Role of Nucleotide Binding Site-Leucine-Rich Repeat (NBS-LRR) Genes in Plant Immunity
Yaparla A, Smith LC
Department of Biological Sciences, George Washington University, 800 22nd Street NW, Washington DC 20052, United States
Abstract
Plants have evolved sophisticated mechanisms to combat pathogen attack and trigger an effective immune response. In addition to having physical and chemical barriers of protection, plants basal immune system can broadly detect pathogens by Microbe-Associated Molecular Patterns (MAMPs) and induce defense. On the other hand, pathogens express a suite of effectors that act to suppress these defenses. The products of Resistance (R) genes in plants interact with these effectors and lead to Effector-Triggered Immunity (ETI), making the effectors Avirulent (Avr) to the host. This R-Avr interaction is explained by two models: (i) The Guard model and (ii) The Decoy Model. Most of the R proteins are characterized by a Nucleotide-Binding Site (NBS) and a Leucine-Rich-Repeat (LRR) sequences, called as NBS-LRR class of proteins. Large repertoires of R genes with diverse recognition specificities have been detected in many plant species, of which NBS-LRR genes are the most abundant. This review throws light on guard and decoy hypothesis and mainly focusses on NBS-LRR genes, from the perspective of their evolution, functional significance and regulation of their gene products.
1. Introduction
Like all other living organisms, plants are also constantly engaged in a struggle for protecting themselves from pathogens, by adapting various immune strategies. The capacity of plants to defend themselves against pathogens, differs largely from mammals primarily due to their fundamental biological differences. Plants are sessile, requiring pathogen mobility where as mammals are mobile and hence are able to spread infections by contact. In addition, mammals have adaptive immunity enabling them to target the pathogens specifically. Intriguingly, no adaptive immune system which is equivalent to the highly complicated vertebrate immune system has been found in plants. Yet, the fact that a plethora of plant species exist till date, evokes an idea that plants in fact, have highly effective system of defense against pathogen invasion and disease. Despite these and other differences, plant disease resistance and mammalian innate immunity share several remarkable similarities. Plants possess physical and chemical barriers such as a thick cell wall made of lignin, a cuticle, secrete chemical compounds such as saponins which offer non-specific protection against a pathogen. It is integral to gain insights into various plant-pathogen interactions from the perspective of both the organisms, to understand how they coevolve for dominance over each other.
Plants are exposed to and invaded by diverse pathogens, including bacteria, fungi, viruses and nematodes. These pathogens cause various diseases, some of which show visible symptoms such as leaf spots, powdered mildew, leaf rust etc. (Himeno et al., 2014). While some diseases result in death of the plant, most of the diseases perturb its growth and development. One of the phenomena of plant disease resistance, is rapid cell death at the site of infection called, the Hypersensitive Response (HR) which is often compared with animal programmed cell death (Wei et al., 2012). Indisputably, this is the most effective process at limiting pathogens that require living host cells. Some resistance responses such as activation of defense gene expression eventually leading to the production of antimicrobial proteins or low molecular weight antibiotics is comparable to that of the innate immune system in insects and vertebrates.
Pathogen-Associated Molecular Patterns (PAMPs), as the name suggests are generic signals of the presence of pathogen and act as elicitors for plant’s basal defense. Some examples include fungal chitin, bacterial lipopolysaccharides, flagellins (Albersheim et al., 1975; Jones et al., 2004) and more recently, these have been coined the term Microbe-Associated Molecular Patterns (MAMPs), as some of the non-pathogenic microorganisms also have PAMPs (Bent at al., 2007). R genes is another mechanism by which plants detect pathogen infection and mount an effective immune response. The products of R genes interact with products of Avr genes that are present in the pathogen, either directly or indirectly. The products of Avr genes cause avirulence in the presence of R gene products and hence the name. The recognition of Avr proteins by plants is called effector-triggered immunity in contrast to (PAMP)-triggered immunity (Chisholm et al., 2006; Jones and Dangl, 2006). In the absence of a cognate R gene, Avr gene products contribute to virulence, although they are not important for pathogen viability. Disease resistance is observed if the product of any particular R gene has recognition specificity for a product produced by a particular pathogen Avr gene.
2. From Guard to Decoy Model
The classic concept of gene for gene relationship, offering plant disease resistance requires the presence of two complementary genes: Avr genes in pathogen and a matching R gene in the host. The direct interaction between Avr gene product and a corresponding R gene product occurs less frequently compared to that of the indirect interaction and can be called as a receptor-ligand model where R protein acts as a receptor for Avr protein (Keen 1990). In fact, direct binding for many R-Avr interactions was found consistent with the receptor-ligand model (Jia et al., 2000; Deslandes et al., 2003; Dodds et al., 2006; Ueda et al., 2006). For a number of R-Avr interactions, failure to detect direct interaction led to the Guard hypothesis which proposes that the target of Avr protein, also called as a guardee is protected by the R protein. When the guardee is modified by the effector target (Avr protein), it activates the R protein, which triggers disease resistance in the host (Figure 1A, Dangl and Jones, 2001). The Guard Model was originally proposed to explain the mechanism of Pseudomonas syringae, AvrPto interaction with the tomato proteins Pto and Prf (Van der Biezen and Jones, 1998) and was later generalized to other effector proteins. Intriguingly, small R gene repertoire can perceive multiple effectors of a broad diversity of pathogens that attack plants (Figure 1B, Vander Hoorn et al., 2008).
The guardees, illustrated in the guard model undergo a series of duplication events and subsequently, evolve independently to become decoys of the effector targets. Studies have shown that these decoys, mimic the guardees and are free from non-immune cellular functions that these proteins originally have (Vandern Hoorn, 2008). Also, they evolve to become better triggers for corresponding guard R protein, thus acting as an effective bait for pathogen effectors (Figure 1C, Vander Hoorn et al., 2008; Ashkenazi, 2002). The decoy itself has no function in the development of disease or resistance. Diverse examples from four cases of effector perception such as Pto in tomato, BS3 in pepper, RPS2 in Arabidopsis, RCR3 as a bait for fungus effector in tomato have aided in supporting the decoy model and have enhanced the proposed mechanism that decoys in fact, evolve independently.
Figure 1: Comparisons of the Guard and Decoy Models.
Figure 1A shows the classical guard model with effector (gray) binding to the guardee (green) with and without R protein (orange). Figure 1B shows a guard model with multiple targets (guarded- green, non-guarded). Figure 1C shows a decoy model where decoy (purple) is a structural mimic of guardee.
3. NBS-LRR genes
Till date, at least five different classes of R genes have been cloned (Van Ooigen et al., 2007). A vast majority of R proteins contain a Nucleotide Binding Site (NBS) and a Leucine-rich Repeat Region (LRR) (Dangl et al., 2001). Typically, LRRs have the ability to evolve to have different binding capacities and are mainly involved in protein-protein interactions (Ellis et al.,2000; Jones et al., 2006). In contrast, NBS have highly conserved motifs such as P loop, Kinase 2 etc. and are mainly involved in signaling (Tan et al., 2012). The NBS-LRR gene products can be classified into two subgroups based on the type of domain at the N-terminal portion: TIR-NBS-LRR (TNL) proteins that contain a Toll like domain and CC-NBS-LRR (CNL) proteins that have a coiled-coil domain. However, this subdivision is not always precise and may greatly differ across species in terms of numbers and organization into subgroups. This is predictable as NBS-LRR genes are one of the most numerous gene families in plants. One striking example is that TNL genes are near-total absent in monocotyledons. In dicotyledons genomes, TNL genes are greater in number than CNL genes. For instance, Arabidopsis thaliana and soybean contain nearly two-fold to six-fold more TNL than CNL genes (Guo et al., 2011; Kang et al., 2012); where as potato (Jupe et al., 2012; Lozano et al., 2012) and Medicago truncatula (Ameline-Torregrosa et al., 2008) have larger number of CNL genes.
4. Genomic Organization of NBS-LRR Genes in the Pant Genome
Through comparative genomic analyses, several hundreds of NBS-LRR proteins have been characterized that are encoded by plant genomes. Also, these encoding sequences gave great diversity in number and division into subgroups, as stated earlier. The evolution of this gene family is influenced by tandem duplications, ectopic duplications followed by local gene arrangements and gene conversion (Porter et al., 2009; Wan et al., 2013). The chromosomal distribution for NBS-LRR genes occur irregularly among different plant species. Also, some chromosomes are characterized by many more NBS-LRR genes than others within the same species. For instance, in Brachypodium dystachion, chromosome 4 consists of nearly one-third of all the NBS-LRR genes that have been detected (Tan et al., 2012). Contrarily, in Brassica rapa, chromosome 3 and 9 contain more NBS-LRR genes and very few are present on chromosome 4 (Mun et al., 2009). Finally, chromosome 2 and 3 In Lotus japonicas; chromosome 16 in soybean genome represent the highest number of NBS-LRR genes (Li et al., 2010).
The NBS-LRR genes can be organized either as isolated genes or linked clusters of varying sizes and studies indicate that latter type is involved in rapid R gene evolution (Hulbert et al., 2001). The clusters are further divided into two types based on phylogenetic relationships: (i) clusters that have undergone tandem duplication (ii) clusters derived from ectopic duplication, transposition, followed by subsequent gene rearrangements (Leister et al., 2004). In general, mixed clusters contain more number of NBS-LRR genes, although it may vary across species. For example, 73% of NBS-LRR genes in potato occur as mixed clusters (Jupe et al., 2012), while M. truncatula has 80% clustered (Ameline-Torregrosa et al., 2008) and B. distachion with 51% mixed clustered NBS-LRR genes. Some clusters may also contain phylogenetically distant NBS genes, although some of them have similar sequences. In fact, about 25% of M. truncatula clusters include both TNL and CNL genes (Ameline-Torregrosa et al., 2008) and also similar ratio is found in potato as well (Jupe et al., 2012).
Pseudogenes, which by definition are genes that resemble known sequences but are incapable of producing functional proteins, originate by the same mechanism as protein coding genes, but become non-functional due to mutations that disrupt the Open Reading Frame (ORF) or due to insertion of a pre-mature stop codon (Chandrasekaran et al., 2008). Many R pseudogenes have been identified in various plant species and typically, they show high sequence similarity with the NBS-LRR gene sequence. Sometimes, these pseudogenes also result in the production of partial proteins. Large deletions could be one of the primary reasons for the occurrence of pseudogenes in many plant species which are produced by transposition events and exon skipping with frame shift mutations. This can be a possible explanation for the production of truncated mRNAs, resulting in partial proteins (Lozano et al., 2012). Depending on the sub-group of NBS-LRR genes (CNL, TNL), the number of pseudogenes may vary. For example, in potato, 87% of the total pseudogenes indentified (179) belong to the CNL type and the rest are TNL type (Lozano et al., 2012). Pseudogenes also act as reservoirs for genetic diversity which can be accessed through recombination or gene conversion (Meyers et al., 1999). Research on plant resistance in potato proposed a role for pseudogenes as adaptor molecules, where by they act as recruiters of or interact with other NBR-LSS proteins (Lozano et al., 2012).
5. Evolution of NBS-LRR genes in plants
The rate of evolution in the NBS and LRR domains are not homogeneous and vary greatly. The gene conversion appears to occur at a much higher rate for LRR domains compared to the NBS domains, resulting in much higher variability in LRR domains (Kuang et al., 2004). This is related to the role of LRR in recognizing specific Avr gene products, to promote plant resistance to pathogens. Also, studies have been conducted in determining the ratio of non-synonymous substitutions (mutations that alter the amino acid sequence) to the synonymous substitutions (mutations that do not alter the amino acid sequence) in these two domains and was seen to be higher for LRR domain (Mondragon-Palomino et al., 2002). Intriguingly, for all the variability across NBS-LRR genes, some genes evolve rapidly with frequent gene conversions and the rest evolve slowly with rare gene conversion events between clades (McHale et al., 2006). Undoubtedly, clustering is one of the factors that influences this gene conversion rate as larger clusters potentially provide more donor sequences.
Two novel classes of NBS-encoding genes, besides the classical CNL and TNL subgroups have been identified. The first was identified in Physcomitrella patens genome which has a protein kinase domain at the N-terminus (PK-NBS-LRR; PNL) (Xue et al., 2012). The second was identified in the Marchantia polymorpha genome which was characterized by an a/b-hydrolase domain at the N-terminus (hydrolase-NBS-LRR; HNL) (Xue et al., 2012). These findings support the notion that NBS-LRR genes evolve to cope with specific pathogens. Often, gene duplication is also associated with functional redundance. Within duplicated regions, genes with redundant functions have been detected in the soybean genome (Kang et al., 2012).
6. Regulation and Activity of NBS-LRR Genes
In order to ensure correct resistance response, regulation of expression of NBS-LRR genes plays a crucial role. This regulation is indeed a complex process that occurs at various levels. Transcriptional level regulation aims at both quantity and quality of mRNA because alternative splicing may occur, giving rise to different transcript forms. Alternative splicing of R genes, mainly involves different combinations of functional domains. Some R gene transcripts also produce truncated proteins with partial domains. For example, Arabidopsis RPS6 which confers resistance to Pseudomonas syringae 6, can produce three alternative transcripts. The full-length protein contains three functional domains (TIR, NBS, LRR) and the alternative forms contain one or two of these domains (Kim et al., 2009). It is very fascinating to see that some of these truncated proteins have positive effects on resistance too. A striking example in this case is AvrRps4 encoded by Arabidopsis TNL gene, which is involved in resistance to bacterial pathogens (Zhang et al., 2003). They may also be functioning as adaptors for downstream signaling events (Duque et al., 2011).
Like other animals, MicroRNAs (miRNAs) are important regulators of gene expression in plants as well. MiRNAs are small RNAs ranging from 21 to 24 nucleotides in length and are processed by the Dicer nuclease from long RNA precursors with base-paired fold back structures (Baulcombe et al., 2004). MiRNAs form ribonuclear complexes with Argonaute (AGO) which together bind to the target RNA (Bartel et al., 2009; Voinnet et al., 2009). They act as mediators for gene silencing by acting as a negative switch and this has been implicated in the regulation of host defenses against pathogens (Yi et al., 2007) For instance, in a study carried out in tomato, datasets of miRNAs were analyzed and a regulatory cascade that affects disease resistance was identified (Shivaprasad et al., 2012). These results besides other, demonstrate that miRNAs have a conserved role in NBS-LRR gene regulation and pathogen resistance.
7. Concluding remarks
Recognition of Avr proteins by R proteins can be either be direct or indirect and the mechanism is proposed by two models: (i) Guard model (ii) Decoy model. Guard model proposes that the virulence target of the Avr protein (guardee), is guarded by the R protein and when the target is perturbed, the R proteins activate resistance. In case of a decoy model, the target is a structural mimic of the guardee. Although there is experimental evidence for both the models, further investigation is required for a deeper understanding of effector perception in plants.
The NBS-LRR proteins which are a class of R proteins in plants, constitute a comprehensive pathogen detection system. The NBS-LRR genes may be clustered or isolated sequences, where clusters share more sequence similarity but some may also contain phylogenetically distant NBS genes. With plant genome-wide studies, duplicated redundant genes have been discovered, which have contributed to provide new insight into expansion of this gene family through gene duplication events. This leads to the generation of novel recognition specificities in the R proteins primarily created through sequence variation in the beta sheet of the LRR domain. It is important to garner insights into biochemistry of NBS-LRR proteins to understand how a pathogen signal is perceived by the functional domains of these proteins. On the other hand, it is equally important to conduct research from the pathogen perspective and investigate its direct host targets.
Acknowledgements
We thank members of the Columbian College of Arts and Sciences at George Washington University, Washington DC, USA for funding.