A detached seedling leaf technique to study resistance to Mycosphaerella graminicola (anamorph Septoria tritici) in wheat
*To whom correspondence should be addressed.†E-mail: firstname.lastname@example.org
A detached seedling leaf technique was developed to screen for resistance to septoria tritici blotch of wheat and to detect specific interactions between cultivars and isolates. Wheat seedlings were inoculated with spore suspensions of Mycosphaerella graminicola. Detached primary leaves were then placed in a clear plastic box such that their cut ends were sandwiched between layers of agar containing benzimidazole, with a gap below the middle of the leaves. Mean levels of disease were affected by light and temperature, and also by the concentration of benzimidazole, such that higher concentrations resulted in less disease. Second leaves were more susceptible than seedling primary leaves. However, none of these factors affected ranking of disease among cultivars or cultivar-by-isolate interactions. Kavkaz–K4500 1.6.a.4, Synthetic 6x and Triticum macha showed specific susceptibility and resistance to different isolates. The detached leaf technique could be a useful complement to field trials and an alternative to whole seedling assays in assessing cultivar resistance and investigating the genetics of the host–pathogen interaction.
Septoria tritici blotch, caused by the ascomycete fungus Mycosphaerella graminicola (anamorph Septoria tritici), is one of the major diseases of wheat worldwide. When severe infections occur, it can cause losses of up to 60% of total yield. There is considerable expenditure on fungicides to control the disease in many countries (Cook, 1999). Compared to other diseases of wheat, there has been little progress in genetic analysis of resistance to septoria tritici blotch. This is partly because, in contrast to work on powdery mildew and rusts, defined isolates of M. graminicola have not been used to identify resistance genes in a repeatable way. Kema et al. (1996a,b) showed that the host–pathogen interactions in septoria tritici blotch involve resistances specifically effective against particular isolates, as well as isolate nonspecific resistances. Successful breeding for durable resistance must take into account these specific interactions to avoid the breakdown of host resistance by specifically virulent isolates. Methods of detecting interactions between wheat and M. graminicola genotypes are therefore required.
Wheat cultivars are usually screened for resistance to M. graminicola in field trials of adult plants. Tests in which seedlings are sprayed with a suspension of spores have also been used (Eyal et al., 1985; Van Ginkel & Scharen, 1988; Kema & Van Silfhout, 1997). These methods require a great deal of time and space. Furthermore, field trials inoculated with single isolates may become contaminated by natural infection, they can only be tested once a year, and results may depend on environmental conditions (Kema & Van Silfhout, 1997). Whole seedling assays require the use of growth cabinets with controllable humidity, light and temperature and may be expensive, even when suitable conditions are available.
Detached seedling leaf tests are used routinely in work with powdery mildew of wheat and barley (Brown & Wolfe, 1990). Such tests were also used for Phaeosphaeria nodorum (anamorph Stagonospora nodorum syn. Septoria nodorum) by Benedikz et al. (1981), who showed that results obtained with this technique had a high correlation with field trial assessments. A disadvantage of these methods is that leaf material is in an artificial situation (Karjalainen, 1984). The methods are considered valuable tools, however, for rapidly evaluating the susceptibility of new wheat cultivars to stagonospora nodorum blotch and powdery mildew.
Previously, Sewell & Caldwell (1960) reported the use of excized leaves to test resistance to S. tritici. More recently, Forni & Zitelli (1979) and Pyzhikova & Karaseva (1985) reported the use of techniques to study sources of resistance. However, Forni & Zitelli (1979) were unable to assess cultivar differences with this test, while Pyzhikova & Karaseva (1985) provided data about the performance of their test for P. nodorum but not for M. graminicola. Benedikz et al. (1981) reported that attempts to apply the technique of using detached seedling leaves to test for resistance to M. graminicola were unsuccessful. Furthermore, in these papers disease was assessed by measuring the length of necrotic lesions and counting the number of lesions. Measuring the coverage of the leaf area by lesions bearing pycnidia shows more pronounced differences between isolates and cultivars than scoring necrotic lesion area, and is therefore a more appropriate method to characterize cultivar resistance and its interaction with M. graminicola isolates (Kema et al., 1996a). A new detached leaf method, more appropriate to M. graminicola, is therefore desirable.
This paper describes a method for using detached seedling leaves to screen for resistance to M. graminicola in wheat cultivars and to detect specific interactions between cultivars and isolates. Experimental conditions required for optimal performance of the test were evaluated and the extent to which the test is robust against environmental variation was assessed. Correlations with results from field trials and whole seedling assays were obtained.
Materials and methods
Plant and fungal material
Experiment series A: environmental conditions
Two M. graminicola isolates, IPO323 and IPO94269 from the Netherlands, obtained from Dr Gert Kema (Plant Research International, the Netherlands) and seven bread wheat lines (Triticum aestivum), previously studied in field trials (Brown et al. 2001), were used to develop the detached leaf method to evaluate resistance to septoria tritici blotch (Table 1).
Table 1. Wheat cultivars and isolates of Mycosphaerella graminicola (IPO323, IPO94269) used in experiment series A with mean levels of septoria tritici blotch in field trials, whole seedling tests and in the standard treatment of detached leaf tests (% of leaf area covered by lesions bearing pycnidia) and correlation coefficients
|Specifically resistant to IPO323|
|Kavkaz–K4500 1·6·a·4 (KK)||0||3||0||1||0||3|
|Correlation coefficients of|
|Field–whole seedling||0·90||0·90|| || || || |
|Field–detached leaf||0·96||0·93|| || || || |
|Whole seedling–detached leaf||0·93||0·85|| || || || |
Experiment series B: further specific interactions
A second set of material, in two parts, was investigated. Part 1 included wheat lines Baldus, Cappelle Desprez, Hobbit sib, Kavkaz–K4500 1.6.a.4 (KK) and Triticum macha (John Innes accession JI 1240001). Part 2 included Baldus, Chinese Spring, Longbow and Synthetic 6x (a synthetic hexaploid wheat: McFadden & Sears, 1946; Sears, 1976). Seven isolates of M. graminicola numbered IPO92001 to IPO92007 from Portugal were used, along with isolates IPO001, IPO290, IPO323, IPO89011, IPO94265 and IPO94269 from the Netherlands.
Detached leaf procedure: standard technique
Several seeds of the cultivars tested were pre-germinated on wet filter paper (Whatman 90 mm, Whatman International Ltd, Hadstone, UK) in darkness at 25°C for 24 h, moved to a refrigerator at 5°C for 48 h, then moved back to the 25°C incubator for another 24 h. Germinated seeds were grown in modified John Innes no. 2 compost in 22 × 17 cm plastic trays in a glasshouse with 16 h of light and a minimum daytime temperature of 15°C until the first or second leaf was fully expanded, usually 12–15 days after germination.
Inoculum was produced from sporulating cultures of M. graminicola, grown on potato dextrose agar (PDA) for 7 days under near-ultraviolet (NUV) light (Philips TL 20 W/05, Edmundson’s Electrical, Norwich, UK) for 16 h per day at 15°C. Cultures were flooded with sterile distilled water and scraped to release conidia. The concentration of the conidial suspension was adjusted to 107 spores mL−1 and polyoxyethylene-sorbitan monolaurate (Tween 20; Sigma-Aldrich Chemie, GmbH, Germany) was added to 0·15% v/v. Wheat seedlings were evenly sprayed with spore suspension to run-off (20–40 mL per tray depending on the number of plants) using a Humbrol paint spray gun kit (Humbrol Ltd, Hull, UK). The leaves were left to dry for 30 min before 3·5 cm sections were cut from the middle of the primary leaves.
Water agar (10 g L−1) containing 100 mg L−1 benzimidazole (Sigma) used to retard senescence, was dispensed in 50 mL aliquots into nonsterile clear polystyrene boxes (8 × 12 × 2 cm). Rectangular sections (3 × 9 cm) were cut from the centre of the agar. The seedling leaf sections were laid, top surface uppermost, across the gap so that the cut ends rested on the agar. The gap below the leaves helped to prevent water soaking and contamination by other microorganisms. Eight to ten leaf sections could be fitted into each box. Strips of agar were then laid over the cut edges of the leaf sections so that they were not exposed, thereby delaying senescence. The boxes were closed and covered with black plastic or foil to keep them in darkness. After incubation at 20°C for 48 h the boxes were uncovered and left under white phosphorescent light (2x Philips TLD 70 W/83) for 12 days at 20°C, then moved to conditions with NUV light at 15°C, 14 days after inoculation to promote sporulation.
The percentage leaf area covered by lesions bearing pycnidia was scored four to five times at intervals of 2–4 days during a period of 19–28 days after inoculation. All assessments were carried out using a dissecting microscope at 40× magnification.
Experiment series A: environmental conditions
To test the limits of the environmental conditions within which the method is reliable, variations of the standard method were examined. Cold treatment, inoculum source, seedling leaf age, temperature and light conditions, benzimidazole concentration, and concentration of the conidial suspension were studied. Cold treatment in a controlled environment (5°C and 70% relative humidity) was given to 11-day-old seedlings for 0 h or 48 h, 3 days before inoculation, to test an observation that leaves may stay green longer if subjected to cold conditions for a period immediately before inoculation. The source of inoculum experiment compared the use of freshly produced inoculum grown for 7 days on PDA plates re-isolated after passage through wheat leaves with frozen spore suspensions kept at −20°C for several months to determine if stored spore suspensions can be used whenever necessary without loss of pathogenicity or viability. In the study of seedling leaf age, primary and secondary seedling leaves were used to test whether variation in leaf age affected cultivar response to infection. In temperature and light treatments, two temperatures, 15°C (15) and 20°C (20) and two types of light, white fluorescent strip light (wh) and NUV were tested. These conditions were combined in six different treatments for a period of 15 days postinoculation (first term) and from day 16 onwards (second term): 15NUV−15NUV, 15wh−15NUV, 20wh−15NUV, 15wh−15wh, 20wh−15wh and 20wh−20wh. Concentrations of 50, 100, 150 and 200 mg L−1 benzimidazole in the agar were investigated to study their effect on infection and disease development. Four different conidial concentrations were tested, 105, 106, 107 and 108 spores mL−1. These conidial suspension concentrations have all been recommended by different authors as being most appropriate for studies of M. graminicola (Eyal et al. 1987).
Whole seedling assay
The same material described for experiment series A was tested. A modified procedure of the seedling assay described by Kema et al. (1996a) was used. The test was conducted in a glasshouse compartment with 16 h of light and a minimum daytime temperature of 20°C. Disease severity was evaluated as the percentage leaf area covered by lesions bearing pycnidia, four times during a period of 20–34 days after inoculation.
All experiments were conducted in randomized complete blocks in layouts generated with the Experimental Design Generator and Randomiser (EDGAR) (Brown, 1997). Part 1 of experiment series B was repeated four times and part 2 once. On each date, there were four replicates of each plant line per isolate. Each plastic box containing detached leaves represented one block with four to eight leaves of different cultivars inoculated with a single isolate. Four additional control blocks per experiment were mock inoculated with sterile water containing Tween 20 surfactant but no spores.
Disease scores were analysed to compare the effects of factors, including cultivar, isolate, treatment and their two- and three-way interactions. The area under the disease progress curve (AUDPC) (Shaner & Finney, 1977) was calculated as the area under the graph of observed disease level plotted against time, from the first to last scoring. The water-inoculated controls always had AUDPC scores of zero; these data were not included in the statistical analysis. Data were analysed by generalized linear modelling of binomial proportions (Genstat 5 Committee, 1993). The variate analysed was the AUDPC as a proportion of the maximum possible AUDPC for each experiment (i.e. all scores were 100%). The analyses were performed using the statistical package GenstatTM for Windows, 4th Edition (Numerical Algorithms Group, Oxford, UK). In experiment series B, median tetrad analysis was used to identify specific resistance or susceptibility of cultivars to particular isolates (Kroes et al., 1999; Brown et al. 2001). The analysis was done on logit-transformed data.
Experiment series A: environmental conditions
The aim of this set of experiments was to determine whether cultivar resistance and cultivar × isolate (CI) interactions, detected in field trials of adult wheat plants and whole seedling assays, were also expressed in detached seedling leaves. The effects of environmental conditions were also investigated. In all experiments, the main effect of the wheat cultivar and the CI interaction was large and highly significant (Table 2). As in field and whole seedling trials, Andante, Baldus and Longbow were susceptible to both isolates, KK was resistant to both isolates, and Arina, Bezostaya 1 and Flame were specifically resistant to IPO323, which caused less disease on these cultivars than IPO94269 (Table 1).
Table 2. Generalized linear modelling analysis of leaf area affected by septoria tritici blotch for various environmental conditions tested
|Error||61||63·9|| ||79||47·1|| ||131||209·5|| ||521||263·9|| ||133||127·5|| ||140||73·3|| |
In the experiments on cold treatment and inoculum source, the main effect of the treatment was not significant and the treatment did not affect the mean level of cultivar resistance [cultivar treatment interaction (CT) in Table 2]. This implies that there was no advantage in using fresh inoculum over stored frozen inoculum or in giving seedlings cold treatment before inoculation.
In the leaf age, temperature and light, benzimidazole concentration, and spore concentration experiments, there were significant differences between the various treatments. There were also significant CT terms in all these experiments, except for that on leaf age in which there was more disease on secondary leaves (mean 48%) than on primary leaves (mean 33%) (treatment term in Table 2). Leaf age, however, had no significant effect on the CI interaction (cultivar × isolate × treatment interaction (CIT) in Table 2). This implies that there was no need to wait for a second leaf to be fully grown to conduct the test because the same ranking of cultivars and the same cultivar-by-isolate interactions were obtained in both conditions.
In the temperature/light experiment, the largest contributor to the CT effect (P < 0·001) was Baldus, which had a relatively high level of disease severity compared to the other cultivars in the 15wh−15wh and 20wh−20wh conditions and relatively low disease severity in the 15wh−15NUV condition (data not shown). Despite this, the CI interaction was not affected significantly by the temperature and light treatments (CIT term in Table 2). This implies that any of the light and temperature conditions used would give reliable results.
At 50 mg L−1 benzimidazole, resistant cultivars had high levels of disease severity with isolate IPO94269 and infection levels were at least two-fold greater than in the standard 100 mg L−1 treatment. At 150 mg L−1 and 200 mg L−1 benzimidazole, even the susceptible cultivars had low disease severity. This implies that the standard concentration of benzimidazole, 100 mg L−1, is most suitable.
In the experiment on conidial suspension concentration, the main contribution to the CT interaction (Table 2) was from results with 105 spores mL−1, which caused low disease severity even on susceptible cultivars. Spore concentrations of 106 and 108 caused similar levels of disease severity to those at the standard concentration of 107 spores mL−1.
The high correlations between the results obtained in field and whole seedling trials and with the methods described in this paper show that, for isolate IPO323, the detached leaf technique replicates field and seedling trials closely in most conditions (Table 3). Isolate IPO94269 had lower correlation coefficients than IPO323 in all conditions, except the temperature/light factors 15NUV−15NUV and 20wh−15NUV for both field and seedling trials and 20wh−15wh for field trials. For tests with IPO94269, the use of secondary leaves and nonstandard benzimidazole concentrations of 50 and 200 mg L−1 had relatively low correlations with field trial and seedling assay results (Table 3).
Table 3. Correlation coefficients between percentage leaf area covered by lesions bearing pycnidia of Mycosphaerella graminicola in field and seedling trials (Table 1) and in various environmental conditions in which detached leaves were tested for isolates IPO323 and IPO94269
|Cold treatment (h)||0*||0·92||0·84||0·81||0·71|
|Inoculum source||Freshly produced*||0·96||0·95||0·78||0·77|
|Seedling leaf age||Primary leaf*||0·76||0·90||0·50||0·79|
|concentration (mg L−1)||100*||0·99||0·95||0·89||0·91|
|concentration (spores per mL)||106||0·98||0·87||0·72||0·61|
Experiment series B: further specific interactions
The purpose of this experiment was to investigate the value of this technique in identifying specific cultivar-by-isolate interactions involving isolates other than IPO323. In both part 1 and part 2 of this experiment with 13 M. graminicola isolates from the Netherlands and Portugal, the main effects of the cultivar and isolate and the CI interactions were highly significant (Table 4). Several specific resistances could be identified (Table 5). To identify these interactions it was necessary to consider each cultivar response to the complete set of isolates and the general level of aggressiveness of each isolate.
Table 4. Generalized linear modelling analysis of area under the disease progress curve (AUDPC) for septoria tritici blotch in experiment series B, parts 1 and 2
|Error||661||190·7|| ||101||96·7|| |
Table 5. Percentage area of detached leaves of wheat cultivars covered by lesions bearing pycnidia of Mycosphaerella graminicola isolates in experiment series B, parts 1 and 2
|From the Netherlands|
|IPO001||48||25||38||9 (38)||4||36||96||58||68||5 (0)||88|
|IPO323||77||62||55||0 (43)||2 (24)||65||100||27 (100)||99||0||100|
|Variety meane||53||32||37||54||12|| ||98||93||81||0|| |
The lines KK (Kema et al., 1996a,b), T. macha, and Synthetic 6x (L.S. Arraiano et al. unpublished results) have been regarded as potential sources of resistance to M. graminicola. However, their resistance was isolate-specific (Table 5). Cultivars Baldus, Cappelle Desprez and Hobbit sib were susceptible to all 13 isolates tested. Several specific interactions, i.e. resistance or susceptibility of cultivars to particular pathogen isolates, were identified by median tetrad analysis. KK was specifically resistant to isolate IPO323, as in the field (Brown et al. 2001) and also to IPO001, IPO94265 and IPO94269. KK was susceptible to all isolates from Portugal except IPO92001, to which it was specifically resistant. T. macha had quantitative resistance to all isolates and, in addition, had specific resistance to isolates IPO323 and IPO89011 from the Netherlands.
Cultivars Baldus and Longbow were susceptible to all isolates tested. Synthetic 6x was completely resistant to four isolates from Portugal and to four isolates from the Netherlands (Table 5). Synthetic 6x had low levels of pycnidial lesions covering the leaf with IPO001 and IPO290 but was susceptible to IPO92006. It appeared to be moderately susceptible to isolates IPO92002 and IPO92003 (Table 5). However, only a single leaf inoculated with each of these isolates was diseased while other replicate leaves had no symptoms. Further tests showed Synthetic 6x to be resistant to these isolates (L. S. Arraiano et al., unpublished results). Chinese Spring had some disease with isolate IPO323 (27%) but much less than would be expected (100%) given its highly susceptible response to other isolates (Table 5). This confirms Chinese Spring's specific resistance to isolate IPO323, also observed in the field (Brown et al. 2001).
The principal purpose of developing this technique was to enable screening for resistance to single isolates of M. graminicola in wheat cultivars at the seedling stage. There was a close relationship between the results of field and seedling trials and those of detached leaf tests in a wide range of conditions tested in experiment series A for environmental conditions (Table 1, 3).
Detached seedling leaf tests have been used extensively in work on powdery mildew, stagonospora nodorum blotch, and Fusarium ear blight (Benedikz et al., 1981; Brown & Wolfe, 1990; Diamond & Cooke, 1999). The advantages of detached leaf methods compared with field and glasshouse trials are the ability to carry out tests on several isolates simultaneously, in controlled conditions at any time of the year, and the relatively low requirement for glasshouse space. A disadvantage of the technique is the artificial conditions in which tests are conducted. The labour required to set up detached leaf tests is balanced by the fact that no labour is required during the course of the test. Unlike whole seedling assays, there is no need to water, fertilize or cut second leaves (Kema et al., 1996a). Although the tests reported here were scored four or five times it is not, in fact, essential to score detached leaf tests so often.
An important difference between the method described here and those described by Forni & Zitelli (1979) and Pyzhikova & Karaseva (1985) for M. graminicola and Benedikz et al. (1981) for P. nodorum is that whole seedlings were sprayed evenly whereas in the earlier work, a drop of inoculum was put on the surface of the detached leaf. Leaves were incubated in darkness for 48 h postinoculation. The value of the dark period was not tested because low light intensity following inoculation increases penetration (Benedict, 1971; Kema et al., 1996c) and infection efficiency (Shaw, 1991) of M. graminicola. A problem mentioned by Benedikz et al. (1981) was that it took a long time for symptoms to appear on detached leaves infected with M. graminicola and that it was difficult to keep the leaves green and healthy for the whole period of the test. In the technique described here, leaves were suspended with only the cut edges lying on, and covered by agar. This allowed air to circulate around the leaf, suppressing contamination, so that mock-inoculated and resistant leaves remained green up to 30 days after inoculation. The best conditions for detached leaf assays are those used as the standard method in this paper. These generally produced results having the highest correlation with results of field and seedling tests. There were higher infection levels on secondary leaves than on primary leaves. However, it is not known if this effect is caused by age of the leaf or its developmental stage. To elucidate this question, tests of primary and secondary leaves of the same age could be conducted. The high correlations between results of detached leaf and whole plant experiments in all temperature and light conditions, for both isolates, implies that any of the conditions tested here should produce acceptable results.
Detached leaf results with isolate IPO323 had higher correlation coefficients with results of field and seedling trials, probably because the cultivar specific avirulence of IPO323 is controlled by a single gene (Kema et al. 2000). This caused large differences in the disease levels on different cultivars (Table 1, 5). Correlations were lower for isolate IPO94269. This may be a consequence of the quantitative, nonspecific nature of resistance to this isolate, which may be more sensitive to differences in environmental conditions than the specific resistance to IPO323 (Table 3).
The detached leaf technique described here allows cultivar-by-isolate specificity to be detected, such that a particular pathogen isolate shows a consistent, stable and repeatable specific pathogenicity towards a particular host cultivar (Johnson, 1992). Cultivar-by-isolate specificity has been described for M. graminicola in field and seedling trials (Eyal et al., 1973; Rosielle & Boyd, 1985; Gilchrist & Velasquez, 1994; Jlibene et al., 1995; Kema et al., 1996a,b; Cordo & Perelló, 1997). In experiment series B, parts 1 and 2, it was shown that such interactions could be identified using a detached leaf technique. Several cultivars, namely Arina, Bezostaya 1, Chinese Spring and Flame, had specific resistance to isolate IPO323 (Table 1, 5). Cultivar KK showed resistance and susceptibility to different isolates (Table 5) confirming results reported by Kema et al. (1996b) and Ezrati et al. (1999). Similarly, at least part of the resistance of Synthetic 6x wheat is isolate-specific. This cultivar was completely resistant to most isolates tested but was susceptible to isolate IPO92006 from Portugal (Table 5).
The detached leaf method has potential for use in studying the genetics of host–pathogen interactions with septoria tritici blotch, as in powdery mildew (Brown & Wolfe, 1990). Nevertheless, it should be emphasized that, although this method replicates field trial results closely, it should be regarded as complementary to field studies of M. graminicola. In other diseases of wheat some resistances are expressed in the adult plant but not in the seedling (Johnson & Taylor, 1972; Hague & Brown, 1996). It is not yet known if any wheat cultivars are resistant to M. graminicola only in the adult stage, so for the time being, trials of adult plants will always have to be conducted. The detached leaf method is an alternative to whole plant assays in glasshouses or controlled environments to test resistance of seedlings.
This research was funded by PRAXIS XXI (BD/9606/96) – Fundação para a Ciência e Tecnologia, Portugal, the European Union Biotechnology programme and the Ministry of Agriculture, Fisheries and Food, UK.