Keywords

2.1 Introduction

Wheat (Triticum spp.) is globally the most important staple food of about 2.5 billion people (33% of the world population) and provides nearly 20% of the daily proteins and calories consumed globally (Breiman and Graur 1995). In terms of food security, it is the second most important food crop in the developing world after rice, with an estimated 80 million farmers relying on wheat for their livelihoods (Curtis et al. 2002). It is globally the most traded export commodity estimated at US$38.8 billion in 2019 (www.worldstopexports.com). Currently, wheat is the most widely grown cereal crop occupying more than 218 million ha. with a global production of 765 million tons, worth approximately US$150 billion (https://www.statista.com/statistics/267268/production-of-wheat-worldwide-since-1990/). For future food security, wheat production increase should target over one billion tons to cater to the needs of the rising population of estimated 9.6 billion people by 2050. Increasing consumption of diversified wheat products and quality profiles across countries will demand increased crop production and productivity in different wheat growing environments (Shewry and Tatham 2016). The continuous effort to increase genetic gains can only be possible by overcoming several of the current barriers such as climate change coupled with a variety of abiotic and biotic stresses that pose significant threat to wheat production both locally and globally. Genetic uniformity in the quest of developing high-performing cultivars has also contributed to vulnerability of rapidly evolving pathogens to the point wherein diseases threaten global wheat production.

2.2 Impact of Biotic Stresses on Wheat Production

On average, about 20% of the wheat production globally is lost due to pests and diseases every year (Anon 2014). Leaf rust became an increasingly important disease after the wheat variety “Thatcher” became susceptible in 1938, destroying millions of hectares in North America and since then it was considered as a damaging disease in USA, former USSR, and China (Chester 1946). Modern wheat cultivars continue to be affected by this disease worldwide. A cost benefit ratio of 1:27 was attributed for leaf rust resistant cultivar development at International Maize and Wheat Improvement Center (CIMMYT) (Simmonds and Rajaram 1988). The disease causes grain yield loss mainly by affecting both the number and weight of wheat kernels (Huerta-Espino et al. 2011). Yield losses between 2000 and 2004 due to leaf rust was estimated at US$350 million in the USA alone. In Mexico, yield losses to leaf rust accounted for US$32 million from 2000 to 2003, and subsequently US$40 million from 2008 to 2009, In South America (Argentina, Brazil, Chile, Paraguay, and Uruguay) between 1996 and 2003, the reported yield loss amounted to US$172 million. The average annual yield loss of 3 million tons was reported in China, whereas in Pakistan 10% yield loss to leaf rust were reported in 1978, estimated at US$86 million. In Australia, potential yield losses annually to leaf rust is estimated at AU$197 million under susceptible cultivars however, the use of resistant cultivars can minimize the loss to about AU$12 million (Huerta-Espino et al. 2011). In the first half of the 20th century, stem rust damaged 20% of wheat production in the USA with repeated epidemics between 1920 and 1960s. Yield loss ranging from 9% to 33% was recorded in Scandinavian countries in 1951 and records of 5–20% loss in eastern and central Europe were reported in 1932. Severe stem rust epidemics were reported in spring wheat grown in northern China and Inner Mongolia in the 1948, 1951, 1952, and 1956 cropping seasons (Sharma 2012). In Australia, sporadic epidemics of stem rust caused losses of £2–3 million (1889), £400,000 (1903), £2 million(1916), £7 million (1947), and the most severe loss was reported in New South Wales with estimated losses of AU$200–300 million (1973), which led to the establishment of the National Rust Control Program (Park 2007). A recent study estimated annual yield losses of wheat due to stem rust could reach 6.2 million tons globally, equivalent to US$1.12 billion (Pardey et al. 2013).

Chen (2005) reported 100% yield losses on stripe rust susceptible cultivars such as “PS 279”, and yield losses can range from 10 to 70% depending on cultivar susceptibility, time of the initial infection and degree of disease progress. In USA, severe losses by yellow rust (YR) was reported in four years (1958–1961, 1974–1978, 1980–1984, and 1999–2005) with significant damage in 2003 estimated at US$300 million. In South Africa, US$2.25 million loss was reported in 1998 two years post introduction. In 2002, China reported losses of 1.3 million tons of wheat grain to stripe rust. Yield loss of 20–40% between 1990s and early 2000s was also reported from Central Asian countries (Chen 2005). The 2003 epidemic of stripe rust in Australia resulted in damage amounting to AU$40 million (Wellings 2011). Fungal pathogens of diseases like rusts (Puccinia ssp.), powdery mildew (Blumeria graminis), Septoria leaf blotch (Septoria spp.) and Fusarium species are ranked among the most important fungal pathogens (Dean et al. 2012) in wheat growing environments where conditions are favorable for pathogen buildup. In low wheat production environments with lack of seed dressing treatments, diseases like smuts and bunts are common (Oerke and Dehne 2004) and in specific wheat-growing areas, fungal pathogens such as Pyrenophora tritici-repentis causing tan spot, Oculimacula spp. causing eyespot of wheat and Cochliobolus sativus causing spot blotchare of significant importance.

2.3 Implications of Climate Change

Climate change can create significant impact not only on the wheat production but also on pathogen dynamics both at regional and global scale. In individual wheat growing environments, shift towards warmer regimes and other climatic conditions such as altered precipitation may result in resurgence and adaptation of older and newer pathotypes of wheat diseases. These changes can also affect the seasonal phenology (better synchronization of pathogen life-cycle events with their host plants), the population dynamics (over-wintering and adaptation to warmer/cooler conditions, and the geographic distribution (expansion or retreat of specific pathogens with increased risk of pathogen incursion) (Chakraborty et al. 1998; Chakraborty and Newton 2011). The impact of climate change on wheat diseases has not yet been extensively studied. Some studies of potential impact of climate change on wheat diseases have already been reported (Chakraborty et al. 1998; Juroszek and von Tiedemann 2013). However, there is an increasing number of studies focusing on specific wheat diseases in relation to climate change, e.g. future changes of Fusarium foot rot in Australia (Backhouse and Burgess 2002), future Karnal bunt risk in Europe (Dumalasová and Bartoš 2009), worldwide changes of rust diseases in future (Chakraborty and Newton 2011), impact of climate change on leaf rust in France (Caubel et al. 2017), stripe rust in central and eastern USA (Lyon and Broders 2017), and very recently with the emergence of new stem rust races (Chakraborty et al. 2011; Juroszek and von Tiedemann 2013; Saunders et al. 2019) and Septoria tritici blotch risk in France (Gouache et al. 2013). These reports suggest that climate change may modify the range of prevalent wheat diseases in some regions that may turn, as a result currently economically less important wheat pathogens into potential threats in the near future (Duveiller et al. 2007).

2.4 Important Biotic Stresses Limiting Wheat Production in Different Environments

The following fungal pathogens are major biotic constraints in intensive wheat production systems worldwide. First, there are the obligate fungal pathogens (biotrophs); Blumeria graminis causing powdery mildew, Puccinia graminis tritici causing stem rust, Puccinia recondita/Puccina triticina causing leaf rust, and Puccinia striiformis causing stripe rust. Second, there are crop residue-borne necrotrophic pathogens; Pyrenophora tritici-repentis causing tan spot, Mycosphaerella graminicola causing Septoria tritici blotch, Phaeosphaeria nodorum causing Septoria nodorum blotch, Cochliobolus sativus causing spot blotch, and Fusarium graminearum and other Fusarium species causing Fusarium head blight or scab. There are many more fungal pathogens which are causing wheat diseases such as soil-borne root rots (Duveiller et al. 2007). In regions with low productivity and without seed dressing, smuts (e.g. common bunt caused by Tilletia caries) and bunts (e.g. Karnal bunt caused by Tilletiaindica) can be of significant importance (Oerke 2006). More than 40 viral diseases of wheat Triticum species have been documented; however, their significance is limited to specific geographic regions causing substantial yield losses. Viruses belonging to the genus Bymovirus (family Potyviridae) or the genus Furovirus (family Virgaviridae) are transmitted by the root-infecting plasmodiophorid Polymyxa graminis Ledingham (Rao and Brakke 1969) and some are insect-transmitted viruses. Insect transmitted viruses belong to the family Luteoviridae causing Barley Yellow Dwarf (BYD) disease transmitted by aphids, and the leafhopper-transmitted Wheat Dwarf Virus (WDV), a member of the genus Mastrevirus within the family Geminiviridae. Furthermore, the mite-transmitted Wheat Streak Mosaic Virus (WSMV) belonging to the genus Tritimovirus within the family Potyviridae are important viral pathogens of wheat. A comprehensive summary of wheat pathogens including fungi, viruses, and bacteria (and economically important animal pests) is well documented (Bockus et al. 2010). Economically important diseases in major wheat growing environments are discussed in detail below.

2.4.1 Rust Diseases

Cereal rust fungi are ubiquitous pathogens, known to occur in most wheat production environments causing substantial yield losses and very recently are considered a serious challenge to wheat production threatening the global wheat supplies (Bhavani et al. 2019). It is estimated that average annual losses to wheat rust pathogens range between US$4.3 to 5.0 billion globally (Beddow et al. 2015). Documented evidence suggest rust diseases could be one of the earliest pathogens wherein spores of stem rust dating back to 1300 BC were detected in Israel and also reported as serious disease of cereals in Italy and Greece (Kislev 1982; McIntosh et al. 1996). There are three wheat rust diseases, namely stem (black) rust, stripe (yellow) rust and leaf (brown) rust, all belong to the family Basidiomycota, genus Puccinia, and named P. graminis f. sp. tritici (Pgt), P. striiformis f. sp. tritici (Pst) and P. triticina (Pt), respectively (McIntosh 1998).

2.4.1.1 Stem Rust

Stem rust (SR), or black rust is common in warmer environments usually detected at later stages of crop growth (Roelfs et al. 1992). SR has the potential to completely destroy a healthy looking crop under epidemic situations and linear yield losses have been observed, with early infections can result in shriveled or no grain fill (Bhavani et al. 2019; Dean et al. 2012). SR epidemics have been significantly curtailed worldwide using various approaches; through eradication of barberry species between 1918 and 1980 in the USA (Singh et al. 2006) and in the UK, with the deployment of wheat germplasm carrying broad effective SR resistance genes and the use of fungicides. After effective control of SR for over three decades, the recent emergence of SR race “Ug99” in East Africa posed a serious threat to global wheat production (Bhavani et al. 2019; Singh 2006; Singh et al. 2015). The race Ug99 (TTKSK) caused widespread damage in Kenya (Pretorius et al. 2000; Singh et al. 2006, 2008a; Wanyera et al. 2006) carrying unique virulence as it was able to overcome over 50% of the known SR resistance genes including widely deployed genes Sr31, Sr38 and many other genes that were effective in different geographies (Singh 2006; Singh et al. 2008a). Ug99 race TTKSK was first identified in Uganda in 1999 and has spread through Africa and the middle east (Singh et al. 2015). The origin of the TTKSK race is unknown, it is genetically distinct from other stem rust races which indicate that this race did not evolve through mutations from other Pgt races (Olivera Firpo et al. 2015; Pretorius et al. 2007; Singh et al. 2015; Visser et al. 2011, 2019). Detection of several new variants within the Ug99 race group with the ability to overcome effective resistance genes substantially increased the vulnerability of varieties not only in East Africa (Jin and Singh 2006; Singh et al. 2008a, 2015; Bhavani et al. 2019) but predicted migration paths threatened production in other wheat growing environments (Singh et al. 2008a). In 2018, another new race with virulence to Sr8155B gene was identified in Kenya (S. Bhavani unpublished data) and currently, seven of the 14 variants within the Ug99 race group have evolved in Kenya, making it the hot spot for evolution of Ug99 race group (D. Hodson pers. communication).

Ethiopia reported devastating localized epidemics of SR on variety “Digalu” in 2013caused byrace TKTTF, a SR race unrelated to the Ug99 race group (Olivera Firpo et al. 2015), also previously reported in Turkey (Mert et al. 2012), Lebanon and Iran (Singh et al. 2015). In addition to Digalu race group, diverse SR races with rare combination of virulence to Sr9e and Sr13 have been found in the central highlands of Ethiopia (Admassu et al. 2009; Olivera Firpo et al. 2012). Unusual SR infections on winter and spring wheat were observed in 2013 season, which triggered concerns if Ug99 had migrated to Europe. Race analyses found six SR races, similar to variants of the Digalu race with additional virulence to Sr7a, Sr45, and SrTt-3 were identified (Olivera Firpo et al. 2017). A race TKKTP with virulence combination for Sr24, Sr36, Sr1A.1R and SrTmp genes has also been identified (Jin and Singh 2006). This race and the TRTTF race (from Yemen and Pakistan) are the only two known races that currently carry virulence to Sr1A.1R (Olivera Firpo et al. 2012). The re-emergence of common barberry has also accounted for SR epidemics in oats in Sweden (Berlin et al. 2013). The race TKTTF has also been detected in Germany, UK, Sweden and Denmark (Lewis et al. 2018). More recently the Sicily SR epidemic of durum wheat was also caused by the race TTRTF (Bhattacharya 2017) and recent studies reported its presence in Eritrea (Patpour et al. 2020).

2.4.1.2 Stripe Rust

Stripe (yellow) rust (YR) is a common disease in almost all wheat-growing environments. Even though YR is known to be well adapted to temperate areas with humid and cool weather (Rapilly 1979), races that are more aggressive and adapted to warmer temperatures have migrated and spread across geographies since 2000 (Ali et al. 2014; Hovmøller et al. 2010; Singh et al. 2015). Race shifts towards higher rates of mutation for virulence within the Pst pathogen (Hovmøller and Justesen 2007) has resulted in the vulnerability of widely deployed cultivars (Milus et al. 2015). Global estimate of yield losses to YR alone is 5.5 million tons per year (Beddow et al. 2015). Production losses in North America alone since 2000 exceeded over one million tons (Wellings 2011) and in China, over 1.8–6.0 million tons yield losses were observed under epidemic conditions (Wan et al. 2007). Similar reports of yield losses to YR in Europe in the recent decade have been attributed largely to the race shifts derived from the Himalayan region (Hovmøller et al. 2016). Historically, impact of newly evolved YR races on wheat productivity have been occasional, however, new incursions have often resulted in widespread damage, e.g. incursion of YR races from Europe into eastern Australia in 1978 (Wellings and McIntosh 1990) and western Australia in early 2002 (Hovmøller et al. 2008; Wellings et al. 2003). Exotic incursions of YR races replaced the existing populations in the USA since 2000 (Markell and Milus 2008; Milus et al. 2009) and race shifts in the European Pst populations in 2011 and 2012 by races from the Himalayan region (Hovmøller et al. 2016; Hubbard et al. 2015) are very good examples of exotic races with different genetic Pst lineages causing significant impact on host susceptibility. A recent study linking both virulence and race structure with recent YR epidemics in different geographies (Ali et al. 2017) suggested different Pst races in distinct genetic lineages, where aggressive strains adapted across diverse environments were spreading across continents, including the more recent outbreak of YR in Argentina (Hovmøller et al. 2008; Carmona et al. 2019).

2.4.1.3 Leaf Rust

Leaf (or brown) rust (LR), is the most common rust disease in both winter wheat and spring wheat growing areas as well as in durum wheat. Yield losses due to LR can be substantial if susceptible varieties are infected at early stages coupled with favorable temperatures and moisture conditions resulting in rapid progress in short time span. Yield losses are largely due to the reduction of kernels per spike and lower kernel weights (Chester 1946). LR shows widespread adaptation from warm to hot weather, such as the great plains of North America, southeast Asia, Russia and central Asia to southeastern US, Mexico, Uruguay, Argentina, Turkey, China and southern Europe. Populations of Pt are specifically adapted to either tetraploid durum wheat or hexaploid common wheat (Anikster et al. 1997) and races conferring virulence to several of the LR genes are prevalent throughout the world (Roelfs et al. 1992). Since the early 2000s, races of Pt that are highly virulent on durum wheat cultivars have spread across South America (Ordoñez and Kolmer 2007), Mexico (Singh et al. 2004a), Europe (Goyeau et al. 2006), the Mediterranean basin, and the Middle East (Kolmer 2001).

On a global scale, most populations of Pt are unique in their virulence and molecular genotypes. Even though the most common mode of evolution is through mutation and selection in a given environment, there is evidence for recent migration of Pt races between different continental regions. Since the mid-1990s, isolates of Pt with virulence to Lr1, Lr3a, and Lr17a and avirulence to Lr28 have increased and spread across the US and Canada (Kolmer 1998; Long et al. 2000). These isolates also had a unique molecular genotype, which indicated that these were likely recently introduced to North America (Kolmer and Anderson 2011). Since the early 2000, these isolates with identical or highly similar virulence and molecular genotypes have been found in Europe (Kolmer et al. 2013), South America (Germán et al. 2007), Ethiopia (Kolmer and Acevedo 2016), Turkey (Kolmer et al. 2013) and Pakistan (Kolmer et al. 2017). Similarly isolates of Pt with virulence to durum wheat that also have identical or highly related molecular genotypes have been found in the Middle East, South America, Europe, Ethiopia, Tunisia, Mexico and the US (Ordoñez and Kolmer 2007).

2.4.2 Non-rust Diseases

2.4.2.1 Powdery Mildew

In contrast to rusts, powdery mildew caused by B. graminis f. sp. tritici is more common in humid rain-fed conditions or irrigated conditions, which favor successful infection. The disease strongly influenced by the amount of nitrogen application, large single doses or excessive multiple applications of N fertilizers can result in serious outbreaks (Chen et al. 2007). Cooler and humid regions of Asia, Japan and North and East Africa, Northern parts of Europe and America are regions where powdery mildew is an important pathogen. Early infection of powdery mildew stimulates non-productive tillers, which reduces food reserves affecting the grain yield and low levels of disease in susceptible varieties can still reduce yield significantly (Bowen 1991; Everts 1992). Reduction in yield was significant when high disease severity was observed at Feekes stage 10 (booting stage) (Large 1954) and at Feekes 9 (expanded flag leaf), susceptibility ratings of the Feekes 1 to 3 leaves were most useful yield predictions. Application of fungicide at Feekes 9 or earlier stages is important if disease has been detected early with faster progress (Royse et al. 1980). Yield losses up to 40% have been observed and losses are related to the reduction in grain size and number per unit area, which largely depend on host resistance/susceptibility (Bowen 1991; Royse et al. 1980). Impact of powdery mildew can result in reduced flour protein but has no significant effect on milling and baking quality (Johnson et al. 1979).

2.4.2.2 Fusarium Head Blight

Fusarium head blight (FHB) is one of the most devastating disease of wheat globally, with major epidemic regions in North America, Europe, East Asia and the Southern Cone of South America. Many species in the genus Fusarium cause FHB, but it is F. graminearum species complex that has global importance and has been found in all major epidemic regions. The disease is favored by warm and humid environment around anthesis, leading to yield reduction and quality deterioration. More importantly, the disease produces a range of mycotoxins, particularly deoxynivalenol (DON, or vomitoxin), which are toxic to humans and animals, raising a serious concern to food and feed safety (Buerstmayr et al. 2020). In many countries, regulations on DON in wheat and its products have been set up, and the market price of wheat grain may be significantly reduced if DON content exceeds a certain threshold. In USA, losses attributable to FHB in wheat and barley between 1993 and 2001 were estimated at US$7.67 billion (McMullen et al. 2012). In China, the epidemic has increased significantly in the last two decades, amounting on average 5.3 Mha and reached 9.9 Mha in the 2012 great epidemic (Zhu et al. 2019). Yield reductions can reach up to 50–70% in Europe and South America (Buerstmayr et al. 2020).

2.4.2.3 Wheat Blast

Wheat blast (WB) caused by the ascomycetes fungus Magnaporthe oryzae pathotype triticum (MoT) is one of the devastating diseases in warm and humid wheat growing regions. It can infect all the aerial parts of wheat, but completely or partially bleached spike is the typical symptom. WB was initially identified in the Parana state of Brazil in 1985; afterwards, its rapid widespread to the neighboring states in Brazil and other countries of South America raising serious concerns. The recent outbreak in Bangladesh in 2016 raised a major concern on wheat production in South Asia (SA), as nearly 7 Mha of the wheat growing areas in SA are vulnerable to WB. More recently, occurrence of WB has been reported from Zambia which can be a major threat for wheat production and trade in Africa (Tembo et al. 2020). Under favorable temperatures of 25–30 °C and high humidity, the disease can cause high yield loss ranging from 10 to 100% depending upon the level of infection (Ceresini et al. 2016).

The long-distance spread of the pathogen occurs through infected seeds, followed by the air transmission; therefore, seed quarantine and chemical treatment can effectively manage the primary inoculum load. For field WB management, foliar fungicides application such as demethylation inhibitors (DMI), quinone outside inhibitors (QoI) and succinate dehydrogenase inhibitors (SDHI) are suggested to be used in combination/rotation so as to reduce the fungal resistance against the fungicides (Cruz and Valent 2017). Various agronomic practices viz. optimizing planting dates, weed management, crop rotation with non-hosts and avoiding excessive nitrogen application are reported to be effective in WB control. However, all these measures do not work well under high disease pressure, thus they should be used in combination with genetic resistance to achieve a better management.

2.4.2.4 Karnal Bunt

Tilletia indica (syn. Neovossiaindica) is a hemibiotrophic fungus which was first described to cause disease in the Indian city of Karnal, hence called ‘Karnal bunt’ (KB). Currently the disease is distributed in parts of Asia (India, Nepal, Pakistan, Iraq, Iran, Afghanistan), Africa (South Africa), and the Americas (USA, Mexico, Brazil). Though the estimated average yield losses due to KB are as low as 0.01–1%, it is an important disease from international trade perspective where many member countries of WTO use it as a non-tariff barrier. KB significantly deteriorates the wheat quality in terms of reduced vitamins, amino acids, weakened dough, and loss in flour recovery, ultimately affecting the human consumption negatively (Bishnoi et al. 2020).

The conducive conditions for disease development are high humidity with cool temperature (<20 °C) favoring teliospore germination. Infected spikes disperse teliospores that become inoculum for the next season, and the teliospores are reported to remain viable for long durations, indicating the spatial and temporal dispersal capability of the disease (Carris et al. 2006). Identifying the disease in field is difficult due to confounding the symptoms with other bunts, thus, making the laboratory and molecular confirmation essential. Laboratory confirmation includes observing teliospores under the microscope for specific morphological characteristics, and molecular characterization adds precision to teliospore morphology with T. indica specific markers (Bishnoi et al. 2020; Kumar et al. 2021).

2.4.2.5 Tan Spot

Tan spot (TS) is caused by the necrotrophic fungus Pyrenophora tritici-repentis (Died.) Drechs. The disease frequently appears in the warm and humid growing regions of bread and durum wheat, especially Canada, Australia, USA and South Africa. Yield and quality losses are common under high disease pressure. Reduced or no-till approaches to prevent soil erosion are important reasons for increased disease pressure. Residue from previous crop carrying pathogen inoculum is considered one of the main inoculum sources. Another major reason that corresponds with increased pathogen virulence is acquisition of a host-selective toxin (HST) PtrToxA by P. tritici-repentis from Stagonospora nodorum via horizontal gene transfer which overcame the resistance of most cultivars carrying Tsn1 gene (Friesen et al. 2007). Based on type of lesion (chlorosis or necrosis) and HSTs produced, P. tritici-repentis is classified into eight races using six differential genotypes (Table 2.1).

Table 2.1 Reaction of eight characterized races of Pyrenophora tritici-repentis on bread and durum wheat differential lines. Resistance and susceptible response are indicated as R and S, respectively
Full size table

2.4.2.6 Septoria NodorumBlotch (SNB)

Stagonospora nodorum, a filamentous ascomycetes fungus, causes wheat leaf and glume blotch and affects wheat yield and quality in the warm and humid areas particularly in Australia, USA, parts of Europe and southern Brazil. Short incubation period enables the pathogen for multiple infection cycles within a season. The fungus can reproduce through asexual conidia and frequent sexual reproduction due to availability of both mating types (MAT1-1 and MAT1-2). Sexual reproduction creates large genetic variability and best-fit strains multiply asexually, having great potential to overcome the effects of resistance genes or fungicides. Therefore, focus should be on the enhancement of quantitative/horizontal resistance in the targeted wheat population, from a breeding perspective (Cowger et al. 2002). SNB produces multiple HSTs, of which 15 have been identified so far. The HSTs (e.g. SnToxA) interact with the corresponding host sensitivity genes (e.g. Tsn1) in an ‘inverse gene-for-gene’ manner that causes infection in the host, just as in tan spot. So far, nine necrotrophic effectors (NE) and sensitive gene interactions viz. SnToxA-Tsn1, SnTox1-Snn1, SnTox2-Snn2, SnTox3-Snn3-B1, SnTox3-Snn3-D1, SnTox4-Snn4, SnTox5-Snn5, SnTox6-Snn6, and SnTox7-Snn7 have been identified in wheat. Three important NE genes in the pathogen viz. SnToxA, SnTox1, SnTox3 and one important host sensitivity gene in wheat viz. Tsn1 have been cloned which has helped in the extensive study of three important interactions viz. SnToxA-Tsn1, SnTox1-Snn1 and SnTox3-Snn3-B1 for better understanding the molecular basis of SNB (Ruud et al. 2019). These studies have indicated that one interaction may enhance or suppress the other interactions depending upon the genetic backgrounds of pathogen/cultivar, which is important from a breeding perspective (Ruud et al. 2017). Tsn1 was identified on chromosome 5BL (Faris et al. 2010), whereas both Snn1 and Snn3-B1 were mapped on 5BS (Ruud et al. 2017). Negative selection of host sensitivity genes during the breeding program would accelerate the breeding progress of resistant varieties.

2.4.2.7 Spot Blotch

Spot blotch (SB) caused by Bipolarissorokiana (telemorph Cochliobolus sativus) is a destructive disease of wheat in the warm and humid growing regions, especially South Asia, Latin America and Southern Africa. The pathogen causes average yield loss of 15–20%, but under favorable environmental conditions yield loss of up to 87% has been detected on the susceptible varieties(Gupta et al. 2018). The pathogen can infect all parts of the wheat plant, but leaf infection is the most typical where infection starts from the older leaves and then progresses upward towards the younger leaves. High temperature (18–32 °C) and humidity (>90%) favors the disease establishment.

2.4.2.8 Septoria Tritici Blotch

Septoria tritici blotch (STB) is caused by the fungal species Zymoseptoria tritici (teleo. Mycosphaerella graminicola). The pathogen is heterothallic with two mating types and thus has frequent sexual reproduction, resulting in a high level of genetic variation (Cowger et al. 2002). Additionally, Z. tritici can make multiple infections during a cropping season, greatly accelerating its evolving speed, leading to a series of problems in STB management, such as break down or erosion of host resistance and fungal resistance to fungicide. Losses to STB can range between 30 and 50% only during severe epidemics in areas with extended periods of cool, wet weather, particularly North America (USA, Canada, Mexico), East Africa (Ethiopia, Kenya), South America (Brazil, Chile, Uruguay, Argentina) and the most damage occurs in Europe and CWANA (Central and West Asia and North Africa) region (Van Ginkel et al. 2002).

Because fungicides usually exhibited low efficiencies in STB control, multiple fungicidal applications are often needed under high disease pressure, leading to high costs to wheat farmers. In Europe, about 70% (US$1.2 billion) of the total fungicide application in cereal crops were for controlling STB. Nevertheless, this disease has been compromised by the emergence of fungicide resistance in Z. tritici, and the new regulation in EU on reducing fungicide application favors the active utilization of other STB management strategies such as host resistance (Torriani et al. 2015).

2.4.2.9 Root Diseases

The healthy root system of a wheat plant is the key for the water and nutrient uptake. Root rots are among the major diseases of wheat resulting in a significant yield loss throughout the world and are found wherever cereal-based farming system dominates (Cook 2001). Because the roots are not typically visible, symptoms of root rot become apparent only when the disease is advanced. Root rot pathogens in cereals include Heterodera species, cereal cyst nematode (CCN), Pratylenchus species, root-lesion nematode (RLN) and many fungal species. Among the latter are Gaeumannomyces, Pythium, Rhizoctonia, Fusarium, Bipolaris genera, and different species of these genera are favored by different soil, cropping system and climate (Cook 2001). The pathogens have a wide host range and can survive in the soil/organic residue for many years. Root rot symptoms are difficult to identify clearly but generally are characterized by discoloration of roots, coleoptiles and stem bases of the infected seedling. Root rot fungi also may attack the upper parts of plants which may result in foliage lesions, head and seedling blight.

Take-all (Gaeumannomyces graminis) is the dominant root disease favored by the moist and cool conditions in winter season followed by the moisture stress during anthesis. There is often a build-up of antagonistic microorganisms following one or two take-all outbreaks, turning soils to be suppressive and subsequently reducing the disease. Registered fungicides might be an option to control the disease (Cook 2001; James Cook 1992). Pythium is a pathogen having a wide host range causing root rot and seedling damping off. Pythium infects root system via root tips and root hairs and can also penetrate the embryo of germinated seed. Main Pythium symptoms are stunting and yellowing of leaf tissue, which sometimes may be misdiagnosed as nitrogen deficiency. Infected roots are stunted and light brown-yellow coloration is seen near the tips. Infected plants develop poorly filled heads and is often misdiagnosed as Rhizoctonia damage. Rhizoctonia disease can prune off the root and causes water and nutrient stress which causes crop damage. It survives in the top of the soil (0–10 cm) on organic matter (Cook et al. 2002). Fusarium spp. especially F. culmorum and F. pseudograminearum cause root diseases on cereals, including foot rot, root rot, and crown rot. Crown rot is the most widely accepted name for this disease, which encompasses symptoms on the lower part of the wheat plant, including the subcrown internode, crown, crown roots and lower stem including nodes and internodes. Diseased plants are characterized by fungal colonization on the wheat stems, crown and root tissues leading to a honey-brown discoloration of the leaf sheaths and lower stem, and necrosis of the crown region (Scherm et al. 2013). Bipolaris spp. especially B. sorokiniana cause common root rot of wheat worldwide, which produces a brown to black discoloration of the sub-crown internode.

Three major species belonging to CCN viz. Heteroderaavenae, H. latipons, and H. filipjevi, are distributed worldwide and cause severe damage in cereals. The Pratylenchus species, especially P. thornei, P. crenatus, P. neglectus and P. penetrans, are widely distributed pathogens for RLN (Dababat et al. 2014). CCN is monocyclic as it completes only one cycle per season while RLN is polycyclic due to a higher multiplication rate of three to five generations per year. RLN causes stunted and poorly tillered plants. The badly damaged roots are thin and poorly branched with short and knotted laterals. CCN can be identified easily through patches and stunted plants. Below-ground symptoms are white females on roots which can be seen with naked eyes in springtime. Identifying which root rot pathogen is present in the field by classical and/or molecular tools is the most important point to tackle the disease (Table 2.2).

Table 2.2 Characteristics of the root rot diseases
Full size table

2.5 Prospects of Genetic Control, Types of Resistance, Strategies to Deploy Different Resistance Mechanisms to Attain Resistance Durability

Even though numerous pests and diseases are known to reduce grain yield and quality in wheat, the three rusts, powdery mildew fungi and other head, foliar and root diseases, continue to be economically important in spite of the extensive use of host resistance and fungicides. The evolution and spread of virulent and aggressive race lineages of rust fungi threaten wheat production worldwide. Fusarium head blight, leaf-spotting diseases, including more recently, wheat blast (in South America and Bangladesh) have become significantly important in recent years. High diversity for race-specific and quantitative resistance is well known for most diseases. Selection through field phenotyping coupled with complementing molecular tools/strategies can offer great promise in achieving durable resistance and enhancing wheat production and productivity. Disease resistance remains a core trait for several plant-breeding programs, and a complete package of high yield, disease resistance, agronomic performance and end-use quality is most preferred in varieties that are released globally.

2.5.1 Types of Resistance

There are two main ways to control diseases in wheat viz. incorporating genetic resistance through breeding and chemical control using fungicides. Genetic control has advantages for environmental and economic reasons, particularly for resource poor farmers in the developing world and the possibility that rust pathogens develop resistance to fungicides (Carmona et al. 2020). Genetic resistance deployed by wheat breeders belong to two general classes of genes based on their phenotypic effects, pathogen race-or strain-specific resistance (R-genes) and adult plant resistance (APR) genes. R-genes mostly function at all growth stages whereas APR genes function mainly at the adult stage. Wheat rust resistance genes of both R and APR classes are designated Lr, Sr, and Yr for leaf, stem, and stripe or YR resistance, respectively.

2.5.1.1 Race-Specific/Seedling Resistance

Race specific, or seedling resistance/all stage resistance/qualitative resistance is effective at all growth stages and belongs to the “R-gene” class (Ellis et al. 2014). R-genes are perceived to confer a major resistance effect/complete resistance, However majority of the R-genes conferring rust resistance do not confer clean phenotype (McIntosh et al. 1996) and some are influenced by varying temperature and light regimes (Chen et al. 2015; Chen 2013; Forsyth 1956). The ease of selecting these genes at both seedling and field stages has made it easier to incorporate such resistance in several wheat breeding programs. However, deployment of single R-genes has often resulted in pathogen acquiring virulence post deployment as varieties in a short period leading to “boom and bust cycles” e.g. widespread virulence for Yr9 and Yr27 genes (Hovmøller et al. 2008), virulence for Sr31 gene and other important SR genes Sr24, Srtmp to the Ug99 race group (Jin et al. 2008; Patpour et al. 2016; Pretorius et al. 2000) and ineffectiveness of LR resistance genes in the United States (Kolmer and Hughes 2015). However, deployment of genes in combination often referred as “pyramiding” can effectively enhance durability of resistance and keep pathogen populations under check.

2.5.1.2 Adult-Plant Resistance (APR) Genes Conferring Pleiotropic Effects

Race-nonspecific resistance often referred as adult plant resistance or partial resistance is effective against wider races of a pathogen species and/or effective against broad range of pathogens. APR is generally quantitative, exhibiting incomplete resistance that is usually expressed at later stages of plant development. These genes help slow the disease progress through increased latency period, reduced infection frequency, reduced pustule size and thus resulting in lower spore production. The phenotypic effects of such genes is relatively minor or inadequate when alone, however, additive effects of such minor APR genes (4–5) in combinations (Knott 1988; Singh et al. 2004b, 2015) can result in enhanced levels of resistance.

Lr34 was first reported in cultivar “Frontana” (Dyck et al. 1966), although it has been a part of wheat improvement since the early 20th century. Wheat cultivars containing Lr34 are widely present and occupy more than 25 million ha in developing countries and is effective in reducing yield losses in epidemic years (Marasas et al. 2003). The Lr34 gene has remained durable as virulence for this gene has not been observed for more than 60 years. Lr34 is located on the short arm of chromosome 7D (Dyck 1987). This gene confers modest levels of resistance and has pleiotropic effects on resistance to multiple diseases such as YR, SR, powdery mildew, barley-yellow dwarf virus and spot blotch (Lr34/Yr18, Sr57, Pm38, Bdv1 and Sb1), respectively (Krattinger et al. 2009b; Lagudah et al. 2006, 2009; Lillemo et al. 2007). Lr34 is associated with a morphological marker expressed as leaf tip necrosis (LTN) on the flag leaves, which can be used as a phenotypic marker (Singh 1992b). Lr34 was cloned and the gene encodes a full-size ATP-binding cassette (ABC) transporter (Krattinger et al. 2009b). Based on the knowledge of the Lr34/Yr18 gene sequence, gene-specific markers were developed and have proven to be highly diagnostic for the Lr34 gene (Lagudah et al. 2009).

Lr46 was first described in 1998 in cultivar “Pavon 76” (Singh et al. 1998) and is located on chromosome 1BL. The latency period of infected adult plants carrying Lr46 was significantly lower compared to the controls without the gene (Martínez et al. 2001). The resistance type conferred by Lr46 is similar to that of Lr34, although smaller in effect and is also known to confer partial APR to YR, SR, powdery mildew with corresponding designations Yr29, Sr58 and Pm39, respectively (Singh et al. 2015; Bhavani et al. 2019). Lr46 is also associated withLTN and is very common in both old and new wheat varieties including durum wheat (Lan et al. 2017a).

The Lr67 gene was identified in the common wheat accession “PI250413” (Dyck and Samborski 1979) and transferred into “Thatcher” to produce the isoline “RL6077” (Thatcher*6/PI250413). Lr67 shows similar pleiotropic effect as Lr34 due to the association with resistance to SR (Dyck et al. 1994) and YR (Singh 1992a). However, Lr67 confers a lower level of LR resistance than that conferred by Lr34 (Hiebert et al. 2010). It was earlier assumed that the gene in RL6077 could be Lr34 translocated from chromosome 7D to a different chromosomal location, however later studies showed that Lr34 is not present in RL6077 (Lagudah et al. 2009). Recent studies mapped Lr67/Yr46/Pm46 on chromosome arm 4DL (Hiebert et al. 2010; Herrera-Foessel et al. 2014). Cloning elucidated that Lr67 gene encodes a hexose transporter (Moore et al. 2015). Lr68 is another APR gene located on chromosome arm 7BL, that confers slow rusting resistance to wheat LR (Herrera-Foessel et al. 2012). This gene was first described in CIMMYT’s spring bread wheat “Parula” (Pedigree: FKN/3/2*Frontana//Kenya 350 AD.9C.2/Gabo 55/4/Bluebird/Chanate). Parula was developed by CIMMYT in 1981 and is also known to carry Lr34 and Lr46 (William et al. 2007) and likely origin of Lr68 is the Brazilian cultivar “Frontana” (Herrera-Foessel et al. 2012). Lr68 showed a weaker effect than Lr34, Lr46 and Lr67 but combined effect of Lr34, Lr46 and Lr68 in Parula resulted in near immunity (Lillemo et al. 2011; Herrera-Foessel et al. 2012), however its effect on other diseases could not be determined.

Stem rust gene Sr2 is one of the most important and widely used gene, confers modest levels of resistance and has been effective until date (over 100 years) even to the more recent and virulent Ug99 and Digalu race groups of SR in East Africa. This gene was transferred from “Hope” and “H-44” into common cultivars (McFadden 1930) and is derived from a tetraploid “Yaroslav” emmer. The Sr2 gene is located on chromosome arm 3BS. This gene was widely used by Dr. N. E. Borlaug when he initiated wheat breeding in 1944 in Mexico, which resulted in varieties such as “Yaqui 50” and several high yielding semi dwarf varieties that were deployed in different wheat programs (Singh et al. 2015). The Sr2 gene shows pleiotropic effects with YR resistance gene Yr30 that also confers moderate resistance. Sr2 gene is also associated with a morphological marker called pseudo-black chaff (PBC) that is expressed as purplish pigmentation on the glumes, internodes and peduncles under favorable conditions. Efforts to combine Sr2 with other minor effect genes to enhance SR resistance in breeding materials at CIMMYT has resulted in several resistant or moderately resistant varieties. Several new uncharacterized slow rusting genes, some potentially pleiotropic, have been identified in the recent years (Rosewarne et al. 2013; Li et al. 2014; Yu et al. 2014) suggesting diversity for APR QTL and their potential in breeding resistant varieties.

Other adult plant resistance genes reported to confer partial or slow rusting include Lr74 (Kolmer et al. 2018b), Lr75 (Singla et al. 2017), Lr77 (Kolmer et al. 2018c), and Lr78 (Kolmer et al. 2018a) for LR, Yr11, Yr12, Yr13, Yr14, Yr16, Yr36, Yr39, Yr52, Yr59, Yr62, Yr68, Yr71, Yr75, Yr77, Yr78, Yr79, Yr80 and Yr82 (Chen and Kang 2017; Feng et al. 2018; McIntosh et al. 2017; Nsabiyera et al. 2018; Pakeerathan et al. 2019) for YR and more recently Sr56 identified in cultivar ‘Arina’ for SR (Bansal et al. 2014).

Currently, over 220 rust resistance genes viz. 79 LR resistance genes (Qureshi et al. 2018a), 82 YR resistance genes (Pakeerathan et al. 2019) and 72 SR genes (Chen et al. 2020; McIntosh et al. 2017) have been formally cataloged and designated of which majority of them confer race specific resistance and only a few genes confer slow rusting/partial adult plant resistance to the three rust diseases.

2.5.2 Enhancing Durable Rust Resistance in Wheat Breeding Germplasm at CIMMYT

Wheat breeding at CIMMYT focuses on small-holder farmers across wheat growing regions of Asia, Africa and Latin America, and strongly emphasizes selecting high-yielding wheat germplasm that possesses good levels of rust resistance based on diverse combinations of multiple pleiotropic resistance genes and other QTLs with significant progress made for all three rusts (Singh et al. 2015; Bhavani et al. 2019). Breeding for rust resistance has been a rigorous exercise owing to the continued evolution and selection of pathogen for new virulence to previously effective resistance genes largely through mutation or sexual recombination, or transboundary migration of races to new wheat production environments. In most developing countries, varieties with genetic resistance are preferred by farmers; therefore, resistance is a required trait for release. Even though several race-specific resistance genes have been identified only a handful of genes are used actively in breeding as several genes are only effective in certain environments and majority are easily overcome in few years of deployment, linkage drag associated with genes transferred from secondary and tertiary gene pools or originating from unadapted genetic backgrounds.

One of the best approaches to utilize these race-specific resistance genes is through pyramiding combinations of multiple effective genes in varieties. Molecular markers linked to some of the effective resistance genes have facilitated the selection for multiple resistance genes and releases of varieties that carry them. However, the lack of diagnostic markers to select genes in different genetic backgrounds leaves no option but to use field-based selections under artificial epidemics, which continues to be the most common practice in several breeding programs.

Other approach is to utilize quantitative APR in breeding, although the individual effects of pleiotropic APR genes and other QTLs are small or moderate in their effect when present alone; near-immune levels of resistance have been achieved by combining 4 to 5 of these genes that often have additive effects (Singh et al. 2008b, 2015). Incorporating such type of resistance has been found to enhance durability and significant progress was made for LR resistance, and more recently for resistance to Ug99 race group and stripe rust resistance in CIMMYT germplasm using a single back cross selected bulk scheme (Singh et al. 2015, 2016). Although breeding for APR is cumbersome initially, additive effect of multiple minor APR genes enables combinations of high disease resistance, which can be simultaneously selected together with high yields with appropriate agronomic traits and the frequency of these genes can be increased within the breeding germplasm. Comparison of grain yield performance of 697 EYT lines (Stage II) 2018–19 derived from Mexico Shuttle and Mexico Kenya Shuttle breeding schemes identified similar frequency of lines that combine high yield potential and SR resistance (Fig. 2.1) and significant progress has been achieved in combining yield potential and rust resistance in CIMMYT breeding lines.

Fig. 2.1
figure 1

Performance of grain yield of 697 EYT lines (Stage II) 2018–19 derived from Mexico Shuttle and Mexico Kenya Shuttle breeding schemes

Full size image

One of the prerequisites for enhancing APR is the absence of epistatic race-specific resistance gene interactions in breeding materials, which enables selection of transgressive segregants with low disease severity under high disease pressure to select for combinations of APR based on their additive effects. The progress in breeding APR to the Ug99 race group was facilitated by extending shuttle breeding scheme and testing between field sites in Mexico and Njoro, Kenya. Combinations of diverse multiple minor genes based APR is especially important to curtail the evolution of new virulent races in most wheat growing environments. Majority of the race specific genes also condition intermediate resistance phenotype and interactions of these moderately effective genes in good APR backgrounds have also enhanced the resistance to the rust diseases in CIMMYT germplasm. We have shown that the success in achieving high levels of complex APR to rusts in the CIMMYT high-yielding germplasm has enhanced resistance durability, provide excellent yield protection, and free up resources to focus on much needed, accelerated yield enhancement and make progress toward resistance to other diseases that are gaining importance. Despite significant progress, incursion and evolution of new races in East African region both for SR and YR has rendered some R-genes carrying varieties susceptible, favourable climatic conditions and incursion of new YR races into Europe, Asia and the America’s has further compounded the biotic stress constraints in the region especially with dependence on R-genes based resistance (Hovmøller et al. 2008, 2016; Milus et al. 2009; Olivera Firpo et al. 2015; Singh et al. 2015; Ali et al. 2017; Lewis et al. 2018).

2.5.3 Genetics and Breeding of Other Wheat Foliar and Root Diseases

Many race-specific resistance genes against powdery mildew are deployed in wheat cultivars. Currently, 68 resistance genes conferring resistance to wheat powdery mildew are known (He et al. 2020). However, cultivars in general carry only one or very few resistance genes which results in selective pressure on the pathogen population to acquire virulence and therefore, resistances are of low durability (Parks et al. 2008; Shah et al. 2018). Several of these race-specific resistances have been easily overcome by simple genetic changes in the pathogen, e.g. Pm8 and Pm17 were overcome already in the 1980s. Pm4a was overcome in some areas of China and Pm21 was overcome after extensive use in Europe (Gao et al. 2012). Therefore, deployment of race-nonspecific pleotropic resistances, such as Lr34[Syn. = Yr18 = Sr57 = Pm38 = Sb1 = Bdv1 = Ltn1], Lr46 [Syn. = Yr29 = Sr58 = Pm39 = Ltn2] and Lr67[Syn. = Yr46 = Sr55 = Pm46 = Ltn3], which is also effective against powdery mildew in combination with other resistance genes, can prevent the emergence of new virulent races and enhance durability. In addition, more than 100 QTLs have been identified and can be employed in marker-based selection procedures (Keller et al. 1999; Marone et al. 2012; Asad et al. 2014). Genome editing technology in the recent years has shown great potential to surpass the bottlenecks of conventional resistance breeding. This technology offers the modification of specific target genes in elite varieties, thus bypassing the whole process of crossing. Recent advances in gene-editing technology can also offer avenues to building resistance durability. Genome editing was found to be effective in improving powdery mildew resistance by editing Mlo homologs in wheat to produce a triple knockout in hexaploid wheat (Wang et al. 2014c). As gene-editing technology develops, site-specific editing of alleles may become practical in the future.

FHB resistance is a typical quantitative trait, conditioned by numerous genes of minor effects. The complexity of FHB resistance also lies on the different resistance mechanisms, e.g. resistance to initial infection (Type I), resistance to fungal spread in the rachis (Type II), resistance to toxin accumulation (Type III), resistance to kernel infection (Type IV) and resistance/tolerance to yield reduction (Type V). Numerous sources of resistance were reported in literature but only a few have been successfully utilized in breeding programs, such as ‘Sumai 3’, ‘Wuhan 1’, ‘Frontana’ etc. (Buerstmayr et al. 2020). Other than these accessions, many wheat lines carrying so-called ‘native’ resistance have been reported and utilized in regional breeding programs (Brar et al. 2019). FHB resistance genes/QTL have been mapped on all the 21 wheat chromosomes, though, only seven QTL have formally been designated as Mendelized genes, of which only Fhb1, Fhb2, Fhb4 and Fhb5 are from common wheat, whereas Fhb3, Fhb6 and Fhb7 are from wild wheat relatives (Bai et al. 2018). So far, only Fhb1 and Fhb7 have been cloned, and their functional markers have been developed for marker-assisted selection (MAS) (Su et al. 2018; Wang et al. 2020).

Breeding for FHB resistance per se is not difficult and high level of resistance comparable to the famous resistant source ‘Sumai 3’ is achievable. However, in breeding practices, FHB resistance is often linked to unfavorable traits such as low yielding, late maturity and high stature (Buerstmayr et al. 2020) and it is often difficult to reconcile FHB resistance and other preferred traits. At CIMMYT, two breeding strategies for FHB are being used, i.e. exploitation of native resistance and introduction of exotic resistance. There is no strong FHB resistance available in the current CIMMYT gene pool, though some moderately resistant lines have been identified and a few QTL with major effects have been mapped. Among those lines are ‘Shanghai3/Catbird’, ‘Mayoor’, ‘Soru#1’, ‘IAS20*5/H567.71’ etc. It is noteworthy that a major QTL on chromosome 2DL has been consistently identified in the first three genotypes and haplotype analysis of a few FHB Screening Nurseries of CIMMYT also demonstrated its high frequencies. Apart from this QTL, others are either of low frequencies or of minor effects but higher level of resistance can still be achieved via accumulating those QTL in elite breeding lines, similar to rust resistance breeding (Singh et al. 2016). The limitation of using native resistance is, however, a lack of QTL/gene with strong Type II resistance, which could be compensated via introduction of exotic FHB resistance genes like Fhb1 and Fhb7. The former is the most well-known FHB resistance gene and has been extensively utilized in China, USA and Canada (Zhu et al. 2019), however, its repulsive linkage with the SR gene Sr2 limited its application in the CIMMYT wheat breeding. To address this problem, several recombinant lines with both Fhb1 and Sr2 were introduced from Australia and included in various crosses with elite CIMMYT breeding lines. Many of the progenies exhibited good agronomic traits as well as promising resistance to FHB and other diseases (Xu et al. 2019).

Since no immunity to FHB has been found in wheat and high level of FHB resistance is difficult to achieve, other disease management strategies are also important in wheat production regions where FHB is a limiting factor. Removal of crop residue and rotation with non-host crops are helpful in reducing inoculum concentration. It is well known that maize-wheat rotation greatly increases the risk of FHB and thus should be avoided, otherwise integrated disease management including deep tillage, fungicide application and growing FHB resistant cultivars are recommended (McMullen et al. 2012).

Genetic resistance to wheat blast involves both qualitative and quantitative loci, with the former being reported under greenhouse experiments. Various major resistance genes viz. Rmg2, Rmg 3, Rmg7, Rmg 8, and RmgGR119 are found to be effective against MoT, whereas, several other resistance genes viz. Rmg1, Rmg4, Rmg5, Rmg6, and RmgTd(t) are effective against non-MoT species (Ceresini et al. 2016; Kumar et al. 2020). Several avirulence (Avr) genes that interact with the host R genes have been detected in non-MoT species, viz. PWT1 (MoO, Oryzae isolate), PWT2 (MoS, Setaria isolate), PWT3 and PWT4 (MoA, Avena isolate). Loss of such Avr genes in non-MoT isolates enables them to become virulent to wheat, just as the case of PWT3, which was likely responsible for the emergence of WB in Brazil (Inoue et al. 2017). It is important to mention that of the five MoT resistance genes, Rmg2, Rmg3, and Rmg7 have been overcome by new MoT isolates, whereas Rmg8 and RmgGR119 exhibited effective resistance in greenhouse but need to be validated in large scale field trials. New technology like Clustered regularly interspaced short palindromic repeats—CRISPR associated protein 9 (CRISPR-Cas9) can also be used in future to silence susceptibility genes for WB when identified, which has already been used in rice blast.

Apart from the Rmg genes, the 2NS/2AS translocation has been widely acknowledged as a stable and effective resistance source, although virulent isolates have emerged recently in South America (Ceresini et al. 2016). The translocation was introduced from Ae. ventricosa and has been widely utilized in wheat breeding due to its resistance against rusts (Yr17, Lr37, Sr38), nematodes (Cre5, Rkn3) and WB. Most well-known WB resistant lines have the 2NS/2AS translocation, e.g. ‘Milan’ and ‘Borlaug #100’ in the CIMMYT germplasm, ‘Sausal CIAT’, ‘CD 116’, ‘Caninde #1’ in South America, ‘BARI Gom33’ in Bangladesh, ‘HD2967’ and ‘DBW189’ in India (He et al. 2020). A recent GWAS involving 1,106 CIMMYT breeding lines identified only one stable QTL on 2NS/2AS, whereas the remaining QTL were of small effects and were detected in only some environments (Juliana et al. 2020). Similar results have been obtained in other germplasm pools too (Singh et al. unpublished data). This highlights the importance of identification and utilization of new WB resistance genes for breeding use, which could alleviate the selection pressure that is being applied to 2NS virulent isolates, to prolong the lifespan of 2NS varieties.

Conventional breeding programs of different countries have succeeded in identifying moderately resistant to resistant lines, but the bottle neck is the reliable WB screening. Early WB resistance breeding in South America depended heavily on natural infection, which was sporadic and unpredictable, with great variation in disease pressure. As for countries being threatened by WB but still do not have the disease (like India), or those have WB but do not have the screening capacity (like Zambia), the request for an international precision phenotyping platform (PPP) is very strong. In collaboration with its national partners, CIMMYT has established three WB PPPs, with one in Jashore, Bangladesh, and two in Bolivia (Quirusillas and Okinawa) to screen germplasm and advanced lines from across the globe (Singh et al. 2016). High quality phenotypic data have been produced from the three PPPs, which greatly facilitated the WB resistance breeding, germplasm screening as well as genetic studies (Juliana et al. 2020). In the early days of KB resistance breeding at CIMMYT, important genetic stocks used were ‘Aldan/IAS58’ from Brazil, ‘Shanghai-7’ from China, and native CIMMYT lines ‘Roek//Maya/Nac’, ‘Star’, ‘Vee#7/Bow’ and ‘Weaver’. To date, screening programs have resulted in the identification of numerous resistant sources for bread wheat and durum wheat from various countries (Bishnoi et al. 2020). Resistant sources have been identified in primary to tertiary gene pools of wheat, durum and triticale, including numerous dipoid, tetraploid and hexaploid species, especially T. urartu (AA) and Ae. tauschii (DD) that have high degree of resistance.

Understanding the epidemiology and population dynamics of T. indica is important for its effective management. Boot emergence to anthesis is the optimum stage for a germinated teliospore to infect, however, an infection can happen as late as at late dough stage (Carris et al. 2006). Treating seed with chlorothalonil or mixture of carboxin and thiaram and foliar spray with propiconazole, triadimefon and carbendazim are the suggested chemical control measures. The natural populations of T. indica have high genetic diversity owing to the sexual recombination occurring between heterothallic fungi generating many recombinants. This results in diversity for virulence of KB strains as well as diversity in the wheat genotypes for resistant/susceptible reaction against the disease (Kumar et al. 2021). The use of a mixture of pathogen isolates as present in the population is advocated as it increases horizontal resistance in the population targeted for breeding resistance (Bishnoi et al. 2020).

Some morphological features have been frequently associated with KB resistance, including presence of pubescence, tight glumes, flat flag leaf angle, low stomata but high hair counts on rachis, high spike compactness and/or narrow glume opening. These traits can be used in the phenotypic selection for KB resistance, but it should be noted that some of the associations might be dependent on genetic background and associated with disease escape rather than actual genetic resistance (Bishnoi et al. 2020).

Currently, it is widely accepted that genetic resistance against KB is governed by polygenes with quantitative inheritance, although gene for gene interaction may exist to some extent. Many genes with small additive effects acting in an additive and epistatic mode impart KB resistance. Since additive genes respond to selection, stacking additive genes along with an eye for significant epistatic gene interactions can enhance levels of KB resistance (Fuentes-Davila et al. 1995). In QTL mapping studies, as expected, majority of the identified QTL had minor effects and only a few major QTL have been identified on chromosomes 4B, 5B, 6B where the one on 4B associated with simple sequence repeat (SSR) marker Xgwm538 was the largest one with phenotypic variation (R2) of 25% (Singh et al. 2007). A GWAS study on 339 accessions from Afghanistan led to the identification of a consistent QTL on chromosome 2BL along with some other novel locations (Gupta et al. 2019).

Tan spot (P. tritici-repentis) is a necrotroph and follows inverse gene-for-gene relationship where recognition of host sensitivity gene by pathogen produced HST results in a compatible (susceptible) interaction, which is opposite to Flor’s classical gene-for-gene model in biotrophic diseases such as mildews and rusts. High level of resistance has been found in several wheat genotypes although immunity is not reported (Faris et al. 2013). Host resistance in wheat against tan spot can be qualitative or quantitative and major/qualitative genes responsive to tan spot are designated as ‘Tsr’, ‘Tsc’, or ‘Tsn’ corresponding to identification of genes by phenotyping assay with only fungal conidial cultures, HST containing fungal culture infiltrates inducing chlorosis and HST fungal culture infiltrate inducing necrosis, respectively. Some of the most well-characterized genes are Tsn1 (interacts with PtrToxA), Tsc2 (interacts with PtrToxB), and Tsc1 (interacts with PtrToxC) (Faris et al. 2013). Tsn1 is the only cloned tan spot resistance gene, which is located on chromosome 5BL and harbors serine/threonine protein kinase (S/TPK), nucleotide binding (NB) and leucine-rich repeat (LRR) domains (Faris et al. 2010). Dominant functional marker Xfcp623 and co-dominant markers, Xfcp394 and Xfcp620 can be used for marker-assisted selection of the resistant allele at Tsn1 locus (Faris et al. 2010). Tsc1 is located on chromosome 1A, being surrounded by markers Xhbd152, XksuM182, XksuM104, Xgwm136 in the distal side and XksuD14 in the proximal side. Tsc2 is located on chromosome 2BS and a PCR-based diagnostic marker XBE444541 is available for MAS. In addition to these three major genes, a recent meta-QTL study identified 19 QTL/loci for resistance to tan spot which can be utilized in wheat breeding programs (Liu et al. 2020).

Resistance breakdown is a major concern in R-genes conferring resistance to biotrophic pathogens as the pathogen Avr genes mutate rapidly. In case of tan spot resistance, if sensitivity genes are knocked-out of mutated, the pathogen cannot evolve as rapidly as biotrophs, so the resistance is more durable. Additionally, the fungus is saprophytic in nature and selection pressure on host would not be as high as in mildews or rusts. Molecular markers associated with major loci conferring susceptibility or resistance are very useful to select for tan spot resistant cultivars. Stacking of multiple QTL (including race non-specific) for tan spot resistance is an important and desirable strategy to manage the disease.

Managing root diseases in the modern farming system is a difficult task due to their hidden nature and when compared to leaf diseases. A variety of management strategies have been studied to control root rots (Cook 2001). Better understanding of the pathogen biology is the first step to apply the best management strategy for targeted root rot disease. Sowing healthy and high-quality seeds at the correct depth and sowing time with adequate levels of nitrogen are main agronomy practices. As these pathogens have a wide range of host crop, rotation with non-host crops may help to reduce inoculum level in the soil (Cook et al. 2002). If there is a registered fungicide, its seed treatment may support stand establishment. ‘Green bridge’ must be broken off, since the volunteer plants or weeds helps the fungi/nematode to survive during offseason, and control of “green bridge” at least four weeks before the seeding can help to reduce the multiplication of the fungus and nematodes (Cook 2001; Dababat et al. 2014).

Using resistant crops of high yield potential is the most efficient and economical way to improve the productivity of the crop and manage root rot diseases, especially in dryland areas. The advantage of using a resistant variety is not only in terms of gained crop yields, but also in reducing inoculum for the next season. Tolerant varieties are also effective in reducing the yield losses; however, they may conduce inoculum build-up/increase in the soil. Wheat and its wild relatives have been screened for resistance against the soil-borne pathogens, and several Cre genes (Cre1 to Cre9, CreX, CreY) against CCN have been identified, which are reported to follow gene for gene hypothesis. The presence of resistance in wheat progenitors has helped in synthesizing a few synthetic wheat derivatives resistant to a variety of root pathogens including CCN and RLN. International collaborative efforts, viz. distribution and utilization of CIMMYT’s International root disease resistance nurseries in the respective national breeding programs, is important to achieve desired resistance in locally adapted wheat varieties (Dababat et al. 2014).

In a nutshell, weakened plants are more vulnerable to infection by root rot fungi. Recognizing the disease by a grower is the most important point to handle the disease. Integrated disease management are likely to be effective for an extended period. Pathologists and breeders should work synergistically to identify resistant germplasm for specific pathogens and preferably sources with multiple diseases resistant.

2.6 Molecular Mapping of Resistance Genes and QTL

In classical breeding programs, selection process is based on the observable phenotypes of the candidate lines but not much knowledge about which genes are going to be selected for what trait. Whereas molecular plant breeding provides breeder with the opportunity to improve existing cultivars and to develop new cultivars using MAS approach using advanced technologies (Moose and Mumm 2008). MAS involve use of molecular markers linked to specific trait of interest in crops (He et al. 2014b).

2.6.1 Molecular Markers

Molecular markers are the specific DNA sequences present at definite locations of the genome and are transferred from one generation to the other by law of inheritance. In contrast to the morphological markers (based on visible traits) and biochemical markers (based on proteins produced by genes), molecular markers are based on DNA assay (Choudhary et al. 2008). DNA markers have been used for the characterization of various traits in wheat over the past two decades (Hoisington et al. 2002).

First plant DNA markers were based on the restriction fragment detection that includes restriction fragment length polymorphism (RFLP) (Botstein et al. 1980). RFLPs were developed about 15 years ago and were used successfully for generating linkage maps of various species but are time consuming and have limited available probes (Bernatzky and Tanksley 1986). With advances in technology, development of PCR-based markers replaced RFLP markers (Collard et al. 2005; Hoisington et al. 2002). Among them, SSR (Litt and Luty 1989; Salimath et al. 1995) were highly useful as genetic markers due to their co-dominant, highly reproducible nature and huge abundance in the genome (Deschamps et al. 2012). Other PCR-based markers included random amplified polymorphic DNA (RAPD) (Williams et al. 1990), amplified fragment length polymorphism (AFLP) (Vos et al. 1995), sequence characterized amplified region (SCAR) (Paran and Michelmore 1993), cleaved amplified polymorphic sequences (CAPS) (Konieczny and Ausubel 1993), sequence tagged site (STS) (Schachermayr et al. 1994) and direct amplification of length polymorphism (DALP) (Desmarais et al. 1998).

With further advancements in the marker technologies, diversity array technology (DArT) and single nucleotide polymorphism (SNP) are considered as the new generation molecular markers and have become main genotyping platforms. DArT is a high-throughput technique with minimum DNA sample requirement that allows the identification of hundreds of markers over the genome in one experiment without any previous DNA sequence information (Jaccoud et al. 2001). SNPs are basically the single base differences in DNA among individuals with alternate nucleotides in a same position (Vignal et al. 2002). SNP loci are present in abundance in the genome which makes it more convenient to develop genetic maps of high density, required for identifying new genes for disease resistance and other valuable traits. By using more modern DNA microarray techniques, thousands of SNPs can be analyzed simultaneously and the analysis is much more effective than any other DNA analysis (Khlestkina and Salina 2006).

Despite some challenges, this high-throughput SNP genotyping platform has been used in wheat research using the Illumina GoldenGate assay (Akhunov et al. 2009). The access to NGS and expressed sequence tags (ESTs) led to the development of new high-throughput non-gel based genotyping methodology, the KBioscience KASP assay (Allen et al. 2011). KASP assay is fast in genotyping a huge set of genotypes for both alleles in a single reaction (He et al. 2014a). Several sequencing platforms have been discovered based on continuous advancement in high-throughput genomic technologies such as development of 90 K SNP iSelect assay by Illumina.

The genome sequence information from other crops has also provided opportunities for comparative mapping. Brachypodium distachyon has replaced rice as a model species due to its high level of collinearity and synteny to various cereal genomes (Yu et al. 2009). The reference sequence of wheat is now available on public domain (Mayer et al. 2014; Appels et al. 2018), thereby serving as an important genomic tool for genetic studies. Another method to sequence specific chromosomes using flow cytometry is becoming popular in allopolyploids due to availability of all the genomic resources adding a new perspective to marker development platforms (Doležel et al. 2012; Mourad et al. 2019a; Nsabiyera et al. 2020).

2.6.2 Mapping Populations

The mapping populations are assessed for variation for the target trait and are developed by crossing resistant and susceptible lines. The size of the population ranges between 50 and 250 lines across many studies depending on the target trait (Mohan et al. 1997). The populations that are used for mapping in case of self-pollinating crop species are F2 (selfed F1 progenies), single backcross (BC; derived from crossing F1 hybrid to the recurrent parent), recombinant inbred lines (RILs; produced through selfing of filial generation F6 or higher), doubled haploid (DH; produced through doubling of F1 embryos of wheat/maize crosses), near isogenic line (NILs). F2 and BC populations are quick to produce but have high level of heterozygosity for segregating loci. In contrast, RIL and DH populations consist of series of homozygous lines representing recombination events and parental types. For mapping, generally F6 generation of RILs is considered good due to attaining high level of homozygosity. NILs are traditionally developed through backcross introgression method and can be used for validating a putative QTL where there are large genomic intervals associated with QTLs. NIL only differs from its parents in one genomic location, where there will be QTL.

2.6.3 Mapping Software

Several mapping software have been used to determine genomic locations of rust resistance genes in wheat such as Map manager QTXb20 (Manly et al. 2001), JoinMap by Kyazma B.V.software from Wageningen University (https://www.kyazma.nl/index.php/JoinMap/) (Van Ooijen 2006), MapDisto 2.0 (http://mapdisto.free.fr/) (Heffelfinger et al. 2017), QTL Cartographer v2.5 (http://statgen.ncsu.edu/qtlcart/WQTLCart.htm) (Wang et al. 2007), QTL IciMapping (http://www.isbreeding.net/) (Meng et al. 2015), MapChart (https://www.wur.nl/en/Research-Results/Research-Institutes/plantresearch/biometris/Software-Service/Download-MapChart.htm) (Voorrips 2002) and Pretzel, a tool to compare genetic and physical maps in wheat (http://plantinformatics.io) (Keeble-Gagnère et al. 2019).

2.6.4 Maps of Different Generations

Genetic linkage maps play vital role in any genomic and genetic studies and have been widely used for the identification of trait specific genetic locus. They provide exceptional framework for various studies including QTL localization, MAS and map-based cloning. The developments in the various DNA marker systems over the time has progressed construction of genetic maps in wheat. Efforts towards genetic mapping in wheat started late 1980s with RFLPs (Chao et al. 1989) but more systematic approach was followed during 1990 through coordination of ITMI. Following the development of microsatellites markers map in wheat (Röder et al. 1995, 1998), many other marker technologies have been developed and employed in molecular mapping (Liu et al. 2015) including integrated or composite maps involving more than one type of molecular markers given by (Somers et al. 2004) and International Triticeae Mapping Initiative (ITMI) maps by (Song et al. 2005). Later, the consensus maps were developed where several different maps were merged into single comprehensive such as map of synthetic W7984 (Syn) X OpataM85 doubled haploid (DH) population produced by (Sorrells et al. 2011). The wheat 9K SNP consensus genetic map based on seven mapping populations was reported by (Cavanagh et al. 2013) using 7504 SNPs. (Saintenac et al. 2013) validated available markers (9K Infinium SNP iSelect array, DArT, SSR and GBS markers) on reconstructed Synthetic X Opata DH population. The development of 90K Infinium SNP iSelect array by Illumina allowed mapping of 40,267 SNPs on combination of six hexaploid mapping populations (Wang et al. 2014b). A high density tetraploid consensus genetic map was also released using both wheat 9K and 90K Infinium arrays from 13 independent bi-parental mapping populations (Maccaferri et al. 2015).

Molecular markers have also been used in the construction of physical maps in wheat for developing a high-quality reference sequence for the wheat genome. These maps allow comparisons between genetic and physical distances of marker in their respective chromosomes. Numerous methods such as deletion mapping (Endo and Gill 1996; Sears 1954), radiation-hybrid mapping (Balcárková et al. 2017; Kalavacharla et al. 2006), in silico (Parida et al. 2006) and bacterial artificial chromosome (BAC) based physical maps have been utilized in wheat. The International Wheat Genome Sequencing Consortium (IWGSC) along with the collaborators have successfully used BAC-based sequencing for the construction of physical maps of individual chromosomes in wheat for generating high quality whole genome sequence of wheat (Appels et al. 2018). These physical maps are providing vital information for improving various traits in wheat breeding programs as well as providing a significant step forward towards cloning of genes.

2.6.5 QTL Mapping

Many genetic studies have showed that most of the important traits in cereals are inherited quantitatively which makes them difficult to detect within the genome. Now with the development of genetic linkage maps, it becomes easier to identify and characterize such quantitative trait loci (QTLs) in many species. QTL mapping is an approach for studying and dissecting quantitative traits that are of complex inheritance i.e. involving minor genes with additive effects and does not follow Mendelian inheritance (Lan et al. 2017b; Semagn et al. 2010). Various QTLs for different traits including disease resistance, grain yield, winter hardiness etc. have been described over the time (Börner et al. 2002). QTL analysis indicates the number of genetic factors involved and their effect in controlling quantitative resistance (Michelmore et al. 1991).

The primary objective of QTL analysis is to restrain quantitative trait loci to narrow down the chromosomal locations as often chromosomal QTL regions are large which may allow the transfer of other undesirable traits that are linked to the desired QTL in plant breeding. QTL mapping requires biparental mapping populations to detect association between a phenotype and a genetic marker. It involves three steps: (i) accessing the phenotypic data across various environments (ii) construction of linkage maps consisting of genetic markers and (iii) to estimate the loci effect affecting the targeted trait using statistical analysis. The linkage maps can be constructed using various platforms i.e. MapMaker (Lander et al. 1987), JoinMap (Van Ooijen 2006) or using the R package ASMap (Taylor and Butler 2017).

QTL analysis can be carried out using different statistical methods to detect associations between phenotypic data and genetic markers. Single marker analysis (SMA) (Soller et al. 1976) based on analysis of variance (ANOVA) was the first simplest method of QTL mapping. Later on, more powerful method for detecting QTL was developed based on maximum likelihood or regression known as interval mapping (IM) or single interval mapping (SIM) (Lander and Botstein 1989). Logarithm or likelihood of odd (LOD) rule was proposed by Lander and Botstein (1989) for providing confidence intervals for QTL positions. SIM is based on the single QTL model and it can be biased in the presence of multiple QTLs (Haley and Knott 1992). To overcome this problem of mapping multiple QTLs, Zeng (1994) proposed a composite interval mapping (CIM) method where SIM was combined with multiple marker regression analysis having control over QTL effects at various genomic regions independently. CIM remains a method a choice for QTL mapping since more than a decade due to its advantages over other methods for mapping single QTLs significantly but the algorithm used in this method cannot ensure epistatic effect of trait on QTL (Li et al. 2007). In order to map multiple QTLs and to be able to study epistatic effect of QTL, multiple intervals needed to be performed simultaneously (Zeng et al. 1999) and that led to the development of multiple interval mapping (MIM) (Kao and Zeng 1997). Later on inclusive composite interval mapping (ICIM) was developed by (Wang 2009) having all the benefits of CIM and MIM without having any increased sampling variance and background marker selection process (Meng et al. 2015). QTL Cartographer v2.5 (Wang et al. 2007), QTL IciMapping (Meng et al. 2015), R/qtl (Broman et al. 2003) and plabqtl (Utz and Melchinger 1996) have been the most used programs for conducting QTL mapping due to their free access.

The result of the QTL analysis is usually calculated using a test statistic score on linkage map called as logarithms of odd (LOD). LOD signifies the likelihood of the evidence for the presence of a QTL, with higher the LOD score, greater would be the evidence that QTL is real. Various computer simulations estimated the minimum LOD threshold of 3.0 to be considered significant in most cases (Lander and Botstein 1989). To further establish the significant LOD score thresholds in a given analysis, a permutation test is run by repeating the original data analysis e.g. up to 500 or 1,000 times by shuffling the phenotypic data across the genome while leaving the genetic data unchanged to assess any false marker-trait associations (Churchill and Doerge 1994).

Once the QTL is identified, it is characterized as major QTL i.e. environmentally stable or minor QTL i.e. environment sensitive on the basis of its phenotypic variation (R2). If a QTL accounts for >10% of a phenotypic variance, it is described as major QTL and if it is less than 10%, it is called as minor QTL (Collard et al. 2005). QTLs can also be described in terms of their significance to ensure that no QTL is missed and to decrease background effects (Lander and Kruglyak 1995).

Up to now, numerious QTLs have been reported for rust resistance in wheat. There have been reports for more than 300 and 200 QTLs for stripe rust and LR resistance, respectively (Wang and Chen 2017; Da Silva et al. 2018). Mendalization and detailed characterization of these QTLs is an on going process for formal naming and development of linked markers for their use in marker-assisted selection. Bansal et al. (2008) reported two major APR QTLs; QSr.Sun-5BL and QSr.Sun-7DS explaining 12% and 26% of phenotypic variation, respectively in Arina/Forno RIL population against SR along with some minor QTLs on chromosome 1AS and 7BL. Later on, a QTL found on chromosome 5BL, QSr.Sun-5BL was mendalized and mapped using Arina/Yitpi RIL population and the locus was permanently named as Sr56 (Bansal et al. 2014).

2.6.6 Marker-Assisted Breeding for Resistance Traits

Transfer and introgression of resistance genes in wheat cultivars is often limited by the practical restrictions of selection methods in conventional breeding selection programs (Bariana 2003; Bariana et al. 2007; Ellis et al. 2014). This limitation was initially overcome by a thorough knowledge of host genetics and pathogen variation, a relatively tedious process for breeding programs, also assisted by the use of morphological markers that were indicative of resistance genes and later the development of DNA markers has facilitated selection of resistance genes known as MAS. MAS provides breeders with the opportunity to select combinations of resistance genes and once the DNA is isolated, markers linked with any other trait can also be used to increase selection efficiency (Weeden et al. 1994; Ribaut and Hoisington 1998; He et al. 2014b). Several approaches are currently being used for application of DNA markers in selection procedures such as markers-assisted backcrossing (MABC) and marker-assisted gene pyramiding (MAGP).

Backcrossing method has always been used in wheat breeding to transfer rust resistance genes from a donor plant into an elite cultivar to capture recurrent parental background, but it is a slow process. MABC has accelerated transfer of rust resistance genes and resulted in rapid recovery of recurrent parental genome in as short as 2–3 backcross generations (Ribaut et al. 2002). MAGP is a process where several genes can be combined into a single genotype. The MAGP provides breeders an efficient method to select multiple traits simultaneously in their breeding programs. It has been used for pyramiding of multiple disease resistance genes and/or along with other traits in wheat. Marker linked with many rust resistance genes have been published in the last decade and are being used for MAGP and MABC of rust resistance genes in the breeding programs for pyramiding of these genes in various combinations. Several breeder-friendly markers linked to rust resistance genes are currently available to the wheat breeding programs such as Yr4 (Bansal et al. 2010), Yr15 (Ramirez-Gonzalez et al. 2015),Yr51 (Randhawa et al. 2014), Lr23 (Chhetri et al. 2017), Lr48 (Nsabiyera et al. 2016), Lr49 (Nsabiyera et al. 2020), Sr2 (Mago et al. 2011), Sr26 (Qureshi et al. 2018c; Zhang et al. 2019b), Yr34/Yr48 (Qureshi et al. 2018b), Lr34/Yr18/Sr57/Pm38 (Lagudah et al. 2009), Lr46/Yr29/Sr58/Pm39 (ES Lagudah unpublished), Lr67/Yr46/Sr55/Pm46 (Moore et al. 2015), Yr47/Lr52 (Qureshi et al. 2017a), Lr24/Sr24 (Bariana et al. 2016) and many more are routinely used in the breeding programs for pyramiding of these genes in various combinations.

Use of molecular markers in selection of rust resistance genes offers various advantages in wheat breeding programs. As compared to the conventional breeding strategies, MAS can increase selection efficiency through enrichment of positive alleles in early generations of breeding (Ribaut and Hoisington 1998; Collard et al. 2005) allowing breeders to conduct series of selections in one year. The success heavily relies on breeder-friendly markers. MAS, especially for disease resistance is independent of time, environment and plant developmental stage making it much more feasible. Some of the breeder friendly KASP markers linked with rust resistance genes in wheat are listed in Table 2.3.

Table 2.3 List of breeder friendly KASP markers linked with rust resistance genes in wheat
Full size table

But the value and use of such markers in MAS, MABC and MAGP heavily depends upon the degree of linkage between markers and the target gene. The process of validation of a marker across number of genotypes is always required to access the reliability of that marker. In some cases, even the reliable marker cannot be diagnostic because of varying level of polymorphisms in different genetic backgrounds.

2.6.7 Map-Based Cloning of Resistance Genes

The developments in the NGS technologies and the availability of sequenced and assembled genome have greatly improved marker development closely linked with the targeted genes in wheat, a plant species with large, complex and polyploid genome (Appels et al. 2018; Poland et al. 2012a). Many molecular markers linked to rust resistance genes in wheat have been developed but they must be diagnostic and proven to be efficient for their use in marker-assisted breeding. Cloning of the targeted gene(s) allows development of diagnostic molecular markers by isolating the resistance genes from the plants. Map-based cloning, also called as positional cloning is one of the traditional gene cloning methods to clone targeted genes without having any prior knowledge of the gene product. Map-based cloning works best for the targeted genes in plants where phenotypes are easily identified such as disease resistance. It requires various steps to enables us to narrow down to the shortest possible genetic interval of targeted gene (fine genetic mapping) and then to identify the candidate genes within corresponding interval on the DNA sequence (physical mapping) (Salvi and Tuberosa 2005). For map-based cloning, a high-resolution fine mapping population is a pre-requisite, which is used for phenotypic scoring and is genotyped with molecular markers developed using various genomic resources that leads to the construction of precise genetic map signifying targeted gene position (Krattinger et al. 2009a; Bettgenhaeuser and Krattinger 2019).

Positional cloning requires large-insert genomic DNA libraries and the vectors and cloning systems associated with the construction of these libraries are improving with time. The yeast artificial chromosome (YAC) cloning system was the first one to be developed by (Burke et al. 1987) to clone larger DNA fragments of up to 1 Mb. But there are several disadvantages of YAC cloning system such as high level of chimerism (~40% of whole library) and instability in the yeast host strain that limited its use (Shi et al. 2011; Umehara et al. 1995). In order to overcome these disadvantages, various bacterial mediated cloning systems such as BAC (Shizuya et al. 1992), transformation-competent artificial chromosome (TAC) (Liu et al. 1999) and P1-derived artificial chromosome (PAC) (Loannou et al. 1994) have been developed. BAC vectors were developed to clone DNA sequences in bacterial cells. The BAC system has been widely used and being instrumental in constructing genomic libraries in plants due to its several advantages such as stability with foreign DNA, cloning inserts of up to 300 kb, relatively easier to purify the plasmid vector and insert DNA from the bacterial host DNA and having higher cloning efficiency (Monaco and Shizuya et al. 1992; Monaco and Larin 1994; Salimath and Bhattacharyya 1999; Ming et al. 2001). In wheat, due to the presence of three highly related homoeologous genomes, chromosome specific BAC library strategy has been successful in sequencing individual chromosomes (Šafář et al. 2004; Paux et al. 2008; Appels et al. 2018).

Once the genomic libraries have been constructed using these vectors, they can be used for chromosome walking or landing approaches. Chromosome walking strategy relies on identifying tightly linked markers to the targeted gene and then taking walking steps (~100–200 kb at a time) to get to the gene via a series of overlapping clones (Han and Korban 2010). These closet flanking markers are used to screen the BAC library and subclones of the identified BACs for positive clones which are then used to isolate insert-ends (Periyannan 2018). These insert-ends are used for screening additional overlapping clones until the contig spanning the target gene is established and the candidate gene is identified (Krattinger et al. 2009a). On the other hand, chromosome landing approach relies mainly on development of molecular markers that are either tightly linked or co-segregating with the targeted gene. In chromosome landing, the distance between the markers and the gene has to be smaller than the average insert length of a genomic library used for gene isolation (Tanksley et al. 1995). These markers are then used to screen the library and isolate the clone carrying the targeted gene (Han and Korban 2010). For successful cloning approaches, genetic complementation of the mutant phenotype with a wild type allele is required. Agrobacterium-mediated genetic transformation and biolistic transformation are the two widely used methods to introduce foreign targeted genes into plant cells (Xia et al. 2012).

2.7 Enabling Genomic Tools in Wheat Breeding

2.7.1 Association Mapping Studies

The main objective of genetic mapping is to identify markers in close proximity of genetic factors affecting quantitative traits usually governed by QTL. Genetic mapping can be performed in two ways: (a) developing bi-parental populations, for “QTL-mapping” or “gene tagging” and (b) using diverse panel of lines called “genome-wide association mapping studies,” or “association mapping (AM)” or “linkage disequilibrium (LD) mapping”. The traditional QTL mapping approach is quite widely used however, it also has some limitations. First, allelic variation in each cross is limited to just two parents used to generate a QTL mapping population. Second, the number of recombination events per chromosome are small when segregating or double haploid populations are used. Third, a typical QTL detected in a specific cross of few hundred inbred lines can range between a few to tens of centimorgan (cM) interval covering several million basepairs. Such large genome regions contain, typically, hundreds to thousands of genes, making gene identification in a QTL region a tedious exercise through map-based cloning (Price et al. 2006).

Association mapping has emerged as a powerful tool in determining the genetic basis of complex traits where large populations are analyzed to determine marker-trait associations using linkage disequilibrium. This approach has advantages over traditional QTL mapping. Firstly, a larger and more representative gene-pool can be examined. Second, overcomes the cost and time of developing mapping populations and facilitates mapping of several traits on one panel of genotypes. Third, a much finer resolution can be achieved, resulting in shorter confidence intervals of the mapped loci compared to conventional mapping, which also necessitates fine-mapping to develop diagnostic markers. Finally, in addition to identifying and mapping QTL, it helps to identify causal polymorphism within a gene that is responsible for the difference in two alternative phenotypes (Yu et al. 2013). However, AM also has challenges of false positives, especially if the experimental design and quality control is not rigorously implemented. For example, population structure has long been known to induce many false positives and accounting for population structure has become one of the main issues when implementing AM in plants (Breseghello et al. 2005). Also, with an increasing number of genetic markers independent validation of identified associations helps in discriminating false positives. With these limitations, AM still shows great promise in understanding the genetic basis of polygenic traits of agronomic importance.

To increase the power and mapping resolution of marker-trait associations, some specialized populations have been developed using a combination of both QTL mapping and AM. For example, NAM (Nested Association Mapping) populations and MAGIC (Multiparent Advanced Generation Inter Cross) populations have been developed in wheat and other crops (Kover et al. 2009; McMullen et al. 2009; Huang et al. 2012; Cavanagh et al. 2013). NAM populations are generated by crossing a set of diverse lines (5–25) to one reference line. F1’s of each cross-are then selfed for multiple generations to develop RIL for each population. MAGIC populations on the other hand are developed by intercrossing for several generations among multiple founder (4–8) lines. Multiple founders are similar to NAM population, which enable capturing more allelic diversity than bi-parental populations, and repeated cycles of intercrossing give greater opportunity of recombination and hence greater precision of QTL mapping. However, generating such specialized populations entails effort, time and investment.

2.7.2 Genotyping/Marker Platforms for Genome-Wide Studies

Most commonly used markers in genome wide association studies include AFLP, DArT, SSR and SNP (Crossa et al. 2007; Honsdorf et al. 2010; Adhikari et al. 2012; Upadhyaya et al. 2013; Gupta et al. 2014). AFLP and DArT markers are easily accessible for all organisms even those lacking genomic data. Similarly, the highly polymorphic, multiallelic and co-dominant nature of SSR markers have made them highly suitable for AM studies in many crops including wheat (Peng et al. 2009; Yao et al. 2009; Liu et al. 2010; Reif et al. 2011; Zhang et al. 2011). However, AFLP and DArT markers being dominant can be challenging especially while estimating population structure or during mapping studies (Ritland 2005). Moreover, the three marker platforms (AFLP, DArT and SSR) are rather expensive and time- consuming technologies and the genomic coverage is also limited.

The rapid development of NGS technologies has allowed unprecedented genotyping capabilities, even for large complex polyploid genomes including wheat (Poland et al. 2012b). The current NGS technologies are capable of analyzing tens of millions of DNA molecules and allow the rapid identification of a large numbers of genetic markers, mainly SNPs (Single Nucleotide Polymorphisms) (Imelfort et al. 2009). SNPs are bi-allelic markers that’s whythe information content per marker is much lower than SSR markers. This, however, is compensated for by a higher genome coverage. Therefore, SNP markers rapidly becoming the marker of choice for most AM studies. SNP markers are also amenable to high-throughput genotyping enabling options of multiplexing or microarray. Several SNP marker platforms have been established in wheat (Akhunov et al. 2009; Wang et al. 2014b) and genotyping of wheat association panels with up to 90,000 SNP markers is now available (Wang et al. 2014b). The potential of SNP markers in determining marker-trait associations is now being widely used across crops including wheat (Lopes et al. 2015).

With further developments in NGS technologies, sequencing today has extended to entire populations enabling simultaneous genome-wide detection (Elshire et al. 2011). This new approach, called “genotyping-by-sequencing” (GBS), uses data from the genotyped populations, thereby removing bias towards a particular population. GBS is a cost-effective technology producing up to a million SNPs per genotype at a low cost. However, one of the challenges associated with GBS is inadequate genome coverage and incomplete datasets (Fu 2014), sometimes with up to 90% missing observations per line (Elshire et al. 2011; Fu and Peterson 2011). Such data cannot be used for AM and filtration should be done to improve the sequence data (Fu 2014). Several methods for imputation include regression-based methods such as random forest (Stekhoven and Bühlmann 2012) and principal component analysis (PCA)-based tools (Stacklies et al. 2007).

2.7.3 Confounding Effects of Population Structure

One of the challenges in using AM to dissect the genetic architecture of complex traits is the risk of detecting false positives due to population structure (Pritchard et al. 2000). The problem of population structure can arise due to the correlation of phenotypic trait with population structure at neutral loci, which can result in an inflated number of false positive associations resulting in Type I errors. Among several methods used to deal with this problem, the ‘genomic control’ (GC) method could be considered useful (Devlin and Roeder 1999). GC estimates association using large number of putative neutral markers or markers that are not thought to be associated with the trait of interest. The distribution of the test statistic is then calculated from these associations for trait of interest and a critical value for desired Type I error rate is chosen from this distribution. Another commonly used method is called structured associations (SA) (Pritchard et al. 2000). SA first queries population for closely associated clusters/subdivisions using a Bayesian approach, and then uses clustering matrices (Q) in AM (by a logistic regression) to correct for false associations. Population structure and shared co-ancestry coefficients between individuals of subdivisions of a population can be effectively estimated with the STRUCTURE program (Pritchard et al. 2000) using several models for linked and unlinked markers.

Principal component analysis (PCA) is widely used as a faster and effective way to diagnose population structure (Chengsong and Jianming 2009). The PCA method makes it computationally feasible to handle a large data sets (tens of thousands of markers) and correct for population stratification. Most widely used programs to calculate PCA are DARwin and EIGENSTRAT (Price et al. 2006).

A mixed linear model (MLM) combining both population structure information (Q-matrix or PCA) and pairwise relatedness coefficients (kinship-matrix) can be used in the analysis. While the Q-matrix explains the structure between groups in a population, the kinship-matrix explains the structure within group. Although MLM approach is computationally intensive, it is very effective in removing the confounding effects of the population in AM (Yu et al. 2006). However, in some cases using MLM + kinship model may result in over correction of the population structure. This could be identified from the QQ-plot when it skews below the reference line. In this case, using generalized linear model and population structure (GLM + PC) will be better in removing the population structure and identifying the markers significantly associated with the studied traits. Similar cases were found in studying disease resistance in wheat and barley indicating the importance of testing both MLM + K and GLM + PC models (Turuspekov et al. 2016; Abou-Zeid and Mourad 2021).

2.7.4 Estimates of LD

Linkage Disequilibrium (LD) refers to the correlation between alleles in a population (Flint-Garcia et al. 2003) and Linkage refers to the correlated inheritance of loci through the physical association on a chromosome but not necessarily on the same chromosome. For AM, it is important to understand the patterns of LD for genomic regions of individual plants and the extent of LD among different populations or groups to design unbiased association mapping studies. Two most widely used statistics to measure LD are r2 (square of the correlation coefficient) and D′ (disequilibrium coefficient). The r2and D′ statistics represent different aspects of LD and perform differently under various conditions. The r2 is affected by both mutation and recombination while D′ is affected by more mutational events of the past.

There are several software programs such as GOLD (Abecasis and Cookson 2000), TASSEL (www.maizegenetics.net) or Powermarker (Liu and Muse 2005) to represent the structure and pattern of LD. Average genome-wide decay of LD can be estimated by plotting LD values (r2 values) obtained from a data set adequately covering an entire genome against the genetic (or) physical distance between markers. The decrease in LD within the genetic distance indicates the portion of LD that is conserved with linkage and proportional to recombination events (Gupta et al. 2014). The decay of LD over physical/genetic distance in a population is a determinant of marker density and coverage needed to perform an association analysis. If rapid LD decay is observed, then a higher marker density is needed to capture markers closely linked to functional sites (Flint-Garcia et al. 2003; Gaut and Long 2003). In wheat, depending on the populations used in study, LD decay have been reported to vary from 0.5 to 40 cM (Chao et al. 2007; Crossa et al. 2007; Somers et al. 2007; Tommasini et al. 2007; Yao et al. 2009; Dreisigacker et al. 2012). The higher distance of LD decay was found in wheat genome D followed by genome A and B, respectively which indicated that lower number of markers are required to identify targeted QTLs in genome D compared with genome A, and B (Liu et al. 2017; Ayana et al. 2018; Mourad et al. 2020).

2.7.5 Association Analysis Programs

GWAS is a very helpful method in detecting QTLs responsible for different traits (Alqudah et al. 2020). It has been used widely in wheat breeding for disease resistance and helped breeders in identifying genes controlling resistance to different races. There are many softwares which could be used in GWAS analysis. Publicly available software using mixed models for AM studies in plants include TASSEL and EMMA/R. Both can analyze moderate to large datasets but only allow single effects (samples or taxa) to be fit as a random effect and all other effects treated as fixed. EMMA relies on the R for data management and visualization which is not limited with TASSEL functions. Other commercial software packages for AM studies include ASREML, JMP Genomics, ASREML, SAS and GenStat. General software such as SAS Proc Mixed and GenStat can perform AM analysis requiring more expertise and programming by the user and JMP Genomics are suited specifically for genetic analysis and can handle models that are more complex.

TASSEL on the other hand uses both GUI (graphical user interface) and CLI (command line interface) versions for detailed analysis and use versions depending on their expertise and consistent results can be obtained independent of the interface. In the latest version of TASSEL (TASSEL 5.0), a compressed MLM method has been developed to compute large datasets. GAPIT-R package is also a very useful software in AM and GWAS studies. It could be applied using different methods such as GLM, MLM and Settlement of MLM Under Progressively Exclusive Relationship (SUPER) (Wang et al. 2014a). For disease resistance, SUPER has been reported as a very useful method in detecting the QTLs significantly associated with the resistance of the targeted disease as it conducts GWAS by extracting a small subset of markers and testing their association with resistance by using Fast-LMM method. This method enables the identification of minor genes controlling the resistance (Mourad et al. 2018a).

2.7.6 Significance Threshold

Significance threshold is set to declare associations as significant in a particular study. Either FDR (false discovery rate) or ‘Bonferroni’ correction can be used to correct for multiple comparisons. The correction factor is needed to test multiple hypotheses simultaneously. FDR controls the proportion of false positives among significant results by defining a threshold from the observed p-value distribution in the data, whereas Bonferroni corrections detect and control false positives (Benjamini and Hochberg 1995). Given the objective of the study, one may consider a high FDR (e.g. dissection of genetic architecture of a trait) or low FDR (e.g. identifying candidate loci for further characterization and validation).

2.7.7 Validation of Association Results

Validation of AM results is an important step before marker information is used for selection decisions, or before identifying causal factors and gene cloning. One way is to compare the AM results with previously published results for the trait; for example, in bi-parental populations, markers in close proximity (<10 cM) to previously reported QTLs/genes, will not only increase the confidence but validate the new genomic target identified for the trait. Secondly, validation can be performed different panels/populations. This is more reliable as the probability of significant associations are confirmed and false positives discarded when validation is carried in two or more populations. Third, AM results point to alleles with opposite effects (favorable/unfavorable alleles) on a trait of interest, multiple F2 populations can be generated from parents carrying contrasting alleles and determine whether phenotype differences co-segregate with the locus of interest.

In addition, testing the LD between the newly identified markers and previously identified markers will give more power to the results obtained from the association tests. For example, testing the LD between the identified SNP markers located on the same chromosome will give an idea if they are controlling the same QTL or different QTLs. If the studied trait or disease has a published accurate SSR marker, the LD between the identified SNPs and the SSR marker will validate the association between the SNP markers and the studied trait (Mourad et al. 2018b, 2019b). Furthermore, gene models harboring the detected QTLs and their functional annotations could be investigated using IWGSC dataset. If the identified marker is located within or near gene model and annotated to improve the targeted trait, this will give more power to the association results. Once tightly linked markers to the target trait are validated, can enhance the speed and cost efficiency of selection in breeding programs.

2.7.8 Genomic Selection in Breeding for Quantitative Disease Resistance in Wheat

With rapid changes in pathogen races and breakdown of major resistance genes, the benefits of marker-assisted selection in selecting for minor gene based quantitative disease resistance in wheat is limited and hence, the focus has shifted to GS (Rutkoski et al. 2011; Poland and Rutkoski 2016). In GS, dense genome-wide markers and trait phenotypes (disease response in this case) are used to obtain the genomic-estimated breeding values (GEBVs) of individuals for the trait, from which selections are made (Heffner et al. 2009; Meuwissen et al. 2001). Since GS models incorporate all the marker information across the genome to estimate marker effects, they are expected to capture well the total additive genetic variance, including the disease variation resulting from minor effect quantitative trait loci (Heffner et al. 2009; Poland and Rutkoski 2016). The potential of GS for disease resistance in wheat has been explored in several studies that have reported different prediction accuracies (PAs, correlations between the predicted and the true breeding values), some of which are discussed below.

The utility of GS for increasing the gains from selection per unit time (Heffner et al. 2010) was first explored for quantitative APR to SR in wheat, where the authors presented a recurrent GS-based breeding scheme including rounds of intermating and GEBV-based selections, with simultaneous evaluation of lines and model updating (Rutkoski et al. 2011). In 2012, a study by Ornella et al. evaluated genomic predictions for stem and YR in five bi-parental populations from CIMMYT and observed maximum within-year CV PAs ranging from 0.56 to 0.75 for SR and from 0.63 for YR in the different populations (Ornella et al. 2012). Genomic prediction for rust resistance was also evaluated in a set of landraces from the Watkins collection and five-fold CV PAs of 0.35, 0.27 and 0.44 were obtained for LR, SR and YR, respectively (Daetwyler et al. 2014).

One of the first studies on comparison of realized gains from GS and phenotypic selection for quantitative SR resistance in wheat, indicated that while both lead to equal rates of gain in the short-term, GS lead to a significantly greater loss in genetic variance compared to phenotypic selection that could reduce the rates of genetic gain in the long-term (Rutkoski et al. 2015). A comparison between genomic prediction and pedigree-based prediction for rust resistance using CIMMYT’s international bread wheat screening nurseries indicated similar accuracies with both the relationship matrices, and the CV genomic PAs ranged between 0.31 and 0.74 for LR seedling resistance, 0.12 and 0.56 for LR APR, 0.31 and 0.65 for SR APR, 0.70 and 0.78 for YR seedling resistance, and 0.34 and 0.71 for YR APR (Juliana et al. 2017a).

Genomic prediction models for FHB resistance in wheat were evaluated using the U.S. cooperative FHB wheat nurseries, and resulted in five-fold within-year cross-validation (CV) PAs of 0.46 and 0.41 for Fusarium damaged kernels (FDK) and deoxynivalenol (DON) content, respectively (Rutkoski et al. 2012). Using a large panel of Central European elite winter wheat lines, Mirdita et al. (2015), reported high five-fold CV PAs of 0.6 for FHB resistance and 0.5 for STB resistance. In another FHB genomic prediction study using breeding lines, high five-fold CV PAs of 0.82 for FDK and 0.64 for DON content were reported (Arruda et al. 2015). A breeding population of spring wheat lines was used to evaluate genomic prediction for FHB and the ten-fold CV PAs for FHB incidence, severity, and DON content were 0.63, 0.43, and 0.42, respectively (Dong et al. 2018). In another study, using elite spring wheat breeding lines from six breeding cycles, Liu et al. (2019) reported genomic PAs ranging from 0.22 to 0.44 for FHB resistance. Genomic prediction for FHB and STB in winter wheat lines indicated PAs of 0.72 and 0.15 for the two traits, respectively (Herter et al. 2019).

Evaluation of genomic prediction for STB, Stagonospora nodorum blotch and tan spot in wheat using CIMMYT’s international bread wheat screening nurseries indicated that the mean CV PAs for STB APR, Stagonospora nodorum blotch seedling resistance, tan spot seedling resistance and tan spot APR were 0.45, 0.55, 0.66 and 0.48, respectively (Juliana et al. 2017b). Comparison of the two whole-genome profiling approaches: genotyping-by-sequencing and diversity arrays technology-sequencing identified that the genotyping-by-sequencing markers performed slightly better than diversity arrays technology sequencing markers and combining markers from the two platforms did not improve the PAs. Another study for the genomic prediction of STB response using European winter wheat varieties, reported a mean five-fold CV PA of 0.44 (Muqaddasi et al. 2019). In a large-scale study involving genomic prediction for several traits, Juliana et al. (2019) reported moderate to high within-panel mean CV PAs of 0.49, 0.5, 0.64 and 0.56 for field resistance to STB, spot blotch, SR and stripe rust. However, when one nursery was predicted from all other panels, they obtained low mean genomic PAs of 0.28, 0.36 and 0.36 for STB, spot blotch and stripe rust, respectively, but a high genomic PA of 0.61 for SR.

Given the promising PAs obtained from genomic prediction for disease resistance in wheat in most of the aforementioned studies, breeding programs can effectively integrate GS in breeding for resistant varieties. It can be especially useful for traits that have a low heritability and are difficult to phenotype. In the case of disease resistance traits that have a high heritability, where phenotypic selection might be the best method to increase genetic gains, GS can be applied to increase the selection intensity in early generations (Poland and Rutkoski 2016). Overall, further research on developing GS-based breeding strategies for wheat disease resistance and using it in combination with other strategies like rapid generation advancement technologies, high-throughput phenotyping and gene editing are important (Voss-Fels et al. 2019).

2.8 Integrating New Tools for Resistance Breeding Presents Opportunities for Wheat Improvement

The proven approach to enhance durability of genetic resistance is the deployment of combinations of multiple effective resistance genes often termed as “pyramiding”. A limitation to stack multiple genes is their segregation when parents possessing different genes are crossed. This requires growing large populations to identify multiple gene combinations and the need to have complementing diagnostic markers tagging the R-genes for ensuring that the desired gene combination achieved. However, incomplete/moderate effect R-genes, race-nonspecific APR genes, or their combinations confers enhanced resistance levels due to additive effects, hence have been shown to be effectively selected in the field under high disease pressures (Singh et al. 2008b, 2015, 2016). New research advances have also facilitated options for combining multiple resistance genes in a single line/variety thereby enhancing resistance durability.

In the last two decades several rust resistance genes have been cloned using various approaches (Table 2.4) viz. eleven SR resistance genes: Sr13 (Zhang et al. 2017a), Sr21 (Marchal et al. 2018), Sr22 (Steuernagel et al. 2016), Sr33 (Periyannan et al. 2013), Sr35 (Saintenac et al. 2013), Sr45 (Steuernagel et al. 2016), Sr46 (Arora et al. 2019), Sr50 (Mago et al. 2015), Sr55 (pleiotropic with Lr67) (Moore et al. 2015), Sr57 (pleiotropic with Lr34) (Krattinger et al. 2009b) and more recently Sr60 (Chen et al. 2020); four LR resistance genes Lr1 (Cloutier et al. 2007), Lr10 (Feuillet et al. 2003), Lr21 (Huang et al. 2003), Lr22a (Thind et al. 2017) and six YR resistance genes Yr5 (Marchal et al. 2018), Yr7 (Marchal et al. 2018), Yr10 (Liu et al. 2014), Yr15 (Klymiuk et al. 2018), YrAS2388R (Zhang et al. 2019a) and Yr36 (Fu et al. 2009). In the last decade, R-gene enrichment sequencing (Ren-Seq) approaches have been widely used to clone resistance genes. Resistance genes from wild relatives can be introgressed to engineer broad-spectrum resistance in domesticated crop species using a combination of association genetics with R-gene enrichment sequencing (AgRenSeq) to exploit pan-genome variation in wild diploid wheat and such approach enabled rapid cloning off our SR resistance genes (Arora et al. 2019) and a relatively new approach called MutRenSeq that combines chemical mutagenesis with exome capture and sequencing has been developed for rapid R-gene cloning and enabled successful cloning of Sr22 and Sr45 from hexaploid bread wheat (Steuernagel et al. 2017). Despite these advances, the availability of the currently effective cloned genes remains limited, therefore requiring a responsible strategy for their deployment.

Table 2.4 Cloned rust resistance genes in wheat
Full size table

The availability of multiple cloned resistance genes opens the possibility to transform wheat lines with a stack or cassette of multiple cloned effective resistance genes. This transgenic approach can help combine multiple resistance genes in a linkage block with one another on a single translocation thereby reducing the chances of segregation upon further breeding processes and up to eleven cloned genes can be stacked (Wulff and Moscou 2014). However, the current regulatory framework in most countries does not allow the cultivation of transgenic, including cisgenic wheat and if future policy decisions favor approval of transgenic-cassettes such approach can be utilized to enhance durable resistance in wheat varieties.

2.9 Gene Editing

R-gene mediated resistance is race-specific though it remains effective in protecting the plant throughout all growth stages. Additionally, mutation in the pathogen avr-locus can also lead to break down of resistance. On the other hand, the APR loci show only partial resistance in adult pants while allowing considerable disease development (Ellis et al. 2014). Plant pathogens exploit host genes and machinery such as sugar transporters to draw nutrient from the host plant or to replicate their genome as in case of viruses. Mutations in some of these genes (susceptibility factors) has no impact on the plantgrowth and phenotype but can restrict pathogen growth (Moore et al. 2015).

Despite the tremendous success, marker-assisted breeding can exhaustively take from 7 to 12 years to introduce a new trait and release an improved variety (Acquaah 2007). Considering the recurrent resistance development against pesticides by several pathogens, emergence of new virulent strains or races, and lack of resistant germplasm against some pathogens, the current breeding methods are unlikely to keep pace with the predicted demand for rapid development of improved disease resistant varieties (Scheben et al. 2017). Redundancy of susceptibility factors due to polyploidy in wheat makes it difficult to identify lines that have all copies of these genes mutated. Breeders have been using chemical mutagenesis, gamma irradiation, fast neutron bombardment, and T-DNA insertion to generate artificial mutants. But low frequency, random and undirected nature of these mutations has impeded its utility in wheat breeding for disease resistance. Site-specific nuclease mediated editing of target genes offers an excellent alternative to precisely mutate a target gene without disturbing rest of the genome.

Until 2013, the dominant genome editing tools were zinc finger nucleases and transcription activator-like effector nucleases (TALENs) and they have been used successfully in many organisms including plants such as wheat (Wang et al. 2014c). The design of ZFNs is challenging due to the complex nature of the interaction between zinc-fingers and DNA as well as limitations imposed by context-dependent specificity. Though the design of TALEN is relatively simpler, the highly repetitive sequences in the construct promote homologous recombination in vivo making it difficult for wide adoption. The RNA-guided genome editing (RGE) using the CRISPR-Cas9 technology has emerged as a simple, versatile and highly efficient tool for editing of target genes in a wide variety of organisms including plants. The CRISPR-Cas system relies on simple Watson–Crick base pairing of a chimeric single guide RNA (sgRNA) that is partly complementary to the target DNA sequence called proto-spacer element. Thus, only 20 nucleotides (nt) in the gRNA need to be modified to recognize a different target. The target sequence must be followed by a protospacer adjacent motif (PAM) such as NGG or NAG (N: any nucleotide) for target recognition (Biswal et al. 2019). Once the target site is recognized, the Cas9 nuclease makes a double stranded break (DSB), three nucleotides upstream of the PAM (Yin et al. 2017). The DSB is immediately repaired by the host cell by the non-homologous end joining (NHEJ), which is erroneous. The NHEJ may insert, remove or even substitute one or a few nucleotides. Removal or insertion of non-triplets can lead to frameshift mutation that can either result in a complete new protein or may introduce a stop codon downstream of the target site resulting in a truncated protein that can be target of subsequent nonsense-mediated decay of the transcript (Shaul 2015). Multiplexed targeting of two different regions of the genome can result in removal of a chunk of DNA flanked by both targets. The host cell can also follow a homology-directed repair (HDR) mechanism, when a suitable template is supplied with homologous flanking arms. HDR method can be applied to replace a faulty gene, to introduce desired mutations or even to introduce a new coding sequence or promoter. Though the efficiency of HDR method is relatively low, HDR enhancers have also been reported (Song et al. 2016). The CRISPR-Cas based prime editors can also be used to replace, delete or introduce a fragment of DNA at the target site without the special supply of HDR template. The prime editing complex uses a Cas9 nickase (nCas9) fused to a reverse transcriptase and a prime editing guide RNA (pegRNA) that specifies the target site as well as encodes the desired edit without introducing DSBs (Anzalone et al. 2019).

Base editing is a newer genome-editing approach that uses a catalytically inactive Cas9 (dCas9) fused to a nucleotide deaminase enzyme (Gaudelli et al. 2017). Base editors can convert C·G to T·A or A·T to G·C in cellular DNA or RNA without making double-stranded DNA breaks. As SNP is one of the most common form of difference observed between resistant and susceptible alleles, base editors can be of immense importance to wheat molecular breeders to directly improve the disease resistance in elite wheat lines. It can also be extended to generate artificial alleles of a target gene.

The CRISPR-Cas mediated genome editing is precise, and modifications can be done at specific locus. More importantly the meiotic segregation of the CRSPR tools leaves final product free of transgenic traces that can be released to farmers with minimum regulatory inhibition. Wheat powdery mildew an important wheat disease, caused by an obligate biotrophic ascomycete fungus Blumeria graminis f. sp. tritici, Bgt, which has a highly selective host range of single-plant genera (Singh et al. 2016). Simultaneous editing of all three homoeoalleles of the MLO locus in hexaploid bread wheat using TALEN and CRISPR-Cas9 has shown heritable resistance to powdery mildew (Wang et al. 2014c). Similarly, simultaneous modification of all homoeologs of TaEDR1 gene by CRISPR-Cas9 technology generated wheat lines with enhanced resistance to powdery mildew (Zhang et al. 2017b). FHB is another important disease of wheat caused by Fusarium graminearum fungus. Deoxynivalenol (DON) is a mycotoxin virulence factor that induces the expression of a transcription factor TaNFXL1, a repressor of F. graminearum resistance. CRISPR-Cas9 mediated knocking out of TaNFXL1 demonstrated increased FHB resistance (Brauer et al. 2020).

The CRISPR/Cas system to dissect the pathogen genetics and to diagnose the disease. The lack of genetic tools to analyze and link the pathogen genes to the disease phenotype and progression is a major impediment in developing disease resistant crops. Recent progress in next generation sequencing has helped to predict functions hundreds of pathogen genes that needs functional validation (Levy et al. 2018). The CRISPR-interference (CRISPRi) system uses a catalytically inactive Cas9 protein (dCas9) and programmable single guide RNAs to modulate the pathogen gene expression that can be employed to dissect the functions of essential and non-essential genes in different pathogen species. Recently, a ‘Mobile-CRISPRi’ system has been developed to analyze antibiotic resistances and host-microbe interactions that uses a modular system and can be transferred to diverse bacterial species by conjugation (Peters et al. 2019). A similar system in plants can be very useful to study plant-pathogen interaction as well as the mechanism of resistance breakdown by new pathotypes.

Rapid and reliable detection of pathogens is important for taking curative measures in order to minimize the crop loss. The CRISPR-Cas technology can provide a versatile tool to detect the pathogen DNA/RNA in plant samples. The SHERLOCK (specific high-sensitivity enzymatic reporter unlocking) system can detect atto-molar concentration of nucleic acids in a solution (Abudayyeh et al. 2019; Gootenberg et al. 2018). The colorimetric lateral flow strips can also be developed to detect certain plant or pathogen genes by using SHERLOCK platform in the field without need of high end instruments or technical expertise (Abudayyeh et al. 2019). The multiplexed SHERLOCK system can also be extended to detect heterozygosity and trait stacking.

2.10 Concluding Remarks

Wheat diseases continue to be a significant challenge in several wheat production environments. Major threat is due to the extreme damage these diseases can cause to susceptible varieties. Although severe epidemics have not been reported in the last two decades, lack of genetic diversity in host and constantly evolving and migrating pathogens can pose a significant risk. Genetic resistance through deployment of both race specific genes and APR though quite widely used in breeding programs, however, faster evolution of new races to overcome race specific genes has resulted in wide spread vulnerability of cultivars, and the increasing importance of some diseases due to changes in cropping systems and crop intensification require reinforcing breeding strategies to develop adequate and durable resistance to multiple diseases for enhancing wheat productivity and farmers’ income worldwide by reducing crop losses. New genomic tools in conjunction with phenotypic selection provides great promise for harnessing ample genetic diversity for resistance that exists in wheat for a number of important diseases. The impact of cost-effective NGS technologies coupled with new tools of rapidly cloning of rust resistance genes alongside the availability of wheat reference genomes can rapidly accelerate pyramiding strategies into desired wheat backgrounds. Progress in genetic mapping techniques and wheat transformation methods can enhance cloning efforts with the possibility of stacking multiple genes or gene cassettes using functional markers Future policy decisions will determine whether transgenic cassettes can be utilized as a new strategy for durable resistance in various countries.