Wheat leaf rust (Puccinia triticina) belongs to the economically most important obligate biotrophic pathogens of wheat (Bolton et al., 2008a; Bolton et al., 2008b; Kolmer, 2013). It is the causative agent of leaf rust, the most common rust of wheat worldwide reducing number of grains and thousand-grain weight, resulting in yield losses of up to 60% (Bancal et al., 2007; Bolton et al., 2008b; Kolmer, 2013). Uredinia, the typical leaf rust fruiting bodies formed during the asexual life cycle, occur on the upper surface and bottom side of leaves on susceptible wheat cultivars with a diameter of up to 1.5 mm (Kolmer, 2005; Kolmer, 2013). These uredinia harbor dikaryotic uredospores of approximately 20 µm. When the leaf epidermis ruptures, the orange-yellow uredospores are spread by the wind to infect new host plants under favorable conditions (i.e., 10°C to 20°C and high humidity) (Bolton et al., 2008a).

At the beginning of a successful penetration process, uredospores germinate 4 to 8 hours after inoculation (hai) and form a germ tube and an appressorium over stomata cells (Bolton et al., 2008a). Within 12–24 h after the formation of the anppressorium, an infection vesicle is generated, and infection hyphae grow between parenchymal cells and form haustorial mother cells (hmc) on the cell walls of mesophyll cells (Bolton et al., 2008b). Next, 24 h after the formation of an appressoria, haustoria start to develop (Serfling et al., 2016). They penetrate the host cells and generate an extrahaustorial membrane (Bolton et al., 2008b).

Race-specific resistance is known to be effective “posthaustorial”, thus after the formation of haustorial mother cells (Niks, 1988; Niks, 1991). Hypersensitive cell death is triggered by a gene-for-gene recognition of effectors (Ji et al., 2022, Kumar et al., 2021) and is contrary to adult plant resistance (APR) already active at the seedling stage. Moreover, race-specific resistance is vulnerable to breakdown by virulent races, which occurs after a mutation of an elicitor or the interacting resistance gene in the host. More than 100 leaf rust resistance genes (Lr-genes) have been described; however, only a few are carried by cultivars due to linkage drag and undesired agronomic properties (Mapuranga et al., 2022). The lifespan of such vertical/qualitative resistance carried by cultivars was calculated by Mapuranga et al. (2022) at 3 to 5 years. Hence, a horizontal/race-independent resistance that has the character of nonhost resistance to leaf rust as a host-specific pathogen could be more durable. This nonhost–pathogen interaction was exemplarily described for barley—P. triticina (Jafary et al., 2008; Haghdoust et al., 2021) and wheat—Blumeria hordei, P. hordei, and Magnaporthe oryzae (Delventhal et al., 2017). Nonhost interactions result in prehaustorial resistance (phr) and have also been observed in T. monococcum–P. triticina interactions (Serfling et al. 2016; Anker et al. 2001; Anker and Niks, 2001). Various studies have shown a different expression of pathogenesis-related (PR-) genes, peroxidases, chitinases, peroxidases, and beta 1,3 glucanase within the first 24 hai. However, the inheritance of nonhost resistance of T. monococcum to P. triticina has not yet been genetically analyzed.

Triticum monococcum accession PI272560 (Serfling et al., 2016) shows complete and nonhost resistance to six investigated leaf rust ( Tables S1A, B ) races (Serfling et al., 2016) and one leaf rust race tested by Niks (1991). This resistance was previously identified as a phr in which leaf rust develops few or no haustorial mother cells after infection because their formation is prevented by an effective defense reaction of the plant (Anker et al., 2001). Serfling et al. (2016) indicate an increased level of phenolic substances, peroxidase, and chitinase activity at the site of infection and pathogenesis-related genes in the first 24 hai in comparison to the partially susceptible T. boeoticum accession (36554). These results demonstrate transcriptome alterations and resistance mechanisms in the background of phr in PI272560. However, the genetic background and the inheritance of this resistance were not elucidated up to now. Therefore, this study aims to identify the genomic regions and candidate genes involved in the phr of T. monococcum based on microscopic analysis of fungal development, the defense reaction, and visual rating in a biparental F₂ mapping population derived from T. monococcum accession PI272560 and T. boeoticum accession 36554. These investigations are complemented by data from transcriptome analysis by massive analysis of cDNA ends (MACE) of the parental accessions to be mapped into QTL regions.

Today, only a few known Lr-genes are used in wheat varieties. Most are seedling resistances, which are generally vulnerable to being broken down by races with a changed virulence pattern (Periyannan et al., 2017; Zetzsche et al., 2020). Only a few non-race-specific resistances have been described, for instance, Lr34 and Lr67, which are quantitative and active at the adult plant development stages only (Hou et al., 2023). In most cases, obligate biotrophic fungal pathogens such as leaf rust are strictly host-specific. Therefore, one possibility to introduce durable leaf rust in wheat cultivars is single genes from alien species. Examples of these are the stem rust resistance genes Sr31, powdery mildew resistance Pm21, and resistance to eyespot disease resistance Pch1 (Ellis et al., 2014; Wulff and Moscou, 2014). Alien introgressions from rye (Secale cereale) contribute several resistance genes for powdery mildew, leaf, stripe, and stem rust, for example, Yr9, Lr25, and Lr26 (Johannson et al., 2020).

Hybridization between wheat and its wild relatives Aegilops sp., Triticum timopheevii, and Thinopyrum ponticum (Keilwagen et al., 2019) occurs naturally and is conducted during the breeding process. However, only a few rust resistances have been introgressed from diploid T. monococcum (Yr34, Chen et al., 2021; Sr22, Kerber and Dyck, 1973). A transfer of leaf rust resistance to wheat was reported by Hussien et al. (1997), whereas Noweiska et al. (2022) and Feuillet et al. (2003) showed that Lr10 is conserved in grass species with similarities to RPM1 in A. thaliana. Most T. monococcum accessions (84%) show a high level of resistance in contrast to most T. boeoticum accessions, including the partially susceptible T. boeoticum accession 36554 (Anker et al., 2001).

Hussien et al. (1997) analyzed three leaf rust resistances derived from T. monococcum, but did not identify their respective genetic positions. Two of them, Lr63 and LrTM16, were mapped on chromosome 3A and 2A in T. monococcum so far (Sodkiewicz et al., 2008; Noweiska et al., 2022). Loci associated to leaf rust resistances on chromosome 5A were identified in wheat based on the Bavarian MAGIC wheat (BMW) population (QLr.jki-5A.1; QLr.cim-5AC, Rollar et al., 2021), in a DH population derived from the Canadian wheat cultivar Carberry (Bokore et al., 2021) and the cultivar Lillian (Bokore et al., 2023). Remarkably, these QTLs explain only a low level of phenotypic variance with regard to the resistance level. Furthermore, they are either effective at the seedling or adult plant stage. In contrast to that, the resistance in Pi272560 is effective at all developmental stages ( Figure 2C ). To our best knowledge, no resistance with such a high efficiency was described on chromosome 5A so far. To understand the background of the prehaustorial leaf rust resistance of T. monococcum accessions, the development of fungal structures was analyzed (Jacobs et al., 1996). Expression studies and microscopical analyses have been performed in different studies (Anker et al., 2001; Sánchez-Martín et al., 2012; Serfling et al., 2016). Recognition of the pathogen by resistant accessions and first defense reactions based on hydrogen peroxide accumulation and antifungal compounds could already be observed 6 hai, so that the generation of hmc, haustoria, hyphae, and pustules was inhibited (Anker et al., 2001; Serfling et al., 2016). The number of hmc at 72 hai and rating data from the F₂ genotypes showed right-skewed or binominal distribution but no segregation known for a single gene in the background. Furthermore, due to a possible nonhost resistance of T. monococcum to wheat leaf rust, an identification of different loci from both parents was expected since 36554 turned out to be partially resistant against P. triticina as well (Serfling et al., 2016).

A QTL analysis was performed to identify the effects of both genotypes on the segregating F₂ population. The comparison of parental lines during the first 24 hai revealed a complex defense reaction comprising various mechanisms leading to an inhibition of the infection process. In accordance with previous studies of the host–pathogen interaction, a higher expression of genes known to be involved in the reaction to leaf rust could be observed. The comparison of the PI272560 and 36554 transcriptome showed clearly that pathogenesis-related genes such as Pr1, β-1,3-glucanases (Pr2), chitinases (Pr3), peroxidases (Pr9), and other Pr-genes were significantly more expressed in PI272560 after inoculation with wheat leaf rust ( Table S7 , Serfling et al., 2016). Remarkably, however, these genes are not located within the corresponding genomic interval (Ling et al., 2018) of the QTLs detected in the course of our study ( Table 4 ). Hence, transcriptome analysis provided information regarding the different expressions of genes but not necessarily about the actual resistance gene in the background of an effective (prehaustorial) resistance.

Such a resistance includes an early onset of hypersensitive response (HR), which could be triggered by genes involved in upstream metabolic processes. This study aimed to combine a MACE approach with the construction of genetic mapping and QTL detection to identify the actual candidate genes for the observed resistance against leaf rust. As expected from the partial resistance of accession 36554, a minor QTL of accession 36554 was identified on chromosome 7A ( Table 4 ), resulting in a reduced number of hmc 48 hai ( Figure 2D ). However, this QTL is of limited importance as it has no impact on pustule development ( Figure 2F ). One reason is most likely the host specificity of leaf rust to wheat, while T. monococcum is almost a nonhost for P. triticina (Dracatos et al., 2018).

Complex defense reactions could be observed shortly after the inoculation with leaf rust together with a high number of differentially expressed genes. According to our results, in a nonhost reaction of wheat to barley leaf rust, Delventhal et al. (2017) could detect 2,498 differentially expressed genes, while Serfling et al. (2016) could identify 311 different defense-related genes. However, these findings offer an overview of the entire transcriptome after infection and do not narrow down regions of the genome where genes are linked to the resistance response. Consequently, an F₂ population of 182 F₂ plants segregating for the hmc generation 48 and 72 hai and a visual rating are suitable to detect QTL resistance.

Remarkably, one QTL that could be detected on chromosome 5A from PI272560 ( Table 4 ) appears to have a major effect on the resistance level ( Figure 2 ). This effect was confirmed within the three different maps and phenotypic data, the hmc (72 hai), and rating data ( Table 4 , Figure 3 ). It is conceivable that owing to the recessive nature of the resistance ( Figure 2C ), the QTL was detected more clearly in the map showing dominant 36554 markers ( Table 4 ) and not in the PI272560 map since repulsion effects of the recessive alleles, within the QTL and the dominant PI272560 markers, might have hampered the QTL identification. Interestingly, the QTL explains the reduced hmc generation of approximately 26% of the phenotypic variance ( Table 4 ). However, almost all—49 of 50 plants—were homozygous for the PI272560 allele of the marker SNP_1364455 and showed phr ( Figure 2C ). The QTLs’ percentage of explained variance (P.E.V.) was likely underestimated because genotypes, being homozygous for the 36554 allele of marker SNP_1364455, were not clearly susceptible ( Figure 2C ). We know that 36554 is not completely susceptible to P. triticina (Serfling et al., 2016); thus, this observation is unsurprising. However, the underestimation of the P.E.V. value in this study is a useful example of how the actual effect of a QTL is determined depending on the parent’s resistance properties.

The identification of the QTL region on chromosome 5A reduced the number of possible candidate genes to 217. After the selection of exclusively expressed tags ( Table 5 ) only six genes are related to defense responses to biotic stress. From these genes, one, coding for a microtubule binding protein (exclusively expressed in PI272560), is known to be involved in hypersensitive response including an accumulation of hydrogen peroxide in an incompatible interaction between wheat and wheat stripe rust (Wang et al., 2016).

The highest tpm could be detected constitutively in 36554 expressed uncharacterized protein with unknown function. Other genes, for instance, coding for a Wall-associated receptor kinase 3 are described as activators of signal cascades and have been identified as involved in leaf rust resistance comparable to APR. Hence, a significant impact on resistance at the seedling stage could not be expected in our investigation. Finally, the greatest differences in expression could be observed between the parental genotypes for an AT-hook motif nuclear-localized protein 10, which was 25.2 times higher expressed in 36554 and a BBE, which was 95.9 times higher expressed in PI272560 at 8 hai than in 36554 ( Figures 4 – 6 ).

BBEs belong to the flavin-dependent oxidoreductases and are described as enzymes that can interact with damage-associated molecular pattern (DAMP) and initiate hypersensitive reactions (Locci et al., 2019). Daniel et al. (2017) termed BBEs in A. thaliana a “treasure trove of oxidative reactions”. A higher expression of BBE was also observed in a nonhost response of barley to wheat powdery mildew (Andrzejczak et al., 2020). In the case of the coffee Hemileia vastatrix (coffee rust) interaction, this enzyme could be identified as a biomarker for the initial phr against the fungus (Guerra-Guimarães et al., 2015; Silva et al., 2022). According to the results of Serfling et al. (2016), during the T. monococcum–wheat leaf rust interaction, as a defense reaction within the first 24 hai in coffee, increased peroxidase activity and PR-like proteins, for instance, chitinases, were detected. The BBE might initiate and trigger hypersensitive cell death, as observed by Serfling et al. (2016). The early time of higher expression could also explain the phr with a strongly reduced number of haustorial mother cells in resistant genotypes of the F₂ population ( Figures 2A–C ).

Hence, the BBE could be one key enzyme for the basal defense response (Guerra-Guimarães et al., 2015), but is also an important enzyme in (nearly) nonhost resistance (Andrzejczak et al., 2020; Wan et al., 2021). As already mentioned by Anker et al. (2001), T. monococcum phenotypically shows almost nonhost resistance to wheat leaf rust. A typical sign of nonhost resistance to leaf rust is an early start of effective resistance reactions before the formation of the haustorium (Anker et al., 2001). In accession PI272560, such reactions and no haustoria could be observed (Serfling et al., 2016), but a BBE could not be detected in the study due to the restriction to specific gene ontology terms. To find suitable markers and detect genes in the background of the phr, a segregating population is a prerequisite. If molecular-assisted selection (MAS), for example, the SNP marker SNP_1364455, could be applied to transmit the resistance of PI272560 in current wheat elite varieties, this would be a possibility to establish a potential nonhost resistance with very long persistence and race independence. Examples of successful alien gene transmission of resistance genes into wheat are, for instance, the stem rust resistance genes Sr21, Sr22 (The, 1973), Sr35 (McIntosh et al., 1995), a leaf rust resistance (Hussien et al., 1997; Anker et al., 2001), and the powdery mildew resistance gene PmTmb (Shi et al., 1998).

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

MD and AS established the plant material, performed the experiments, and wrote the manuscript. AS and FO conceived the basic idea of this research project and helped to improve the manuscript. All authors contributed to the article and approved the submitted version.