To be or not to be a nonhost species: A case study of the Leptosphaeria maculans and Brassica carinata interaction Authors: Julie M. Noah, Mathilde Gorse, Carole‐Anne Romain, Elise J. Gay, Thierry Rouxel, Marie‐Hélène Balesdent, Jessica L. Soyer Published: 2024 ABSTRACT Abstract Leptosphaeria maculans is one of the major fungal pathogens on oilseed rape (Brassica napus), causing stem canker disease. The closely related Brassica species B. nigra, B. juncea, and B. carinata display extreme resistance toward stem canker. In this study, we demonstrate the nonhost status of B. carinata toward L. maculans in France through field experiments and inoculations performed in controlled conditions. A few isolates moderately adapted to B. carinata in controlled conditions were recovered in the field on B. nigra leaves, allowing us to investigate the unusual B. carinata–L. maculans interactions using molecular, macroscopic, and microscopic analyses. A cross between a L. maculans isolate adapted to B. napus and an isolate moderately adapted to B. carinata allowed the generation, in the lab, of recombinant L. maculans strains better adapted to B. carinata than the natural parental isolate obtained from B. nigra, and highlighted the polygenic determinism of the adaptation of L. maculans to B. carinata and B. napus. This biological material will allow further investigation of the molecular determinants of the adaptation of L. maculans to nonhost species and elucidate the genetic resistance basis of B. carinata. INTRODUCTION Phytopathogenic fungi are one of the major problems in agriculture, causing economic and food production problems (Pennesi, 2010). However, most interactions between a potential plant pathogen and a potential host plant do not lead to disease. Plants exhibit two types of disease resistance: nonhost resistance (NHR) (Heath, 2000) and host‐specific resistance, with NHR being the most widespread as it prevents most plants from being infected by potential pathogens. In NHR, all genotypes of a considered plant species are resistant to all genotypes of a potential pathogen species. Host resistance is genotype‐specific, as some genotypes of a given plant species are resistant toward some genotypes of the adapted pathogen species (Niks & Marcel, 2009). Mechanisms responsible for NHR remain to be fully deciphered, but some resistance mechanisms involved in host‐specific and nonhost resistances converge (Gill et al., 2015). In NHR, two types of resistance can be distinguished: type‐I and type‐II nonhost resistances (Schulze‐Lefert & Panstruga, 2011). Type‐I NHR is a pre‐invasive resistance mechanism characterized by a complete absence of visible symptoms, thus involving passive defence plant mechanisms. On the contrary, type‐II NHR, a post‐invasive resistance mechanism, is characterized by a local cell death, called hypersensitive response (HR; Gill et al., 2015). While the defence mechanisms set up in host‐specific or nonhost resistance can involve the same actors, the difference between both mechanisms lies in the level at which the resistance is expressed (either all genotypes of a given species are resistant or only some genotypes are). To date, NHR is considered more durable than host‐specific resistance in the management of plant diseases (Lee et al., 2017; Panstruga & Moscou, 2020) and many examples can be found of interspecies transfer of NHR genes into host‐species in breeding programs (Kawashima et al., 2016; Wang et al., 2018; Zhao et al., 2005; see more examples in the review of Ayliffe & Sørensen, 2019). Leptosphaeria maculans is a Dothideomycete fungus that causes stem canker of oilseed rape (Brassica napus). It is one of the main fungal pathogens of B. napus leading to important yield losses, ca. 20% worldwide (Fitt et al., 2006; Rouxel et al., 2011). This hemibiotrophic pathogen has a complex life cycle tightly associated with that of its host plant, during which it alternates between different nutrition modes (Rouxel & Balesdent, 2005). Firstly, it survives and develops as a saprophyte on infected crop residues, on which sexual reproduction occurs. Ascospores are then disseminated by the wind and infect seedlings (e.g., in autumn on winter oilseed rape in Europe) after the penetration of hyphae through stomata or natural wounds (Hammond & Lewis, 1987). The asymptomatic development in young leaves is followed by a short necrotrophic phase, resulting in the development of greyish spots on leaves leading to asexual reproduction. L. maculans then initiates systemic and biotrophic colonization from the petiole to the stem basis, progressing in and around the vascular bundles (Hammond et al., 1985). After several months of asymptomatic colonization in the stem, necrosis appears at the stem basis, causing plant lodging and resulting in yield losses (Rouxel & Balesdent, 2005). In the field, L. maculans is often associated with the closely related species Leptosphaeria biglobosa (West et al., 2001), and these two Leptosphaeria species can be found together on the same host and even in the same plant tissues (Jacques et al., 2021). They both cause necrotic lesions and colonize host plant tissues, but their impacts on crops are different as it is postulated that only L. maculans can cause lodging and damage to the crops (West et al., 2001). The favoured control methods against stem canker of oilseed rape relies on the identification of effective host resistance sources (mostly resistance genes) and the breeding of resistant cultivars. This genetic resistance mechanism relies on a monogenic determinism following the gene‐for‐gene relationship (Flor, 1955): interaction of the product of a host‐specific resistance gene (R) with the product of an avirulence gene (AVR) causes HR, which thus blocks the progression of the pathogen. Until now, 21 specific major resistance genes (Rlm), obeying the gene‐for‐gene relationship, have been identified in different Brassica species, and some have been introduced into B. napus (see Cantila et al., 2021 for a detailed list of the identified resistance genes toward blackleg Brassica spp). For instance, B. rapa and B. juncea have been used in breeding programs in various countries (Cantila et al., 2021; Raman et al., 2013) to select resistance genes and introduce them into oilseed rape. Nevertheless, L. maculans has a great capacity to adapt, for example via the loss, inactivation, or point mutations of avirulence genes leading to an absence of recognition of the avirulence protein by the host surveillance machinery (Rouxel & Balesdent, 2017). This enables L. maculans populations to overcome resistance deployed in the field (Daverdin et al., 2012; Zhang et al., 2015), illustrating the need to find new sources of resistance that would be more durable than resistances currently used in the field. L. maculans is a specialist fungus with a limited host range. This fungus can infect a few Brassicaceae species including species from the U triangle (Nagaharu, 1935). The U triangle is a schematic representation of the interspecific hybridization between three ancestral Brassica species: Brassica rapa (turnip, Chinese cabbages; A genome), Brassica nigra (black mustard; B genome), and Brassica oleracea (cabbages; C genome). These three diploid species are natural progenitors of the three allotetraploid Brassica species: Brassica napus (oilseed rape, canola; AC genome), Brassica juncea (Indian mustard; AB genome), and Brassica carinata (Ethiopian mustard; BC genome) (Nagaharu, 1935). The species B. napus, B. oleracea, and B. rapa are susceptible host species of L. maculans (Gugel et al., 1990). On the contrary, much rarer examples are available in the literature of isolates able to attack B. juncea, B. nigra or B. carinata (Gugel et al., 1990; Johnson & Lewis, 1994; Li et al., 2005; Roy, 1978; Somda et al., 1999; Zhu et al., 1993). In analyses from Gugel et al. (1990), Johnson and Lewis (1994), Somda and Brun (1995) or Li et al. (2005), isolates able to infect B. carinata corresponded to “weakly” virulent isolates of L. maculans; these “weakly” virulent isolates were later identified as isolates from the L. biglobosa species (Shoemaker & Brun, 2001). However, two L. maculans Australian isolates, isolated from B. napus in the field, were reported moderately virulent on one B. carinata genotype (Li et al., 2005). We analyse here the extreme resistance of B. carinata toward L. maculans by (i) evaluating the field behaviour of B. carinata in French conditions and (ii) evaluating the adaptive ability of L. maculans to this species using phytopathological, cytological, genetic, molecular, and transcriptomic approaches. A collection of natural L. maculans isolates, including isolates collected on B. nigra (Somda & Brun, 1995) were then evaluated on several B. napus and B. carinata genotypes to assess their capacity to infect these species. One of the isolates, HB10.19, causing limited symptoms on B. carinata cotyledons, was used to screen a large collection of accessions of B. carinata. A cross between the reference isolate of L. maculans, JN3, causing virulence symptoms on B. napus while non‐adapted to B. carinata, and HB10.19 was done. The progeny isolates displayed a continuum of aggressiveness on cotyledons of both B. napus and B. carinata, suggesting a complex and multi‐genic determinism of the adaptation to the hosts. A few progeny isolates had a transgressive behaviour and could infect B. carinata cotyledons in controlled conditions. However, these could not proceed to the stems in which they only caused minor symptoms. Altogether, these data indicate that B. carinata is a nonhost species of L. maculans in our conditions; they nevertheless suggest that recombination of traits at meiosis can allow the emergence of transgressive L. maculans isolates able to infect either B. napus and B. carinata cotyledons. EXPERIMENTAL PROCEDURES Plant and fungal material Sixteen plants from the B. carinata line SD199947, provided by Dr. Silva Dias, were further self‐pollinated in bulk to produce the B. carinata line D5.6.12 used in this experiment; 52 accessions of B. carinata (Table S1) were retrieved from the IPK collection (IPK, Gatersleben, Germany). The D5.6.12 and IPK genotypes were evaluated at the seedling stage for their resistance toward six L. maculans isolates. A total of 53 L. maculans isolates, present in the INRAE collection of isolates, and from diverse plant and geographical origins, were used in pathogenicity tests (Table S2). Field experiments Experimental fields were set up for five cropping seasons as described in Degrave et al. (2021) and Jacques et al. (2021) with a reinforced inoculum (Kerdraon et al., 2020). In 2008–2009, 2009–2010, 2010–2011, the B. carinata line D5.6.12, B. napus cv. Goéland (susceptible to French L. maculans populations), and the B. napus cv. Fu‐Eurol‐Rlm10 (in which the B. nigra resistance gene Rlm10 was introgressed) were sown at eight locations in France (Degrave et al., 2021). In 2016–2017 and 2017–2018, the genotypes were sown at the INRAE experimental field station at Grignon, France with cv. Darmor instead of cv. Goéland (Jacques et al., 2021). Fields were regularly surveyed for the development of L. maculans or L. biglobosa symptoms and leaf spots were collected to recover Leptosphaeria isolates. Isolates (culture, sporulation, and in vitro cross) Leptosphaeria isolates were recovered from single pycnidia on symptoms developed on Brassica leaf collected in field experiments as described in Balesdent et al. (2023). Isolates of L. maculans and L. biglobosa were grown on V8‐agar at 24°C for sporulation, and stored for long term at 4°C on 1% malt‐2% agar slant tubes to establish a collection of isolates. Mycelium of the isolates was collected and placed in 2 mL Eppendorf® tubes for further DNA extraction. An in vitro cross between the isolates JN3 and HB10.19 was performed as previously described (Balesdent et al., 2001). Progeny of the cross was recovered as described in Plissonneau et al. (2016). Fungal cultures and conidia production were performed as described in Ansan‐Melayah et al. (1995). Pathogenicity assays Pathogenicity assays at the cotyledon stage were performed as described in Balesdent et al. (2001), on 12‐days‐old seedlings of genotypes of B. napus and genotypes of B. carinata. Two methods were used for inoculation, with or without wounding of the cotyledons. In the former, cotyledons were wounded by a needle and 10 μL of a 107/mL pycnidiospores suspension was put on the wound; in the latter, Tween 20 was added in the pycnidiospore suspension (0.1% [v/v] Tween® 20; Roth, Karlsruhe, Germany) and the suspension was applied with a 0.6‐mm brush on the upper part of the cotyledon. After 48 h of incubation under high humidity and darkness at room temperature, plants were transferred in a climate chamber at 24°C (day) and 16°C (night) with a photoperiod of 16 h of light and 80% relative air humidity. Symptoms were scored using the IMASCORE rating scale with six infection classes (IC; Balesdent et al., 2001) in which we have taken into consideration the “intermediate resistance phenotypes” (Neik et al., 2022; Van de Wouw et al., 2009; Yang & Fernando, 2021). Symptoms were scored nine and 12–13 days post‐inoculation (dpi); IC ≤ 2 indicates that the isolate is avirulent (hence the plant is resistant), IC 4–6 indicate that the isolate is virulent (hence the plant is susceptible), IC 2 to < 4 are characteristic of the intermediate interaction phenotypes. A Wilcoxon–Mann–Whitney statistical test was performed using RStudio (RStudio Team, 2020) software to evaluate significant differences of symptoms caused by the progeny isolates compared to the symptoms caused by both parental isolates on each Brassica species. Cotyledon inoculation of B. napus cv Yudal and B. carinata line D5.6.12 with ascospores from naturally infected stubbles was done using the ascospore shower method (Marcroft et al., 2012). Naturally infected stubbles from susceptible B. napus cultivars bearing mature pseudothecia were humidified and placed over 10‐days old plants of B. napus or B. carinata for 48 h in the dark, as described in Gay et al. (2021). Then, the plants were transferred to growth chambers as described above. Symptoms were surveyed during 2 weeks in order to recover isolates from symptoms, when present (Plissonneau et al., 2017). For the analysis of stem infection, pathogenicity assays were performed as described in Gervais et al. (2017). Three isolates (JN3, HB10.19, and V77.1.11) were used to inoculate the stem of B. napus (cv. Yudal‐RlmSTEE98; Jiquel et al., 2021) and B. carinata (line D5.6.12). For each isolate, 16 plants of each species were inoculated. Plants were inoculated at the three‐leaf stage (in our experiment, 22 days post‐sowing), the petioles were horizontally cut 1 cm from the leaf insertion and 10 μL of a 107/mL pycnidiospores suspension was put on the wound. After 48 h of incubation under high humidity and darkness, plants were transferred in a climate chamber at 24°C (day) and 16°C (night) with a photoperiod of 16 h and 80% relative air humidity. The length of the external necrosis of the stem were measured 38 dpi with a calliper then the stem basis was longitudinally cut 5 cm long with a scalpel checking for sign of inner necrosis. A Wilcoxon–Mann–Whitney statistical test was performed using RStudio (RStudio Team, 2020) software to evaluate significant differences in terms of external necrosis caused by L. maculans on each Brassica species. Identification of the species from leaf symptoms using amplification and sequencing of the ITS locus Twenty milligrams of lyophilised mycelium was used for DNA extraction as described in Jacques et al. (2021). Characterization of isolates was based on PCR amplification and sequencing of the Internal transcribed spacer (ITS) locus (Mendes‐Pereira et al., 2003; Table S3), allowing their assignation to L. maculans or L. biglobosa species. Biomass quantification analysis Fungal biomass in infected cotyledons of B. carinata (line D5.6.12) and B. napus (cv. Yudal) was estimated as described in Jacques et al. (2021). Five, 7, 9, and 12 dpi, a 0.25 cm2 area of around 50 infected plant material was collected around the inoculation point and pooled in a falcon tube, flash frozen in liquid nitrogen, and stored at −80°C until further analysis. The infected material was ground before DNA extraction and 80–100 mg of material was used for DNA extraction as described by Jacques et al. (2021). For each condition tested, two DNA extractions of two independent biological samples were done. L. maculans species‐specific qPCR were performed using primers to amplify LmEF1‐α, ACTIN and a probe (Table S3) as described in Jacques et al. (2021). Calibration ranges using DNA from the JN3 isolate as a reference DNA were performed from 5 ng to 5 × 10−3 ng in triplicate, as described in Jacques et al. (2021). Cytological assays Cotyledons from two plants per interactions (i.e., two cotyledons of B. carinata line D5.6.12 or B. napus cv. Yudal with JN2, HB10.19 and V77.1.11, independently) were sampled at seven and 15 dpi following wounded inoculation and maintained in glass vials. Three millliters of trypan blue lactophenol solution prepared as follows: (0.05% [w/v], 0.02 g trypan blue, and 10 mL lactic acid (85% [w/v]), 10 mL phenol (TE buffer equilibrated, pH 7.5–8.0), 10 mL of glycerol (≥99%), and 10 mL distilled water) was added to the sample and placed in boiling water for 3 min. The trypan blue solution was then discarded and replaced with a chloral hydrate solution (3 mL of a 2.5 g/mL solution); this solution was changed every 3 days with fresh chloral hydrate solution until the solution and tissue became colourless. The samples were mounted between slides in 50% glycerol before observation with the Leica DM5550 microscope. RNA‐seq data and analysis Naturally infected leaves of B. napus (cv. Darmor) and B. carinata (line D5.6.12) were collected on November 16 and 29, 2017, from the same experimental field at Grignon, France. Symptoms were classified as “atypical”, that is, presumably due to L. biglobosa, “typical”, that is, presumably due to L. maculans, or asymptomatic. Leaf discs (1 cm diameter) were collected from the samples when infection of B. napus susceptible material was well established, centred on symptoms (or randomly from asymptomatic leaf tissues), for RNA‐seq analysis, as described by Gay et al. (2021). Raw RNA‐seq reads were mapped with the STAR software version 020201 (Dobin et al., 2013) on the reference genomes of L. maculans (isolate JN3) and L. biglobosa (isolate G12‐14) (Dutreux et al., 2018). As isolates of L. biglobosa and L. maculans can be present simultaneously in the same samples, the mapping parameters have been optimized as follows: we chose a maximum number of mismatches of two and a mapping on a concatenated genome of both species. According to the maximum intron size found in the genomes, we allowed an intron size of 10,000 bp. Other parameters have been used as follows: outFilterMultimapNmax: 100; SeedSearchStartLmax: 12; alignSJoverhangMin: 15; alignIntronMin: 10. Then, we selected the properly paired reads in BAM files with Samtools v1.6 (Gay et al., 2023; Li et al., 2009). RESULTS Brassica carinata displays an extreme resistance to Leptosphaeria maculans Attempts were performed to recover L. maculans isolates from B. carinata leaves in the field for three consecutive years at eight sites in France and during two other consecutive years (autumn 2016 and 2017) in fields at Grignon, France (Table 1; Figure S1). While L. maculans was easily isolated from leaf spots on a susceptible or a Rlm10‐B. napus genotype—respectively 94.5% and 85.5% of the isolates—mainly L. biglobosa (98%) isolates were recovered from B. carinata leaf symptoms. The three L. maculans isolates recovered from B. carinata were not able to produce symptoms on B. carinata after re‐inoculation in controlled conditions, while 57% of the isolates recovered from B. napus with the specific resistance gene Rlm10 were found virulent after inoculation on B. napus‐Rlm10 (Table 1). TABLE 1Comparison of Leptosphaeria populations recovered from Brassica napus and Brassica carinata under field conditions.Sampling year B. napus susceptible cvs. a B. carinata line D5.6.12 B. napus cv. Fu‐Eurol74 (Rlm10)Nb of isolates collectedLb b Lm b vir Lm c Nb of isolates collectedLb b Lm b vir Lm c Nb of isolates collectedLb b Lm b vir Lm c 20081012 (2%)99 (98%)na d 3232 (100%)0na554 (7%)51 (93%)36 (71%)200966066 (100%)na2119 (91.5%)2 (9.5%)04910 (20%)39 (79.6%)15 (30.6%)2010122 (7%)10 (83%)na2424 (100%)0nandndndnd2016ndndndnd87 (88%)1 (12%)0ndndndnd201716315 (9%)148 (91%)nd17 + 40 e 17 + 40 (100%)00ndndndndTotal34219 (5.5%)323 (94.5%)141138 (98%)3 (2%)010414 (13.5%)90 (85.5%)51 (57%) a In 2008–2010, cv. Goéland (Rlm9); in 2017, cvs. Yudal (no Rlm gene) and Darmor (Rlm9). b The isolates recovered from leaf symptoms were classified as Leptosphaeria maculans (Lm) or L. biglobosa (Lb) following PCR amplification of the ITS locus. c Number of L. maculans isolates (% of L. maculans isolates) showing virulence when inoculated on cotyledons of the genotypes they were isolated from (B. carinata or B. napus‐Rlm10 genotypes). d na, not applicable (the genotype is susceptible or no L. maculans was isolated); nd, not done. e Number of isolates collected on B. carinata via the “ascospore shower” method (see Section 2) or in field conditions, respectively. Comparison of Leptosphaeria populations recovered from Brassica napus and Brassica carinata under field conditions. In 2008–2010, cv. Goéland (Rlm9); in 2017, cvs. Yudal (no Rlm gene) and Darmor (Rlm9). The isolates recovered from leaf symptoms were classified as Leptosphaeria maculans (Lm) or L. biglobosa (Lb) following PCR amplification of the ITS locus. Number of L. maculans isolates (% of L. maculans isolates) showing virulence when inoculated on cotyledons of the genotypes they were isolated from (B. carinata or B. napus‐Rlm10 genotypes). na, not applicable (the genotype is susceptible or no L. maculans was isolated); nd, not done. Number of isolates collected on B. carinata via the “ascospore shower” method (see Section 2) or in field conditions, respectively. An additional attempt was performed to recover L. maculans isolates on B. carinata with the “ascospore shower" method. This method allowed us to place plantlets of B. carinata under residues of oilseed rape bearing mature pseudothecia. Ascospores released from the pseudothecia produced leaf spots when ejected on B. napus plants, but no typical L. maculans symptoms appeared on the B. carinata D5.6.12 line. Seventeen isolates were nevertheless recovered from the symptoms developed on cotyledons of B. carinata. All these isolates corresponded to L. biglobosa (Table 1). Experiments performed in the field and in controlled conditions failed to recover a L. maculans population adapted to B. carinata. No evidence of gene expression of Leptosphaeria maculans was detected from leaves of Brassica carinata in field conditions We set up a transcriptomic analysis from leaves of B. carinata and B. napus cv. Darmor harvested in the field in autumn (Figure 1). Leaves were collected at two time points after sowing (one and two and a half months) when symptoms of L. maculans on plantlets of B. napus, sown alongside B. carinata, were observed. In B. napus leaf samples with typical L. maculans leaf spots (Figure 1C), Leptosphaeria reads represented around 3.4% (first sampling time) to 7% (second sampling time) of the RNA‐seq read numbers (Figure 1A). In these samples, more than 97.5% of the reads were assigned to the L. maculans species (Figure 1B). Although a very limited number of fungal reads were uncovered from “asymptomatic” and “atypical” B. napus leaves, L. maculans (and to a lesser extent, L. biglobosa) RNA‐seq reads were detected, indicating that the two species were infecting these plant tissues. Even from asymptomatic samples of B. napus, L. maculans accounted for the majority of the transcriptomic activity detected, as 80%–96% of the reads corresponded to L. maculans. In B. napus samples with “atypical” symptoms (Figure 1C), both L. maculans and L. biglobosa were detected in variable proportions, with L. maculans reads representing 15%–73% of the Leptosphaeria reads. In B. carinata leaf samples, a very limited amount of RNA‐seq reads were assigned to either Leptosphaeria species (less than 1%–2%; Figure 1A), and these mostly (more than 99%) corresponded to L. biglobosa (Figure 1B). Proportion of RNA‐seq reads assigned to Leptosphaeria maculans or Leptosphaeria biglobosa within field‐infected leaves of Brassica napus (cv. Darmor) or Brassica carinata (line D5.6.12). (A) Proportions of L. biglobosa (violet) and L. maculans (red) reads in each plant sample. (B) Among fungal reads, proportions of reads mapping to L. biglobosa (violet) or L. maculans (red) genomes. (C) Examples of symptoms observed for each type of sample sequenced. Leaf disks were collected at two sampling dates (1, 2 months; 2, 2.5 months after sowing); three independent replicates (different field plots) were sequenced for each time‐point and plant material. The finding of a natural isolate of Leptosphaeria maculans showing partial adaptation to Brassica carinata A collection of 53 isolates originating from diverse plant species and geographical origins (Table S2) were inoculated on B. napus cv. Westar and B. carinata D5.6.12 line in controlled conditions. Among them, 37 isolates originating from B. napus or Raphanus sp., and the reference isolates, JN2 and JN3, were virulent on B. napus cv. Westar (mean score = 4.65 ± 0.56) but failed to infect B. carinata (mean score = 1.05 ± 0.14), with the rapid appearance of localized necrosis directly at the point of infection (Table S2; Figure 2A for an example of the symptom on B. carinata line D5.6.12). In contrast, four isolates among the 10 collected on B. nigra cv. Junius produced small symptoms on B. carinata line D5.6.12 corresponding to intermediate resistance phenotypes (mean scores ranging from 2.8 to 3; Table S2). These four isolates were unable to infect B. napus cv. Westar (mean score = 1). The isolate HB10.19, displaying an intermediate phenotype on B. carinata (mean score = 2.9 ± 0.6; Figure 2A), was selected for further experimentations. Comparison of the infection process of three Leptosphaeria maculans isolates on Brassica carinata (line D5.6.12) and Brassica napus (cv. Yudal). (A) Macroscopic symptoms at 12 days post‐inoculation (dpi) of the three isolates inoculated on B. napus and B. carinata. (B) Microscopic observation of the infection. Isolates were coloured with trypan blue. (C) Biomass quantification of the isolates. Quantification from samples recovered 5, 7, 9, and 12 dpi. Dark blue, isolate JN3; green, isolate HB10.19; yellow, isolate V77.1.11. R1, replicate 1; R2, replicate 2. St, stomata, hyp, hyphae, pyc, pycnidia. Genetics of the interaction show a complex determinism of the adaptation of Leptosphaeria maculans on Brassica carinata and Brassica napus The isolate HB10.19 was crossed with the L. maculans reference isolate JN3, and a progeny of 102 isolates, hereinafter referred as to “V77.i.j”, was recovered. The progeny and parental isolates were tested for pathogenicity on B. napus cvs. Westar, and Yudal, and B. carinata D5.6.12 line (Figure 3). The disease scores in the progeny did not display a bimodal distribution corresponding to parental phenotypes, as expected for a monogenic inheritance of the phenotype (e.g., Balesdent et al., 2001), whatever the plant genotype. Instead, a continuous distribution of the mean disease scores, ranging from 1 ± 0 to 5.5 ± 0.57, was observed on B. carinata (Figure 3C). A continuous distribution of the mean disease scores was also observed on the two B. napus cultivars tested here (Figure 3A,B). These results suggest a polygenic determinism of the adaptation of L. maculans to B. carinata and B. napus. Symptoms of the V77.i.j progeny isolates showing a continuum of aggressiveness on (A) cv. Yudal, (B) cv. Westar of Brassica napus, and (C) Brassica carinata line D5.6.12. Infections and symptoms were evaluated on cotyledons following the IMASCORE rating scale (Balesdent et al., 2001). Dark blue bar, isolate JN3; green bar, isolate HB10.19; grey bars, V77.i.j progeny isolates, with the V77.1.11 isolate highlighted in yellow. The red star above the horizontal line indicates that progeny strains display significantly different symptoms on B. carinata compared to both parental isolates (p < 0.05). Among the 102 progeny isolates, 15 displayed a positive transgressive phenotype on B. carinata D5.6.12 (Wilcoxon test, p < 0.05); that is, they produced more severe symptoms on this genotype than the parental isolate HB10.19. Four progeny isolates were identified as able to cause severe symptoms on both B. napus (cv. Yudal) and B. carinata (line D5.6.12) with symptom scores ranging from 4.6 to 5.5 on B. carinata and from 4 to 4.9 on B. napus (Figure 3A,C). Among them, strain V77.1.11 was selected for further analysis; this progeny isolate was able to generate typical symptoms of L. maculans infection on cotyledons of both B. carinata and B. napus, with a greyish lesion, no black margin on B. carinata contrary to the parental isolate HB10.19, and extensive production of pycnidia on both species (Figure 2A,B). Screening of a collection of Brassica carinata accessions confirmed the nonhost status of this species toward Leptosphaeria maculans A collection of 52 B. carinata accessions recovered from the IPK genebank was screened with six isolates of L. maculans, including JN3, HB10.19, and the transgressive strain V77.1.11 (Table S1). All the L. maculans isolates failed to infect any B. carinata accession except for the genotype BRA_2124, which is suspected to be a cabbage based on the aspect of the cotyledons, on its susceptibility to the L. maculans isolates adapted to B. napus and on its resistance to both the parental isolate moderately adapted to B. carinata as well as to the transgressive isolate. Even though HB10.19 was able to moderately infect the B. carinata D5.6.12 line, all the IPK accessions tested here were fully resistant (with a mean score of 1.45) to this natural isolate. The transgressive progeny strain V77.1.11, induced susceptibility symptoms (i.e., mean score > 3) on 25 out of the 52 genotypes of B. carinata tested. However, a few B. carinata lines (n = 6, mean score < 2) were resistant to V77.1.11, as infection induced a rapid local cell death typical of a HR. Conversely, some accessions were fully resistant to HB10.19 (mean score < 1.5), while they were fully susceptible (mean score > 4) to V77.1.11. As all accessions of B. carinata tested here were resistant to the five natural L. maculans isolates (excluding the strain V77.1.11 originating from a cross in laboratory), this suggests an extremely resistant status of B. carinata toward field isolates. Cotyledon‐inoculations using two different approaches highlighted that only the selected progeny strain V77.1.11 gave rise to large susceptibility symptoms on Brassica carinata Two inoculation approaches were performed under controlled conditions: inoculation after wounding, and inoculation with a brush, using a pycnidiospore suspension (Section 2). For inoculations after wounding, the virulence symptoms caused by JN3 on B. napus cv. Yudal were characterized by big grey leaf spots on which pycnidia (i.e., black dots) developed 9–14 dpi (Figure 2A). On B. carinata line D5.6.12, the infection with JN3 resulted in the rapid appearance of local cell death, directly at the site of infection, typical of a HR caused by L. maculans on B. napus carrying a resistance gene (Figure 2A). The isolate HB10.19 could not infect cotyledons of B. napus, as displayed by the occurrence of a black necrosis, which was nevertheless more extensive than that caused by JN3 on B. carinata (Figure 2A). In contrast, HB10.19 was able to infect moderately B. carinata, causing a grey leaf spot with pycnidia surrounded by a black margin, suggesting that the fungus can initially spread within cotyledons of B. carinata but is eventually blocked in its colonization. The transgressive progeny strain, V77.1.11, was capable of infecting both Brassica species, causing extensive grey leaf spots with pycnidia on each species (Figure 2A). Symptoms observed following brush‐inoculated cotyledons were consistent with results obtained from inoculations following wounding: on cotyledons of B. carinata line D5.6.12, most infections with the L. maculans isolate adapted to B. napus were followed by the appearance of local cell death (for 10 out of 16 inoculated cotyledons; Figure S2); the isolate HB10.19 exhibited a moderate adaptation (out of 12 inoculated cotyledons, six exhibited moderate symptoms, including three showing the development of pycnidia, and six infections resulted in the development of a black symptom; Figure S3); and most infections with V77.1.11 exhibited typically adapted phenotype (out of 18 interactions, 15 developed typical virulent symptoms, including the development of pycnidia on 12 interactions, and one black symptom was identified; Figure S4). For brush‐inoculation of B. napus cv. Yudal, the results were also consistent with the wounded‐inoculation method, although fewer symptoms were overall observed (Figures [Link], [Link], [Link]). To further assess the interaction between the different isolates of L. maculans and B. napus or B. carinata, infection was monitored via microscopy analyses to follow the spread of the isolates in cotyledons of B. carinata or B. napus. Cotyledons of B. napus and B. carinata infected with the isolates JN2, HB10.19, or V77.1.11 were collected 7‐ and 15 dpi and stained with Trypan Blue (Section 2). For all interactions, pycnidiospores deposited on the surface of the cotyledon at the inoculation site after wounding were capable of germinating, and hyphae subsequently developed and were visible at 7 dpi (Figure 2B). At 15 dpi, adapted interactions (i.e., HB10.19 infecting B. carinata, JN2 infecting B. napus, V77.1.11 infecting B. napus or B. carinata) were characterized by extensive colonization of the cotyledon, development of pycnidia (visualized by the large dark round structure) and contacts between hyphae and stomata were observed, suggesting a possible penetration of stomata, as is the case in adapted/compatible interactions. In the case of a non‐adapted interaction (i.e., HB10.19 infecting B. napus, and JN2 infecting B. carinata), although limited hyphal growth was observed in the cotyledons, no pycnidia developed (Figure 2B). The transgressive strain V77.1.11 colonized extensively both B. napus and B. carinata cotyledons, produced large amounts of pycnidia 15 dpi, and was able to penetrate stomata (Figure 2B). Altogether, our observations showed that all isolates, adapted or non‐adapted, were able to initiate the first stages of infection (i.e., pycnidiospore germination, extensive hyphal growth on the cotyledon surface) yet the non‐adapted isolates failed to initiate their asexual reproduction, as is typically the case in adapted interactions, and seemed to be impaired in stomatal penetration. Although microscopic analyses show that the HB10.19 isolate was able to spread and infect successfully cotyledons of B. carinata, symptoms caused by this isolate on B. carinata were less prominent than those caused by the infection of B. napus by an adapted isolate, as a black margin surrounded the symptoms generated by HB10.19 (Figure 2A). The natural isolate moderately adapted to Brassica carinata, and the progeny isolate inducing severe susceptibility symptoms on cotyledons of both Brassica species show a partial adaptation to Brassica carinata We performed a biomass analysis to monitor fungal growth by harvesting cotyledons of B. napus and B. carinata infected by isolates JN3, HB10.19, and V77.1.11 at 5, 7, 9 and 12 dpi. On B. napus, the biomass of JN3 was very low 5‐ and 7‐dpi then increased between 7‐ and 12‐dpi to reach more than 270 ng of DNA in 10,000 ng of total DNA (Figure 2C). Following infection of B. carinata by JN3, the amount of fungal material remained very low throughout the course of the experiment (the fungal biomass reached 13 ng; Figure 2C). Likewise, the biomass of HB10.19 in infected cotyledons of B. napus remained low between 5‐ and 12‐dpi (ranging between 5.35 and 9.35 ng DNA for a total DNA amount of 10,000 ng; Figure 2C). The biomass of V77.1.11 on B. napus was lower than that of JN3 on the same host, although the pattern of colonization was similar to that of JN3 (a very limited biomass at 5‐ and 7‐dpi and an increase between 7‐ and 12‐dpi), suggesting that V77.1.11 was not able to develop as extensively as JN3 on B. napus (Figure 2C). On B. carinata, the amount of V77.1.11 increased at 12 dpi and reached 227 ng per 10,000 ng of total DNA at the highest, suggesting this fungus is better adapted to this species than the parent HB10.19. As for the development of this isolate on B. napus, although the fungal biomass increased, it did not reach the same level as in the interaction between JN3 and its host. However, V77.1.11 was able to generate a similar amount of biomass on cotyledons of both species. To summarize, there was an increase in fungal biomass during the adapted interactions, even though the increase was limited for interactions between HB10.19 and B. carinata. We also compared stem infections between adapted and non‐adapted interactions. Following inoculation of stems of B. carinata, no external or internal necrosis was observed with isolates JN3 and HB10.19 (Figure 4A). For the stem infection on B. napus, JN3 induced external necrosis on the stem (median symptom length, 13.5 mm; Figure 4B), and internal necrosis was systematically observed (Figure S8). On the contrary, HB10.19 caused no symptoms on the outer part of the stem of B. napus or B. carinata (Figure 4A,B). V77.1.11 caused typical adapted macroscopic symptoms on cotyledons of B. carinata and B. napus (Figure 2A), yet limited external necrosis was observed on both Brassica species (Figure 4), and inner symptoms were observed on only one of the 16 inoculated plants (data not shown). Altogether, biomass analysis and stem inoculations highlighted a limited ability to colonize B. carinata by HB10.19 and the transgressive strain V77.1.11. Symptoms of Leptosphaeria maculans on the stem basis of (A) Brassica carinata (line D5.6.12) and (B) Brassica napus (cv. Yudal). Symptoms were evaluated by the size of the external necrosis measured in millimetre. Dark blue, isolate JN3; green, isolate HB10.19; yellow, isolate V77.1.11. Letters ‘a’ and ‘b’ indicate the significant differences between the size of the external necrosis caused by each isolate on the stem of both Brassica species (p < 0.05). DISCUSSION In this study, we highlight the extreme resistance of B. carinata to L. maculans, which can be considered a nonhost species of L. maculans in our conditions. This result is supported by different experiments performed in the field for several years and in controlled conditions. The field experiments were performed in Northern France with B. carinata associated with diverse B. napus genotypes to maximize the likelihood of contact between natural L. maculans isolates and B. carinata. Macroscopic assessment of lesions developing on B. carinata cotyledons and transcriptomic analyses from leaves failed to identify evidence of colonization, even asymptomatic, of B. carinata by L. maculans. In controlled conditions, the 52 genotypes of B. carinata tested were resistant to the five tested genotypes of L. maculans. We challenged the resistance of B. carinata in controlled conditions with different approaches (i.e., different techniques of inoculation, using wounding or a brush, and different inoculum, either pycnidiospores or ascospores). We failed to recover L. maculans from the cotyledons of B. carinata with the “ascospore shower” method, although this method ensures direct ascospore ejection on the cotyledons. Moreover, although the moderately adapted isolate (HB10.19) and the transgressive adapted strain (V77.1.11) infect cotyledons of B. carinata, they fail to successfully infect the stem basis of B. carinata, showing that their adaptation to this species is only partial and restricted to specific organs. Stem infection is an important step of the L. maculans lifecycle on B. napus as necrotic lesions of the stem bases can eventually cause lodging, and passage of the fungus through the stem is essential for the fungus to complete its sexual cycle. In field conditions, these partially adapted strains would not be able to undergo sexual reproduction, and would not be maintained in the population of L. maculans isolates. Therefore, the absence of sexual reproduction might prevent the genetic determinism allowing adaptation to be maintained in the population. The infection assays within the L. maculans/B. napus pathosystem in controlled conditions are routinely performed using miniaturized assays in which cotyledons are inoculated with a droplet of pycnidiospore suspension after wounding (Badawy et al., 1991; Mengistu et al., 1991). This miniaturized test is rapid and easy to handle. Nevertheless, it bypasses the natural penetration of the host via stomata. Therefore, we also performed brush inoculation to compare symptoms obtained following different inoculation methods. This assay has demonstrated that, even without wounding, L. maculans pycnidiospores can induce visible HR symptoms on B. carinata, indicating that B. carinata displays a type II‐NHR toward L. maculans. In addition, during the interaction between any of the natural isolates of L. maculans and B. carinata (line D5.6.12) tested in field experiments, the rapid development of a local black necrotic symptom was identified, as observed during a hypersensitive reaction within incompatible interactions between L. maculans and resistant B. napus genotypes (Hammond & Lewis, 1987; Šašek et al., 2012). This response suggests that the fungus can penetrate the plant stomata but is actively recognized by the plant defence mechanism, preventing the fungus from spreading further. Type II‐NHR is a post‐invasive resistance characterized by a hypersensitive reaction. It is suggested to be triggered by the response to pathogen elicitors, as is effector‐triggered immunity (ETI) (Gill et al., 2015). Type II‐NHR involves active recognition and defence of the plant and relies on NB‐LRR (primarily R genes) or genes involved in hormone defence pathways (Fonseca & Mysore, 2019). We postulate that B. carinata type II‐NHR, observed during infection of natural L. maculans isolates could be based on the recognition of L. maculans effector proteins by B. carinata resistance proteins. Moreover, the effector repertoire is one of the key determinants of pathogen host range, including adaptation to different host genotypes or host species (Dong et al., 2015; Frantzeskakis et al., 2018; Ma et al., 2010; Raffaele et al., 2010; Spanu et al., 2010). The continuum of aggressiveness caused by the V77 progeny strains on B. carinata suggests that the rearrangement of traits, including effectors, between isolates JN3 and HB10.19 allows the progeny strains to be better adapted to B. carinata. These progeny strains may lack effectors allowing them to escape the plant recognition. The advantage of type II‐NHR is that the deciphering of the underlying defence mechanisms in plants displaying such resistance can be used to identify control methods against pathogens. Among the B. carinata genotypes recovered from the IPK and tested here, we identified accessions displaying susceptibility toward the transgressive strain V77.1.11 while all are resistant to the parental isolates used for the cross. Understanding the mechanisms underlying the extreme resistance of B. carinata will enable identification of new sources of resistance that may be more durable than the host resistances against stem canker usually used in breeding programs. In the B. carinata/L. maculans interaction, microscopic analysis evidences that pycnidiospores germinate at the surface of B. carinata cotyledons both in the adapted and the non‐adapted interactions. In adapted interactions, contact between hyphae and stomata were observed, suggesting penetration of B. carinata stomata, as is the case for adapted isolates on B. napus. Yang and Fernando (2021) performed microscopic analyses between various isolates of L. maculans and different genotypes of B. napus, showing that an adapted interaction is associated with hyphal growth, colonization of intercellular spaces, and pycnidia formation, consistent with previous analyses (Hammond et al., 1985; Li et al., 2008). In contrast, neither pycnidia formation nor extensive necrosis was observed in the non‐adapted interactions, as in an incompatible interaction. Chen and Howlett (1996) showed that for an incompatible interaction between L. maculans and B. napus, guard cell necrosis and tissue browning appear when hyphae have reached the mesophyll layer. We observe here that L. maculans is still viable on the surface of cotyledons in non‐adapted interactions, however, further microscopic analyses are required to better characterize the nonhost interaction between L. maculans and B. carinata and to identify when the infection process is impaired. Among the 53 natural isolates tested on B. carinata in controlled conditions, we identified four isolates from our collection that could moderately infect one line; among these, HB10.19 was selected for further analysis. This isolate was recovered from B. nigra leaves in the field. B. carinata results from natural hybridization between B. oleracea and B. nigra, ca. 11,000–29,900 years ago (Yim et al., 2022); hence B. nigra and B. carinata share the B‐genome. It has been demonstrated that the B‐genome is an important resource for high level of resistances against biotic and abiotic stresses, including stem canker (Christianson et al., 2006; Fredua‐Agyeman et al., 2014; Rashid et al., 2018; Yim et al., 2022). Further analysis of the B. carinata genome will undoubtedly pinpoint genomic features associated with this resistance. To investigate the extreme resistance of B. carinata toward L. maculans, we performed a cross between a reference isolate JN3, adapted to B. napus, and the isolate HB10.19, moderately adapted to B. carinata. In the offspring of the cross, several transgressive strains were identified, showing greater aggressiveness on one of the two Brassica species than the parental isolate adapted to this Brassica. Extreme phenotypes within progenies have critical roles in species evolution (Rieseberg et al., 1999). Four progeny isolates were able to cause symptoms on both Brassica species. These transgressive phenotypes were also observed when we quantified the amount of biomass of the three L. maculans isolates JN3, HB10.19, and V77.1.11 on B. napus and B. carinata. Although V77.1.11 produces more biomass than HB10.19 on B. carinata during infection, this biomass does not reach the biomass of L. maculans adapted to B. napus. Transgression segregation has been well described in plants (Rieseberg et al., 1999, 2003; Stelkens & Seehausen, 2009), nevertheless, to date, only a few transgressive phenotypes were formally laid out in filamentous fungi (Caffier et al., 2022; Cumagun et al., 2004; Gibson et al., 2014; Voss et al., 2010). The transgressive isolates obtained in the progeny indicate that after only one sexual reproduction event, it is possible to generate strains of L. maculans adapted to B. carinata cotyledons. We hereby demonstrate that the adaptation of L. maculans to B. carinata is under a polygenic determinism. Our present study represents a valuable biological material to investigate underlying factors of the adaptation of L. maculans to different Brassica species. The transgressive strains and the cross between JN3 and HB10.19 will allow us to identify the genetic mechanisms underlying adaptation of L. maculans to different hosts and identify whether adaptation to nonhost resistance of B. carinata involves the same underlying molecular events as those involved in the circumvention of B. napus resistance genes. Despite NHR being considered more durable than host resistance (Ayliffe & Sørensen, 2019; Fonseca & Mysore, 2019), it is not insurmountable. Even though B. carinata is not a widely cultivated species, it has been introduced into North America and Argentina for industry uses (Ban et al., 2017; Song et al., 2021; Taylor et al., 2010). The sympatry of B. napus and B. carinata may favour a host shift of L. maculans, or an expansion of its host range, toward B. carinata. Several pathogens have adapted due to the proximity of their host with another plant species (Bettgenhaeuser et al., 2014; Xia et al., 2018). Consequently, the risk of emergence of a L. maculans population adapted to B. carinata must be considered if oilseed rape crops are grown alongside B. carinata crops. Besides, environmental conditions (i.e., temperature, photoperiod, humidity, and developmental stage of the plant) need to be considered and can affect the nonhost status of a plant toward pathogens (Nnadi & Carter, 2021; Panstruga & Moscou, 2020). The context of global climate change represents an increasing threat of the emergence of new and better‐adapted pathogen species (e.g., see Nnadi & Carter, 2021). Thus, global climate change could also create conditions favouring host shifts in L. maculans. Altogether, our analyses show that the L. maculans/B. carinata pathosystem is a promising nonhost framework to decipher the underlying mechanisms responsible for the adaptation of L. maculans to different hosts, and to investigate the genetic determinants involved in the NHR of B. carinata toward L. maculans. Nevertheless, we highlighted the risk of the emergence of better‐adapted isolates to this Brassica species, which might be emphasized by agricultural practices and climate change. AUTHOR CONTRIBUTIONS Julie M. Noah: Investigation; writing – original draft; validation; visualization; writing – review and editing. Mathilde Gorse: Investigation; validation; writing – review and editing. Carole‐Anne Romain: Investigation; validation; visualization; writing – review and editing. Elise J. Gay: Investigation; validation; visualization; writing – original draft; writing – review and editing. Thierry Rouxel: Conceptualization; investigation; funding acquisition; writing – original draft; writing – review and editing; validation. Marie‐Hélène Balesdent: Conceptualization; investigation; funding acquisition; writing – original draft; validation; visualization; writing – review and editing; supervision. Jessica L. Soyer: Conceptualization; investigation; funding acquisition; writing – original draft; validation; visualization; writing – review and editing; supervision. CONFLICT OF INTEREST STATEMENT The authors declare no conflicts of interest. Supporting information FIGURE S1. Symptoms recovered on leaves of Brassica napus (cv. Darmor) and Brassica carinata (line D5.6.12) in a field experiment. Leaves of (A) B. napus cv. Darmor and (B) B. carinata line D5.6.12. Examples of symptoms observed on leaf disks collected 2.5 months after sowing on (C), (D) B. napus and (E), (F) B. carinata. Symptoms identified from disks (C), and (E) were suspected to be caused by L. biglobosa; while (D) and (F) were suspected to be caused by L. maculans. FIGURE S2. Symptoms observed on Brassica carinata line D5.6.12 following brush inoculation with the isolate JN2 of Leptosphaeria maculans. The upper line displays details of the experiment; with, from left to right, JS_2018_3, the reference of the inoculation; the dates of start and end of the experiment (i.e., dates of inoculation and observation); AF425, the reference of the B. carinata line used for the inoculation (i.e., line D5.6.12); Line 1, the line of the plants used for inoculation; JN2, the name of the L. maculans isolate brush‐inoculated on the cotyledons (Section 2). FIGURE S3. Symptoms observed on Brassica carinata line D5.6.12 following brush inoculation with the isolate HB10.19 of Leptosphaeria maculans. The upper line displays details of the experiment; with, from left to right; JS_2018_3, the reference of the inoculation; the dates of start and end of the experiment (i.e., dates of inoculation and observation); AF425, the reference of the B. carinata line used for the inoculation (i.e., line D5.6.12); Line 2, the line of the plants used for inoculation; HB10.19, the name of the L. maculans isolate brush‐inoculated on the cotyledons (Section 2). FIGURE S4. Symptoms observed on Brassica carinata line D5.6.12 following brush inoculation with isolate V77.1.11 of Leptosphaeria maculans. The upper line displays details of the experiment; with, from left to right; JS_2018_3, the reference of the inoculation; the dates of start and end of the experiment (i.e., dates of inoculation and observation); AF425, the reference of the B. carinata line used for the inoculation (i.e., line D5.6.12); Line 3, the line of the plants used for inoculation; V77.1.11, the name of the L. maculans isolate brush‐inoculated on the cotyledons (Section 2). FIGURE S5. Symptoms observed on Brassica napus cv. Yudal following brush inoculation with the isolate JN2 of Leptosphaeria maculans. The upper line displays details of the experiment; with, from left to right; JS_2018_3, the reference of the inoculation; the dates of start and end of the experiment (i.e., dates of inoculation and observation); Yudal, the reference of the B. napus cultivar used for the inoculation; Line 4, the line of the plants used for inoculation; JN2, the name of the L. maculans isolate brush‐inoculated on the cotyledons (Section 2). FIGURE S6. Symptoms observed on Brassica napus cv. Yudal following brush inoculation with the isolate HB10.19 of Leptosphaeria maculans. The upper line displays details of the experiment; with, from left to right; JS_2018_3, the reference of the inoculation; the dates of start and end of the experiment (i.e., dates of inoculation and observation); Yudal, the reference of the B. napus cultivar used for the inoculation; Line 5, the line of the plants used for inoculation; HB10.19, the name of the L. maculans isolate brush‐inoculated on the cotyledons (Section 2). FIGURE S7. Symptoms observed on Brassica napus cv. Yudal following brush inoculation with the isolate V77.1.11 of Leptosphaeria maculans. The upper line displays details of the experiment; with, from left to right; JS_2018_3, the reference of the inoculation; the dates of start and end of the experiment (i.e., dates of inoculation and observation); Yudal, the reference of the B. napus cultivar used for the inoculation; Line 6, the line of the plants used for inoculation; V77.1.11, the name of the L. maculans isolate brush‐inoculated on the cotyledons (Section 2). FIGURE S8. Internal necrosis caused by three Leptosphaeria maculans isolates on stems of Brassica napus and Brassica carinata. Transversal sections of stems of B. napus, upper panel, and B. carinata, lower panel; (A) and (E) uninfected stems; (B) and (F) JN3‐infected stems; (C) and (G) HB10.19‐infected stems; (D) and (H) V77.1.11‐infected stems. Uninfected stems were used as a negative control. TABLE S1. Accessions of Brassica carinata retrieved from the IPK genebank and tested for their resistance toward different isolates of Leptosphaeria maculans. aAcquisition date from the IPK Genebank. bCountry of origin: ETH, Ethiopia; PAK, Pakistan; SWE, Sweden; ZMB, Zambia. cMean of the symptoms assessed using the IMASCORE rating scale (Balesdent et al., 2001). dOrigin of the L. maculans isolates: HB10.19 and V77.1.11, this study; IBCN14 and IBCN80 are reference isolates of L. maculans included in the International Blackleg of Crucifer Network collection (Balesdent et al., 2005). TABLE S2. List and characteristics of Leptosphaeria maculans isolates used and their pathogenicity on Brassica napus and Brassica carinata. aSymptoms were evaluated on 10–12 plantlets of B. napus and B. carinata and were evaluated using the IMASCORE rating scale with six infection classes (IC; Balesdent et al., 2001) in which we have taken into consideration the “intermediate resistance phenotypes” (Neik et al., 2022; Van de Wouw et al., 2009; Yang & Fernando, 2021). Symptoms were scored nine and 12–13 days post‐inoculation; the symptoms with IC < 2 indicates that the isolate is avirulent (hence the plant is resistant), IC 4–6 indicate that the isolate is virulent (hence the plant is susceptible), IC 2 to < 4 are characteristic of the intermediate interaction phenotype. bJN2 and JN3 are sister isolates originating from a cross performed in the lab (Attard et al., 2002). TABLE S3. Primers used for PCR and qPCR amplification in this study. For, forward primer; Rev., reverse primer. Probes are used to improve the specific quantification of a species (Jacques et al., 2021).