner, D.R., Peacock, W.J. and Dennis, E.S. 1990.Chitinase,β-1,3-glucanase, osmotin, and extensin are expressed

438

in tobacco explants during flower formation. Plant Cell 2: 673–684.

Nicholson, R.L. and Hammerschmidt, R. 1992. Phenolic com-pounds and their role in disease resistance. Annu. Rev Phytopath.30: 369–389.

Reuveni, M., Tuzun, S., Cole, J.S., Siegel, M.R. and Kúc, J. 1986.The effects of plant age and leaf position on the susceptibil-ity of tobacco to blue mold caused byPeronospora tobacina.Phytopathology 76: 455–458.

Sambrook, J., Fritsch, E.F. and Maniatis, T. 1989. MolecularCloning: A Laboratory Manual, 2nd ed., Cold Spring HarborLaboratory Press, Plainview, NY, pp. 9.31–9.57, E6.

Stoessl, A., Stothers, J.B. and Ward, E.W.B. 1976. Sesquiterpenoidstress compounds of the Solanaceae. Phytochemistry 15: 855–872.

Tuner, J.G. and Novacky, A. 1974. The quantitative relation betweenplant and bacterial cells involved in the hypersensitive reaction.Phytopathology 64: 885–890.

Van den Ackerveken, G.F.J.M., Vossen, P. and de Wit, P.J.G.M.1993. Theavr9 race-specific elicitor ofCladosporium fulvumisprocessed by endogenous and plant proteinases. Plant Physiol.103: 91–96.

Van Gijsegem, F., Genin, S. and Boucher, C. 1993. Evolution-ary conservation of pathogenicity determinants among plant andanimal pathogenic bacteria. Trends Microbiol. 1: 175–180.

von Heijine, G. 1990. The signal peptide. J. Membrane Biol. 115:195–201.

Wyatt, S.E., Pan, S.Q. and Kuc, J. 1991.β-1,3-Glucanase, chiti-nase, and peroxidase activities in tobacco tissues resistant andsusceptible to blue mould as related to flowering, age and suckerdevelopment. Physiol. Mol. Plant Path. 39: 433–440.