Can effectoromics and loss-of-susceptibility be exploited for improving Fusarium head blight resistance in wheat?

Andrii Gorsh*, Rit Armonien? Keml Kzn

aInstitute of Agriculture,Lithuanian Research Centre for Agriculture and Forestry(LAMMC),Instituto av.1, Akademija LT-58344,Lithuania

bCommonwealth Scientific and Industrial Research Organisation (CSIRO)Agriculture and Food,Brisbane,QLD 4067,Australia

cQueensland Alliance for Agriculture and Food Innovation(QAAFI),The University of Queensland, Brisbane,QLD 4067,Australia

Keywords:Effectoromics Susceptibility genes Fusarium head blight Fusarium graminearum Wheat breeding for resistance

ABSTRACT Bread wheat(Triticum aestivum L.),which provides about 20%of daily calorie intake,is the most widely cultivated crop in the world,in terms of total area devoted to its cultivation.Therefore,even small increases in wheat yield can translate into large gains. Reducing the gap between actual and potential grain yield in wheat is a crucial task to feed the increasing world population.Fusarium head blight(FHB)caused by the pathogenic fungus Fusarium graminearum and related Fusarium species is one of the most devastating wheat diseases throughout the world. This disease reduces not only the yield but also the quality by contaminating the grain with mycotoxins harmful for humans, animals and the environment. In recent years, remarkable achievements attained in “omics” technologies have not only provided new insights into understanding of processes involved in pathogenesis but also helped develop effective new tools for practical plant breeding.Sequencing of the genomes of various wheat pathogens,including F.graminearum,as well as those of bread and durum wheat and their wild relatives,together with advances made in transcriptomics and bioinformatics, has allowed the identification of candidate pathogen effectors and corresponding host resistance(R)and susceptibility(S)genes.However,so far,FHB effectors and wheat susceptibility genes/factors have been poorly studied.In this paper, we first briefly highlighted recent examples of improving resistance against pathogens via new techniques in different host species. We then propose effective strategies towards developing wheat cultivars with improved resistance to FHB.We hope that the article will spur discussions and interest among researchers about novel approaches with great potential for improving wheat against FHB.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2

1.1. A brief history of FHB research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2

1.2. What have we learned? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2. Lifestyle of the FHB pathogen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

3. Effector-assisted selection of resistant germplasm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

3.1. Can effectoromics be used for breeding against FHB in wheat? . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

4. Loss-of-function of S genes as a novel strategy for resistance breeding . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

5. Molecular techniques used for down-regulating host S genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

6. Concluding remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

CRediT authorship contribution statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

1.Introduction

Fusarium head blight (FHB), caused mostly by fungal pathogens Fusarium graminearum and F.culmorum,is one of the most destructive wheat diseases in the world. The United States Department of Agriculture (USDA) rates FHB as the second most devastating wheat disease only after stem rust and its destructive epidemics in 1950s [1]. Fusarium graminearum can also infect other cereals such as barley and maize.Symptoms of FHB in wheat include necrosis and bleaching of infected spikelets or the entire heads with shrivelled or shrunken pinkish or chalky white “tombstone” kernels, reduction of yield and contamination of grains with harmful mycotoxins.F. graminearum is known to have both sexual (Gibberella zeae)and asexual stages. Life cycles of about 20% of Fusarium species consist of both sexual and asexual stages, while only asexual stage is known for F. culmorum and most other Fusarium species [2]. Whereas, the presence of G. zeae mating gene homologs in F. culmorum may indicate that the sexual stage in this fungus is either undetected or has been lost recently [3,4]. Asexual spores (macroconidia) and sexual spores(ascospores)are major inoculum sources for infection.Macroconidia are dispersed at short distances by rain splash,whereas airborne ascospores can be dispersed at relatively long distances by wind,rain and insects.

Severe epidemics of FHB occur when warm and humid weather coincide with abundance of spores during wheat anthesis[5,6].Crop debris infected with FHB pathogens serves as the main reservoir of pathogenic inoculum [7]. Noncultivated grasses may serve as an additional source of inoculum, genetic recombination and gene flow between different F. graminearum populations for cultivated wheat [8].Indeed, it was recently shown that F. graminearum is most likely associated with native grasses in North America before becoming a major pathogen of wheat and related cereal species [9]. The growing practice of monoculture wheat cultivation has probably increased the rate of pathogen evolution and forced the natural balance shifting in favour of the pathogen,thereby demanding interference by breeders to elevate plant resistance. Moreover, recent changes in cropping systems and climate have favoured the development of FHB, resulting in more frequent and severe epidemics, even in regions where FHB has not been previously reported[10,11].

In this review article, we will first introduce a brief overview of the main achievements and challenges in improving wheat resistance against FHB. Thereafter, the biology of FHB pathogen and plant defence mechanisms against the pathogen will be described. We will briefly review the concepts of effectoromics and loss-of-function of susceptibility genes and demonstrate successful applications of these phenomena to different pathosystems. Finally, we will discuss potential exploration of these concepts for improving wheat's resistance against FHB.

1.1. A brief history of FHB research

With the emergence of FHB as an important disease in North America at the early 1900s, initial research to improve FHB resistance in wheat had been conducted at the University of Minnesota from 1920 to 1950. Afterwards, the focus of wheat research had to be changed due to the emergence of severe stem rust epidemics around the world in the 1950s [12]. Major FHB research was resumed only in the early 1980s by the International Maize and Wheat Improvement Center(CIMMYT) in cooperation with Chinese researchers, and subsequently germplasm exchanges and screenings of thousands of germplasm under artificial inoculation initiated.Despite global-scale research and international cooperation and continuous efforts to curb FHB, the disease has reemerged in the 1990s and had a devastating effect in the USA, Canada, China, Europe, and South America [13-15]. The range of severe FHB epidemics between 1991 and 1997 caused direct economic losses of 1.3 billion USD for wheat and barley owing to yield losses and price discounting and about of 4.8 billion USD of direct and secondary economic losses were recorded in the USA alone [16,17]. The direct and indirect economic losses from FHB were estimated to be at around about 2.7 billion USD from 1998 to 2000 for wheat and barley in the Great Plains of the USA and the Canadian Prairies [18].Apart from economic losses, FHB causes negative consequences on human, animal and environmental health. The occurrence of trichothecene mycotoxins in food and feed can cause a number of toxic effects, including weight loss,vomiting, diarrhoea and even death. To limit the problems associated with deoxynivalenol (DON) mycotoxin contamination, many countries have imposed restrictions on the maximum rates of DON mycotoxins for daily intake in small cereal grains and their derivative products for humans and animals [19-21]. Subsequently, CIMMYT declared that FHB is the main limiting factor for wheat production throughout the world [12]. The re-emergence of FHB epidemics was partially explained by expansion of the cultivated area of maize together with the adoption of minimal tillage or no-tilling agricultural practices, which favour the accumulation of fusarium conidia and provide favourable conditions for pathogen development [22].

1.2. What have we learned?

International collaboration on Fusarium research was initiated in the early 1980s by CIMMYT.Over the ensuing years,a global network of research institutes in the USA,China,Japan,Europe, South America and Australia has been developed for the improvement of FHB resistance in wheat[23].Initially,the initiative has focussed on large-scale screening of wheat germplasm from breeders and different gene banks for FHB resistance. Wheat resistance to FHB is a multi-gene trait heavily influenced by environmental factors. Therefore,reproducibility of results from different groups was initially a significant challenge. To overcome this problem, screenings were conducted in multiple disease-conducive locations in Mexico. The mutual FHB initiative included shuttle breeding and germplasm exchange programs between CIMMYT and Chinese research institutes in the 1980s[24,25].The resistance of local wheat germplasm was improved via introduction of FHB resistant Chinese germplasm. Previously, no resistant or moderately resistant wheat genotypes were available while only 0.4% of CIMMYT's wheat genotypes were moderately susceptible. After the introduction of FHB resistant Chinese wheat germplasm into crossing schemes, the situation has been quickly improved and the frequency of moderately susceptible and moderately resistant genotypes increased to 13.5% and 2.4%, respectively, in 1987 [26,27]. Quick improvement of resistance was mainly owing to the introduction and pyramiding of new resistance QTL not present in the CIMMYT's germplasm pool. However, the introduced exotic resistance was not adapted outside of Chinese environments or for intensive agricultural practices due to low yield performance and undesirable agronomic traits. Therefore,time-consuming backcrossing programs were required to eliminate the traits negatively affecting the agronomic performance[28,29].

Genetic resistance based on monogenic inheritance follows the gene-for-gene interaction model and is more suitable for manipulation in plant breeding. In contrast, genetic resistance against necrotrophs and hemibiotrophs shows complex multigenic inheritance, making breeding efforts challenging. Comparing the resistance of winter wheat cultivars released from 2002 to 2017,Schweizer[30]concluded that resistance to biotrophic pathogens such as the rusts and powdery mildews has been improved significantly, while resistance to the tan spot disease caused by the necrotrophic pathogen Pyrenophora tritici-repentis, and FHB caused by the hemibiotrophic F. graminearum remained almost the same in Germany.Furthermore, Mesterházy [31]proposed that wheat resistance against the rusts and powdery mildews is easier to improve than that against Septoria tritici leaf spot and FHB caused by hemibiotrophic pathogens. Nevertheless, development of highly resistant commercial wheat cultivars to FHB remains an elusive goal.So far,more than 200 QTL associated with FHB resistance in wheat have been identified and new QTL are found every year[13,32].Although genetic techniques such as marker assisted selection (MAS) and genomic selection (GS) can facilitate the accumulation of desirable QTL into an elite genotype, this task is difficult to achieve because of the relatively small effects afforded by these QTL and in some cases their linkages with undesirable traits.Nevertheless,the major resistance locus Fhb1 from the wheat cultivar ‘Sumai 3' provides relatively large effect (about 30%-40% increased FHB resistance). Stacking multiple QTL conferring FHB resistance could also additively contribute to FHB resistance and improve the reliability of disease phenotyping.For instance, the complete resistance of cultivar‘Wangshuibai'to FHB is conferred by about 11 QTL[32,33].

Despite half a century of intensive wheat breeding for resistance to FHB in global scale,none of the new commercial wheat cultivars show higher FHB resistance than the old cultivar‘Sumai 3',which was developed in 1970.Sumai 3 has been used globally as a key source of resistance against FHB.This cultivar was developed via conventional breeding methods by crossing the susceptible Italian cultivar ‘Funo'with the moderately susceptible Chinese landrace ‘Taiwan wheat',and a resistant genotype was selected in a segregating population under infection pressure in the field[34].Remarkable resistance of‘Sumai 3'to FHB may be due to a successful transgressive segregation where the performance of progeny can exceed that of the parents.FHB resistance observed in the Canadian cultivar‘Emerson'is another example of successful transgressive segregation, as the parents of this cultivar do not carry any known QTL conferring resistance to FHB[35,36].However,in general,it is difficult to obtain such transgressive segregation events which confer strong resistance to FHB.For instance, Mesterházy [31]screened several thousand wheat genotypes from general breeding programs and from crosses with exotic resistance sources ‘Sumai 3' and ‘Nobeoka Bozu'between 2010 and 2016.These authors found that the level of FHB resistance and the frequency by which FHB resistant genotypes could be recovered was much higher, as expected,in breeding programs where resistant sources were used.However, it was also possible to detect genotypes with moderate or even highly resistant genotypes from general breeding programs by phenotypic screenings conducted under high disease pressure[31].

So far,breeding for FHB resistance has been hindered by both the complexity of the FHB resistance,which, as stated above, is controlled by many genes with relatively small effects, and the lack of knowledge of exact location of these genes and mechanisms of resistance[35,37].Although many QTL conferring FHB resistance were identified in different wheat germplasm,so far only few QTL, Fhb1 and Fhb2, were genetically dissected to clone the genes underlying these QTL [38,39]. Dhokane [38]identified six putative resistance-associated genes located within or near the Fhb2 locus. These genes were proposed to be responsible for cell-wall fortification and restriction of the pathogen spread in rachis within a floret as well as for detoxification of the DON mycotoxin [38]. Subsequently, two independent research teams, Li [40]and Su [41], found recently that Fhb1-mediated resistance is due to a critical 752-bp deletion within the histidine-rich calcium-binding-protein gene designated as TaHRC or His. However, these two groups proposed different explanations for the underlying resistance mechanism.Li[40]found that a deletion in the resistant allele HisRresulted in the generation of a protein 14 amino acids longer than that produced by the susceptible allele HisS, suggesting that the resistance is a result of gaining a new function. In contrast, Su[41]showed that overexpression of the putative HisRdid not lead to increased FHB resistance, but knockout of a HisSallele conferred FHB resistance in susceptible wheat genotypes, suggesting that HisSallele is a susceptibility factor promoting successful development of the pathogen.Lagudah and Krattinger[42]overviewed these two contradictory hypotheses in detail and concluded that further research is needed to clarify the nature of His resistance.However,it is evident through these recent studies that His gene is an important contributor to FHB resistance and its conservation in cereal crops can provide potential new options for improvement of plant resistance to FHB in wheat and other plant species[27,40,43].As discussed in more detail below,more recently,the gene underlying another QTL(Fhb7)mediating FHB resistance has also been cloned and functionally characterised.

2.Lifestyle of the FHB pathogen

Pathogens have evolved various strategies and lifestyles to invade plants.Generally,pathogens are divided into two main categories, biotrophs and necrotrophs, based on their lifestyles. The biotrophs obtain their nutrients from living host cells and cannot survive in dead cells. In contrast, the necrotrophs kill host cells and feed on them [44]. F.graminearum and F. culmorum are considered hemibiotrophs,the class of pathogens which penetrate into the plant as biotrophs, but later become necrotrophs[45-47].

Unlike mammals, plants do not possess adaptive immunity, but they have evolved innate immune system, which consists of multiple layers of active and passive defence mechanisms [48,49]. The first layer of plant defence against FHB is the so-called passive resistance.This type of resistance may consist of morphological and physiological traits such as kernel density,flowering time and duration,and plant height.These traits provide more or less favourable environmental conditions for disease development by affecting microclimatic conditions around wheat heads. Similarly, wax coating of spikelets can reduce water availability required for Fusarium conidial growth and penetration while the degree of anther extrusion from spikelets can increase the likelihood of successful infections by the pathogen [13,50]. Active plant defences consist of two parts; basal resistance or pathogenassociated molecular pattern(PAMP)-triggered immunity and effector-triggered immunity or R gene-mediated resistance.Pattern recognition receptors (PRRs) located at cellular membranes recognise conserved pathogen patterns such as bacterial flagellin and fungal chitin microbe associated molecular patterns or MAMPs also known as PAMPs. PAMP-triggered immunity(PTI), which is initiated by PRRs,provides the second layer of defence [51,52]. After pathogen recognition, several defence mechanisms are activated, including stomatal closure to prevent pathogen penetration; callose deposition to strengthen cell walls; oxidative burst (reactive oxygen species or ROS production) necessary for defence signalling and further defence reactions (e.g. fortification of cell walls, DNA mutations and protein oxidations that eventually cause cell death);mitogen-activated protein kinase(MAPK)pathways to enhance signalling and activate defencerelated genes[52-57].

To invade host cells, manipulate their physiology and metabolism, and overcome PTI, pathogens secrete effector proteins[58].This triggers a second layer of plant defence that involves intracellular receptors or plant resistance (R) proteins.Direct or indirect sensing of pathogen effectors by plant R proteins activates the effector-triggered immunity (ETI).However, pathogens can produce modified or mutated effectors, which can suppress PTI or escape detection by R proteins [59,60]. To date, numerous effectors and cognate R genes have been identified from plants [58,61,62]. Biotrophic pathogens are mainly detected by the plant leucine-rich repeat (NBS-LRR) class of R proteins, triggering a hypersensitive cell death (HR) response. HR enables to isolate the pathogen by surrounding it with dead host cells. It is an effective measure to control biotrophic pathogens which cannot survive on dead cells. However, HR may promote the development of necrotrophic pathogens that feed on dead cells[63-65].

Hemibiotrophs combine certain features of the two previous classes of pathogens, and often possess a more complicated arsenal used to penetrate and parasitize plant cells.Being isolated by dead cells, hemibiotrophs transform their lifestyle into a necrotrophic one during host infection. F.graminearum and related Fusarium spp., produce mycotoxins,especially after transformation into necrotrophic lifestyle,that can contribute to plant cell death. Finally, a third layer of plant defence which involves the production of RR resistance-related proteins and metabolites is triggered to encounter pathogen attack(Fig.1)[45,56,66].

3. Effector-assisted selection of resistant germplasm

Harold Flor[67]proposed the gene-for-gene hypothesis,which is based on the specific interactions between pathogen avirulence (Avr) genes and plant R genes [67]. It has taken more than half a century to establish the molecular basis of this hypothesis [68]. The new term “effector” emerged in the 1990s and is now used much more broadly than the old term avirulence (Avr) [69]. Subsequently, hundreds of pathogen effectors were identified over the last two decades,contributing to our understanding of host-pathogen interactions.It was found that effectors can suppress host defences in different ways such as by interfering with host chloroplast functions and/or signalling pathways.Some effectors destroy or silence host RNA or bind DNA to suppress transcription and downregulate defence gene expression [57,70-73]. Development of molecular tools, which enable the identification and utilisation of effectors, has consequently been heralded as a new breeding strategy. This strategy called effectoromics is a novel approach which focuses on the identification of pathogen effectors and using them as functional markers to define plant R genes.Such approach is considered potentially useful to speed up resistance breeding[57,74].

Fig.1-A model illustrating different layers of plant defences employed by wheat against Fusarium graminearum.Passive plant defence,which consists of range of morphological and physiological traits,constitutes the first layer of defence against FHB.Passive plant defences limit or escape the infection through physical/morphological barriers or physiological traits such as early or delayed flowering times.Layers 2 and 3 represent active plant defences. Layer 2 enables resistance to pathogen invasion through pathogen detection and signalling while layer 3 is mediated via multiple pathways through accumulation of number of resistance-related proteins and metabolites limit the spread of the pathogen within the spike.PRR,pattern recognition receptors;PAMP,pathogen-associated molecular pattern;ROS,reactive oxygen species;MAPK,mitogen-activated protein kinase;DON,deoxynivalenol;RR,resistance-related proteins and metabolites.

Quick identification of effectors can be facilitated by the availability of specific sequence motifs or “fingerprints”shared by effector families. For instance, a marker for detecting effectors of oomycete class pathogens(Phytophthora infestans, Hyaloperonospora arabidopsidis, etc.) is the conserved RXLR pattern/motif (Arg-X-Leu-Arg) [75-77]. Rapid detection of effectors by the presence of the RXLR motif and consequent development of a complete catalogue of P. infestans effectors speeded up the identification of R genes in potato (Solanum tuberosum L.), Arabidopsis and lettuce (Lactuca sativa) [78-85].However,so far,such universal sequence motifs could not be commonly found for effectors of fungal pathogens.Currently,effectors are predicted using a few criteria such as small size,cysteine content and whether they are secreted out of fungal cells. However, some exceptions to these criteria were also found.For instance,Phytophthora effectors proteins,CRN1 and CRN2, consisting of about 450 amino acids, are larger than typical effector proteins [86]. It was also proposed that the definition of the term effector should be broadened to include any pathogen-derived molecule including toxins and enzymes involved in the degradation of host cell wall and detoxifying plant defence molecules such as phytoalexins[59]. However, the integrated knowledge of the sequence similarities, general criteria and functional evaluation together with computational tools enable the discoveries of candidate effectors for fungal pathogens from different genera. A number of candidate effectors from wheat pathogens have been identified, and some were functionally validated [87-90].

Initially, effectoromics was applied for detecting nucleotide-binding site leucine rich repeat (NBS-LRR) class of R genes of potato and Arabidopsis against biotrophic pathogens.Therefore,although the term effector was often used for small secreted proteins from biotrophic pathogens in particular, as stated above, this term could be extended to any pathogen-derived molecules associated with host-infection.Nowadays,the effectoromics is implemented not only against biotrophic, but also hemiobiotrophic and necrotrophic pathogens of various crops [90-93]. In wheat, effector-assisted selection was studied in necrotrophic pathosystems. Mycotoxins of necrotrophic pathogens are well known effectors that are easily detectable [93-95]. Screening of germplasm using pathogen effectors can be performed through agroinfiltration, potato virus X (PVX) agroinfection (it is well established for Solanum species) or infiltration of effectors produced using the expression system of Escherichia coli or Pichia pastoris[92,96-98].

In wheat, susceptibility to tan spot caused by the necrotrophic pathogen Pyrenophora tritici-repentis is strongly correlated with effector ToxA sensitivity. ToxA is a host selective necrotrophic effector that triggers strong necrosis in wheat genotypes that carry the Tsn1 disease susceptibility gene. A high-throughput screening procedure for evaluating wheat genotypes through the infiltration of ToxA into wheat leaves has been developed and used in Australia[90,99]. This method helped quickly eliminate the S gene Tsn1 from commercial cultivars. As a result, the area sown with ToxA-sensitive cultivars was reduced from 30.4% to 16.9% during the three-year period following the use of this screening system [90,100]. However, the application of this method to other pathogens such as Parastagonospora nodorum could be more complicated, because all three homoeologous of the susceptibility gene must be eliminated to attain wheat resistance to the disease. In addition, undiscovered effectors that can differ between different regional populations may exist in the pathogen. Notwithstanding, effector-assisted selection can be an effective way for determining weak and environment-depended QTL [90,101]. Moreover, this method enables to dissect components of quantitative resistance,develop diagnostic markers and fine-map susceptibility genes. Such markers can be converted into Kompetitive Allele-Specific PCR (KASP) markers for rapid selection of desirable alleles. Thus, applying effector-assisted selection for developing diagnostic markers can increase the pace of resistance breeding in wheat against FHB and other plant pathogens[100,101].

3.1. Can effectoromics be used for breeding against FHB in wheat?

As stated above, resistance to FHB is a quantitative trait controlled by many genes with relatively small effects.Screening of germplasm for FHB resistance largely depends on the precise phenotyping, which is determined by the degree of gene expression and environmental conditions.Many efforts have been made for improving the accuracy of evaluating wheat for FHB resistance. Despite this, multilocation and multi-year trials cannot completely eliminate the inaccuracy of FHB phenotyping due to specific biology of the pathogen and the environmental variability encountered in field trials. For instance, even slight variation of flowering time between different genotypes to be phenotyped requires that the inoculations be conducted at different times, contributing to the variation observed. Artificial mist irrigation systems provide uniform humidity for wheat heads during inoculations at early stages of the disease development;however, it is not easy to control microclimatic conditions and temperatures in the field. As a result, the pathogen may produce different amounts of effectors depending on the environment[102].

Reproducibility of F. graminearum inoculations from one year to the next is much less accurate than that of biotrophic pathogens[13,103].The same isolate may cause significantly different disease development, which is correlated with its different DON producibility at different years. Several studies indicate a high correlation between isolate aggressiveness and its mycotoxin producibility. However, Mesterházy [102], based on the decade-long research results, found that the same isolate may show up to ten-fold difference in its ability to produce DON on the same wheat genotype in different years.According to the study, the level of DON concentration significantly correlated with the amount of precipitation for the 11-day period after inoculations[102].It is well known that Fusarium isolates differ by the type and proportion of different trichothecene mycotoxins they produce, and the isolates, which produce DON and 3-acetyldeoxynivalenol (3-AcDON) mycotoxins, are much more aggressive than those producing T-2 or nivalenol type mycotoxins. However, the proportion of different mycotoxins produced by the pathogen may also depend on the environment.Mesterházy [102]observed that the same isolate was nivalenol producing isolates in one year, can turn into DON producers in another year.This variability in producing different amounts and types of mycotoxins depending on the environments can contribute to the inaccuracy of phytopathological tests. In contrast, the infiltration with effectors (e.g. purified proteins)eliminates the influence of the environment on pathogen and enables the precise dosing of the effector protein and highthroughput phenotyping [90,99]. Moreover, an accurate phenotyping may increase the efficiency of other genetic technics such as marker-assisted selection(MAS),genomic selection(GS),and genome wide association analysis(GWAS)for precise defining of host resistance genes[101,104].

Evaluation of wheat resistance to FHB through the infiltration of leaves tissues with Fusarium mycotoxins at seedling stage is an attractive preliminary screening method that could reduce the number of laborious head inoculations (Table 1).Associations between the resistance of different wheat tissues determined under controlled conditions and field resistance to FHB have been studied for several decades. A high association between resistance of wheat coleoptile tissue segments to Fusarium mycotoxins and field resistance to FHB was noticed in the 1980s [105]. However, in a number of ensuing studies, a significant correlation between resistance of different wheat tissues (seedlings leaves, coleoptile segments, anther-derived callus and anther-derived embryos)and field FHB resistance could not be found[106-109].In other studies, relatively high correlations were found between resistance to DON during wheat seed germination and field FHB resistance [110]. A high correlation was found between DON resistance detected by infiltrating wheat spikelets with DON during anthesis and field resistance to FHB[111-112]and several candidate wheat genes for DON resistance were identified [112].

It is well known that F.graminearum mutants with partially or completely abolished DON production show significantly reduced ability to spread within the head,while their ability to penetrate and cause infection in the infected florets was not altered [113-116]. Therefore, it was concluded that DON is a virulence factor (effector) which is not necessary for initial penetration[56,117].Development of transgenic wheat plants expressing DON-toxification enzymes can be a powerful strategy to elevate FHB resistance. Two recent examples of such strategy are transgenic wheat expressing a barley UDP-glycosyltransferase (HvUGT13248) gene and the Fhb7 gene from Thinopyrum elongatum[118,119].Fhb7 encodes a glutathione S-transferase (GST) and confers broad resistance to Fusarium species by detoxifying trichothecenes through deepoxidation[119].Nevertheless,resistance to DON is only one component of a complex interaction between wheat and Fusarium. Paranidharan [120]performed metabolite analysis of wheat heads inoculated with F. graminearum and DON and determined that among RRI (resistance-related induced)metabolites only few of them were common for both inoculations. The obtained results evidence that there are many of effectors with relatively small effects in pathogen's arsenal[120].

?

One of the major difficulties in applying effectoromics for FHB resistance breeding is the obscure knowledge available on R genes in wheat.Traditionally,R genes refer to those that are directly or indirectly involved in sensing of effectors of biotrophic pathogens. No such R gene against FHB has so far been identified in wheat.Therefore,resistance against FHB is a quantitative trait that may be associated with both developmental (e.g. spike morphology), molecular (e.g. the production of pathogenesis- and resistance-related metabolites and/or absence of effector targets). In addition, F.graminearum effector repertoire is poorly defined and the research has been focused on the trichothecene mycotoxin effectors [57]. The hemibiotrophic lifestyle of F. graminearum indicates that the pathogen may possess a broad arsenal of effectors which enable fungal penetration as a biotroph and then spreading as a necrotroph [57,121,122]. Identifying F.graminearum effector repertoire remains a challenging task;however new achievements in this research field can provide new options for improving and speeding up wheat breeding for durable resistance to FHB in wheat.

4. Loss-of-function of S genes as a novel strategy for resistance breeding

Traditionally, breeding for disease resistance has been focused on the introgression of plant resistance genes(R genes)identified through germplasm screening. However, remarkable examples of durable resistance for several decades owing to the loss-of-function of S genes have also been demonstrated. This irrefutably shows the significance and potential contribution of this alternative strategy in resistance breeding[123]. Susceptibility genes/factors (S genes) are broadly defined as host factors which enable compatibility with pathogens and facilitate their development [91,121]. Accordingly,loss-of-function mutations in S genes can make pathogens unable to invade the plant and cause disease. Such a loss-offunction, which can occur naturally [124]or be induced through mutagenesis [121,125], typically leads to recessive resistance.Although resistance generated by loss-of-function of S genes has been used for more than half a century, the susceptibility gene concept was established after the identification of PMR6 in Arabidopsis in 2002 [126]. The term“susceptibility gene” was then coined by Eckardt [127]and subsequently used by other authors in both academic and practical sense as a novel breeding strategy [51,121,128-132].One of the best described examples of S genes is the loss-offunction of the Mlo gene in barley. Barley mutants with recessive non-functional mlo were found in the late 1930s as spontaneously occurring mutations in Ethiopia. Since then,mutagen-induced barley mlo mutants have been obtained repeatedly [49,133,134]. Barley varieties with recessive mlo loss-of-function mutants were cultivated extensively since 1990 for about 700,000 ha each year in several European countries. Despite this, mlo-mediated resistance is still effective and provides durable resistance against the powdery mildew pathogen Blumeria graminis f.sp.hordei[49,135].

This resistance mediated by loss of S genes is considered to be more durable than R gene mediated resistance,because the former type of resistance is not based on the recognition of pathogen effectors that can change rapidly to overcome R gene-mediated resistance. Interestingly, some S genes is conserved between monocots and dicots. For instance, after the discovery of barley Mlo orthologs in Arabidopsis by Consonni [136], the homologs of barley Mlo were found in almost all important agricultural commodities such as cereals(wheat [137,138]and rice [139]), vegetables (tomato, pepper[140], cucumber [141], and melon [142]), legumes (peas[143,144]and lentils [145]), fruit trees and shrubs (apples[146], grapevines [147], peaches, and strawberries [148]), and flowers (petunia [149]and roses [150]). It is still unknown whether all these Mlo homologs can also act as susceptibility genes in their respective hosts.

The nature of susceptibility factors and why S genes have been retained during evolution is still not well understood.However, several hypotheses can be considered:

1) S genes may function as negative regulators of immune responses to avoid collateral effects such as spontaneous lesion development and reducing unnecessary physiological expenses that could otherwise be allocated into plant growth[49,51];

2) While conferring susceptibility to a specific pathogen such as a biotroph,an S gene(e.g.Mlo)may confer resistance to a different pathogen such as a necrotroph[63,64,151];

3) It is possible that S genes perform certain essential functions in various biological processes. For example,the S gene Xa13 of rice is essential for pollen development[152].

Importantly, several examples (e.g. Mlo, Pmr4, and Dm6)demonstrated that the loss-of-function of some specific S genes can provide durable resistance without significant or visible fitness cost [123,135,138,153]. However, in other cases,impairments of S genes can cause negative pleotropic effects on important agronomical traits[49].

The model plant Arabidopsis is the first plant species extensively studied for functional characterization of S genes [153]. Huibers [135]showed that loss-of-function mutations in tomato (Lycopersicon esculentum) orthologs of the two known S genes of Arabidopsis PMR4 and DMR1 provides resistance against Oidium neolycopersici, which causes powdery mildew in tomato. Sun [153]performed protein sequence comparisons based on the highest level of homology of amino acid sequences to identify potato homologs of 11 known Arabidopsis S genes. Silencing of six putative S genes resulted in complete or partial resistance in potato against P. infestans [153]. These findings suggest that S-gene function is conserved across different plant species against the pathogens of the same genera,related pathogens or even pathogens which belong to a different genus and displaying a different lifestyle. For example, the loss-of-function of PMR4 and DMR1 provided resistance not only to the biotrophic downy mildew fungus Hyaloperonospora parasitica but also the hemibiotrophs F.graminearum and F. culmorum in Arabidopsis [154]and P.infestans in potato [153]. However, apart from pathogens,MLO influences the beneficial plant-fungus interactions,reducing the colonization of a root endophyte in barley[155]and facilitating colonization of arbuscular mycorrhizae fungus in barley and wheat [156].

5. Molecular techniques used for down-regulating host S genes

Assuming that S genes are functionally conserved among different plant species; targeted mutations can be induced in such genes using recently developed technologies. So far,mutagenesis has been the most broadly used method for inactivating S genes in plants. For instance, application of Targeting Induced Local Lesions in Genomes (TILLING), a method of using chemical mutagenesis and a highthroughput detection of mutations in specific regions of interests [157], enabled the development of a series of genotypes with inactivated S genes in different plant species[39,132,138,149,158-164].However,TILLING for polyploid plant species such as wheat can be more challenging than diploids,because independent mutants for a gene of interest need to be identified in all three sub-genomes and combined by crossing to generate a full knock-out. Acevedo-Garcia [163]identified ethyl-methanesulfonate (EMS) induced missense mutations in all three hexaploid wheat TaMlo homoeologues. Homozygous triple mutant lines (Tamlo-aabbdd) showed partial lossof-function of the Mlo gene and enhanced resistance to the powdery mildew pathogen Blumeria graminis f. sp. tritici [163].Combining EMS-induced missense mutations in Mlo homoeologues of tetraploid durum wheat had the strongest effect on resistance to powdery mildew in mutagenized plants[138]. Another caveat is that random mutagenesis can generate many other mutations across the genome. Some of these mutations may cause negative effects on plant vitality and fitness unless the mutant plants are extensively backcrossed to the parental line to eliminate background mutations [165-167]. Nevertheless, plants carrying nontargeted mutations in S genes are not considered GM(genetically modified)in many jurisdictions.

RNA interference (RNAi) is another powerful method to silence S genes. RNAi-mediated gene silencing can be triggered by introducing a double-stranded RNA (dsRNA) construct complementary to the endogeneous gene whose expression is to be silenced as described originally in Caenorhabditis elegans [168]. In addition, small RNAs (smallinterfering RNAs (siRNA) and micro-RNAs (miRNAs)) play a role in down-regulation of gene expression[169,170]and have been used widely for plant pest and pathogen control[171-176]. Host-induced gene silencing (HIGS) is an approach where transgenic host crop expresses dsRNAi molecules targeting virulence-associated genes in the pathogen. HIGS has been successfully used in different plant species against different pathogens including FHB [177-180]. RNAi based S gene silencing was demonstrated in barley[181],Arabidopsis,tomatoes [135], potatoes [153], grapevines [147]and apples[146]. Although RNAi-based plant protection is an effective strategy to control diseases,its application remains restricted due to weak public acceptance and strict legislation of GMOs in food and feed production in many countries. However,exogenous application of dsRNA or siRNAs, which is called spray-induced gene silencing (SIGS), can be an alternative to RNAi-based approaches. SIGS is considered a non-transgenic approach and thus can be promising for fungal disease control[176,182,183].

New gene editing techniques such as clustered regulatory interspaced short palindromic repeats/CRISPR-associated protein 9(CRISPR/Cas9)and transcription activator-like effector nucleases (TALEN) allow precise silencing of S genes in different plant species[165].Targeted genome modification of all three TaMlo homoeologues [137]and Enhanced Disease Resistance1 (TaEDR1) [184]conferred resistance to powdery mildew. Recently, Brauer [185]showed that virus-induced gene silencing and CRISPR-mediated genome editing of DON induced transcription factor TaNFXL1 confers resistance to F.graminearum in wheat.

Despite these exciting possibilities, using CRISPR tools is subject to certain limitations for commercialisation in most countries [186]. The European Court of Justice has imposed the same strict regulations for genetically edited plants as for conventional GMOs [187,188]and the negative effect of GMO legislation will continue to impact new gene-editing technologies and commercial uptake in the EU.

6.Concluding remarks

Deciphering molecular crosstalk between F. graminearum and wheat has recently gained momentum because of the availability of the genome sequences of both components of this pathosystem [189-191]. As discussed in this review article,effectoromics and the loss-of-function of S genes are new approaches with considerable potential to improve durability of FHB resistance in wheat. Recent emergence of effector-assisted breeding has accelerated the identification of new resistance genes and their introduction into commercial cultivars of potato and tomato [78,90,92]. Recent effectoromics studies in other pathosystems provided new high-throughput approaches for accurate and effective phenotyping and mining of resistance genes in plants.However,our knowledge about FHB effectors and wheat susceptibility factors is still restricted compared to other host-pathogen systems.

Recent advances in “omics” and bioinformatics have allowed quick identification of large number of candidate effectors from a variety of plant pathogens. However, so far,the majority of candidate effectors have not been functionally validated because of the laboriously time-consuming nature of this process [87]. We propose that the data generated from transcriptomic, proteomic, metabolomic and genomic studies complement each other and provide an efficient toolbox for identification of new candidate effectors and corresponding R and/or S genes of wheat (Fig. 2). Understanding of plantpathogen “cross-talks”and effector biology of Fusarium species can provide new means for developing durable resistance to FHB in wheat. Genome editing tools such as CRISPR/Cas9 and TALENs are new genetic techniques that enable precise genome editing and “switching-off” S genes in elite wheat genotypes minimize the time required for selection. In view of sophisticated pathogen biology and quantitative wheat resistance,these new approaches have a great potential for improving wheat resistance against FHB, providing sustainable wheat growing and healthy environment[192,193].

Fig.2- Complementary approaches proposed based on the exploitation of omics technologies in both the host and the pathogen to improve Fusarium head blight resistance in wheat.

CRediT authorship contribution statement

Andrii Gorash wrote the manuscript. Rita Armonien? gave input on some of the parts of this review and contributed to designing of the Fig. 2. Kemal Kazan critically reviewed,suggested major revisions, edited and contributed to overall improvement of the manuscript. All authors read and approved the final manuscript.

Declaration of competing interest

Authors declare that there are no conflicts of interest.

Acknowledgments

This work was carried out within the framework of LAMMC long-term research program “Genetics and direct genotype development of agricultural and forestry plants”. This work was partially funded by the Research Council of Lithuania,grant No.DOTSUT-218(01.2.2-LMT-K-718-01-0065).