During the past decades it has become clear that plants have a unique and complex defense system that consists of several layers, which enables them to avoid, suppress, or actively defend against pathogens from all kingdoms like fungi, bacteria, nematodes, and viruses. Of all plant viruses known, only a few cause serious diseases and, if so, mostly limited to a very small number of crops. In general, most viruses have a limited (natural) host range and the number of so-called non-hosts exceeds those of hosts. In those plants that are hosts, viruses encounter different mechanisms of defense. Some act general against all viruses and this response is part of the innate immune system, while others are virus-specific and involve resistance genes. Triggering of the latter simultaneously mediates a rapid necrosis at the site of virus entry and prevents further spread of the virus throughout the host. In several cases resistance genes do not confer absolute resistance and low levels of virus replication can still be observed. In those cases the genes are referred to as partial resistance genes or tolerance genes.

While throughout the years reviews on resistance genes have appeared with regular intervals, these mostly had their main focus on fungal and bacterial resistance genes, primarily due to the large amount of data available. This review aims to present an overview on the current status of resistance genes against plant viruses, with emphasis on single dominant resistance genes. The very basis of plant pathogens (among others plant viruses) not being able to infect all plants is due to a mechanism called non-host resistance (NHR) (For an extensive review on this, see Uma et al., 2011). NHR holds for all plant pathogens and is a generic, nonspecific resistance that can be divided into two main types, distinguished by the mechanism and mode of recognition (Mysore and Ryu, 2004). Type 1 is the most pre-dominant type of NHR and presents a basic defense mechanism that prevents pathogen invasion, e.g., thickening of the cell-wall, secondary metabolite production, etc. This type of resistance usually is symptomless. In contrast, type 2 NHR is associated with induction of necrosis at the site of infection, and is induced when pathogens overcome type 1 resistance. Here, the pathogen is recognized through specific structures or proteins that are associated with the pathogen. The recognition of these structures/proteins, so called microbe associated molecular patterns (MAMPs) or PAMPS (Pathogen), takes place by pattern recognition receptors (PRRs) on plant plasma membranes. These PRRs recognize conserved structures of pathogens, like flaggelin from the flagella of bacteria or chitin from the cell wall of fungi, and induce a so called PAMP triggered immunity (PTI) response (Jones and Dangl, 2006). Since plant viruses need to overcome the physical barrier of a cell wall, they enter their host cells either via mechanical inoculation or the infection is mediated by vectors like insects, nematodes, or even fungi. Direct recognition of viruses probably does not occur in the apoplast. However, a study recently reported on the possible involvement of (intracellular) receptor-like kinases (RLKs), of the like that are involved in PAMP recognition by PRRs, in plant-virus interactions (Kørner et al., 2013).

One of the first innate immune responses all plant viruses encounter when invading a host consists of antiviral RNA silencing [also called RNA interference (RNAi) and in the very early days post-transcriptional gene silencing (PTSG)]. RNA silencing is a host response triggered by double stranded (ds)RNA. These molecules thus act as a MAMP/PAMP and in which RNAi can be regarded as PTI. The main difference with pathogens such as fungi and bacteria is that recognition of viral MAMPs/PAMPs occurs intracellularly (Ding and Voinnet, 2007).

RNA silencing consists of two major “branches”; the first one is that of small-interfering (si)RNAs, and one of the hallmarks for antiviral RNAi, and the second one is that of (host-gene encoded) micro (mi)RNAs involved in gene regulation. The antiviral RNAi response is induced by viral double stranded (ds)RNA molecules that arise from replicative intermediates or secondary RNA folding structures. These structures are sensed by a host RNase type III-like enzyme called Dicer-like (DCL) protein and cleaved into short interfering (si)RNA of 21–24 nucleotides (nt) in size (Sharma et al., 2013). The siRNAs generated are unwound and only one strand, the so-called guide-strand, is uploaded into a functional protein complex termed RNA-induced silencing complex (RISC). This activated complex next surveils and subsequently degrades (viral) RNA target molecules with sequence complementarity to the guide-strand. Degradation of the target RNA is mediated by slicer, the core component of RISC, which is represented by a member of the Argonaut (AGO) family of proteins (Vaucheret, 2008; Sharma et al., 2013). After primary siRNAs have been generated, in plants an amplification of siRNAs follows, which is required to mount an RNAi response to effectively combat virus infections locally and systemically. This amplification involves host RNA dependent RNA polymerases (RDRs) that are able to convert (aberrant) viral (m)RNAs into dsRNA in a siRNA-dependent and -independent manner (Csorba et al., 2009). Their subsequent processing by DCL leads to the generation of secondary siRNAs that correspond to sequences outside the primary target sequence, a process also called transitive silencing (Sijen et al., 2001). The antiviral RNAi response acts against all RNA and DNA viruses (Incarbone and Dunoyer, 2013), but in general is a relatively slow process that does not lead to complete clearance of viral infections. For an extensive description of RNAi readers are referred to nice reviews from e.g., Ding (2010) and Sharma et al. (2013).

Besides RNAi, viruses may also run into another, second layer of defense that involves resistance genes. While most of these are triggered by and confer resistance to a specific virus only, some act against several (related) viruses. The major class of these genes represent single dominant resistance genes (and of which the biggest group consists of the NB-LRR type), while others are recessive, tolerance, or partial resistance genes. A very nice example of a dominant resistance gene of the latter case has recently been described with the cloning and characterization of the Ty-1 resistance gene from tomato against Tomato yellow leaf curl geminivirus (TYLCV) This gene encodes an RNA-dependent RNA polymerase (RdRp) and is proposed to confer resistance to TYLCV by amplifying the RNAi signal (Verlaan et al., 2013). Tomato plants containing Ty-1 do not show symptoms upon a challenge with TYLCV, but low levels of virus can still be detected.

Recessive resistance (Truniger and Aranda, 2009) acting against viruses, relies on the observation that viruses require host factors (also called susceptibility factors) to enable an infection. The inability of interaction between such host factor and the virus leads to resistance. Since susceptibility factors are dominant, a resistance based on these requires all gene copies to be in the (resistant) recessive state. This explains why such resistance is generally termed recessive resistance. The majority of the recessive resistance genes known against plant viruses have been reported for potyviruses (Kang et al., 2005) and encode translation initiation factors of the 4E or 4G family (eIF4E/eIF4G) (Truniger and Aranda, 2009). The latter proteins need to interact with the cap-structure on (viral) transcripts, to allow for translation. Potyviral transcripts do not contain a cap structure, but provide a VPg (Virus-protein genome linked) to render their transcripts translatable in a cap-independent manner. Potyvirus infection leads to host shut off of cap-dependent transcripts, but only allow the cap-independent transcripts to be translated mediated by a subgroup of translation initiation factors; eIF(iso)4E/G. Viruses that encode their own cap-like structure (like potyviruses: VPg) require interaction with the translation initiation factors eIF4E/eIF4G for translation, this in turn induces a selection pressure on the host to escape the interaction between VPg and eIF4e, leading to recessive resistance. Recessive resistance genes toward other pathogens, such as fungi and bacteria have only been described to a limited extent and their encoded susceptibility factors (S genes) are proposed to provide a more durable resistance than dominant R genes. However due to their functions they may cause pleiotropic effects when knocked out from the host genome (Gawehns et al., 2013).

Dominant R genes trigger a hypersensitive response (HR) or an extreme response (ER) in case the reaction occurs in a single cell. Both involve a programmed cell death (PCD) response that rapidly kills infected cells and prevents systemic spread of the (virus) pathogen. An induced HR is quite characteristic and involves the activation and expression of SA, jasmonic acid (JA), nitride oxide (NO), ethylene, reactive oxygen species (ROS), and Ca²⁺, and expression of Pathogenesis Related (PR)-genes. While each component has a specificity toward certain pathogens, only SA, ROS, and Ca²⁺ seem to be effective against viruses (Loebenstein, 2009; Carr et al., 2010).

In the past, an HR was considered to be part of the resistance response, however, recent insights into R protein downstream signaling indicate that programmed cell death (HR) and resistance are distinct physiological pathways (Bendahmane et al., 1999; Bai et al., 2012). One of the best examples in support of this comes from studies on Rx-based resistance against PVX. The Rx gene product is a CC-NB-LRR protein from potato that is triggered by the PVX structural CP protein. The Rx protein localizes in the cytoplasm while shuttling to and from the nucleus thereby triggering resistance (Slootweg et al., 2010). Although an HR is monitored, this response can be knocked out without affecting Rx-mediated resistance against PVX (Bendahmane et al., 1999). Another example is the N-gene mediated resistance against TMV as described above in section Model of R Gene Recognition (Bhattacharjee et al., 2009). Similar observations have been made by others (Cole et al., 2001; Cawly et al., 2005; Genger et al., 2008; Bulgarelli et al., 2010; Bai et al., 2012) and indicate that the actual resistance response is different from an HR, although both mostly are triggered and may act in concert to clear viral invasions. Whether both are triggered by a pathogens' Avr determinant or whether HR is sequentially triggered following the R gene response is not clear.

While several interacting proteins have been identified that control R protein activity in the absence of pathogens (e.g., RAR1, SGT1, WRKY1, TPR1, Hsp90), more recently it has been found that there are also proteins that modulate the strength of defense responses (RanGAP, EDS1-PAD4) (Lu et al., 2003; Wiermer et al., 2005; Sacco et al., 2009). The benefit for the plant in a modulated fine-tuning of the ETI response to specific pathogens lies in improved effector sensing and minimizing the fitness costs involved with certain defense responses (free radical production, defense protein synthesis, cell death) (Padmanabhan and Dinesh-Kumar, 2010). While R gene mediated defense is taking place locally at the site of entry, it is also able to induce defense signaling responses in distally located tissues, known as systemic acquired resistance (SAR) (Vlot et al., 2008). For both the N gene in tobacco and Rx1 in potato, SAR has been demonstrated (Delaney et al., 1994; Liu et al., 2010) and in both cases this response is mediated by the SA-dependent pathway as a mobile signal. SAR also prevents infection by other pathogens in the host by activating PR genes in the systemic tissue, which are used as a hallmark of SAR and were shown to have antimicrobial activity, although a direct inhibition on virus replication has not been shown (Durrant and Dong, 2004; Loebenstein, 2009; Carr et al., 2010).

Viruses are pathogens to many different organisms and, irrespective of the host species they infect, often share similarities in genome organization and functions of encoded proteins. A good example of this is exemplified by viruses from the Bunyaviridae family where all members infect animals with the exception of those from the Tospovirus genus that, besides infecting their thrips vector, are plant pathogenic and are postulated to have evolved from a common ancestor. Likewise, as a result of co-evolution driven by host-pathogen interactions, plants and animals show some similarities in their innate immune sensory systems. While in plants the aforementioned R genes are important in mounting an ETI response, in animals two major classes are distinguished that (partially) share similarity to these R genes, however both function as PRRs in the PTI response. The first major class present the “nucleotide-binding domain and leucine-rich repeat”-proteins (NLRs) and the second class is that of Toll like receptors (TLRs), which are all found to function as PRRs in the PTI response. Both are immune receptors aimed at detecting “foreign” structures and activating downstream defense responses. The family of NLRs share the most homology, as evidenced when looking at R genes from plants and NACHT-LRR encoding genes from the animal kingdom (NAIP–CIITA–HET-E–TP1 domain) (Leipe et al., 2004; Takken et al., 2006; Maekawa et al., 2011b). They both contain a nucleotide binding domain and a LRR (Maekawa et al., 2011b) (Figure 2). Additionally, plant R proteins also share homology at their N-termini with animal TLRs, membrane-bound immune receptors that function as sensors in pathogen recognition across membranes.

The NLRs of both plant and animal kingdom share homology through the presence of the Leucine-rich repeats (LRR) in these proteins. The most prevalent type of R proteins in plants belong to the NB-LRR protein structural class, from which the central nuclear binding domain (NBS) exhibits similarity to the nucleotide binding domain in several metazoan apoptosis regulating proteins like Apaf-1 from mammals and CED-4 from C. elegans. Due to the latter the NBS domain is also often referred to as NB-ARC domain (from Apaf1–R-protein–CED4) (Van Der Biezen and Jones, 1998; Takken et al., 2006). The N-terminal domain furthermore separates different classes of R genes; TIR-NB-LRRs harbor a Toll/Interleukin-1 Receptor domain with similarity to metazoan TLRs (Burch-Smith et al., 2007; Bernoux et al., 2011; Maekawa et al., 2011b; Hao et al., 2013). CC-NB-LRRs contain coiled coil domain forming the more irregular shaped intertwined alpha-helices (Lupas, 1997). Parallel to the discovery of many NB-LRR encoding R genes in plants in the recent years, the search for homology to Apaf-1 and CED-4 resulted in the recognition of the NACHT-LRR protein family in vertebrates (Koonin and Aravind, 2000; Leipe et al., 2004). Animal NLRs activate caspase-1 leading to activation and release of the cytokine interleukin-1 beta (Case, 2011), which subsequently induces local and systemic immune reactions. Similar to plant NB-LRR proteins, NLRs were found to act as higher-order active complexes, e.g., NLRP1-3 and NLRC4 form a complex often termed the inflammasome (Maekawa et al., 2011b).

TLRs represent the best studied family of PRRs in mammals so far. They are transmembrane glycoprotein receptors with an extracellular PAMP-binding domain consisting of multiple LRR that fold into a “horseshoe” structure. Additionally, it possesses intracellular signaling regions that have similarity to the intracellular domain of the Interleukin-I receptor ((IL-1R), which is referred to as Toll/IL-1R (TIR) domain that mediates downstream signaling upon activation of the receptor. TLRs initiate signal cascades involving the activation of nuclear factor kappa b (NF-κ B), mitogen-activated protein kinase (MAPK), and interferon regulatory factors (IRFs). This subsequently leads to a concerted expression of interferons, cytokines, and chemokines. Finally, inflammatory processes, cell cycle arrest, and cell death are induced (Honda et al., 2005; Kaisho and Akira, 2006). In humans, 10 TLRs have been identified of which TLR2, -3, -4, -7, and -8 are involved in sensing structural components of RNA viruses like double-stranded RNA, single-stranded RNA and viral glycoproteins (Bowie and Unterholzner, 2008). While most TLRs are involved in extracellular recognition of PAMPs, TLR3, -7, and -8 are primarily restricted to intracellular compartments (endoplasmic-reticulum (ER), endosomes etc.) where they sense structural components of viral RNA. Besides TLRs cytosolic retinoic acid-inducible gene I (RIG-I)-like receptors (RLRs; RIG-I, MDA5, and LSP2) have been identified as sensors of RNA viruses and are involved in the very early response of some RNA viruses (Bowie and Unterholzner, 2008; Gerlier and Lyles, 2011; Jensen and Thomsen, 2012).

Pathogen recognition in both animal and plant kingdoms involves the LRR domain, which binds the ligand in its horseshoe shape, often followed by activation of a signaling cascade through kinase phosphorylation. Structural similarities between animal TLR/NLR and plant NB-LRR proteins point to a convergent evolution of these defense-related pathways (Yue et al., 2012). However innate immunity in animals and plants differs substantially in their downstream defense response, with interleukin/interferon-activated inflammatory responses combined with activating the adaptive immune system in mammalian systems and a resistance response (as explained before) often seen as a programmed cell death response in form of HR in plants.

Dominant resistance against plant viruses are of increasing interest to breeders and scientists in the past years. More and more is known about the molecular mechanisms behind R gene mediated defense and the induction of the HR. However, the number of well-studied examples is still very limited, and therefore makes it difficult to extrapolate to other less studied R genes. In depth studies on other R genes from different crops providing resistance against different phytopathogens will have to show whether there is common mechanism of defense shared between all R genes, or whether specialization between different classes of R genes occurs. Knowledge on mammalian innate immunity sensors and their mode of action may provide interesting and cross pollinating views in this. The rapid generation of resistance breaking virus isolates against dominant resistance genes already indicates the importance for alternative resistance genes, to provide more durable, and effective resistances.