Early events in resistant and susceptible interactions between Brassica napus and Leptosphaeria maculans
 
 
 
 
 
 
 
 
 

Peijun Zhang¹, Jinxing Tu² and Brian Fristensky
 
 
 

Department of Plant Science, University of Manitoba, Winnipeg, R3T 2N2, Canada

¹ Present address: AAFC/ECORC 960 Carling Avenue, Ottawa, K1A 0C6 Canada

² Present address: Department of Agronomy, Huazhong Agricultural University, Wuhan 430070, China
 
 
 
 
 

Corresponding author: B. Fristensky. E-mail: frist@cc.umanitoba.ca
 

GenBank Accession #: Ypr1.1, U64806; Ypr1.2, U70666

ABSTRACT
 

Despite the agricultural importance of blackleg disease caused by Leptosphaeria maculans, little is known about the early cellullar events leading to resistance or susceptibility. We now show that compatible pathogenicity group PG3 germinates within 3 days postinoculation, with extensive hyphal proliferation by 9 days. In contrast, incompatible isolate PG2 shows little evidence of germination until 5 days postinoculation, with only limited growth apparent by 9 days. To probe the events prior to spore germination, we have cloned two cDNAs encoding basic PR1 proteins from the interaction of Brassica napus cv. Glacier with incompatible isolate PG2. With incompatible isolate PG2, PR1 gene expression is apparent within 8 hours postinoculation, and remains strong for at least 96 hours, while expression in response to compatible isolate PG3 declines after 12 hr. A similar induction pattern was observed for -1,3-glucanase. Surprisingly, no induction of defensin or PR5 was detected, suggesting that these genes are regulated through a different pathway than PR1 and -1,3-glucanase. Heat-killed pycnidiospores elicit only weak PR1 expression. When different pathogenicity groups were compared, strongest PR1 expression was found in single gene resistance to PG2 in Glacier and PG3 in Quinta, compared to compatible interactions. Weaker PR1 expression was seen in non-host resistance to non-agressive isolate PG1. Salicylic acid, but not wounding or heat shock, can also induce PR1 expression in B. napus. Negligible expression was seen in healthy cotyledons, young leaves, mature leaves, flowers, and young siliques. These results show B. napus responds differentially to compatible and incompatible isolates within 12 hours after inoculation, long before fungal germination occurs.
 

Additional keywords: Brassica, Leptosphaeria maculans, blackleg, pathogenesis-related proteins, PR1
 

INTRODUCTION

In most fungal diseases of plants, the events which determine resistance or susceptibility occur long before symptoms become apparent. Despite the agricultural importance of blackleg disease in crucifers, only a small number of studies have examined the cellullar events leading to resistance or susceptibility to the causal agent, Leptosphaeria maculans (Phoma lingam). While L. maculans can attack most plant organs, the most important phase is blackleg, or stem canker, resulting from an initial infection of leaves. In Brassica napus, hyphae invade through wounds or stomates, and grow biotrophically in intercellular spaces. The blackleg fungus propagates through the petioles into the stem (Hammond et al. 1985). In susceptbile interactions, blackleg switches to a necrotrophic phase, resulting ultimately in visible symptoms. Resistance to blackleg in B. napus occurs during this biotrophic phase. In contrast, resistance in B. juncea appears to be mediated through necrosis of guard cells, preventing the initial invasion of mesophyl (Chen and Howlett, 1996)

Resistance to blackleg in B. napus is controlled by single or multiple major genes (Pang and Halloran 1996a; 1996b; Rimmer and Van den Berg 1992). Both avirulent and virulent isolates in L. maculans have been identified. At least one avirulence gene, AvrLm1, has been genetically defined (Ansan-Melayah et al. 1995), indicating that the interaction may follow a gene-for-gene mechanism. Based on the disease reaction profiles in three differential cultivars, Mengistu et al. (1991) grouped L. maculans isolates into four pathogenicity groups: PG1 - PG4.

Macroscopically, blackleg resistance in these differentials can only be distinguished from susceptibility five to 10 days postinoculation. Only a few studies have begun to dissect the molecular events preceeding visual symptom development. Roussel et al. (1999) found that resistance to avirulent L. maculans in B. napus is characterized by a hypersensitive response (HR) as well as fragmentation of chromatin in the nucleus. As early as 6 hpi, pectin-like material began to be detectible in the lumen of vascular gels, so that by 72 hpi intense histocytochemical staining of these tissues was evident. The authors hypothesize that pectic gels could inhibit the movement of hyphe through the plant. Rasmussen et al. (1992) cloned a basic chitinase that was more strongly induced by L. maculans (Phoma lingam) in a resistant cultivar, compared to a susceptible cultivar. In the same plant/fungus interaction, antibodies specific for tobacco -1,3-glucanase (PR2), class IV chitinase (PR-Q) and a thaumatin/osmotin-like protein (PR-S) detected proteins that accumulated in both B. napus and B. nigra plantlets within 26 hours after inoculation with L. maculans (Dixelius, 1994). -1,3-glucanase was also induced in B. rapa (B. campestris) inoculated with Xanthomonas campestris (Newman et al., 1994). Pathogenesis-related gene PR1 was found to be expressed in senescent B. napus leaves, but expression in response to pathogens was not studied (Hanfrey et al., 1996).

Although majority of the data obtained in the blackleg system has previously been obtained at times when macroscopic symptom development was already evident, the present study seeks to identify early events which distinguish resistance from susceptibility. We have determined that germination of infiltrated spores is first seen within 3 dpi in compatible interactions, but not until 5 dpi in incompatible interactions. To enable the study of defense gene activation, we have cloned two PR1 cDNAs from B. napus, and demonstrate that PR1 gene expression is significantly stronger in an incompatible interaction compared to a compatible interaction within 24 hpi. Induction of PR1 requires live inocula, and evidence from uninoculated plants suggests that PR1 is controlled by pathogenesis-related, rather than developmentally regulated pathways.
 

METHODS AND MATERIALS

Plant materials and pathogen isolates

Brassica seeds and L. maculans (Desm.) Ces. et De Not. isolates were kindly supplied by Dr. Roger Rimmer and Dr. Rachael Scarth, Dept. of Plant Science, Univ. of Manitoba. Cultivars used were Glacier, Quinta, Westar. Pycnidiospores were prepared according Mengistu et al. (1991). All plants were grown in Jiffy Pots filled with Metromix (W.R. Grace & Co. Ltd., Ajax, Ontario, Canada) in the growth chamber conditioned at 20/16^oC day/night temperature and a 16 hr photoperiod. Pathogen isolates were 84-10 (PG1), 86-12 (PG2), 89-18 (PG3), and LIFOLLE 5 (PG4). For experiments in which killed inocula was used, pycnidiospore suspensions were heated in a boiling water bath for 25 min.
 

Plant inoculation and chemical/physical treatments

Pycnidiospore suspension (2 10⁷ spores/ml), sterile distilled water, and 2 mM sodium salicylate were infiltrated into cotyledons or leaves using a 1-cc syringe without a needle. The outlet of the syringe was covered with a section of Tygon tubing. Infiltration was conducted by shooting inocula directly into cotyledons or leaves through the abaxial face while supporting the cotyledons or leaves on the operator's gloved index finger. One or two inoculated plants were kept in the growth chamber for at least two weeks to verify interaction phenotype development. Wounding was done by puncturing leaves repeatedly with a needle. Heat shock was conducted by incubating plants at 40^oC for 2 hr. SA, wounded, and heat shocked cotyledons were harvested 24 hr after treatment.
 

Microscopy

Calcofluor white M2R stain (Roringer et al., 1977): Inoculated cotyledons were fixed in ethanol/glacial acetic acid (3:1) for 48 hours at room temperature, cleared by boiling for 2 minutes in chloral hydrate (2.4 g/ml) and then washed 2 x 15 min. with 50% ethanol, 2 x 15 min. with 0.05 M NaOH, 3x with water, and placed in 0.1M Tris-Cl, pH8.5 for 30 min. Cotyledons were stained with Fungi-Fluor solution A and B (Sigma, No. F6259) overnight. Finally, tissue was washed 4 x 10 min. in water and 1 x 30 min. with 25% aqueous glycerol. Cotyledons were mounted in glycerol containing a trace of lactophenol as a preservative, and examined with a Zeiss Research Microscope.
 

Aniline Blue stain: Inoculated cotyledons were fixed in ethanol/glacial acetic acid (3:1) for 48 hours at room temperature, and immersed in clearing-staining solution (2ml/cm² of tissue) [Bruzzese and Hasan, 1983] for 48 hours at room temperature. They were then placed in chloral hydrate (2.5 g/ml) for 24 - 48 hours, and rinsed in distilled water. Cotyledons were examined by light microscopy.
 

cDNA library screening

Construction of the cDNA library using total RNA from B. napus cv. Glacier leaves inoculated with L. maculans isolate PG2 for 48 hr. was described previously [Fristensky et al., 1999].

The library was screened with A. thaliana PR1 clone pBSPR1 (GB::M90508) [Uknes et al. 1992] using DuPont Colony/Plaque Screen Hybridization transfer membranes and probed with Arabidopsis PR1 cDNA. After two rounds of screening, positive clones were picked and the pBluescript SK phagemids were in vivo excised using Strategene In Vivo Excision System according to the manual. The sizes and authenticity of the inserts were checked by DNA hybridization.
 

DNA sequencing and analysis

Inserts were digested at internal restriction sites and subcloned. Sequencing of both strands was done using Applied Biosystems 373A DNA Sequencer at the Plant Biotechnology Institute, Saskatoon, Canada. Sequence analysis was done using the BIRCH facility (Fristensky, 1999). All programs were run from the Genetic Data Environment (GDE 2.3) [Smith et al. 1994].
 

RNA and DNA isolation and analysis

Plant materials were frozen in liquid N₂ immediately at harvest. Frozen tissues were used directly for RNA or DNA extraction, or stored at - 70^oC (for RNA extraction) or freeze dried and stored at -20^oC (for DNA extraction). Total RNA was extracted from plant cotyledons or leaves using the method of Kim et al. [1992]. Genomic DNA was extracted by using a CTAB protocol (Ausubel et al. 1994). For RNA hybridization analysis, RNAs (10 g/lane) were separated by 1.2% agarose-formaldehyde gels and blotted onto Bio-Rad Zeta-Probe GT membranes. For DNA hybridization, DNA samples digested with restriction enzymes were separated by 1.0 % agarose gels and blotted onto the same membrane. Hybridization was conducted according to the manufacturer's manual. Probes used included B. napus Ypr1.1 clone pPZ234-4 (GB::U64806), PR2 (-1,3-glucanase) clone pBSAPR2 (GB::M90509), PR5 clone pBSPR5 (GB::M90510) (Uknes et al., 1992); and radish defensin clone pFRG1 (GB::U18557, Terras et al., 1995). Autoradiographs were scanned using a pdi325oe Biological Imaging System. Digitized data were normalized with the average of all signals on each autoradiograph.
 

RESULTS

Inoculation by infiltration results in infection phenotypes that coincide with known pathology

The ability to study the early events in the Brassica-L. maculans interaction requires an inoculation method in which a large number of cells can be challenged. Standard pinprick inoculation methods for blackleg do not challenge a large enough area of tissue to obtain amounts of RNA adequate for Northern blotting. At the same time, resistance occurs while L. maculans is in its intercellular biotrophic phase (Hammond et al. 1985). We therefore inoculated leaves or cotyledons by infiltration, which challenges a large area of cells in a uniform fashion, allowing for a more synchronous infection system than would be possible with surface inoculation. Either cotyledons or the first two fully expanded true leaves of B. napus cv. Glacier were inoculated with pycnidiospore suspensions (2 10⁷ spores/ml) of L. maculans PG2 (incompatible) or PG3 (compatible) by applying the suspension through a 1-cc syringe to which no needle was attached (Fig. 1A). Similar interaction phenotypes were observed in both cotyledons and leaves. In compatible interactions, brown to dark gray necrosis develops, accompanied by tissue collapse. By 14 days postinoculation, black pycnidia are clearly seen. In incompatible interactions, chlorosis develops, followed by a yellowish necrosis at later times (Fig. 1B). Furthermore, we found two striking differences in symptom development between the two interactions. 1) Macroscopical symptoms are evident by day 4 in incompatible interactions but day 5 in compatible interactions; 2) lesions are limited to the inoculated areas in incompatible interactions but not in compatible interactions (Fig. 1C).

In order to determine whether or not live inocula are required for these interaction phenotypes, heat-treated pycnidiospore suspensions were used to inoculate Glacier cotyledons by infiltration (Fig. 1D). While interaction phenotypes developed normally in the cotyledons infiltrated with live PG2 and PG3 spores, no response was observed in the cotyledons inoculated with the heat-killed spores.
 

Spore germination occurs later in incompatible interactions

One goal was to determine the differences between germination and growth of compatible isolate PG3 with incompatible isolate PG2. Two types of staining were tested to visualize germinating spores: aniline blue and calcofluor. Although ungerminated spores were often visible with aniline blue under transmitted white light, their small size, along with the fact that many host cellular components took up strain, made it difficult to distinguish spores from other components (data not shown). A much better contrast between spores and host cells was seen with calcofluor fluorescence. As shown in Fig. 2A, the primary background in water-inoculated control tissue comes from host cell walls. As shown in Fig. 2B-E, fungal spores and mycelia show up brightly against the background fluorescence.

Germination of compatible isolate PG3 was seen within 3 days postnoculation (Fig. 2C), whereas incompatible PG2 spores appeared ungerminated at this time (Fig. 2B). Germination was first detected with PG2 (Fig. 2D) at 5 d.p.i., at which time prolitic hyphal growth was seen with PG3 (Fig. 2E). Extensive spread of hyphae is seen by 9 d.p.i. in the susceptible interaction with PG3 (Fig. 2G). In contrast, far less hyphal growth is seen in resistant interaction with PG2 (Fig. 2F).
 

At least two distinct PR1 orthologues are found in B. napus

A cDNA library derived from Glacier leaves inoculated with PG2 (48 h. p. I.) [Fristensky et al., 1999] was screened with Arabidopsis PR1 cDNA (Uknes et al. 1992). Sequencing of 3' ends of nine positive clones showed at least two distinct classes of PR1 genes. One clone from each class, pPZ234-4 and pPZ234-19, was fully sequenced, and the genes corresponding to these sequences were designated as Ypr1.1 and Ypr1.2, respectively (Fig. 3). Conceptual translation revealed that Ypr1.1 and Ypr1.2 encode 161 AA and 162 AA polypeptides, respectively, each with a predicted 26-AA hydrophobic signal peptide similar to that found in tomato PR1 homologue p14 (Lucas et al., 1985). Ypr1.1 and Ypr1.2 encode mature basic proteins whose predicted pIs are 9.02 and 8.16, respectively. The two PR1 cDNAs are 88.9% identical at the nucleotide level and 92.0% at the protein level.

DNA hybridization was conducted to estimate the copy number for PR1. Genomic DNA from B. napus cv. Westar was digested with DraI, EcoRI, BamHI, or HindIII, none of which digest Ypr1.1 and Ypr1.2 cDNAs. A blot probed with Ypr1.1 is shown in Fig. 3. The four restriction enzymes produce one to four bands, suggesting that Brassica PR1 is a small multigene family of at least two to four copies per haploid genome.
 

PR1 genes are induced in the B. napus-L. maculans interaction

The results in Fig. 1 show that the incompatible and compatible interactions in the Brassica-L. maculans pathosystem are strikingly different. However, this macroscopic difference can only been seen at 5 d.p.i. or later. On the assumption that resistance or susceptibility are determined earlier, we examined the induction of PR1 in leaves during incompatible and compatible interactions. The first two fully expanded leaves of Glacier were inoculated with pycnidiospore suspensions of PG2, PG3 or water. Total RNAs extracted from inoculated leaves 48 h. p. i. were analyzed by RNA hybridization using Ypr1.1 cDNA as probe. The results are shown in Fig. 5. In water-treated leaves, no significant signal was detected. To verify that PR1 induction is the result of an interaction between plant and fungus, we compared the PR1-eliciting activity of live and dead inocula. Glacier cotyledons were inoculated with the live and heat-killed spore suspensions. As shown in Fig. 5, dead spores induce only minimal PR1 expression at 48 h. p. i., as evidenced by RNA gel blot analysis.

To determine how early PR1 is induced by L. maculans, Glacier cotyledons were inoculated with PG2, PG3, or water and harvested at several times from 0 to 96 h. p. i. Hybridization analysis of RNAs from these timepoints is shown in Fig. 6. As above, water inoculation did not induce PR1 at any time. Induction of PR1 was detected at all times in fungus-inoculated samples, even as early as 5 h. p. i. (data not shown). In the first 12 h. p. i., PR1 is induced in both the Glacier-PG2 and Glacier-PG3 interactions. Induction in the Glacier-PG3 interaction is slightly stronger than in the Glacier-PG2 interaction. At 24 h. p. i., expression decreases significantly in both treatments. However, from 36 to 96 hr.p.i, PR1 is expressed strongly in the incompatible Glacier-PG2 interaction but relatively weakly in compatible Glacier-PG3 interactions, in which signal returns to near control levels by 96 h. p. i.
 

PR2, but not PR5 and defensin, are also induced in the Brassica-L. maculans interaction

To detect possible coordinate expression of PR1 with other defense genes, timecourse RNAs were also hybridized with probes for A. thaliana -1,3-glucanase (PR2), A. thaliana thaumatin/osmotin-like protein PR5and radish defensin (Fig. 7). The expression pattern of PR2 is similar to that of PR1, in that PR2 is strongly-expressed in both interactions in early hours, but expression drops to near control levels by 24 h.p.I. in the compatible interaction. One difference, in comparision to PR1 is that within the first 12 h. p. i., the induction of PR2 is stronger in PG3-inoculated plants (compatible) than in PG2-inoculated ones (incompatible). In contrast, defensin mRNA was seen even in the water control, with no obvious differences between PG2- and PG3-inoculated plants. Expression of PR5 could not be detected in any interaction (data not shown).
 
 
 

PR1 expression differs among infection types

Several blackleg isolates were used to compare PR1 expression in different infection types. Since previous experiments showed substantial difference at 36 h.p.I., B. napus leaves were inoculated for 36 hours with L. maculans isolates PR1, PG2, PG3, PG4 or water (Fig. 8). Most B. napus cultivars express a non-host resistance to PG1, which is thought to be a separate species from L. maculans (Taylor et al., 1991). While all 3 cultivars are resistant to PG1, cv. Glacier showed only moderate PR1 expression in response to PG1, while cv. Quinta showed very little PR1 expression, and cv. Westar showed no expression. Both Glacier and Quinta express single gene resistance to PG2 (Rimmer and van den Berg, 1992), and both show strong PR1 expression, in contrast to weak PR1 expression in cv. Westar. Only Quinta expresses resistance to the more agressive isolate PG3, and this cultivar also shows the strongest PR1 expression. None of the 3 cultivars is resistant to PG4, and all show weak PR1 expression.
 

Salicylic acid can stimulate PR1 expression in B. napus

Defense gene expression has often been shown to be influenced by developmental or environmental stimuli other than pathogens. For this reason, expression of Brassica PR1 was assayed by RNA gel blots using RNA from cv. Glacier leaves treated with 2 mM sodium salicylate (SA), wounding or heat shock. As shown in Fig. 9, SA can stimulate expression of PR1, though the hybridization signal is weaker than that in the Glacier-PG2 interaction. However, no detectable hybridization signals were found in heat shocked or wounded Glacier leaves. As well, no PR1 mRNA was detected in healthy cotyledons, mature leaves, flowers or young siliques.
 

DISCUSSION

Infiltration inoculation

In commonly-used L. maculans inoculation protocols, interaction phenotypes are not apparent until several days postinoculation. At the same time, these methods produce very small lesions at pinpricks on the cotyledons of resistant cultivars. To follow the molecular events occurring in the early hours postinoculation, a large number of cells must be challenged uniformly. Though infiltration is commonly used for inoculation with bacterial pathogens, it has seldom been used with fungal pathogens partially due to the large sizes of most fungal inocula. However, L. maculans pycnidiospores are quite small (3-5 1.5-2 m, Williams, 1992), allowing them to pass into Brassica cotyledons, often through the stomates, a natural site of entry.

The interaction phenotypes we observed using infiltration are quite similar to these produced by the pinprick inoculation (Mengistu et al., 1991). However, some subtle features of symptom development can only be observed in infiltration-inoculated plants. Chlorosis in the incompatible interactions can be observed by naked eyes 4 d.p.i., one day earlier than in the compatible interactions. Presumably, chlorosis in incompatible interactions is the macroscopic manifestation of a hypersensitive response. Two observations supports this hypothesis. 1) Heat-killed spores and spore-free inoculum (ultrafiltrate of pycnidiospore suspension, data not shown) produced no chlorosis; 2) the chlorosis was always delimited in the inoculated areas in incompatible interactions. Hammond and Lewis' (1987) observations also support this hypothesis. They found that host cell necrosis took place beyond hyphal front in incompatible B. napus-L. maculans interaction.
 

Brassica PR1 genes and proteins

We have shown that PR1 is a small gene family in Brassica, containing at least two classes of PR1 genes, represented by Ypr1.1 and Ypr1.2. Mutational distance between the two Brassica PR1s is small, suggesting a recent divergence from a single gene. PR1 proteins are targeted either extracellularly or to vacuoles. In tobacco, acidic and basic PR1s are targeted to apoplast and vacuoles, respectively (Bol et al. 1990). In other species, targeting of PR1 does not always follow the tobacco model. For example, two barley basic PR1 proteins (HvPR-1a, HvPR1b) are detected extracellularly (Bryngelsson et al. 1994). Tomato basic PR1 protein p14 is present in both intercellular spaces and in vacuoles (Vera et al. 1989). It was also reported that an acidic tobacco PR1 (PR-1b) could be detected within crystal idioblast vacuoles (Dixon et al. 1991). Other PR proteins, such as basic tomato PR5 (NP24) (King et al. 1988), potato basic chitinase (Kombrink et al. 1988) are also secreted extracellularly. Both vacuole-targeted and apoplast-secreted pre-PR1s contain hydrophobic signal peptides about 25-AA long. Some PR proteins contain C-terminal extensions. Both Brassica PR1s contain a 26-AA long signal peptide but neither contains the C-terminal extension. It was postulated that neither the signal peptides nor the C-terminal extensions control PR1 targeting (Bol et al. 1990), but the explicit motifs controlling PR1 protein targeting are unknown at present. The targeting of the two Brassica PR1s remains to be demonstrated experimentally.

Tobacco and tomato PR1 proteins display antifungal activity both in vivo and in vitro (Niderman et al. 1995). However, antifungal activity of PR1 proteins from other plants, including Ypr1.1 and Ypr1.2 from B. napus, has not yet been reported.
 

Expression of Brassica PR1 preceeds spore germination

In L. maculans-Brassica interactions, macroscopic interaction phenotypes cannot be determined until several days postinoculation. In our experiments, the induction of PR1 could be detected as early as five h. p. i. (data not shown). This indicates that the earliest molecular response to L. maculans must take place within a few hours postinoculation.

At the same time, spore germination was not observed in our experiments until at least 3 d.p.I. in the compatible interaction, and 5 d.p.I. in the incompatible interaction, while no PR1 induction was seen in mock-inoculated plants. These data indicate that B. napus is capable of detecting the presence of ungerminated spores, probably through the presence of fungal-derived elicitors. Roussel et al. (1999) detected the production of pectin-like gels in vascular tissue within 24 h.p.i. as well as nuclear modifications, such as chromatin fragmentation, within 72 h.p.I. In their experiments, similar results were obtained using either L. maculans pycnidiospores, or the Phytophthora-derived elicitor cryptogein.

In the first 12 h. p. i. PR1 and PR2 expression are similar between compatible and incompatible interactions. At later times, these genes are more strongly-expressed in the incompatible interaction. Two hypotheses can explain this difference. One hypothesis is that the compatible PG3 suppresses defense mechanisms at later times, including PR1 and PR2. The other possibility is that the initial induction seen with PG3 until 24h.p.I. represents a non-specific response, which is augmented in the incompatible PG2 by some additional specific response. The latter would be more consistent with a dominant resistance gene/avirulence gene mechanism.

Non-specific activation of defenses in early hours postinoculation, followed by a second wave of induction specific to the incompatible interaction, has been observed in other systems. Components of the oxidative burst in response to bacterial infection are activated in tobacco and tomato cells within the first h. p. i. with either compatible or incompatible pathovars, while a second, longer-lived response occurs 1.5 to 3 hr. p.i. in incompatible interactions only (Baker et al., 1991; Keppler et al., 1989). In pea tissue inoculated with Fusarium solani races, PR10 and DRR206 genes are induced during the first 8 h. p. i., with both compatible or incompatible races, but a second wave of induction 12 - 48 h. p. i. is only seen with the incompatible race (Fristensky et al., 1985). A comparable pattern of induction was seen for PAL, 4CL, and PR10 in parsley leaves inoculated with Phytophthora megasperma var sojae (Reinold and Hahlbrock, 1996).

Induction of PR1 can result from the interaction between the plant and the growing pathogen, but it might also be a response to elicitors carried in inocula or on spore surfaces. Our data showed clearly that elicitation of a defense response requires a live pathogen, because spore-free inocula and heat-killed spores failed to elicit a hypersensitive response or PR1 expression. This observation is interesting in light of the fact that defense transcripts such as PR1 are seen within a few hours postinoculation. If these early responses were due to pre-formed elicitors on spore surfaces, then killed spores should induce defenses equally well. These results are most consistent with some production of elicitors by spores prior to germination.

Typically, pathogenic attack by fungi activates a large number of genes, suggesting that a successful defense response needs synergistic actions of many defense proteins. PR1 is often induced along with other defense genes. For instance, in Arabidopsis, acidic PR1, PR2, and PR5 and basic PR3 are rapidly induced by turnip crinkle virus infection (Dempsey et al. 1993). Defensin, an antifungal protein found in crucifer seeds (Terras et al. 1995), is coordinately induced with PR1 by Alternaria brassicicola but not by chemical (SA and 2,6-dichloroisonicotinic acid) treatments in Arabidopsis (Penninckx et al. 1996). In B. napus, we have found that the induction pattern of PR2 is similar to that of PR1. However, we could not detect constitutive or induced expression of PR5. For defensin, only weak constitutive expression but no induced expression were detected. This suggests that several distinct regulatory pathways function in Brassica-L. maculans interactions. Even when PR1 alone is considered, regulation of PR1 was distinct among different infection phenotypes. The strongest expression was seen with the more aggressive pathogenicity groups PG2, PG3 and PG4, and weaker with non-aggressive PR1. The strongest response of PR1 to agressive isolates occurred in single-gene resistance to PG2. A similar distinction between pathogenicity groups was observed in B. napus cv. Quinta by Roussel et al. (1999). In Quinta plants expressing single-gene resistance to PG3, toluidine blue staining of vessels was seen as early as 6 h.p.I., in contrast to 48 h.p.I seen with compatible isolate PG4.

Systemic acquired resistance (SAR) is an important resistance mechanism in some plants. An indicator of SAR is induction of defense genes by SA. For example, in Arabidopsis and tobacco PR1 is strongly induced by SA (Uknes et al., 1992; Yalpani et al., 1993). In B. napus PR1 is moderately induced by SA in B. napus, implying that SAR may be involved in defense in B. napus. These results agree with previous observations that the sizes of lesions produced by highly virulent blackleg isolates could be reduced by pre-inoculation of weakly virulent isolates (Mahuku et al. 1996). In addition, it was reported that SA could induce the accumulation of glucosinolates in B. napus leaves (Kiddle et al. 1994). However, recently it was suggested that SA is also involved in gene-for-gene resistance and modulation of disease severity (Bent 1996; Delaney 1997). It is therefore possible that the responsiveness of PR1 to SA treatment might be part of a gene-for-gene resistance response.

We have also assayed PR1 expression in the contexts of non-pathogenic stimuli or developmental control. No induction of PR1 by wounding and heat shock was detected. Only negligible levels of constitutive expression of PR1 have been detected in various healthy parts, such as cotyledons, young and mature leaves, flowers, and young siliques. These data suggest that PR1 is primarily inducible by pathogenic or related stimuli.
 

ACKNOWLEDGMENTS

This work was supported by the Western Grains Research Foundation and the NSERC/AAFC Research Partnership Program. Arabidopsis plasmids pBSPR1, pBSAPR2 and pBSPR5 were provided by Dr. John Ryals (Ciba-Geigy Corp.) and radish defensin AFP cDNA plasmid pFRG1 by Dr. W. Broekaert (F. A. Janssens Laboratory of Genetics, Katholieke Universiteit Leuven, Belgium). J. Tu was supported by the Canadian International Development Agency.