Quantitative resistance to symptomless growth of Leptosphaeria maculans (phoma stem canker) in Brassica napus (oilseed rape)

Authors


*E-mail: yong-ju.huang@bbsrc.ac.uk

Abstract

Quantitative resistance to Leptosphaeria maculans in Brassica napus was investigated in field and controlled environments using cultivars Darmor (with quantitative resistance) and Eurol (without quantitative resistance). In field experiments, numbers of phoma leaf spot lesions in autumn/winter and severity of stem canker the following summer were assessed in three growing seasons. There were no differences between Darmor and Eurol in number of leaf lesions in autumn/winter. However, stem cankers were less severe on Darmor than Eurol at harvest the following summer. In controlled-environment experiments, development of leaf lesions at different temperatures (5–25°C) and wetness durations (12–72 h) was investigated using ascospore inoculum; symptomless growth of L. maculans along leaf petioles towards the stem was quantified using quantitative PCR and visualized using GFP-expressing L. maculans; growth of L. maculans within stem tissues was investigated using GFP-expressing L. maculans. There were more leaf lesions on Darmor than Eurol, although there was no difference between Darmor and Eurol in L. maculans incubation period. There were no differences between Darmor and Eurol in either distance grown by L. maculans along leaf petioles towards the stem or quantity of L. maculans DNA in leaf petioles, but L. maculans colonized stem tissues less extensively on Darmor than Eurol. It was concluded that quantitative resistance to L. maculans operates during colonization of B. napus stems by the pathogen.

Introduction

Many important fungal plant diseases have a symptomless phase between infection and appearance of symptoms/production of new inoculum (Shankar et al., 1996; Fitt et al., 1998; Chartrain et al., 2004; Xu et al., 2005). It is often difficult to control pathogens with long symptomless phases in their life cycles because it is difficult to breed cultivars with resistance to them or to correctly time the application of fungicides. For effective control of plant pathogens with a long symptomless phase, fungicides may need to be applied before development of symptoms on crops. For example, in the UK Pyrenopeziza brassicae (light leaf spot) infects winter oilseed rape (B. napus) leaves in autumn, 3–5 months before visible symptoms appear; to control epidemics, a fungicide spray must be applied in autumn when no symptoms have appeared on leaves (Gilles et al., 2000). To improve control of many fungal plant diseases with a symptomless phase, whether by breeding disease-resistant cultivars or by use of fungicides, there is a need to understand the symptomless phase.

The pathogen Leptosphaeria maculans, cause of phoma stem canker in Brassica napus (oilseed rape, canola, colza), has two periods of symptomless growth. The first symptomless period (5–15 days in winter oilseed rape in Europe) occurs in leaves after the penetration of stomata by hyphae (produced from airborne ascospores) before the appearance of leaf lesions (Biddulph et al., 1999; Toscano-Underwood et al., 2001; Huang et al., 2003). The second symptomless period (5–6 months) occurs between appearance of leaf lesions and appearance of cankers on stems (Hammond et al., 1985; West et al., 1999; Huang et al., 2005). Phoma stem canker is economically the most important disease of oilseed rape worldwide and causes annual yield losses worth more than $900 m (Fitt et al., 2008). Globally, L. maculans has spread to new areas over the last 30 years and it now threatens oilseed rape production in China. Furthermore, it is predicted that the range and severity of phoma stem canker epidemics will continue to increase under climate change scenarios (Evans et al., 2008). Therefore, it is essential to exploit host resistance to L. maculans, which has been the most economical and effective method to control the disease (Delourme et al., 2006).

Two types of resistance to L. maculans have been identified in B. napus. The first type is a qualitative resistance operating in cotyledons and leaves during the first symptomless phase immediately after penetration of leaves by hyphae from the ascospores (Ansan-Melayah et al., 1998; Balesdent et al., 2001). Qualitative resistance to L. maculans is considered as single-gene race-specific complete resistance that is effective in protecting plants only if the corresponding avirulent allele is predominant in the local L. maculans population (Balesdent et al., 2001; Rouxel et al., 2003). Qualitative resistance often loses its effectiveness within three growing seasons of widespread use in commercial cultivars because L. maculans populations change to render ineffective the plant resistance gene (Rouxel et al., 2003; Sprague et al., 2006). Furthermore, effectiveness of such major gene-mediated qualitative resistance is often influenced by environmental factors such as temperature (Huang et al., 2006a). With recent reports of loss of effectiveness of major gene resistance in B. napus in France and Australia (Li et al., 2003; Rouxel et al., 2003; Sprague et al., 2006), breeding cultivars with durable quantitative resistance has become a priority.

The second type of resistance is a quantitative resistance operating during symptomless growth between initial leaf infection and the formation of stem cankers (Pilet et al., 1998; Delourme et al., 2006). By contrast with qualitative resistance, quantitative resistance (often mediated by many genes) is partial and is considered to be race-non-specific and more durable than qualitative resistance (Boyd, 2006; Delourme et al., 2006). However, compared with qualitative resistance, little is known about the operation of quantitative resistance to L. maculans. Since it mainly operates during the long period of symptomless growth between appearance of leaf lesions and appearance of cankers on stems, selecting cultivars for quantitative resistance currently relies on field experiments in which stem canker is assessed before harvest (Pilet et al., 1998; Fitt et al., 2006). This reliance on assessments made at the end of the growing season has made it difficult to investigate the operation of quantitative resistance to L. maculans. Whereas seedlings can be screened for qualitative resistance in cotyledon tests (Balesdent et al., 2001), it is not clear whether it is possible to screen for quantitative resistance during the first period of symptomless growth at the seedling stage.

The availability of green fluorescent protein (GFP)-transformed isolates of L. maculans (Eckert et al., 2005; Huang et al., 2006a) now provides an opportunity to study this symptomless growth to improve understanding of quantitative resistance to L. maculans. Using GFP-transformed isolates now makes it possible to visualize the growth of L. maculans in oilseed rape tissues when there are no symptoms. Furthermore, the development of quantitative PCR (qPCR) techniques to quantify pathogen biomass in asymptomatic infections of several fungal pathogens (Fraaije et al., 2002) provides another method to assess the operation of quantitative resistance. It was suggested that the qPCR method might be used to assess quantitative resistance of winter oilseed rape cultivars by autumn quantification of symptomless L. maculans growth in leaf petioles as an alternative to pre-harvest stem canker assessments the following summer (Kenyon et al., 2004). Recently, qPCR was used to assess quantitative resistance to Stagonospora nodorum (leaf and glume blotch of wheat) in wheat by measuring the fungal biomass in the flag leaf (Oliver et al., 2008). These tools provide a specific, sensitive means to investigate operation of quantitative resistance to L. maculans during the symptomless growth of L. maculans.

The long period of symptomless growth of L. maculans after appearance of phoma leaf spots may be considered in two parts: that in leaf petioles before the pathogen reaches the stem and that in stem tissues before the appearance of stem canker symptoms (West et al., 1999; Fitt et al., 2006). It is not clear when quantitative resistance to L. maculans operates during these periods of symptomless growth. Evidence obtained during construction and validation of a weather-based model to describe development of stem canker epidemics, using data from more than 60 winter oilseed rape experiments over a 15-year period, suggested that components of quantitative resistance operate between first appearance of leaf lesions in autumn and development of stem cankers in spring/summer (Evans et al., 2008). However, that study did not provide direct experimental evidence about the operation of such resistance. This paper describes a series of experiments, using GFP-expressing L. maculans and qPCR and two cultivars differing in quantitative resistance to L. maculans, to investigate the operation of quantitative resistance to L. maculans during symptomless pathogen growth in leaf lamina, leaf petiole and stem tissues.

Materials and methods

Eight experiments were done in controlled environments and three experiments in field conditions, using winter oilseed rape cvs Darmor (with quantitative resistance) and Eurol (without quantitative resistance) (Pilet et al., 1998) to investigate the relationship between quantitative resistance and symptomless growth of L. maculans in oilseed rape leaves, petioles and stems (Table 1). Although both cultivars have major genes for qualitative resistance against L. maculans (Darmor has Rlm9; Eurol has Rlm2 and Rlm3), these resistance genes are no longer effective, since the frequencies of the virulent alleles avrLm2, avrLm3 and avrLm9 in L. maculans populations in Europe, including Rothamsted, are 100% (Stachowiak et al., 2006). Ascospores from natural populations of L. maculans at Rothamsted or conidia of GFP-expressing L. maculans isolate ME24/3·13 (Eckert et al., 2005; Huang et al., 2006a) (carrying avrLm2, avrLm3 and avrLm9) were used as inoculum for these experiments.

Table 1.  Methods used in eight controlled-environment (CE) experiments and three field experiments to examine phoma leaf spotting and stem canker development on oilseed rape (Brassica napus). Cultivars Darmor (with quantitative resistance) and Eurol (without quantitative resistance) were inoculated with ascospores or conidia (GFP-transformed isolate) of Leptosphaeria maculans, under a range of temperatures and wetness durations, and different disease assessment methods were used
Exp.InoculumaPlant grown inTemperature (°C) (Wetness (h))bAssessment method
  • a

    Leptosphaeria maculans ascospores obtained from naturally infected oilseed rape stem base debris collected in August 2004 at Rothamsted; GFP conidia were produced by GFP-transformed isolate ME24/3·13.

  • b

    Experiments CE1–3 and 6–8 were done in growth cabinets; CE4–5 were done in a growth room. All CE conditions were 12 h light/12 h darkness.

  • c

    Experiments were done on the Rothamsted farm, UK in the 2003/04, 2004/05 and 2005/06 growing seasons.

CE1–3AscosporesPot5, 10, 15, 20, 25 (12, 24, 48, 72)Leaf lesions counted
CE4–5GFP conidiaPot18 (72)Distance grown along leaf petiole (GFP)
CE6–7GFP conidiaMagenta vessel15, 25 (72)Distance grown along leaf petiole and canker development in stem (GFP)
CE8AscosporesPot20 (48)L. maculans DNA in leaf petiole (qPCR)
Field 2003–2006cAscosporesFieldNatural conditionsLeaf lesion and stem canker scores

Field experiments

Winter oilseed rape cvs Darmor and Eurol were sown in late August in field plots (14 × 3 m) arranged in a randomized block design as part of larger experiments at Rothamsted in the 2003/04 and 2004/05 (three replicates) and 2005/06 (four replicates) growing seasons. These plots did not receive any fungicide treatments. Ten plants of each cultivar from each plot were randomly selected and labelled using plastic tags. Labelled plants were assessed for phoma leaf spot lesions weekly from autumn until spring, by counting numbers of phoma leaf spot lesions on each plant. The severity of stem base canker on labelled plants was assessed every 2 weeks, from late spring until harvest, on a 0–4 scale, based on external symptoms [0 = healthy; 1 = 1–49% stem circumference girdled; 2 = 50–74% girdled, stem firm; 3 = 75–100% girdled, stem weak; 4 = 100% girdled, plant dead or lodged; Zhou et al. (1999); Huang et al. (2005)]. Just before harvest (in July), stems of these marked plants were cut at the crown (stem base) to assess the internal severity of stem canker by estimating the percentage of stem cross-sectional area with necrosis.

Controlled-environment experiments with ascospore inoculum

To investigate whether quantitative resistance in L. maculans can be detected during the first period of symptomless growth, three controlled-environment (CE) experiments were done using ascospore inoculum (CE1–3, Table 1). Plants of cvs Darmor or Eurol were grown in pots (5 cm diameter) containing peat-based compost and a soluble fertilizer. Plants were initially grown in a glasshouse (20–23°C) with one plant per pot and then placed in seed trays (37 × 23 cm) with 14 plants (seven plants each of Darmor and Eurol) per tray. Three weeks after sowing, when each plant had three expanded leaves, the trays were transferred to a growth cabinet (12 h light/12 h darkness, light density 210 µe m−2 s−1) at 15°C for 16 h before inoculation.

Plants were inoculated with L. maculans ascospores using an ‘ascospore shower’ method (Huang et al., 2006b). Stem base debris with mature L. maculans pseudothecia was cut into pieces 2–3 cm long that were mixed thoroughly. Six stem pieces were chosen at random and evenly attached to the underside of a tray lid (37 × 23 × 14 cm) with Vaseline (Chesebrough-Pond's Ltd). In total, 20 tray lids (for five temperature × four wetness duration treatments) were prepared. The pieces of stem were sprayed with distilled water until run-off to induce release of ascospores and the lids with attached pseudothecia were placed over the trays with plants to allow ascospores to be naturally deposited onto the leaves from 10 cm above the plants, at 15°C. After 2 h, the pieces of debris with pseudothecia were removed from the tray lids.

The inoculated plants were sprayed with distilled water using a laboratory sprayer and covered immediately with tray lids for wetness-duration treatments. Four trays chosen at random were immediately placed in each growth cabinet at 5, 10, 15, 20 or 25°C, respectively. Trays (wetness treatments) were arranged in a completely randomized design in each growth cabinet (temperature treatment). Plants received wetness periods of 12, 24, 48 or 72 h at each temperature, after which the tray lids were removed. One tray (seven plants of each cultivar) was used for each wetness treatment. The number of phoma leaf spot lesions was counted daily on each plant until no new lesions appeared or the leaf senesced. At each temperature, the time from inoculation until the first lesion was recorded. The incubation period was determined as the time from inoculation to the appearance of the first lesion (IP_first) or the time from inoculation to the appearance of 50% of the maximum number of lesions (IP_50).

Controlled-environment experiments with GFP-expressing conidial inoculum

To investigate if quantitative resistance operates during symptomless growth of L. maculans from leaf lesions along leaf petioles towards the stems or during initial colonization of the stems, controlled-environment experiments were done with GFP-expressing L. maculans. A conidial suspension of GFP-expressing L. maculans isolate ME24/3·13 was prepared from a 12-day-old culture on V8 agar and the concentration of conidia adjusted to 107 mL−1.

In experiments CE4–5 (Table 1), plants of cvs Darmor and Eurol were grown in 9-cm-diameter pots in a growth room at 18°C (12 h light/12 h darkness). Plants were inoculated when each plant had three fully expanded leaves. For inoculation, the lower part of the leaf lamina was wounded using a sterile pin and a 10-µL droplet of conidial suspension (107 conidia mL−1) was placed on each wound. Each leaf had two inoculation sites, one on each side of the main vein. After inoculation, plants were covered with tray covers to maintain high humidity for 72 h. In experiment CE4, the first three leaves of each plant were inoculated and 16 plants of each cultivar were used. In experiment CE5, the second and third leaves of each plant were inoculated and 25 plants of each cultivar were used. Pots were arranged in a completely randomized design.

At 22 days post-inoculation (dpi), the inoculated leaves were detached at the point where the petiole joined the stem. The petiole length of each leaf was measured, then leaves were viewed using a Leica MZ FLIII stereomicroscope. For observation of GFP fluorescence, filter GFP2 from Leica Microsystems was used. The distance from the inoculation site on the leaf lamina to the leading GFP L. maculans hyphal tip in the leaf petiole was measured on each leaf. Photomicrographs were taken with a digital camera (Leica DC 300FX) operated with IM50 software (Leica DC Twain, v. 4·1·5·0).

To investigate whether differences in distance grown by L. maculans along the leaf petiole between cultivars were similar under different growing conditions, two experiments (CE6–7) (Table 1) were done in 15 and 25°C growth cabinets (12 h light/12 h darkness) with plants grown in containers, consisting of pairs of Magenta vessels (each 77 × 77 × 97 mm) joined with a Magenta coupler (one plant per container). The plants were kept in a 15°C growth cabinet until each plant had three fully expanded leaves. The first three leaves of each plant were inoculated as described for CE4–5. After inoculation, 14 plants of each cultivar were kept in 15 or 25°C growth cabinets. In each cabinet, seven plants of each cultivar were randomly selected to assess symptomless growth of L. maculans in petioles and the remaining seven plants were used to assess growth in stem tissues. At 15 dpi (25°C) or 20 dpi (15°C), the inoculated leaves (in total 21 leaves from each cultivar at each temperature) were detached. The distance along the leaf petiole from the inoculation site to the leading GFP-L. maculans hyphal tip was measured on each leaf. At 47 dpi, stems of the remaining seven plants at each temperature were harvested and the internal stem canker severity on each stem was assessed as percentage of stem cross-sectional area with necrosis. The stem cross-sections were also viewed to assess the extent of colonization of different stem tissues by the GFP-L. maculans hyphae (i.e. presence of GFP-L. maculans in the outer cortex or central pith tissues).

Controlled-environment experiment to quantify symptomless L. maculans growth in leaf petioles

To investigate whether quantitative resistance in B. napus can be detected during the symptomless growth of L. maculans in the leaf petiole by quantification of the pathogen DNA using quantitative PCR (qPCR), an experiment (CE8, Table 1) was done with Darmor and Eurol. Plants were grown in 9-cm-diameter pots in a 20°C growth cabinet. For each plant, the lower parts of the first three leaves were inoculated at two sites, one on each side of the main vein. Each site was inoculated with a 15-µL drop of ascospore suspension (103 ascospores mL−1) without wounding the leaf. There were three replicates arranged in a randomized block design with five plants in each replicate (15 inoculated leaves in each replicate for each cultivar). Plants were kept at high humidity for 48 h. Inoculated leaves were detached at 18 dpi. Petiole length was measured on each leaf, then the leaf petiole was cut and placed in a 15-mL tube to be freeze-dried. Freeze-dried individual leaf petioles were ground into powder using a mortar and pestle.

DNA was extracted from a 20-mg sub-sample (from each ground individual leaf petiole) using a DNA extraction kit (DNAMITE Plant Kit, Microzone Ltd) and quantified on a Nanodrop ND-1000 spectrophotometer (Labtech International, UK). The amount of L. maculans DNA was quantified using a SYBR green quantitative PCR with the primers LmacA and LmacRev (Liu et al., 2006). A Stratagene Mx3000P Real Time PCR machine was used. Each reaction volume was 20 µL, including 0·6 µL of primers at a final concentration of 300 nm, 10 µL of SYBR Green JumpStart Tag ReadyMix (Sigma), 0·08 µL of ROX internal reference dye and 2·5 µL of DNA sample. There were two replicates for each sample. To produce a standard curve, a total of five 10-fold dilutions (ranging from 104 pg µL−1 to 1 pg µL−1) of DNA from a pure culture of L. maculans were run, with two replicates of each dilution on each plate. The amount of L. maculans DNA in each sample was determined by comparing the data to the standard curve. The thermocycling profile consisted of an initial cycle of 95°C for 2 min, followed by 40 cycles of 95°C for 15 s, 60°C for 30 s, 72°C for 45 s and a read step at 83°C for 15 s, then a dissociation stage (thermal profile: 95°C 1 min, 60°C 1 min, 95°C 15 s).

Statistical analysis

Field experiment data for the number of phoma leaf spots, and external and internal stem canker severity were analysed to compare the differences between cvs Darmor and Eurol by analysis of variance. Since external stem canker severity data were scores, these data were also logit-transformed before analysis of variance. For data from controlled-environment experiments with ascospore inoculum, analyses of variance were done to assess the effects of temperature and leaf wetness duration on incubation period and number of lesions on leaves of cvs Darmor and Eurol. Linear regressions of incubation period (IP_first or IP_50) against temperature were calculated separately for cvs Darmor and Eurol. Analyses of position and parallelism were done to assess whether the data were fitted best by single straight lines, pairs of parallel lines or pairs of non-parallel lines for Darmor and Eurol. In experiments with GFP-expressing conidial inoculum, the differences between Darmor and Eurol in the distances grown by L. maculans along the leaf petiole towards the stem and for stem canker severity were compared by analysis of variance. In experiments with qPCR, analyses of variance were done to compare cvs Darmor and Eurol for the leaf petiole length and the amounts of L. maculans DNA in the leaf petiole. All the analyses were done using genstat statistical software (Payne et al., 2007).

Results

Field experiments

In 2003/04, there was no difference between cvs Darmor and Eurol in number of phoma leaf spot lesions in autumn/winter (Fig. 1a). There was also no difference between Darmor and Eurol in number of leaf spot lesions in 2004/05 (Fig. 1b) or 2005/06 (Fig. 1c). However, before harvest in all three seasons, external stem canker severities were greater on Eurol than on Darmor (Fig. 1): 0·5 for Darmor and 1·7 for Eurol in 2003/04; 1·5 for Darmor and 2·4 for Eurol in 2004/05; 1·6 for Darmor and 2·2 for Eurol in 2005/06. Analysis of logit-tansformed data (not presented) confirmed these differences. For the internal stem canker severities, the percentages of stem cross-sectional area with necrosis were greater on Eurol than on Darmor in all three seasons (Fig. 2). In 2003/04, when the phoma leaf spot epidemic started in December (which was later than in the other two seasons), the subsequent stem canker epidemic was less severe than those in 2004/05 and 2005/06.

Figure 1.

Cumulative numbers of phoma leaf spot lesions per plant and external stem canker severity scores on winter oilseed rape cultivars Darmor (inline image, with quantitative resistance) and Eurol (inline image, without quantitative resistance) in the 2003/04 (a), 2004/05 (b) and 2005/06 (c) growing seasons at Rothamsted. Phoma leaf spot lesions were assessed weekly from appearance of the first lesion in autumn until spring and stem canker was assessed every 2 weeks from April until harvest (inline image, Darmor; inline image, Eurol) on a 0–4 scale (external) on 30 or 40 marked plants per cultivar. Vertical bars represent standard errors of differences (degrees of freedom for leaf lesions and external stem canker severity, respectively: 4 and 8 for 2003/04, 3 and 34 for 2004/05, 8 and 23 for 2005/06).

Figure 2.

Internal stem canker severity of winter oilseed rape cultivars Darmor (with quantitative resistance) and Eurol (without quantitative resistance) at Rothamsted in the 2003/04, 2004/05 and 2005/06 growing seasons under natural infection conditions. At harvest, 30 or 40 plants were assessed for internal canker symptoms by taking a cross-section of the crown (stem base) and scoring for percentage of stem cross-sectional area with necrosis. Vertical bars represent standard errors of differences (degrees of freedom: 47 for 2003/04, 52 for 2004/05, 75 for 2005/06).

Controlled-environment experiments with ascospore inoculum

In experiments CE1–3, there was no difference in the visual appearance of phoma leaf spot symptoms and there was no visible difference in size of lesions between Darmor and Eurol at temperatures of 5–25°C. On both Darmor and Eurol for all temperature and wetness-duration treatments, typical phoma leaf lesions developed. There was no difference between wetness treatments in incubation period IP_first or IP_50. However, there were differences in IP_first (P < 0·001; SED = 0·14; 8 d.f.) and IP_50 between temperature treatments. IP_first decreased from 18 days at 5°C to 6 days at 25°C and IP_50 decreased from 21 days at 5°C to 7 days at 25°C for both Darmor and Eurol. However, there was no difference between Darmor and Eurol in either IP_first or IP_50. The relationships between incubation period (f) or 1/f and temperature (T) were fitted best by single lines for both Darmor and Eurol (Fig. 3). For example, the relationship between incubation period (f) and temperature (T) for IP_first was best described by the function f = 19·3 –0·62T (R2 = 0·85).

Figure 3.

Effects of temperature on incubation period (a, b) and inverse of incubation period (c, d) of Leptosphaeria maculans in controlled-environment experiments. Incubation period was estimated as time from inoculation with L. maculans ascospores to the appearance of first lesions (a) or 50% of lesions (b) on leaves of oilseed rape cvs Darmor (inline image, with quantitative resistance) or Eurol (inline image, without quantitative resistance). The relationships between incubation period (f) and temperature (T) were fitted best by single straight lines for time to first lesions (a, f = 19·3 – 0·62T, R2 = 0·85) and time to 50% of lesions (b, f = 22·6 – 0·71 T, R2 = 0·89) for Darmor and Eurol. The relationships between 1/f and T were best fitted by single straight lines for first lesions (c, 1/f = 0·027 + 0·006T, R2 = 0·95) and 50% of lesions (d, 1/f = 0·021 + 0·005T, R2 = 0·98) for Darmor and Eurol. Data presented are means of four wetness treatments in three experiments (CE 1–3).

Temperature (P < 0·01; SED = 4·6; 8 d.f.) and wetness duration (P < 0·001; SED = 4·1; 30 d.f.) both affected the maximum number of phoma leaf spot lesions that developed on leaves of Darmor and Eurol. There was no interaction between temperature and wetness duration. On average, more lesions developed on Darmor than on Eurol (P < 0·001; SED = 4·9; 40 d.f.). On Darmor, more lesions developed at 15 and 20°C than at 5, 10 and 25°C with wetness durations of 48 and 72 h (Fig. 4). For both Darmor and Eurol, the number of lesions decreased as leaf wetness duration decreased from 72 to 12 h; very few lesions developed with a wetness duration of 12 h at 5 or 10°C (Fig. 4).

Figure 4.

Effects of temperature and wetness duration on maximum number of leaf spot lesions on leaves of oilseed rape cvs Darmor (D, with quantitative resistance) or Eurol (E, without quantitative resistance). Plants were inoculated with ascospores of Leptosphaeria maculans with wetness durations of 12, 24, 48 or 72 h. Data presented are means of three experiments (CE 1–3) (standard error of difference 12·82, 75 degrees of freedom).

Controlled-environment experiments with GFP-expressing conidial inoculum

In experiment CE4, there was no significant difference in distance grown by L. maculans along the leaf petioles towards the stems between Darmor and Eurol. There was no difference in leaf petiole length between Darmor and Eurol. In CE5, there were also no differences in leaf petiole length or distance grown by L. maculans along the leaf petiole between Darmor and Eurol.

In experiments CE6–7, hyphal growth of a GFP-expressing isolate along the leaf petiole towards the stem and spread within the stem was observed in both Darmor and Eurol (Fig. 5). There was no significant difference between Darmor and Eurol in the distance grown by L. maculans along the leaf petiole at 15°C or 25°C (Fig. 6a). At 47 dpi, stem cankers developed on stems of both Darmor and Eurol at positions of leaf scars of the inoculated leaves (Fig. 5d,e). After GFP-L. maculans reached the stem, it spread both upwards along the stem and downwards to the tap root from the positions of the leaf scars of the inoculated leaves (Fig. 5d,e). At 15°C, the internal severity of stem canker on Eurol (38·7% stem cross-sectional area with necrosis) was greater than on Darmor (6·3% stem cross-sectional area with necrosis), while at 25°C there was no significant difference in internal severity of stem canker between Darmor and Eurol (Fig. 6b). At 15°C, the growth of L. maculans in Darmor was restricted mainly to the outer stem cortex, while in Eurol it had often spread to colonize the central stem pith (Fig. 5f,g). Leptosphaeria maculans spread to the stem pith on 75% of Eurol plants compared with 33% of Darmor plants.

Figure 5.

Growth of Leptosphaeria maculans from phoma leaf spot lesions along leaf petioles of oilseed rape into the stems to cause stem canker in controlled-environment experiments at 15°C (CE 6–7). Leaf (a, b) of cv. Eurol inoculated with conidia of L. maculans isolate ME24/3·13 expressing GFP at 16 dpi (days post-inoculation), viewed with brightfield illumination (a) or a GFP2 filter (b); leaf petiole of Eurol (c) at 20 dpi, viewed with brightfield illumination (c-1) or a GFP2 filter (c-2); stems of cvs Darmor (d) and Eurol (e) at 47 dpi, viewed with brightfield illumination (d-1, e-1) or a GFP2 filter (d-2, e-2) (red arrows: leaf scars of the inoculated leaves; white arrows: direction of spread of L. maculans after reaching the stem); stem cross-sections of Darmor (f) and Eurol (g) at 47 dpi, viewed with brightfield illumination (f-1, g-1) or a GFP2 filter (f-2, g-2). Bars, 200 µm (a, b), 1·5 mm (c) or 2 mm (d, e, f, g).

Figure 6.

(a) Distance grown by Leptosphaeria maculans from sites of inoculation with conidia of GFP L. maculans along leaf petioles towards the stems in controlled-environment experiments (CE 6–7) at 15 and 25°C on oilseed rape cvs Darmor (inline image, with quantitative resistance) or Eurol (inline image, without quantitative resistance) at 20 days post-inoculation (dpi) (15°C) or 15 dpi (25°C); and (b) internal stem canker severity at 47 dpi scored for percentage of stem cross-sectional area with necrosis. Vertical bars represent standard errors of differences (degrees of freedom: 12 for growth in petiole and 21 for internal stem canker severity).

Controlled-environment experiment to quantify symptomless L. maculans growth in leaf petioles

In experiment CE8 with Darmor and Eurol grown in pots at 20°C and inoculated with L. maculans ascospores, there was no difference in leaf petiole length between Darmor and Eurol. There was also no significant difference in amount of L. maculans DNA in the leaf petiole between Darmor and Eurol.

Discussion

This work provides the first direct experimental evidence that quantitative resistance to L. maculans operates during colonization of B. napus stem tissues by the pathogen. Controlled-environment experiments suggested that this resistance does not operate during the development of phoma leaf spots, since they demonstrated that more phoma leaf lesions developed on Darmor (with quantitative resistance) than on Eurol (without quantitative resistance) over a range of temperature/wetness regimes. There were no differences in incubation period between Darmor and Eurol, estimated as either the time from inoculation to the appearance of the first lesions or to the appearance of 50% of lesions, suggesting that quantitative resistance to L. maculans does not operate during the first short period of symptomless growth in the leaf lamina. Investigation of the second period of symptomless growth in leaf petioles with GFP-expressing L. maculans showed that there were no differences between Darmor and Eurol in distance grown by L. maculans along the leaf petiole or the amount of L. maculans DNA detected in the petiole. However, differences were observed between Darmor and Eurol during the growth of L. maculans in stem tissues. In controlled-environment experiments at 15°C, GFP-expressing L. maculans often spread into the central pith of Eurol stems, but was confined to the outer cortex of Darmor stems (Fig. 5f,g). Differences between Darmor and Eurol in the extent of colonization of stem tissues by L. maculans suggested that quantitative resistance to L. maculans may operate by impeding the growth of L. maculans within the stem tissues. This containment of the pathogen in the stem may be associated with a more rapid lignification of the pith cells of the resistant cultivars (Hammond & Lewis, 1987). Quantitative resistance to other plant pathogens is often expressed by reducing the growth of the pathogen within the host. Quantitative resistance to Mycosphaerella graminicola (septoria tritici blotch) in wheat operates by reducing the leaf area covered by lesions with pycnidia (Chartrain et al., 2004), while quantitative resistance to Magnaporthe grisea (rice blast) operates by reducing both the number and size of leaf lesions (Talukder et al., 2004). For both these pathogens, yield loss is associated with damage to the leaves, whereas for L. maculans it is associated with damage to the stems. Therefore, breeding programmes selecting lines that suffer less yield loss have selected for quantitative resistance operating in different host tissues for different pathogens.

The field experiments provided indirect evidence that B. napus quantitative resistance to L. maculans operates during colonization of stem tissues by the pathogen. In these experiments, there were consistently less severe stem cankers on Darmor than on Eurol in summer at harvest in all three growing seasons, although there were no differences between Darmor and Eurol in the number of phoma leaf lesions in the previous autumn/winter. This suggests that quantitative resistance to L. maculans operates after the leaf spot stage by decreasing the subsequent symptomless growth of L. maculans in stem tissues. These results were supported by results from controlled-environment experiments in which more severe stem cankers were observed on Eurol than on Darmor, although there was no difference in distance grown by symptomless L. maculans in their leaf petioles. This conclusion is also supported by indirect evidence from 60 winter oilseed rape experiments that showed no evidence of differences between resistant and susceptible cultivars in timing of phoma leaf spotting, but differences between them in the timing of stem canker development and severity of stem canker (Evans et al., 2008). However, it is not clear what is the crucial period for operation of resistance against the spread of L. maculans within stem tissues. In studies of root colonization by L. maculans, the spread of the pathogen into root tissue was observed after onset of flowering (Sprague et al., 2007). Further detailed time-course investigations of the development of stem canker after the pathogen has reached the stem are needed.

The direct evidence that quantitative resistance operates by preventing or restricting the spread of L. maculans within stem tissues was produced using a GFP-expressing isolate. The use of GFP as a reporter gene enabled the visualization of symptomless growth of L. maculans, firstly from leaf lesions along leaf petioles towards the stems and secondly in stem tissues. The use of GFP-L. maculans also showed that the pathogen could spread from the leaf scar of the inoculated leaf both up the stem and down towards the tap root, confirming observations reported in susceptible spring oilseed rape (Sprague et al., 2007). In controlled environments, although more severe stem cankers developed on Eurol than on Darmor at 15°C, there were no significant differences between Darmor and Eurol in stem canker severity at 25°C. This suggests that temperature may affect the operation of quantitative resistance to L. maculans, as with operation of major gene Rlm6-mediated resistance (Huang et al., 2006a). This fits with recent predictions that the range and severity of phoma stem canker epidemics will increase under global warming associated with climate change (Evans et al., 2008).

The development of quantitative PCR (qPCR) provides a convenient alternative method to investigate symptomless fungal growth. Although GFP-transformed fungal isolates are valuable tools for studying symptomless pathogen growth, their use has limitations because they can be used only under controlled conditions. By contrast, qPCR can be used to quantify symptomless fungal growth in tissues of plants grown either in controlled or field conditions. Furthermore, a large number of samples can be assessed quickly using qPCR. Work is now needed to investigate the use of qPCR to assess the development of stem canker before the appearance of cankers on stems. Use of qPCR could help to accelerate the process of breeding cultivars with quantitative resistance, which currently relies on assessment of stem canker at harvest (Pilet et al., 1998; Delourme et al., 2006; Fitt et al., 2006).

Acknowledgements

We thank the UK Biotechnology and Biological Sciences Research Council (BBSRC, IPA project, BB/E001610/1), the Department for Environment, Food and Rural Affairs (Defra), the European Union (SECURE project, QLRT-2001-01813), DuPont, HGCA, KWS-UK Ltd, the Royal Society and the British Council for supporting the work. We thank Alan Todd and Rodger White for statistical analyses of the data, and Ian Crute, Pietro Spanu, Graham Jellis, Peter Werner and Olu Latunde-Dada for advice. All work involving the use and storage of genetically modified L. maculans was done under Defra licence no. PHL 174E/5443(01/2007).

Ancillary