Genetics of resistance to Zymoseptoria tritici and applications to wheat breeding

1Resistance to S. tritici blotch (STB; caused by Zymoseptoria tritici, formerly Mycosphaerella graminicola) became a significant target trait in wheat breeding much more recently than diseases such as the rusts and powdery mildew. The potential threat posed by STB was drawn to international attention by a very damaging epidemic in North Africa in 1968–1969, which followed the introduction of semi-dwarf wheat cultivars and increased use of artificial fertiliser. Subsequently, STB increased in importance, particularly in semi-dwarf cultivars given high rates of nitrogen fertiliser and is now a potentially damaging disease throughout the temperate regions. Early work on genetics (reviewed by ) focussed on the discovery of sources of resistance for breeding and on cultivar-by-isolate interaction but substantial progress has been made in the last 20 years in the genetics of resistance. This is giving breeders a deeper understanding of effective approaches to improving resistance and will allow resistance genes to be isolated and their functions revealed.

2As in many other plant diseases, wheat has essentially two types of resistance to STB, as shown in a large study of 236 wheat cultivars grown in the UK in the 1990s and their progenitors. Qualitative resistance is strong and is usually controlled by major genes with a large effect. These genes are generally effective against avirulent pathogens but not against other, virulent isolates. Their pattern of interaction with Z. tritici accords with the gene-for-gene relationship, as has been demonstrated for Stb6. Quantitative resistance, by contrast, has a partial phenotype and is controlled by several-to-many genes with moderate-to-small effects. In many instances but not always, it is effective against all Z. tritici genotypes. Even though STB is almost always scored as a quantitative trait or at least on an ordinal scale, segregation of a qualitative gene can give rise to a large difference between two groups of progeny, resistant and susceptible (e.g. Stb6; ). This can be obscured, however, by segregation of minor genes which alter the level of STB symptoms and lead to intermediate phenotypes.

3In certain cultivar-by-isolate interactions, adult-plant responses to Z. tritici do not necessarily reflect responses of seedlings to the pathogen. Many genotype-specific, qualitative resistances are independent of growth stage whereas the expression of partial resistance may depend on the plant’s growth stage. Stb17 is an example of a gene with a quantitative effect on disease which is expressed in adult plants but not seedlings, while genes on chromosome arm 5BS of Hobbit sib increased susceptibility only in adult plants.

4In this review, we survey the genes for STB-resistance reported to date (Fig. 1), including both qualitative (Table 1) and quantitative resistance (Table S1), beginning with the first genes to be named, Stb1, Stb2 and Stb3. Earlier reports of sources and genetics of resistance were summarised by. We also describe how knowledge of the genetics of resistance can be applied to wheat breeding.

5As with genetic analysis of any trait, study of STB-resistance requires a method of scoring the phenotype which can be applied to large populations. Tests can be conducted at the seeding stage, with inoculation typically when seedlings are two weeks old. Both whole-seedling and detached leaf assays are used, and require conditions with high relative humidity. They generally take around four weeks following inoculation. The advantage of using whole seedlings is that many plants can be tested but a disadvantage, especially in greenhouse trials, is that environmental conditions may not be strictly controlled, which can affect the development of disease. Detached-leaf tests are particularly suitable when many isolates are to be tested, although this can also be done in a greenhouse. Alternatively, plants can be tested at the adult stage in the glasshouse or field. Given the genotype-specificity of most qualitative resistances (see below), identification of single genes in field conditions requires plants to be inoculated with a Z. tritici isolate at a dose sufficient to make contamination by natural inoculum comparatively negligible. The expression of symptoms following inoculation in adult plants proceeds at a broadly similar rate as in seedlings. When genetic analysis is conducted on naturally-infected trials, the genes identified are effective against the current local Z. tritici population. This is generally relevant to the practice of plant breeding but may have implications for repeatability if there is polymorphism at avirulence loci corresponding to segregating resistance genes.

6In most genetic analyses of STB-resistance, the phenotype studied is formation of pycnidia, the asexual fruiting bodies of Z. tritici which form within necrotic tissue on the leaf. In this case, the data are fractions of leaf area covered by necrotic lesions bearing pycnidia. Some studies have reported other phenotypes in addition to pycnidium formation, including total necrotic leaf area with or without pycnidia, latent period and disease progress. These traits are scored visually, either by eye or, more recently, by computer-aided image analysis.

7Please refer to Table 1 for details of genes for STB-resistance in bread wheat (Triticum aestivum) and to Fig. 1 for their locations.

8Stb1, Stb2 and Stb3 were the first genes for STB-resistance to be named. Before then, it was generally thought that resistance to Z. tritici was a quantitative, polygenic trait. Although quantitative resistance is indeed considerably more important than qualitative resistance in wheat breeding, the discovery that significant amounts of resistance can be controlled by major genes opened the way to genetic analysis of STB and may offer an opportunity to improve resistance by ‘stacking’ or ‘pyramiding’ several Stb genes (; see Section 6.1).

9Stb1, Stb2 and Stb3 have been mapped to chromosome arms 5BL (i.e. the long arm of chromosome 5B; ), 1BS (the short arm of 1B; ) and 7AS respectively. Stb2 and Stb3 were originally mapped to chromosomes 3BS and 6DS respectively but those locations were corrected for the reasons given in the subsequent papers. Stb2 was found to map to the same region of 1BS as Stb11 but no test of allelism of the two genes has yet been done.

10The sources of Stb1, Stb2 and Stb3 – cvv. Bulgaria 88, Veranopolis from Brazil, and Israel 493 respectively – all have Stb6 in addition. The resistance of Bulgaria 88 was described as durable by but it is not known if this refers to Stb1, Stb6 or both.

11Stb4 in cv. Tadinia was the first gene to be identified by controlled inoculation with a single isolate of Z. tritici, CA30 from California. It was subsequently mapped to chromosome arm 7DS (7DS; ), close to the locus of Stb5. Again, the allelism of Stb4 and Stb5 has not been tested.

12The first gene for STB-resistance to be mapped was Stb5, which originated from a highly resistant synthetic hexaploid line, Synthetic 6x, derived from Triticum dicoccoides (AABB genomes) and Triticum tauschii (also known as Aegilops squarrosa; DD). Stb5 mapped to the pericentromeric region of chromosome arm 7DS, close to where Stb4 was mapped subsequently. The mapping work used Z. tritici isolate IPO94269 from The Netherlands but Stb5 conferred resistance to all but one of the isolates tested. The location of Stb5 was greatly facilitated by the use of precise cytogenetic stocks, which form a unique resource for wheat genetics to identify chromosomes carrying genes for traits of interest.

13Synthetic hexaploids are a rich source of qualitative genes for resistance to STB and other diseases. Stb8 was identified in another synthetic line, W7984, bred by CIMMYT (the International Maize and Wheat Improvement Centre). It conferred resistance to an isolate from the USA and mapped to the long arm of chromosome 7B.

14Two further genes, Stb16q and Stb17, were discovered in the synthetic hexaploid line M3. Stb16q on chromosome 3DL was designated as a quantitative (q) locus because it was not possible to determine if there was indeed a single gene at the locus. However, it controlled a high proportion of variation in necrotic leaf area, leaf area bearing pycnidia and latent period, and, alone among major Stb genes reported so far, conferred resistance at the seedling stage to all Z. tritici isolates tested, of which there were 20. It may be better regarded as a type of qualitative resistance. Stb17 on 5AL was detected only at the adult-plant stage and was less potent than Stb16q. The phenotype of Stb17 may fall between the qualitative and quantitative classes.

15The only qualitative gene for STB-resistance which has been shown to control a gene-for-gene relationship is Stb6, at the distal end of the short arm of chromosome 3A. This gene, which confers resistance to a Dutch Z. tritici isolate, IPO323, was first identified in the UK cvv. Flame and Hereward.

16Stb6 is an especially notable gene as it was found to be present in most of the well-known sources of STB-resistance studied previously. Analysis of alleles of a simple-sequence repeat (SSR or microsatellite), Xgwm369, closely linked to Stb6, allied to analysis of wheat breeding pedigrees indicated that Stb6 had been introduced on at least six separate occasions into modern European germplasm, and was also present in Chinese Spring, a selection from a landrace which has been widely used in genetic studies of wheat. It is the second most frequent Stb gene in European wheat, present in about 15% of cultivars tested.

17At least five analyses of quantitative trait loci (QTL; ) as well as an association genetic analysis have mapped field resistance to STB close to the Stb6 locus. A study of a large panel of UK and continental European cultivars found Stb6 to be associated with a reduction in STB symptoms in field conditions. This is consistent either with a minor gene for partial resistance to STB being closely linked to Stb6 or with Stb6 itself having a residual effect on field resistance even though virulence to Stb6 is almost fixed in the European population of Z. tritici (J.K.M.B., unpublished data).

18Avirulence (AVR) to Stb6 was shown to be controlled by a single gene in Z. tritici IPO323 in a cross with IPO94269 in which virulence to several cultivars and breeding lines co-segregated. As resistance to IPO323 maps to the Stb6 locus in all these cultivars, it is concluded that Stb6 is present in all of them and that it corresponds to a single AVR gene in IPO323.

19Stb7 on chromosome 4AL was first identified in cultivar ST6 as conferring resistance to the Canadian isolate MG2. It was mapped close to SSR locus Xwmc313. A gene for resistance to the Uruguayan isolate IPO87019 which mapped in the same location in the Portuguese line TE9111 (later released as cv. Nabão) was thought to be Stb7, an allele of Stb7 or a closely linked gene. As the resistance gene in TE9111 was mapped by QTL analysis a precise location for the gene could not be achieved.

20The CIMMYT breeding Kavkaz-K4500 L.6.A.4 (KK) carries Stb12, which is closely linked to Stb7 on chromosome 4AL. Stb12 provides resistance to isolate Isr398 from Israel but not to IPO87019. It was mapped by QTL analysis and was closer to Xgwm219 than to Xwmc313, which are ∼3cM apart. Of the 94 single-seed descent progeny of the cross of KK with the Isr398-susceptible cv. Shafir, four were resistant to ISR398 but susceptible to IPO97019, demonstrating the existence of two genes in that region. This is an example of Stb genes being clustered, a common feature of genes involved in plant defence.

21Stb9 on chromosome 2BL was mapped in the spring wheat cvv. Courtot and Tonic. It confers resistance to the Dutch isolate IPO89011.

22Stb10 was also discovered in KK. Like Stb5, it conferred resistance to IPO94269 but it was clearly a different gene, located near the centromere of chromosome 1D.

23Stb11 on chromosome 1BS was identified and mapped in TE9111 and reported to confer resistance to isolate IPO90012 from Mexico but it may be widespread in global spring wheat breeding. When remapped, Stb2 was located close to or at the Stb11 locus. StbWW, identified in three populations in Australia, was also mapped on chromosome arm 1BS at or near Stb11. These genes may all be Stb11, which may have spread in global wheat breeding by the movement of elite breeding lines from CIMMYT.

24Stb13 on chromosome 7BL and Stb14 on 3BS were discovered in the Canadian cv. Salamouni. Both genes conferred resistance to MG2, like Stb7, while Stb13 also provided resistance to MG96-36. Salamouni also has a third gene, designated StbSm3, which maps close to the Stb6 locus on chromosome 3AS but apparently distal to it. No test of the allelism of StbSm3 and Stb6 has yet been conducted.

25Stb15 on chromosome 6AS was identified as providing resistance to the Ethiopian isolate IPO88004. It is very common in European winter wheat, present in about 60% of cultivars tested but, unlike the other widespread gene, Stb6, it is not associated with resistance in field conditions.

26Stb18 on 6DS confers genotype-specific resistance in the French winter wheat cv. Balance. It was expressed at the seedling stage but inconsistently in adult plants, being detected in one of two years of field trials of a population produced from Apache x Balance.

27Although STB is a severe disease of modern cultivars of durum wheat (Triticum durum), especially in North Africa, the genetics of STB-resistance in T. durum are poorly understood. In a search for sources of resistance in older, landrace cultivars, resistance to Z. tritici isolate Tun06 in a selection from the Agili landrace segregated as a single major gene. This gene was associated with AFLP markers but has not yet been assigned to a chromosome.

28The diploid emmer wheat, Triticum monococcum, is highly resistant to Z. tritici. All accessions tested varied from very resistant to immune both to artificial resistance as seedlings and in five years of field trials. The genetics of resistance were studied in one accession, MDR043, and the gene TmStb1 was mapped to chromosome 7AmS (Table 1; ).

29Please refer to Table S1 for details of these genes and to Fig. 1 for their locations.

30In field trials, resistance to STB generally appears as a quantitative trait, largely additive in nature with some dominance, controlled by an oligogenic or polygenic system with moderate to high heritability in both durum wheat and bread wheat (; and papers cited in Section 4 and Table S1). QTL for resistance to STB at both seedling and adult stages are distributed throughout the genome of wheat (Table S1a). To date, 167 QTL of resistance against STB have been detected in a total of nineteen bi-parental mapping populations. From seven of these populations, 27 meta-QTL, i.e. refined QTL from multiple individual QTL, have been identified, integrating 105 individual QTL (; Table S1b). Of 89 regions identified, 62 QTL and 27 meta-QTL, 27 were detected at the seedling stage, 48 at the adult stage and 14 at both stages. They included genome regions involved in the control of necrosis, pycnidium development and disease progress estimated as area under the disease progress curve (AUDPC). Two minor QTL controlling latent period have also been identified.

31All chromosomes except 5D carry at least one QTL or meta-QTL for STB-resistance. Nineteen QTL or meta-QTLs co-localised with genes involved in plant height (Rht: reduced height), heading date (Ppd: photoperiod-insensitivity) or both, among which six mapped closely to the Rht8 and Ppd-D1 (2DS), Rht-D1 (4DS), Ppd-A1 (2AS) and Rht-B1 (4BS). Three chromosome arms, 3BL, 6BS and 7DL, were especially involved in quantitative resistance to STB according to the number of QTLs identified. There are probably co-localisations with qualitative Stb genes for 22 QTL and 6 meta-QTL. QTL have frequently been mapped to the regions where Stb6 (3AS), Stb5/Stb4 (7DS) and Stb11/Stb2/StbWW (1BS) are located and less frequently to the regions of Stb1 (5BL), Stb9 (2BL), Stb7 and Stb12 (4AL), Stb13 (7BL), Stb14 (3BS) and Stb18 (6DS). Except for eight QTL identified in synthetic hexaploid wheat the chromosome substitution line ‘Chinese-Spring’ (T. aestivum subsp. spelta 7D) and a line from the USA, all QTL listed in Table S1 originated from European germplasm.

32Association mapping studies have highlighted the presence of many regions of the genome in cultivated wheat and landraces associated with resistance to STB, including both some previously associated with STB-resistance and some not. From spray-inoculated field trials conducted over two years, detected 68 SSR significantly associated with adult resistance in a panel of 372 European lines. Nine loci were significantly associated with all phenotyping parameters. Association QTL mapped to the loci of Stb1, 4, 6 and 8, implying that these genes or alleles of them may be present in European cultivars. In addition, several traits related to STB-resistance mapped at or near QTL identified previously.

33In a study of 1055 elite hybrids and their corresponding 87 parental lines trialled in two locations which either had natural infection or were inoculated by spraying with a mixture of isolates, identified eight single-nucleotide polymorphisms (SNP) associated with STB resistance. Although half the SNP were not genetically mapped, the others were located on chromosomes 1B, 2B, 5B and 6A. The 5B locus may represent Stb1 or a QTL identified in the population Arina x Forno and Steele-ND x ND 735.

34Finally, seven SNP at four loci were significantly associated with resistance in an association mapping study of a panel of 528 spring wheat landraces of worldwide origin phenotyped at the adult stage in growth chambers. These SNP mapped to chromosomes 3B, 6B and 7B and most likely relate to new resistance genes.

35In a cytogenetic analysis, the 5BS arm of Hobbit sib (Dwarf A) was found to promote susceptibility to STB in adult plants but not in seedlings. Formally, the data were also consistent with this chromosome arm carrying genes which suppress resistance but Hobbit sib has no known STB-resistance genes. The same chromosome arm has genes which increase resistance to yellow (stripe) rust and powdery mildew, which implies that there may be a trade-off between breeding for resistance to STB and to biotrophic fungi.

36Strong specificity in the interaction between cultivars of bread wheat (T. aestivum) and Z. tritici isolates was reported by and confirmed in subsequent studies on seedlings and adult plants. These interactions are akin to gene-for-gene relationships but this has only been demonstrated in Stb6 resistance to IPO323.

37Specialisation of Z. tritici isolates to either T. aestivum or T. durum has been reported by some workers ( and papers cited therein; ) but not others. Cultivar-by-isolate specificity within T. durum has also been reported. In a cross of aestivum-adapted and durum-adapted isolates of Z. tritici, the AvrStb6 locus for avirulence to Stb6 was associated with part of the variation in ability of progeny isolates to infect T. durum. This suggests that resistance of these two wheat species to inappropriate specialised forms of Z. tritici may be controlled in part by qualitative resistance genes.

38Most Z. tritici isolates used in research on STB are virulent to almost all Stb genes although some isolates with more than one functional avirulence phenotype are known. IN95-Lafayette-1196-WW-1-4 was avirulent to both Stb1 and Stb4 while the Paskeville isolate was avirulent to Stb2 and Stb3. Among the isolates avirulent to Stb5 in Synthetic 6x were IPO323, which is also avirulent to Stb6, IPO89011 (avirulent to Stb9: ), IPO94269 (avirulent to Stb10: ) and IPO001, which is avirulent to some UK cultivars. MG2 was avirulent to Stb7, Stb13 and Stb14. IPO323 may detect a second resistance gene in KK in addition to Stb6. Otherwise, the high frequency of virulence implies that there is little obstacle to Z. tritici mutating to virulence on Stb genes which control gene-for-gene interactions.

39As in other diseases with a gene-for-gene system, the specificity of qualitative resistance to avirulent Z. tritici genotypes leads to selection for virulence (loss of avirulence). In Oregon, USA, such a ‘breakdown’ of resistance due to pathogen adaptation happened rapidly, with a catastrophic effect on disease control, in cv. Gene in the 1990s and more gradually in cv. Foote in the 2000s. Cv. Gene is resistant to IPO323 and IPO94269, which are avirulent to Stb6 and Stb10 respectively but it is not known if either of these genes was the one which was overcome by the fungus. reported that the resistance of cv. Obelisk became less effective in The Netherlands during the 1980s. Virulence to Stb4 evolved in Z. tritici in California at some time before 2000.

40A superficially attractive option for breeding for resistance to STB, as in many other plant diseases, is to use qualitative genes with large effects on the pathogen. Synthetic hexaploid wheat may be a rich source of such genes. A persistent difficulty in applying this strategy in plant breeding is that many (but not all) single genes which confer strong resistance conform to the gene-for-gene relationship, while most gene-for-gene resistances (but not all) are readily overcome by the target pathogens.

41Breeding for resistance to STB can benefit greatly from the long history of breeding crops to control other diseases. While reliance on qualitative genes may reduce STB in the short-term, this approach is unlikely to provide durable resistance. A relevant comparison is with the successive use of gene-for-gene resistances against powdery mildew of barley or yellow (stripe) rust of wheat, a strategy which has been far from durable. The useful lifetime of qualitative genes can be extended by supporting them with high levels of ‘background’, usually durable, quantitative resistance.

42It is striking that some notable sources of STB-resistance, such as KK, Salamouni and TE9111 have several qualitative resistances. This suggests that ‘stacking’ or ‘pyramiding’ several Stb genes might improve the effectiveness of resistance, a strategy which has sometimes been effective in controlling crop diseases. Against this, the fact that most known Z. tritici isolates are virulent to most Stb genes (Section 5.1) suggests that the resistance achieved by gene-stacking may not be durable.

43Known individual Stb genes are not currently effective against Z. tritici populations in Europe (see Section 5.1; also ) and have not been durable (Section 5.2), although some are associated with minor quantitative resistance (Section 4.1). The majority of variation in field resistance to STB, therefore, must be controlled by quantitative resistance, as defined in Section 1, and the progress in breeding for STB-resistance over the last 30 years presumably happened by the gradual accumulation of minor genes. This type of resistance appears to be more durable than qualitative resistance. Its effectiveness may be gradually eroded but this happens much more slowly and to a lesser extent than the rapid evolution of virulence in a gene-for-gene interaction.

44A significant problem in the genetics of STB is that, when a gene has a large effect on resistance to the current pathogen population and therefore seems desirable as a source of resistance in breeding, there is currently no way of determining from its phenotype or underlying mechanism whether or not it might be durable. Qualitative genes which control detection of a specific pathogen genotype in a gene-for-gene relationship are much less likely to be durable than those that enhance downstream defences. The latter class of gene includes Lr34 in wheat against biotrophic pathogens, STV11 in rice against Rice stripe virus and several others. In rusts and powdery mildews, gene-for-gene interactions generate a hypersensitive response and reduce the infection type (IT) of pustules or colonies, whereas quantitative resistance tends to reduce the extent of symptoms rather than the IT. No such distinction can yet be made between the phenotypes of genotype-specific qualitative resistance and other, potentially more durable forms of resistance to STB. As Z. tritici is initially endophytic, becoming necrotrophic in its pathogenic phase, the only distinction between compatible and incompatible interactions known at present is the amount of disease visible on the leaf.

45If a plant resistance gene introduced from a wild population or a genetically diverse landrace population is effective against the current pathogen population, it will not necessarily be durable. If it follows the gene-for-gene relationship, virulence may be rare because the resistance gene has a fitness cost or the disease is not severe at the place of origin of the resistance gene. Stb16q is intriguing and potentially useful because it has a strong effect against all the large number of Z. tritici isolates with which it has been tested. Until it is isolated, however, or unless an isolate virulent to Stb16q is discovered, it will not be possible to tell whether it controls a gene-for-gene resistance which is effective against the current pathogen population, or is part of the plant’s downstream defences and therefore may be durable.

46Greater knowledge about mechanisms of STB-resistance would support wheat breeding, particularly by characterising the difference between gene-for-gene resistance and other kinds of resistance which may be more durable. This would help breeders to make informed decisions about the likely durability of resistance without the lengthy process of isolating the gene. A method of selecting both effective qualitative resistance, which may not be durable, and a good level of quantitative, possibly durable resistance in the same cultivar would be especially useful.

47In addition to disease resistance, STB levels can also be reduced by traits that contribute to disease escape, which limits the spread of fungal inoculum within crops. This typically happens in cultivars which are taller and later-heading, as both traits reduce the spread of spores to the upper leaves. Escape traits can be undesirable, however, because they can be maladaptive in terms of agronomic properties and yield.

48A general approach to increasing quantitative resistance (to any disease) stems from viewing plant breeding as a greatly accelerated form of natural selection, in which variation in traits is selected by breeders and inherited by the next generation of cultivars. The three essential requirements for breeding for effective, durable STB-resistance are diverse germplasm, efficient breeding processes which generate new combinations of genes, and field trial sites with high levels of STB at which resistant cultivars with good agronomic properties can be selected reliably and consistently. Once these fundamentals are in place, targetted selection of cost-effective genes or combinations of genes can contribute to raising the level of STB-resistance in new wheat cultivars.