Abstract. The cell walls of two near-isogenic lines of

bean (Phaseolus vulgaris L.) seedlings, susceptible or

resistant to the bean anthracnose pathogen Colletotri-

chum lindemuthianum, were digested with the pure

endopolygalacturonase (endoPG; EC 3.2.1.15) isolated

from the fungus. The solubilized pectic fragments were

separated according to their charge and size. Analysis of

their uronic acid contents showed that their elution

patterns were quite dissimilar, depending on whether

they originated from the resistant or the susceptible host

plant. Their sugar compositions revealed that neutral

sugars were more abundant in the fragments released

from the resistant plant than from the susceptible one,

while the reverse was true for acidic residues. The

fragments solubilized from the resistant plant induced an

increase of pathogenesis-related (PR) proteins when

challenged on resistant or susceptible bean seedlings,

both at the transcript and enzyme-activity levels. On the

other hand, pectic fragments released from susceptible

bean cell walls exhibited either no signi®cant activity or

only a weak elicitor e€ect on the defence of susceptible

or resistant bean seedlings. The di€erential elicitor e€ect

observed between pectic fragments was inversely corre-

lated to their acidity. Thus, endoPG-released pectic

fragments from bean cell walls exhibited the same ability

as the endoPG itself (C. La®tte et al., 1993, Mol Plant-

Microb Interact 6: 628±634) to elicit defence responses in

a cultivar-speci®c manner.

Key words: Defense reactions ± Elicitor ± Endopoly-

galacturonase ± Pectic fragments ± Phaseolus

Introduction

The concept of endogenous elicitors has evolved from

the initial reports that microbial pectinases induce

defence responses in plants (Lee and West 1981; Davis

et al. 1984). In-vitro bioassays have con®rmed that

pectic oligomers function as biological signals that

stimulate a wide array of defence responses, ranging

from early events at the cell surface to the subsequent

biosynthesis of defence molecules such as phytoalexins,

pathogenesis-related (PR) proteins, proteinase inhibi-

tors, lignin (CoÃte and Hahn 1994), and hydroxyproline-

rich glycoproteins (Boudart et al. 1995). In addition, it

has been shown that these oligomers also regulate

development and growth (Eberhard et al. 1989; MarfaÁ

et al. 1991). Although active fragments are generally

oligogalacturonides with a degree of polymerization

(DP) of galacturonic acid (GalA) of 9±15, more-complex

galacturonides can also act as elicitors (Brecht and

Huber 1988). Recently, the presence of active fragments

was detected in vivo in extracts of ripening tomato

pericarp (Melotto et al. 1994), thus substantiating the

biological signi®cance of endogenous oligosaccharide

signalling.

Previously, we have shown that the pure endopoly-

galacturonase (endoPG) of Colletotrichum lindemuthia-

num race b, a fungal pathogen of bean plants, induces

defence responses in a speci®c manner, i.e. more rapidly

in resistant (R) than in susceptible (S) cultivars to this

race (La®tte et al. 1993). Induction is very ecient since

picomole amounts of endoPG per seedling are sucient

to induce defence at the whole-plant level. This e€ect

mimics the response of these cultivars to infection by the

same race of the fungus. Cellular and molecular studies

have revealed that endoPG gene expression occurs in

planta during the parasitic life of Colletotrichum (Ben-

hamou et al. 1991; Centis et al. 1997). The di€erential

e€ect of endoPG might result either from the protein

itself, or from the nature and perception of the

fragments it releases from the cell walls. Indeed, a

polygalacturonase inhibitory protein (PGIP) has been

characterized in resistant seedlings (La®tte et al. 1984;

Abbreviations: DP = degree of polymerisation; endoPG ˆ en-

dopolygalacturonase; galA ˆ galacturonic acid; HPAEC ˆ high-

performance anion-exchange chromatography; PGiP ˆ polyga-

lacturonase-inhibitory

protein;

PR ˆ pathogenesis

related;

R ˆ resistant; S ˆ susceptible
Correspondence to: G. Boudart; E-mail: [e-mail];

Fax: [phone]

Planta (1998) 206: 86±94

Di€erential elicitation of defense responses by pectic fragments

in bean seedlings

Georges Boudart, Claude La®tte, Jean Paul Barthe, Denis Frasez, Marie-TheÂreÁse EsquerreÂ-TugayeÂ

Centre de Biologie et Physiologie VeÂgeÂtales, UMR 5546 CNRS-UPS, Universite Paul Sabatier, 118 route de Narbonne,

F-31062 Toulouse Cedex, France

Received: 18 July 1997 / Accepted: 23 January 1998

Cervone et al. 1987). The protein PGIP contains leucine-

rich repeats (LRR) closely related to LRR domains in

resistance gene products (Jones et al. 1994; Staskawicz

et al. 1995). Thus, the association of endoPG with PGIP

might resemble the interaction between resistance and

avirulence gene products which leads to the rapid

activation of defence in resistant plants. Alternatively,

or additionally, lowering of the endoPG activity by

PGIP might increase the half-life and pattern of active

fragments.

The aim of the present work was to generate and

characterize the pectic fragments obtained through the

action of endoPG on cell walls of susceptible or resistant

beans. Prior to hydrolysis by endoPG the cell walls were

free from PGIP, in order to prevent interference with the

enzyme. The fragments that were recovered were then

assayed for their e€ect on defence markers, notably PR-

protein gene expression and ethylene.

Materials and methods

Plant material. Two near-isogenic lines of Phaseolus vulgaris L. cv.

Processor (I.N.R.A, Versailles, France) resistant or susceptible to

race b of Colletotrichum lindemuthianum, were grown in a growth

chamber as described by La®tte et al. (1984).

Endopolygalacturonase. The enzyme was puri®ed to homogeneity

from the culture ®ltrate of Colletotrichum lindemuthianum race b

according to a previously described procedure (Barthe et al. 1981).

Purity of the enzyme was assessed by SDS-PAGE and silver

staining of the gel. The amount of endoPG corresponding to 1 nkat

enzyme activity was 0.4 lg of the pure protein. One nanokatal

corresponds to the release of 1 nmol GalA equivalent á s

)1

from

0.1% polygalacturonic acid (Sigma, Saint Quentin Fallavier,

France) in 0.2 M acetate bu€er (pH 5.2) at 30 °C for 30 min.

Bean pectic fragments. The cell walls of the two isolines were

prepared at 0±4 °C according to the method of Nevins et al. (1967).

It was previously checked that preparing the cell walls by successive

washings with 0.5 M K-phosphate bu€er (pH 6.8) and methanol-

chloroform eliminated the endogenous PGIP activity of the cell

wall. Solubilization of cell wall pectic fragments by the pure

endoPG of Colletotrichum lindemuthianum and their fractionation

by chromatography according to their charge and size were done

following a previously described procedure (Boudart et al. 1995).

Brie¯y, the protocol included two steps: (i) separation of the

solubilized material by anion-exchange chromatography on QAE-

Sephadex (Pharmacia, Orsay, France) and stepwise elution with

125, 300, 600 mM and 1 M imidazole-HCl at pH 6.0; and (ii) gel

permeation chromatography of the active QAE fractions on

Sephacryl S200 HR (Pharmacia). The recovered fractions were

dialyzed against ultrapure water with Spectra/Por CE (Poly Labo,

Strasbourg, France; cut-o€ MW 1000) and used in bioassays.

Oligogalacturonides. A mixture of oligogalacturonides of DP 10 to

15 was prepared from autoclaved polygalacturonic acid according

to the protocol described by Mathieu et al. (1991). The released

oligogalacturonides were made with 100 mM imidazole-HCl (pH

6.0) and separated by anion-exchange chromatography on a QAE

Sephadex column (15 cm long, 2 cm i.d.) equilibrated in 100 mM

imidazole-HCl (pH 6.0). Oligogalacturonides were stepwise eluted

in four fractions with 400 mM imidazole-HCl (pH 6.0; 2 ´ 50 ml)

and 600 mM imidazole-HCl (pH 6.0; 2 ´ 50 ml). Analysis of their

galacturonic acid content by high-performance anion-exchange

chromatography (HPAEC) with the Dionex system (Dionex,

Sunnyvale, Calif., USA) and pulsed amperometric detection

(Hotchkiss and Hicks 1990) showed that they corresponded to

oligogalacturonides with DPs 2, 3±12, 10±21 and 18±31, respec-

tively. The fraction enriched in oligogalacturonides of DP 10±21

was further fractionated by HPAEC on a semi-preparative column

(CarboPac PA1; Dionex; 250 mm long, 9 mm i.d.) using the

Dionex system equipped with a precolumn (CarboPac 10±32).

Fractionation was achieved at pH 5.0 with a linear sodium acetate

gradient (220±820 mM) at room temperature for 60 min at a ¯ow

rate of 1 ml á min

)1

. Oligogalacturonides of DP 10±15 were

identi®ed according to the previous determination of their retention

time. They were collected in a single fraction which was desalted by

repeated precipitation with ethanol. The precipitate was ®nally

dissolved in distilled water and the remaining ethanol was

evaporated under vacuum. Finally, oligogalacturonides were taken

up into water, analyzed for their GalA content (Blumenkrantz and

Asboe-Hansen 1973) and bioassayed.

Bioassays. Bean pectic fragments and oligogalacturonides were

bioassayed on bean cuttings, using a previously described protocol

(La®tte et al. 1993). Brie¯y, each cutting (about 0.8 g) was allowed

to absorb pectic material (250 lg GalA equivalent) diluted in

0.3 ml of distilled water through the cut section of the petiole. This

amount was retained on the basis of published work (Darvill et al.

1992) showing that pectic fragments display optimum activity at

concentrations most often ranging over 10±50 lg GalA per ml cell-

suspension culture, i.e. per 80 mg average fresh mass cells.

Absorption of the samples was completed after 4±6 h at 22 °C in

a growth chamber under light. The cuttings were then supplied with

water plus streptomycin (0.1 mg á ml

)1

) until harvest. Controls

received only water plus streptomycin under the same conditions.

Elicited and control cuttings were harvested at 48 h and frozen in

liquid nitrogen. They were then ground at 4 °C in 1 M NaCl,

50 mM Na-acetate bu€er (pH 5.2; 8 ml á g

)1

fresh weight) in the

presence of Polyclar AT. The homogenates were spun, and

supernatants were dialyzed overnight against 50 mM Na-acetate

bu€er (pH 5.2). The dialyzed extracts were adjusted to 10 ml per

gram fresh weight with 50 mM Na-acetate bu€er (pH 5.2) and used

as enzyme sources for the measurement of b-1,3-glucanase and

chitinase activities.

b-1,3-Glucanase and chitinase assays. Activity of b-1,3-glucanase

was determined by measuring the production of reducing sugars

from laminarin (Sigma). The standard assay (1 ml) contained the

enzyme extract (100 ll), 2.5 mg laminarin in 50 mM Na-acetate

bu€er at pH 5.2 (500 ll) and 400 ll of the same bu€er. The

reaction mixture was incubated for 30 min at 50 °C. Total reducing

sugars were assayed by the colorimetric method of Somogyi (1952)

and expressed as glucose equivalents.

Chitinase activity was estimated by the release of reducing

sugars from chitin (Sigma) which was previously prepared accord-

ing to the method of Rupley (1964). The standard assay (2 ml)

contained the enzyme extract (150 ll), 10 mg chitin in 1 ml of

50 mM Na-acetate bu€er (pH 5.2) and 850 ll of the same bu€er.

After incubation for 4 h at 37 °C under continuous stirring, 50 ll

of 5 N KOH was added to the reaction mixture. The medium was

centrifuged and the supernatant estimated for its content of

reducing sugars, expressed as N-acetyl glucosamine equivalents,

by the Somogyi (1952) procedure. Activities of b-1,3-glucanase and

chitinase in elicited seedlings were calculated as nmole sugars

released in 1 s per g fresh weight, and expressed over the

corresponding activities estimated in control plants.

Ethylene. For the measurement of ethylene, each cutting was

enclosed in a 125-ml Wheaton ¯ask ®tted with a small vial

containing either a solution of pectic fragments plus 0.1%

streptomycin, or water plus 0.1% streptomycin in the case of

controls. The ¯asks were capped with rubber plugs, sealed with

Para®lm immediately after introducing the cuttings into the vials,

and transferred to a growth chamber at 22 °C under light for a

period of 10 h which corresponded to the time necessary for

complete absorption of the solutions under these conditions. Air

G. Boudart et al.: Di€erential elicitation of defense responses by pectic fragments in bean seedlings

87

samples (1 ml) were withdrawn from the ¯asks and their ethylene

content was determined on a gas chromatograph equipped with an

alumina column and a ¯ame ionization detector, according to

Rickauer et al. (1989).

Sugar analysis. The glycosyl-residue composition of the fractions

recovered from Sephacryl S200 chromatography was determined

by Dionex HPAEC-PAD (Dionex) after hydrolysis of the samples

in 2 N tri¯uoroacetic acid (TFA) at 120 °C for 1 h. Hydrolysis was

performed in 0.1 N TFA at 100 °C when 2-keto-3-deoxy-

D

-manno-

octulosonic acid and apiose were to be determined. Sugars were

quanti®ed by reference to standards; the values were corrected

according to the stability of each sugar during TFA hydrolysis.

Analysis of GalA C6-esteri®cation. The procedure described by

Maness et al. (1990) was used to determine the degree of

esteri®cation of the pectic fragments recovered from Sephacryl

S200 chromatography. The protocol is based on the reduction of

C6-esteri®ed GalA residues and their conversion to galactose.

Reduction was achieved by treating the fragments (50 lg GalA

equivalents) with 5 M NaBH

4

in 1 M imidazole-HCl (pH 7.0)

(200 ll), for 1 h in an ice bath. The reduced pectic fragments were

exhaustively dialyzed against distilled water with Spectra/Por CE

(cut-o€ MW 1000) for 96 h at 4 °C, lyophylized and hydrolyzed in

2 N TFA at 120 °C for 1 h. Galactose residues which account for

both native galactose and reduced C6-esteri®ed GalA were quan-

ti®ed by Dionex HPAEC-PAD. The degree of esteri®cation was

determined from the increase in galactose observed after reduction

of the fragments over the amount of galactose present in native

fragments. In the case of fragments containing only oligogalactur-

onides of low DP, the dialysis step was omitted; Na

+

and imidazole

cations were removed by ion-exchange chromatography on AG

50 W X16, H

+

form resin, and the remaining material was

evaporated to dryness. Borate was eliminated by successive

washings with methanol. The desalted residue was then submitted

to TFA hydrolysis as described above.

Analysis of acetylation. The fragments recovered from Sephacryl

S200 chromatography (500 lg GalA equivalents) were saponi®ed

with 1 N NaOH (500 ll) at pH 12 for 1 h at room temperature and

the released acetate was quanti®ed using the commercially available

test-combination for the determination of acetic acid (Boehringer

Mannheim, Germany), according to the instructions of the

manufacturer. The degree of acetylation of the pectic fragments

was expressed in mole percent total sugar residues.

Isolation of RNA and gel blot hybridization. Total RNA was

isolated from control and elicited bean cuttings using the ready-for-

use Extract-All (Eurobio, Les Ulis France) kit; RNA concentra-

tions were measured photometrically at 260 nm. The RNA samples

(7 lg) were denatured at 50 °C for 1 h with glyoxal/dimethyl

sulfoxide and subjected to electrophoresis in a 1.2% agarose gel in

10 mM Na

2

-phosphate bu€er, pH 7.0. The RNA gels were blotted

overnight onto a nylon membrane (Hybond N+; Amersham, UK),

®xed at 80 °C for 2 h and analyzed by northern blot hybridization.

Equal loading of the gels was checked by staining the membrane

with methylene blue according to Sambrook et al. (1989). The

membranes were prehybridized at 42 °C for 2 h in ``High SDS

Bu€er'' (Boehringer Mannheim, Germany) supplemented with

0.005% yeast RNA (Boehringer Mannheim), then hybridized

overnight at 42 °C in the same solution containing the radioactive

probe. The basic chitinase and basic b-1,3-glucanase probes were

obtained by polymerase chain reaction (PCR) ampli®cation of bean

genomic

DNA

using

the

primers

5¢-GCAGTGTAG-

GAGTGGTGTGGATGC-3¢ and 5¢-TAGCAGTCAAGGTTGTT

GCCATAACC-3¢ for chitinase (Broglie et al. 1989); 5¢-GGTGT

GTTATGGCATGATGGGC-3¢ and 5¢-GGGTACTTCTTCTCT

TTGCTGGG-3¢ for glucanase (Edington et al. 1991). Fragments

were ligated in pGEM-T (Promega, Madison, Wis., USA) and

ampli®ed in the Escherichia coli strain XL1 blue according to

standard procedures (Sambrook et al. 1989). Nucleotide sequencing

of the inserts (Sanger et al. 1977) con®rmed that they corresponded

to the above-mentioned chitinase and b-1,3-glucanase clones, as

well as to the b-1,3-glucanase isoform which was previously

isolated from infected bean seedlings and partly sequenced

(Daugrois et al. 1992). Radioactive labelling of the probes was

performed with [a

32

P]dCTP by random priming as described by

Feinberg and Vogelstein (1983). After hybridization, the mem-

branes

were

then

successively

washed

in

2 ´ SSC

(1 ´ SSC ˆ 0.15 M NaCl, 0.015 M Na

3

-citrate, pH 7), 0.1%

SDS for 10 min at room temperature, 2 ´ SSC, 0.1% SDS at

50 °C for 30 min and exposed to an autoradiography-®lm (NEN

Life Science Products, Le Blanc Mesnil., France) at )80 °C for 2 d.

Results

Recovery of pectic fragments. Digestion of bean cell walls

by the pure endoPG yielded pectic material containing

38 ‹ 11 mg GalA equivalent per g cell wall of the

resistant isoline, and 37 ‹ 10 mg GalA equivalent per g

cell wall of the susceptible isoline. The endoPG-solubi-

lized material was then fractionated by anion-exchange

chromatography on QAE Sephadex, and mainly recov-

ered in the two fractions which eluted at 300 mM and

600 mM of the gradient. On the basis of its uronic acid

content (Fig. 1), the material recovered in the 300 mM-

QAE fraction from the resistant isoline represented

about 70% of the eluted pectic material; the 600 mM-

QAE fraction contained most of the remaining eluted

pectic compounds. The material solubilized from the

susceptible isoline was distributed in nearly equal

amounts in the 300 mM- and 600 mM-QAE fractions,

averaging 80% of the total eluted material; the remain-

der was mainly recovered in the 1M-QAE fraction.

The endoPG-solubilized material eluted in the

300 mM- and 600 mM-QAE fractions was submitted

Fig. 1. Proportions of galacturonic acid-containing fragments in the

fractions recovered from the cell walls (4 g) of resistant (Pr.R) and

susceptible (Pr.S ) bean seedlings after hydrolysis with the endoPG

(200 nkat) of Colletotrichum lindemuthianum. The hydrolysate was

applied on a QAE Sephadex column (8 cm long, 2 cm i.d.) and

separation was achieved by stepwise elution with 125, 300, 600 mM

and 1 M imidazole bu€er, pH 6.0. The fractions were assayed for their

uronic acid content and amounts are expressed as percent of total

GalA eluted from the column

88

G. Boudart et al.: Di€erential elicitation of defense responses by pectic fragments in bean seedlings

to gel-permeation chromatography on a Sephacryl S200

HR column. The column eluate was recovered in four to

seven GalA-containing fractions (A±G), corresponding

to fragments of decreasing size (Fig. 2). For conve-

nience, the Sephacryl S200 HR fractions derived from

the 300 and 600 mM QAE fractions of resistant and

susceptible host cell walls will be referred to as R300 and

R600, S300 and S600, respectively. From preparation to

preparation, the recorded elution pro®les only presented

minor variations. Figure 2 shows that the R300 and

R600 QAE fractions were composed of a mixture of

large-, intermediate- and low-size fragments, with a

relative abundance of the latter. Previous analysis had

shown that the low-size fraction R300-G only contained

di-, and tri-GalA (Boudart et al. 1995). The elution

pattern of the fragments recovered from the susceptible

plant was di€erent. Thus, on the basis of the same

amount of GalA-containing material, the S300 fraction

was composed of higher amounts of large-size fragments

(0.1 < kav < 0.3) and smaller amounts of low-size

oligomers (0.8 < kav < 1). The S600 fraction, instead,

was characterized by a main peak of low-size fragments

(Fig. 2). Each fraction was exhaustively dialyzed against

distilled water before being analyzed for its sugar

content and biological activity.

Sugar composition. Of the eleven sugar monomers that

were recorded in the hydrolysates of the pectic fractions,

only GalA, galactose (Gal), arabinose (Ara), and

rhamnose (Rha) showed appreciable variations (Ta-

ble 1). Within any given set (R or S) of fragments, GalA

was always the most abundant residue, and its propor-

tion increased throughout the fractions (A to D, to E, or

G); the fractions of low size were only composed of low-

DP oligogalacturonides. The level of the major neutral

sugars, Gal and Ara, was higher in the large-size

fractions than in the intermediate-size fractions.

Comparing the di€erent sets of fractions showed that

the proportions of Gal, Ara, and Rha were higher, and

the proportion of GalA was generally lower, in the R300

and 600 fractions than in the S300 and 600 fractions. In

addition, whatever their origin (R or S), the 300 mM

fractions contained more Gal and Ara than the 600 mM

fractions, while the reverse was observed for Rha.

Altogether, these data indicated that the R-fragments

were enriched in the branched rhamnogalacturonan

domains of pectin, as compared to the S-fragments

which were characterized by a higher proportion of low-

size acidic fragments. The other sugar residues, com-

prising fucose, xylose, glucose, 2-O-methyl-xylose, glu-

curonic

acid,

apiose,

and

3-deoxy-

D

-manno-2-

octulosonic acid were found in very low amounts and

did not vary appreciably within and between the various

sets of fractions.

Further analysis showed that a substantial propor-

tion of GalA was esteri®ed in most fractions and that

they all contained acetyl groups. The ratio of GalA to

C6-esteri®ed GalA indicated that the S600 fractions

were less esteri®ed on their carboxyl groups than R600

and that both were also less esteri®ed than the corre-

sponding R300 and S300 fractions. Overall, the data

show that the pectins of the two isolines of P. vulgaris,

resistant or susceptible to Colletotrichum lindemuthia-

num, di€er from each other. The greater proportion of

carboxyl-esteri®ed GalA in the resistant cultivar ac-

counts, at least in part, for this di€erence. According to

the mode of action of endoPG which only cleaves

unesteri®ed homogalacturonans, it follows that frag-

ments of di€erent size, hence of di€erent charge, are

generated through the action of the enzyme. The higher

proportion of the two neutral sugars Gal and Ara in the

R-fragments is consistent with an enrichment in

branched-rhamnogalacturonan structures resulting from

a di€erent pattern of cleavage by endoPG.

Biological activity of R- and S-pectic fragments. The

fragments were ®rst bioassayed on their respective

isolines. Each bean cutting (0.8 g) was allowed to absorb

the fragments (250 lg GalA equivalent). Measurements

of PR activity were made after 48 h for the purpose of

comparison with the previously reported di€erential

e€ect of endoPG. Ethylene, a hormonal signal that often

precedes and induces PR proteins was determined after

10 h. As shown on Fig. 3A (a), the R300 pectic fractions

elicited hydrolytic activities in resistant bean seedlings.

Depending on the fraction, b-1,3-glucanase was en-

hanced by 1 to 3 times and chitinase by 2 to 4 times over

Fig. 2. Gel-permeation on a Sephacryl S200 HR column (45 cm long,

2.2 cm i.d.) of the 300 mM and the 600 mM fractions recovered from

QAE Sephadex chromatography of the endoPG-released cell wall

material. The QAE fractions originating from the resistant and

susceptible isolines are referred to as R300, R600, and S300, S600,

respectively. The column was loaded with 16 mg GalA equivalents of

pectic fragments. Two-milliliter fractions were collected and assayed

colorimetrically for their uronic acid content. Depending on the

elution pro®le, the column eluates were divided into several fractions

(A to D, to E, or G), which were then exhaustively dialyzed against

distilled water and bioassayed. Calibration of the column was

achieved with 680 kDa Leuconostoc dextran (Vo determination,

kav ˆ 0), apple pectin (AP), and galacturonic acid (GA, kav ˆ 1)

G. Boudart et al.: Di€erential elicitation of defense responses by pectic fragments in bean seedlings

89

the respective controls, the fractions of intermediate size

being the most active. Ethylene production was also

enhanced, but the array of active fractions did not

exactly match the pattern of fractions that elicit the two

PR proteins. By comparison, the R600 fractions

[Fig. 3A(b)] exhibited a much weaker elicitor e€ect on

b-1,3-glucanase, chitinase and ethylene than the R300

fractions. However, the array of active fractions was

similar in terms of size and charge.

In the susceptible isoline supplied with either the S300

or S600 pectic fractions, b-1,3-glucanase and chitinase

were not, or only weakly, enhanced, i.e less than 50%

over the control [Fig. 3A (c) and (d)]. Ethylene was not

produced in response to any fraction.

In order to interpret the striking di€erences observed

in elicitor activity between the fragments recovered from

the resistant and the susceptible cultivar cell walls,

experiments were carried out in which the susceptible

isoline was challenged with the R300 and R600 pectic

fractions and the resistant isoline with the S300 and S600

pectic fractions. It was found that supplying susceptible

plants with the R300 fractions [Fig. 3B(a)] elicited

defence responses with a pattern similar to that observed

in resistant plants challenged with the same fractions

[Fig 3A(a)]. A twofold increase in b-1,3-glucanase

activity was measured in response to most R300 pectic

fractions. Similarly, chitinase activity was enhanced by 3

to 4 times. Ethylene was also increased, although to a

lesser extent. The R600 fractions were much less active

[Fig. 3B(b)], mimicking their e€ects on the resistant

plant. In contrast, supplying resistant plants with the

S300 and S600 fractions did not change much the

responses recorded on susceptible plants [Fig. 3B, (c)

and (d)]. b-1,3-Glucanase and chitinase activities were

increased by 40% to 60% over the control. Ethylene

was released at a level representing twice that of the

control.

For comparison, the activity of a mixture of DP 10±

15 oligogalacturonides was assessed on resistant and

susceptible bean seedlings. Figure 4 shows that no

signi®cant elicitor e€ect on any defense marker could

be measured in either resistant or susceptible bean, in

response to these oligogalacturonides, whatever the

amounts (50 lg or 250 lg) that were used.

b-1,3-glucanase and chitinase gene expression. In order to

further compare the activity of R300 and S300 frag-

ments, a kinetic study of their e€ect on basic b-1,3-

Table 1. Sugar composition and degree of esteri®cation and acetylation of the R300, R600, S300, S600 pectic fragments separated by gel

permeation chromatography on Sephacryl S200 HR. The fragments were hydrolyzed with tri¯uoroacetic acid and their sugar composition

was determined by Dionex HPAEC-PAD analysis. The degree of esteri®cation and of acetylation were calculated as described in Materials

and methods

Fractions

Sugars (mole %)

eGalA

b

GalA/eGalA

c

Ac

d

GalA

Galactose

Arabinose

Rhamnose

Other sugars

a

R300

A

61.9

19.6

9.3

3.9

5

16

3.9

12

B

70.8

13.5

6.2

3.4

5.4

32

2.2

10

C

77.4

8.5

4.2

3.5

6.2

37

2.1

8

D

81.2

5.8

3.1

3.1

7

37

2.2

7

E

82.9

4.5

3.1

3.3

6.1

38

2.2

5

F

87.6

2.7

2.5

2.8

4.3

39

2.2

5

G

99.1

nd

e

nd

nd

nd

6

16.5

3

R600

A

58.7

17.3

8.6

6.2

9.1

16

3.7

6

B

74.5

5.3

3.3

6.4

9.9

29

2.6

7

C

69.5

6.2

6.2

8.3

9.8

28

2.5

6

D

66

6

7.7

8

12.4

20

3.3

5

E

99.2

nd

nd

nd

nd

5

19.9

7

S300

A

78.7

10.5

4.2

2

5

32

2.5

7

B

73.7

8.1

4.3

2.9

10.8

23

3.2

7

C

84

2.1

1.7

1.5

7.1

38

2.2

6

D

86.4

1.9

1.6

1.7

5.3

33

2.6

4

E

97.8

nd

nd

nd

nd

3

32.6

2

S600

A

73.8

5.8

4.2

6.3

9.7

26

2.8

7

B

80.6

2.6

3.3

4.9

8.3

25

3.2

6

C

85.2

1.7

2.1

2.9

7.3

30

2.8

6

D

98.7

nd

nd

nd

nd

0.4

246

8

a

Other sugars: fucose, glucose, xylose, apiose, glucuronic acid, 2-O-methyl-xylose and 3-deoxy-

D

-manno-2-octulosonic acid were present

b

% C6-esteri®ed GalA residues were determined as described in Materials and methods

c

Ratio of GalA/C6 esteri®ed GalA residues

d

Acetic acid content is expressed as mole percent total sugar residues

e

nd, not detected

90

G. Boudart et al.: Di€erential elicitation of defense responses by pectic fragments in bean seedlings

glucanase and chitinase gene expression was undertaken

in the resistant isoline. The fragments retained for these

experiments were composed of a mixture of the R300-

D,E,F, and of the S300-B,C, fractions, respectively.

Northern blot analysis indicated that a single transcript

of approximately 0.9 kb was increased for each gene in

elicited seedlings, more quickly and with a greater

intensity in response to the R300 than to the S300

fractions (Fig. 5). Expression of neither gene was

constitutive, as revealed by the absence of hybridization

signal at time zero of the experiment. Transcripts weakly

accumulated in controls, possibly as a result of wound-

ing. Chitinase gene expression was stimulated earlier

than b-1,3-glucanase, i.e 6 h after the beginning of the

elicitor treatment, while b-1,3-glucanase transcripts were

not detected before 12 h. Accumulation of transcripts

reached a maximum at 24±36 h, and decreased after-

wards.

Fig. 3A,B. Elicitor e€ect of the

various R- and S-fractions re-

covered by gel permeation

chromatography (Fig. 2) on PR

hydrolases and ethylene of their

respective resistant (A, a and b;

Pr.R) or susceptible (A, c and d;

Pr.S) isoline. Experiments

where S cuttings were treated

with the various R-fractions (B,

a and b) and R cuttings with the

various S-fractions (B, c and d)

were carried out simultaneous-

ly. Activity is expressed as per-

cent variation over the control;

weak suppressor e€ects (nega-

tive values, black symbols) were

recorded for some S-fractions

(A, c and d). Each cutting was

supplied with 250 lg GalA

equivalents of each fraction and

harvested after 48 h for b-1,3-

glucanase and chitinase mea-

surements. Ethylene was re-

corded 10 h after start of

treatment. Data represent the

means ‹ SD for six indepen-

dent experiments

G. Boudart et al.: Di€erential elicitation of defense responses by pectic fragments in bean seedlings

91

Discussion

The race-cultivar-speci®c interaction between Phaseolus

vulgaris and Colletotrichum lindemuthianum is charac-

terized by early induction of defence responses in

resistant cultivars of the host plant upon inoculation

with avirulent races of the pathogen (Bell et al. 1986;

Daugrois et al. 1990). In a previous work, we reported

that the endoPG of C. lindemuthianum race b elicits b-

1,3-glucanase in two cultivars of Phaseolus vulgaris with

the same speci®city as the fungus, i.e. more rapidly in

resistant (R) than in susceptible (S) isolines to this race

(La®tte et al. 1993). This led us to hypothesize that

discrimination of resistant and susceptible seedlings by

endoPG might result either from a di€erential control of

the enzyme activity by its inhibitor (PGIP) in planta, or

from the chemical nature and perception of the pectic

fragments released from cell walls of resistant and

susceptible plants. In order to look for the intrinsic

elicitor activity of pectic fragments, we prepared PGIP-

free cell walls and submitted them to endoPG hydrolysis.

The present study shows that the fragments released

from the cell walls of resistant and susceptible bean

seedlings to race b of the fungus are quite di€erent in

terms of chemical composition and elicitor activity.

Thus, the fragments recovered from the resistant cultivar

are much more active than the fragments of the

susceptible cultivar whose elicitor e€ect on PR expres-

sion (b-1,3-glucanase, chitinase) and ethylene is very

low. Activity correlates to an enrichment of the most-

active fractions in the two neutral sugars, galactose and

arabinose, thereby indicating the presence of

rhamnogalacturonan-I (RG-I) side chains. Such a com-

position reveals that the pure endoPG does not cleave R

and S pectic polysaccharides at the same locations. The

fact that the enzyme only cleaves unesteri®ed GalA-

containing regions with a minimum GalA DP of 4,

suggested that the presence and patterns of esteri®ed

uronosyl residues might di€er between resistant and

susceptible seedlings. Indeed, analysis of the C6-esteri-

®cation degree of the various fragments showed that the

R-fractions of high GalA content (>70%) i.e. most

fractions, were generally more esteri®ed than their S-

homologs. Consistant with their enrichment in RG-I

regions, the R-fragments were also enriched in acetyl

residues. Taken together, the data indicate that the

composition of pectic fragments released from bean cell

walls by the endoPG closely re¯ects the extent of GalA

C6 esteri®cation, and that pectic polysaccharides are less

acidic in resistant than in susceptible seedlings. This was

recently con®rmed by measuring the cation-exchange

capacity of the cell walls from the two sources (data not

shown). Thus, the di€erent compositions of R and S

pectic fragments and their di€erential activity are

sucient to account for the di€erential elicitor e€ect of

endoPG on R and S seedlings. The high density of

endoPG molecules at sites of infection (Benhamou et al.

1991) allows us to predict that such fragments are also

produced in vivo. This does not exclude the possibility

that other factors, notably PGIP and remorin, the

recently isolated uronide-binding protein (Reymond

et al. 1996), may modulate the endogenous elicitation

phenomenon, but this must await further investigation.

Since the fragments were partly puri®ed, their struc-

ture-function relationship was not investigated, and the

precise distribution of ester and acetyl groups was not

determined. However, additional experiments where the

R-fragments were de-esteri®ed by saponi®cation (data

not shown), almost completely abolished their elicitor

activity. This is in favor of GalA-containing regions as

the active domains. Why, then, were the S-fragments less

active? It is conceivable that a certain degree of

esteri®cation would prevent autoassembly of the gala-

cturonan portions of the fragments via calcium ions,

thus enabling better accessibility to the host plant.

Alternatively, it might limit further hydrolysis of the

fragments by the previously reported endogenous poly-

galacturonase activity of bean tissues (Pressey and

Avants 1977). The fact that oligogalacturonides of DP

10±15 are have poor activity strengthens this hypothesis.

Fig. 4. Elicitor e€ect of a mixture of oligogalacturonides of DP 10±15

on chitinase, b-1,3-glucanase and ethylene in resistant (Pr.R) or

susceptible (Pr.S) bean seedlings. Each cutting was supplied with

50 lg or 250 lg GalA equivalent oligogalacturonides. Data represent

the means ‹ SD of three independent experiments

Fig. 5. Time course study of the accumulation of basic b-1,3-

glucanase and basic chitinase transcripts in resistant (Pr.R) bean

seedlings elicited with either R300 or S300 pectic fragments (250 lg

GalA equivalents per cutting). The controls were treated with water

plus streptomycin. Total RNA (7 lg) denatured in glyoxal/dimethyl

sulfoxide was separated by electrophoresis in an agarose gel,

transferred to nylon membrane and hybridized with a radiolabelled

probe corresponding to basic b-1,3-glucanase and chitinase

92

G. Boudart et al.: Di€erential elicitation of defense responses by pectic fragments in bean seedlings

The latter result does not con®rm most literature reports

which state that the optimum DP for oligogalacturonide

activity ranges between 10 to 15 GalA residues. A

possible explanation to this discrepancy might be that

most oligogalacturonide bioassays are performed on

model systems, most often cell-suspension cultures or

tissue explants, whose cell surface structures and acces-

sibility are quite di€erent from the parameters encoun-

tered at the seedling level. On the other hand the

occurrence of active, complex pectin oligomers contain-

ing GalA and neutral sugars has already been reported

in other systems (Aldington and Fry 1994; Rong et al.

1994). Even though the pectic fragments used in this

study are partly composed of oligogalacturonides, our

results present evidence that complex uronides are much

more active than oligogalacturonides per se.

The contribution of endogenous elicitor signals to the

outcome of the bean-Colletotrichum interaction during

infection will be addressed in the future. Strains of C.

lindemuthianum where the endoPG genes and enzymes

are no longer functional either by disruption or by

mutagenesis, would be useful for this purpose. Towards

this end, we have cloned the two endoPG genes of the

fungus (Centis et al. 1996, 1997), and are currently

engineering new strains.

Besides endoPG, one should also consider the

enzymes which mediate esteri®cation and de-esteri®ca-

tion of the host pectin. Pectin methylesterases which are

present in the host plant and in the fungus, might

facilitate accessibility of endoPG to its substrate. The

recent report that a potato genotype susceptible to

Erwinia carotovora is characterized by a lower content of

esteri®ed pectin than its resistant counterpart, and

correlatively by the presence of an additional pectin

methylesterase isoform (Marty et al. 1997), reinforces

our assumption that the concerted action of pectin

methylesterases and endoPGs play important roles in

cell surface signaling.

We are grateful to our colleague Dr. B. Dumas (UMR 5546 CNRS-

UPS, Toulouse, France) for help in the preparation of the

molecular probes.

References

Aldington S, Fry SC (1994) Rhamnogalacturonan-II ± a biolog-

ically active fragment. J Exp Bot 45: 287±293

Barthe JP, Cantenis D, Touze A (1981) Puri®cation and charac-

terization of two polygalacturonases secreted by Colletotrichum

lindemuthianum. Phytopathol Z 100: 162±171

Bell JN, Ryder TB, Wingate VPM, Bailey JA, Lamb CJ (1986)

Di€erential accumulation of plant defense gene transcripts in a

compatible and incompatible plant-pathogen interaction. Mol

Cell Biol 6: 1615±1623

Benhamou N, La®tte C, Barthe JP, EsquerreÂ-Tugaye MT (1991)

Cell surface interactions between bean leaf cells and Colletotri-

chum lindemuthianum. Cytochemical aspects of pectin break-

down and fungal endopolygalacturonase accumulation. Plant

Physiol 97: 234±244

Blumenkrantz N, Asboe-Hansen G (1973) A new method for

quantitative determination of uronic acids. Anal Biochem 54:

484±489

Boudart G, Dechamp-Guillaume G, La®tte C, Ricart G, Barthe JP,

Mazau D, EsquerreÂ-Tugaye MT (1995) Elicitors and suppres-

sors of hydroxyproline-rich glyprotein accumulation are solu-

bilized from plant cell walls by endopolygalacturonase. Eur J

Biochem 232: 449±457

Brecht JK, Huber DJ (1988) Products released from enzymically

active cell wall stimulate ethylene production and ripening in

preclimateric tomato (Lycopersicon esculentum Mill.) fruit.

Plant Physiol 88: 1037±1041

Broglie KE, Biddle P, Cressman R, Broglie R (1989) Functional

analysis of DNA sequences responsible for ethylene regulation

of a bean chitinase gene in transgenic tobacco. Plant Cell 1:

599±607

Centis S, Dumas B, Fournier J, Marolda M, EsquerreÂ-Tugaye MT

(1996) Isolation and sequence analysis of Clpg1, a gene coding

for an endopolygalacturonase of the phytopathogenic fungus

Colletotrichum lindemuthianum. Gene 170: 125±129

Centis S, Guillas I, SeÂjalon N, EsquerreÂ-Tugaye MT, Dumas B

(1997) Endopolygalacturonase genes from Colletotrichum

lindemuthianum: cloning of CLPG2 and comparison of its

expression to that of CLPG1 during saprophytic and parasitic

growth of the fungus. Mol Plant-Microbe Interact 10: in press

Cervone F, De Lorenzo G, Degra L, Salvi G, Bergami M (1987)

Puri®cation and characterization of a polygalacturonase-inhib-

iting protein from Phaseolus vulgaris L. Plant Physiol 85: 631±

637

CoÃte F, Hahn MG (1994) Oligosaccharins: structure and signal

transduction. Plant Mol Biol 26: 1379±1412

Darvill A, Augur C, Bergmann C, Carlson RW, Cheong J,

Eberhard S, Hahn MG, Lo VM, Marfa V, Meyer B, Mohnen

D, O'Neill MA, Spiro MG, van Halbeek H, York WS,

Albersheim P (1992) Oligosaccharins ± oligosaccharides that

regulate growth, development and defence responses in plants.

Glycobiology 2: 181±198

Daugrois JH, La®tte C, Barthe JP, Touze A (1990) Induction of b-

1,3-glucanase and chitinase activity in compatible and in

incompatible interactions between Colletotrichum lindemuthia-

num and bean cultivars. J Phythopathol 130: 225±234

Daugrois JH, La®tte C, Barthe JP, Faucher C, Touze A, EsquerreÂ-

Tugaye MT (1992) Puri®cation and characterization of two

basic b-1,3-glucanases induced in Colletotrichum lindemuthia-

num-infected bean seedlings. Arch Biochem Biophys 292: 468±

474

Davis KR, Lyon GD, Darvill AG, Albersheim P (1984) XXV.

Endopolygalacturonic acid lyase isolated from Erwinia car-

otovora elicits phytoalexin accumulation in soybeans. Plant

Physiol 74: 52±60

Eberhard S, Doubrava N, MarfaÁ V, Mohnen D, Southwick A,

Darvill A, Albersheim P (1989) Pectic cell wall fragments

regulate tobacco thin-cell layer explant morphogenesis. Plant

Cell 1: 747±755

Edington BV, Lamb CJ, Dixon RA (1991) cDNA cloning and

characterization of a putative 1,3 b-glucanase transcript in-

duced by fungal elicitor in bean cell suspension cultures. Plant

Mol Biol 16: 81±94

Feinberg AP, Vogelstein B (1983) A technique for labelling DNA

restriction endonuclease fragments to high speci®c activity.

Anal Biochem 132: 6±13

Hotchkiss AT, Hicks KB (1990) Analysis of oligogalacturonic acids

with 50 or fewer residues by high-performance anion-exchange

chromatography and pulsed amperometric detection. Anal

Biochem 184: 200±206

Jones DA, Thomas CM, Hammond-Kosack KE, Balint-Kurti PJ,

Jones JDG (1994) Isolation of the tomato Cf-9 gene for

resistance to Cladosporium fulvum by transposon tagging.

Science 266: 789±793

La®tte C, Barthe JP, Montillet JL, Touze A (1984) Glyco-

protein

inhibitors

of

Colletotrichum

lindemuthianum

endopolygalacturonase in near isogenic lines of Phaseolus

vulgaris resistant and susceptible to anthracnose. Physiol Plant

Pathol 25: 39±53

G. Boudart et al.: Di€erential elicitation of defense responses by pectic fragments in bean seedlings

93

La®tte C, Barthe JP, Gansel X, Dechamp-Guillaume G, Faucher

C, Mazau D, EsquerreÂ-Tugaye MT (1993) Di€erential

induction by endopolygalacturonase of b-1,3-glucanases in

Phaseolus vulgaris isoline susceptible and resistant to Col-

letotrichum lindemuthianum race b. Mol Plant-Microb Interact

6: 628±634

Lee SC, West CA (1981) Polygalacturonase from Rhizopus

stolonifer, an elicitor of casbene synthetase activity in castor

bean (Ricinus communis L.) seedlings. Plant Physiol 67: 633±639

Maness NO, Ryan JD, Mort AJ (1990) Determination of the

degree of methyl esteri®cation of pectins in small samples by

selective reduction of esteri®ed galacturonic acid to galactose.

Anal Biochem 185: 346±352

MarfaÁ V, Gollin DJ, Eberhard S, Mohnen D, Darvill A,

Albersheim P (1991) Oligogalacturonides are able to induce

¯owers to form on tobacco explants. Plant J 1: 217±225

Marty P, Jouan B, Bertheau Y, Vian B, Goldberg R (1997) Charge

density in stem cell walls of Solanum tuberosum genotypes and

susceptibility to blackleg. Phytochemistry 44: 1435±1441

Mathieu Y, Kurdjian A, Xia H, Guern J, Koller A, Spiro MD,

O'Neill M, Albersheim P, Darvill A (1991) Membrane respons-

es induced by oligogalacturonides in suspension-cultured to-

bacco cells. Plant J 1: 333±343

Melotto E, Greve LC, Labavitch JM (1994) Cell wall metabolism in

ripening fruit. VII. Biologically active pectin oligomers in

ripening tomato (Lycopersicon esculentum Mill.) fruits. Plant

Physiol 106: 575±581

Nevins DJ, English PD, Albersheim P (1967) The speci®c nature of

plant cell wall polysaccharides. Plant Physiol 42: 900±906

Pressey R, Avants JK (1977) Occurrence and properties of

polygalacturonase in Avena and other plants. Plant Physiol

60: 548±553

Reymond P, Kunz B, Paul-Pletzer K, Grimm R, Eckerskorn C,

Farmer EE (1996) Cloning of a cDNA encoding a plasma

membrane-associated, uronide binding phosphoprotein with

physical properties similar to viral movement proteins. Plant

Cell 8: 2265±2276

Rickauer M, Fournier J, EsquerreÂ-Tugaye MT (1989) Induction of

proteinase inhibitors in tobacco cell suspension culture by

elicitors of Phytophthora parasitica var. nicotianae. Plant

Physiol 90: 1965±1970

Rong L, Carpita NC, Mort A, Gelvin SB (1994) Soluble cell

wall compounds from carrot roots induce the picA and pgl

loci of Agrobacterium tumefaciens. Mol Plant-Microbe Interact

7: 6±14

Rupley JA (1964) The hydrolysis of chitin by concentrated

hydrochloric acid and the preparation of low molecular weight

substrates for lysozyme. Biochim Biophys Acta 83: 245

Sambrook J, Fritsh EF, Maniatis T (1989) Molecular cloning: a

laboratory manual, 2nd edn. Cold Spring Harbor Laboratory

Press, New York

Sanger F, Nicklen S, Coulson AR (1977) DNA sequencing with

chain-terminating inhibitors. Proc Natl Acad Sci USA 74: 5463

Somogyi M (1952) Notes on sugar determination. J Biol Chem 195:

19±23

Staskawicz BJ, Ausubel FM, Baker BJ, Ellis JG, Jones JDG (1995)

Molecular genetics of plant disease resistance. Science 268: 661±

667

94

G. Boudart et al.: Di€erential elicitation of defense responses by pectic fragments in bean seedlings