Genetic approaches to dissect plant nonhost resistance mechanisms

Abstract

Nonhost resistance (NHR) refers to the immunity of most tested genotypes of a plant species to most tested variants of a pathogen species. Thus, NHR is broad spectrum and durable in nature and constitutes a major safety barrier against invasion of a myriad of potentially pathogenic microbes in any plants including domesticated crops. Genetic study of NHR is generally more difficult compared to host resistance mainly because NHR is genetically more complicated and often lacks intraspecific polymorphisms. Nevertheless, substantial progress has been made towards the understanding of the molecular basis of NHR in the past two decades using various approaches. Not surprisingly, molecular mechanisms of NHR revealed so far encompasses pathogen‐associated molecular pattern‐triggered immunity and effector‐triggered immunity. In this review, we briefly discuss the inherent difficulty in genetic studies of NHR and summarize the main approaches that have been taken to identify genes contributing to NHR. We also discuss new enabling strategies for dissecting multilayered NHR in model plants with a focus on NHR against filamentous pathogens, especially biotrophic pathogens such as powdery mildew and rust fungi.

This review discusses new strategies for dissecting multilayered immunity (i.e., nonhost resistance) in plants against nonadapted filamentous pathogens, such as powdery mildew and rust fungi.

1. INTRODUCTION

In nature, plants are resistant to most potential phytopathogens. Conversely, a majority of phytopathogens are capable of only infecting certain plant species, which defines their host ranges. The effective resistance or immunity exhibited by an entire plant species to all genetic variants of a phytopathogen species is described as nonhost resistance (NHR) (Heath, 1977, 2000; Nurnberger & Lipka, 2005; Thordal‐Christensen, 2003). NHR is a phenomenological term that also reflects the inability of a nonadapted but potentially pathogenic microbe to complete its asexual or sexual life cycle on a particular plant species (Heath, 2000; Nurnberger & Lipka, 2005; Panstruga & Moscou, 2020). Because NHR means robust and complete resistance, there is strong motivation to understand the molecular basis of NHR with the hope of using the NHR mechanisms for improvement of disease resistance against adapted and aggressive pathogens (Borlaug, 2000; Fan & Doerner, 2012; Gill et al., 2015). However, there are two major constraints that limit genetic dissection of NHR. First, there is no or rarely polymorphism in NHR as a genetic trait within a plant species, hence conventional forward genetics (which is often based on segregation of the target trait in an intraspecific population for identifying genes controlling the trait) is not easily applicable to genetic studies of NHR. Interspecific segregating populations can be useful for studying NHR, but sexual hybridization between plant species is often prohibited due to various barriers. When sexual hybridization is feasible in some cases, there may not be a suitable pathogen that has adapted to one plant species but not the other, although exceptions do exist and have been used for studying NHR (Bartaula et al., 2018, 2019; Newcombe, 2005). Second, the inability of a potential pathogen to conquer resistance of an entire plant species implies that NHR consists of multiple layers of effective immunity that collectively can block even accidental successful invasion from the most capable individuals of the pathogen population. This ability of NHR may explain why only incremental erosion of NHR rather than complete breakdown of NHR has been achieved in most cases through single rounds of random mutagenesis (Collins et al., 2003; Hematy et al., 2020; Lipka et al., 2005; Stein et al., 2006) or virus‐induced gene silencing (VIGS) (Ramu et al., 2020; Rojas et al., 2012). This is especially the case for filamentous pathogens, whose reproductive success requires brutal penetration of plant cells to derive adequate nutrients from the attacked plants. Hence, the high (theoretical) potential of NHR contrasted with its poor mechanistic characterization makes the question “what is the molecular basis of NHR” among the top 10 unanswered questions in the field of molecular plant–microbe interactions (Harris et al., 2020).

Despite the aforementioned inherent difficulty, there have been considerable efforts in NHR studies in the past two decades and around 30 genes involved in NHR have been identified in Arabidopsis and other plants using various forward and reverse genetics approaches (Fonseca & Mysore, 2019; Lee et al., 2017; Panstruga & Moscou, 2020). Diverse and probably layered defence mechanisms have been shown to contribute to NHR (Figure 1). Preformed physical barriers such as leaf cuticular layers (Yu et al., 2019) and preformed chemical barriers such as constitutively synthesized secondary metabolites of antimicrobial activity (Bowyer et al., 1995; Papadopoulou et al., 1999) help certain plants prevent infection from nonadapted pathogens. Apart from the preformed barriers, NHR is built on the same molecular basis of host resistance against adapted pathogens (Panstruga & Moscou, 2020). In other words, pathogen‐associated molecular pattern (PAMP)‐triggered immunity (PTI) and/or effector‐triggered immunity (ETI), the two well‐characterized immune mechanisms responsible for host resistance (Jones & Dangl, 2006; Ngou et al., 2022), probably also underlie NHR in most cases (Lee et al., 2017; Panstruga & Moscou, 2020; Senthil‐Kumar & Mysore, 2013). A complete genetic dissection of most robust NHR such as rice's immunity to rust and powdery mildew, if possible, may also lead to discovery of unexpected novel mechanisms that either may not exactly fit the PTI–ETI framework or may reveal how PTI or ETI is regulated at an unknown level(s), or host factors that are strictly required for pathogenesis (Figure 1). This prospect has motivated and will continue to motivate researchers to take novel and enabling approaches to dissect multilayered NHR. Given that there are many comprehensive reviews on the concept and possible mechanisms of NHR (Ayliffe & Sorensen, 2019; Bettgenhaeuser et al., 2014; Fonseca & Mysore, 2019; Lee et al., 2017; Panstruga & Moscou, 2020; Schulze‐Lefert & Panstruga, 2011; Senthil‐Kumar & Mysore, 2013), we will focus our review on summarizing the approaches employed for NHR studies and propose improved strategies for dissecting multilayered NHR against biotrophic filamentous phytopathogens in model dicot (Arabidopsis) and monocot (rice) plants.

FIGURE 1.

Multilayered defence mechanisms underlying nonhost resistance (NHR) may be dissectible. For any filamentous pathogens to achieve reproductive success in a particular plant species, they must overcome several spatiotemporally distinct or connected defence barriers of the plant (a). Likely defence layers of NHR include (1) the epicuticular wax, (2) the cell wall, (3) antimicrobials by pathogen‐associated molecular pattern (PAMP)‐triggered immunity (PTI), (4) antimicrobials and hypersensitive response (HR) by effector‐triggered immunity (ETI), and/or (5) possible novel but currently unknown mechanisms. Like an onion, multilayered NHR against a nonadapted filamentous pathogen is dissectible using appropriate strategies and tools (b). Red stars in (a) indicate the defence layers where the microbial invasion is halted; grey lines in (b) between layers imply possible mechanistic connections

2. APPROACHES USED TO ELUCIDATE NHR MECHANISMS

To understand the molecular genetic basis of a biological trait, one needs to identify the gene(s) that controls the target trait. This entails genetic association (linkage) studies with a population(s) that possesses phenotypic polymorphism (i.e., trait segregation) for the identification of a candidate gene(s). Indeed, over the past decades, researchers have utilized various strategies to identify natural or create polymorphisms in NHR against a particular phytopathogen for the identification of genes and specific cellular mechanisms contributing to NHR. Because NHR is typically exhibited by an entire plant species, natural polymorphism in NHR is rare and only found in special circumstances. Major efforts have been focused on creating plant (or pathogen in some cases) mutants and/or conditions in which NHR is partially or significantly eroded or completely abolished by chemical inhibitors, random genome‐wide mutagenesis, or targeted mutagenesis. Below, we summarize the main approaches that have been used for genetic studies of NHR with a few of the most relevant examples and propose a few new improved strategies.

2.1. Chemical genetics to understand major requirement of NHR

Chemical genetics is the investigation of the function of genes or their engaged signalling pathways by using small molecules to perturb a particular cellular function (Specht & Shokat, 2002). Pharmacological inhibitors were used to infer an important role of the actin cytoskeleton for NHR of plants against nonadapted powdery mildew (Kobayashi et al., 1997a, 1997b), rust (Song et al., 2012; Wang et al., 2015a), and Colletotrichum species (Shimada et al., 2006). For example, treatment of eds1‐2 mutant leaves with cytochalasin E, an actin polymerization inhibitor, significantly compromised Arabidopsis NHR against barley powdery mildew (Blumeria graminis f. sp. hordei; Bgh) (Yun et al., 2003). Likewise, inactivation of myosins by pharmacological inhibitors in Arabidopsis leaves compromised penetration resistance against Bgh, which was further substantiated by genetic mutations of genes encoding myosin XI proteins (Yang et al., 2014). Similarly, chemical inhibition of phosphatidic acid production by phospholipase D (PLD) increased Bgh penetration success on wild‐type Arabidopsis leaves, which led to the identification of PLDδ to be important for NHR against nonadapted powdery mildew. Miklis et al. (2007) found that functional actin cytoskeleton is required for both NHR and mlo‐mediated resistance against powdery mildew in barley (Miklis et al., 2007). In a recent screen for chemical suppressors of mlo‐mediated resistance to powdery mildew, Wu and colleagues found that alloxan (5,5‐dihydroxyl pyrimidine‐2,4,6‐trione) and some of its structural analogues can partially suppress mlo‐mediated resistance as well as NHR to powdery mildew, probably through destabilization of the cytoskeletal architecture (Wu et al., 2017, 2020). These results further suggest that mlo‐mediated resistance and NHR share certain common cellular mechanisms (Humphry et al., 2006). Notably, Qin et al. (2021) recently provided genetic evidence that corroborates the observations made with chemical inhibitors. These authors found that genetic disruption of the ARP2/3 complex and formins, two actin‐nucleating systems, compromised Arabidopsis penetration resistance to nonadapted fungal pathogens (Qin et al., 2021). Besides the findings concerning the roles of actin cytoskeleton and phospholipases in NHR, chemical inhibitors of protein kinases and protein phosphatases were also found to impair pea NHR to a bean pathogen, Fusarium solani f. sp. phaseoli, although the specific cellular mechanism(s) impacted is not known (Hartney et al., 2007).

With the availability of relevant chemicals, screening for chemicals impacting NHR via leaf infiltration is simple to perform in model plants such as Arabidopsis thaliana and Nicotiana benthamiana. However, due to potential chemical toxicity and/or the lack of precise mechanistic information about the cellular targets of most chemical inhibitors, the utility of chemical genetics in NHR studies is rather limited.

2.2. Exploring natural variations to identify genes underpinning NHR in near‐hosts

In nature, host adaptation of most pathogens is thought to be gradual and continuous, although host jumps do happen (Morris & Moury, 2019; Panstruga & Moscou, 2020; Thines, 2019). Hence, in many cases, there may be no strict demarcation of host and nonhost. Accordingly, plant species that are evolutionarily close to a host plant species may exhibit less robust NHR to a corresponding pathogen, therefore they can be described as “near‐hosts” (Figure 2a). Conversely, such a pathogen may be in a “near‐pathogen” state with respect to its poor or limited adaptation to the near‐host. Plants as near‐hosts have been used for genetic studies of NHR against rust and powdery mildew fungi (Atienza et al., 2004; Delventhal et al., 2017; Dracatos et al., 2016; Jafary et al., 2008). One common strategy for using near‐hosts to understand NHR is to find intraspecific or interspecific variation in NHR and identify genes underlying the differences through genetic mapping. For example, in a recent report, Haghdoust et al. (2021) tested 492 barley accessions and two mapping populations with pathogenically diverse cereal rust (Puccinia) isolates representing distinct formae speciales adapted to different cereal species. The authors found that about 80% of the barley accessions were either immune or near immune to all pathogens tested and identified major quantitative trait loci (QTLs) in barley that may contribute to NHR (Haghdoust et al., 2021). These observations support the notion that there is a continuum between NHR and host resistance, and in some cases the resistance phenotypes are quantitative traits (Bettgenhaeuser et al., 2014). Through a similar strategy, Wang et al. found that the lectin receptor‐like kinases encoded by orthologous genes at the Rphq2 locus from cultivated barley and wild barley confer resistance to rust pathogens. Interestingly, the resistance in either background is much stronger to their respective nonadapted rust compared to their respective adapted rust, indicating that the host status of a particular barley genotype to different leaf rust fungi is quantitatively affected by the same orthologous receptor kinase (Wang et al., 2019). In rare cases where natural or deliberately constructed interspecific hybrids between a near‐host and a host are fertile, NHR in the near‐host plant may be amenable to genetic studies for mapping genes controlling the NHR. For example, barberry (Berberis spp.) is known to be the alternate host of grass rust fungi (Puccinia spp.). A natural population of Berberis xottawensis, derived as an interspecific hybrid of B. vulgaris (host of Puccinia graminis) and B. thunbergia (nonhost of Puccinia graminis), was used for the identification of several QTLs that are associated with the NHR against P. graminis in B. thunbergia (Bartaula et al., 2018, 2019). Another example concerns triticale, a man‐made interspecific hybrid between wheat (Triticum) and rye (Secale) made in the 19th century (Longin et al., 2012). Triticale used to be resistant to powdery mildew (Blumeria graminis) formae speciales either adapted to wheat (B. graminis f. sp. tritici) or rye (B. graminis f. sp. secalis) (Walker et al., 2011), suggesting that genes contributing to NHR from wheat and rye are dominant and complementary. Interestingly, Menardo and colleagues discovered that the powdery mildew forma specialis that has overcome the resistance in triticale (and wheat but not rye) is a sexual hybrid of the wheat powdery mildew and rye powdery mildew (hence named B. graminis f. sp. triticale) (Menardo et al., 2016). This finding not only sheds insight onto evolutionary mechanisms of new powdery mildew speciation through hybridization but also implies an important role of powdery mildew effectors in host adaptation via suppressing immunity genes contributing to NHR.

FIGURE 2.

Major strategies for dissecting genetic components contributing to nonhost resistance (NHR). Distinct defence layers are depicted by different shapes; different components in the same layer are indicated by different colours, with grey being indicative of functional suppression by the invading pathogen; functional impairment of unknown immunity genes by mutagenesis is indicated by a dark “x”, while targeted ablation of known genes is indicated by grey “x”. Waved lines depict linear mechanistic relationships, dotted lines depict interconnection between different mechanisms, “=” indicates functional redundancy, and circled “S” indicates a host susceptibility factor. Pathogen‐associated molecular pattern (PAMP)‐triggered immunity (PTI); effector‐triggered immunity (ETI); virus‐induced gene silencing (VIGS). (a) Using natural near‐hosts. NHR in nonhost α and nonhost β is completely or largely intact and thus is not easily amenable to genetic studies. In contrast, NHR in a near‐host is conferred by only one or a few genes and thus is genetically tractable using various approaches as shown. (b) Employing “top‐down” forward and/or reverse genetics. Ethyl methanesulfonate (EMS) mutagenesis was used to identify pen mutants compromised in penetration resistance; the pen mutant could be used to identify genes involved in postpenetration resistance by further EMS or CRISPR mutagenesis. Arabidopsis–powdery mildew interaction is used as an example. Bgh, barley powdery mildew Blumeria graminis f. sp. hordei. (c) Employing “bottom up”‐ and/or “nullify knowns”‐based forward genetics. Removal of postpenetration resistance can allow more sensitized and saturated forward genetic screens for novel pen mutants, while next‐round forward and/or reverse genetics screens in the background of a higher‐order, immunocompromised mutant can reveal unknown residual defence mechanisms contributing to NHR. (d) Breaking rice's NHR to powdery mildew and rust fungi. One or more rounds of multiplexed CRISPR mutagenesis of all known or candidate immunity genes may create rice mutants susceptible to the nonadapted fungi. If not, a susceptible factor(s) may be inadequate or missing in rice. In this case, the known barley S factor Mildew locus O may be expressed in the rice mutant to enable pathogenesis of Bgh or other cereal powdery mildew

Definitive determination of genes underlying NHR in a near‐host was provided by Cevik and colleagues in 2019. The authors screened 593 inbred lines from an Arabidopsis multiparent advanced generation intercross (MAGIC) mapping population derived from 19 accessions and identified two transgressive segregants that are susceptible to Albugo candida, indicating an absence of gene(s) critical for NHR (Cevik et al., 2019). Through gene mapping using F₂ progenies derived from crossing these two individuals back to the 19 MAGIC parental accessions, the authors found that resistance to A. candida race 2 (Ac2V) can be explained in each accession by at least one of the four NLR (nucleotide‐binding site leucine‐rich repeat) genes identified. These findings demonstrated the utility of A. thaliana as a near‐host of A. candida for revealing a critical role of ETI in NHR (Cevik et al., 2019). Using wheat as a near‐host to powdery mildew adapted to related cereals, Bourras and colleagues analysed the allelic polymorphisms at the wheat resistance gene locus Pm3 (encoding an NLR) and those of the corresponding Avr genes in the adapted and the nonadapted powdery mildew formae speciales and concluded that Pm3‐enabled ETI also explains NHR against the nonadapted powdery mildew (Bourras et al., 2019).

In addition, the molecular basis of NHR in a particular near‐host species can also be inferred from the genetic polymorphisms and pathogenesis patterns among related pathogen species (including nonadapted and adapted) or their sexual hybrids. By comparing the effector repertoires of Pseudomonas syringae DC3000 (adapted to Arabidopsis) and P. syringae T1 (adapted to tomato but not Arabidopsis), Sohn et al. (2012) identified two effector genes, avrRpt2 and hopAS1, responsible for ETI in nonhost A. thaliana (Sohn et al., 2012). Interestingly, while hopAS1 is broadly present in P. syringae strains contributing to virulence in tomato, all tested P. syringae strains that are pathogenic in Arabidopsis carry truncated hopAS1 variants (presumably to avoid ETI). These observations suggest that ETI plays an important role in Arabidopsis NHR against a broad range of P. syringae strains (Sohn et al., 2012). More recently, Laflamme et al. (2020) constructed the pan‐effectorome from 494 P. syringae strains based on the pan‐genomic analyses, screened for core effectors that can elicit ETI in A. thaliana, and identified 59 ETI‐eliciting effectors with orthologues distributed among 96.8% of P. syringae strains. Interestingly, ETI in Arabidopsis was found to be activated by only a small number of NLRs, with CAR1 and ZAR1 being able to recognize orthologous effectors present in about 95% of P. syringae strains, many of which are not adapted to Arabidopsis (Laflamme et al., 2020). The new findings further provide genome‐scale evidence that ETI underpins NHR of Arabidopsis as a near‐host to many nonadapted P. syringae biotypes that are close relatives of the adapted P. syringae strains.

The above examples demonstrate that NHR in a near‐host against bacterial or fungal pathogens may be genetically tractable and that typical ETI may underpin NHR under such pathocontexts. These findings also imply that PTI in near‐hosts has been partially or largely overcome by the invading pathogens, hence the delivery of effectors and ETI. Obviously, the availability of near‐host/near‐pathogen accessions/biotypes is a prerequisite for successful employment of the above‐mentioned strategies to study the genetic basis of NHR. Infection tests of large segregating populations and/or whole‐genome sequence analyses are required to map and identify genes contributing to NHR.

2.3. Applying mutagenesis‐aided forward genetics to identify components of NHR

In cases where NHR consists of more than one layer with each being effective in halting the invasion of pathogens, the resistance is unlikely breakable by the nonadapted pathogens. For example, NHR of dicot plants such as Arabidopsis against powdery mildew fungi adapted to monocot plants is probably multilayered and insurmountable by the pathogens (Figure 2a, nonhost α). The inducible defences of the NHR against filamentous pathogens can be divided into two major layers, penetration resistance and postpenetration resistance. Logically, removal of the first layer (i.e., cell wall‐based penetration resistance) of NHR via mutagenesis is considered the very first step towards dissecting the genetic mechanisms of NHR.

Indeed, ethyl methanesulfonate (EMS) mutagenesis of the A. thaliana Col‐0 accession followed by screening mutagenized M₂ seedlings with barley powdery mildew (Bgh) has led to the identification of a series of penetration (pen) mutants with increased penetration success of the nonadapted powdery mildew (Figure 2b). The cloning and functional characterization of PEN1 to PEN4 genes made a significant contribution to our understanding of plant cell wall‐based penetration resistance (Collins et al., 2003; Hematy et al., 2020; Lipka et al., 2005; Stein et al., 2006). Using the same strategy, Fukunaga and colleagues identified two allelic necrotic spotted lesion (nsl) mutants on inoculation of nonadapted fungi and identified the causal mutations in NSL1, which is predicted to encode a protein containing a putative membrane‐attack complex/perforin (MACPF) domain without known function (Fukunaga et al., 2017). Interestingly, the nsl1 mutants also developed necrotic spotted lesions on flg22 treatment and the cell death phenotype was abrogated in the absence of PEN2 or when the biosynthesis of salicylic acid (SA) is defective. This suggests that PAMPs from bacteria or fungi can enhance production of a PEN2‐dependent, tryptophan metabolism‐derived secondary metabolite that can induce programmed cell death via the SA‐dependent pathway in the absence of NSL1 (Fukunaga et al., 2017). A similar approach was also used for identifying genes contributing to NHR to nonadapted bacterial pathogens. Zhou and colleagues identified NHO1, a gene encoding glycerol kinase, to be important for NHR and host resistance against nonadapted P. syringae pv. phaseolicola (Kang et al., 2003; Lu et al., 2001). For inducing random mutations in DNA, besides chemical mutagen treatments, other mutagenesis methods, including radiation mutagenesis, T‐DNA insertion, and transposon tagging, can also be used to introduce mutations into genes associated with NHR. For example, using Medicago truncatula Tnt1 retrotransposon insertion lines, Uppalapati and colleagues revealed a role of abaxial leaf epicuticular wax of M. truncatula in NHR against nonadapted rust fungi (Uppalapati et al., 2012).

Random, genome‐wide mutagenesis allows nonbiased forward genetic screens to identify key genes that control a specific phenotype. The efficiency of this approach depends on the efficacy of DNA mutagenesis and identification of phenotypic alteration. If the defect in NHR is minor and requires microscopy or other laborious assays for phenotyping, mutant screening and subsequent identification of causal mutations by positional cloning can be time‐consuming, hence mutant screening and gene identification may be difficult to reach saturation. Also, because a single round of forward genetic screen cannot circumvent genetic redundancy, this approach is ineffective in identifying genes that are functionally redundant.

2.4. Using reverse genetics to identify components of NHR

This approach has been used in several laboratories to search for genetic components contributing to NHR against nonadapted bacterial pathogens. The success of this strategy depends on the establishment of a relatively efficient gene‐manipulation method such as virus‐induced gene silencing (VIGS) and/or targeted mutagenesis method such as CRISPR in a near‐host plant (Figure 2a). Using VIGS, a role of SGT1 in both NHR and host resistance was identified (Peart et al., 2002; Wang et al., 2010). Notably, by using VIGS in N. benthamiana for initial gene identification (Senthil‐Kumar et al., 2013) and functional studies of orthologous/homologous genes in Arabidopsis, the Mysore laboratory identified a dozen novel genes involved in NHR of N. benthamiana and Arabidopsis against the nonadapted bacterial pathogen P. syringae pv. tomato T1 (Fonseca et al., 2020; Fonseca & Mysore, 2019; Pant et al., 2020; Ramu et al., 2020), demonstrating the power of reverse genetics. These studies have collectively revealed how the leaf apoplast environment is modified during NHR to restrict the survival and proliferation of nonadapted bacteria.

With the development of the next‐generation sequencing technologies, candidate genes involved in NHR can be more efficiently identified through (pan)genomics and comparative transcriptomics studies (Gangurde et al., 2021; Iven et al., 2012; Tufan et al., 2009). When such gene expression information is coupled with VIGS and CRISPR, the reverse genetics approach for identification of genes involved in NHR can become more efficient. An obvious advantage of this approach is that when multiple members of a gene family are targeted simultaneously, this can reveal relevant mechanisms of NHR where genetic redundancy exists.

More and more mutant lines of model plants have been generated, characterized, and made available to the research community. One may collect a panel of mutants with mutations in genes belonging to a gene family or involved in a particular signalling or metabolic pathway that may serve a role in immunity and test them with one or a group of nonadapted pathogens. This more targeted reverse genetics approach may identify causal mutations that compromise NHR. For example, infection tests with a panel of phosphoinositide signalling‐related Arabidopsis mutants with T‐DNA insertions led to the identification of PLDδ in both penetration and postpenetration resistance to nonadapted powdery mildew (Pinosa et al., 2013; Zhang et al., 2018).

2.5. Combining stepwise forward and reverse genetics to identify novel components of NHR

To dissect multilayered NHR once a major component of NHR is discovered, the corresponding mutant can be used to identify genes that function in the next layer via another round of forward or reverse genetic screen using the same or other nonadapted pathogens (Figure 2b). Because of the essential role of PEN2‐dependent glucosinolate metabolic pathway in penetration resistance against a broad range of pathogens (Bednarek et al., 2009; Clay et al., 2009; Lipka et al., 2005; Sanchez‐Vallet et al., 2010), most stepwise genetic studies described in published literature were conducted in pen2 mutant backgrounds. For example, Kopischke et al. (2013) further mutagenized the Arabidopsis pen2‐1 mutant and screened for new mutants with an altered response to infection by the nonadapted pathogen Phytophthora infestans. Characterization of the first new mutant led to the identification of a phospholipid: sterol acyltransferase (PSAT1) in negative regulation of mesophyll cell death associated with excessive callose deposits (Kopischke et al., 2013). Characterization of an additional mutant from the same screen implicated EDR1 in negative regulation of PTI response induced by flg22 and elfl8 (Geissler et al., 2015). Using a stepwise reverse genetics strategy, Langenbach et al. (2013) found that BRT1, encoding a UDP‐glucosyltransferase in the phenylpropanoid pathway, was highly induced in Arabidopsis plants lacking pen2 on inoculation with soybean rust Phakopsora pachyrhizi. They constructed the pen2 brt1 double mutant and found that it could support more mesophyll haustorium formation, despite the fact that the haustoria are not functional enough to support further fungal development (Langenbach et al., 2013). Interestingly, the brt1 mutation in the wild‐type (Col‐0) background exhibited normal NHR, suggesting that BRT1 plays a role in postpenetration resistance activated only when PEN2‐dependent penetration resistance is breached (Langenbach et al., 2013).

By introducing the pad3‐1 mutation, which impairs camalexin production (Zhou et al., 1999), into the pen2‐1 mutant background, Schlaeppi and colleagues found that the pen2‐1 pad3‐1 displayed susceptibility to nonadapted Phytophthora brassicae, indicating that NHR of Arabidopsis against P. brassicae relies on a combined action of PEN2‐dependent production of 4‐methoxyindol‐3‐ylmethylglucosinolate and PAD3‐dependent production of camalexin, both of which are derived from the tryptophan metabolic pathway (Schlaeppi et al., 2010). Since then, more specific studies have shown that PTI signalling components, including transcription factors and MPK3/6, play important roles in regulating the production of camalexin during NHR (Frerigmann et al., 2016; Pastorczyk et al., 2020; Saga et al., 2012; Xu et al., 2016; Yang et al., 2020). Using a similar approach, SOBIR1, a co‐receptor for PAMP recognition, was found to be required for NHR against Pyricularia oryzae in Arabidopsis (Takahashi et al., 2016, 2018), and BAK1, another PAMP co‐receptor, has recently been shown to play a role in Arabidopsis postpenetration resistance against Alternaria brassicicola independent of tryptophan‐derived metabolites (Kosaka et al., 2021).

Notably, the pen2 mutant was also tested with different nonadapted pathogens for exploring new mechanisms of NHR. For example, it was recently found that in epidermal cells of the pen2 mutant, atypical small chloroplasts act as defence‐related motile organelles by specifically positioning immune components in the plant epidermis. Blocking the distinct steps of such responses decreases NHR against nonadapted fungi (Irieda & Takano, 2021).

2.6. Going forward: Improved/new strategies for identifying novel genes contributing to NHR

The efficiency of any forward genetics approach depends on how readily the desirable mutants can be identified. A typical “top‐down” forward genetic screen (Figure 2b) was employed to identify pen1 to pen4 mutants based on the detection of impaired penetration resistance as a top layer of NHR by microscopy (Collins et al., 2003; Hematy et al., 2020; Lipka et al., 2005; Stein et al., 2006). Although extremely successful in setting the early milestones for understanding penetration resistance as part of NHR, this screen by design was laborious and thus difficult to reach saturation. Moreover, the identification of the causal mutations using microscopy‐based phenotyping for positional cloning was also time‐consuming, which probably in part explains why the characterization of PEN4 was only recently published (Hematy et al., 2020).

Inspired by an early observation that the Arabidopsis pen2 pad4 sag101 triple mutant supported profuse sporulation visible to the naked eye from the nonadapted pea powdery mildew (Lipka et al., 2005), we can envisage a sensitized “bottom‐up” forward genetics strategy to more efficiently identify genes contributing to NHR in penetration resistance (Figure 2c). Specifically, we made an Arabidopsis mutant in which key components of postpenetration resistance, EDS1, PAD4, and SID2, were mutated and found the eds1 pad4 sid2 triple mutant to be supersusceptible to the adapted Arabidopsis powdery mildew isolate Golovinomyces cichoracereaum UCSC1 (Zhang et al., 2018). Interestingly, the triple mutant still possesses largely intact penetration resistance to a strawberry powdery mildew pathogen Podosphaera aphanis (pathogen β in Figures 2c and 3a). Knocking out PEN2 in the eds1 pad4 sid2 pen2 background (Figure 3c) by CRISPR resulted in profuse sporulation of P. aphanis on the quadruple mutant plant visible to the naked eye (Figure 3b). Therefore, one may anticipate that screening for genes involved in penetration resistance as part of NHR against nonadapted cell wall‐penetrating filamentous pathogens using an EMS‐mutagenized eds1 pad4 sid2 population would be efficient and possible to reach saturation because potential pen mutants can be recognized by the naked eye instead of having to use microscopy (Figure 2c).

FIGURE 3.

Making it visible: a sensitized “bottom‐up” forward genetics strategy for identifying genes involved in penetration resistance against nonadapted powdery mildew fungi. The immunocompromised eds1‐2 pad4‐1 sid2‐2 (eps) triple Arabidopsis mutant is supersusceptible to the adapted powdery mildew isolate Golovinomyces cichoracearum UCSC1 (Zhang et al., 2018). However, the eps mutant remains resistant to the nonadapted strawberry powdery mildew pathogen Podosphaera aphanis, as evidenced by the lack of fungal mass visible to the naked eye at 8 days postinoculation (dpi) (a, left) and restricted fungal growth revealed by trypan blue staining followed by microscopy (a, right). CRISPR‐targeted mutagenesis using the pHEE401E vector containing Cas9 under control of an egg cell‐specific promoter (Wang et al., 2015b) was deployed to mutate PEN2. Inoculation of 6‐week‐old 20 T₁ plants with P. aphanis identified two plants to be susceptible to this nonadapted pathogen as shown by profuse fungal sporulation visible to the naked eye (b, left), which was confirmed by trypan blue staining and microscopy at 8 dpi (b, right; red arrows indicate conidiophores). Sequencing the target site of PEN2 revealed a single nucleotide insertion in the second exon of PEN2 in plant #1 and a 211 nucleotide deletion in plant #2 (c), both of which result in an early stop codon (TGA). All plants were grown under 22°C, 65% relative humidity, short‐day (8 h light at 125 μmol/m²/s, 16 h dark) for 6 weeks before fungal infection. The above results demonstrate that an immunocompromised mutant with defective postpenetration resistance can be used to effectively identify genes involved in penetration resistance against nonadapted cell wall‐penetrating filamentous pathogens through saturated mutagenesis followed by efficient mutant screening with the naked eye

To efficiently identify novel genes involved in postpenetration resistance, we can also envisage a “nullify‐knowns”‐based forward genetics strategy that requires the construction of an Arabidopsis mutant in which key components of both penetration resistance (e.g., PEN1, PEN2, and PEN3) and postpenetration resistance (e.g., EDS1, PAD4 and SID2) are mutated. If a higher‐order Arabidopsis mutant such as eds1 pad4 sid2 pen1 pen2 pen3 is still resistant to cereal powdery mildew fungi or any other filamentous pathogens (pathogen α in Figure 2c), a genetic screen with the EMS mutagenized M₂ population of such a mutant line may enable identification of mutants that permit growth of nonadapted pathogen α visible to the naked eye (Figure 2c).

CRISPR/Cas9‐targeted mutagenesis has been widely used for knocking out genes in several model plant species (Gaillochet et al., 2021). The ability to simultaneously knock out up to 15 genes by multiplexed CRISPR system in Arabidopsis, N. benthamiana, and rice (Stuttmann et al., 2021; Zhang et al., 2021) has created unprecedented opportunities to perform highly efficient reverse genetic screens targeting genes and gene families in single or multiple signalling networks involved in control of complex biological traits such as NHR. This can not only circumvent the functional redundancy constraint inherent to conventional forward genetic screens but also should greatly facilitate the “nullify knowns”‐based forward genetics strategy described above. By taking advantage of rice's high efficiency in both transformation and CRISPR‐mutagenesis, Meng and colleagues constructed the first large‐scale CRISPR/Cas9 mutant library in rice that is of high quality, with good coverage and uniform distribution (Meng et al., 2017). This rice library consists of 14,000 independent transgenic rice lines expressing Cas9 and one or more small guide (sg) RNAs targeting one or more of 12,802 rice genes, which can be used for screening mutants with biological phenotypes of interest, including compromised NHR. Such genetic screens and subsequent gene identification seem to blur the boundary of forward genetics and reverse genetics, and should make gene discovery much easier using the sgRNA sequence information. We further envision generation of a rice mutant population expressing Cas9 and a library of multiplexed (e.g., 16×) sgRNA targeting all possible known or predicted rice immunity genes (>500). Such a rice mutant library may be very useful for identification of high‐order rice mutants with compromised NHR to cereal rust and powdery mildew fungi (Figure 2d), and if so, we can address perhaps the most challenging and symbolic question regarding NHR (i.e., why is rice immune to these cereal fungi?).

However, it is also possible that a host factor(s) required for pathogenesis of either one of the two biotrophic pathogens may be missing or insufficient in rice and, as such, immunocompromised higher‐order rice mutants may be unable to support infection of the nonadapted fungi. Given that one or more host Mildew locus O (MLO) proteins are strictly required for host‐entry of powdery mildew (Buschges et al., 1997; Consonni et al., 2006) and that overexpression of barley MLO can increase host penetration by a nonadapted powdery mildew in barley (Elliott et al., 2002), it is possible that expression of a cognate rice MLO is inadequate in rice epidermal cells, hence contributing to NHR. Therefore, expressing barley MLO from its native promoter or the rice ubiquitin promoter in an immunocompromised rice mutant may further increase the chances for breaking down rice's NHR to barley or other cereal powdery mildew (Figure 2d).

Furthermore, nonadapted pathogens may lack a mechanism for nutrient acquisition from nonhost plants, (partly) accounting for their failure in reproduction or proliferation. For example, phytobacteria Xanthomonas spp. use transcription activator‐like (TAL) effectors to target and manipulate the transcription of plant SWEET sucrose transporter genes for increasing sugar availability in the apoplast (Bezrutczyk et al., 2018). If there are no binding sites of TAL effectors from a Xanthomonas species in the promoter region of any SWEET genes of a nonhost plant under attack, the bacteria may remain nonpathogenic even if the plant is immunocompromised. Ectopic expression of a heterologous SWEET gene known to be targeted by a TAL effector of the pathogen (thus such a SWEET gene is a susceptibility factor) in an immunocompromised nonhost plant can help ascertain if lack of adequate nutrient acquisition could also attribute to NHR against some phytobacteria.

3. CONCLUSIONS AND PERSPECTIVES

While genetic control of NHR in near‐hosts is relatively simple and thus genetically amenable, NHR in most cases is often multilayered and thus dissection of its genetic basis may require multiple rounds of forward and/or reverse genetic studies. The increased understanding of plant immune mechanisms using various host–pathogen systems, together with the development of advanced enabling technologies, including various OMICS and CRISPR technologies, should greatly facilitate future studies towards a complete dissection of the most canonical NHR such as rice's immunity to powdery mildew and rust fungi. As discussed in a recent review (Panstruga & Moscou, 2020), NHR is just a phenomenological term, and genetic/molecular mechanisms underlying NHR may vary depending on the context of the plant–pathogen interaction under study. However, the mechanisms of defence programmes induced by nonadapted pathogens during NHR most likely are the same as those occurring in host resistance that encompasses PTI and ETI. Therefore, while NHR studies may not lead to discoveries of entirely new immune mechanisms independent of PTI and ETI, elucidation of multilayered NHR may help reveal hidden components or regulatory nodes of PTI and ETI, thereby advancing our understanding of the plant immunity architecture in general and realizing the great potential of NHR for crop improvement by engineering effective immunity in susceptible crop plants.

CONFLICT OF INTEREST

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as potential conflicts of interest.

Wu, Y. , Sexton, W. , Yang, B. & Xiao, S. (2023) Genetic approaches to dissect plant nonhost resistance mechanisms. Molecular Plant Pathology, 24, 272–283. Available from: 10.1111/mpp.13290

DATA AVAILABILITY STATEMENT

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