Introduction

Blumeria graminis f. sp. avenae (Bga), the causal agent of powdery mildew disease in oats, is a major limiting factor for oat production and yield particularly in Northern Europe (Montilla-Bascón et al., 2015; Roderick et al., 2000). The use of resistant cultivars is the most sustainable method of disease control in cereal crop production (Berlin et al., 2018). There are at least two modes of reaction by which plants defend against powdery mildew attack: penetration resistance and post-penetration resistance (Sánchez-Martín et al., 2011). Pattern-recognition receptors (PRRs) localised at the cell surface recognise pathogen-associated molecular patterns (PAMPs), for example fungal chitin, to trigger PAMP triggered immunity (PTI). Penetration resistance represents a major element of PTI including defence mechanisms such as papillae and cell wall appositions that form under the site of attempted fungal ingress to act as barriers to penetration (Wei et al., 2020).

Adapted pathogen isolates have evolved to overcome PTI by secreting virulence effectors that enable infection. Therefore post-penetration resistance is dependent on intracellular NLR (nucleotide-binding domain and leucine-rich repeats) receptors, that recognise fungal effectors to induce effector triggered immunity (ETI). ETI is typically associated with the post-penetration hypersensitive response (HR) characterised by localised programmed cell death (Kang et al., 2020). Encasement of the fungal haustoria in callose and phenolic compounds also contributes to resistance by preventing the uptake of nutrients from the plant and, in turn, limits fungal growth (Ramming et al., 2011). Both penetration and HR-mediated resistance mechanisms can be histologically assessed using microscopy (Fondevilla et al., 2006; Wei et al., 2020). The frequency of these resistance mechanisms can be used as indicators to characterise genetic disease resistance (Sánchez-Martín et al., 2011).

To date, 13 powdery mildew Resistance (R)-genes, denoted as Pm, have been catalogued in oats (Ociepa et al., 2020; Ociepa & Okoń, 2022; Schurack et al., 2024). These genes have been sourced from several wild oat species: Pm1, Pm3, Pm11 and Pm12 were sourced from Avena sterilis accessions (Hsam et al., 1998; Hsam et al., 2014; Ociepa et al., 2020; Ociepa & Okoń, 2022); Pm2 from A. hirtula (Herrmann & Mohler, 2018); Pm4 from A. barbata (Hsam et al., 2014); Pm5 from A. macrostachya (Yu & Herrmann, 2006); Pm6, Pm9 and Pm10 from A. byzantina (Herrmann & Mohler, 2018; Hsam et al., 2014); Pm7 from A. eriantha (pilosa) whereas Pm8 was mapped in the cultivar Rollo (Hsam et al., 2014) and Pm13 in the cultivar Husky (Schurack et al., 2024).

However, only a limited number of these R-genes have been deployed in commercial cultivars possibly due to the yield penalties associated with them (Herrmann & Mohler, 2018). There are existing virulent Bga isolates against a range of Pm-genes. Oats carrying Pm1, Pm3, and Pm8 are susceptible to both Irish and Polish Bga isolates. UK Bga isolates exist that are virulent on oat cultivars with mildew resistance (OMR) groups 1–4 which correspond to Pm6, Pm1, Pm3 and Pm4 respectively (Roderick et al., 2000). A German isolate is also virulent to Pm1, Pm3 and Pm6 and several Polish isolates are virulent on Pm6 (Reilly et al., 2024; Schurack et al., 2024). Finally, there is a low frequency of virulent Polish Bga isolates reported on Pm7 and Pm12 cultivars (Cieplak & Okoń, 2023; Okoń et al., 2018,b). For these reasons, new sources of disease resistance need to be identified and characterized (Sowa & Paczos-Grzęda, 2019).

Quantitative resistance is typically incomplete, durable and conferred by multiple quantitative genes each contributing to resistance that is not easily overcome by new races of pathogens (Sánchez-Martín et al., 2011). These genes can be involved in PTI responses, or in some cases the loss of susceptibility genes. ETI may also contribute to quantitative resistance where several R-genes are eroded by the evolution of virulent pathogen genotypes (Cowger & Brown, 2019).

Plant breeding for cereal crops began in Ireland under state ownership in 1905. Oat breeding and research was focused on the establishment of new native oat varieties with higher grain yields. While imported varieties were also tested for suitability to Irish climatic conditions (Caffrey, 1948). However, the availability of new farming technologies, and international breeders, put pressure on the generation and production of Irish varieties. In 1973 Ireland joined the EU which brought increased interest in the Irish market from large international breeders and by 1996 oat breeding was no longer taking place in Ireland. Internationally bred, commercial oat cultivars were therefore introduced and imported for production (O’Donovan, 2019). Although they are no longer grown commercially, a collection of heritage oat accessions has been maintained in a gene bank by the Department of Agriculture, Food and the Marine (DAFM), Co. Kildare in an effort to preserve agro-biological diversity. The collection of Irish historical oats includes both native accessions and some varieties imported from Austria, and the United Kingdom (EURISCO Catalogue, http://eurisco.ecpgr.org). These oats may have been overlooked as potential sources of disease resistance. Therefore, we assessed powdery mildew resistance in these heritage oats.

Materials and methods

Plant material

Plant materials consisted of 29 Irish heritage oat accessions that are registered on the EURISCO Catalogue, the catalogue of ex-situ collections in Europe (http://eurisco.ecpgr.org). All are spring oats with the exception of Wexford Tawny. Seeds from this collection were used with permission from the Irish Department of Agriculture, Food and Marine (DAFM). Commercial oat cultivars Barra and Elison were used as susceptible and resistant controls, respectively (AHDB, 2021; DAFM, 2022). Oat seeds were placed on petri dishes with moistened filter paper, then stratified for three days at 4 °C and then incubated at room temperature in darkness for three days to synchronise germination. Of the 29 heritage accessions available from DAFM, 22 germinated (Table 1). Two seedlings were grown together in 1 litre pots filled with John Innes No.2 compost.

Table 1 Oat accessions, Irish heritage and modern accessions (marked with an asterisks), origin, pedigree, date and details
Full size table

Glasshouse screen for Bga resistance

Plants were inoculated with a single spore isolate of Bga, which was collected from the cultivar Barra on the University College Dublin research farm, Lyons, Co. Kildare in 2017 (labelled W23ENY2, for the corresponding Eircode). Barra was the predominant cultivar grown in Ireland from 1985 to 2022 (DAFM, 2022). The isolate was maintained on Barra plants in a sealed containment unit under natural light and at room temperature (~22 °C). The isolate was propagated by transfer to new Barra seedlings every 2 weeks. For fungal inoculations, leaves after the first true leaf were infected with fresh pustules of Bga by touch-inoculation from infected plants (Reilly et al., 2024).

Pots with germinated seeds were placed under glasshouse conditions at the Rosemount Environment Research Station, Belfield, UCD, under natural light (13-16 h daylength) and temperature conditions (12–22 °C). To assess potential seedling resistance, two seedlings of each accession were grown to growth stage (GS) 13 (14 days old). Three leaves on each of the two plants (together in the same pot) for each accession were inoculated with the single spore Bga isolate W23ENY2 and marked with a waterproof marker for future assessment. The three leaves of inoculated plants were subsequently scored 28 days post-inoculation (dpi) for percentage leaf area infected with Bga. Data were analysed using the R statistical programme (R Core Team), with packages ‘lme4’, ‘emmeans’, ‘multcomp’, ‘multcomView’ and ‘ggplot2’. Disease severity means were analysed using a generalized linear model (glm) with a quasibinomial distribution. Individual oat accessions were treated as levels of the fixed factor “Oat Accession”. Data from all levels of the factor were assessed. Differences in means was determined by the Tukey’s HSD test to compare differences between the accessions. All figures were generated using the R statistical package “ggplot2” (R Core team, 2020). Three independent experiments were performed sequentially.

Histological analysis of seedling resistance

To characterise seedling resistance to Bga, the accessions that exhibited resistance (i.e., disease severity scores not significantly different to the resistant control Elison and/or significantly different to the susceptible control Barra during glasshouse screens) were used for further histological assays. For this, the first and second true leaves from GS 13 plants were touch-inoculated with the Bga isolate, W23ENY2. Segments (2–3 cm) of infected leaves were harvested 48 hours post-inoculation (hpi) and boiled in Trypan Blue solution for 20 minutes (Koch & Slusarenko, 1990; Reilly et al., 2024). Leaf segments were subsequently placed in chloral hydrate (2.5 g/ml H2O) overnight. Segments were rinsed with water and mounted on microscope slides with 80% glycerol. Slides were examined using a Leica DM5500B microscope under bright field light conditions to identify the infection stage of Bga conidiospores. Spores were then examined under fluorescent light using the Leica DM5500B microscope with filter cube D setting (UV + violet excitation range) to detect the presence/absence of a resistance response (Carver et al., 1992), defined as follows: (1) appressorium-stage spore with no penetration, papillae formation; (2) successfully penetrated spore, with haustorial encasement; (3) successfully penetrated spore, with epidermal cell death; (4) successfully penetrated spore with no resistance response (Fig. 1). Images in bright field and fluorescent light were recorded with a Leica DFC310FX digital camera. A minimum of 50 germinated spores were quantified on each leaf segment (two leaf segments from two separate plants per accession on each microscope slide, per independent experiment). Three independent experiments were performed, for a total of 300 spores assessed per oat accession. Percentage data was calculated from microscopy counts and fitted to a generalized linear mixed model (glm) with a quasi-binomial distribution. The oat accessions (Barra, Elison, Sandy and Tyrone 1994) and the resistance responses (papillae, cell death and encasements) were treated as levels of the fixed factors “Oat Accession” and “Resistance Response”, respectively. Data from all levels of each factor were assessed. Multiple comparison tests were performed to compare differences between the frequencies of each resistance response expressed in each oat accession using Tukey HSD test. This was achieved using packages “multcomp”, “multcompView” and “emmeans”. All figures were generated using the R statistical package “ggplot2” (RCore team, 2020).

Fig. 1
figure 1

Leaves infected with Bga spores were stained with trypan blue and visualized under (a) brightfield light and (b) fluorescent (UV + violet) light, counted and categorised into four groups: (1) appressorium-stage spore with no penetration, papillae formation; (2) successfully penetrated spore, haustorial encasement; (3) successfully penetrated spore, epidermal programmed cell death (PCD); (4) successfully penetrated spore with no resistance response. Asterisks indicates conidium; white line for site of papillae, encasement and PCD. Images were recorded with a Leica DFC310FX digital camera

Full size image

Host-isolate tests to determine resistance factors

Host-pathogen tests, previously described by Okoń and Kowalczyk (2012), were carried out on the heritage oat accessions resistant to W23ENY2, Sandy and Tyrone 1994, together with 12 control oat genotypes that were previously found to carry different Pm-genes (Supp. Table 1). At the same time we also compared the response of the 12 control oat genotypes to the commercial oat cultivars Barra, Binary, Delfin, Elison and Yukon which was published previously. Therefore the bordered part of the Table 2 is reproduced with permission and was published in Crop Protection, 176, Reilly et al., Breadth of resistance to powdery mildew in commercial Oat cultivars available in Ireland, 106517, Copyright Elsevier (2024).

Table 2 Seedling responses of different oat genotypes to different single spore isolates of Bga PL 1–9 (Bialka 2014; Choryn 2015; Choryn 6 2018; Chorynskowice 3 2018; Felin 4 2018; Modzuirow 4 2018; Polanowice 7 2018; Strzelce 2015 (1) pk; Zabno 2015 pk) and the Irish isolate W23ENY2. The mean percentage powdery mildew disease coverage is shown with SEM in brackets; green indicates resistance (0–10%), yellow indicates moderate susceptibility (11–30%) and red, susceptibility (>30%). Detached leaves of 10-day old seedlings were scored 10 days post-inoculation. The mean disease severity results are shown in the table below with SEM in brackets. The black bordered part of the table is reproduced with permission and was published in Crop Protection, 176, Reilly et al., Breadth of resistance to powdery mildew in commercial Oat cultivars available in Ireland, 106517, Copyright Elsevier (2024).
Full size table

Two, 2-3 cm leaf segments of 10-day old seedlings of each accession and control genotype were harvested and placed into petri dishes with benzimidazole agar. Petri dishes were individually inoculated with single spore isolates of Bga collected in Poland, Bga_PL 1–9 (Bialka 2014; Choryn 2015; Choryn 6 2018; Chorynskowice 3 2018; Felin 4 2018; Modzuirow 4 2018; Polanowice 7 2018; Strzelce 2015 (1) pk; Zabno 2015 pk) (Cieplak et al., 2021) and the Irish isolate W23ENY2 as previously described by Okoń and Kowalczyk (2012). These selected Bga isolates have different virulence profiles against oat cultivars with known resistance genes (Supp. Table 1). Bga isolates were maintained on leaf segments in sealed containers on the oat cultivar Fuchs with the exception of W23ENY2 which was maintained on Barra. Leaves were scored for percentage leaf area infected with Bga at 10 dpi. The experiment was conducted twice. Resistance levels to these isolates were then grouped into three categories; R (resistant) from 0 to 10% infection; I (intermediately resistant) from 11 to 30%; and S (susceptible) above 30% (Draz et al., 2019; Reilly et al., 2024).

Results

Seedling resistance to Bga

The Irish heritage accessions were grown to GS 13 in the glasshouse and subsequently scored 28 days post-inoculation (dpi) with Bga for percentage leaf area infected. Of the 22 oat heritage accessions assessed, two accessions (Sandy and Tyrone 1994) were found to express seedling resistance to Bga (Fig. 2). Tyrone 1994 had similar levels of Bga infection (4.3%, ± 0.9 SEM) to the resistant control, commercial cultivar Elison (3.4%, ± 1.0 SEM), and both had significantly (p = 0.0001) lower levels of infection compared to the susceptible commercial control cultivar Barra (17.8%, ± 2.2 SEM). Levels of Bga infection on Sandy (9.3%, ± 1.0 SEM) were significantly (p = 0.02) lower than that on Barra (17.8%, ± 2.2 SEM), but significantly (p- = 0.04) higher than Elison (3.4%, ± 1.0 SEM). All other 20 heritage oat accessions did not have statistically different levels of Bga infection to the susceptible control Barra and are therefore susceptible to the single spore isolate, of Bga (W23ENY2) used in this glasshouse study (Fig. 2).

Fig. 2
figure 2

Differences in mean disease severity scores of oat heritage seedlings grown under glasshouse conditions. Three-leaf seedlings were inoculated with a single-spore isolate of Bga W23ENY2 and scored 28 days post-inoculation. Three leaves on the main tiller of each plant were assessed for percent leaf area infected with Bga. Three independent experiments were performed. Means with the same letters are not significantly different to the susceptible and resistant controls, Barra and Elison, respectively; determined by Tukey test (p ≤ 0.05). Error bars represent confidence intervals

Full size image

Histological analyses of seedling resistance

To characterise the mechanisms contributing to resistance in accessions Sandy and Tyrone 1994, Bga-inoculated leaves were microscopically examined at 48 hpi (Fig. 1). Barra and Elison were included in this assessment as susceptible and resistant controls, respectively. The assessment revealed that resistance in the heritage accession Sandy was predominantly conferred by an encasement response. The frequency of encasements in Sandy (58.7%, ± 2.7 SEM) was significantly (p ≤ 0.05) higher compared to Barra (23.3%, ± 3.4 SEM), Elison (36%, ± 3.8 SEM) and Tyrone 1994 (39.3%, ± 4.7 SEM). Elison did not have statistically different frequencies of encasements to Tyrone 1994 and Barra. The resistance to Bga infection in the accession Tyrone 1994 was conferred by a combination of both encasements and the formation of papillae. Tyrone 1994 formed significantly (p < 0.05) higher frequencies of papillae (30.3%, ± 2.3 SEM) than Barra (8.7%, ± 2.0 SEM), Elison (10.3%, ± 2.3 SEM) and Sandy (15.7%, ± 3.5 SEM). The frequencies of papillae formation between Barra, Elison and Sandy were not significantly different to each other (Fig. 3).

Fig. 3
figure 3

Frequency of defence responses to Bga penetration attempts on different oat cultivars. The frequency of these responses indicates levels of cultivar resistance or susceptibility; the presence of programmed cell death, haustoria encasements and papillae. Observed data are based on 100 germinated spores where two leaves from two separate plants for each oat accession were scored for 50 germinated spores. This was repeated three times independently for a total of 300 germinated spores per accession. The frequency of each defence response was individually compared between each cultivar (i.e., Cell death frequency was compared between cultivars; frequency of encasements was compared between cultivars; papillae frequency was compared between cultivars). Response means sharing the same letter are not significantly different to each other, determined by Tukey test (p ≤ 0.05)

Full size image

The resistant control Elison had a significantly (p = 0.0001) higher frequency of cell death responses to Bga infection than the other oats accessions assessed. The frequency of cell death responses in Sandy (0.7%, ± 0.3 SEM) was statistically similar to the susceptible control Barra (0.3%, ± 0.3 SEM). Cell death in Tyrone 1994 (3.0%, ± 0.6 SEM) was significantly (p < 0.05) higher than both Sandy (p = 0.02) and Barra (p = 0.01) but significantly lower than Elison (39.7%, ± 0.9 SEM) (p = 0.0001) (Fig. 3).

Host-isolate tests to determine resistance genes

To further characterise if the resistance in Tyrone 1994 and Sandy is broad spectrum, host-pathogen screens were completed using a collection of oat cultivars with ‘known’ R-genes (Supp. Table 1) and a panel of single-spore Bga isolates (Okoń & Ociepa, 2018; Reilly et al., 2024). By comparing the infection patterns between the heritage accessions and oat cultivars with known R-genes, it may be possible to determine if resistance genes are present in Sandy and Tyrone 1994 or if the resistance is broad spectrum. Means (± SEM values) of powdery mildew disease levels are outlined in Table 2.

The oat accession Sandy was susceptible to isolate Bga_PL 4 and Bga_PL 8 with powdery mildew disease coverage of 37.5% ±2.5 and 37.3 ± 5.4 respectively (Table 2). Sandy was completely resistant to Bga_PL 7 (0.5% ± 0 disease coverage) and W23ENY2 (0.3% ± 0.3 disease coverage) but had intermediate resistance to all other isolates tested (12.5–25% disease coverage).

Tyrone 1994 was resistant to isolates Bga_PL 5, Bga_PL 7 and W23ENY2 with disease coverage of 8% ±7, 5% ± 5 and 0.9% ± 0.4 respectively. Intermediate resistance was observed to all other isolates tested (12.5–20% disease coverage) (Table 2).

For comparison, the commercial cultivar Elison used as a resistant control in the glasshouse and microscopy assay was resistant to all Bga isolates tested, with a low disease coverage of between 0 and 2.8% ± 2.3. While Barra was used as a susceptible control in the glasshouse and microscopy assay, it was only susceptible to Bga_PL 2 and W23ENY2 with disease coverage of 32.5% ± 17.5 and 32% ± 12.2 respectively. Barra was resistant to Bga_PL 1; Bga_PL 5; Bga_PL 7 with low disease coverage of up to 0.5% ± 0.5 and had intermediate resistance to Bga_PL 3; Bga_PL 4; Bga_PL 8 and Bga_PL 9 (12.5–25% disease coverage) (Table 2).

Discussion

This study aimed to identify and characterise resistance to powdery mildew in a heritage oat collection. Heritage collections of crops have been previously defined as old varieties from recent centuries that are associated with specific regions, cultures and people who have taken the care to conserve and maintain them (Wolfe & Ceccarelli, 2020). The Irish heritage oat collection listed on EURISCO, contains a number of predominantly native oat varieties that were bred in Ireland, Northern Ireland, as well as a number from Scotland, and England (Table 1). Although some of these accessions are not originally native to Ireland, their importance to Irish agriculture in recent centuries has earned their inclusion in this collection (Caffrey, 1933, 1941, 1948; Caffrey & Carroll, 1938).

As heritage and landrace populations are genetically diverse, they can be used as a source of unique alleles for modern plant breeding programs where beneficial alleles, such as those for resistance, are introgressed into elite backgrounds (Acquaah, 2015; Monteagudo et al., 2019). However, the use of landraces or heritage varieties in breeding programs can be limited by linkage drag when undesirable alleles are introduced along with the beneficial alleles (Monteagudo et al., 2019).

Due to this and the self–pollination of these individual accessions to bulk seed stock, the segregation of genes may have resulted in minor phenotype differences among individual seeds within an accession (Mazurkievicz et al., 2019). However, as these accessions were compared to commercial ‘uniformed’ genotypes such as Barra and Elison, we can determine their level of Bga resistance.

Of the 22 heritage accessions tested in this study, two (Tyrone 1994 and Sandy) were shown to express resistance to the Irish Bga isolate (W23ENY2). In agreement with our previous studies this Bga isolate causes high levels of disease on Barra (the predominant oat cultivar used in Ireland between 1985 to 2022). However, W23ENY2 causes intermediate disease on cultivars with Pm1, Pm3 and Pm8 and has not broken Pm4, Pm5, Pm6, Pm7, Pm9 or Pm10 (Reilly et al., 2024). From initial glasshouse screens, Tyrone 1994 has similar low disease scores to Elison, the resistant commercial control cultivar. Elison is a descendent of the cultivar Canyon the first released cultivar with the major R-gene Pm7 (Brodführer et al., 2023). Sandy, had higher levels of disease compared to Elison, but lower disease levels than found in the susceptible control Barra.

Powdery mildew penetration can be blocked via papillae formation. However, if penetration is successful, subsequent feeding by the haustorium can be ‘cut off’ by encasement in phenolic compounds, callose or through HR induction (Berkey et al., 2012). Tyrone 1994 did not have a similar response to any of the control oat-isolate interactions with known Pm-genes. However, four Pm-genes were not tested in these screens (Pm2, Pm11, Pm12 and Pm13). Pm2 confers resistance to a range of Bga isolates and therefore would not be expected to be present in Tyrone 1994 or Sandy (Okoń & Ociepa, 2017). The inclusion of additional Pm oat genotypes in future host-pathogen tests may provide evidence for the presence or absence of Pm-genes. Alternatively, studying the ancestral lineage of Sandy and Tyrone 1994 may give clues to the source of and/or identity of Pm resistance.

The lack of HR to Bga isolate W23ENY2 together with the intermediate resistance of Sandy and Tyrone 1994 to the majority of Bga isolates tested would support a quantitative resistance. Quantitative resistance is diverse, multigenic and in some cases may overlap with race-specific resistance (Kang et al., 2020). The high levels of encasements in Tyrone 1994 and Sandy may therefore be remnants of overcome R-gene(s). While Tyrone 1994’s resistance may also be based on components of PTI mediating the higher levels of papillae formation.

In this study, we identified Bga resistance in two Irish historical oat accessions Sandy and Tyrone 1994. The mechanisms responsible for resistance were papillae formation and/or haustorial encasements. The intermediate resistance of Sandy and Tyrone 1994 may provide components for achieving durable resistance.