The Open Agriculture Journal

: Background: At all stages of their development, plants are in permanent contact with causative agents of various diseases. Mechanisms of disease resistance and its durability in crops largely depend on the pathogen’s lifestyle, namely the nutrition mode and host range. Objective: The objective


INTRODUCTION PLANT IMMUNITY MECHA-NISMS WITH CONSIDERATION FOR PATHOGEN'S LIFESTYLE
Although the main goals of crop breeding are high yield and quality, they cannot be attained without consideration for In addition, it was detected that some plant-parasitic nematode groups produce pheromones named ascarosides, which are recognized by the plant defense system as signaling molecules with nematode-associated molecular patterns (NAMPs) [12,13]. NAMPs trigger the activation of innate immune responses in plants and enhance resistance to viruses, bacteria, fungi, oomycetes, and root-knot nematodes in some plant species [12].
The second tier of the plant immune system involves intracellular receptors with a central nucleotide-binding (NB) domain and a C-terminal leucine-rich repeat (LRR) domain (NLRs or NB-LRR proteins) detecting effectors (race-specific elicitors) produced by specialized pathogens [5 -8]. There are two major types of NLRs depending on their N-terminal domain: the coil-coiled (CC) type (CNLs) and the Toll/interleukin-1 receptor-like (TIR) type (TNLs), TNLs being absent in monocots and some dicots, e.g., Beta vulgaris L. [14]. NLRs account for effector-triggered immunity (ETI). NLRs are products of classical R genes in Flor's gene-for-gene model in the case of biotrophic pathogens [15,16], where effectors are avirulence factors.
Both pattern receptors and NLRs initiate signaling cascades involving multiple participants -mitogen-activated protein kinases, hormones, calcium, G-proteins, ubiquitin, and transcription factors, triggering the expression of genes for defense responses. Such responses include hypersensitive response, production of reactive oxygen species, cell wall enforcement, and production of various resistance-related proteins and metabolites [7,8,17].
Although many responses involving activation of the surface and intracellular receptors are similar, in contrast to NLRs, PRR activation is not associated with programmed cell death, and the defense responses caused by them are not so prolonged [6,18]. In addition, recent investigations of resistance against the bacterial pathogen Pseudomonas syringae in Arabidopsis thaliana L. have demonstrated the mutual potentiation between the PRR and NLR recognitiondependent defense pathways to activate strong defense against the pathogen [19].
According to the invasion model of Cook et al. [20], both elicitors and effectors are considered invasion patterns (IP), which are detected by plant IP receptors triggering different IP responses depending on the pathogen's (invader's) nature. As to the circular model of the plant innate immune system of Andolfo and Ercolano [21], hormone-regulated signaling defense pathways play a central role in plant immunity modulation, resulting in the resistance response specific to the pathogen's lifestyle. Among plant hormones, salicylic acid (SA) is the main hormone involved in resistance responses to biotrophic and hemibiotrophic pathogens acting via the product of the nonexpressor of pathogenesis-related genes 1 (NPR1) as a transcriptional activator of defense-related genes [2], whereas NPR3/NPR4 act as redundant transcriptional repressors [18]. Another important hormone that works along with SA in both local and systemic immunity is N-hydroxypipecolic acid [18], which plays a key role in SA biosynthesis. Jasmonic acid and ethylene are major hormones involved in plant response to necrotrophic pathogens [2].
Defense against biotrophic pathogens is largely based on gene-for-gene resistance due to R genes (NLRs), which commonly leads to hypersensitive response -a rapid localized cell death at the pathogen penetration site restricting biotrophic pathogen's access to water and nutrients [2,17]. On the contrary, for necrotrophs, programmed cell death is beneficial as they exploit this mechanism through NLRs for their expansion. According to Mengiste [22], toxins, necrosisinducing proteins and related molecules are equivalents of effectors in necrotrophs. The necrotrophic fungi Parastagonospora nodorum and Pyrenophora tritici-repentis produce the effector Toxin A, which causes susceptibility in wheat lines with the toxin sensitivity gene Tsn1. Tsn1 encodes a disease resistance gene-like protein with serine/threonine protein kinase and NB-LRR domains [23]. Thus, this NLR is required for susceptibility to the pathogens, whereas inactive alleles of Tsn1 provide insensitivity to Toxin A. In addition, Tsn1 also accounts for sensitivity to Toxin A produced by the necrotrophic fungus Bipolaris sorokiniana [24,25]. Another well-known example is the A. thaliana gene LOV1, which encodes a typical NLR and confers sensitivity to the fungal toxin victorin, an effector of the necrotroph Cochliobolus victoriae required for pathogenesis [26]. Victorin binds to thioredoxin TRX-h5, activates LOV1 and elicits host cell death, thus conferring disease susceptibility. So, depending on the lifestyle of the pathogen, the role of ETI in resistance may be ambiguous.
It is ETI provided by NLRs (R genes) associated with the hypersensitive response that is referred to as qualitative resistance, or vertical resistance considered in terms of complete resistance or susceptibility [7,27]. Quantitative resistance is defined as resistance expressed in the reduction of the disease rather than its absence [27] or reduced susceptibility [7]. NLRs generally provide race-specific resistance, whereas quantitative resistance is race-nonspecific, and some quantitative resistance genes may provide resistance against multiple pathogens. Poland et al. [27] considered qualitative and quantitative resistance as only two ends of the continuum with R genes and quantitative resistance loci lying towards each end of the spectrum, as implied by the phenomenon of residual resistance of some defeated classical R genes [28 -31]. Moreover, according to Kushalappa et al. [7], all genes involved in plant defense processes may be considered R genes (genes responsible for the synthesis of resistance proteins and resistance metabolites) and thus employed for plant improvement.
The most important practical question is the durability of disease resistance. According to Johnson [32], durable resistance is resistance that remains effective while a cultivar possessing it is widely cultivated. Nicks et al. [33] proposed the revision of this definition as resistance that remains effective with the deployment of a certain R gene (combination). Parlevliet [3] considered durable resistance as a quantitative trait ranging from not durable (ephemeral or transient) to highly durable. Ephemeral resistance occurs against specialists -fungi, oomycetes, and bacteria of biotrophic and hemibiotrophic nature with high evolutionary potential when resistance is largely mediated by R (NLR) genes causing the hypersensitive response, and pathogens regain their virulence due to a loss mutation in a respective avirulence (effector) gene [3,34]. Durable resistance against biotrophic and hemibiotrophic phytopathogens is, for the most part, quantitative, without a hypersensitive response. Resistance against generalists (pathogens with a wide host range) is durable and quantitative in nature. Resistance against pathogens with intermediary host range is also considered more durable than that against specialists [3]. On the contrary, resistance against viruses is often durable, irrespective of the level of specialization, even if resistance is monogenic by the hypersensitive type via NLRs [3,35], largely due to the small size of the viral genome and low fitness of genotypes with mutations for avoiding resistance genes.
Thus, qualitative R genes against evolutionarily active pathogens are frequently defeated after their large-scale deployment. A prominent example is the massive use of bread wheat cultivars with the wheat-rye 1BL.1RS translocation from the rye Petkus, which shortly led to the global defeat of disease resistance genes on this translocation: the leaf rust resistance gene Lr26, the powdery mildew resistance gene Pm8, and the stripe rust resistance gene Yr9 [36 -38]. On the contrary, the stem rust resistance gene Sr31 on 1BL.1RS remained effective for about 30 years against all stem rust races until the appearance of races of the Ug99 lineage, first described in 1999 in Uganda [39]. In addition, Sr31 further remained important for European countries, as was demonstrated by the outbreak of the highly virulent stem rust race TTTTF (later named TTRTF) in Sicily [40]. This race shows complex virulence but is avirulent to Sr31, and it appeared in 2014 in the territory of Georgia, which turned out to be a hot spot for the formation of new stem rust races via sexual recombination [41]. TTRTF was also detected in Eritrea in 2016 and in the south of Iran in 2019 [42]. Sr31 also provided resistance against recent stem rust races revealed in Germany [43]. However, in a short time, a new stem rust race with virulence to Sr31, TKHBK, appeared in Spain [44]. So, the lack of durability of important R genes presents a permanent problem for crop cultivation on a global scale, implying the search for new approaches to provide durable resistance. Main advances in the production of genotypes with durable disease resistance are associated with insights into molecular mechanisms of plant immunity and the use of biotechnological approaches such as marker-assisted selection (MAS), employment of monogenic quantitative resistance genes, including adult resistance genes, alterations of expression of certain genes associated with resistance or introduction of foreign genes via transgenic plant production, mutagenesis, including that for silencing susceptibility alleles, and genome editing. In this paper, results of studies for providing durable resistance will be considered for globally important food crops such as wheat, rice, and potato, as well as for barley. Wheat, rice, and potato are leading crops grown for human consumption on a global scale (Table 1) [45,46]. Their proportion used for human food is the highest among all the crops and comprises about 66% for wheat, 67% for potato and as high as 82% for rice. As can be seen from Table 1, their consumption per person in 2019 was 65.9, 32.4, and 80.5 kg, respectively [46]. For comparison, in 2019, the consumption of such an important crop as maize was much lower, 19.02 kg per person and the food proportion of this crop was only about 13% [46].

WHEAT AND BARLEY
Diseases contributing to serious yield losses in wheat and barley on a global scale predominantly include rusts, blotches and head blight, as well as powdery mildew and viral infections [47,48]. Yield losses from some major diseases are summarized in Table 2. However, as mentioned above, fungal biotrophic pathogens show the highest evolutionary potential, and resistance to such pathogens is commonly non-durable, in contrast to resistance to necrotrophs [3]. This part of the review considers different approaches for providing durability of resistance to mainly biotrophic pathogens. In addition, cases of ambiguity of resistance are described when a gene providing resistance to a disease at one developmental stage confers susceptibility either at another stage or to another pathogen.

Durable Resistance against Fungal Pathogens
Among a vast diversity of resistance genes to the most widespread biotrophic pathogens of wheat causing leaf rust (Puccinia recondita), stem rust (P. graminis sp. tritici), stripe rust (P. striiformis var. tritici) [45], and powdery mildew (Blumeria graminis), most of the genes are "classical" qualitative race-specific R genes, with a low durability potential [55]. These genes are referred to as major genes or seedling or all-stage resistance genes as they could be effective at all growth stages, from seedlings to adult plants. A much smaller group is represented by race-nonspecific resistance genes, which provide only moderate but durable resistance. Such genes are called adult plant resistance (APR) genes, as they commonly provide resistance only in adult plants [56,57]. Due to their effect on the development of rust fungi, they are termed slow-rusting genes as they are associated with the prolonged latent period of development of the disease, the smaller number and size of uredinia in the first two weeks after the infection in comparison with plants lacking the APR [56]. Another beneficial feature of certain APR genes is that many confer multiple resistances, are effective against different pathogens [57]. APR genes are perfect targets for MAS because they provide only a moderate level of resistance and show insufficient ability to withstand artificial infectious backgrounds [58]. The list of wheat APRs is compiled in Table  3.
The most prominent wheat APR gene is Lr34, described in 1977 by Dyck [59], who later assigned it to chromosome 7D [78]. Further studies showed that this gene was located on arm 7DS [79]. The Lr34 gene was found to coincide with the gene Yr18 for moderate resistance to stripe rust [80,81], the Pm38 gene for powdery mildew resistance [82], Bdv1 for tolerance to barley yellow dwarf virus [83], and the Sr57 gene for racenonspecific moderate resistance to stem rust [78]. The Lr34 gene is associated with leaf tip necrosis (LTN) [84]. Moreover, the association of Lr34 and LTN with resistance to spot blotch disease (QTL QSb.bhu-7D) caused by the necrotrophic fungus B. sorokiniana was revealed by Kumar et al. [85]. It should be noted that Lr34 also provides seedling resistance to leaf rust at low temperatures [86,87]. Krattinger et al. [60] performed physical and genetic mapping of the Lr34/Yr18/Pm38 /Sr57/Bdv1 gene (Lr34) and revealed that it encodes a pleiotropic drug resistance-like ATP-binding cassette transporter. The nucleotide sequence of Lr34 is 11805 bp in length, consisting of 24 exons [60]. The resistance and susceptibility alleles of the Lr34 gene differ by a single nucleotide polymorphism in intron 4, a deletion in exon 11, and a single nucleotide polymorphism in exon 12 [60,88]. Abscisic acid turned out to be the substrate of the ABC transporter encoded by Lr34 [89]. Transformation with Lr34 proved to be beneficial for durum wheat, but in barley, it induced rapid developmental leaf senescence [90,91].
The most pronounced level of APR against leaf rust is provided by the Lr22a gene from Aegilops tauschii Coss. on chromosome 2DS [92,93], which accounts for the increased latent period and reduced sporulation but not a reduction in the number of pustules per unit area. The gene was cloned and turned out to be 2,739 bp in length, consisting of a single exon and coding for a 912-amino acid NLR protein of the CC type [61]. However, the Lr22a protein showed only low sequence homology to other cloned wheat NLRs, and its closest homolog in Arabidopsis is RPM1 conferring resistance to the bacterial pathogen P. syringae expressing either avrRpm1 or avrB. RPM1 is a peripheral membrane protein residing on the cytoplasmic surface of the plasma membrane. Its activation leads to hypersensitive response and growth restriction of P. syringae strains expressing AvrRpm1 or AvrB [94,95].
Another APR gene, Lr46/Yr29/Pm39/Sr58/Ltn2, confers partial resistance to many biotrophic pathogens and is also associated with LTN [62, 96,97]. The gene was localized in the distal region of chromosome arm 1BL and shown to confer the same type of resistance as Lr34/Yr18/Pm38/Sr57/Bdv1 but at a lower level [98].
The Lr67/Yr46/Sr55/Pm46/Ltn3 APR gene originated from accession PI250413 [99] and was transferred onto chromosome 4DL of the cultivar Thatcher producing line RL6077 [63]. No yield penalty is associated with the resistance allele of Lr67 [63]. It confers partial resistance to all three wheat rust pathogens and powdery mildew and is also associated with LTN [100]. The Lr67 gene was found to encode a hexose transporter that differs from the susceptible form by two amino acids and alters hexose transport [64]. The Lr68/Ltn4 gene on chromosome 3BS is another APR conferring resistance against leaf rust only; it was first described in the spring bread wheat Parula and probably originated from the cultivar Fontana [65]. Other broad-spectrum leaf rust APRs were further identified in wheat: Lr74 on chromosome 3BS of the cultivar Caldwell [66], Lr75 on 1BS of the cultivar Forno [67], Lr77 on chromosome 3BL of the cultivar Santa Fe [68], and Lr78 on 3DS of the cultivar Toropi [69].
The Yr36 gene conferring moderate APR against only stripe rust was introgressed from emmer wheat Triticum dicoccum Schrank ex Schübl. onto chromosome 6BS [70]. The gene is temperature-dependent (the resistance is expressed under higher temperatures) and encodes a protein with a kinase domain and a putative START lipid-binding domain being essential for the resistance [71]. Several other high-temperature adult-plant (HTAP) resistance genes such as Yr52 [72] and Yr59 on 7BL [73], Yr62 on 4BL [74], as well as APRs Yr78 on 6BS [75] and Yr80 [76] on 3BL, have been identified but their presence in many cultivars over the world is questionable. The Yr15 gene showing broad-spectrum all-stage resistance to stripe rust races was also transferred to bread wheat from emmer wheat. It is located on 1BS and encodes a kinasepseudokinase protein designated as wheat tandem kinase 1 [101].
One of the most widely used broad-spectrum stem rust genes is Sr2 [102], transferred from the emmer wheat cultivar Yaroslav in the 1920s to produce the cultivar Hope [103]. The Sr2 gene is moderately effective against all races of stem rust, including the group of Ug99 races, although its expression in the field is often suppressed [104,105]. Sr2 was localized on chromosome 3BS [104,106]. It is also pleiotropic with the juvenile race-specific leaf rust resistance gene Lr27 as well as with partial resistance to powdery mildew and stripe rust (as Yr30) [77]. Surprisingly, Sr2-mediated resistance to stem rust and powdery mildew turned out to be associated with the death of photosynthetic cells around the infection site, similarly to hypersensitive response-type necrosis [107], although no NLR genes reside at the Sr2 locus [108]. Among all-stage resistance genes, an interesting case is the stem rust Sr26 gene, which retained its durability despite its deployment in a number of Australian cultivars since 1971 [109]. Sr26 was introgressed from Thinopyrum ponticum (Podp.) Z. -W. Liu & R. -C. Wang as the T6AS.6AL-6Ae#1 translocation and proved to be an NLR of the CNL type; the same expectations are for another Th. ponticum gene, Sr61, which also encodes a CNL [109]. The gene Sr62 from the wild diploid wheat relative Aegilops sharonensis Eig, which encodes a non-NLR protein (a tandem protein kinase) and provides broad stem rust resistance, is also promising [110]. However, not all such tandem kinase R proteins provide broad resistance. For example, the Triticum monococcum L. gene Sr 60 encoding a tandem kinase (wheat tandem kinase 2) proved to be only race-specific [111].
Moreover, interactions between the broad-spectrum resistance genes such as Sr2/Yr30, Lr34/Yr18/Sr57 and Lr68 were shown to confer enhanced adult plant resistance to rust diseases in some bread wheat genotypes [112]. To provide broad-spectrum resistance to stem rust employing all-stage resistance genes (NLR genes), a transgenic approach was used by introducing a transgene cassette of five resistance genes into bread wheat (Sr45, Sr55, Sr50, Sr35, and Sr22) as a single locus [113]. However, it is still possible that new races with virulence to most of those Sr genes might appear [41].
In barley, there are three genes conferring APR to the leaf rust pathogen P. hordei Rph20, Rph24, and Rph23 (Table 4) conferring high, moderate, and low levels of APR, respectively, of which Rph20 and Rph24 are objects of MAS [114]. The most prominent durable resistance source employed in barley breeding is the race-nonspecific powdery mildew resistance gene mlo, which is a recessive loss-of-function allele of the corresponding dominant allele Mlo [117,118]. Mlo encodes a plasma membrane-localized protein with seven transmembrane domains, in which the N-terminus is located extracellularly and the C-terminus intracellularly [115,119]. Among more than 40 mlo mutant alleles, only two (natural mlo11 from Ethiopian accessions and induced mutant mlo9) have been involved in spring barley cultivars since 1979, providing powdery mildew immunity [117,118]. Upon infection, barley mlo genotypes usually show the arrest of fungal pathogenesis at early stages, without the formation of haustoria and secondary hyphae. Local cell-wall callose appositions in host epidermal cells beneath attempted fungal penetration sites are formed accompanied by accumulation of defense-associated compounds, including hydrogen peroxide resulting in localized cell death and formation of necrotic leaf spots [118,120]. In addition, the host genetic background determines the efficiency of mlo resistance [118,120]. At the same time, the presence of mlo is considered to be associated with reduced resistance to some pathogens with necrotrophic developmental stages, such as enhanced sensitivity to B. sorokiniana toxins in comparison with nonmutant genotypes of barley [121], higher susceptibility to Magnaporthe oryzae [122], and Ramularia collo-cygni causing Ramularia leaf spot, which became a major barley disease in Europe [123]. In a series of mlo mutant plants of bread wheat produced by TALEN (transcription-activator-like nuclease) and TILLING (targeted induced local lesions in genomes) methods, the stronger powdery mildew resistance was also shown to be correlated with enhanced susceptibility to M. oryzae pathotype Triticum [124].
Three new APR genes (Rbgh1, Rbgh2, and Rbgh3) for powdery mildew resistance have been recently identified in barley landraces (Eth069 from Azerbaijan and HOR3270 from Turkey) in the terminal regions of chromosomes 5HL, 7HS, and 1HS, respectively [116]. As opposed to mlo, the presence of those genes was not associated with spontaneous necrosis and mesophyll cell death, and resistance was localized to the site of the attempted penetration of the fungus and cytologically involved cell wall appositions and cytosolic vesicle-like bodies, without strong induction of reactive oxygen species. One may expect that such powdery mildew APRs would not be associated with higher susceptibility to necrotrophic pathogens.
Factors of resistance against necrotrophic fungi of the genus Fusarium are considered to be quantitative, i.e., their additive effect providing the measured level of resistance in comparison with the plants lacking such factors [125]. Fusarium head blight (FHB), caused predominantly by Fusarium graminearum, is one of the most devastating diseases accompanied by the production of mycotoxins, which are harmful to humans and animals. The most effective and durable gene for FHB resistance by type II (resistance against symptom spread in the head) is Fhb1 on chromosome 3BS. This gene was first described in the Chinese cultivar Sumai 3 developed in 1972 [126]. The Fhb1 gene encodes the histidinerich calcium-binding-protein (His), and the resistance allele resulted from a 752-bp deletion involving exon 3, leading to the change of the translation start codon [127,128]. The deduced resistance variant of His is 14 residues longer than the wild-type protein and differs by 21 N-terminal amino acid residues [127]. This nucleus-localized protein is presumed to be involved in calcium signaling [127]. On the other hand, the FHB-resistant variety Sumai 3 shows a Fusarium seedling blight-susceptible reaction (resistance inversion) [129]. Apart from Fhb1, there are a number of other FHB resistance genes, including Fhb7 introgressed from Th. elongatum, which encodes glutathione S-transferase and acts via detoxifying trichothecenes [130], as well as the susceptibility factor on 4DS [131], which could be targets of MAS for pyramiding resistance genes.
An important role in the regulation of plant interaction with pathogens belongs to the NPR1 gene [132]. It was shown that in A. thaliana, the functional product of the gene plays a key role in the PR1 gene expression and switching between jasmonate-dependent and salicylate-dependent defense response [133]. Diethelm et al. [134] detected that certain alleles of homoeologous NPR1-like genes on wheat chromosomes 2D and 2A (TDF_076_2D and TDF_076_2A) conferred type II resistance to F. graminearum and F. culmorum at the level of 14.2% and 3%, respectively. Transformation of FHB-susceptible wheat cultivars with either Arabidopsis or Secale cereale NPR1 genes led to improvement of FHB resistance [135,136]. However, transferring the NPR1 gene from A. thaliana into wheat caused increased susceptibility to Fusarium asiaticum at the juvenile stage, while in adult plants, on the contrary, it conferred resistance [137]. At the same time, knocking out NPR1 genes (Ta7ANPR1) on homoeologous group 7 chromosomes in wheat increased resistance to stem rust [138]. Similarly, RNAi (RNA interference)-mediated stable silencing of the gene TaCSN5 (for constitutive photomorphogenesis 9 (COP9) signalosomea regulator of plant growth and development) conferred broadspectrum resistance to the stripe rust pathogen, which indicated that TaCSN5 as a candidate susceptibility gene could be the object of genome editing for providing stripe rust resistant genotypes [139].

Advances in Virus Resistance Due to Biotechnological Approaches
Genetically modified plants produced via transgene-based host-induced gene silencing (HIGS) based on RNAi is a promising biotechnological approach to increase resistance to various pathogens, including viruses, fungi, and nematodes [140,141]. RNAi silencing has been demonstrated to be especially effective for producing genotypes with virus resistance. For example, wheat plants transformed with a construct involving the sequence of a portion of the coat protein of wheat streak mosaic virus (WSMV) proved to be WSMV resistant [142]. Transformation of wheat plants with a polycistronic amiRNA (artificial microRNA) construct targeting various conserved regions in the WSMV genome resulted in WSMW immunity [143]. Similarly, barley plants transformed with a polycistronic amiRNA precursor construct based on the conservative sequence elements of several wheat dwarf virus (WDV) strains expressing three amiRNAs simultaneously resulted in highly efficient resistance to WDV [144]. Moreover, RNAi silencing of the endogenous wheat genes TaeIF(iso)4E and TaeIF4G encoding initiation factors induced resistance to WSMV, Triticum mosaic virus, soilborne wheat mosaic virus and a significant reduction in barley yellow dwarf virus infection [145]. Further development of this approach led to the strategy of spray-induced gene silencing (SIGS) based on spraying double-stranded RNAs (dsRNAs) and small RNAs (sRNAs) targeting essential pathogen genes on plant surfaces, which does not require genetic modification of plants [141,146].
Thus, durable resistance to fungal biotrophic pathogens in wheat and barley could be conferred by APRs. However, they provide only a moderate level of resistance and are to be supplemented by qualitative R genes with major effects. Silencing of susceptibility alleles, common for providing resistance to necrotrophs [ 23 ], or certain regulatory genes via mutagenesis or biotechnological approaches proved to be another way to increase resistance to biotrophic pathogens as well [ 118,138,139]. There are cases when the presence of some resistance factors to a certain pathogen may be associated with susceptibility to other pathogens [118] or increased susceptibility at juvenile stages [129,137], which indicates their involvement in a complex regulatory network. In addition, HIGS based on RNAi is considered to be a promising universal approach for increasing resistance to various pathogens in crop species.

RICE
Among multiple diseases of rice, the most destructive and widespread are blast caused by the hemibiotrophic fungus M. oryzae ( Pyricularia oryzae ) and bacterial blight caused by Xanthomonas oryzae pv. oryza e ( Xoo ), a biotrophic bacterium [ 147,148] (Table 5). In this part, the main genes conferring durable resistance to primarily these diseases are reviewed as well as some biotechnological approaches to improve resistance are mentioned.

Durable Resistance against the Fungus M. oryzae
As M. oryzae is a fungus with the biotrophic phase of development, R genes providing complete resistance to blast are commonly defeated in 1-7 years after the release of resistant varieties [150]. Because of this, genes conferring broad-spectrum resistance are highly valuable. The most prominent gene for durable blast resistance in rice is the racenonspecific recessive resistance gene pi21 identified in the Japanese upland variety Owarihatamochi on chromosome 4 [151] ( Table 6). The wild-type susceptibility allele encodes a proline-rich protein with a putative heavy metal-binding domain and putative protein-protein interaction (proline-rich) motifs. The recessive resistance allele is a result of a loss-offunction mutation due to the deletion of the 18-and 48-bp sequences in the proline-rich region containing PxxPxxP, which is a "core motif" for protein-protein interaction [152]. It is assumed that pi21 plays a role in the pre-penetration plantpathogen interaction through elicitor-triggered immunity and ethylene signaling [150].
Another durable broad-spectrum resistance gene, Ptr, was identified in the resistant tropical japonica variety Katy on chromosome 12. This gene encodes an untypical resistance protein with four Armadillo repeats, which may represent a non-typical E3 ligase [153]. However, E3 ligase activity in vitro was not detected, so the authors assumed that the Ptr protein is more likely involved in protein-protein interactions. The Ptr gene is expressed constitutively and codes for two isoforms localized mainly in the cytoplasm. Ptr is also required for broad-spectrum blast resistance conferred by the NLR genes Pi-ta and Pi-ta2 [153].  It should be noted that many broad-spectrum quantitative blast resistance genes turned out to be NB-LRR proteins with certain peculiarities in their structure or expression. The dominant race-nonspecific gene Pi35 on chromosome 1 from the Japanese breeding line Hokkai 188, which retains its effectiveness for 60 years, is a typical R gene encoding an NB-LRR protein bearing multiple functional polymorphisms with respect to the race-specific allele Pish [154]. Pi35+ pi21 was detected to be the most effective combination for the suppression of leaf blast [163]. In the Japanese upland rice variety Kahei, the major blast resistance QTL Pikahei-1(t) on chromosome 4 involves the gene Pi63 encoding a typical NB-LRR protein, which differs from its susceptibility variant not only in amino acid sequence but also by the higher expression level [155]. The durable panicle blast 1 (Pb1) gene derived from the indica variety Modan shows adult panicle blast resistance, which has remained effective for about 40 years: it does not provide resistance at young vegetative stages, but the resistance level increases with plant growth and is retained even after heading [164]. This gene encodes an atypical CC-NB-LRR protein in whose NB domain the P-loop is absent and some motifs are degenerated; Pb1 expression increases during the development [156]. It was further detected that Pb1 resistance was negatively dependent on three QTLs located on chromosomes 7, 9 and 11 and positively dependent on one QTL on chromosome 8 [165]. The broad-spectrum resistance gene Pi39 derives from the Chinese cultivar Haonaihuan [166]. Based on the Pi39 candidate cDNA sequences, an InDel-based marker for Pi39 gene selection was developed [167]. An interesting case of durable resistance to M. oryzae is the Chinese rice variety Gumei 4, which has been employed as a blast resistance donor for more than 50 years. This variety carries the Pigm resistance locus containing a cluster of NLR genes, among which PigmR confers broadspectrum resistance, and PigmS competitively attenuates PigmR homodimerization to suppress PigmR-mediated resistance. It turned out that epigenetic regulation of PigmS fine-tunes disease resistance; PigmS increases yield by increasing seed setting and so counteracts yield penalties induced by PigmR [157].
Using sodium azide as a mutagen, the mutant line SA0169 showing broad-spectrum blast resistance was produced from the blast-susceptible Taiwan japonica cultivar Tainung 67. This mutant line retains its broad-spectrum blast resistance for about 20 years. The combination of two regions was detected to confer blast resistance in this mutant: a 1.16-Mb region on chromosome 6 (Pi169-6(t)) and a 2.37-Mb region on chromosome 11 (Pi169-11(t)) involving 2 and 7 candidate R genes in those regions, respectively [168].
Thus, in the rice gene pool, several genes provide durable broad-spectrum resistance to blast, which can be extended by mutagenesis and its new version, genome editing, or by genetic transformation with foreign resistance genes. For example, rice blast resistance was improved via CRISPR/Cas9-targeted knockout of the ERF (ethylene responsive factor) transcription factor gene OsERF922 in the japonica rice variety Kuiku131, and the silencing did not affect agronomic traits [169]. Transgenic rice plants of the japonica rice cultivar Nipponbare expressing the wheat Lr34 gene showed improved resistance against multiple isolates of M. oryzae [170].

Durable Resistance against the Bacterium Xoo
Bacterial blight caused by Xoo is also a highly damaging disease in rice ( Table 5). Currently, more than 40 Xoo R genes have been described, some of which are broad-spectrum ones and some provide only race-specific resistance, which nevertheless proved to be rather durable [171]. For example, the race-specific gene Xa4 has been conferring durable Xoo resistance since the early 1970s. This gene encodes a cell wallassociated kinase; moreover, it strengthens the cell wall by promoting cellulose synthesis and suppressing cell wall loosening, thus increasing the mechanical strength of the culm and improving lodging resistance [158] ( Table 6). The Xa21 gene, which encodes a leucine-rich repeat receptor kinase-like protein, is a broad-spectrum resistance gene derived from O. longistaminata [159]. Xa21 expression increases with age providing full resistance only in adult plants, but overexpression of this gene in transgenic rice plants provides resistance at both seedling and adult stages [172]. The Xa3 (Xa26) gene also encodes a protein of the LRR receptor kinase type; its expression also gradually increases from the early seedling stage to the adult stage. It shows a higher expression level in the background of the japonica rice, which results in enhanced expression of defense-responsive genes, ultimately providing a higher level and spectrum of Xoo resistance as compared to the indica rice [160]. Overexpression of Xa3 in transgenic rice plants enhanced resistance in both indica and japonica backgrounds [160].
The Xa13 susceptibility gene belongs to the SWEET (Sugars Will Eventually be Exported Transporter) family encoding putative sugar transporters induced by transcription activator-like (TAL) effectors of Xoo [173]. Among others, this family includes such susceptibility genes as OsSWEET11, OsSWEET12,OsSWEET13,OsSWEET14,and OsSWEET15 [174]. The resistance protein encoded by the recessive resistance allele xa13 differs from the susceptibility variant by only one amino acid but resistance is provided via expressional non-reaction of xa13 to Xoo infection due to alterations in the promoter region [161]. The promoter region of Xa13 contains an upregulated transcription activator-like 1 (UPT) effector box, which is involved in the activation of expression by Xoo race 6 (PXO99). The induction of site-specific mutations into the UPT box using CRISPR/Cas12a technology to hinder TAL protein binding and gene activation resulted in the production of the genome-edited rice with improved bacterial blight resistance [175].
The Xa7 gene confers bacterial blight resistance for more than 10 years and, importantly, remains effective at high temperatures and drought. It turned out to be a small orphan gene encoding a protein of only 113 amino acid residues, which is distinct from any other resistance proteins [162]. The XA7 protein is anchored in the endoplasmic reticulum membrane and induces programmed cell death. The Xa7 promoter contains the 27-bp effector binding element, which is essential for AvrXa7-inducing expression [162]. According to Luo et al. [176], Xa7 belongs to executor R genes and acts as a guard against pathogen's exploitation of the rice major susceptibility gene SWEET14. Lines with broad-spectrum Xoo resistance were produced by CRISPR/Cas9 technology via the generation of InDels in the TAL effector-binding element of the promoter of OsSWEET genes involved in disease susceptibility in rice plants [177].
Despite advances in the production of resistant forms via genetic transformation and genome editing, pyramiding Xoo R genes with resistance to different races using MAS remains an important approach to provide durable broad-spectrum resistance to bacterial blight of rice [148,166,178].

Durable Virus Resistance
In the context of resistance durability, the Stvb-i gene providing durable resistance to rice stripe virus (RSV), an RNA virus causative of rice stripe disease, should be mentioned. This gene encodes a 1,649-amino acid protein that lacks a NB-LRR domain but possesses a domain homologous to the histidine kinase/HSP90-like ATPase superfamily protein and is expressed mainly in meristematic tissues. It was suggested that Stvb-i may be involved in the protection of the meristematic tissue not only from RSV multiplication but also from heat stress [179], which is of importance in connection with global warming.
Thus, in rice, like in wheat, most cloned durable resistance genes against Xoo and M. oryzae are distinct from NLRs or, in the case of the latter pathogen, some encode untypical NB-LRR proteins or show peculiarities in their expression. Likewise, silencing of some susceptibility genes also improves resistance [ 164,171,173,174,177]. However, a unique case of achieving resistance to a number of different pathogens has been reported [180,181]. Transgenic rice lines overexpressing the rice BSR1 (BROAD-SPECTRUM RESISTANCE 1) gene, which encodes a putative receptor-like cytoplasmic kinase, turned out to be highly resistant to Xoo and M. oryzae [180]. In addition, they showed resistance to the bacterium Burkholderia glumae, which causes bacterial seedling rot and bacterial grain rot, as well as to the necrotrophic fungus Cochliobolus miyabeanus, causing brown spot [181]. Moreover, rice plants with BSR1 overexpression showed slight resistance even to RSV [181]. Thus, overexpression of one gene, BSR1, is a unique and promising case as it could confer resistance to multiple diseases caused by pathogens of different trophic classes: biotrophs, hemibiotrophs and necrotrophs.

POTATO
Potato ( Solanum tuberosum L.), the third most important food crop, is threatened by many different pathogens: bacteria, fungi, oomycetes, viruses, viroids, nematodes, and phytoplasmas, which affect both crop yield and quality. Moreover, because of the vegetative propagation of potatoes, pathogens could be transmitted via tubers to subsequent generations. This review focuses on advances in durable resistance to late blight caused by the oomycete Phytophthora infestans and to the economically important viruses, which cause high yield losses ( Table 7).

Approaches to Provide Durable Resistance to the Oomycete P. infestans
Late blight caused by the hemibiotrophic oomycete P. infestans is one of the most devastating diseases of potatoes. The pathogen rapidly overcomes R genes, so approaches to achieve durable resistance include deployment of quantitative resistance genes or multiple R genes simultaneously [184,185]. The potato cultivar Sarpo Mira showed resistance to late blight for more than a decade after its release [186]. This cultivar was detected to carry the qualitative R genes R3a, R3b, R4, and Rpi-Smira1, as well as the quantitative resistance gene designated Rpi-Smira2, which confers partial field resistance [187]. However, later it was demonstrated that the Rpi-Smira2 gene is located at the same position as the R8 gene from the wild potato species S. demissum on chromosome 9 and both show recognition of the AVR8 effector, implying that R8 and Rpi-Smira2 are allelic [188]. R8 encodes a typical NLR protein of the CC type (CNL) with 89% homology to Sw-5, tomato spotted wilt virus resistance R protein [188] ( Table 8). The R8 gene was also shown to coincide with the previously identified QTL dPI09c for late blight resistance on potato chromosome 9 [189,190], indicating that a single major disease resistance gene can be responsible for the QTL providing durable resistance.  P. infestans R8 (Rpi-Smira2) 9 S. demissum CC-NB-LRR [188 -190] Rpi-vnt1.1 9, cisgene S. venturii CC-NB-LRR [191 -193] Rpi-sto1 8, cisgene S. stoloniferum CC-NB-LRR [191,192] Rpi-blb1 8, cisgene S. bulbocastanum CC-NB-LRR [192 -194] Rpi-blb2 6, cisgene S. bulbocastanum CC-NB-LRR [192,193] Rpi-chc1 10, cisgene S. chacoense CC-NB-LRR [192] Rpi-blb3 4

S. bulbocastanum CC-NB-LRR [195]
Rpi-amr1 11 S. americanum NRC helper-dependent CC-NB-LRR protein [196] ELR 12, cisgene S. microdontum receptor-like protein ELR (elicitin response) [197] StDMR6-1 S. tuberosum loss-of-function mutation of the susceptibility gene encoding salicylic acid 5-hydroxylase [198] StCHL1 S. tuberosum loss-of-function mutation of the susceptibility gene encoding a transcription factor involved in brassinosteroid hormone signalling, [198] PVY  [204,205] Rl adg 5 S. tuberosum ssp . andigena [206,207] An effective biotechnological approach to combine R genes in the same genotype is cisgene stacking, i.e., the introduction of stacks of cloned R genes from crossable wild potato species to existing varieties by genetic modification technology [191,192]. To achieve broad-spectrum resistance, stacks of two genes, , or three genes,  were introduced into potato varieties. Within the framework of the research project on Durable Resistance in potatoes against Phytophthora (DuRPh), four varieties were transformed with one to three R cisgenes aimed to attain durable resistance to late blight [192].
Non-transgenic biotechnological approaches for the transfer of R genes from crossable wild species include somatic hybridization of cultivated potato with a wild species and identification of resistance genes using gene-specific markers [194,195]. For example, backcross clones of potatoes with broad-spectrum late blight resistance due to the introgressed resistance genes Rpi-blb1 and Rpi-blb3 from S. bulbocastanum were produced with those methods [195].
The potential of wild potato species as sources of new late blight resistance genes seems to be not fully studied, as resistant accessions were identified among several species that were never previously reported to be late blight resistant: Solanum albornozii, S. agrimoniifolium, S. chomatophilum, S. ehrenbergii, S. hypacrarthrum, S. iopetalum, S. palustre, S. piurae, S. morelliforme, S. neocardenasii, S. trifidum, and S. stipuloideum [208]. A new R gene, Rpi-amr1, from S. americanum encoding an NRC helper-dependent CC-NLR protein with broad-spectrum resistance representing a family of nine resistant alleles is promising for the transfer of broadspectrum durable resistance against P. infestans to S. tuberosum [196]. Du et al. [197] cloned a gene for the receptor-like protein ELR (elicitin response) from S. microdontum. This protein is involved in extracellular recognition of the elicitin domain representing a conserved molecular pattern in Phytophthora species. Transformation of cultivated potatoes with the ELR gene resulted in enhanced resistance to P. infestans and the authors proposed to pyramid ELR with cytoplasmic NLRs to maximize the potential for disease resistance durability.
Genome editing technology could be effective for knocking out susceptibility genes in potatoes. Using a CRISPR/Cas9 system with co-expression of two guide RNAs, tetra-allelic deletion mutants with functional knockouts of the susceptibility genes StDND1, StCHL1 and DMG400000582 (StDMR6-1) were generated, and the edited plants showed increased resistance against late blight [198]. Of them, the authors report StDMR6-1 and StCHL1 as promising S-gene targets for late blight resistance breeding as they do not affect plant growth phenotypes.
Analysis of QTLs may be helpful in identifying durable quantitative genes for late blight resistance. The study of the relationship of 65 candidate genes with late blight resistance QTLs in three diploid potato populations PCC1, BCT, and PD detected three significant cases: the locus of a putative receptor-like protein kinase b on chromosome 11, the Lox gene on chromosome 3 and two protein phosphatase loci in a QTL with the largest effect on chromosome 12 [209]. The association mapping study of S. tuberosum Group Phureja revealed two late blight resistance QTLs with the candidate genes encoding a potato homolog of thylakoid lumen 15 kDa protein (StTL15A) and a stem 28 kDa glycoprotein (StGP28) with the 7% and 11% effects, respectively [210]. Juyo Rojas et al. [211] identified 16 organ-specific QTLs conferring resistance to late blight, explaining 13.7% to 50.9% of the phenotypic variance. In silico analysis revealed that four candidate genes for resistance to late blight have no functional genome annotation (including those for QTLs with 50.9% and 38.4% effects), while eleven candidate genes encode diverse proteins, including a leucine-rich repeat kinase.
Many QTLs for late blight resistance were identified in the wild species S. microdontum and S. pampasense [212]. The effects of those QTLs ranged from 16.9 to 47.5%, and they can be employed for introgression into cultivated potatoes [212].

Advances in Virus Resistance
The greatest advances with respect to durable resistance of potatoes were achieved for virus resistance due to natural resistance genes as well as genetic transformation and genome editing. There are two main types of resistance against potato viruses: hypersensitive resistance (HR), which is a rapid defense response resulting in the programmed cell death (necrosis) at the site of infection, and extreme resistance (ER), which is characterized by the absence of symptoms and prevention of virus multiplication at the early stage of infection [199,213,214]. HR genes are strain-specific, whereas ER genes are effective against all strains of the virus.
Among potato viruses, potato virus Y (PVY) is the most economically serious [183,215] (Table 7), with wild species being the reservoir of many PVY resistance genes. HR against PVY is conferred by Ny genes, and ER is mediated by Ry genes [199]. The ER genes Ry adg , Ry sto and Ry chc , which are employed in breeding programs, originate from S. tuberosum subsp. andigena, S. stoloniferum, and S. chacoense, respectively, and many PCR markers are available for MAS [199].
The extreme resistance gene Ry sto on chromosome 12 encodes an NLR protein with an N-terminal TIR domain, which recognizes the PVY coat protein ( Table 8); Ry stodependent extreme resistance is temperature-independent, requires EDS1 (Enhanced Disease Susceptibility 1) and NRG1 (N requirement gene 1 proteins), and is epistatic to Ny-1mediated HR [200]. Another PVY extreme resistance gene derived from S. stoloniferum, Ry-f sto , had been mapped on potato chromosome 12 [216]. Still, later Ry-f sto was considered as the allele of Ry sto . The alignment of sequences of the Ry sto and Ry-f sto alleles revealed their 100% sequence identity in the coding and non-coding regions of the 4.85 kb amplified product [200]. Potato leafroll virus (PLRV) is also among the most destructive viruses of potatoes ( Table 7). The gene Rlr etb confers PLRV resistance due to reduced PLRV accumulation in foliage and the inhibition of the systemic spread of PLRV from infected foliage to tubers. It was introgressed from the non-tuber-bearing wild potato species S. etuberosum [204,205]. The Rl adg gene conferring high resistance level and low accumulation of PLRV was derived from S. tuberosum ssp. andigena [206,207]. QTL analysis of resistance to PLRV accumulation revealed the major QTL PLRV.1 on potato chromosome 11, explaining 50-60% of the phenotypic variance, which can be traced using molecular markers [222,223].
Expression of the full-length PLRV replicase gene in transgenic potato plants of the cultivar Russet Burbank provided a high level of field resistance to PLRV [224]. Marker-free transgenic PLRV-resistant plants were generated using an inverted repeat construct corresponding to a PLRV coat protein gene segment employing the heat inducible Cre-loxP system to excise the nptII antibiotic resistance marker gene [225]. Inhibition of expression of the gene for movement protein (MP) via RNAi was also proposed as a method for the production of PLRV-resistant plants, and, as MP homologues are present in most plant viruses, the authors suggested that this technology could be used for generating virus-resistant plants of other species [226].
Thus, in cultivated potatoes, late blight resistance is mostly based on pyramiding NLRs, primarily those from wild relatives, achieved mainly by MAS and genetic transformation of cloned genes, and the nature of this resistance does not show much promise as to their longevity. Currently, genome editing is used to improve late blight resistance by silencing susceptibility genes, which could potentially provide more durable resistance. At the same time, cloning of virus resistance genes showed that NLRs provide durable resistance against potato viruses. Reliable advances in virus resistance are also associated with transgenic plants.

CONCLUSION
Different mechanisms of plant resistance highly depend on the nutrition types of pathogens and may be quite ambiguous. Molecular markers proved to be of much help for traditional breeding of the main food crops to pyramid monogenic quantitative resistance genes involved in durable resistance and qualitative R genes, significantly reducing time and cost spent to obtain lines or cultivars with the genes of interest. More importantly, direct manipulations with genotypes are becoming more widely and successfully used for understanding the resistance and susceptibility mechanisms, as well as engineering crop genotypes with broad-spectrum durable resistance, employing both transgenesis and genome editing, primarily the CRISPR/Cas technologies. Durability of resistance is the main challenge in the case of biotrophic and hemibiotrophic fungi, oomycetes, and bacteria with high evolutionary potential. However, in wheat, barley, and rice, there are a number of broad-spectrum durable resistance genes to such specialist pathogens that are different from NLRs, in contrast to potato. Related species remain a source of promising broad-spectrum resistance genes against such diseases, especially in the case of potatoes. Traditional mutagenesis and novel gene-editing technologies are of importance for improving disease resistance, in particular, by the production of loss-of-function alleles. Silencing of some, often regulatory, genes or overexpression of some host genes, as well as the introduction of foreign genes, by genetic modification are promising biotechnological ways for the production of genotypes with durable broad-spectrum resistance.

CONSENT FOR PUBLICATION
Not applicable.

FUNDING
This work was supported by the National Academy of Agrarian Sciences of Ukraine (the project 0121U000082).