The initial processes of extreme resistance are best-characterized in the potato Rx1-PVX system. Also referred to as Rx, Rx1 is a coiled-coil NLR protein that confers extreme resistance to the Potexvirus, PVX, which can cause yield losses of up to 20% and is a common problem for potato growers (Adams et al., 1985). The Rx1 gene is located on chromosome XII and has been introgressed from the wild species of potato, Solanum andigena, into some commercial potato cultivars (e.g., cv. Cara, Atlantic; Ritter et al., 1991; Bendahmane et al., 1997). The Rx1 gene shares high sequence complementarity and is functionally redundant to another NLR, Rx2, which is found on potato chromosome V and was introgressed into potato from Solanum acuale (Ritter et al., 1991; Bendahmane et al., 1997). Despite similarities between the two genes, Rx1 has received a much greater research focus than Rx2.

Previous research on NLR protein function indicates that most, if not all, NLRs rely on chaperone proteins for stability and proper function (Kadota et al., 2010). A triad of highly conserved NLR chaperone proteins, SUPRESSOR OF G2 ALLELE OF SKP1 (SGT1), HEAT SHOCK PROTEIN 90 (HSP90), and REQUIRED FOR MLA12 RESISTANCE (RAR1) are essential for disease resistance. Many studies involve the expression of Rx1 in the genetically tractable model plant, Nicotiana benthamiana (Nb). Silencing of NbSGT1 suppressed Rx1-mediated extreme resistance to PVX in Nb, and indicated that the extreme resistance conferred by Rx1 relies on proteins similar to non-extreme resistance conferring NLRs (Botër et al., 2007; Figure 1). Despite no known nuclear or cytoplasmic localization signals, Rx1 must be distributed in both the cytoplasm and nucleus for an extreme resistance immune response to occur (Slootweg et al., 2010; Tameling et al., 2010). The Rx1 coiled-coil domain is necessary for accumulation in the nucleus, while the LRR domain is required for cytoplasmic localization (Slootweg et al., 2018). Binding between Rx1 and N. benthamiana RAN GTPase ACTIVATING PROTEIN (NbRG2) results in the retention of Rx1 in the cytoplasm, which is required for their dual roles in pathogen sensing and gene activation (Rairdan and Moffett, 2006; Sacco et al., 2007; Tameling et al., 2010).

The function of NbRG2 in the NbRG2-Rx1 complex is not clear, although a few possible roles are apparent (Slootweg et al., 2010; Hao et al., 2013). RanGAP proteins, like NbRG2, facilitate the GTPase activity of Ran GTPase proteins, which aid in the transport of protein complexes into and out of the nucleus (Dasso, 2002). There are no known transport activities of RanGAP proteins themselves and NbRG2 hydrolazing activity is not required for a successful resistance response (Tameling and Baulcombe, 2007). Many NLR proteins have evolved to monitor host proteins that are vulnerable targets of pathogen effector proteins (Collier and Moffett, 2009). Slootweg et al. (2010) suggest that NbRG2 may act as a decoy target of the PVX coat protein, in which Rx1 would be “monitoring” NbRG2 until bound by the PVX coat protein, after which Rx1 could initiate an immune response (Slootweg et al., 2010). This claim is buttressed by the findings that Gpa2, an NLR protein that shares a high amount of amino acid conservation with Rx1, guards NbRG2 and initiates an immune response only when NbRG2 is bound by effector proteins secreted by nematodes during infection (Sacco et al., 2009). Interactions between NbRG2 and the PVX coat protein could produce a conformational change in NbRG2, which could in turn allow Rx1 to reach an activated state (Hao et al., 2013). Many NLR proteins work in pairs with other NLRs or helper NLRs to effectively transmit immune signals (Adachi et al., 2019a). A study by Wu et al. (2017) found that silencing of three genes that encode helper NLR proteins, NLR REQUIRED FOR CELL DEATH 1 (NRC1), NRC2, NRC3, disabled Rx1-conferred extreme resistance, but only if all three genes were silenced simultaneously (Wu et al., 2017). These results indicate levels of redundancy and possible robustness to interference by pathogens within plant immune signaling.

Binding between Rx1 and PVX coat protein occurs in the cytoplasm (Bendahmane et al., 1995). Recognition of PVX coat protein by Rx1 and subsequent binding causes the release of NbRG2 from Rx1, allowing Rx1 to translocate to the nucleus through a yet unknown process, as there are no detectable nuclear localization signals within Rx1 (Tameling and Baulcombe, 2007). However, nuclear localization signals are notoriously difficult to predict and can be hidden within secondary structure. Although many studies have focused on NLR functionality in the cytoplasm, there are other examples of NLR nuclear localization and function (Burch-Smith et al., 2007; Shen et al., 2007). The barley NLR, MLA10, interferes with repression of defense genes by binding to WRKY transcription factors in the nucleus (Chang et al., 2013). Experiments in which nuclear exclusion signals or nuclear localization signals were added to an Rx1-GFP fusion protein indicated that the DNA binding capabilities of Rx1 are contingent upon Rx1 recognition of the PVX coat protein in the cytoplasm, which is followed by movement of Rx1 to the nucleus. Resistance, but not cell death responses were compromised in experiments in which Rx1 was localized predominantly to the nucleus or the cytoplasm (Bendahmane et al., 1999; Knip et al., 2019; Richard et al., 2021). These results indicate that Rx1 must be activated (ATP-bound) in order to successfully bind DNA and likely must be able to translocate from the cytoplasm to the nucleus in order to initiate the extreme resistance response. Activation of Rx1 and other NLRs occurs after binding effector proteins, in this case the PVX coat protein. In vitro binding assays indicate that the DNA binding activity of Rx1 is inhibited, while in an inactivated state (ADP-bound), while the activated (ATP-bound) Rx1 can bind DNA (Fenyk et al., 2015). Further, recognition of the PVX coat protein likely results in a perturbed binding between the LRR and ARC2 (Apaf-1, R proteins, and CED-4) domains of Rx1, a process which may play a role in initiation of resistance pathways. Rairdan and Moffett (2006) suggest that the LRR domain can repeatedly dissociate and reassociate with the ARC2 domain after recognition of the PVX coat protein, and that this iterative process may serve to amplify the resistance signal and could play a key role in the extreme resistance response (Rairdan and Moffett, 2006).

Upon entering the nucleus, the activated nucleotide-binding domain of Rx1 allows for binding and melting of double-stranded DNA in a non-sequence specific manner, but with a higher affinity for DNA topologies similar to transcription start site bubbles (Finzi and Dunlap, 2010; Tang et al., 2011; Townsend et al., 2018). The DNA binding activities of Rx1 likely become sequence-specific when in complex with NbGLK1 (a Golden 2-like transcription factor), although this remains to be definitively proven (Townsend et al., 2018). The activation state of Rx1 likely determines the DNA binding activity of NbGLK1, as inactivated Rx1 in complex with NbGLK1 does not bind DNA in planta (Townsend et al., 2018). Golden 2-like transcription factors preferentially bind DNA sequences with GLK-like motifs and are known to regulate the transcription of genes involved in abscisic acid (ABA) signaling (Ahmad et al., 2019). Although the genes that NbGLK1 regulates in response to PVX infection are not known, GLK-like transcription factors play a role in resistance to cucumber mosaic virus and fungal pathogens in Arabidopsis (Savitch et al., 2007; Murmu et al., 2014; Han et al., 2016).

There are likely other proteins that interact with Rx1 and NbGLK1 in the nucleus. Sukarta et al. (2020) recently described a direct interaction between Rx1 and a DNA-binding bromodomain containing protein, NbDBCP (Sukarta et al., 2020). The precise role(s) of NbDBCP remains unclear, though it likely acts as a repressor of Rx1-mediated resistance signaling. Silencing of NbDBCP as well as co-expression of non-functional NbDBCP decreased PVX coat protein accumulation during Potato virus X infection in plants expressing Rx1, indicating that NbDBCP may negatively regulate extreme resistance responses. Binding of DNA by NbDBCP occurs in situ, but not when co-expressed with Rx1 or during PVX infection. Size exclusion chromatography results indicate that Rx1, NbDBCP, and NbGLK1 may form a transient complex; however, this idea remains theoretical and untested. These results conservatively indicate a negative regulatory role of NbDBCP on Rx1-mediated extreme resistance, although its exact role(s) require more research (Sukarta et al., 2020).

Intriguingly, overexpression of NbGLK1 in N. benthamiana confers immunity to PVX even in the absence of Rx1, and this immunity does not result in HR (Townsend et al., 2018). These results may signal that NbGLK plays a role in controlling gene expression that is important for extreme resistance, but that role is likely independent or upstream of HR/cell death. Additionally, NbDBCP overexpression, in the presence of Rx1 and during PVX infection, resulted in increased cell death. Expression of a non-functional NbDBCP resulted in decreased cell death, lending credence to the idea that (1) NbDBCP negatively regulates the extreme resistance pathway and (2) extreme resistance and HR/programmed cell death are largely separate or sequentially activated pathways (Sukarta et al., 2020). Similar separation of cell death and resistance has been reported in N. benthamiana plants expressing the barley NLR, MLA10, with a nuclear localization tag (Bai et al., 2012). The MLA10 gene provides resistance to the barley powdery mildew fungus, indicating that nuclear functions of NLRs may not be specific to virus resistance.

Extreme resistance conferred by Rx1 does not limit viral spread through the phloem. Grafting experiments revealed that PVX moved from a susceptible rootstock through the phloem of a middle, resistant scion and into another upper, susceptible scion and caused infection (Bendahmane et al., 1999). A reaction similar to Rx1-mediated extreme resistance can occur in protoplasts, while HR does not (Otsuki et al., 1972; Baulcombe et al., 1984; Goulden et al., 1993; Kohm et al., 1993; Bendahmane et al., 1995). These results may indicate that intercellular signaling components or cell wall components may not be necessary for Rx1-conferred extreme resistance responses, whereas they are for HR responses. It is likely that the mechanisms controlling extreme resistance occur rapidly in the cell, as extreme resistance prevents viral replication and spread beyond the initial point of inoculation.

Overexpression of Rx1 in N. benthamiana results in HR, regardless of whether its elicitor, the PVX coat protein, is present or not. Transformation of Rx1 under its native promoter into N. benthamiana and Nicotiana tabacum results in a typical, symptomless extreme resistance to PVX, indicating that functionality and possible downstream interacting elements are conserved between species (Bendahmane et al., 1999). Overexpression of PVX coat protein in N. benthamiana expressing Rx1 under its native promoter results in HR. Bendahmane et al. (1999) suggest that the continued production of the PVX coat protein after the initial recognition event and extreme resistance activation may signal to the cell that extreme resistance has been overcome and that further immune action may be warranted, hence the subsequent HR (Bendahmane et al., 1999). This hypothesis is supported by the finding that extreme resistance conferred by Rx1 is epistatic to HR, as plants expressing both Rx1 and N, an NLR that provides resistance with an HR phenotype to Tobacco mosaic virus, were resistant to Tobacco mosaic virus infection but did not display HR when the virus was engineered to express both the PVX coat protein and the protein elicitor of N during infection. The addition of a nuclear localization signal to NbRG2 caused Rx1 to accumulate almost solely in the nucleus and prevented HR from occurring, even when auto-active Rx1 mutants were overexpressed (Slootweg et al., 2010). Thus, it is likely that Rx1 must be located in the cytoplasm in order for HR and concurrent signaling to occur. This conclusion is congruent with cytosolic location of PVX replication and PVX coat protein detection by Rx1 (Slootweg et al., 2010).

The possible interconnectedness between HR and extreme resistance responses underscores the need for more sensitive resistance assays. To better understand extreme resistance, it is paramount that researchers try to replicate the native expression levels of Rx1 (and other genes that confer extreme resistance) when experimenting outside of its native potato system. Slootweg et al. (2010) expressed Rx1 from a vector with a second start codon inserted upstream and out of frame of the Rx1 start codon. The resulting “leaky” expression of Rx1 led to protein levels in N. benthamiana that were 5–10x lower than expression driven by a typical CaMV35S promoter, and a much more sensitive assay (Slootweg et al., 2010). Studies in A. thaliana also indicate the expression level of NLR proteins may in part determine the phenotype of the resistance response. For example, resistance to the yellow strain of Cucumber mosaic virus is conferred by RCY1, a coiled-coil NLR, in Arabidopsis ecotype C24. Resistance to Cucumber mosaic virus (Y) via RCY1 is normally accompanied by a hypersensitive response. Transgenic Arabidopsis lines that over-expressed RCY1 at high levels (i.e., ~100x greater than its native promoter) exhibited the extreme resistance phenotype. Transgenic plants that expressed RCY1 at moderately elevate levels (i.e., ~20x greater than native expression) exhibited enhanced resistance with very small areas of cell death (“micro-HR”; Sekine et al., 2008). The hypersensitive response was observed in transgenic lines with levels of RCY1 expression similar to native expression. None of the RCY1 transformed Arabidopsis lines became systemically infected with Cucumber mosaic virus (Y). These results indicate that RCY1 expression levels, at least in part, govern virus-resistance phenotypes, possibly by determining the type of the subsequent immune response.

Other publications studying the NLR, HRT, which confers resistance to Turnip crinkle virus, have noted similar results (Cooley et al., 2000). However, levels of resistance protein expression are likely not the only factor governing immune responses. Overexpression of the Turnip crinkle virus coat protein, which is the binding target of HRT, resulted in severe HR, similar to the reaction that occurs after overexpression of the PVX coat protein plants expressing Rx1 (Cooley et al., 2000). Expression and activity levels of NLR proteins in plants are regulated in many ways (e.g., transcriptionally, post-transcriptionally, post-translationally, etc.; Borrelli et al., 2018). It is possible that increased expression of HRT or RCY1 is sufficient to overcome some negative regulation, resulting in faster immune responses.

The mechanistic details of Rx1 conferred resistance restricting PVX viral replication and spread are not yet known, however, experiments indicate that the translational arrest of the PVX transcripts is likely a major component of these resistance processes (Richard et al., 2021). By employing an inducible effector protein expression system and nuclear- and cytoplasm-localized Rx1 expression, Richard et al. (2021) demonstrate that Rx1-conferred extreme resistance likely relies on PVX transcript-specific translational arrest and that this response occurs within a few hours after infection (Meteignier et al., 2016; Richard et al., 2021). These data also demonstrate that nuclear- or cytoplasm-localized Rx1 expressed individually or together, results in HR or trailing necrosis (i.e., HR that trails viral spread throughout the plant) after 4 hours of the induction of PVX coat protein transcription, but does not induce extreme resistance. These results further support that upon recognition of the PVX coat protein in the cytoplasm, Rx1 must translocate to the nucleus in order to initiate the extreme resistance response. Translational arrest is a common host antiviral strategy (Machado et al., 2017). Rx1-expressing, PVX-infected potato protoplasts did not support replication of either Tobacco mosaic virus or Cauliflower mosaic virus, indicating that the Rx1-mediated antiviral response was a general antiviral response (Goulden et al., 1993; Bendahmane et al., 1995). Further, Rx1 may not be unique in this regard, as recognition of the resistance elicitor, Tobacco mosaic virus p50, by the tobacco NLR gene, N, can initiate an immune response that prevents translation of PVX transcripts in N. benthamiana, but only in the presence of RNA containing the PVX coat protein coding sequence (Bhattacharjee et al., 2009).

These results collectively suggest that a conserved characteristic of viral RNAs, possibly secondary structure, may be specifically targeted by NLR-mediated translational inhibition responses and that this mechanism may play a key role in the extreme resistance response. Although the factor(s) governing translational arrest are not known, it is interesting to note that virus-induced gene silencing of ARGONAUTE 4 (AGO4) disabled symptomless resistance responses, and in turn allowed systemic PVX infection in N. benthamiana (Bhattacharjee et al., 2009). Similarly, in N. benthamiana plants in which the RNA interference (RNAi) suppressor proteins from Beet western yellows virus and Turnip crinkle virus, P0 and P38, were expressed, the antiviral response was also disabled. The Turnip crinkle virus P0 protein targets and induces degradation of Argonaute proteins (Baumberger et al., 2007; Bortolamiol et al., 2007). The expression of other suppressors of RNAi, including potyviral Helper-component protease (HC-Pro), did not prevent an antiviral response from occurring, although HC-Pro disables RNAi through sequestration of virus-derived small RNAs, not through the degradation of Argonaute (Mallory et al., 2002). Another RNA virus, Tobacco rattle virus (TRV), expresses a suppressor of RNAi silencing protein, 16k, which binds AGO4 and ago4 mutant N. benthamiana plants are more susceptible to infection (Ma et al., 2015; Fernández-Calvino et al., 2016).

ARGONAUTE 4 is well known for its roles in transcriptional gene silencing and the RNA-directed DNA methylation pathway, as well as methylation-based antiviral defense against plant viruses with DNA genomes (Zilberman et al., 2003; Li et al., 2006; Raja et al., 2008, 2014; Gao et al., 2010; Greenberg et al., 2011; Dowen et al., 2012; Wierzbicki et al., 2012; Ye et al., 2012). These well-characterized roles of AGO4 all occur in the nucleus. Interestingly, cytoplasm-localized AGO4 is necessary for resistance to the potexvirus virus, Plantago asiatica mosaic virus, in Arabidopsis. This resistance does not involve other protein components of the RNA-directed DNA methylation pathway (e.g., DICER-LIKE 3, RNA POLYMERASE IV, and RNA POLYMERASE 5), indicating that AGO4 antiviral activity in this case is likely independent of its DNA methylation activity (Brosseau et al., 2016). The importance of AGO-4 non-methylation-based defense is not limited to antiviral responses, as silencing of AGO4 in Arabidopsis plants without functional RNA-directed DNA methylation pathways increased susceptibility to Pseudomonas syringae (Agorio and Vera, 2007).

Research on Rx1-conferred extreme resistance has illuminated the potential nuclear functions of Rx1 and laid a framework for future studies. In particular, gaining an understanding of the DNA sequences targeted by the Rx1-NbGLK1 complex and the possible transcriptional changes that occur after recognition of the PVX coat protein will aid in the identification of other genes and mechanisms involved in Rx1-conferred extreme resistance. Future experiments employing RNA sequencing, chromatin-immunoprecipitation sequencing, and translatome analysis would increase understanding of the regulation of this defense system. Additional studies are required to determine the role(s) of bromodomain containing proteins, to identify the DNA sequences that are targeted by the Rx1-GLK complex, and if targeted gene(s) are responsible for the next stages of the Rx1 conferred extreme resistance response. Further, studies that dissect the antiviral translational repression response and possible antiviral roles of AGO4 would provide a greater understanding of NLR-mediated immunity.

The Ry genes in potato (e.g., Ry_sto, Ry_fsto, and Ry_adg) provide resistance to particular strains of PVX and PVY. The Ry_sto gene (Resistance to PVY from S. stoloniferum), which conveys resistance to a broad spectrum of strains of PVY and Potato virus A (PVA) in potato and tobacco, is the only one of the genes controlling extreme resistance that has been isolated from the Ry loci (Cockerham, 1970; Barker, 1996; Grech-Baran et al., 2020). Global potato production is reliant on pathogen-free seed tubers, which are vulnerable to generational buildup and spread of pathogens, particularly viruses. Various PVY strains (including PVY^NTN and PVY^N-Wi) are the most economically harmful viral pathogens involved in potato production and genetic resistance to PVY is a major focus of breeding programs (Karasev and Gray, 2013). Wild potato varieties and landraces are sources of PVY-specific NLR resistance genes that can be introgressed into commercial potato cultivars. Loci conferring extreme resistance to PVY have been mapped in Solanum chacoense (Ry_chc), S. tuberosum group Andigena (Ry_adg), and S. stoloniferum (Ry _sto; Herrera et al., 2018). Grech-Baran et al. (2020) employed resistance enrichment sequencing (RenSeq) to isolate the gene conferring Ry_sto-mediated extreme resistance from the commercial potato cultivar, Alicja. Introgressed from S. stoloniferum, Ry_sto is a Toll-interleukin receptor (TIR) NLR protein, similar to other potato virus resistance genes (e.g., N, Pvr4, Y-1, etc.). The broad-spectrum resistance conferred by Ry_sto lends it an attractive trait for breeders of Solanaceous plants. The Ry_sto gene is present in various commercial potato cultivars, including American cultivars Payette Russet and Castle Russet and European cultivars Alicja, White Lady, and Pirola. The Ry_sto protein either directly or indirectly recognizes or binds the coat protein of PVY and PVA to elicit the extreme resistance response (Grech-Baran et al., 2020).

The Ry_sto gene has been cloned and expressed in PVY-susceptible Solanaceous plants. Challenge of transgenic plants expressing Ry_sto under its native promoter with PVY usually results in an extreme resistance response (i.e., no infection, no symptoms), but can cause veinal necrosis or HR in response to some isolates of PVY^O (Hinrichs-Berger et al., 1999). Co-expression of Ry_sto and the PVY coat protein under control of a CaMV35S promoter results in HR in potato and N. benthamiana. Expression of Ry_sto in tobacco and subsequent challenge with PVY produced some localized necrosis in inoculated leaves. Grech-Baran et al. (2020) suggest that establishment of either extreme resistance or HR depends on at least three variables: expression level of the resistance gene; abundance of the cognate effector protein; and the genetic background of the host. Extreme resistance conferred by Ry_sto relies on at least two other genes for successful immune activation: the lipase-like ENHANCED DISEASE SUSCEPTIBILTY (EDS1) and the CC-NLR, N REQUIREMENT GENE 1 (NRG1). These dependencies corroborate studies on other TIR-NLRs (Grech-Baran et al., 2020).

Similar to the Rx1-PVX system, the genes controlling Ry_sto-mediated extreme resistance downstream of virus recognition are not known, although some results have hinted at the involvement of proteins that interact with plasmodesmata. β-1,3-glucanase proteins aid in plant virus infection, likely by hydrolyzing callose and increasing the size exclusion limit of plasmodesmata, thus allowing for cell to cell spread of the virus (Iglesias and Meins, 2000). Callose is a polysaccharide that influences the size exclusion limit of the plasmodesmata and also serves as a deposition site for defense compounds (Zavaliev et al., 2011). Overexpression experiments of β-1,3-glucanase (class III) proteins were carried out in potato cultivars Santé (which displays extreme resistance to PVY) and Désirée (PVY-susceptible) within the context of PVY^NTN infection. Resistant Santé plants overexpressing β-1,3-glucanase exhibited modest, transient increases in PVY^NTN that dissipated within days of infection and the virus did not spread beyond the inoculated leaf. Susceptible Désirée plants that overexpressed β-1,3-glucanase may have exhibited slightly faster systemic infection of PVY^NTN, although a relatively small sample size precluded more definitive conclusions (Dobnik et al., 2013).

Callose deposition is also targeted by PVY during infection, as PVY-encoded HC-Pro suppresses callose deposition during PVY^O infection through an unknown mechanism (Chowdhury et al., 2020). Callose deposition also occurs in cells surrounding HR activity during PVY infection and those cells can harbor viable PVY, thus further indicating that cell death is not the primary driver of resistance during HR and that callose deposition alone is not effective at arresting viral spread (Lukan et al., 2018). Cells undergoing HR/cell death processes may release signals to surrounding cells to initiate immune responses or that may act as defense compounds themselves (Lamb and Dixon, 1997). These signals may include reactive oxygen species (ROS), which are a common component of plant immune responses (Qi et al., 2017). Whether or not reactive oxygen signaling acts a component of extreme resistance is not known, although there are examples of symptomless resistance to Tobacco mosaic virus in tobacco plants induced by application of ROS (Kuenstler et al., 2016). The timing of foliar treatments was key to inducing resistance to Tobacco mosaic virus, as resistance did not occur in plants that were treated with reactive oxygen species 3 days after virus inoculation, but an HR-like reaction and cell death did.

The NLR resistance gene, Ny-1, also provides resistance to PVY, but the immune response is accompanied by HR. Extreme resistance conferred by Ry_sto is epistatic to Ny-1-mediated HR, as plants expressing both Ny-1 and Ry_sto exhibit an extreme resistance phenotype but lack HR when challenged with a PVY^NTN isolate that is recognized by both Ry_sto and Ny-1 (Grech-Baran et al., 2020). For many TIR-NLR proteins, including Ny-1, resistance breaks down at high or low temperatures, while Ry_sto function is not limited by high ambient temperatures. Modulation of defense responses by temperature is likely controlled by NLR proteins, as point mutations in NLR genes can decrease nuclear accumulation of NLR proteins at higher temperatures and reduce NLR immune function (Zhu et al., 2010). At higher temperatures plants may preferentially activate pattern-triggered immunity (PTI), rather than NLR-dependent immunity (Cheng et al., 2013). Since NLR proteins confer resistance to plant pathogens that disable host pattern-triggered immune responses, a greater understanding of NLR thermosensitivity is needed as global temperatures continue to rise (Trebicki, 2020).

The expression levels of both the genes conferring extreme resistance and the viral elicitor protein seems to be a critical factor if HR or extreme resistance occurs in both Rx1-PVX and Ry_sto-PVY systems. The cellular distribution of Ry_sto both before and during resistance responses is not known but should be the focus of further research given the potential nuclear functions of Rx1. The advent and expanded use of RenSeq, which allows for the expedited identification of NLR genes conferring specific resistance phenotypes, and CRISPR technologies, which allow for precise genome editing, should facilitate faster identification and breeding of resistance genes (Witek et al., 2016; Zhang et al., 2019). As extreme resistance is largely characterized by a lack of infection symptoms, there may be a pool of genes that confer extreme resistance to viruses that are yet to be discovered within landraces or wild Solanaceous species. For example, the PVR4 gene, which encodes an NLR protein that originated in a landrace of hot pepper (Capsicum annum), confers extreme resistance to multiple potyviruses, including many PVY strains, Pepper mottle virus, and Pepper severe mosaic virus (Kim et al., 2015). The PVR4 protein recognizes the potyviral RNA-dependent RNA polymerase (NIb) to elicit the extreme resistance response (Kim et al., 2015). Global potato production relies on labor-intensive seed tuber certification programs to prevent pathogen accumulation, particularly viruses, with a large focus on PVY strains. Given that many NLR genes can be shuttled between Solanaceous species without a loss of function, transferring multiple, broad-spectrum NLR genes that target different potyviral components (i.e., Ry_sto, PVR4) to potato could provide durable and sustainable immunity to potyviruses (Rivero et al., 2012; Zhu et al., 2012; Jo et al., 2014).

Potyviruses are not the only problematic viruses of potato production. TRV, of the genus Tobravirus, causes necrosis in potato tubers, a disease commonly referred to as corky ringspot disease or “spraing.” Corky ringspot disease can render tubers unmarketable. Tobacco rattle virus is vectored by various species of nematodes and can infect many different plant species. Although cultivars that display symptoms of TRV infection in tubers are often labeled as susceptible, the ringed, necrotic tuber tissue characteristic of corky ringspot disease are likely the result of HR in response to viral infection in the tuber (Xenophontos et al., 1998; Sahi et al., 2016). Some frequently grown cultivars are tolerant to TRV infection and can contain high levels of virus throughout the plant and yet remain largely asymptomatic, although these tolerant infections can have negative effects on yield and tuber size and also serve as inoculum sources (Dale et al., 2000; Brown et al., 2009). The potato cultivar, Saturna, displays an extreme resistance response in leaves when inoculated with the TRV isolate, PpK20, and no symptoms or detection of TRV in tubers. A yet undescribed protein in Saturna recognizes the TRV movement protein, 29K, to initiate an extreme resistance response (Ghazala and Varrelmann, 2007). Similar to potato cultivars that display extreme resistance to PVY and PVX, overexpression of the elicitor protein (29k, in this case) leads to HR in the infected leaf.

Corky ringspot disease is a significant problem for growers in Europe, while TRV is a small but growing problem in the United States. Once TRV/corky ringspot disease occurs in a field it can be difficult to eliminate completely. Soil fumigation can reduce the presence of stubby root nematode vectors, but often does not completely eliminate TRV presence. Stubby root nematodes can feed on a wide variety of plants (i.e., potato, barley, corn, peas, brassicas, various common weeds, and others), often rendering crop rotations in infested fields non-effective at lowering vector pressure. Given that corky ringspot disease/spraing is caused by an HR-like immune response in tubers, then the largely symptomless immune response provided by extreme resistance would be a valuable asset to growers. Further, isolation of the gene(s) that confer extreme resistance to TRV in potato would be beneficial for breeding resistant cultivars.

The Rsv1 locus encodes multiple NLR genes, is highly complex, and is mapped to soybean chromosome 13 (Widyasari et al., 2020). The extreme resistance phenotype conferred by the Rsv1 locus is contingent upon the strain of the infecting virus and the Rsv1 allele present. There are 10 identified alleles of the Rsv1 locus, which are associated with strain-specific resistance to SMV (Klepadlo et al., 2017; Wu et al., 2017). Various resistance phenotypes occur in plants containing the Rsv1 locus that are infected with SMV strains G1–G7. The dominant Rsv1 allele, which shares its name with the locus itself, confers extreme resistance to SMV strains G1–G6, but does not provide resistance to isolates of strain SMV-G7 and experimentally evolved SMV-G7d (Hajimorad et al., 2006).

Understanding of Rsv1-SMV interactions is limited because the gene(s) controlling Rsv1-conferred extreme resistance have not yet been identified. Gene silencing experiments concurrently targeting the expression of three non-Toll interleukin receptor NLR genes from the Rsv1 allele of the Rsv1 locus resulted in viral foci formation in resistant plants (cultivar L78–379), similar to that seen in susceptible plants (a near isogenic line of L78–379; Zhang et al., 2012). It is likely that at least two of those three targeted genes are necessary for Rsv1-conferred extreme resistance (Wen et al., 2013). Individual silencing of each gene was not possible because of high sequence similarity between the three (Zhang et al., 2012). Extreme resistance conferred by the Rsv1 locus may be dependent on host recognition of two viral proteins, as mutations in both the HC-Pro and Protein 3 (P3) protein coding regions of the SMV genome are needed to break Rsv1-conferred extreme resistance (Hajimorad et al., 2005, 2006, 2008; Zhang et al., 2009; Wen et al., 2013).

Studies that utilized recombinant inbred soybean lines derived from the Rsv1-allele containing PI969893 line have enhanced understanding of Rsv1-conferred extreme resistance. One recombinant inbred line, L800, contains one NLR gene from the Rsv1 locus, denoted as 3gG2. Another recombinant inbred line, L943, contains five NLRs from the Rsv1 locus, but does not contain 3gG2. Interestingly, both lines are resistant to SMV-N, but L943 recognizes the HC-Pro from SMV-N to induce resistance, while L800 recognizes P3 from SMV-N to induce resistance, suggesting that the Rsv1 locus may contain at least two resistance genes that recognize separate SMV proteins to induce resistance (Wen et al., 2013). It is interesting that line L943, which contains five NLRs from the Rsv1 locus, allowed limited virus replication, as a few viral foci were evident by GUS staining after inoculation with SMV-N expressing GUS, but virus did not spread beyond the inoculated leaf and the foci did not grow. In contrast, line L800, which only contains one NLR (3gG2) displayed no GUS expression when inoculated with SMV-N expressing GUS (Wen et al., 2013). These small GUS foci are similar to those seen when comparing Rsv1 and Rsv3 extreme resistance in Zabala et al. (2009) and Zhang et al. (2009). These results could indicate that separate mechanisms may inhibit viral replication and viral spread and that both may be induced by the extreme resistance response.

Mutations in the SMV-N HC-Pro and P3 coding regions, which are recognized by Rsv1 proteins, enabled virus replication in L943 and L800 soybean lines, but not in P196983, which contains the entire Rsv1 locus. An additional mutation to SMV-N HC-Pro resulted in productive infections in L800, L943, and PI96983 soybean lines, further indicating that the Rsv1 locus likely contains multiple NLR or other genes that induce extreme resistance SMV. This idea is also supported by data indicating that SMV-N more easily evolves to evade variants that infect the single NLR (i.e., 3gG2) containing L800 soybean line than the L943 soybean line that contains five NLRs from Rsv1 or the full Rsv1 locus-containing PI96983 line (Hajimorad et al., 2003; Wen et al., 2013).

The HC-Pro and P3 genes are next to each other in the potyviral genome and on the resulting polyprotein before self-cleavage. It is not known if the protein(s) from the Rsv1 locus recognize the SMV polyprotein or mature HC-Pro or P3 (Hajimorad et al., 2008; Wen et al., 2013). In another example of extreme resistance outside of potato or soybean, the cowpea cultivar, Arlington, displays extreme resistance to Cowpea mosaic virus by recognizing the enzymatically active 24K-protease as it cleaves the polyproteins of the Cowpea mosaic virus’ segmented genomes (Fan et al., 2011). It is also possible that closely related NLR genes from the Rsv1 locus guard host proteins that are targeted by HC-Pro/P3 during the early stages of SMV infection (Hajimorad et al., 2008). It is also plausible that Rsv1 contains multiple genes that confer variable levels of SMV resistance. For example, extreme resistance in the 3gG2 containing L800 soybean line could involve SMV-N P3 recognition, while other genes within the Rsv1 locus could confer less effective resistance phenotypes (i.e., limited viral replication, but no spread) through SMV-N HC-Pro recognition. Further, line PI96983 could contain a yet unidentified NLR that recognizes a separate region of HC-Pro to induce extreme resistance.

Additional studies provide some indirect support that NLR proteins are likely involved in Rsv1-conferred extreme resistance to SMV, as silencing of genes that interact with NLR proteins resulted in increased virus load in resistant plants. Heat shock proteins often serve as molecular chaperones and HEAT SHOCK PROTEIN 90 (HSP90) acts as a chaperone to NLR proteins in plants and animals (Kadota et al., 2010). Virus induced gene silencing (VIGS) of HSP90 resulted in SMV infection foci in the leaves of resistant cultivars, resulting in a phenotype similar to SMV infection in a susceptible cultivar (Zhang et al., 2012). Two other genes, RAR1 and SGT1, serve as co-chaperones to HSP90 to stabilize NLR proteins. Silencing of RAR1 and SGT1 in two independent publications provided conflicting results as to if either are involved in extreme resistance (Fu et al., 2009; Zhang et al., 2012). These differences are likely explained by both differences in experimental design and in that gene silencing assays rarely result in a complete loss of target gene expression.

Experimental VIGS was also used to target a suite of other defense-related genes, including soybean homologs of EDS1, PHYTOALEXIN DEFICIENT 4 (PAD4), ENHANCED DISEASE RESISTANCE 1 (EDR1), and JASMONIC ACID-AMINO SYNTHETASE 1 (JAR1). Reducing expression of the aforementioned genes resulted in SMV infection foci in the inoculated leaves of extreme resistant soybean cultivar L78–379 (Zhang et al., 2012). These infections were phenotypically similar to infections in SMV-infected leaves of susceptible cultivars, thus indicating that the silenced genes are likely components of the extreme resistance defense response. The EDS1 protein family includes EDS1, PAD4, and SENESCENCE-ASSOCIATED GENE 101 (SAG101). Heterodimers between EDS1/PAD4 and EDS1/PAD4/SAG101 are required for effector triggered immunity (ETI) in most seed plants. These heterodimers act downstream of pathogen recognition but upstream of transcription of defense genes. Recent research indicates that EDS1/PAD4 promote salicylic acid (SA) biosynthesis via the isochorismate pathway, but also control and preserve SA signaling through an alternative, parallel pathway. A third, separate salicylic acid signaling pathway relies on MAPK signaling (Cui et al., 2017). Salicylic acid is a primary signaling component of pattern-triggered and effector-triggered plant immune responses to many biotrophic pathogens and is therefore a primary target of plant pathogen interference. Downstream signaling and transcriptional reprogramming during ETI is controlled in part by the EDS1/PAD4 complex, which is in turn negatively regulated by the MAPKK kinase, EDR1 (Neubauer et al., 2020). Salicylic acid-mediated disease resistance is negatively regulated by EDR1 and Arabidopsis EDR1 mutants are sensitive to ABA. These results are surprising because reducing expression of EDR1 would not be expected to affect the resistance phenotype. Jasmonic acid (JA) signaling, which modulates defense responses to herbivory and often acts antagonistically to the salicylic acid pathway, is controlled in part by JAR1 (Suza and Staswick, 2008).

Other genes targeted by VIGS within the context of Rsv1-conferred extreme resistance to SMV infection of soybean plants include WRKY transcription factors. WRKY proteins are among the largest transcription factor families in plants and largely regulate gene expression in response to abiotic and biotic stressors. A large scale VIGS study targeting 62 separate WRKY transcription factors revealed two genes, GmWRKY30 and GmWRKY6, that compromised Rsv1-mediated resistance in soybeans that had been challenged with SMV (Zhang et al., 2012). Arabidopsis WRKY6 is induced upon infection with a variety of viruses, which may suggest a conserved role for WRKY6 within the context of virus infection across plant species (Whitham et al., 2003). GmWRKY30 shares sequence similarity with Arabidopsis WRKY3, which is induced by pathogen infection and salicylic acid treatment in Arabidopsis (Liu et al., 2004; Lai et al., 2008; Pandey and Somssich, 2009). Further research is needed to understand which genes are regulated by GmWRKY30 and GmWRKY6 in response to viral infection and if similar antiviral roles are played by WRKY homologs in other plant species.

The factors controlling the relationship(s) between resistance, SMV strain, and Rsv1 locus are not well understood and raise some interesting questions. Namely, why do some Rsv1 alleles provide extreme resistance to particular SMV strains and others do not? For example, the Rsv1 allele, Rsv1-m, provides extreme resistance to SMV strains G1, G4, and G5, but exhibits systemic necrosis when infected with G2, G3, G6, and G7 (Chen et al., 1991). The Rsv1 allele confers resistance to SMV strains G1–G6, but systemic necrosis occurs in response to SMV-G7 infection. These differences may be attributed to allelic differences in elicitor binding/recognition efficiencies but may also be due to non-NLR host factors (Yuan et al., 2020). The cause(s) of systemic necrosis is not well understood but may be due to the delayed activation of the immune response/HR (Hajimorad and Hill, 2001; Hajimorad et al., 2003; Seo et al., 2009). In studies by Chen et al. (2017) and Wu et al. (2017), Rsv-1-containing soybean plants in which EUKARYOTIC TRANSLATION INITIATION FACTOR 5A was silenced exhibited increased virus accumulation and less necrosis, indicating a possible role for EUKARYOTIC TRANSLATION INITIATION FACTOR 5A in defense responses.

These plants also exhibited lower expression of other defense genes and genes involved in ROS signaling. Genetic variation in the RNA interference machinery between species and cultivars can result in differential susceptibility to virus infection. For example, the ARGONAUTE 2 (AGO2) gene in A. thaliana limits PVX infection, but N. benthamiana AGO2 does not. Further, AGO2 exhibits a high incidence of polymorphism between A. thaliana accessions, some of which affect antiviral activity. The AGO2 sequences contain signatures of selective pressure, possibly due to co-evolution with viruses (Brosseau et al., 2020). Variation among non-NLR host factors may explain some of the differences observed between resistance phenotypes of different species or cultivars (i.e., plants that develop viral foci in inoculated leaves but not systemic infection) but more research is needed to understand the impact of these differences on antiviral defense in plants.

Future studies involving the cloning and testing of individual genes within the Rsv1 locus and identification of possible plant protein binding partners and other host factors involved in antiviral defense could further our understanding of Rsv1-conferred extreme resistance. Isolation of genes within the Rsv1 locus could also provide access to a suite of genes that confer varying levels of resistance, which may be experimentally valuable.

The gene GLYMA.14g204700 (referred to hereafter as simply “Rsv3”), which encodes a coiled-coil NLR, is likely the gene responsible for Rsv3-mediated extreme resistance in the L29 cultivar (Tran et al., 2018). The soybean cultivar, L29, exhibits extreme resistance to isolates of SMV strains G5, G6, G7, and G5H via the Rsv3 locus but is susceptible to the SMV-G7H isolate. The cylindrical inclusion protein of SMV is indirectly recognized by Rsv3 to initiate an immune response. Virus strain SMV-G7H escapes Rsv3-mediated host detection via amino acid substitutions in the cylindrical inclusion protein (Seo et al., 2009).

Transcriptomic responses of the L29 soybean cultivar to SMV strains have been the focus of recent studies investigating the mechanism of extreme resistance to SMV (Seo et al., 2014; Alazem et al., 2018, 2019). Alazem et al. (2018) postulate that the extreme resistance response may inhibit SMV replication and spread in three successive steps: (1) virus recognition by an NLR, resulting in rapid callose deposition at the plasmodesma of infected cells; (2) dsRNA detection, which induces viral genome destruction via RNAi; and (3) clearance of remaining viral proteins via the transient induction of autophagocytosis. Seo et al. (2014) compared the transcriptomic responses of L29 soybean plants when infected with either a virulent SMV strain (SMV-G7H) or an avirulent strain (SMV-G5H) at 8, 24, and 54 hours post infection (hpi). Analysis of differential gene expression revealed that genes encoding Type 2C protein phosphatases were among the most differentially expressed in L29 soybean plants that exhibited extreme resistance to avirulent SMV infection. Type 2C protein phosphatases are a large class of serine/threonine phosphatases that are key regulators of the ABA signaling network in plants (Fuchs et al., 2013). Abscisic acid signaling plays key roles in developmental regulation, stress responses, and likely defense responses (Flors et al., 2005; Fan et al., 2009; Alazem et al., 2017; Alazem and Lin, 2017; Xie et al., 2018; Pasin et al., 2020). L29 plants that overexpressed PROTEIN PHOSPHATASE2C 3A (GmPP2C3a) were not systemically infected when inoculated with virulent SMV, whereas all non-transgenic plants were systemically infected. Further analyses indicated that Type 2C protein phosphatases are likely regulators of the Rsv3-mediated extreme resistance response (Seo et al., 2014).

The aforementioned transcriptomic data was further analyzed in Alazem et al. (2018). Genes involved in the ABA pathway exhibited increased expression at 8 and 24 hpi in plants infected with avirulent SMV, but expression of these genes was not increased by virulent SMV infection. The induction of genes involved in ABA signaling in plants infected with avirulent SMV dissipated by 54 hpi. Topical application of ABA to L29 soybean leaves 24 h before infection with virulent SMV reduced virus accumulation by about 50% compared to non-treated plants, although virus replication was not completely eliminated. The specific mechanism(s) of ABA mediated virus reduction are an active area of research and are not well understood. It is likely that ABA signaling works in concert with other defense pathways including RNA interference and PTI. Foliar ABA application of N. benthamiana plants induced the expression of genes involved in RNAi in response to Bamboo mosaic virus infection, specifically the argonaute protein encoding genes, AGO2 and AGO3 (Alazem et al., 2017). There are similar reports of increased expression of RNAi pathway components in Arabidopsis and soybean (Chen et al., 2013; Alazem et al., 2014; Alazem and Lin, 2017).

One effect of greater cellular ABA concentrations is increased callose deposition at plasmodesmata (Nishimura et al., 2003; Flors et al., 2005; Li et al., 2012). Abscisic acid negatively regulates β-1-3-glucanase expression, which encodes for proteins that break down callose tissue. The number of β-1-3-glucanase transcripts was reduced at all three time points in plants infected with avirulent SMV. Increased callose deposition at inoculation sites occurred in L29 plants infected with avirulent SMV and in L29 plants overexpressing GmPP2C3a and infected with virulent SMV. Similar increases in callose deposition were not observed in non-transgenic plants infected with virulent SMV. Foliar treatments of L29 leaves with the callose synthesis inhibitor, 2-deoxy-D-Glucose (DDG), resulted in increased avirulent SMV abundance in treated leaves. These experiments elegantly illustrate the importance and interplay of ABA signaling, GmPP2C3a expression, and callose deposition to the extreme resistance response as conferred by Rsv3 in soybean.

Cellular concentrations of ABA increased dramatically in L29 plants that exhibited extreme resistance, but the abundance of another plant defense signaling molecule, SA, did not change (Seo et al., 2014). The importance of SA signaling to antiviral defense in plants is well documented (Vleesschauwer et al., 2014). Increased SA accumulation occurred in L29 plants infected with virulent SMV but remained unchanged in plants infected with avirulent SMV, possibly suggesting that the antiviral role(s) of SA are not activated until after extreme resistance mechanisms have been broken. The ABA and SA pathways seem to interact in dueling antagonistic manner (i.e., high cellular ABA concentrations reduce SA biosynthesis and high SA concentrations inhibit ABA signaling; Nishimura et al., 2003; Zabala et al., 2009; Manohar et al., 2017). Interestingly, type 2C protein phosphatases (PP2Cs) bind both ABA and SA and may be important in modulating ABA and SA signaling (Manohar et al., 2017). ABA signaling is negatively regulated by Type 2C protein phosphatases which, in the absence of ABA, dephosphorylate ABA signaling kinases. Increased ABA concentration enhances binding between PP2Cs and PYR1-like regulatory elements, which inhibit PP2C dephosphorylation activity, resulting in the autophosphorylation of ABA signaling kinases and the expression of ABA responsive genes (Manohar et al., 2017).

Many SA-binding proteins have been documented. It is likely that a multitude of proteins are directly involved in regulating ABA and SA signaling. Staining leaves with DAB (3,3'-diaminobenzidine), which indicates reactive oxygen species presence, revealed that the Rsv3-mediated extreme resistance response does not rely on reactive oxygen species production. This result is contradictory to evidence indicating that reactive oxygen production is an important component of other plant immune responses, including extreme resistance. Therefore, it supports the idea that Rsv3-mediated extreme resistance differs mechanistically from Rsv1-mediated extreme resistance (Zhang et al., 2009). JA is also an important defense signaling molecule during plant-pathogen interactions, particularly during interactions with herbivorous insects (Suza and Staswick, 2008). The relative importance of JA signaling during viral infection remains unclear. Genes involved in the JA pathway exhibited either no change or decreases in expression in reactions with avirulent SMV but saw increased expression during virulent SMV infection at all timepoints. These data suggest that the induction of JA signaling may be important for establishing infection or may act as another layer of defense signaling. Another recent study by Alazem et al. (2019) illustrated that the effects of ABA treatment can be strain-dependent in cultivars lacking the Rsv3 locus. Topical treatment of leaves with ABA reduced the severity of virulent SMV infection but promoted avirulent SMV infection in an Rsv3-lacking soybean cultivar. The presence of Rsv3 (along with other proteins involved in its network, potentially) is necessary to fully recapitulate the extreme resistance response. It is plausible that avirulent SMV (strain G5) has evolved to manipulate components of the ABA pathway for its own benefit, but more research is needed to understand these differences (Alazem et al., 2019).

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