Research advances in the Pyrenophora teres–barley interaction

Summary Pyrenophora teres f. teres and P. teres f. maculata are significant pathogens that cause net blotch of barley. An increased number of loci involved in P. teres resistance or susceptibility responses of barley as well as interacting P. teres virulence effector loci have recently been identified through biparental and association mapping studies of both the pathogen and host. Characterization of the resistance/susceptibility loci in the host and the interacting effector loci in the pathogen will provide a path for targeted gene validation for better‐informed release of resistant barley cultivars. This review assembles concise consensus maps for all loci published for both the host and pathogen, providing a useful resource for the community to be used in pathogen characterization and barley breeding for resistance to both forms of P. teres.

One P. teres form is often dominant within a geographical area. P. teres f. maculata has become more prevalent in North Dakota (Liu and Friesen, 2010) and Idaho, USA (Marshall et al., 2015) and Victoria, Australia (McLean et al., 2010a), but this likely depends on the susceptibility of the predominant cultivars planted in each region. Multiple differential barley sets have been proposed to classify global pathotypes of P. teres f. teres (Afanasenko et al., 2009(Afanasenko et al., , 1995Fowler et al., 2017) and P. teres f. maculata isolates in local populations (McLean et al., 2014(McLean et al., , 2010b. The taxonomy, population diversity, symptoms, life cycle, infection cycle and control of, and preliminary insights into the P. teres-barley interaction were previously reviewed by Liu et al. (2011), and therefore these aspects will not be covered here. In this review, we have collated the recent literature that lays the foundation for the functional characterization of P. teres-barley interactions.

Genetics of barley reactions to P. teres f. teres
Resistance to P. teres f. teres was first shown to be quantitatively inherited by Geschele (1928). Schaller and Wiebe (1952) screened barley lines for resistance to P. teres, but it is unknown which form was used, concluding that lines, primarily of Manchurian origin, contained enough resistance for introgression into elite barley lines. The first resistance locus implicated in the P. teres-barley interaction was harboured by Tifang in a Tifang × Atlas cross, inherited as a single gene trait and effective against Californian P. teres isolates (Schaller, 1955). Mode and Schaller (1958) subsequently designated this locus Pt1 and identified two additional loci, designated Pt2 and Pt3 (Table 1). The lines Ming, Manchurian and Harbin harboured Pt2 as a single locus, whereas the lines Canadian Lake Shore and CI4922 carried Pt2 and Pt3 (Mode and Schaller, 1958). The Pt1 and Pt2 loci were shown to be closely linked with 2.57% recombination by using resistant by resistant crosses (Mode and Schaller, 1958). An additional locus, designated Pt a , was discovered conferring resistance to the Australian isolate W.A.-2 that was distinct from Pt1, Pt2 and Pt3 (Khan and Boyd, 1969a). Khan and Boyd (1969a) concluded that the genetic makeup of resistance in Harbin, Manchuria, Tifang and Ming was different depending on the isolate used based on the existing research at the time. Following the trisomic analysis of Bockelman et al. (1977) that failed to separate Pt1 and Pt2, and the revision of the barley gene nomenclature, the Pt1 and Pt2 loci were collapsed into the Rpt1 locus on the long arm of chromosome 3H with Rpt1.a and Rpt1.b alleles from Tifang/Harbin and CI9819, respectively (Table 1). Additional subsequent loci that were collapsed into the Rpt1 locus included Pta, QRpts3L and QNFNBSLR.Al/S-3H from Igri (Graner et al., 1996), Arapiles  and Alexis (Lehmensiek et al., 2007), respectively (Table 1). The chromosome 3H resistance locus in Canadian Lake Shore mapped to the short arm of chromosome 3H (Dinglasan et al., 2019), indicating that the Canadian Lake Shore resistance is not Rpt1 (Fig. 1). Confusion over the Pt2 locus occurred when Rpt2.c reported by Bockelman et al. (1977) was assumed to be Pt2 reported by Mode and Schaller (1958). However, Rpt2.c in CI9819 is distinct from Pt2, which is allelic to other sources of resistance that map to the Rpt1 locus (Bockelman et al., 1977) (Table 1). Manninen et al. (2006) reported a major and a minor effect locus effective against different P. teres f. teres isolates on chromosomes 6H and 1H, respectively, in a Rolfi × CI9819 population, corroborating Rpt2.c on chromosome 1H (Table 1).
The Rpt5 locus has been reported by multiple studies and is considered to be a complex locus that is highly important in the P. teres f. teres-barley interaction (Abu Qamar et al., 2008;Adawy et al., 2013;Adhikari et al., 2019;Cakir et al., 2011Cakir et al., , 2003Dontsova et al., 2018;Emebiri et al., 2005;Friesen et al., 2006;Grewal et al., 2012Grewal et al., , 2008Gupta et al., 2011Gupta et al., , 2010Koladia et al., 2017a;Liu et al., 2015;Ma et al., 2004;Manninen et al., 2000Manninen et al., , 2006Martin et al., 2018;Molnar et al., 2000;Novakazi et al., 2019;O'Boyle et al., 2014;Raman et al., 2003;Rau et al., 2015;Richards et al., 2016;Richter et al., 1998;Rozanova et al., 2019;Spaner et al., 1998;St. Pierre et al., 2010;Steffenson et al., 1996;Wonneberger et al., 2017bWonneberger et al., , 2017a. Abu Qamar et al. (2008) found that Rika was resistant to P. teres f. teres isolate 15A and susceptible to isolate 6A, whereas for Kombar the reciprocal reactions occurred in that Kombar was resistant to 15A and susceptible to 6A. The two alleles rpt5.k. and rpt5.r were originally designated as recessive resistance due to a 1:3 resistant:susceptible segregation ratio and therefore the requirement of a homozygous state of alleles for resistance (Abu Qamar et al., 2008). The Rpt5 region Rpt4 (Williams et al., 1999(Williams et al., , 2003 QRpt7 (Grewal et al., 2008) QNFNBAPR Rpt4.e 7H P. teres f. teres P. teres f. maculata Rpt5 (Manninen et al., 2006) Spt1 (Richards et al., 2016) Pt a (Khan and Boyd, 1969a) Pt,,d (Graner et al., 1996) QRpts6L  QRpt (Embiri et al., 2005) rpt.r/rpt.r (Abu Qamar et al., 2008) QRpt6 (Grewal et al., 2008(Grewal et al., , 2012  that conferred susceptibility was delimited to a c. 5.9 cM region in a 118-progeny doubled haploid (DH) population derived from Rika × Kombar (Abu Qamar et al., 2008). While attempting to map the susceptibility loci in Rika/Kombar, a single Rika × Kombar progeny line was resistant to both 15A and 6A, and a single progeny line was susceptible to both 15A and 6A. These putative recombinants led to the hypothesis that two closely linked genes, rpt.r and rpt.k were held in repulsion c. 1.8 cM apart and were responsible for susceptibility (Abu Qamar et al., 2008). Under a two-gene scenario, rpt.r and rpt.k would each be responsible for initiating a susceptible response to an effector present in each parental isolate (Abu Qamar et al., 2008). Using the same Rika × Kombar DH population (Abu Qamar et al., 2008), 15 expressed sequence tag-derived markers were added to the genetic map to refine the chromosome 6H region to a c. 3.3 cM genetic interval and the two suspected genes to a c. 1.6 cM genetic interval (Liu et al., 2010). Liu et al. (2010) concluded that the rpt.r/rpt.k region was most likely located close to the centromere on the   long arm of chromosome 6H with high synteny to chromosomes 2 and 3 of rice and Brachypodium, respectively. The annotation list of the putative rpt5 region contained two leucine-rich repeat (LRR) receptor-like kinases, which are often associated with plant disease resistance (Liu et al., 2010). In addition, the Rpt5 locus has also been designated Susceptibility to P. teres 1 (Spt1), indicating the dominant nature of susceptibility to P. teres f. teres in some barley lines (Richards et al., 2016) (Table 1). The standard nomenclature indicated Reaction to P. teres (Rpt), with capital R indicating dominant resistance and lower case r indicating recessive resistance (Franckowiak and Platz, 2013), and did not allow for the designation of dominant susceptibility. Therefore, both designations have been used for the Rpt5/Spt1 locus since the locus may provide dominant resistance in the case of CI5791 or dominant susceptibility in Rika/Kombar depending on the allele-effector combination (Abu Qamar et al., 2008;Manninen et al., 2006;Richards et al., 2016;Shjerve et al., 2014). Using immortal critical recombinants (ICRs) of the previously defined Rpt5/Spt1 locus, the region was fine mapped to a c. 0.24 cM interval, equating to a c. 466 kb genomic region of Brachypodium with 62 annotated genes (Richards et al., 2016). The genomic region of Brachypodium translated to c. 9.5 Mb of physical sequence in the barley genome containing 39 high-confidence genes, leading to the hypothesis that the locus contained a single, two tightly linked or an 'island' of susceptibility genes that interact with effectors to varying specificities depending on the model (Richards et al., 2016). A total of six of the genes within the Rpt5/Spt1 locus coded for immunity receptor-like proteins that could be responsible for a susceptible response (Richards et al., 2016). However, allelic analysis of candidate genes suggested that a divergent allelic series of a single gene rather than the two tightly linked genes or a 'susceptibility island' could be responsible for different effector specificities for the complex interactions with P. teres (Richards et al., 2016). In addition, the putative Rika × Kombar recombinants that were either resistant or susceptible to both 15A and 6A were later found to be suspect; therefore, this result no longer ruled out the hypothesis of one allelic gene at the Rpt5/Spt1 locus as concluded by Richards et al. (2016).
A separate approach lending support that Rpt5/Spt1 encoded a disease resistance gene that was targeted by a proteinaceous necrotrophic effector was shown by Liu et al. (2015) where the protein designated PttNE1 interacted with a gene at the same 6H (Rpt5/Spt1) locus and designated SPN1. In this study, the PttNE1-SPN1 interaction accounted for 31% of the disease variation . The isolates 15A, 0-1, LDN07 Pt-5, ND89-19 and NB022 all exhibited a compatible reaction mapping to the SPN1 locus . In addition, the major QTL responsible for the isolate JPT9901 did not co-localize with SPN1 but distally at a distinct locus, suggesting additional genes are harboured on chromosome 6H . This single Rpt5/Spt1 locus exhibits interactions that are reminiscent of the gene-for-gene model (Flor, 1971(Flor, , 1955 and inverse gene-for-gene model (Friesen et al., 2007) in parallel. Therefore, the interaction between barley and P. teres can follow typical biotrophic models such as the gene-for-gene model (Flor, 1971(Flor, , 1955 resulting in effector-triggered immunity, as well as an inverse gene-for-gene model (Friesen et al., 2007) where necrotrophic effectors lead to necrotrophic effector-triggered susceptibility Richards et al., 2016).
A total of seven conventional genome-wide association studies (GWAS) have been performed to investigate P.  Table S1). Richards et al. (2017) revealed that the majority of the markers significantly associated with NFNB resistance localized to the centromeric region of chromosome 6H, further indicating the importance of this region. Vatter et al. (2017) performed a nested association mapping (NAM) study of barley resistance to P. teres f. teres. Vatter et al. (2017) used reaction type (Tekauz, 1985) and average ordinate to measure the infection of P. teres f. teres on the Halle Exotic Barley 25 (HEB-25) population (Maurer et al., 2015), identifying 24 QTLs (Fig. 2). In both reaction type and average ordinate phenotyping, a QTL on chromosome 2H was responsible for the most phenotypic variation at 9.23% and 14.88%, respectively (Vatter et al., 2017), but this was not the same as the Rpt3 locus. The Rpt5/Spt1 locus was identified in the NAM approach but only contributed 0.08% and 0.39% of the phenotypic variation for reaction type and average ordinate phenotyping, respectively (Vatter et al., 2017). The low level of phenotypic variation explained by the Rpt5/Spt1 locus could be because few of the founding barley lines used in the HEB-25 population contained alleles that provided resistance to P. teres f. teres.
The most novel approach to mapping resistance QTLs in barley to P. teres f. teres used exome QTL-seq (Hisano et al., 2017). For exome QTL-seq, 10 highly resistant and 10 susceptible lines were pooled separately prior to exome capture library construction and sequencing (Hisano et al., 2017). Plotting the single nucleotide polymorphisms (SNPs) for QTL mapping identified two QTLs on chromosome 3H and one on chromosome 6H from barley line H602 (Hisano et al., 2017). Hisano et al. (2017) found that there were twice as many SNPs on chromosomes 3H and 6H within the resistant bulk sample compared to the number within the susceptible bulk sample, suggesting that these loci were undergoing diversifying selection resulting in resistant lines, potentially to avoid recognition of P. teres effectors that result in programmed cell death. The results of Hisano et al. (2017) are similar to those of Gupta et al. (2010), Koladia et al. (2017a) and Liu et al. (2015). The sheer number of reports of QTLs located at the centromere of chromosome 6H, now referred to as Rpt5/Spt1, warrants validation and characterization. Research to determine the alleles and validation of Rpt5/Spt1 in an array of barley lines is currently underway (Brueggeman et al., unpublished).

Genetics of barley reactions to P. teres f. maculata
Originally, the genetics conferring resistance to P. teres f. maculata contained a total of three major designated loci and therefore had been considered less complex in comparison to the P. teres f. teres-barley interaction. The continuous distribution of P. teres f. maculata infection responses indicate that quantitative loci in barley play a part in P. teres f. maculata resistance (Burlakoti et al., 2017) and occasionally previously identified resistance loci to P. teres f. teres have been found effective against P. teres f. maculata, although it is unconfirmed if the same genes are conferring resistance/susceptibility. All three of the major barley loci conferring resistance to P. teres f. maculata confer seedling resistance (Liu et al., 2011) and are designated Rpt4, Rpt6 and Rpt8, respectively (Table 1). The Rpt4 locus Fig. 2 Consensus map of loci that are associated with net form net blotch (NFNB, red), spot form net blotch (SFNB, green) or both (blue) resistance/susceptibility across the barley genome identified using association mapping. Barley chromosome numbers are indicated at the top of each chromosome in sequential order. Designated loci are indicated by sequential numbering whereas previously undesignated loci are marked by an underscore (see also Table S3). located on chromosome 7H was found in Galleon, CI9214, Keel, Tilga, Chebec (Williams et al., 2003(Williams et al., , 1999, PI67381, PI84314 (Tamang et al., 2019) and possibly TR251 (Grewal et al., 2008). The Rpt6 locus was found on the short arm of chromosome 5H in CI9819 (Burlakoti et al., 2017;Manninen et al., 2006) and Rpt8 was found on chromosome 4H in Q21861 (Franckowiak and Platz, 2013;Friesen et al., 2006).
Breeding lines in Australia that used Rpt4 resistance were found to have lower levels of resistance to P. teres f. maculata than expected, indicating that additional genes were required for adult plant resistance (APR) (Williams et al., 2003). Although Rpt4 was found to contribute to APR in a Galleon × Huruna Nijo population, Rpt4 had a much smaller effect in the adult stage when compared to Rpt4 seedling resistance (Williams et al., 2003). At least three APR QTLs were mapped, one QTL on each of chromosomes 4H and 5H, and one to three QTLs on chromosome 7H (Williams et al., 2003) (Table S2). The barley lines Galleon, VB9104, CI9214, Kell and Chebec were found to have an APR locus on chromosome 7H, whereas Galleon and VB9104 also contained the APR loci on chromosomes 4H and 5H (Williams et al., 2003). The one to three APR loci on chromosome 7H mapped distal to Rpt4 (Williams et al., 2003), but based on the interval mapping approach used, all three loci are closely linked or overlap so it is difficult to conclude whether there are one, two or three APR loci in addition to Rpt4.
The previously identified loci included QRpts4, QRpt6 corresponding to Rpt5/Spt1 (Grewal et al., 2008), Rpt4, Rpt6, Rpt7 (Williams et al., 2003) and QTLs on 4H (Friesen et al., 2006) corresponding to Rpt8 . The Rpt7 locus reported by Tamang et al. (2015) on 7H was postulated to be the undesignated APR locus discovered by Williams et al. (2003), and therefore should remain an undesignated locus as the Rpt7 locus was previously designated on chromosome 4H (Table 1). In addition, the QRpt7 locus was previously reported by Grewal et al. (2008) and was established to be Rpt4 on chromosome 7H. QRpts4 (Grewal et al., 2008) and APR on chromosome 4H (Williams et al., 2003) may be the same locus as Rpt7 on chromosome 4H. The APR QTL located on chromosome 5H (Williams et al., 2003) is not Rpt6 and remains undesignated.
Four highly significant QTLs (QRptm7-4, QRptm7-6, QRptm7-7 and QRptm7-8) map to a 36 cM region that approximately encompasses the Rpt4 locus on chromosome 7H , similar to results reported by Williams et al. (2003). However, it is unknown if QTLs on chromosome 7H are multiple linked QTLs, multiple alleles of a single QTL or an artefact of population structure . Analysis of linkage decay of the QTLs located on chromosome 7H suggests that the four significant QTLs are independent of each other, but further verification is required .
Burlakoti et al. (2017) studied the effect of two-and six-row barley, concluding that the proportion of six-row barley (43%) resistant to P. teres f. maculata was higher than that of the two-row barley (13%) tested. This difference is in contrast to Wang et al. (2015), who had comparable numbers of resistant lines (40-65%) in Australian two-and six-row barley. The most significant QTL identified was novel, mapped to chromosome 2H and accounted for 4.3-6.6% of the phenotypic variation (Burlakoti et al., 2017). Daba et al. (2019) identified a single novel QTL on chromosome 6H, but this is attributed to most lines being susceptible to P. teres f. maculata. The association mapping studies of Tamang et al. (2015), Wang et al. (2015) and Burlakoti et al. (2017) indicated that the genetic interaction of P. teres f. maculata and barley was more complex than originally hypothesized. In addition, novel QTLs and therefore sources of resistance can be identified with the power of association mapping to evaluate more accessions in parallel.
Recently, studies to identify new sources of resistance to P. teres f. maculata have been undertaken by Çelik Oğuz et al. (2017) and Gyawali et al. (2018). Çelik Oğuz et al. (2017) found two six-row landraces in Turkey that were resistant to all (three P. teres f. teres and three P. teres f. maculata) isolates tested that could be used in breeding programmes. Gyawali et al. (2018) screened 340 diverse ICARDA barley lines using one P. teres f. maculata isolate to identify 12 lines that were highly resistant and stable across environments to be used for resistance breeding. However, without mapping and subsequent allelic analysis it is unknown what resistance loci or alleles are present.

Effectors of P. teres
Effectors are molecules that interact with the host to manipulate the host response, and encompass secondary metabolites, peptides, small secreted proteins, enzymes and more recently small RNAs (Toruño et al., 2016). Effectors can be identified in multiple ways, including forward and reverse genetic approaches, including traditional genetic mapping, genomic mining and proteomics/metabolomic approaches. A genetic mapping approach to effector discovery is labour-intensive and includes population development, genotyping and phenotyping but can generate highly useful candidate gene lists to functionally validate effectors that segregate within a given population. The use of genetic mapping is less effective in finding highly conserved effectors that do not segregate between parents of a population or those that become fixed within a natural population. Genomic mining or proteomic/metabolomic approaches can be used to identify these highly conserved effectors but result in large candidate lists of genes that need to be functionally validated.
For the successful colonization by necrotrophic fungi such as P. teres f. teres and P. teres f. maculata, necrosis and chlorosis of plant tissue is associated with a compatible reaction (reviewed in Liu et al., 2011). Chlorosis surrounding the necrotic lesion is often associated with diffusible phytotoxic secondary metabolites (Keon and Hargreaves, 1983;Sarpeleh et al., 2007). Four toxins, N-(2-amino-2-carboxyethyl)-aspartic acid, anhydroaspergillomarasmine A, aspergillomarasmine A and aspergillomarasmine B, were isolated and their pathways subsequently identified (Bach et al., 1979;Friis et al., 1991;Sarpeleh et al., 2009;Smedegård-Petersen, 1977) and used to screen for resistance to P. teres (Sharma, 1984;Weiergang et al., 2002). All four toxins identified were of the marasmine class that chelate iron ions and have been shown to be lightand temperature-dependent, indicating a potential role targeting organelles such as the chloroplast (Sarpeleh et al., 2009). Secondary metabolites do not appear to be isolate-specific and little host specificity exists apart from lines being either sensitive or insensitive (Sarpeleh et al., 2009(Sarpeleh et al., , 2007. Multiple studies have shown that different virulence profiles exist within P. teres populations around the globe (Akhavan et al., 2016;Arabi et al., 2003;Cromey and Parkes, 2003;Fowler et al., 2017;Gupta and Loughman, 2001;Harrabi and Kamel, 1990;Jalli and Robinson, 2000;Jonsson et al., 1997;Khan and Boyd, 1969b;Liu et al., 2012;McLean et al., 2010b;Robinson and Jalli, 1996;Sato and Takeda, 1997;Steffenson and Webster, 1992;Tekauz, 1990;Wu et al., 2003). Virulence variation among isolates on different barley lines appears to originate from their segregating effector repertoire, many of which are now believed to be proteinaceous in nature Richards et al., 2016). Virulent P. teres f. teres and P. teres f. maculata isolates have increased production of proteases, possibly for faster establishment of the pathogen and increased sugar acquisition compared to less virulent isolates (Dikilitas et al., 2018). Two papers from the same group proposed that variations in virulence could be explained by fungal growth, ability to deliver effectors and differentially expressing effectors (Ismail et al., 2014a(Ismail et al., , 2014b. Ismail et al. (2014a) used six P. teres f. teres isolates and found that the four more virulent isolates had greater conidial germination and appressoria formation compared to the less virulent isolates yet extracted proteinaceous effectors from all isolates were able to induce necrosis. Three proteinaceous effectors were postulated in the virulent P. teres f. teres isolate 32/98 as PttXyn11A, PttCHFP1 and PttSP1 that showed homology to proteins involved in plant disease interactions (Ismail et al., 2014b). Using a combination of one-and two-dimensional electrophoresis, Ismail and Able (2016) identified 63 proteins that were produced by 28 virulent isolates. Ismail and Able (2017) subsequently identified a transition between colonization and necrotrophy at approximately 48 h based on the in planta gene expression analysis of 222 proteins. As no functional validation was reported on these candidate genes/proteins, it is not yet known whether any of these are biologically relevant to the P. teres f. teres-barley interaction.
A proteinaceous necrotrophic effector, PttNE1, was isolated from intercellular wash fluid by Liu et al. (2015) using the susceptible barley line Hector inoculated with P. teres f. teres isolate 0-1. PttNE1 exhibited direct association with disease and interacted with SPN1, and therefore Liu et al. (2015) concluded that host sensitivity resulted from necrotrophic effector-triggered susceptibility, but the pathogen gene PttNE1 was not mapped in P. teres f. teres. The first mapped effector implicated in the P. teres f. teresbarley system, designated AvrHar, was identified by Weiland et al. (1999) in the P. teres f. teres 15A × 0-1 population. The 15A allele of AvrHar was proposed to contribute low virulence on the barley line Harbin (Weiland et al., 1999) (Table 2). In an expanded population of 15A × 0-1, the 15A allele of the AvrHar locus also conferred avirulence on Tifang and Canadian Lake Shore (Lai et al., 2007). In addition, two QTLs, designated AvrPra1 and AvrPra2, from 0-1 were mapped and found to be functionally redundant to confer avirulence on Prato (Lai et al., 2007) (Table 2). The AvrHar and AvrPra2 loci mapped to the same locus on linkage group 7 (chromosome 5) (Lai et al., 2007;Wyatt et al., 2018), but whether the locus contains two tightly linked genes or alleles of the same gene is yet to be confirmed (Fig. 3). The AvrPra1 locus was mapped to linkage group 1 (Lai et al., 2007), currently designated chromosome 9 based on the 0-1 assembly (Fig. 3, Wyatt et al., 2018). A separate avirulence gene, designated Avr Heartland , was identified in a 67-progeny mapping population with the Canadian avirulent isolate WRS 1906 and the virulent isolate WRS 1607 (Beattie et al., 2007) (Table 2). Afanasenko et al. (2007) proposed the presence of a suppressor of avirulence genes where virulence arose in 30 out of 84 progeny from an avirulent by virulent cross on the barley line Diamond. Based on an avirulent: virulent segregation ratio of 9:7, Afanasenko et al. (2007) suggested the presence of two avirulence genes and one suppressor gene. Based on phenotypic ratios alone, Afanasenko et al. (2007) proposed that up to four avirulence genes and up to two suppressor genes in various combinations were present in four P. teres f. teres crosses, but the corresponding loci were not mapped. Shjerve et al. (2014) was the first study to begin using a restrictionsite associated-digest-genotyping-by-sequencing (RAD-GBS) approach detailed by Leboldus et al. (2015)   isolate is known to be avirulent on, additional lines may exist, a allele 1, b allele 2. § QTL effects containing ranges mean that the experiments were performed for multiple locations or multiple isolates and the effects for individual treatments fall into this range. ‖ Inferred through current genome assembly and chromosome numbering .-No information is available for this entry.
rapid construction of genetic maps by subsequent studies including Koladia et al. (2017b) and Carlsen et al. (2017) in P. teres f. teres and P. teres f. maculata, respectively. The RAD-GBS approach facilitated the mapping of four QTL regions that have now been assigned chromosomal designations based on the recently updated 0-1 reference genome assembly  including VK1 and VK2 (Virulence on Kombar 1 and 2) and VR1 and VR2 (Virulence on Rika 1 and 2). VK1 localized to chromosome 3, VK2 and VR1 to separate loci on chromosome 2 and VR2 on chromosome 10 (Fig. 3). Isolate 15A harboured the virulence alleles at VK1 and VK2 with each conferring virulence on Kombar but lacking virulence on Rika (Shjerve et al., 2014) (Table 2). The isolate 6A harboured virulence alleles at VR1 and VR2 and the reciprocal occurred, where VR1 and VR2 confer virulence on Rika but lacked virulence on Kombar (Shjerve et al., 2014) (Table 2). In addition, Shjerve et al. (2014) further characterized the progeny isolates of the 15A × 6A population to discover that VK1, VK2 and VR2 all targeted the same Spt1 host region on chromosome 6H. Isolates containing VK1 or VK2 alone were each virulent on Kombar containing Spt1.k, whereas the isolate that contained VR2 alone was avirulent (Shjerve et al., 2014). Conversely the isolate that contained VR2 alone was virulent on Rika that contained Spt1.r but isolates that contained VK1 and VK2 alone were avirulent (Shjerve et al., 2014). Koladia et al. (2017b) identified nine unique QTLs in the P. teres f. teres cross between the Danish isolate BB25 and the North Dakotan isolate FGOH04Ptt-21, of which only one of the virulence alleles was conferred by BB25 and eight were conferred by FGOH04Ptt-21 (Table 2). A major QTL (PttTif1) with the virulent allele from FGOH04Ptt-21 found on linkage group 10.1 (chromosome 1) contributed as much as 74% of the phenotypic variation on three lines (Manchurian, Tifang and CI4822) (Koladia et al., 2017b) (Fig. 3). The second major locus (PttBee1) conferring virulence from FGOH04Ptt-21 was found on linkage group 1.1 (chromosome 1), contributing 56% of the phenotypic variation on the barley line Beecher (Koladia et al., 2017b) (Fig. 3). The BB25 isolate was only virulent on barley line Pinnacle with the BB25 virulence allele (PttPin1) located on linkage group 5.1 (chromosome 3), which explained 49% of the phenotypic variation (Koladia et al., 2017b) (Fig. 3). The virulence alleles of the remaining six loci all contributed less than 18% of the phenotypic variation and were distributed across multiple chromosomes (Koladia et al., 2017b). Currently, effector candidate genes underlying these major QTLs are being evaluated. Carlsen et al. (2017) is currently the only study to map P. teres f. maculata virulence. The FGOB10Ptm-1 × SG1 population was inoculated on commonly used SFNB differential barley lines Skiff, TR326, 81-82/033 and PI392501 with a total of six loci identified across five of the 12 linkage groups (Carlsen et al., 2017) (Fig. 3 and Table 2). FGOB10Ptm-1 and SG1 contributed the virulence allele at five and one QTL, respectively (Carlsen et al., 2017). The QTL with the virulence allele contributed by SG1 (vQTL4, PtmWel1) on linkage group 1.1 (chromosome 2) accounted for 30-37% of the phenotypic variation on barley lines 81-82/033 and PI392501, whereas the largest effect QTL with the virulence allele contributed by FGOB10Ptm-1 (vQTL5, PtmWel2) on linkage group 5.1 (chromosome 3) accounted for 26-34% of the phenotypic variation on barley lines 81-82/033, TR326 and PI392501 (Carlsen et al., 2017) (Fig. 3 and Table 2). We have assigned tentative designations for this review until a robust naming nomenclature is established.

Advancement of genomic resources in barley and P. teres
The first assembled genome of barley was published by IBGSC (2012) using the American six-row malting cultivar Morex. Again, Morex was used for a high-quality reference assembly using chromosome conformation capture (Beier et al., 2017;Mascher et al., 2017), with subsequent analysis in the repetitive elements . Additional lines that have been sequenced include the American two-row spring feed barley line Bowman, the German tworow spring malting barley line Barke, the Scottish two-row winter malting barley line Igri (IBGSC, 2012), the Japanese two-row spring malting barley line Haruna Nijo (IBGSC, 2012;Sato et al., 2016), and the Tibetan hulless cultivars Lasa Goumang (Zeng et al., 2015) and Zangqing320 (Dai et al., 2018;Nyima et al., 2018), but no published studies have compared these released barley genomes (Monat et al., 2018). Multiple studies have revealed genomic signatures of domestication using exome sequencing of 267 barley accessions (Russell et al., 2016), targeted resequencing of 433 barley accessions (Pankin et al., 2018) and whole-genome sequencing of 172 barley accessions . In the future we will be able to generate pan-genomic resources that consolidate resistance loci to P. teres from different studies into their respective allelic series. Currently, large-scale barley genome sequencing for pan-genomic resources is hampered by the large amount of repetitive content in the barley genome (Monat et al., 2018). In this review, we consolidated mapped loci that have not currently been proven to be separate loci.
A wealth of P. teres genomic resources has been generated since the first genetic marker studies evaluating virulence/avirulence. The first P. teres genome was of the Canadian P. teres f. teres isolate 0-1 using Illumina paired-end reads (Ellwood et al., 2010). The size of 0-1 was initially estimated at 41.95 Mb with an average sequence coverage of 20×, confirming the presence of at least nine chromosomes based on pulsed-field gel electrophoresis and germ tube burst visualization (Ellwood et al., 2010). Wyatt et al. (2018) subsequently confirmed 12 chromosomes with a reference quality assembly of the same 0-1 isolate using PacBio long-read single-molecule real-time (SMRT) sequencing. The 0-1 assembly produced a reported genome size of 46.5 Mb at an average coverage of 200× . Current genome assemblies include an additional eight P. teres f. teres isolates (W1-1, NB29, NB85, NB73, 15A, 6A, FGOH04Ptt-21 and BB25) and two P. teres f. maculata isolates (FGOB10Ptm-1 and SG1) (Syme et al., 2018;Wyatt et al., 2019). Genome sizes currently are in the range of 39.27-41.28 Mb for P. teres f. maculata and 46.31-51.76 Mb in P. teres f. teres, concluding that P. teres f. maculata has a smaller genome due mostly to repetitive elements (Syme et al., 2018).
Due to the diversity of the natural P. teres f. teres population, evident from the mapped effectors mentioned earlier, a pangenomic sequencing approach of four P. teres f. teres isolates 15A, 6A, FGOH04Ptt-21 and BB25 was used to evaluate the genomic architecture and gene content in comparison to 0-1 (Wyatt et al., 2019). Each P. teres f. teres isolate was predicted to contain approximately 200 effectors using EffectorP (Wyatt et al., 2019), but EffectorP prediction is known to underestimate effectors due to stringent criteria. The majority of published virulence QTLs have been identified in subtelomeric regions where chromosome rearrangements and repetitive elements are found at higher frequency, allowing rapid development of polymorphisms (Wyatt et al., 2019). Wyatt et al. (2019) showed that 14 of 15 currently published P. teres f. teres QTLs span accessory genomic compartments with ten of 14 accessory genomic compartment QTLs specifically localizing to subtelomeric regions. This observation highlights the importance of the subtelomeric regions as drivers of virulence for P. teres f. teres. Additionally, a whole chromosome fusion was found between chromosomes 1 and 2 of isolate BB25, which is predicted to be a recent event and not inherited through ancestry and lost in other lineages (Wyatt et al., 2019). Chromosome fusions have been recently identified in the closely related species P. tritici-repentis in reference isolate M4 and may be a common feature in the Pyrenophora genus (Moolhuijzen et al., 2018). Liu et al. (2011) stated that the following characteristics facilitate the use of P. teres as a candidate model organism: the worldwide economic importance of the disease, the ability to produce biparental mapping populations and clone genes from the pathogen, the amenability of the pathogen to transformation for gene validation, relatively abundant genomic resources in both the host and pathogen, and the large numbers of host mapping populations to decipher the interactions and diversity of both the host and pathogen. These points were in their infancy and have been built on since Liu et al. (2011). As noted, P. teres has become an increasing problem worldwide, leading to the expansion of pathogen and host mapping studies that have identified loci important for pathogen virulence and host resistance/susceptibility using biparental and association mapping studies. In summary, over 340 and 140 QTLs have been published in relation to barley responses to P. teres f. teres and P. teres f. maculata isolates, respectively, of which eight loci are sequentially designated from Rpt1 to Rpt8 on chromosomes 3H, 1H, 2H, 7H, 6H, 5H, 4H and 4H, respectively (Table 1). Despite fine-mapping of loci such as Rpt5/Spt1 (Richards et al., 2016), delimiting loci within the barley pan-genome remains difficult due to the multitude of studies that report large overlapping genomic regions (Figs 1 and 2). Additionally, a large proportion of identified QTLs are collapsed into 39 consensus loci, resulting in a total of 47 loci (Figs 1 and 2). In contrast, a total of 22-23 loci have been published in P. teres that follow a tentative nomenclature of (a)virulence on a specific barley line (Table 2). An exponential growth of genomic resources for P. teres has occurred since the first genome assembly reported by Ellwood et al. (2010), with reference quality genome assemblies and analysis (Syme et al., 2018;Wyatt et al., 2019). Adoption of the P. teres-barley pathosystem as an agriculturally significant model has allowed rapid advancement in the abundance of genomic sequence, biparental and natural populations, and marker trait associations/QTLs to validate and characterize genes in both host and pathogen. Despite some overlap in functional resistance/susceptibility loci to P. teres f. teres and P. teres f. maculata in barley, each form of P. teres has a novel repertoire of effectors at their disposal. Therefore, breeding for resistance in barley to P. teres f. teres and P. teres f. maculata should proceed separately based on the most effective loci to the respective local P. teres populations.

Data availability statement
Data sharing is not applicable to this article as no new data were created or analysed in this study.

SUPPORTING INFORMATION
Additional supporting information may be found in the online version of this article at the publisher's web site: Table S1 Studies mapping barley resistance/susceptibility genes to Pyrenophora teres f. teres. Different populations are indicated by alternating grey scale with the parent contributing the resistance allele in bold. Plant stage indicates whether the resistance is effective at the seedling or adult stage. If a single genotype isolate is used, this is indicated by the name and country of origin, whereas natural infection is indicated by specific location. Barley chromosome, phenotypic variation and designation of each locus are displayed, if available, from the corresponding reference. The inferred locus designation is reported based on markers obtained from the relevant publication and collapsed into loci using a maximum distance of 10 Mb of the Morex reference genome  using BarleyMap (Cantalapiedra et al., 2015) or T3/Barley (Fig. 1). genotype isolate is used this is indicated by the name and country of origin, whereas natural infection is indicated by specific location. Barley chromosome, phenotypic variation and designation of each locus are displayed if available from the corresponding reference. The inferred locus designation is reported based on markers obtained from the relevant publication and collapsed into loci using a maximum distance of 10 Mb of the Morex reference genome  using BarleyMap (Cantalapiedra et al., 2015) or T3/Barley (Fig. 1). Table S3 Genomic positions of markers reported in biparental or association mapping studies that could be anchored to the Morex barley reference genome , including the locus designation (from publication), the corresponding marker, the assigned chromosome and chromosome position, and the reference.