NBS Domain Genes: The Molecular Sentinels of Plant Immunity and Their Biomedical Potential

Abstract

This article provides a comprehensive analysis of Nucleotide-Binding Site (NBS) domain genes, the largest class of plant resistance (R) genes that form the core of the plant innate immune system. We explore the foundational biology of NBS-LRR proteins, their role in Effector-Triggered Immunity (ETI), and their remarkable structural diversity across plant species. The content details cutting-edge computational and experimental methodologies for NBS gene identification, from traditional domain-based searches to modern deep learning tools like PRGminer. It further addresses the challenges in characterizing these complex genes and outlines robust validation frameworks using transcriptomics, qPCR, and functional assays. Designed for researchers and scientists in plant biology and drug development, this review synthesizes recent genomic discoveries to illuminate how understanding plant immune receptors can inform broader biomedical research and sustainable crop protection strategies.

The Architecture and Evolutionary Dynamics of Plant NBS Immune Receptors

Plants inhabit an environment teeming with potentially pathogenic microorganisms, including bacteria, fungi, oomycetes, viruses, and nematodes. Unlike animals, they lack a mobile immune system and have consequently evolved a sophisticated, multi-layered innate defense network [1] [2]. This system relies on the capacity to detect invading pathogens and mount an effective immune response. The foundational model for understanding this process is the two-tiered plant immune system, comprising Pattern-Triggered Immunity (PTI) and Effector-Triggered Immunity (ETI) [1] [3]. This conceptual framework, formally articulated by Jones and Dangl in 2006, revolutionized the understanding of plant-pathogen interactions by introducing a dynamic, zig-zag model of escalating offense and defense [2]. Within this system, nucleotide-binding site (NBS) domain genes, particularly those encoding nucleotide-binding and leucine-rich repeat receptors (NLRs), play a critical role as the central mediators of ETI [4] [5]. This technical guide provides an in-depth analysis of PTI and ETI, framing their mechanisms within the context of NBS gene evolution, function, and their burgeoning application in crop engineering.

Pattern-Triggered Immunity (PTI): The First Layer of Defense

Core Components: PAMPs and PRRs

PTI constitutes the first and broadest layer of inducible plant defense. It is activated upon recognition of conserved microbial molecules, historically termed Pathogen-Associated Molecular Patterns (PAMPs) but more accurately described as Microbe-Associated Molecular Patterns (MAMPs), as they are present in both pathogenic and non-pathogenic microbes [6]. These molecular patterns are indispensable for microbial viability and include bacterial flagellin, elongation factor Tu (EF-Tu), fungal chitin, and oomycete glucans [1] [6].

Detection of MAMPs is mediated by Pattern Recognition Receptors (PRRs), which are typically plasma membrane-localized receptor complexes [1] [7]. PRRs primarily belong to two classes: Receptor-Like Kinases (RLKs), which contain an extracellular ligand-binding domain, a transmembrane domain, and a cytoplasmic kinase domain; and Receptor-Like Proteins (RLPs), which lack the cytoplasmic kinase domain and require interaction with adapter kinases for signaling [2] [6]. Well-characterized examples include:

• FLS2: An RLK that recognizes bacterial flagellin via a conserved epitope (flg22) [6].
• EFR: An RLK that recognizes bacterial EF-Tu [6].
• CEBiP: An RLP involved in chitin perception in rice [2].

Signaling and Defense Responses

Upon MAMP perception, PRRs rapidly associate with co-receptors, such as the LRR-RK BAK1/SERK3, to initiate a robust intracellular signaling cascade [6]. Key early events include:

• An influx of extracellular Ca²⁺ into the cytosol [1].
• The activation of Mitogen-Activated Protein Kinase (MAPK) cascades [1] [2].
• A burst of Reactive Oxygen Species (ROS) [1] [7].
• Nitric oxide production [1].

This signaling network leads to extensive transcriptional reprogramming and the activation of downstream defense responses [1]. These include:

• Reinforcement of cell walls through callose deposition [1].
• Production of antimicrobial compounds, such as phytoalexins [8].
• Synthesis of defense hormones, including salicylic acid (SA), jasmonic acid (JA), and ethylene [1] [2].

Table 1: Key Molecular Components and Events in PTI Activation

Component/Event	Description	Function in PTI
MAMP/PAMP	Conserved microbial molecules (e.g., flg22, chitin)	Serves as the initial "danger signal" for pathogen presence
PRR	Plasma membrane receptor (e.g., FLS2, EFR)	Binds MAMPs to initiate immune signaling
Co-receptor (BAK1)	Somatic embryogenesis receptor kinase	Forms complex with PRRs to amplify and transduce signal
Calcium Influx	Rapid movement of Ca²⁺ into the cell	Acts as a secondary messenger
MAPK Cascade	Series of phosphorylation events	Transduces signal to the nucleus for transcriptional activation
ROS Burst	Production of reactive oxygen species	Direct antimicrobial action and signaling

Effector-Triggered Immunity (ETI): The Second Layer of Defense

Pathogen Effectors and the Role of NLRs

Successful pathogens have evolved to suppress PTI by secreting a repertoire of effector proteins directly into the plant cell apoplast or cytoplasm [1] [8]. This leads to Effector-Triggered Susceptibility (ETS). In response, plants have evolved intracellular immune receptors to recognize these effectors and activate a more potent defense response known as Effector-Triggered Immunity (ETI) [1] [3]. The primary mediators of ETI are the Nucleotide-binding and Leucine-rich Repeat receptors (NLRs), which are encoded by one of the largest and most diverse gene families in plants [1] [4] [5]. NLRs are also known as NBS-LRR proteins, highlighting the central role of the Nucleotide-Binding Site (NBS) domain [4].

NLR Structure, Classification, and Activation

A typical NLR protein features a conserved tripartite architecture [1] [3]:

• N-terminal Domain: Determines signaling pathway and classifies NLRs into:
  • TNLs: Contain a Toll/Interleukin-1 Receptor (TIR) domain.
  • CNLs: Contain a Coiled-Coil (CC) domain.
• Central NBS/ NB-ARC Domain: Serves as a molecular "on/off" switch, regulated by nucleotide (ADP/ATP) binding and hydrolysis [1] [3].
• C-terminal LRR Domain: Involved in auto-inhibition and effector recognition; evolves rapidly under diversifying selection to detect new effectors [4].

NLR activation functions as a molecular switch. In the resting state, the NLR is auto-inhibited, often with ADP bound to the NBS domain. Effector recognition, either direct or indirect, triggers nucleotide exchange (ADP for ATP), inducing a conformational change that leads to oligomerization into large signaling complexes called "resistosomes" [1] [3]. CNL resistosomes can form calcium-permeable channels in the plasma membrane, while TNL resistosomes often act as enzymes to produce small signaling molecules that activate downstream helpers [3].

Models of Effector Recognition

NLRs employ sophisticated strategies to detect pathogen effectors, balancing the need for specificity with the limited number of NLR genes against a vast number of potential effectors [1].

• Direct Recognition (Receptor-Ligand Model): The NLR directly binds the effector protein via its LRR domain. This is a straightforward gene-for-gene interaction but can be evolutionarily constrained [1].
• Indirect Recognition: The NLR monitors ("guards") the integrity of host proteins that are targeted by pathogen effectors.
  • Guard Model: The NLR detects effector-mediated modification of a true host virulence target (the "guardee") [1] [4]. A classic example is the activation of the NLRs RPM1 or RPS2 upon phosphorylation or cleavage of the guardee protein RIN4 by Pseudomonas syringae effectors [1].
  • Decoy Model: The NLR detects effector-mediated modification of a host protein that mimics the real virulence target but has no essential function other than to attract effectors. This confounds the pathogen [1]. For example, the decoy protein PBS1 is cleaved by the effector AvrPphB, triggering RPS5-mediated immunity [1].
  • Integrated Decoy Model: A domain that mimics a effector target is integrated directly into the NLR structure itself. For instance, the RRS1-RPS4 NLR pair contains an integrated WRKY domain that is targeted by effectors, enabling self-monitoring and activation [1].

Quantitative Genomics of NBS Domain Genes

The NBS-LRR gene family exhibits remarkable quantitative variation across the plant kingdom, reflecting its dynamic evolution and adaptation to diverse pathogenic pressures.

Table 2: Genomic Distribution of NBS-LRR Genes in Selected Plant Species [4]

Plant Species	Total NBS-LRR Genes	TNL Genes	CNL Genes	Notable Features
Arabidopsis thaliana	149-159	94-98	50-55	Model dicot with more TNLs
Oryza sativa (rice)	553-653	-	-	Monocot; lacks canonical TNLs
Glycine max (soybean)	319	-	-	Highly duplicated genome
Solanum tuberosum (potato)	435-438	65-77	361-370	High number of CNLs
Brachypodium distachyon	126	0	113	Monocot with no TNLs
Medicago truncatula	333	156	177	Balanced TNL/CNL distribution

A recent pan-genomic study identified 12,820 NBS-domain-containing genes across 34 plant species, from mosses to monocots and dicots, which were classified into 168 distinct domain architecture classes [5]. This diversity includes both classical (e.g., TIR-NBS-LRR) and novel, species-specific patterns (e.g., TIR-NBS-TIR-Cupin_1), underscoring the extensive diversification of this gene family [5]. NBS-LRR genes are often organized in clusters at specific chromosomal loci, a genomic arrangement thought to facilitate rapid evolution through tandem duplication and gene conversion, generating new pathogen recognition specificities [4].

PTI-ETI Synergy and Integrated Immune Signaling

The historical view of PTI and ETI as separate, linear pathways has been supplanted by a model of extensive crosstalk and synergy [1] [7]. While ETI responses are generally more robust and rapid, often culminating in the Hypersensitive Response (HR)—a localized programmed cell death that confines the pathogen—both systems activate an overlapping set of downstream defense responses [1] [8].

Recent research demonstrates that PTI and ETI potentiate each other [7]. ETI can enhance the amplitude and duration of PTI-related signals, such as the ROS burst and MAPK activation. Conversely, PTI components are often required for the full execution of ETI [1] [7]. This synergistic interaction ensures a robust, amplified defense output that is more effective than either system alone. The signaling networks converge on the production of defense hormones and the establishment of Systemic Acquired Resistance (SAR), a long-lasting, whole-plant immunity against secondary infections [2].

Diagram 1: Two-tiered plant immune system with PTI-ETI synergy.

Experimental Approaches and Research Reagents

The study of NBS genes and plant immunity leverages a wide array of molecular and genomic techniques. Key experimental workflows and reagents are essential for advancing this field.

Key Experimental Protocols

1. Identification and Evolutionary Analysis of NBS Genes:

• Method: Perform a genome-wide identification of NBS-encoding genes using HMMER-based searches (e.g., PfamScan) with the NB-ARC (PF00931) domain hidden Markov model (HMM) as a query against plant genome assemblies [5].
• Downstream Analysis: Classify genes based on domain architecture (TNL, CNL, etc.). Conduct evolutionary studies using OrthoFinder for orthogroup inference, construct phylogenetic trees with FastTreeMP, and analyze duplication events (tandem vs. whole-genome) [5].

2. Functional Validation via Virus-Induced Gene Silencing (VIGS):

• Purpose: To rapidly assess the function of a candidate NBS gene in plant resistance [5].
• Workflow: A fragment of the target NBS gene is cloned into a VIGS vector (e.g., based on Tobacco Rattle Virus). The recombinant vector is introduced into plants (e.g., resistant cotton) via Agrobacterium-mediated infiltration. Silenced plants are then challenged with a pathogen, and disease symptoms and pathogen titer are compared to control plants [5].

3. Interfamily Transfer of NLR Pairs:

• Purpose: To engineer disease resistance in a crop plant by transferring NLR sensor-helper pairs from a non-host donor species [3].
• Protocol: Isolate the genes encoding a known sensor NLR (e.g., Rpi-amr3 from Solanum americanum) and its cognate helper NLR (e.g., NRC2). Co-transform the susceptible recipient crop's genome (e.g., soybean) with both genes. Validate effector-dependent immune activation and resistance to the target pathogen (e.g., Phytophthora) [9] [3].

Diagram 2: Functional characterization workflow for NBS genes.

The Scientist's Toolkit: Essential Research Reagents

Table 3: Key Research Reagents and Solutions for Investigating NBS Genes and Plant Immunity

Reagent / Solution	Function / Application	Example Use-Case
PAMP Elicitors	Synthetic peptides (e.g., flg22, elf18) used to artificially activate PTI in experimental settings.	Studying early PTI signaling events like MAPK activation and ROS bursts [6].
VIGS Vectors	Viral vectors (e.g., TRV-based) designed to silence endogenous plant genes.	Rapid loss-of-function validation of candidate NBS genes in resistant plants [5].
Heterologous Expression Systems	Platforms like Nicotiana benthamiana for transient gene expression via Agrobacterium (Agroinfiltration).	Testing protein-protein interactions, NLR oligomerization, and cell death induction [3].
NLR Sensor-Helper Pairs	Cloned gene pairs (e.g., Bs2 + NRC3/4, Rpi-amr3 + SaNRC2) from donor species.	Engineering resistance in susceptible crops through interfamily transfer [3].
Protein Interaction Tools	Yeast-two-hybrid systems, Co-Immunoprecipitation (Co-IP) kits.	Validating direct binding of NLRs to effectors or host guardee/decoy proteins [1].
AI Prediction Tools	Computational models trained on protein datasets (e.g., from ANNA database).	Predicting novel pathogen-NLR interactions and optimizing receptor engineering [10].
CRBN ligand-10	CRBN ligand-10, MF:C13H12N2O2, MW:228.25 g/mol	Chemical Reagent
10-Methyltetradecanoyl-CoA	10-Methyltetradecanoyl-CoA, MF:C36H64N7O17P3S, MW:991.9 g/mol	Chemical Reagent

The two-tiered plant immune system, with its foundational PTI layer and highly specific ETI layer, represents a sophisticated defense network. The NBS domain genes, as the coding platform for NLR receptors, are central to ETI and are one of the most dynamic and diverse gene families in plants, shaped by a continuous evolutionary arms race with pathogens [4] [5]. Current research has moved beyond viewing PTI and ETI as isolated pathways, focusing instead on their synergistic integration, which provides a comprehensive and amplified defense output [1] [7].

Future directions in the field are being driven by advanced technologies. AI and machine learning are being used to predict plant-pathogen interactions and optimize immune receptors for broader recognition [10]. Synthetic biology approaches, such as engineering "Pikobodies" (where NLR recognition domains are replaced with nanobodies), are creating novel resistance specificities [10]. Furthermore, overcoming Restricted Taxonomic Functionality (RTF) by co-transferring sensor and helper NLR pairs from non-host plants into crops is a breakthrough strategy for engineering durable resistance, as demonstrated by conferring resistance to bacterial leaf streak in rice [9] [3]. A deep understanding of NBS gene evolution, PTI-ETI synergy, and the application of these novel engineering strategies is paramount for developing next-generation crops with resilient, broad-spectrum disease resistance.

Plant immunity relies on a sophisticated two-layered immune system, with Nucleotide-Binding Site Leucine-Rich Repeat (NBS-LRR) proteins serving as the primary intracellular receptors responsible for effector-triggered immunity (ETI). These proteins, also termed NLRs (NOD-like receptors), constitute one of the largest and most diverse gene families in plants, with approximately 80% of cloned disease resistance (R) genes encoding NBS-LRR proteins [11] [12]. They function as specialized intracellular sensors that detect pathogen effector molecules, initiating robust immune responses that often include a hypersensitive response (HR) and programmed cell death (PCD) to restrict pathogen spread [11] [13]. Recent structural and functional studies have revealed that NBS-LRR proteins operate not merely as simple receptors but as complex molecular switches that assemble into large signaling complexes called resistosomes, enabling them to function as genuine intracellular hubs for immune signaling integration and amplification [13].

The significance of NBS-LRR proteins extends beyond fundamental plant immunity to practical agricultural applications. Breeding programs increasingly leverage these proteins to develop disease-resistant crops, while their unique structural features offer insights for novel resistance gene design [13] [14]. This technical guide comprehensively examines the domain architecture, classification, activation mechanisms, and experimental methodologies for studying NBS-LRR proteins, providing researchers with a foundation for advancing both basic science and translational applications in plant immunity.

Domain Architecture and Classification of NBS-LRR Proteins

Conserved Domain Structure

NBS-LRR proteins are characterized by a conserved tripartite domain architecture that forms the structural basis for their immune functions. These large proteins, ranging from approximately 860 to 1,900 amino acids, contain at least four distinct domains joined by linker regions [15].

Table 1: Core Domains of NBS-LRR Proteins

Domain	Structural Features	Functional Role
Amino-Terminal Domain	Variable domain containing TIR, CC, or RPW8 motifs	Involved in protein-protein interactions and initiation of downstream signaling pathways
NBS (NB-ARC) Domain	Conserved nucleotide-binding site with multiple motifs (RNBS-A, RNBS-B, etc.)	Functions as a molecular switch through ATP binding and hydrolysis; regulates activation state
LRR Domain	Tandem leucine-rich repeats forming solenoid structure with parallel β-sheet	Mediates pathogen recognition specificity; involved in autoinhibition and protein-protein interactions
Carboxy-Terminal Domain	Variable non-conserved region	Potential regulatory functions; less characterized

The NBS domain (also called NB-ARC for NOD-like receptor Apaf-1, R proteins, and CED-4) contains several defined motifs characteristic of the 'signal transduction ATPases with numerous domains' (STAND) family of ATPases [15]. This domain functions as a molecular switch through specific binding and hydrolysis of ATP, with conformational changes between ADP-bound (inactive) and ATP-bound (active) states regulating downstream signaling [15] [16].

The LRR domain typically consists of tandem repeats of 20-30 amino acids with a characteristic leucine-rich motif, forming a curved solenoid structure with a parallel β-sheet lining the inner concave surface that serves as the putative binding interface [15] [12]. This domain exhibits the highest sequence diversity among NBS-LRR proteins, with evidence of diversifying selection acting on solvent-exposed residues, reflecting its role in specific pathogen recognition [15].

Classification Systems

NBS-LRR proteins are classified based on their domain composition into typical and irregular groups, with further subdivision according to N-terminal domain type.

Table 2: Classification of NBS-LRR Proteins in Selected Plant Species

Classification	Domain Composition	Nicotiana benthamiana [17] [16]	Salvia miltiorrhiza [11]	Vernicia species [14]
TNL	TIR-NBS-LRR	5 members	2 members	3 members in V. montana
CNL	CC-NBS-LRR	25 members	61 members	9 members in V. montana
RNL	RPW8-NBS-LRR	Not specified	1 member	Not detected
NL	NBS-LRR	23 members	Not specified	12 members in V. fordii
TN	TIR-NBS	2 members	7 members	7 members in V. montana
CN	CC-NBS	41 members	87 members	37 members in V. fordii
N	NBS only	60 members	29 members	29 members in V. fordii

The typical NBS-LRR proteins (TNL, CNL, NL) contain all three major domains and function primarily in pathogen recognition [17] [16]. In contrast, the irregular group (TN, CN, N), which lacks the LRR domain, typically functions as adaptors or regulators for the typical types [17]. The distribution of these subfamilies varies significantly across plant lineages, with TNLs completely absent from cereal genomes and showing marked reduction in some eudicot species like Salvia miltiorrhiza [11] [15].

Molecular Activation Mechanisms

Pathogen Recognition Strategies

NBS-LRR proteins employ sophisticated molecular strategies for pathogen detection, primarily through direct and indirect recognition mechanisms:

• Direct Recognition: Some NBS-LRR proteins physically bind pathogen effector proteins through their LRR domains. Examples include the rice Pi-ta protein binding to the fungal effector AVR-Pita [12], and flax L proteins interacting directly with fungal AvrL567 effectors [12]. This strategy typically involves high specificity, with single amino acid changes in either partner sufficient to disrupt recognition [12].

• Indirect Recognition (Guard Hypothesis): Many NBS-LRR proteins detect pathogens indirectly by monitoring the status of host proteins that are modified by pathogen effectors [12]. The Arabidopsis RIN4 protein represents a classic example, which is targeted by multiple bacterial effectors (AvrRpm1, AvrB, AvrRpt2) and monitored by the RPM1 and RPS2 NBS-LRR proteins [12]. Similarly, the Arabidopsis RPS5 guards the PBS1 kinase, detecting its cleavage by the bacterial protease AvrPphB [12].

• Integrated Decoy Model: Recent evidence suggests that some NBS-LRR proteins incorporate domains that mimic effector targets, serving as integrated decoys that trigger immunity upon effector binding [13].

From Recognition to Resistosome Assembly

Upon pathogen recognition, NBS-LRR proteins undergo profound conformational changes that initiate immune signaling:

• Initial Activation: Effector perception, whether direct or indirect, induces conformational alterations in the LRR and amino-terminal domains [12]. These changes promote nucleotide exchange in the NBS domain, replacing ADP with ATP [12] [16].

• Oligomerization: ATP binding triggers the oligomerization of NBS-LRR proteins into large multimeric complexes termed resistosomes [13]. This represents a critical step in activation, analogous to the oligomerization of mammalian NOD proteins [15].

• Resistosome Function: Structural studies have revealed distinct mechanisms for CNL and TNL resistosomes:

  • CNL Resistosomes: Proteins like ZAR1 and Sr35 form calcium-permeable cation channels upon oligomerization, suggesting channel activity directly contributes to immune signaling [13].
  • TNL Resistosomes: TNL proteins such as RPP1 and ROQ1 assemble into complexes with NADase activity, generating nucleotide-derived signaling molecules that are sensed by EDS1–PAD4 or EDS1–SAG101 complexes [13].

• Downstream Signaling: Resistosome formation initiates multiple defense pathways, including:

  • Calcium influx and reactive oxygen species (ROS) production
  • Activation of helper NLRs (e.g., ADR1s, NRG1s)
  • Defense gene expression and phytohormone signaling
  • Hypersensitive response and programmed cell death at infection sites

Figure 1: NBS-LRR Protein Activation Pathway. Pathogen recognition triggers conformational changes that promote nucleotide exchange and resistosome formation, leading to immune signaling.

Subcellular Localization and Compartmentalization

NBS-LRR proteins exhibit diverse subcellular localizations that correspond to their specific functions in pathogen detection. In Nicotiana benthamiana, predictions indicate 121 NBS-LRRs localized to the cytoplasm, 33 to the plasma membrane, and 12 to the nucleus [17] [16]. This compartmentalization enables surveillance of different cellular spaces and targeting of pathogen effectors with distinct subcellular localization patterns. Nuclear localization is particularly important for NBS-LRR proteins that detect effectors targeting host nuclear processes, such as RRS1-R which interacts with the PopP2 effector in the nucleus [12].

Experimental Methods for NBS-LRR Characterization

Genome-Wide Identification and Bioinformatics

Comprehensive characterization of NBS-LRR families begins with systematic genome-wide identification using conserved domain searches:

• HMMER Search: Hidden Markov Model searches using the NB-ARC domain (PF00931) from the Pfam database with stringent E-value cutoffs (e.g., E-values < 1*10⁻²⁰) [17] [16]. This approach identified 156 NBS-LRR homologs in Nicotiana benthamiana and 196 in Salvia miltiorrhiza [17] [11].

• Domain Validation: Candidate sequences are validated using multiple domain databases including SMART, Conserved Domain Database (CDD), and Pfam to confirm complete domain architecture with E-values below 0.01 [17] [16].

• Classification Pipeline: Validated sequences are classified into subfamilies based on presence/absence of TIR, CC, RPW8, and LRR domains using a combination of HMMER and CDD searches [18].

Figure 2: Experimental Workflow for NBS-LRR Gene Identification and Characterization. The pipeline progresses from bioinformatic identification through phylogenetic analysis to functional validation.

Phylogenetic and Structural Analysis

• Multiple Sequence Alignment: Tools like Clustal W and MUSCLE generate alignments of complete NBS-domain genes under default parameters [17] [18].

• Phylogenetic Tree Construction: Maximum likelihood methods in MEGA7/MEGA11 with bootstrap analysis (1000 replicates) based on models like Whelan and Goldman + Freq model [17] [18].

• Motif Discovery: MEME Suite analysis with motif count set to 10 and width lengths from 6-50 amino acids identifies conserved motifs beyond canonical domains [17] [16].

• Gene Structure Analysis: TBtools visualization of exon-intron structures from GFF3 annotation files reveals structural patterns across subfamilies [17].

Functional Characterization Approaches

• Expression Profiling: RNA-seq analysis of NBS-LRR genes under pathogen infection and stress conditions. Differential expression analysis using tools like Cufflinks with FPKM normalization [18].

• Virus-Induced Gene Silencing (VIGS): Powerful functional validation approach, as demonstrated in Vernicia montana where VIGS of Vm019719 compromised Fusarium wilt resistance [14].

• Heterologous Expression: Expressing NBS-LRR genes in susceptible backgrounds to confirm function, such as improved Pseudomonas syringae resistance in Arabidopsis expressing maize NBS-LRR genes [18].

• Promoter Analysis: Identification of cis-regulatory elements using PlantCARE database interrogation of 1500 bp upstream sequences [17] [16].

Research Reagent Solutions for NBS-LRR Studies

Table 3: Essential Research Reagents for NBS-LRR Characterization

Reagent/Tool	Specifications	Application	Example Use
HMMER Software	v3.1b2 with PF00931 (NB-ARC)	Genome-wide identification of NBS domains	Initial identification of 156 NBS-LRRs in N. benthamiana [17]
MEME Suite	v5.5.4 with motif count=10	Discovery of conserved protein motifs	Identification of 10 conserved motifs in N. benthamiana NBS-LRRs [17] [16]
TBtools	v2.0 with visualization modules	Gene structure mapping and motif visualization	Exon-intron structure analysis of NBS-LRR genes [17]
PlantCARE Database	Online platform with cis-element library	Promoter analysis and regulatory element prediction	Identification of 29 shared cis-elements in NBS-LRR promoters [17]
VIGS Vectors	TRV-based silencing systems	Functional validation through gene silencing	Confirmation of Vm019719 role in Fusarium wilt resistance [14]
CELLO v.2.5 & Plant-mPLoc	Multi-localization prediction tools	Subcellular localization prediction	Prediction of 121 cytoplasmic, 33 membrane, 12 nuclear NBS-LRRs [17]

Evolutionary Dynamics and Genomic Distribution

NBS-LRR genes exhibit remarkable evolutionary dynamics characterized by rapid birth-and-death evolution. They are frequently organized in clusters resulting from both segmental and tandem duplications [15]. This genomic architecture facilitates the generation of diversity through unequal crossing-over, sequence exchange, and gene conversion [15]. The evolution of different domains is heterogeneous, with the NBS domain subject to purifying selection while the LRR region shows evidence of diversifying selection, particularly in solvent-exposed residues that likely interact with pathogen components [15].

The number of NBS-LRR genes varies substantially across plant species, reflecting lineage-specific expansions and contractions. For example, Arabidopsis thaliana contains approximately 150 NBS-LRR genes, Oryza sativa over 400, and Triticum aestivum as many as 2151 [15] [18]. This variation results from species-specific evolutionary pressures and differences in pathogen exposure.

Comparative genomics reveals distinct evolutionary patterns in NBS-LRR subfamilies. TNL genes are completely absent from cereal genomes and show marked reduction in some eudicot lineages like Salvia species [11] [15]. In contrast, CNL genes are widespread across angiosperms, suggesting the early angiosperm ancestors possessed multiple CNLs [15]. These lineage-specific distributions reflect complex evolutionary histories including gene loss, subfunctionalization, and adaptive radiation.

NBS-LRR proteins represent sophisticated intracellular hubs that integrate pathogen perception with defense activation through complex molecular mechanisms. Their modular domain architecture enables dual functionality in pathogen recognition and signaling initiation, while their capacity to form resistosomes provides a structural basis for signal amplification. The extensive diversification of this gene family across plant lineages reflects continuous evolutionary arms races with pathogens.

Future research directions include elucidating the complete structural diversity of resistosomes, understanding the signaling networks connecting different NBS-LRR subtypes, and exploiting natural and engineered diversity for crop improvement. The integration of structural biology with genome editing approaches promises to accelerate the development of designer R genes with novel recognition specificities. As our understanding of NBS-LRR activation mechanisms deepens, so too will our ability to engineer durable disease resistance in crop plants, reducing reliance on chemical pesticides and enhancing global food security.

Plants employ a sophisticated, two-layered innate immune system to defend against pathogens. The second layer, known as effector-triggered immunity (ETI), is primarily mediated by intracellular nucleotide-binding site-leucine-rich repeat (NLR) receptors that detect pathogen-derived effector molecules, initiating robust immune responses [19] [20]. These NLR proteins constitute one of the largest and most variable gene families in plants, often representing nearly 1% of all annotated genes in a genome [4]. NLRs are modular proteins typically consisting of three core domains: a variable N-terminal domain, a central nucleotide-binding site (NBS) domain that acts as a molecular switch, and a C-terminal leucine-rich repeat (LRR) domain responsible for pathogen recognition [20] [5]. Based on their N-terminal domain and phylogeny, plant NLRs are classified into three major subfamilies: coiled-coil (CC) domain-containing NLRs (CNLs), Toll/interleukin-1 receptor (TIR) domain-containing NLRs (TNLs), and RESISTANCE TO POWDERY MILDEW 8-like CC (CCR) domain-containing NLRs (RNLs) [19] [5]. This classification reflects not only structural differences but also distinct functional specializations and signaling mechanisms, which form the focus of this technical guide.

Structural Classification and Genomic Distribution of NLR Subfamilies

Domain Architecture and Phylogeny

The classification of NLRs into CNL, TNL, and RNL subfamilies is defined by their distinct N-terminal domains, which dictate specific signaling functions and interaction partners.

• CNLs (Coiled-Coil NLRs): Characterized by an N-terminal coiled-coil (CC) domain. The CC domain in several characterized CNLs (e.g., ZAR1, Sr35) forms a helical bundle that, upon activation, oligomerizes to form a resistosome structure functioning as a calcium-permeable cation channel at the plasma membrane [19] [20] [13].
• TNLs (TIR Domain NLRs): Feature an N-terminal Toll/interleukin-1 receptor (TIR) domain. The TIR domain possesses enzymatic NADase (nicotinamide adenine dinucleotidase) and ADPR (adenosine diphosphate-ribose) polymerase-like activities. Upon effector recognition, TNLs oligomerize, and their TIR domains generate specific signaling molecules that activate downstream immune components [19] [13].
• RNLs (RPW8-like CC NLRs): Contain an N-terminal CC domain that is phylogenetically distinct from that of CNLs and is similar to the CC domain found in RPW8 (Resistance to Powdery Mildew 8) proteins [19] [5]. RNLs are further divided into two conserved subclades: the ACTIVATED DISEASE RESISTANCE 1 (ADR1) family and the N REQUIREMENT GENE 1 (NRG1) family [19]. They typically function as helper NLRs, acting downstream of sensor CNLs and TNLs to activate full immunity [19] [20].

The phylogenetic analysis of NLRs from various plant species reveals that these three subfamilies form distinct, well-supported clades, indicating an ancient divergence before the separation of angiosperms [19] [11].

Genomic Distribution and Variation Across Species

The number and proportion of NLR subfamilies vary dramatically across the plant kingdom, influenced by evolutionary pressures and lineage-specific adaptations. The table below summarizes this genomic distribution in selected species.

Table 1: Genomic Distribution of NLR Subfamilies in Selected Plant Species

Plant Species	Total NLR Genes	CNL Count (%)	TNL Count (%)	RNL Count (%)	Key References
Arabidopsis thaliana	149-159	~55 (35%)	~98 (62%)	5 (3%)	[19] [4]
Nicotiana benthamiana	156	25 (16%)	5 (3%)	4 (3%)	[16]
Salvia miltiorrhiza	62*	61 (98%)	0 (0%)	1 (2%)	[11]
Oryza sativa (Rice)	553-653	~550 (>99%)	0 (0%)	Limited	[11] [4]
Triticum aestivum (Wheat)	2151	Majority	0 (0%)	Limited	[18] [11]
Nicotiana tabacum	603	274 (45.5%)	15 (2.5%)	Included in CC-types	[18]

Note: *Number of typical NLRs with complete N-terminal and LRR domains out of 196 identified NBS-domain genes. *Percentages based on broad categories (CC-NBS, TIR-NBS) from the source data.*

Key observations from genomic studies include:

• Monocot-Dicot Divergence: Monocot species like rice and wheat have completely lost the TNL subfamily, while CNLs have massively expanded [11] [4] [21].
• Lineage-Specific Reduction: Some dicot families, such as Lamiaceae (e.g., Salvia miltiorrhiza), also show a marked reduction or loss of TNLs and RNLs [11].
• Expansion Mechanisms: NLR repertoires expand primarily through whole-genome duplication (WGD) and tandem duplication, with genes often organized in clusters that facilitate rapid evolution and generation of new pathogen specificities [5] [4] [18].

Signaling Mechanisms and Pathways

CNL Signaling: Resistosome Channel Formation

Sensor CNLs, such as Arabidopsis ZAR1 and wheat Sr35, initiate immunity through a well-characterized mechanism of oligomerization into resistosomes.

Table 2: Key Experimental Findings on CNL Resistosomes

CNL Protein	Pathogen Trigger	Oligomeric State	Function	Key Experimental Evidence
ZAR1 (Arabidopsis)	Pseudomonas syringae effectors via RKS1	Pentameric	Ca²⁺-permeable non-selective cation channel	Cryo-EM structure; channel activity in Xenopus oocytes and plant cells; channel activity required for cell death and immunity [19] [20] [13].
Sr35 (Wheat)	Wheat stem rust effector	Pentameric	Ca²⁺-permeable non-selective cation channel	Cryo-EM structure; channel activity in Xenopus oocytes; sufficient to confer resistance [19] [20].

The canonical CNL activation pathway involves:

• Effector Recognition: Direct or indirect detection of pathogen effectors, often via integrated decoy domains or associated proteins.
• Nucleotide Exchange: Conformational change in the NBS domain from ADP-bound (inactive) to ATP-bound (active) state.
• Oligomerization: Assembly into a wheel-like pentameric resistosome.
• Channel Formation: The N-terminal CC domains form a funnel-shaped α-helical barrel that inserts into the plasma membrane, creating a calcium-permeable channel [20] [13] [22].
• Immune Activation: Calcium influx disrupts ion homeostasis, activates other defense components, and triggers transcriptional reprogramming and often a hypersensitive response (HR) [19] [22].

Figure 1: CNL Signaling Pathway via Resistosome Formation

TNL Signaling: Enzymatic Activity and Bifurcated Helper Recruitment

TNLs employ a distinct, more complex signaling mechanism that involves enzymatic activity and downstream helper components. Key characterized TNLs include Arabidopsis RPP1 and Nicotiana benthamiana Roq1 [20].

TNL Activation and Signaling Workflow:

• Effector Recognition and Oligomerization: Similar to CNLs, effector perception induces ATP binding and TNL oligomerization into resistosomes [20] [13].
• Enzymatic Activity: The oligomerized TIR domain acts as an NADase, hydrolyzing NAD+ and generating a mix of nucleotide-based signaling molecules (e.g., ADP-ribose isomers, cyclic ADP-ribose) [19] [13].
• EDS1 Heterodimer Activation: These signaling molecules are perceived by two exclusive heterodimers of the lipase-like protein EDS1: EDS1-PAD4 and EDS1-SAG101 [19].
• Helper NLR Activation:
  • The EDS1-PAD4 heterodimer physically associates with and activates ADR1 family RNLs [19].
  • The EDS1-SAG101 heterodimer physically associates with and activates NRG1 family RNLs [19].
• Execution of Immunity: The activated RNLs form calcium-permeable channels at the plasma membrane, ultimately leading to defense gene activation and cell death [19]. ADR1s are particularly involved in transcriptional reprogramming, while NRG1s are more specialized in triggering cell death [19].

Figure 2: TNL Signaling via Enzymatic Activity and Helper RNL Activation

RNL Signaling: Helper NLRs as Common Signaling Hubs

RNLs function as essential signaling nodes downstream of multiple immune receptors. The Arabidopsis genome encodes 3 ADR1 and 2 NRG1 full-length genes that act partially redundantly [19].

Key Functional Characteristics of RNLs:

• Convergence Points: The ADR1 subfamily, in particular, is required not only for TNL and some CNL signaling but also for immune signaling initiated by certain cell surface pattern recognition receptors (PRRs), positioning the EDS1-PAD4-ADR1 module as a convergence hub for pattern-triggered and effector-triggered immunity [19].
• Plasma Membrane Localization: RNL localization to the plasma membrane is mediated by the interaction of positively charged residues in their CCR domain with phosphatidylinositol-4-phosphate lipids. This localization is critical for their cell death function [19].
• Resistosome Formation: Like CNLs, activated RNLs (both ADR1 and NRG1) self-associate and form high-molecular-weight complexes (resistosomes) at the plasma membrane [19] [22].
• Cation Channel Function: Autoactivated NRG1.1 and ADR1 have been shown to promote non-selective cation influx in plant and human cells, leading to cell death independent of other plant proteins, confirming their function as executioners of immunity [19].

Experimental Methodologies for NLR Research

Genome-Wide Identification and Classification

A standard pipeline for identifying and classifying NLR genes leverages the conserved NBS (NB-ARC) domain.

Table 3: Standard Protocol for Genome-Wide NLR Identification

Step	Method/Tool	Key Parameters	Purpose	Validation
1. Domain Search	HMMER v3.1b2	HMM profile PF00931 (NB-ARC), E-value < 1e-20 [11] [16] [18]	Initial identification of NBS-containing genes	Manual verification via Pfam/CDD
2. Domain Annotation	Pfam Scan / SMART / NCBI CDD	Profiles for TIR (PF01582), CC, LRR (PF00560, etc.) [16] [18]	Classify into CNL, TNL, RNL, and atypical subtypes	Confirm domain integrity and architecture
3. Phylogenetic Analysis	MUSCLE (Alignment), MEGA11 (Tree)	Neighbor-joining or Maximum Likelihood, 1000 bootstraps [11] [18]	Visualize evolutionary relationships and subfamily clades	Check clustering with known NLRs from model species
4. Genomic Distribution	MCScanX	Self-BLASTP, synteny analysis [5] [18]	Identify tandem/segmental duplications and gene clusters	Compare with known duplication history

Functional Characterization Techniques

Several key experimental approaches are used to delineate the function of specific NLRs and their signaling mechanisms.

Table 4: Key Functional Assays in NLR Research

Assay Type	Methodology	Application Example	Readout
Genetic Requirement	Reverse genetics (Knockout mutants, VIGS)	Demonstrate that RNLs (ADR1s/NRG1s) are required for TNL immunity [19]	Loss of resistance/HR in mutant
Biochemical Activity	In vitro enzymatic assays	Show TIR domains of TNLs have NADase activity [13]	NAD+ hydrolysis, product formation
Protein Complex Analysis	Co-immunoprecipitation (Co-IP), FRET, SEC-MALS	Confirm EDS1-PAD4 interaction with ADR1s [19]	Physical association of proteins
Channel Function	Electrophysiology (e.g., in Xenopus oocytes)	Demonstrate ZAR1 resistosome is Ca²⁺-permeable channel [20] [22]	Ion current measurement
Functional Validation	Virus-Induced Gene Silencing (VIGS)	Silencing of GaNBS in cotton reduces virus resistance [5]	Increased pathogen titer/symptoms
Structural Studies	Cryo-Electron Microscopy (Cryo-EM)	Solve structures of ZAR1, RPP1, ROQ1 resistosomes [20] [13]	Atomic-level 3D structure

The Scientist's Toolkit: Essential Research Reagents

Table 5: Key Research Reagents for NLR Signaling Studies

Reagent / Material	Function / Application	Specific Examples / Notes
HMM Profile PF00931	Hidden Markov Model for identifying NBS domains in genomic sequences	Critical first step for genome-wide NLR identification [11] [16]
VIGS Vectors	Virus-Induced Gene Silencing for rapid transient loss-of-function studies	Used to validate NBS gene function in cotton and tobacco [5] [16]
Heterologous Systems (e.g., Xenopus oocytes)	For electrophysiological characterization of NLR channel activity	Confirmed cation channel function of ZAR1 and NRG1 [19] [22]
Anti-EDS1 / Anti-PAD4 Antibodies	Immunoprecipitation and protein complex analysis	Essential for probing EDS1 heterodimer interactions [19]
Cryo-EM Infrastructure	High-resolution structural determination of NLR resistosomes	Revealed oligomeric structures of ZAR1, Sr35, RPP1 [20] [13]
Mutant Plant Lines	Genetic analysis of NLR function (e.g., T-DNA knockouts, CRISPR-Cas9)	Arabidopsis adr1, nrg1, eds1 mutants define immune hierarchy [19]
Cyclohex-1,4-dienecarboxyl-CoA	Cyclohex-1,4-dienecarboxyl-CoA, MF:C28H38N7O17P3S-4, MW:869.6 g/mol	Chemical Reagent
(25S)-3-oxocholest-4-en-26-oyl-CoA	(25S)-3-oxocholest-4-en-26-oyl-CoA, MF:C48H76N7O18P3S, MW:1164.1 g/mol	Chemical Reagent

The classification of plant NLRs into CNL, TNL, and RNL subfamilies reflects a fundamental functional specialization within the plant immune system. CNLs and TNLs primarily act as sensor NLRs that directly or indirectly recognize pathogen effectors, but they activate immunity through distinct mechanisms: CNLs via cation channel formation and TNLs via enzymatic production of small signaling molecules. RNLs function as conserved helper NLRs that transduce signals from both TNLs and some CNLs/PRRs, ultimately executing defense responses through a similar channel-based mechanism.

Future research will likely focus on several frontiers:

• Structural Dynamics: Understanding the full conformational landscape of NLR activation from resting to active states.
• Network Integration: Elucidating how complex NLR networks, with multiple sensors and helpers, are regulated to avoid autoimmunity while ensuring robust defense.
• Engineering Applications: Leveraging structural and mechanistic insights to design novel synthetic NLRs with tailored resistance specificities, offering powerful strategies for crop improvement and sustainable agriculture [13].

The precise knowledge of CNL, TNL, and RNL signaling domains and pathways not only deepens our fundamental understanding of plant immunity but also provides the essential toolkit for engineering disease resistance in the era of climate change and emerging plant pathogens.

The nucleotide-binding site (NBS) domain genes represent a cornerstone of plant innate immunity, encoding intracellular immune receptors that recognize diverse pathogens and trigger robust defense responses [23] [24]. These genes, predominantly belonging to the nucleotide-binding leucine-rich repeat (NLR) family, exhibit remarkable genomic architecture characterized by dynamic arrangements and extensive diversification mechanisms [25] [26]. Their genomic organization is not random but follows distinct patterns that facilitate rapid evolution in response to changing pathogen pressures. This technical guide examines the structural and evolutionary principles governing NBS gene families, with particular emphasis on how tandem duplication events, gene cluster formation, and various selective pressures collectively generate the diversity necessary for effective plant immunity. Understanding these organizational paradigms provides crucial insights for harnessing NBS genes in crop improvement programs and developing sustainable disease management strategies.

Genomic Distribution and Architectural Patterns of NBS Genes

Chromosomal Arrangement and Cluster Formation

NBS genes display non-random distribution patterns across plant genomes, with significant clustering observed in specific chromosomal regions. Studies across multiple species reveal that these genes are frequently concentrated near telomeric regions, where they form complex arrays conducive to rapid evolution [23]. In pepper (Capsicum annuum), chromosome 09 harbors the highest density of NLR genes, with 63 identified members, while chromosome 08 also shows significant clustering [23]. Similarly, research on barley (Hordeum vulgare) indicates that duplication-prone regions containing NBS and other defense-related genes are located primarily in subtelomeric regions across all seven chromosomes [26].

The propensity for NBS genes to cluster in specific genomic regions creates architectural frameworks that facilitate evolutionary innovation. These arrangements allow for the coordinated evolution of functionally related genes and enable the generation of novel recognition specificities through various recombination mechanisms. The physical proximity of NBS genes within these clusters promotes sequence exchanges and the emergence of new gene variants through non-allelic homologous recombination, contributing to the extensive diversity observed in plant immune receptors.

Presence-Absence Variation and the Core-Adaptive Model

Pan-genomic studies have revealed extensive presence-absence variation (PAV) for NBS genes among different accessions of the same species, supporting a "core-adaptive" model of resistance gene evolution [25]. This model distinguishes between:

• Core subgroups: Evolutionarily conserved NBS genes present across most or all individuals, exemplified by ZmNBS31 in maize, which demonstrates high expression under both stressed and control conditions, suggesting a fundamental role in basal immunity [25].
• Adaptive subgroups: Highly variable NBS genes exhibiting significant PAV, such as the ZmNBS1-10 and ZmNBS43-60 subgroups in maize, which likely represent recent evolutionary adaptations to specific pathogen pressures [25].

This genomic plasticity enables plant populations to maintain a diverse repertoire of resistance specificities, with structural variants (SVs) associated with altered motif structures and significantly impacted gene expression profiles [25].

Table 1: NBS Gene Distribution and Classification Across Plant Species

Plant Species	Total NBS Genes	Subfamily Composition	Genomic Features	Reference
Capsicum annuum (pepper)	288 canonical NLRs	CNL, TNL, RNL, and truncated variants	Significant clustering on Chr09 (63 genes) and near telomeric regions	[23]
Nicotiana tabacum (tobacco)	603 NBS members	45.5% N-type, 24.9% CN-type, 12.3% CC-NBS-LRR, 10.6% CC-NBS	76.62% traceable to parental genomes (N. sylvestris and N. tomentosiformis)	[27]
Nicotiana benthamiana	156 NBS-LRR homologs	5 TNL, 25 CNL, 23 NL, 2 TN, 41 CN, 60 N-type	0.25% of annotated genes; RPW8 domain in only four NBS-LRRs	[16]
Zea mays (maize)	Multiple subgroups	Distinct "core" (e.g., ZmNBS31) and "adaptive" (e.g., ZmNBS1-10) subgroups	Extensive presence-absence variation across 26 inbred lines	[25]

Evolutionary Mechanisms Driving NBS Gene Diversification

Duplication Modes and Their Evolutionary Impacts

NBS gene families expand and diversify primarily through three duplication mechanisms: tandem duplication, segmental duplication, and whole-genome duplication (WGD), each contributing distinct evolutionary dynamics [23] [27] [26].

Tandem duplication serves as the primary driver of NLR family expansion in several plant species. In pepper, approximately 18.4% (53/288) of NLR genes originated through tandem duplication events, predominantly on chromosomes 08 and 09 [23]. These recent, species-specific expansions generate localized clusters of homologous genes that undergo rapid sequence diversification, enabling adaptation to emerging pathogen strains.

Whole-genome duplication contributes significantly to NBS gene content in allopolyploid species such as Nicotiana tabacum, where WGD-derived genes typically exhibit strong purifying selection (low Ka/Ks ratio), preserving essential immune functions [25] [27]. In contrast, genes arising through tandem and proximal duplications often show signs of relaxed or positive selection, indicating directional selection for new functions [25].

Different NBS gene subtypes demonstrate distinct preferences for duplication mechanisms. In maize, canonical CNL/CN genes largely originate from dispersed duplications, while N-type genes are enriched in tandem duplications [25]. This subtype-specific duplication bias influences evolutionary rates and functional diversification across different NBS gene classes.

Selection Pressures and Evolutionary Innovation

NBS genes experience varied selection pressures across their protein domains, reflecting their functional constraints and evolutionary flexibility. The LRR (leucine-rich repeat) domains typically display the highest variability, often showing signatures of positive selection that fine-tune pathogen recognition specificities [23]. In contrast, the NBS (nucleotide-binding site) domains generally evolve under purifying selection, conserving essential functions in signal transduction [23].

This domain-specific evolution enables NBS proteins to maintain conserved signaling machinery while diversifying their pathogen recognition capabilities. The "birth-and-death" evolutionary model characterizes NBS gene family dynamics, with continuous gene duplication, functional diversification, and pseudogenization generating extensive structural and functional variation over evolutionary time [26].

Table 2: Evolutionary Characteristics of NBS Genes Across Duplication Types

Duplication Mechanism	Evolutionary Rate (Ka/Ks)	Selection Pressure	Functional Implications	Examples
Tandem Duplication	Variable, often high	Frequent positive selection	Rapid generation of novel recognition specificities	Pepper NLRs on Chr08/09 [23]
Segmental Duplication	Moderate	Primarily purifying selection	Expansion of functional gene clusters	Maize NBS subgroups [25]
Whole-Genome Duplication	Low (Strong purifying selection)	Strong purifying selection	Preservation of essential immune functions	Nicotiana tabacum NBS genes [27]
Dispersed Duplication	Subtype-dependent	Varies by gene type	CNL/CN gene expansion in maize	Maize canonical CNL/CN genes [25]

Experimental Methodologies for NBS Gene Identification and Characterization

Genomic Identification and Annotation Pipeline

Comprehensive identification of NBS gene families requires integrated bioinformatics approaches combining multiple computational tools and databases. The following workflow represents a standardized pipeline for NBS gene annotation:

Step 1: Initial Identification

• Perform HMMER searches (v3.1b2/v3.3.2) against the target proteome using the NB-ARC domain (PF00931) from the Pfam database with E-value cutoffs of <1×10⁻²⁰ [23] [27] [16].
• Conduct BLASTP searches using known NLR protein sequences from related species as queries [23].

Step 2: Domain Validation and Classification

• Confirm domain composition using NCBI Conserved Domain Database (CDD) for NB-ARC (cd00204), CC, and other domains [23] [27].
• Validate with Pfam batch search or InterProScan for additional domain annotation (TIR: PF01582, LRR: PF00560, PF07723, PF07725, PF12779, etc.) [23] [27].
• Classify genes into subfamilies (TNL, CNL, NL, TN, CN, N) based on domain architecture [27] [16].

Step 3: Manual Curation

• Remove redundant sequences and pseudogenes.
• Verify domain completeness and structural integrity.
• Extract gene sequences and genomic coordinates for downstream analyses.

Evolutionary and Expression Analysis Methods

Evolutionary Analysis:

• Construct phylogenetic trees using Multiple Sequence Alignment with Muscle v5 or ClustalW, followed by Maximum Likelihood analysis with IQ-TREE or MEGA7/11 (1000 bootstrap replicates) [23] [16].
• Identify gene duplication events using MCScanX with BLASTP-based self-comparisons and synteny analysis [23] [27].
• Calculate non-synonymous (Ka) and synonymous (Ks) substitution rates using KaKs_Calculator 2.0 with Nei-Gojobori model to determine selection pressures [27].

Expression Profiling:

• Analyze RNA-seq data by mapping clean reads to reference genomes using Hisat2 [23] [27].
• Perform differential expression analysis with DESeq2 or Cufflinks/Cuffdiff, applying thresholds of |log2 Fold Change| ≥ 1 and FDR < 0.05 [23] [27].
• Validate expression patterns through RT-qPCR under pathogen inoculation conditions [23].

Regulatory Element Analysis:

• Extract promoter regions (up to 2 kb upstream of transcription start site).
• Identify cis-regulatory elements using PlantCARE database, focusing on defense-related motifs (SA/JA-responsive elements, W-boxes) [23] [16].

Table 3: Key Research Reagents and Computational Tools for NBS Gene Analysis

Resource Type	Specific Tool/Database	Primary Function	Application Example
Domain Databases	Pfam (PF00931), NCBI CDD (cd00204)	NBS domain identification and validation	Confirming NB-ARC domain in candidate sequences [23] [16]
Bioinformatics Tools	HMMER v3.1b2/v3.3.2, MEME, MCScanX	Sequence search, motif discovery, synteny analysis	Identifying conserved motifs, tandem duplication events [23] [27] [16]
Phylogenetic Software	MEGA11, IQ-TREE, ClustalW	Multiple sequence alignment, tree construction	Evolutionary relationship inference among NBS subfamilies [23] [27] [16]
Selection Pressure Analysis	KaKs_Calculator 2.0	Ka/Ks calculation	Determining purifying/positive selection on duplicated genes [27]
Genome Browsers/ Databases	NCBI, Sol Genomics Network, PlantCARE	Genome annotation, cis-element prediction	Retrieving promoter sequences, identifying regulatory elements [23] [16]
Expression Analysis	Hisat2, DESeq2, Cufflinks	RNA-seq mapping, differential expression	Identifying pathogen-responsive NBS genes [23] [27]

Computational Approaches and Emerging Technologies

Advanced Bioinformatics Frameworks

Traditional domain-based bioinformatics pipelines are increasingly supplemented with machine learning (ML) and deep learning (DL) approaches for improved R-protein prediction [28]. These computational strategies address limitations of conventional methods, particularly for identifying divergent NBS genes with atypical domain architectures.

Specialized computational tools have been developed specifically for resistance gene annotation, including:

• DRAGO2/3 and RGAugury for comprehensive R-gene analysis
• RRGPredictor utilizing machine learning classifiers
• NLR-Annotator and NLRtracker for NLR-specific annotation [28]

These tools enable more accurate genome-wide identification of NBS genes and facilitate comparative genomic analyses across species, revealing evolutionary patterns and functional relationships.

Genomic Technologies for NBS Gene Manipulation

CRISPR-Cas systems have emerged as powerful tools for functional characterization and improvement of NBS genes, enabling:

• Precise genome editing to enhance function of R genes
• Disruption of susceptibility (S) genes frequently exploited by pathogens
• Gene pyramiding strategies to develop durable resistance [24]

RNA interference (RNAi) provides a non-transgenic approach for disease control by silencing essential pathogen genes, leveraging the plant's innate RNAi machinery to target specific pathogen mRNA sequences [24].

Additionally, high-throughput sequencing technologies facilitate metagenomic pathogen identification and tracking of disease outbreaks, supporting the discovery of novel NBS gene functions and pathogen recognition specificities [24].

The genomic organization of NBS genes represents a sophisticated evolutionary adaptation that balances structural conservation with functional diversification. Tandem duplication events, gene cluster formation, and varied selection pressures collectively generate the diversity necessary for plant immunity. The intricate relationship between duplication mechanisms, structural variations, and selection pressures shapes the evolution of NBS genes across plant species, enabling continuous adaptation to evolving pathogen challenges. Future research leveraging advanced genomic technologies, computational approaches, and functional characterization methods will further illuminate the complex dynamics of NBS gene families, facilitating their strategic application in crop improvement programs and sustainable agriculture. The organized complexity of NBS gene genomic architecture stands as a testament to the remarkable evolutionary innovation underlying plant-pathogen interactions.

Leucine-rich repeat (LRR) domains in plant nucleotide-binding site (NBS)-LRR proteins represent a striking example of evolutionary innovation in pathogen recognition. These domains evolve through positive selection that preferentially targets solvent-exposed residues, generating the diversity necessary for recognizing rapidly evolving pathogen effectors. Genomic analyses across multiple plant species reveal that LRR domains are hotspots for nonsynonymous substitutions, indel variations, and domain shuffling. This review synthesizes current understanding of the molecular evolutionary forces driving LRR diversification and their functional implications for plant immunity, providing a framework for leveraging this knowledge in crop improvement strategies.

Plant nucleotide-binding site (NBS) domain genes encode the largest family of intracellular immune receptors that confer resistance to diverse pathogens including viruses, bacteria, fungi, oomycetes, nematodes, and insects [29]. The majority of cloned plant disease resistance (R) genes encode NBS-leucine rich repeat (LRR) proteins characterized by a central NBS domain and C-terminal LRR region [30] [29]. These proteins function as sophisticated surveillance systems that directly or indirectly recognize pathogen effector molecules, triggering robust defense responses such as the hypersensitive response (HR) [31].

The NBS-LRR family is subdivided into two major classes based on N-terminal domains: TIR-NBS-LRR (TNL) proteins containing Toll/interleukin-1 receptor domains and CC-NBS-LRR (CNL) proteins containing coiled-coil domains [29] [32]. A third minor class, RPW8-NBS-LRR (RNL), has also been identified in some species [33]. These proteins exhibit a modular architecture where different domains perform specialized functions: the N-terminal domain mediates downstream signaling, the central NBS/ NB-ARC domain functions as a molecular switch regulated by nucleotide binding and hydrolysis, and the LRR domain is primarily responsible for pathogen recognition specificity [27] [31].

The genomic architecture of NBS-LRR genes reflects their evolutionary dynamics. They frequently reside in clusters throughout plant genomes, with copy numbers varying significantly across species—from approximately 150 in Arabidopsis thaliana to over 400 in Oryza sativa and more than 700 in Arachis hypogaea [29] [34]. This clustered arrangement facilitates rapid evolution through unequal crossing-over, gene conversion, and tandem duplications, enabling plants to keep pace with evolving pathogen populations [35].

Evolutionary Patterns of Positive Selection in LRR Domains

Molecular Evidence for Positive Selection

Comparative genomic analyses provide compelling evidence that LRR domains in NBS-LRR proteins undergo positive selection. A genome-wide study of Arabidopsis NBS-LRR genes found substantial evidence of positive selection, with positively selected positions disproportionately located in the LRR domain (P < 0.001) [30]. The same study identified a nine–amino acid β-strand submotif within LRRs that is likely solvent-exposed and particularly targeted by positive selection.

The signature of positive selection is detected through elevated ratios of nonsynonymous to synonymous nucleotide substitutions (ω = dN/dS). When ω > 1, positive selection is inferred, indicating that amino acid-changing mutations are favored by natural selection [30]. This pattern contrasts with purifying selection (ω < 1) observed in constrained regions and neutral evolution (ω = 1). Maximum likelihood methods applied to NBS-LRR gene families have identified specific amino acid residues under positive selection, with the majority clustering in the LRR region [30].

Structural Basis for Diversification

The tertiary structure of LRR domains explains why specific residues become targets for positive selection. Based on structural determinations of diverse LRR-containing proteins including porcine ribonuclease inhibitor, individual LRRs form repeats of β-strand-loop and α-helix-loop units with non-leucine residues in the β-strands exposed to solvent [30]. These solvent-exposed residues potentially interact with pathogen ligands and thus determine recognition specificity [30].

Table 1: Distribution of Positively Selected Sites in NBS-LRR Proteins

Protein Domain	Proportion of Positively Selected Sites	Primary Evolutionary Force	Functional Implications
LRR domain	~70%	Positive selection/diversifying selection	Pathogen recognition specificity; binding surface diversification
NBS/NB-ARC domain	~25%	Purifying selection	Signal transduction switch function; ATP binding/hydrolysis
N-terminal domain (TIR/CC)	~5%	Purifying selection with some positive selection	Downstream signaling specificity
Specific LRR Submotifs
β-strand residues	Highly enriched for positive selection	Diversifying selection	Direct interaction with pathogen effectors
Between β-sheet regions	Indel variation common	Relaxed selection	Alters binding surface orientation

Beyond point mutations, LRR domains also exhibit substantial indel variation, creating elasticity in LRR length that could further influence resistance specificity [30]. This structural flexibility allows for continuous reshaping of the binding interface to track evolving pathogen ligands.

Comparative Evolutionary Genomics

The evolutionary patterns observed in LRR domains extend across plant species. In cassava, 228 NBS-LRR genes were identified, with 63% occurring in 39 clusters across the chromosomes [32]. Similarly, in peanut, 713 full-length NBS-LRR genes showed evidence of genetic exchange events both within and between subgenomes [34]. These studies consistently find that LRR domains evolve more rapidly than other protein regions and show signatures of adaptive evolution.

Relaxed selection pressure on LRR domains has been documented in cultivated species. In Arachis hypogaea, LRR domains were preferentially lost compared to its diploid ancestors, potentially explaining the lower disease resistance of the cultivated peanut [34]. This pattern highlights the trade-offs between maintaining diversity and potential fitness costs of highly polymorphic recognition systems.

Molecular Mechanisms of LRR Domain Evolution

Genomic Processes Generating Diversity

The remarkable diversity of LRR domains arises through several interconnected genomic processes:

• Gene duplication: Both segmental and tandem duplications create copies of NBS-LRR genes that subsequently diverge. Whole-genome duplication significantly contributes to NBS gene family expansion, as observed in Nicotiana species where 76.62% of N. tabacum NBS members could be traced to parental genomes [27].

• Unequal crossing-over: Within gene clusters, unequal crossing-over generates copy number variation and novel combinations of LRR sequences. This process maintains a diverse array of genes to retain advantageous resistance specificities [35].

• Gene conversion: Sequence exchange between homologous genes creates new LRR variants. Type I genes in lettuce evolve rapidly with frequent gene conversions, while Type II genes evolve more slowly with rare conversion events [29].

• Domain shuffling: Recombination events can create novel domain combinations, as evidenced by the discovery of proteins containing both TIR and CC domains in A. hypogaea, unlike its diploid ancestors [34].

Population-Level Dynamics

At the population level, NBS-LRR genes follow a birth-and-death model of evolution where gene duplication creates new copies (birth), while deleterious mutations or functional redundancy leads to pseudogenization and loss (death) [29]. This dynamic process maintains a reservoir of genetic variation that can be rapidly recruited when new pathogen strains emerge.

The rate of evolution varies significantly even within individual clusters, creating heterogeneous evolutionary patterns. For example, some NBS-LRR lineages evolve rapidly with frequent sequence exchange, while others evolve slowly with strong purifying selection, suggesting different functional constraints or recognition specificities [29].

Experimental Evidence: Structure-Function Relationships

Domain Interaction Studies

Functional studies of the potato Rx protein (a CNL) provide mechanistic insights into how domain interactions govern activation. Surprisingly, co-expression of the LRR and CC-NBS as separate domains resulted in a coat protein (CP)-dependent hypersensitive response, demonstrating that functional complementation can occur in trans [31]. Similarly, the CC domain complemented a version of Rx lacking this domain (NBS-LRR).

Co-immunoprecipitation experiments confirmed physical interactions between these domains: the LRR domain interacted physically with CC-NBS, and the CC domain interacted with NBS-LRR [31]. Both interactions were disrupted in the presence of the pathogen elicitor (CP), suggesting that effector recognition initiates conformational changes through sequential disruption of intramolecular interactions.

Table 2: Experimental Approaches for Studying LRR Evolution and Function

Method Category	Specific Techniques	Key Applications in LRR Research
Evolutionary Analysis	Maximum likelihood models for ω (dN/dS) estimation	Identifying sites under positive selection [30]
	Phylogenetic analysis	Reconstructing evolutionary relationships among NBS-LRR genes [34] [32]
	Population genetics	Assessing selection pressures in natural populations
Functional Characterization	Domain complementation assays	Testing functional interactions between separate domains [31]
	Co-immunoprecipitation	Detecting physical interactions between protein domains [31]
	Transient expression systems	Assessing hypersensitive response activation [31]
Genomic Approaches	Hidden Markov Model searches	Genome-wide identification of NBS-LRR genes [27] [33] [32]
	Synteny analysis	Tracing evolutionary history across related species [27]
	RNA-seq expression profiling	Identifying differentially expressed NBS-LRR genes during infection [27] [33]

Mechanistic Model of Activation

Based on experimental evidence, a refined model for NBS-LRR activation has emerged. In the resting state, intramolecular interactions between the LRR and other domains maintain the protein in an autoinhibited state. pathogen recognition induces conformational changes that disrupt these interactions, allowing the protein to adopt an active signaling state [31]. The precise molecular mechanisms differ between CNL and TNL proteins, but both classes appear to use related principles of autoinhibition and activation.

The LRR domain plays dual roles in both recognition and regulation. Beyond determining specificity, the LRR region maintains the protein in an inactive state until pathogen detection. This dual functionality creates evolutionary constraints that shape the diversification patterns observed in LRR sequences.

Research Methodologies and Experimental Protocols

Genomic Identification of NBS-LRR Genes

Standard protocols for genome-wide identification of NBS-LRR genes involve:

• HMMER searches: Using hidden Markov model profiles (e.g., PF00931 for the NB-ARC domain from the Pfam database) to scan predicted protein sequences [27] [33] [32]. Typical parameters include an E-value cutoff of 1×10⁻⁵ or more stringent.

• Domain annotation: Confirming identified candidates through additional domain searches against TIR (PF01582), LRR (PF00560, PF07723, PF07725, PF12799), and RPW8 (PF05659) profiles [32]. Coiled-coil domains are typically identified using Paircoil2 or similar tools with a P-score cutoff of 0.03 [32].

• Manual curation: Verifying domain architecture and filtering out false positives, particularly removing proteins with kinase domains but no relationship to NBS-LRR genes [32].

• Classification: Categorizing genes based on domain architecture into CNL, TNL, NL, RNL, and other subclasses [27] [33].

Detecting Positive Selection

Experimental workflow for identifying positive selection in LRR domains:

Figure 1: Experimental Workflow for Detecting Positive Selection in LRR Domains

The maximum likelihood approach implemented in programs such as CodeML (PAML package) differs substantially from earlier methods that partitioned codons a priori into predicted solvent-exposed and buried regions [30]. The ML method identifies specific amino acid residues under positive selection without prior assumptions about protein structure.

Functional Complementation Assays

Protocol for testing domain interactions and complementation:

• Construct design: Create expression vectors encoding separate protein domains (e.g., CC-NBS and LRR) with appropriate tags (e.g., HA epitope tag) [31].

• Transient expression: Co-express domain combinations in heterologous systems such as Nicotiana benthamiana leaves using Agrobacterium-mediated transformation [31].

• Phenotypic scoring: Assess hypersensitive response activation following elicitor treatment, typically within 24-72 hours post-infiltration [31].

• Interaction validation: Confirm physical interactions between domains through co-immunoprecipitation and western blotting [31].

The Scientist's Toolkit: Research Reagent Solutions

Table 3: Essential Research Reagents for Studying LRR Domain Evolution and Function

Reagent Category	Specific Examples	Applications and Functions
Bioinformatics Tools	HMMER v3.1b2 with PF00931 model	Identifying NBS-LRR genes in genome sequences [27] [32]
	MCScanX	Analyzing gene duplication and synteny [27] [33]
	KaKs_Calculator 2.0	Calculating nonsynonymous/synonymous substitution rates [27]
	MEME Suite	Identifying conserved protein motifs [33] [32]
Molecular Biology Reagents	Agrobacterium tumefaciens strains (GV3101)	Transient expression in plants [31]
	Epitope tags (HA, FLAG, Myc)	Protein detection and co-immunoprecipitation [31]
	Gateway or Golden Gate cloning systems	Modular vector construction for domain swapping
Analysis Software	MEGA11	Phylogenetic tree construction [27] [32]
	IQ-TREE 2.0.3	Maximum likelihood phylogenetics [33]
	DESeq2	Differential expression analysis from RNA-seq [33]
Database Resources	Pfam database	Protein domain identification [27] [33] [32]
	NCBI Conserved Domain Database	Domain verification [27] [32]
	Plant genome databases (Phytozome)	Genomic sequence retrieval [32]
6-hydroxyoctanoyl-CoA	6-hydroxyoctanoyl-CoA, MF:C29H50N7O18P3S, MW:909.7 g/mol	Chemical Reagent
3,5,7-Trioxododecanoyl-CoA	3,5,7-Trioxododecanoyl-CoA, MF:C33H52N7O20P3S, MW:991.8 g/mol	Chemical Reagent

The LRR domains of plant NBS-LRR proteins exemplify how positive selection drives molecular diversification in host-pathogen interactions. The evolutionary patterns observed—concentrated positive selection in solvent-exposed residues, indel variation creating length elasticity, and birth-and-death evolution in genomic clusters—collectively generate the recognition diversity necessary for plant immunity.

Future research directions should focus on integrating evolutionary knowledge with protein engineering approaches. The identification of positively selected sites provides targets for focused diversification in crop improvement programs. Additionally, understanding the balance between diversity generation and functional constraints will inform synthetic biology approaches to design novel resistance specificities.

The modular nature of NBS-LRR proteins, with separable recognition and signaling domains, offers opportunities for creating custom resistance genes by combining engineered LRR domains with appropriate signaling modules. As structural information becomes available for more plant NBS-LRR proteins, computational design of LRR domains with tailored specificities may become feasible, potentially revolutionizing approaches to crop disease management.

The study of LRR domain evolution thus provides not only fundamental insights into plant-pathogen coevolution but also a roadmap for engineering durable disease resistance in agricultural systems.

Nucleotide-binding site-leucine-rich repeat (NBS-LRR or NLR) genes constitute the largest family of plant disease resistance genes, playing crucial roles in effector-triggered immunity. Recent comparative genomic analyses across diverse angiosperms have revealed dynamic evolutionary patterns of NLR gene subfamilies, characterized by striking lineage-specific expansions and losses. This whitepaper synthesizes current understanding of TIR-NBS-LRR (TNL), CC-NBS-LRR (CNL), and RPW8-NBS-LRR (RNL) subfamily distributions across major plant lineages, highlighting the convergent reduction of TNL genes in monocots and specific dicot families, as well as the conservative evolution of RNL genes. The findings presented herein offer insights into the co-evolution between plants and their pathogens and provide a framework for targeted disease resistance breeding in crop species.

Plant immunity relies on a sophisticated surveillance system where nucleotide-binding site-leucine-rich repeat (NBS-LRR or NLR) proteins serve as critical intracellular immune receptors [36] [11]. These proteins detect pathogen effector molecules and initiate robust defense responses, culminating in effector-triggered immunity (ETI) [11] [37]. Angiosperm NLR genes are phylogenetically classified into three major subclasses: TIR-NBS-LRR (TNL) characterized by an N-terminal Toll/Interleukin-1 receptor domain, CC-NBS-LRR (CNL) featuring a coiled-coil domain, and RPW8-NBS-LRR (RNL) containing a Resistance to Powdery Mildew 8 domain [36] [38].

The evolution of these NLR subfamilies across angiosperms exhibits remarkable dynamism, with evidence of both rapid expansion and contraction in specific lineages [39] [40]. Genomic studies have revealed that NLR gene content can vary up to 66-fold among closely related species, reflecting continuous evolutionary arms races between plants and their pathogens [40]. This technical review synthesizes current comparative genomic evidence to elucidate patterns of lineage-specific expansion and loss of NLR gene subfamilies across angiosperms, with particular emphasis on monocot-dicot divergences.

Results

Dynamic Evolution of NLR Subfamilies Across Angiosperms

Comprehensive genomic analyses across diverse angiosperm species have revealed that the NLR gene family has experienced dynamic evolutionary patterns, including significant gene content variation and subclass composition shifts.

Table 1: NLR Gene Distribution Across Representative Angiosperm Species

Species	Total NLRs	TNL	CNL	RNL	Special Features	Citation
Arabidopsis thaliana	189-207	Present	Present	Present	Model dicot with all subclasses	[16] [11]
Oryza sativa (rice)	505	Absent	Present	Present	Complete TNL loss	[11]
Zea mays (maize)	-	Absent	Present	Present	Complete TNL loss	[11]
Triticum aestivum (wheat)	2151	Absent	Present	Present	Massive CNL expansion	[5] [11]
Euryale ferox (basal angiosperm)	131	73	40	18	TNL dominance	[38]
Nicotiana benthamiana	156	5	25	4	All subclasses present	[16]
Salvia miltiorrhiza	196	2	75	1	Extreme TNL/RNL reduction	[11]
Pinus taeda (gymnosperm)	311	89.3% of typical NLRs	-	-	TNL predominance	[11]

The evolutionary history of angiosperm NLR genes traces back to 3 anciently separated classes - RNL, TNL, and CNL - with evidence suggesting that 23 ancestral NBS-LRR lineages gave rise to current diversity through dynamic expansions [36]. Phylogenetic analysis of NLR genes from 22 angiosperm genomes supports this early divergence, with each subclass exhibiting distinct evolutionary patterns [36].

Recent studies have identified convergent NLR reduction in association with specific ecological adaptations. Aquatic, parasitic, and carnivorous plants consistently show contracted NLR repertoires, resembling the limited NLR expansion observed in green algae prior to land colonization [40]. This pattern suggests that ecological factors significantly influence NLR evolution independent of phylogenetic relationships.

Lineage-Specific Patterns in Monocots

Monocots exhibit the most striking example of lineage-specific NLR evolution, characterized by the complete absence of TNL genes in all investigated species. Genomic analyses of rice (Oryza sativa), maize (Zea mays), and wheat (Triticum aestivum) consistently demonstrate this TNL deficiency [11]. Wheat represents an extreme case with 2,151 NLR genes identified, all belonging to CNL and RNL subclasses [5] [11].

This TNL loss in monocots coincides with specific modifications in immune signaling components. Research has revealed a co-evolutionary pattern between NLR subclasses and plant immune pathway components, suggesting that deficiencies in TNL-specific signaling pathways may have facilitated TNL loss [40]. In particular, the EDS1-SAG101-NRG1 module, which is essential for TNL signaling, shows modifications in monocots that may explain this evolutionary pattern.

Expansion and Contraction Patterns in Dicot Families

Dicot species generally maintain all three NLR subclasses, though significant variation exists between families. The Salvia genus (Lamiaceae) demonstrates extreme reduction of TNL and RNL subfamilies, with Salvia miltiorrhiza possessing only 2 TNL and 1 RNL genes among 196 identified NLRs [11]. Comparative analysis across five Salvia species revealed complete absence of TNL subfamily members and limited RNL copies (1-2), significantly fewer than in other angiosperms like Arabidopsis thaliana and Vitis vinifera [11].

In Apiaceae species, comparative genomic analysis revealed dynamic NLR gene evolution with significant variation between species. Coriandrum sativum possesses 183 NLR genes, nearly double the number identified in Angelica sinensis (95 NLRs) [39]. Phylogenetic analysis demonstrated that these NLR genes derived from 183 ancestral NLR lineages and experienced different levels of gene loss and gain events during speciation [39].

Table 2: Evolutionary Patterns of NLR Subfamilies in Plant Lineages

Plant Group	TNL Evolution	CNL Evolution	RNL Evolution	Driving Forces
Basal Angiosperms	Moderate expansion	Moderate expansion	Conservative	Ancient pathogen pressures
Monocots	Complete loss	Extensive expansion	Conservative	Co-evolution with signaling pathways
Eudicots	Variable: expansion to near-complete loss	Generally expanded	Conservative	Lineage-specific adaptations
Aquatic Plants	Contracted	Contracted	Conservative	Reduced pathogen pressure
Carnivorous/Parasitic	Contracted	Contracted	Conservative ecological adaptations

Analysis of basal angiosperms provides insights into early NLR evolution. In Euryale ferox (Nymphaeales), TNL genes dominate the NLR repertoire (73 of 131 genes), suggesting early diversification of this subclass [38]. Gene duplication analysis revealed that segmental duplications acted as the major mechanism for NLR gene expansion in E. ferox, except for RNL genes, which were scattered without synteny loci, suggesting ectopic duplications [38].

Methodologies for Comparative Genomic Analysis of NLR Genes

Genome-Wide Identification of NLR Genes

Standardized protocols for NLR identification enable comparative analyses across species. The fundamental approach involves:

Hidden Markov Model (HMM) Searches: Using the NB-ARC domain (Pfam: PF00931) as query with optimized E-value thresholds (typically 1.0 for initial search, 0.0001 for verification) [27] [38] [39]. Multiple studies employed HMMER software (v3.1b2 or later) for this purpose [27].

Domain Verification and Classification: Candidate sequences are subjected to comprehensive domain analysis using:

• Pfam database (http://pfam.sanger.ac.uk/) for NBS, TIR, and LRR domains
• NCBI Conserved Domain Database (CDD) for CC domains
• SMART tool (http://smart.embl-heidelberg.de/) for additional domain validation [16]

Classification System: NLR genes are classified based on domain architecture into eight subfamilies: CN, CNL, N, NL, RN, RNL, TN, and TNL [27]. This detailed classification enables precise evolutionary comparisons.

Figure 1: Workflow for Genome-Wide Identification and Classification of NLR Genes

Phylogenetic Reconstruction and Evolutionary Analysis

Sequence Alignment and Phylogenetic Tree Construction:

• Multiple sequence alignment of NBS domains using ClustalW or MUSCLE [27] [38]
• Phylogenetic analysis with maximum likelihood method (IQ-TREE or MEGA) [38] [39]
• Model selection using ModelFinder and branch support with UFBoot2 (1000 replicates) [38]

Gene Duplication Analysis:

• Whole-genome duplication (WGD) analysis via self-BLASTP [27]
• Tandem and segmental duplication detection using MCScanX [27] [5]
• Syntenic block identification through reciprocal BLASTP searches [27]

Selection Pressure Analysis:

• Calculation of non-synonymous (Ka) and synonymous (Ks) substitution rates
• Use of KaKs_Calculator 2.0 with Nei-Gojobori evolutionary model [27]

Expression and Functional Analysis

Transcriptomic Analysis:

• RNA-seq data processing (Trimmomatic for quality control, Hisat2 for alignment) [27]
• Differential expression analysis (Cufflinks/Cuffdiff with FPKM normalization) [27]
• Validation via qRT-PCR for key NLR candidates [37]

Functional Validation Approaches:

• Virus-Induced Gene Silencing (VIGS) to assess gene function [5] [16]
• Heterologous expression in model systems (e.g., Arabidopsis) [27]
• Protein-protein interaction studies (yeast two-hybrid, co-immunoprecipitation) [5]

Table 3: Essential Research Reagents and Resources for NLR Gene Studies

Category	Specific Tool/Resource	Application	Key Features	Citation
Database Resources	ANNA (Angiosperm NLR Atlas)	Comparative genomics	>90,000 NLR genes from 304 angiosperm genomes	[5] [40]
	Pfam Database	Domain identification	Curated HMM profiles (e.g., PF00931 for NB-ARC)	[27] [16]
	NCBI CDD	Domain verification	Comprehensive domain database	[27] [16]
Bioinformatics Tools	HMMER Suite	Domain searches	Hidden Markov Model-based searches	[27] [38]
	MCScanX	Duplication analysis	Gene duplication and synteny analysis	[27] [39]
	OrthoFinder	Orthogroup analysis	Pan-genome comparative analysis	[5]
Experimental Resources	VIGS Vectors	Functional validation	Virus-Induced Gene Silencing	[5] [16]
	RNA-seq Platforms	Expression profiling	Transcriptome analysis under stress	[27] [5]
	PhasiRNA/miRNA tools	Regulatory studies	sRNA sequencing and analysis	[41]

Figure 2: Simplified NLR Immune Signaling Network

Discussion

The comparative genomic analyses synthesized in this review demonstrate that NLR gene evolution in angiosperms is characterized by dynamic birth-and-death processes with significant lineage-specific adaptations. The differential evolutionary patterns observed among NLR subclasses reflect their distinct functional roles in plant immunity.

The conservative evolution of RNL genes across angiosperms aligns with their role as "helper" NLRs involved in signal transduction downstream of "sensor" NLR activation [36] [38]. This functional constraint likely limits radical changes in RNL composition. In contrast, the extensive diversification of TNL and CNL genes corresponds to their function as pathogen sensors directly engaged in evolutionary arms races with rapidly evolving pathogen effectors.

The complete absence of TNL genes in monocots represents one of the most significant lineage-specific NLR adaptations. This loss appears to be associated with modifications in the EDS1-SAG101-NRG1 signaling module essential for TNL function [40]. However, exceptions exist, as some basal dicots like Aquilegia coerulea and asterid species in the lamiales lineage also lack TNL genes [36]. The long-term contraction of TNL genes during early angiosperm evolution may have facilitated their complete loss in certain lineages [36].

The convergent NLR reduction observed in aquatic, parasitic, and carnivorous plants suggests that ecological factors significantly influence NLR repertoire size independent of phylogenetic relationships [40]. This pattern indicates that changes in pathogen pressure associated with specialized lifestyles can drive NLR contraction.

Lineage-specific expansion and loss of NLR gene subfamilies represents a fundamental aspect of angiosperm evolution, reflecting continuous adaptation to pathogen pressures and ecological niches. The contrasting evolutionary patterns between monocots and dicots, particularly the complete absence of TNL genes in monocots, highlights the plasticity of plant immune systems. These findings have significant implications for disease resistance breeding, suggesting that strategies must be tailored to the specific NLR composition of target species. Future research should focus on elucidating the functional consequences of these lineage-specific NLR profiles and their interactions with signaling pathway components.

From Genomes to Genes: Computational and Functional Pipelines for NBS Discovery

In plant immunity research, nucleotide-binding site-leucine-rich repeat (NBS-LRR) genes constitute the largest and most important class of disease resistance (R) genes, enabling plants to detect pathogens and trigger robust defense responses. The identification and characterization of these genes rely heavily on bioinformatics workflows centered on domain-based analysis. This technical guide details established methodologies using HMMER, Pfam, and InterProScan for comprehensive identification of NBS-LRR genes, with direct application to plant immunity studies. We provide experimental protocols from recent genome-wide investigations, visual workflows, and reagent solutions to equip researchers with practical tools for resistance gene discovery.

Plants employ a sophisticated two-layered immune system where intracellular NBS-LRR proteins mediate effector-triggered immunity (ETI), recognizing pathogen-secreted effectors to activate defense responses often accompanied by hypersensitive cell death [11] [12]. These proteins characteristically contain a conserved nucleotide-binding site (NBS) domain and a C-terminal leucine-rich repeat (LRR) domain, with additional variable N-terminal domains defining major subfamilies [15]. The NBS domain facilitates ATP binding and hydrolysis, functioning as a molecular switch for immune activation, while the LRR domain is primarily responsible for pathogen recognition through protein-protein interactions [4] [12].

The NBS-LRR family is one of the largest gene families in plants, with significant variation in size and composition across species. For example, Arabidopsis thaliana contains approximately 150-159 NBS-LRR genes, while rice (Oryza sativa) possesses 505-653, and tobacco (Nicotiana benthamiana) has 156 [11] [4] [16]. This expansion and diversification reflects an evolutionary arms race between plants and their pathogens, making the identification and characterization of these genes crucial for understanding plant immunity and developing disease-resistant crops.

Traditional bioinformatics workflows for NBS-LRR identification leverage complementary tools in a pipeline that progresses from initial sequence identification to detailed domain annotation:

• HMMER: Performs sensitive sequence searches using Hidden Markov Models to identify candidate NBS-containing proteins in genomic or proteomic datasets [16].
• Pfam: Provides curated multiple sequence alignments and HMMs for protein domains, including the essential NB-ARC domain (PF00931) used for initial identification [16].
• InterProScan: Integrates multiple protein signature recognition methods from various databases, enabling comprehensive domain architecture analysis and functional prediction [42].

Research Reagent Solutions

Table 1: Essential Bioinformatics Resources for NBS-LRR Identification

Resource Name	Type	Primary Function	Key Identifier/Database
Pfam Database	Protein Family Database	Provides curated HMM profiles for domain identification	NB-ARC domain (PF00931) [16]
NCBI CDD	Conserved Domain Database	Confirms domain presence and completeness	CDD accession numbers [18]
InterPro	Integrated Database	Unifies protein family, domain, and functional site information	InterPro entries [42]
SMART	Protein Domain Annotation	Validates domain composition and architecture	Domain boundaries [16]
PlantCARE	cis-Element Database	Identifies regulatory elements in promoter regions	Hormone and stress-responsive elements [11] [16]

Experimental Workflows and Protocols

Genome-Wide Identification Protocol

Recent studies across multiple plant species have established a standardized workflow for NBS-LRR identification:

• HMMER Search Implementation

  • Retrieve the NB-ARC (PF00931) HMM profile from the Pfam database [16].
  • Perform HMMsearch against the target proteome using HMMER v3.1b2 or later with an expectation value (E-value) threshold of <1×10⁻²⁰ [16] [18].
  • Extract candidate sequences containing the NBS domain for further analysis.

• Domain Validation and Classification

  • Submit candidate sequences to the Pfam database, SMART tool, and NCBI CDD for domain verification [16] [18].
  • Confirm the presence of complete NBS domains with E-values <0.01 [16].
  • Classify sequences into subfamilies (CNL, TNL, NL, CN, TN, N) based on the presence/absence of N-terminal (CC, TIR, RPW8) and C-terminal (LRR) domains [11] [16].

• Manual Curation

  • Remove duplicate entries and fragments lacking complete domain structures.
  • Verify atypical NBS-LRR proteins (lacking either N-terminal or LRR domains) through multiple database searches [11].

Table 2: NBS-LRR Classification Based on Domain Architecture

Subfamily	N-Terminal Domain	NBS Domain	LRR Domain	Functional Role
TNL	TIR (Toll/Interleukin-1 Receptor)	Present	Present	Pathogen recognition & immunity signaling [15]
CNL	CC (Coiled-Coil)	Present	Present	Pathogen recognition & immunity signaling [15]
RNL	RPW8 (Resistance to Powdery Mildew 8)	Present	Present	Helper NLR in defense signaling [11]
NL	None or undefined	Present	Present	Pathogen recognition with divergent N-terminus [16]
TN	TIR	Present	Absent	Potential adaptor/regulator [16]
CN	CC	Present	Absent	Potential adaptor/regulator [16]
N	None	Present	Absent	Truncated forms, function not fully characterized [16]

Workflow Visualization

Case Studies in Plant Immunity Research

Medicinal Plant Analysis: Salvia miltiorrhiza

A 2025 study identified 196 NBS-LRR genes in the medicinal plant Salvia miltiorrhiza, with only 62 possessing complete N-terminal and LRR domains [11]. The workflow employed:

• HMM profiles from InterPro for initial identification
• Phylogenetic analysis integrating NLRs from model plants
• Promoter analysis revealing cis-acting elements related to plant hormones and abiotic stress
• Expression profiling linking specific SmNBS-LRRs to secondary metabolism

This study revealed a marked reduction in TNL and RNL subfamily members in Salvia species compared to other angiosperms, suggesting lineage-specific evolution of immune receptors [11].

Model Plant Analysis: Nicotiana benthamiana

A comprehensive 2025 analysis of Nicotiana benthamiana identified 156 NBS-LRR homologs using HMMsearch with the NB-ARC domain (PF00931) [16]. The experimental protocol included:

• HMMER search with E-values <1×10⁻²⁰ followed by manual Pfam verification
• Classification into 5 TNL-type, 25 CNL-type, 23 NL-type, 2 TN-type, 41 CN-type, and 60 N-type proteins
• Subcellular localization prediction showing 121 NBS-LRRs in cytoplasm, 33 in plasma membrane, and 12 in nucleus
• Gene structure analysis revealing most NBS-LRR genes contained few introns
• Promoter analysis identifying 29 shared cis-element types and 4 unique to irregular-type NBS-LRR genes

Comparative Genomics in Nicotiana Species

A recent multi-species study identified 1,226 NBS genes across three Nicotiana genomes, with 603 in allotetraploid N. tabacum and approximately 45.5% containing only the NBS domain [18]. The methodology featured:

• HMMER searches with PF00931 followed by NCBI CDD validation
• Phylogenetic analysis using MUSCLE and MEGA11
• Whole-genome duplication analysis via MCScanX
• Expression profiling during resistance to black shank and bacterial wilt

This study demonstrated that 76.62% of NBS members in N. tabacum could be traced to parental genomes, with whole-genome duplication significantly contributing to NBS family expansion [18].

Advanced Analysis and Integration

Phylogenetic and Structural Analysis

Following identification, comprehensive characterization of NBS-LRR genes includes:

• Multiple Sequence Alignment: Using Clustal W or MUSCLE with default parameters [16] [18]
• Phylogenetic Tree Construction: Implementing maximum likelihood method in MEGA7/MEGA11 with 1000 bootstrap replicates [16] [18]
• Motif Analysis: Predicting conserved motifs with MEME suite (motif count=10, width=6-50 amino acids) [16]
• Gene Structure Visualization: Using TBtools or similar platforms to examine exon-intron organization [16]

Expression and Functional Analysis

Integration of functional data enhances the interpretation of NBS-LRR genes:

• Transcriptome Analysis: Mapping RNA-seq reads with Hisat2, quantifying expression with Cufflinks, and identifying differentially expressed genes with Cuffdiff [18]
• Subcellular Localization: Predicting localization using CELLO v.2.5 and Plant-mPLoc [16]
• Physicochemical Characterization: Calculating molecular weight and pI using EXPASY ProtParam [16]
• Promoter Analysis: Identifying cis-regulatory elements with PlantCARE database (1500bp upstream of ATG) [16]

Traditional bioinformatics workflows centered on HMMER, Pfam, and InterProScan provide robust, standardized methodologies for domain-based identification of NBS-LRR genes in plant immunity research. The integration of these tools enables comprehensive characterization of this crucial gene family, from initial identification through structural, evolutionary, and expression analyses. Recent applications in species ranging from medicinal plants to model organisms demonstrate the continued utility of these approaches for elucidating plant immune systems and identifying potential resistance genes for crop improvement. As genomic resources expand, these established workflows remain fundamental to advancing our understanding of plant-pathogen interactions and developing sustainable disease control strategies.

Plant immunity relies on a sophisticated innate immune system where Resistance genes (R-genes) play a pivotal role in effector-triggered immunity (ETI), enabling plants to recognize specific pathogen effectors and mount a robust defense response [43] [44]. Among the major classes of R-genes, those encoding nucleotide-binding site leucine-rich repeat (NBS-LRR or NLR) proteins constitute the largest and most prominent family [4] [35]. These intracellular immune receptors are characterized by a central nucleotide-binding site (NBS or NB-ARC) domain, which acts as a molecular switch regulated by ADP-ATP exchange, a C-terminal leucine-rich repeat (LRR) domain involved in pathogen recognition, and a variable N-terminal domain that dictates downstream signaling pathways [44] [35]. The NBS domain itself contains several conserved motifs—P-loop, kinase-2, kinase-3a, and GLPL—that facilitate nucleotide binding and are crucial for the conformational changes that activate defense signaling [44].

The genomic architecture of NBS-LRR genes reveals their dynamic evolutionary history. They are frequently organized in clusters of closely duplicated genes distributed unevenly across plant genomes, a arrangement that facilitates rapid evolution of new pathogen specificities [4] [35]. This gene family exhibits remarkable diversity across plant species, ranging from approximately 50 members in papaya to over 650 in rice (Oryza sativa), reflecting varying evolutionary pressures and pathogen landscapes [4]. Furthermore, NBS-LRR genes are classified into distinct subclasses based on their N-terminal domains, primarily TIR-NBS-LRR (TNL) containing a Toll/interleukin-1 receptor domain and CC-NBS-LRR (CNL) featuring a coiled-coil domain, with TNL genes being predominantly absent from monocot genomes [4] [35]. Understanding the function and diversity of these NBS domain genes is therefore fundamental to unraveling the genetic basis of plant resistance and developing sustainable crop protection strategies.

The Computational Challenge of R-Gene Identification

Limitations of Traditional Identification Methods

Traditional methods for identifying NBS-LRR genes have primarily relied on alignment-based approaches using tools such as BLAST, InterProScan, HMMER, and various motif identification programs [43]. While these methods have been invaluable in early gene discovery efforts, they face significant limitations, particularly when dealing with newly sequenced plant genomes. Similarity-based methods often fail to identify R-genes with low sequence homology to known references, which is particularly problematic given the rapid evolution and diversification of this gene family [43]. The unique genomic structure of R-genes further complicates their identification, as they are often organized in clusters with numerous similar sequences that can challenge genome assembly and lead to fragmented annotations [43]. Additionally, their typically low expression levels make transcriptome-based prediction unreliable, and they can be misannotated as repetitive elements during genome annotation processes [43].

The Promise of High-Throughput Computational Approaches

The limitations of traditional methods, coupled with the exponential growth of genomic data, have created an pressing need for more sophisticated, high-throughput computational approaches for R-gene discovery. High-throughput screening (HTS) refers to the rapid automated testing of large numbers of samples or data points, with capabilities ranging from 10,000 to over 100,000 assays per day in biological contexts [45] [46]. The adaptation of HTS principles to computational genomics enables the systematic analysis of entire genomes for R-gene content, dramatically accelerating the pace of discovery compared to labor-intensive experimental approaches. Machine learning and deep learning represent the next evolution of these high-throughput capabilities, offering the potential to identify complex, non-linear patterns in sequence data that escape detection by traditional homology-based methods [43] [47]. These approaches can extract higher-level features from raw protein sequences, enabling classification based on learned characteristics rather than explicit similarity thresholds [43].

PRGminer: A Deep Learning Framework for High-Throughput R-Gene Prediction

PRGminer represents a cutting-edge deep learning-based tool specifically designed for high-throughput prediction of plant resistance genes [43]. Implemented as a two-phase classification system, PRGminer first identifies candidate resistance proteins from input sequences, then categorizes them into specific R-gene classes, providing a comprehensive solution for genome-scale R-gene annotation.

Table 1: PRGminer's Two-Phase Prediction Architecture

Phase	Function	Classification Categories	Key Features
Phase I	Initial R-gene identification	R-gene vs. Non-R-gene	Binary classification using dipeptide composition; Filters out non-R-genes
Phase II	R-gene categorization	8 major R-gene classes: CNL, TNL, KIN, RLP, LECRK, RLK, LYK, TIR	Multi-class classification based on domain architecture

The workflow begins with Phase I, where input protein sequences are classified as either R-genes or non-R-genes using a deep learning model trained on dipeptide composition features [43]. Sequences identified as non-R-genes are excluded from further analysis, while predicted R-genes proceed to Phase II. In this second phase, the tool performs fine-grained classification into eight distinct R-gene classes based on their domain architectures and sequence characteristics: CNL (Coiled-coil, Nucleotide-binding site, Leucine-rich repeat), TNL (Toll/interleukin-1 receptor, NBS, LRR), KIN (Kinase domain), RLP (Leucine-rich repeat and Transmembrane domains with cytoplasmic region), LECRK (Lectin, Kinase, and Transmembrane domains), RLK (Extracellular Leucine-rich repeat and Kinase domains), LYK (LysM domain, Kinase, and Transmembrane domains), and TIR (Toll/interleukin-1 receptor domain) [43].

Performance and Validation

PRGminer has demonstrated exceptional performance in both training and independent validation tests. In Phase I, using dipeptide composition features, the tool achieved an accuracy of 98.75% in k-fold training/testing and 95.72% on independent testing, with high Matthews correlation coefficient values of 0.98 and 0.91 respectively [43]. Phase II classification maintained this high standard with an overall accuracy of 97.55% in k-fold training/testing and 97.21% on independent testing, with MCC values of 0.93 and 0.92 respectively [43]. These results indicate that PRGminer outperforms traditional alignment-based methods and previous machine learning approaches, providing a robust and accurate solution for large-scale R-gene prediction.

Figure 1: PRGminer two-phase workflow for R-gene prediction and classification

Experimental Framework for NBS-LRR Gene Analysis

Genome-Wide Identification of NBS Domain Genes

Comprehensive identification of NBS domain genes across plant species requires systematic bioinformatics protocols. The following methodology has been validated in large-scale comparative studies analyzing over 12,000 NBS-domain-containing genes across 34 plant species [5]:

• Data Collection: Obtain latest genome assemblies from public databases (NCBI, Phytozome, Plaza) for target species. Selection should consider phylogenetic diversity and ploidy levels.
• Domain Screening: Use PfamScan with the Pfam-A.hmm model and an e-value cutoff of 1.1e-50 to identify genes containing the NB-ARC domain (PF00931). All genes with this domain are considered NBS genes for further analysis.
• Architecture Classification: Analyze domain architecture of identified NBS genes using a standardized classification system. Group genes with similar domain patterns into classes, identifying both classical (NBS, NBS-LRR, TIR-NBS, TIR-NBS-LRR) and species-specific structural patterns.
• Evolutionary Analysis: Perform orthogroup clustering using OrthoFinder with DIAMOND for sequence similarity searches and MCL for clustering. Construct phylogenetic trees using maximum likelihood methods (FastTreeMP) with 1000 bootstrap replicates.

Functional Validation of Candidate NBS-LRR Genes

Once candidate NBS-LRR genes are identified through computational approaches, experimental validation is essential to confirm their function in plant immunity. The following protocols provide a framework for functional characterization:

• Expression Profiling: Analyze expression patterns under biotic stress using RNA-seq data from databases (IPF, CottonFGD, Cottongen). Compare susceptible and resistant varieties to identify differentially expressed NBS genes. Calculate FPKM values and categorize expression into tissue-specific, abiotic stress-specific, and biotic stress-specific profiles.

• Genetic Variation Analysis: Identify sequence variants between susceptible and tolerant accessions. Focus on non-synonymous mutations in NBS domains that may affect protein function. For example, comparative analysis between cotton accessions identified 6,583 unique variants in a tolerant line versus 5,173 in a susceptible line [5].

• Protein Interaction Studies: Conduct protein-ligand and protein-protein interaction assays to validate interactions with pathogen effectors. For viral pathogens like cotton leaf curl disease, demonstrate strong interaction between NBS proteins and core viral proteins [5].

• Functional Genetic Tests: Implement virus-induced gene silencing (VIGS) to knock down candidate NBS genes in resistant plants. Monitor subsequent changes in disease susceptibility and pathogen titers. For example, silencing of GaNBS (OG2) in resistant cotton demonstrated its essential role in limiting virus accumulation [5].

Domain Architecture and Molecular Mechanisms of NBS-LRR Proteins

Structural Domains and Their Functions

NBS-LRR proteins consist of three primary domains that work in concert to recognize pathogens and activate defense signaling. The table below details the structure-function relationships of these core domains.

Table 2: Functional Domains of Plant NBS-LRR Proteins

Domain	Key Motifs/Features	Function in Plant Immunity	Conservation
NBS (NB-ARC)	P-loop, kinase 2, kinase 3a, GLPL	Serves as a molecular switch; ADP/ATP binding regulates activation state	Highly conserved across plant species
LRR	LxxLxLxxN/CxL repeats (x=any amino acid)	Pathogen recognition through protein-protein interactions; determines specificity	Highly diverse, under positive selection
N-terminal (TIR/CC)	TIR (~175 amino acids) or CC (coiled-coil)	Mediates downstream signaling; determines signaling pathway requirements	TIR mainly in dicots; CC in both monocots and dicots

The NBS domain functions as a molecular switch regulated by nucleotide exchange. In the inactive state, the domain binds ADP, maintaining the protein in an auto-inhibited conformation. Upon pathogen recognition, ADP is exchanged for ATP, triggering conformational changes that activate downstream signaling [44]. The LRR domain, with its solvent-exposed residues, is the primary determinant of recognition specificity and is under diversifying selection to evolve new pathogen specificities [4] [44]. The N-terminal domain dictates signaling specificity, with TIR domains activating pathways typically requiring EDS1 (Enhanced Disease Susceptibility 1) proteins, while CC domains can signal through both EDS1-dependent and independent pathways [44] [35].

Activation Mechanisms and Immune Signaling

The activation of NBS-LRR proteins involves a sophisticated molecular mechanism that translates pathogen recognition into defense activation. Research on the potato Rx protein (a CNL-type NBS-LRR) demonstrates that functional recognition can occur through interactions between separately expressed domains, where the CC-NBS and LRR regions expressed as separate molecules can complement each other in trans to confer a coat protein-dependent hypersensitive response [31]. This suggests that the intact Rx protein maintains autoinhibition through intramolecular interactions between its domains, which are disrupted upon pathogen recognition.

Figure 2: NBS-LRR protein activation mechanism from recognition to defense response

The current model of NBS-LRR activation proposes that effector recognition initiates a sequence of conformational changes that disrupt intramolecular interactions, particularly between the LRR and NBS domains, and between the CC and NBS domains [31]. This disruption enables nucleotide exchange from ADP to ATP, transitioning the protein from an inactive to an active state. The activated NBS-LRR protein then initiates downstream signaling cascades that culminate in defense responses such as the hypersensitive response (HR), production of reactive oxygen species, and expression of pathogenesis-related (PR) genes [31] [44]. Different NBS-LRR proteins localize to specific cellular compartments—including the cytoplasm, nucleus, plasma membrane, and endocytic vesicles—to recognize effectors and activate defense in the appropriate context [44].

Table 3: Research Reagent Solutions for R-gene Studies

Resource Category	Specific Tools/Databases	Application in R-gene Research
Gene Prediction Software	PRGminer, FINDER, AUGUSTUS, GeMoMa, GeneMark	Ab initio and homology-based gene prediction; Specialized R-gene identification
Genomic Databases	Phytozome, Ensemble Plants, NCBI Genome, Plaza	Source of genome assemblies and annotated genes for multiple species
Domain Analysis Tools	PfamScan, HMMER, InterProScan, Phobius, TMHMM	Identification of NBS, LRR, TIR, CC, and other domains in protein sequences
Expression Databases	IPF Database, CottonFGD, Cottongen, NCBI BioProject	Tissue-specific and stress-responsive expression patterns of NBS-LRR genes
Validation Tools	VIGS vectors, RNAi constructs, CRISPR-Cas9 systems	Functional characterization through gene silencing or genome editing

The integration of machine learning and deep learning approaches, exemplified by tools like PRGminer, represents a transformative advancement in the high-throughput prediction of plant resistance genes. By overcoming the limitations of traditional alignment-based methods, these computational frameworks enable rapid, accurate identification and classification of NBS-LRR genes across diverse plant species. As our understanding of NBS domain gene evolution, domain architecture, and activation mechanisms continues to deepen, the integration of multi-omics data with artificial intelligence will further accelerate the discovery of novel resistance genes. These advancements hold tremendous promise for guiding targeted breeding efforts and developing durable disease resistance strategies to enhance global food security in the face of evolving pathogen threats. The continued refinement of deep learning models, coupled with experimental validation, will be essential to fully harness the potential of plant immune receptors for sustainable agriculture.

Plant immunity relies on a sophisticated innate immune system where nucleotide-binding site-leucine-rich repeat (NBS-LRR) proteins serve as critical intracellular immune receptors. These proteins, encoded by one of the largest and most variable gene families in plants, are responsible for detecting pathogen effector proteins and initiating robust defense responses, including hypersensitive response and systemic acquired resistance. The genome-wide identification of these genes provides a comprehensive framework for understanding plant-pathogen co-evolution and enables the development of disease-resistant crops through marker-assisted breeding and genetic engineering.

This technical guide examines the methodologies and outcomes of genome-wide NBS-LRR identification in two biologically distinct but scientifically valuable systems: the medicinal plant Salvia miltiorrhiza and the model plant Nicotiana benthamiana. These case studies exemplify how genomic approaches can reveal structural and functional diversity in plant immune receptors across species with different biological characteristics and economic applications.

Biological Significance of NBS-LRR Genes in Plant Immunity

Domain Architecture and Classification

NBS-LRR proteins constitute the largest class of plant resistance (R) proteins, characterized by a conserved tripartite domain architecture:

• N-terminal domain: Either a Toll/Interleukin-1 receptor (TIR) domain, coiled-coil (CC) domain, or resistance to powdery mildew 8 (RPW8) domain
• Central nucleotide-binding site (NBS): Contains conserved motifs (P-loop, kinase 2, kinase 3a, GLPL) that bind and hydrolyze nucleotides
• C-terminal leucine-rich repeats (LRR): Highly variable region responsible for pathogen recognition specificity [44]

Based on their N-terminal domains, NBS-LRR proteins are classified into three major subfamilies: TNL (TIR-NBS-LRR), CNL (CC-NBS-LRR), and RNL (RPW8-NBS-LRR). This structural classification reflects functional specialization in immune signaling pathways and pathogen recognition mechanisms [28].

Molecular Mechanisms in Effector-Triggered Immunity

NBS-LRR proteins function as intracellular surveillance sensors that detect pathogen effector proteins through either direct binding or indirect monitoring of host components ("guard model"). Upon effector recognition, these proteins undergo conformational changes from ADP-bound inactive states to ATP-bound active states, exposing their N-terminal domains to initiate downstream signaling cascades. This activation leads to defense responses including reactive oxygen species bursts, phytohormone signaling changes, and often localized programmed cell death (hypersensitive response) to restrict pathogen spread [12] [44].

Table 1: Plant NBS-LRR Gene Family Sizes Across Species

Plant Species	Total NBS-LRR Genes	TNL Subfamily	CNL Subfamily	RNL Subfamily	Reference
Arabidopsis thaliana	149-159	94-98	50-55	-	[4]
Oryza sativa (japonica)	553	0	~553	-	[4]
Nicotiana benthamiana	156	5	25	4 (with RPW8)	[16]
Salvia miltiorrhiza	196	2	61	1	[11]
Vernicia montana	149	3	98*	-	[14]
Solanum tuberosum	435-438	65-77	361-370	-	[4]

*Includes genes with CC domains across different structural classes

Computational Methodologies for Genome-Wide Identification

Sequence Identification and Domain Validation

The standard pipeline for genome-wide identification of NBS-LRR genes involves sequential filtering based on conserved domain structures:

Figure 1: Workflow for identifying NBS-LRR genes

The Hidden Markov Model (HMM) profile for the NB-ARC domain (PF00931) serves as the primary search query against proteome or genome sequences. Candidate sequences meeting E-value thresholds (typically < 1e-20) undergo subsequent validation using multiple domain databases (SMART, Conserved Domain Database, Pfam) to confirm domain architecture and classify genes into subfamilies based on N-terminal and C-terminal domains [16].

Phylogenetic and Structural Analysis

Following identification, comprehensive analysis of gene family characteristics includes:

• Phylogenetic reconstruction using maximum likelihood methods (ClustalW, MEGA) to elucidate evolutionary relationships
• Conserved motif discovery with MEME suite to identify characteristic sequence patterns
• Gene structure analysis examining exon-intron organization using genomic and GFF3 annotation files
• Chromosomal mapping and synteny analysis to identify gene clusters and rearrangement events
• Promoter analysis (1500bp upstream) using PlantCARE to identify cis-regulatory elements [16]

These analytical approaches reveal evolutionary patterns including tandem duplications, segmental duplications, and gene loss events that have shaped the NBS-LRR repertoire in different plant lineages.

Case Study 1: Salvia miltiorrhiza

Genome-Wide Identification and Characteristics

Salvia miltiorrhiza (Danshen), a renowned medicinal plant in traditional Chinese medicine, produces valuable bioactive compounds (tanshinones and phenolic acids) and serves as a model for studying disease resistance in medicinal plants. A recent genome-wide analysis identified 196 NBS domain-containing genes in the S. miltiorrhiza genome, representing approximately 0.42% of all annotated protein-coding genes [11].

Among these, only 62 genes encoded complete NBS-LRR proteins with both N-terminal and LRR domains. The S. miltiorrhiza NBS-LRR family demonstrates remarkable subfamily distribution:

• CNL subfamily: 61 members
• RNL subfamily: 1 member
• TNL subfamily: 2 members

This distribution reveals a distinct subfamily reduction, particularly in TNL and RNL subfamilies, compared to other dicotyledonous plants like Arabidopsis thaliana which contains 94-98 TNL genes [11].

Chromosomal Distribution and Expression Patterns

NBS-LRR genes in S. miltiorrhiza show non-random chromosomal distribution with clustering in specific genomic regions, suggesting tandem duplication events as a major evolutionary mechanism. Expression profiling revealed that several SmNBS-LRR genes associate with secondary metabolism, indicating potential crosstalk between defense signaling and biosynthesis of medicinal compounds [11].

Promoter analysis identified abundant cis-acting elements related to plant hormone responses (jasmonic acid, salicylic acid, abscisic acid) and abiotic stresses, providing molecular evidence for the integration of defense signaling with environmental adaptation. This finding has significant implications for cultivating disease-resistant medicinal plants without compromising production of valuable secondary metabolites [11].

Case Study 2: Nicotiana benthamiana

Genome-Wide Identification and Structural Features

Nicotiana benthamiana, an established model plant for plant-pathogen interactions, possesses 156 NBS-LRR homologs in its genome, representing only 0.25% of its 61,328 annotated genes. Phylogenetic analysis classifies these genes into three major clades with the following distribution [16]:

• Clade I: 30 members
• Clade II: 49 members
• Clade III: 54 members

Structural classification reveals a diverse repertoire of NBS-LRR types:

• TNL-type: 5 proteins
• CNL-type: 25 proteins
• NL-type: 23 proteins (NBS-LRR without recognized N-terminal domain)
• TN-type: 2 proteins (TIR-NBS without LRR)
• CN-type: 41 proteins (CC-NBS without LRR)
• N-type: 60 proteins (NBS domain only)

Subcellular Localization and Physicochemical Properties

Subcellular localization predictions indicate that the majority of N. benthamiana NBS-LRR proteins (121) localize to the cytoplasm, with smaller numbers targeted to the plasma membrane (33) and nucleus (12). This diverse subcellular distribution reflects the multiple strategies employed by NBS-LRR proteins to detect pathogens in different cellular compartments [16].

Physicochemical characterization reveals substantial variation in molecular weight (31.48-220.15 kDa) and theoretical isoelectric points (pI 4.97-9.34), indicating functional specialization. Gene structure analysis shows that most N. benthamiana NBS-LRR genes contain few introns (0-2), consistent with the general characteristic of rapidly evolving resistance gene families [16].

Table 2: Comparative Analysis of NBS-LRR Families in Case Study Species

Characteristic	Salvia miltiorrhiza	Nicotiana benthamiana
Total NBS-containing genes	196	156
Typical NBS-LRR (with N-terminal & LRR)	62	53 (TNL+CNL+NL)
CNL subfamily	61	25
TNL subfamily	2	5
RNL subfamily	1	4 (with RPW8 domain)
Irregular types (missing domains)	134	103
Genome percentage	0.42%	0.25%
Chromosomal distribution	Non-random, clustered	Non-random, three phylogenetic clades

Experimental Validation of NBS-LRR Function

Functional Characterization Techniques

Several experimental approaches validate the function of identified NBS-LRR genes:

• Virus-Induced Gene Silencing (VIGS): A powerful reverse genetics approach to assess gene function by knocking down candidate NBS-LRR genes and evaluating changes in disease resistance phenotypes [14]
• Dual-Luciferase Transactivation Assays: Measure the ability of transcription factors to activate promoters of downstream defense-related genes [48]
• Hairy Root Transformation: An efficient system for functional characterization, particularly in species like S. miltiorrhiza [48]
• Expression Profiling: Quantitative RT-PCR analysis of gene expression patterns in different tissues and under various stress conditions [48]

Case Example: Functional Validation of VmNBS-LRR

In a related study on Vernicia montana resistance to Fusarium wilt, researchers identified an NBS-LRR gene (Vm019719) that was upregulated in resistant plants. Functional validation through VIGS demonstrated that silencing this gene significantly compromised resistance to Fusarium wilt, confirming its essential role in defense responses. Further analysis revealed that the expression of this resistance gene was activated by the transcription factor VmWRKY64, illustrating the complex regulatory networks controlling NBS-LRR gene expression [14].

Figure 2: NBS-LRR mediated immune signaling pathway

Research Reagent Solutions

Table 3: Essential Research Reagents for NBS-LRR Gene Characterization

Reagent/Tool	Application	Specific Examples
HMMER Software	Identification of NBS-domain containing genes	Search with NB-ARC (PF00931) HMM profile [16]
MEME Suite	Discovery of conserved protein motifs	Identification of P-loop, kinase 2, kinase 3 motifs [16]
Phylogenetic Tools	Evolutionary relationship analysis	ClustalW, MEGA7, FastTreeMP [5] [16]
VIGS Vectors	Functional characterization through gene silencing	TRV-based vectors for N. benthamiana [14]
Dual-Luciferase Systems	Promoter activation assays	Measurement of transcription factor activity [48]
Hairy Root Transformation	Functional studies in recalcitrant species	Agrobacterium rhizogenes-mediated transformation [48]

Genome-wide identification of NBS-LRR genes in Salvia miltiorrhiza and Nicotiana benthamiana reveals both conserved features and species-specific innovations in plant immune receptor repertoires. The remarkable reduction in TNL genes in S. miltiorrhiza compared to other dicots, and the diversity of irregular-type NBS genes in N. benthamiana, highlight the dynamic evolution of this gene family.

These case studies demonstrate that integrated computational and experimental approaches enable comprehensive characterization of plant immune receptor families. Future research should focus on:

• Elucidating the specific pathogen recognition spectra of identified NBS-LRR genes
• Engineering synthetic NBS-LRR genes with expanded recognition capabilities
• Understanding the coordination between different NBS-LRR subfamilies in immune signaling networks
• Exploring the potential trade-offs between disease resistance and specialized metabolism in medicinal plants

The methodologies and findings presented here provide a framework for similar studies in other plant species and contribute to the broader understanding of plant immunity mechanisms. This knowledge enables more precise breeding and engineering of disease-resistant crops, reducing reliance on chemical pesticides and enhancing agricultural sustainability.

The nucleotide-binding site-leucine-rich repeat (NBS-LRR) gene family represents the largest and most critical class of plant disease resistance (R) genes, serving as intracellular immune receptors that activate the plant immune system upon pathogen recognition [49] [11]. These genes encode proteins characterized by a conserved NBS (NB-ARC) domain and a C-terminal LRR domain, with N-terminal domains classifying them into distinct subfamilies: TNL (TIR-NBS-LRR), CNL (CC-NBS-LRR), and RNL (RPW8-NBS-LRR) [11] [16]. The NBS domain functions in ATP/GTP binding and signal transduction, while the LRR domain is responsible for specific pathogen recognition through direct or indirect interactions with pathogen effector molecules [49] [50].

Plants constantly face threats from biotic and abiotic stresses, necessitating sophisticated defense mechanisms. The plant immune system operates through a two-layered innate immune response: Pattern-Triggered Immunity (PTI) activated by cell-surface pattern recognition receptors (PRRs) recognizing pathogen-associated molecular patterns (PAMPs), and Effector-Triggered Immunity (ETI) mediated primarily by NBS-LRR proteins that detect specific pathogen effectors, often culminating in a hypersensitive response (HR) and programmed cell death to restrict pathogen spread [51] [50]. Recent research has revealed that NBS-LRR genes not only function in pathogen detection but are also intricately connected to hormonal signaling pathways and diverse stress responses, positioning them as critical integrators of plant defense signaling [52] [50].

This technical guide explores state-of-the-art methodologies for expression profiling and promoter analysis of NBS-LRR genes, establishing robust connections between their regulatory mechanisms and downstream stress responses. By integrating quantitative expression data, cis-element mapping, and experimental validation approaches, we provide researchers with comprehensive frameworks for deciphering the complex roles of NBS genes in plant immunity.

Expression Profiling of NBS-LRR Genes

Transcriptomic Approaches for Global Expression Analysis

RNA sequencing (RNA-Seq) provides a powerful, unbiased method for comprehensively profiling NBS-LRR gene expression across different tissues, developmental stages, and stress conditions. The typical workflow begins with RNA extraction from samples of interest, followed by cDNA library preparation and high-throughput sequencing. Bioinformatic analysis involves quality control, read alignment to a reference genome, and quantification of gene expression levels [49] [53].

Table 1: Key NBS-LRR Expression Changes Under Various Stress Conditions

Plant Species	Stress Condition	NBS-LRR Genes Regulated	Expression Pattern	Reference
Brassica oleracea (Cabbage)	Fusarium oxysporum infection	14 TNL genes	9 upregulated, 5 downregulated	[49]
Dendrobium officinale	Salicylic acid treatment	6 NBS-LRR genes (e.g., Dof020138)	Significantly upregulated	[53]
Glycine max (Soybean)	Phytophthora sojae infection	GmTNL16	Induced expression	[52]
Lathyrus sativus (Grass pea)	Salt stress (50 and 200 μM NaCl)	9 LsNBS genes	Majority upregulated, 3 downregulated at high concentration	[54]

In cabbage, RNA-Seq analysis revealed that 37.1% of TNL genes display highly specific or elevated expression in roots, with particularly strong root-specific expression for genes located on chromosome 7 (76.5%) [49]. This tissue-specific expression pattern suggests specialized roles for certain NBS-LRR clusters in soil-borne pathogen resistance. Following Fusarium oxysporum infection, expression profiling identified 14 TNL genes with significant transcriptional changes, providing candidates for further functional characterization [49].

In medicinal plants like Salvia miltiorrhiza, transcriptome analysis has established connections between SmNBS-LRR expression and secondary metabolism, suggesting potential crosstalk between defense pathways and the production of bioactive compounds such as tanshinones and phenolic acids [11].

Targeted Expression Validation

While RNA-Seq provides global expression profiles, quantitative real-time PCR (qPCR) offers precise, sensitive validation of specific NBS-LRR gene expression changes. The qPCR workflow includes RNA extraction, DNase treatment, cDNA synthesis using reverse transcriptase, and amplification with gene-specific primers using SYBR Green or TaqMan chemistry [54].

In grass pea, researchers selected nine LsNBS genes for qPCR validation under salt stress conditions. Most genes showed upregulation at both 50 and 200 μM NaCl, though three genes (LsNBS-D18, LsNBS-D204, and LsNBS-D180) displayed reduced or drastic downregulation at higher concentrations, revealing potential functional specialization within the NBS-LRR family for abiotic stress response [54].

For salicylic acid response studies in Dendrobium officinale, RNA-Seq identified 1,677 differentially expressed genes (DEGs) under SA treatment, including six significantly upregulated NBS-LRR genes. Weighted Gene Co-expression Network Analysis (WGCNA) further pinpointed Dof020138 as closely connected to pathogen recognition pathways, MAPK signaling, plant hormone signal transduction, and energy metabolism pathways [53].

Diagram 1: Experimental workflow for NBS-LRR gene expression profiling, integrating RNA-Seq and qPCR validation approaches under various experimental conditions.

Promoter Analysis of NBS-LRR Genes

Cis-Element Identification and Characterization

Promoter analysis represents a critical approach for understanding the transcriptional regulation of NBS-LRR genes. This process typically involves extracting ~2,000 bp sequences upstream of the translation start site and analyzing them using tools like PlantCARE or PLACE databases to identify cis-regulatory elements [49] [16].

Table 2: Common Cis-Elements in NBS-LRR Gene Promoters and Their Functions

Cis-Element	Sequence	Function	Plant Species Where Identified
SA-responsive elements	TTCACC	Salicylic acid responsiveness	Cabbage, Sweet orange, Tobacco
JA-responsive elements	TGACG	Jasmonic acid responsiveness	Sweet orange, Tobacco
ABA-responsive elements	ACGTG	Abscisic acid responsiveness	Grass pea, Sweet orange
Auxin-responsive elements	TGTCTC	Auxin responsiveness	Tobacco
Defense and stress responsiveness	TCA	Defense and stress signaling	Multiple species
Wound responsiveness	AAATTC	Wound-induced expression	Tobacco
MYB binding sites	TAACTG	Drought inducibility	Multiple species
MYC recognition sites	CATGTG	Dehydration response	Multiple species

In tobacco, analysis of 156 NBS-LRR genes revealed 29 shared types of cis-elements across typical and irregular-type NBS-LRR genes, with 4 unique elements specifically present in irregular-type NBS-LRR promoters, suggesting distinct regulatory mechanisms for different NBS-LRR subfamilies [16]. Similarly, promoter analysis in sweet orange identified abundant cis-elements related to plant hormones and abiotic stress in SmNBS genes [11].

The promoter analysis of grass pea NBS-LRR genes identified 103 transcription factors in upstream regions that govern the expression of nearby genes affecting plant excretion of salicylic acid, methyl jasmonate, ethylene, and abscisic acid, highlighting the complex regulatory networks controlling NBS-LRR gene expression [54].

Hormone-Responsive Cis-Elements and Signaling Integration

The presence of hormone-responsive cis-elements in NBS-LRR promoters provides a molecular basis for the integration of different defense signaling pathways. Research across multiple species consistently shows that NBS-LRR gene promoters are enriched for elements responsive to salicylic acid (SA), jasmonic acid (JA), ethylene (ET), and abscisic acid (ABA) [11] [16] [50].

In sweet orange, comprehensive promoter analysis of 111 NBS-LRR genes revealed complex profiles of cis-elements, with many genes containing multiple hormone-responsive elements that enable integrated responses to different pathogens and stress conditions [55]. This pattern is consistent with the known antagonistic and synergistic relationships between defense hormones, where SA-mediated pathways typically respond to biotrophic pathogens, while JA/ET pathways defend against necrotrophic pathogens and herbivores [50].

Diagram 2: Hormonal regulation of NBS-LRR genes through cis-elements and transcription factors, showing the complex interplay between different signaling pathways.

Linking NBS-LRR Genes to Hormonal Pathways

Salicylic Acid-Mediated Regulation

Salicylic acid (SA) serves as a primary defense hormone against biotrophic pathogens and establishes systemic acquired resistance (SAR). Multiple studies demonstrate direct connections between SA signaling and NBS-LRR gene regulation [53] [52] [50].

In soybean, the GmTNL16/gma-miR1510 regulatory pair participates in defense response against Phytophthora sojae through both JA and SA pathways. RNA sequencing analysis revealed that upon pathogen infection, reduced miR1510 expression enables induced expression of GmTNL16, leading to activation of SA pathway-associated genes including TGA transcription factors and PR (pathogenesis-related) genes [52]. This demonstrates how NBS-LRR genes can be integrated into established SA signaling cascades.

Similarly, in Dendrobium officinale, SA treatment significantly upregulated six NBS-LRR genes, with Dof020138 showing particularly strong connections to SA-mediated defense activation. This gene appears to function at the convergence point of multiple pathways, including pathogen recognition, MAPK signaling, and plant hormone signal transduction [53].

Jasmonic Acid and Ethylene Signaling Cross-Talk

The JA/ET defense pathways typically confer resistance against necrotrophic pathogens and herbivorous insects. While traditionally considered antagonistic to SA signaling, there is growing evidence of complex cross-talk between these pathways in regulating NBS-LRR gene expression [52] [50].

In the soybean-P. sojae interaction system, the GmTNL16/gma-miR1510 module activates both SA and JA pathways, with RNA-seq data showing enrichment of differentially expressed genes in both hormonal pathways. JA pathway components such as JAZ repressors and COI1 receptors respond to P. sojae infection in conjunction with NBS-LRR activation, suggesting coordinated rather than antagonistic regulation in certain pathosystems [52].

The presence of both SA-responsive and JA-responsive cis-elements in many NBS-LRR gene promoters enables this flexible signaling integration, allowing plants to fine-tune their immune responses based on the nature of the invading pathogen [16] [50].

Experimental Protocols

Comprehensive Promoter Analysis Workflow

Objective: Identify and characterize cis-regulatory elements in NBS-LRR gene promoters to link them with hormonal pathways and stress responses.

Materials and Reagents:

• Genomic DNA extraction kit
• PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/)
• PLACE database (https://www.dna.affrc.go.jp/PLACE/)
• Bioinformatics tools: TBtools, MEME suite, HMMER
• PCR reagents for promoter amplification

Methodology:

• Sequence Extraction: Obtain ~2000 bp upstream sequences of translation start sites from genomic databases [49]
• Cis-Element Screening: Submit promoter sequences to PlantCARE or PLACE databases for identification of regulatory elements
• Motif Analysis: Use MEME suite to identify conserved motifs with parameters: motif width 6-50 amino acids, maximum 10 motifs [16]
• Transcription Factor Binding Prediction: Identify potential TF binding sites using domain-specific databases
• Hormone-Responsive Element Mapping: Categorize elements by hormone responsiveness (SA, JA, ET, ABA)
• Comparative Analysis: Compare promoter architectures across NBS-LRR subfamilies and between species

Validation Approaches:

• Promoter-reporter fusions (GUS, GFP) in transgenic systems
• Electrophoretic mobility shift assays (EMSAs) for TF binding confirmation
• Chromatin immunoprecipitation (ChIP) with specific TF antibodies

Integrated Expression Profiling Protocol

Objective: Quantify NBS-LRR gene expression changes under hormonal treatments and stress conditions.

Materials and Reagents:

• RNA extraction kit (TRIzol or column-based)
• DNase I for genomic DNA removal
• Reverse transcriptase and primers for cDNA synthesis
• SYBR Green qPCR master mix
• Gene-specific primers for target NBS-LRR genes
• RNA-Seq library preparation kit (for transcriptomic approaches)

Methodology:

• Experimental Design:
  • Include multiple time points (0, 6, 12, 24, 48 hours post-treatment)
  • Apply hormonal treatments: SA (0.5-2 mM), JA (100 μM), ABA (100 μM)
  • Include pathogen inoculation: Fusarium oxysporum, Phytophthora sojae, etc. [49] [52]
  • Implement abiotic stresses: salt (50-200 mM NaCl), drought, temperature extremes

• RNA Extraction and Quality Control:

  • Extract total RNA using standard protocols
  • Verify RNA integrity (RIN > 8.0) and quantify precisely
  • Treat with DNase I to remove genomic DNA contamination

• Expression Analysis:

  • For RNA-Seq: Prepare libraries, sequence on Illumina platform, align reads, quantify expression [53]
  • For qPCR: Synthesize cDNA, perform amplification with gene-specific primers, use reference genes for normalization [54]
  • Include three biological replicates and three technical replicates

• Data Analysis:

  • Identify differentially expressed genes (fold-change > 2, FDR < 0.05)
  • Perform cluster analysis for co-expressed genes
  • Conduct promoter enrichment analysis for co-regulated genes
  • Integrate expression data with cis-element profiles

The Scientist's Toolkit: Essential Research Reagents

Table 3: Key Research Reagent Solutions for NBS-LRR Gene Analysis

Reagent/Category	Specific Examples	Function/Application	Technical Notes
Bioinformatics Tools	HMMER v3.1b2, MEME suite, PlantCARE, TBtools	Domain identification, motif discovery, promoter analysis, visualization	HMMER E-value cutoff < 1e-10 for NBS domain identification [49]
Sequencing & Analysis	Illumina platforms, Local TBLASTN, MUSCLE alignment	RNA-Seq, sequence similarity analysis, phylogenetic reconstruction	BLAST parameters: 90% similarity, 600 nt length threshold [54]
Expression Validation	SYBR Green qPCR reagents, gene-specific primers, reference genes	Quantitative expression validation	Normalize using multiple reference genes; include no-template controls
Hormone Treatments	Salicylic acid, Jasmonic acid, Abscisic Acid, Ethylene precursors	Elicitor treatments for signaling pathway analysis	Use appropriate concentrations: SA (0.5-2 mM), JA (100 μM) [53] [50]
Cloning & Transformation	Gateway cloning systems, Agrobacterium strains, GUS reporter assays	Promoter-reporter fusions, functional characterization	Use ~2000 bp promoter fragments for comprehensive cis-element coverage [49]
Domain Analysis	Pfam database, SMART tool, CDD database, Paircoil2	Protein domain identification and classification	Use Pfam NB-ARC domain (PF00931) for NBS identification [16]
trans-19-methyleicos-2-enoyl-CoA	trans-19-methyleicos-2-enoyl-CoA, MF:C42H74N7O17P3S, MW:1074.1 g/mol	Chemical Reagent	Bench Chemicals
11-Methyltridecanoyl-CoA	11-Methyltridecanoyl-CoA, MF:C35H62N7O17P3S, MW:977.9 g/mol	Chemical Reagent	Bench Chemicals

The integration of expression profiling and promoter analysis provides powerful insights into the regulation of NBS-LRR genes and their connections to hormonal signaling pathways and stress responses. The consistent identification of hormone-responsive cis-elements in NBS-LRR promoters across diverse plant species underscores the evolutionary conservation of these regulatory mechanisms. Meanwhile, expression studies demonstrate the precise transcriptional control of specific NBS-LRR genes in response to both biotic and abiotic challenges.

The experimental frameworks presented in this technical guide offer comprehensive approaches for unraveling the complex regulatory networks controlling plant immunity. By applying these integrated methodologies, researchers can advance our understanding of how NBS-LRR genes serve as central hubs in plant defense signaling, potentially leading to innovative strategies for crop improvement and sustainable disease management. The continuing expansion of genomic resources and analytical tools will further enhance our ability to decipher the intricate relationships between NBS-LRR regulation, hormonal pathways, and stress responses across the plant kingdom.

Plant immunity relies on a sophisticated surveillance system where nucleotide-binding site-leucine rich repeat (NBS-LRR) proteins serve as critical intracellular immune receptors. These proteins, encoded by the largest family of plant resistance (R) genes, are responsible for detecting pathogen effector molecules and initiating robust defense responses, including the hypersensitive response and systemic acquired resistance [28]. The NBS domain serves as a molecular switch within these proteins, binding and hydrolyzing ATP/GTP to regulate activation states, while the LRR domain facilitates pathogen recognition [56] [57]. Understanding the subcellular localization and physicochemical properties of NBS proteins is fundamental to elucidating their function in plant immunity, as these characteristics directly influence their interaction capabilities, activation mechanisms, and downstream signaling pathways.

The expanding genomic resources for numerous plant species, from model organisms to crops and medicinal plants, have enabled comprehensive genome-wide identification and characterization of NBS-LRR genes [16] [11] [58]. This guide provides researchers with integrated computational and experimental methodologies for determining key characteristics of NBS proteins, focusing on subcellular localization and physicochemical properties within the context of plant immunity research.

NBS Protein Classification and Structural Characteristics

Domain Architecture and Phylogenetic Classification

NBS-LRR proteins are categorized based on their N-terminal domains into distinct subclasses that influence their function and signaling pathways. The major classifications include:

• TNL proteins: Contain an N-terminal Toll/Interleukin-1 receptor (TIR) domain and typically signal through components like EDS1 and PAD4 [28].
• CNL proteins: Feature a coiled-coil (CC) domain at the N-terminus [28].
• RNL proteins: Possess an RPW8 domain and often function in downstream signaling [59].

Additionally, atypical or "irregular" NBS proteins exist that lack complete domain structures, such as TN (TIR-NBS), CN (CC-NBS), NL (NBS-LRR), and N (NBS-only) proteins, which may serve as adaptors or regulators for typical NBS-LRR proteins [16].

The distribution of these subclasses varies significantly across plant species, reflecting evolutionary adaptations to different pathogen pressures. For example, comprehensive genome-wide analysis of Nicotiana benthamiana identified 156 NBS-LRR homologs, comprising 5 TNL-type, 25 CNL-type, 23 NL-type, 2 TN-type, 41 CN-type, and 60 N-type proteins [16]. In contrast, Secale cereale (rye) possesses 582 NBS-LRR genes with a striking predominance of CNL subclass members (581) and only one RNL representative [58]. Medicinal plants like Salvia miltiorrhiza show similar trends, with 61 CNLs and only one RNL protein among 62 typical NLRs, indicating marked reduction or loss of TNL and RNL subfamilies [11].

Conserved Motifs and Functional Domains

The NBS domain contains several conserved motifs that facilitate nucleotide binding and are crucial for protein function. These include:

• P-loop: Involved in phosphate binding during nucleotide hydrolysis [56]
• Kinase-2 and Kinase-3a motifs: Collectively referred to as the NB subdomain [57]
• GLPL motif: Additional conserved region of unknown function [56]

MEME suite analysis typically identifies 8-10 conserved motifs dispersed throughout NBS protein sequences in both typical and irregular-type NBS-LRRs [16] [58]. These motifs demonstrate high conservation in their order and amino acid sequences across plant species, reflecting their functional importance [59].

Table 1: Distribution of NBS-LRR Genes Across Selected Plant Species

Plant Species	Total NBS Genes	TNL	CNL	RNL	Atypical	Reference
Nicotiana benthamiana	156	5	25	-	126	[16]
Hordeum vulgare (barley)	96	-	-	-	-	[56]
Secale cereale (rye)	582	-	581	1	-	[58]
Salvia miltiorrhiza	196	0	61	1	134	[11]
Akebia trifoliata	73	19	50	4	-	[59]

Computational Prediction of Subcellular Localization

Methodology and Tools

Predicting the subcellular localization of NBS proteins provides crucial insights into their function, as different compartments (cytoplasm, nucleus, plasma membrane) determine their accessibility to pathogen effectors and interaction partners. The following computational pipeline represents a standard approach for localization prediction:

Step 1: Primary Prediction with CELLO v.2.5

• Input: Protein sequence in FASTA format
• Method: CELLO employs a support vector machine (SVM)-based system trained on known protein localization patterns
• Parameters: Default settings typically used
• Output: Predicted localization with confidence scores

Step 2: Secondary Prediction with Plant-mPLoc

• Input: Same protein sequence
• Method: Plant-mPLoc incorporates functional domain information and evolutionary features using a multi-label classifier
• Parameters: Specify "plant" as the organism
• Output: Predicted localization with probability scores

Step 3: Results Comparison and Consensus Building

• Compare results from both tools
• Resolve discrepancies by favoring predictions with higher confidence scores
• Consider domain architecture (e.g., nuclear localization signals, transmembrane domains)

Application of this integrated approach to Nicotiana benthamiana NBS-LRR proteins predicted 121 proteins localized to the cytoplasm, 33 to the plasma membrane, and 12 to the nucleus [16]. This distribution aligns with the known functions of NBS proteins in perceiving intracellular pathogens and initiating signaling cascades.

Localization-Function Relationships

The subcellular localization of NBS proteins is not merely a structural characteristic but fundamentally linked to their biological function:

• Cytoplasmic localization: Enables interaction with cytoplasmic pathogen effectors and signaling components
• Nuclear localization: Facilitates transcription regulation in immunity, as demonstrated by certain NBS proteins that function as transcriptional regulators
• Plasma membrane association: May enable interaction with membrane-bound receptors or pathogen recognition at cell periphery

These localization patterns are not static and may change upon pathogen recognition, leading to re-localization that is essential for signal transduction [31].

Determination of Physicochemical Properties

Key Parameters and Computational Analysis

Physicochemical properties of NBS proteins influence their stability, interaction capabilities, and functional dynamics. The EXPASY ProtParam tool serves as the primary resource for calculating these properties from protein sequences:

Protocol for EXPASY ProtParam Analysis:

• Input Preparation: Protein sequences in FASTA format (without ambiguous residues)
• Analysis Parameters:
  • Molecular weight calculation based on average isotopic masses of amino acids
  • Theoretical pI computation using pKa values of amino acids
  • Instability index calculation predicting protein stability
  • Aliphatic index estimating thermostability
  • Grand average of hydropathicity (GRAVY) assessing hydrophobicity
• Interpretation:
  • Instability index < 40 predicts stable protein
  • Higher aliphatic index indicates greater thermostability
  • Negative GRAVY values suggest hydrophilic character

Table 2: Key Physicochemical Properties of NBS Proteins and Their Functional Implications

Property	Calculation Method	Functional Significance	Typical Range for NBS Proteins
Molecular Weight	Sum of amino acid residues	Influences diffusion rates, complex formation	50-150 kDa
Theoretical pI	pH where net charge is zero	Affects solubility, interaction partners	Varies by subclass
Instability Index	Amino composition stability	Predicts in vivo half-life	<40 (stable) to >40 (unstable)
Aliphatic Index	Relative volume occupied by aliphatic side chains	Indicates thermal stability	Varies by species adaptation
GRAVY	Hydropathicity average	Suggests membrane association potential	Negative values typical

Property Variations Across NBS Subclasses

Comparative analysis of physicochemical properties across NBS subclasses reveals both common features and unique characteristics. Studies on Nicotiana benthamiana NBS-LRR proteins demonstrated significant variation in properties like molecular weight, isoelectric point, and instability indices among different subclasses (TNL, CNL, NL, TN, CN, and N-types) [16]. These variations likely reflect functional specialization and adaptation to different pathogen recognition roles.

The molecular weights of NBS proteins typically range from 50-150 kDa, influenced by domain composition and LRR repeat numbers. Theoretical pI values show considerable diversity, potentially affecting protein-protein interaction specificities under different physiological conditions. Instability indices provide insights into protein turnover rates, which may be regulatory mechanisms in plant immunity.

Integrated Workflow for Comprehensive NBS Gene Characterization

From Genome to Functional Prediction

A complete characterization pipeline for NBS genes integrates multiple bioinformatic tools and databases to move from genomic sequences to functional predictions:

Comprehensive Characterization Protocol:

• Gene Identification

  • Tool: HMMER package with NB-ARC domain (PF00931) HMM profile
  • Parameters: E-value cutoff 1.0 for initial search
  • Validation: HMMscan against Pfam database (E-value < 0.0001)

• Domain Architecture Analysis

  • Tools: SMART, Conserved Domain Database (CDD), Pfam
  • Domains: CC (predicted by Coiledcoil with threshold 0.5), TIR, RPW8, LRR
  • Classification: Assign to TNL, CNL, RNL, or atypical categories

• Phylogenetic Analysis

  • Alignment: ClustalW for multiple sequence alignment
  • Tree Construction: MEGA software with maximum likelihood method
  • Validation: Bootstrap analysis (1000 replicates)

• Subcellular Localization

  • Tools: CELLO v.2.5 and Plant-mPLoc
  • Integration: Consensus prediction from both tools

• Physicochemical Properties

  • Tool: EXPASY ProtParam
  • Parameters: Molecular weight, pI, instability index, aliphatic index, GRAVY

• Gene Structure and Motif Analysis

  • Tools: MEME Suite for motif discovery, TBtools for visualization
  • Parameters: 10 motifs, width 6-50 amino acids

Advanced Computational Tools for NBS Protein Prediction

Recent advances in machine learning and deep learning have produced specialized tools for R gene prediction that complement traditional methods:

• PRGminer: A deep learning-based tool that predicts R proteins from sequence data with 98.75% accuracy in training/testing, utilizing dipeptide composition features [43]
• prPred: A support vector machine (SVM)-based predictor achieving 93.5% accuracy, integrating CKSAAP and CKSAAGP features [60]
• DRPPP: Another SVM-based tool specifically designed for plant R protein prediction [57]

These tools represent the next generation of prediction methods that can identify divergent R genes that might be missed by traditional domain-based searches.

Table 3: Essential Research Reagents and Computational Tools for NBS Protein Characterization

Category	Resource/Tool	Specific Function	Application in NBS Research
Bioinformatic Tools	HMMER v.3.0	Domain identification using hidden Markov models	Initial identification of NBS domains using PF00931 profile
	MEME Suite	Discovery of conserved protein motifs	Identification of P-loop, kinase-2, and other NBS motifs
	MEGA software	Phylogenetic analysis	Evolutionary relationships among NBS subclasses
	TBtools	Genomic data visualization	Gene structure, domain architecture visualization
Databases	Pfam Database	Protein family classification	Verification of NBS and other domain boundaries
	NCBI CDD	Conserved domain identification	Comprehensive domain architecture analysis
	PRGdb	Plant Resistance Gene database	Reference data for comparative analysis
Prediction Servers	CELLO v.2.5	Subcellular localization prediction	Determining cytoplasmic, nuclear, or membrane localization
	Plant-mPLoc	Plant-specific localization prediction	Enhanced accuracy for plant proteins
	EXPASY ProtParam	Physicochemical parameter calculation	Molecular weight, pI, stability indices
Experimental Validation	Confocal Microscopy	Protein localization visualization	Validation of computational localization predictions
	Bimolecular Fluorescence Complementation	Protein-protein interaction studies	Investigating NBS protein interactions in signaling

The integrated computational approaches described in this guide provide powerful methodologies for predicting subcellular localization and physicochemical properties of NBS domain genes. These predictions form the foundation for hypothesizing protein functions in plant immunity and designing targeted experimental validation. As genomic resources continue to expand across diverse plant species, and machine learning approaches become increasingly sophisticated, our ability to correlate sequence features with biological function will continue to improve. This knowledge is crucial for advancing fundamental understanding of plant immunity and for developing novel strategies for crop improvement through molecular breeding approaches.

The Nucleotide-Binding Site Leucine-Rich Repeat (NBS-LRR) gene family represents the largest and most critical class of plant disease resistance (R) genes, forming the cornerstone of the plant immune system. These genes enable plants to recognize diverse pathogens and initiate robust defense responses, playing an indispensable role in crop protection. Approximately 80% of cloned plant resistance genes belong to this family [61] [62], highlighting their predominant role in pathogen recognition. These genes encode proteins that function as intracellular immune receptors within the plant's Effector-Triggered Immunity (ETI) system [63]. During plant-pathogen interactions, NBS-LRR proteins directly or indirectly recognize specific pathogen effector molecules, triggering a complex defense signaling cascade that often culminates in a hypersensitive response (HR) to restrict pathogen growth and spread [64].

The NBS-LRR protein structure typically consists of three fundamental domains: a variable N-terminal domain that determines specific subfamily classification, a central nucleotide-binding site (NBS) domain responsible for energy transduction, and C-terminal leucine-rich repeats (LRRs) that facilitate pathogen recognition. Based on their N-terminal domain structures, NBS-LRR genes are classified into several major subfamilies: TIR-NBS-LRR (TNL) containing Toll/Interleukin-1 receptor domains, CC-NBS-LRR (CNL) featuring coiled-coil domains, and RPW8-NBS-LRR (RNL) with resistance to powdery mildew8 domains [64] [27]. The CNL and RNL subfamilies are collectively referred to as non-TNL (nTNL) [61]. This classification system reflects functional specialization within the plant immune system, with different subfamilies often employing distinct signaling pathways to activate defense responses.

Table 1: Major NBS-LRR Subfamilies and Their Characteristics

Subfamily	N-terminal Domain	Prevalence in Monocots vs. Dicots	Key Functional Role
CNL	Coiled-Coil (CC)	Abundant in both monocots and dicots	Pathogen recognition and defense activation
TNL	TIR (Toll/Interleukin-1 Receptor)	Primarily in dicots; absent in most monocots	Defense signaling with different pathway requirements
RNL	RPW8 (Resistance to Powdery Mildew 8)	Less abundant; present in both groups	Signal transduction helpers; often work with other NBS-LRRs

The genomic organization of NBS-LRR genes reveals important evolutionary patterns that directly impact breeding strategies. These genes are typically distributed unevenly across plant chromosomes, with approximately 54% forming gene clusters driven by tandem duplications and genomic rearrangements [61]. This clustering dynamic facilitates the rapid evolution of new recognition specificities through gene duplication and diversification, enabling plants to keep pace with evolving pathogen populations. Understanding these fundamental aspects of NBS-LRR gene structure, classification, and genomic organization provides the essential foundation for developing effective gene pyramiding strategies for durable disease resistance in crops.

The Rationale for Gene Pyramiding in Crop Improvement

Addressing the Durability Challenge in Plant Resistance

Conventional plant breeding approaches often rely on deploying single major resistance genes, which typically confer complete but race-specific resistance. While effective initially, this strategy has a critical limitation: it creates strong selection pressure that favors pathogen strains with corresponding virulence effectors, leading to frequent breakdown of resistance [65] [66]. This cyclical pattern of resistance deployment and failure has been described as an evolutionary "arms race" between crops and their pathogens, necessitating continuous identification and introgression of new resistance genes—a process that is both time-consuming and resource-intensive [66].

Gene pyramiding addresses this fundamental durability challenge by stacking multiple resistance genes with complementary functions into a single genotype. This approach provides several strategic advantages. First, it increases the genetic complexity required for pathogens to overcome host resistance, as pathogens must simultaneously accumulate multiple virulence mutations to successfully infect the plant [67] [65]. Second, pyramids incorporating genes with different modes of action—such as major R-genes, quantitative trait loci (QTLs), and genes regulating different defense signaling pathways—create a more robust, multi-layered defense system that is less vulnerable to evolutionary bypass by pathogen populations [66]. Research has demonstrated that pyramiding four quantitative trait loci (QTLs) in rice, each controlling different responses to Magnaporthe oryzae, conferred strong, non-race-specific, and environmentally stable resistance to blast disease [66].

Enhanced Efficacy and Stability of Pyramided Resistance

Beyond improving durability, gene pyramiding significantly enhances the efficacy and stability of disease resistance. When single QTLs are deployed individually, their resistance is often incomplete, environmentally sensitive, and exhibits substantial variation across different environments [66]. For example, in rice, individual QTLs showed coefficients of variation ≥15% across different field environments, demonstrating their instability when used alone [66]. However, when these same QTLs are combined through pyramiding, they exhibit cumulative effects that result in stronger, more consistent resistance with reduced environmental modulation [66].

The combination of different resistance mechanisms through pyramiding creates a synergistic defense network that is more difficult for pathogens to overcome. For instance, pyramids may include genes involved in different defense signaling pathways (e.g., salicylic acid, ethylene, or jasmonic acid pathways), creating a more comprehensive immune response [66]. Additionally, pyramiding allows breeders to combine genes with different recognition specificities, expanding the spectrum of pathogen strains that can be effectively recognized and controlled [67]. This multi-mechanism approach was successfully demonstrated in rice, where pyramiding achieved lesion areas of ≤1%—comparable to the durable resistant donor cultivar—while significantly reducing variation across environments and pathogen isolates [66].

Molecular Toolkit for NBS Gene Pyramiding

Marker-Assisted Selection Technologies

Marker-Assisted Backcross Breeding (MABB) has emerged as a powerful methodology for precise introgression of multiple NBS-LRR genes into elite crop varieties. This approach enables breeders to combine desired resistance genes while rapidly recovering the recurrent parent genome, significantly reducing the time required for variety development compared to conventional breeding methods [67]. The MABB process typically involves several generations of backcrossing with continuous marker-assisted selection to ensure the incorporation of target genes and minimize linkage drag.

The successful implementation of MABB relies on the availability of reliable molecular markers tightly linked to target NBS-LRR genes. These markers include:

• PCR-based co-dominant markers like the pTA248 marker for the Xa21 bacterial blight resistance gene in rice [67]
• Functional markers developed from polymorphic sites within resistance gene sequences, such as those available for the blast resistance gene Pi54 [67]
• SSR (Simple Sequence Repeat) markers for background selection to accelerate recovery of the recurrent parent genome [67]
• Gene-specific markers designed from characterized NBS-LRR gene sequences for precise selection [68]

Recent advances in sequencing technologies have further enhanced marker development, allowing for high-throughput genotyping and more efficient selection of optimal combinations of NBS-LRR genes in breeding programs.

Research Reagent Solutions for NBS Gene Pyramiding

Table 2: Essential Research Reagents for NBS Gene Pyramiding

Reagent/Category	Specific Examples	Function/Application
NBS Gene Donor Lines	IRBB60 (carrying xa5, xa13, Xa21), Tetep (carrying Pi54, qSBR7-1, qSBR11-1, qSBR11-2) [67]	Source of validated resistance genes for pyramiding programs
Molecular Markers	pTA248 for Xa21, gene-specific markers for Pi54 [67]	Tracking and selecting target genes during backcrossing
Polymorphic SSR Markers	Genome-wide distributed SSR markers [67]	Background selection to accelerate recovery of recurrent parent genome
Pathogen Isolates	Diverse races of Magnaporthe oryzae, Xanthomonas oryzae strains [66]	Phenotypic validation of resistance specificity and durability
HMM Profiles	PF00931 (NB-ARC domain) [64] [27]	Bioinformatics identification of NBS-LRR genes in genome sequences
Phthalimide-PEG1-amine	Phthalimide-PEG1-amine, MF:C12H14N2O4, MW:250.25 g/mol	Chemical Reagent
Linoelaidyl methane sulfonate	Linoelaidyl methane sulfonate, MF:C19H36O3S, MW:344.6 g/mol	Chemical Reagent

Experimental Framework for Pyramiding NBS Genes

Gene Identification and Marker Development Workflow

The initial phase of any pyramiding program involves comprehensive identification and characterization of candidate NBS-LRR genes. The following workflow outlines the standard experimental approach for this critical first step:

Figure 1: Experimental workflow for NBS-LRR gene identification and marker development

The process begins with Hidden Markov Model (HMM) searches using the PF00931 profile to identify NB-ARC domain-containing genes in target genomes [64] [27]. Candidate sequences then undergo comprehensive domain verification using databases such as Pfam and NCBI Conserved Domain Database (CDD) to confirm the presence of characteristic NBS-LRR domains (TIR, CC, LRR) [64] [27]. Following identification, genes are classified into subfamilies based on their domain architecture, and phylogenetic analyses are conducted to understand evolutionary relationships and inform optimal gene combinations for pyramiding [64] [69]. Finally, gene-specific markers are developed and validated on both donor and recipient genotypes to ensure robust selection in subsequent breeding generations.

Marker-Assisted Backcross Breeding Protocol

The implementation of marker-assisted backcross breeding for gene pyramiding follows a systematic protocol that ensures precise introgression of multiple target genes while maintaining the desirable genetic background of elite varieties. The following detailed protocol outlines the key steps:

Step 1: Initial Crosses

• Cross the recurrent parent (elite but susceptible variety) with donor parents containing target NBS-LRR genes
• For multiple gene pyramiding, this may involve sequential crosses with different donors or use of intermediate lines already containing some target genes [67]

Step 2: Marker-Assisted Selection in Segregating Generations

• Screen F1 and subsequent generations using gene-specific markers to confirm heterozygosity/homozygosity for target genes
• In each backcross generation, select plants positive for all target genes for further backcrossing
• Implement background selection using polymorphic SSR markers to accelerate recovery of the recurrent parent genome [67]

Step 3: Selfing and Homozygote Selection

• After 2-3 generations of backcrossing, self-select plants to generate homozygous lines
• Confirm homozygosity for all target genes using molecular markers
• In BC₃F₃ generation, select improved pyramided lines carrying all target genes/QTLs through molecular and phenotypic assays [67]

Step 4: Phenotypic Validation

• Evaluate homozygous pyramided lines for resistance against target pathogens under controlled and field conditions
• Assess agronomic performance and quality traits to ensure no negative impacts from gene pyramiding [67]

This protocol was successfully implemented in rice to pyramid seven genes/QTLs (xa5 + xa13 + Xa21 + Pi54 + qSBR7-1 + qSBR11-1 + qSBR11-2) into popular cultivars ASD 16 and ADT 43, resulting in lines exhibiting high degrees of resistance to bacterial blight, blast, and sheath blight diseases while maintaining the phenotypes of recurrent parents [67].

Case Studies: Successful Implementation of NBS Gene Pyramiding

Multi-Disease Resistance in Rice

One of the most comprehensive examples of NBS gene pyramiding comes from rice breeding for multiple disease resistance. Researchers successfully introgressed three bacterial blight resistance genes (xa5, xa13, and Xa21), one blast resistance gene (Pi54), and three sheath blight resistance QTLs (qSBR7-1, qSBR11-1, and qSBR11-2) into the genetic background of popular South Indian cultivars ASD 16 and ADT 43 [67]. This ambitious pyramiding program involved several strategic phases:

First, homozygous three-gene bacterial blight pyramided lines (xa5 + xa13 + Xa21) were developed in BC₃F₃ generation through MABB. These lines were then crossed with the donor Tetep to combine blast (Pi54) and sheath blight (qSBR7-1, qSBR11-1, and qSBR11-2) resistance [67]. The resulting improved pyramided lines carrying a total of seven genes/QTLs were selected through molecular and phenotypic assays, followed by rigorous evaluation under greenhouse conditions. The outcome was the development of nine lines in ASD 16 background and fifteen lines in ADT 43 background that exhibited high degrees of resistance to all three diseases while maintaining the desirable phenotypes of the recurrent parents [67].

This case study demonstrates several important principles of successful gene pyramiding: (1) the feasibility of stacking multiple resistance genes with different functions; (2) the importance of molecular markers for tracking multiple genes simultaneously; and (3) the necessity of comprehensive phenotypic validation to confirm resistance efficacy and maintain agronomic performance.

Quantitative Trait Locus Pyramiding for Durable Blast Resistance

Another compelling case study involves pyramiding quantitative trait loci (QTLs) for durable blast resistance in rice. Researchers developed near-isogenic lines representing all possible combinations of four QTL alleles (pi21, Pi34, qBR4-2, and qBR12-1) from the durably resistant cultivar Owarihatamochi in the genetic background of the susceptible cultivar Aichiasahi [66]. This systematic approach enabled precise evaluation of each QTL's individual contribution and their combined effects in a homogeneous genetic background.

The results demonstrated that while individual QTLs conferred incomplete resistance with substantial environmental sensitivity (coefficient of variation ≥15%), their combinations produced additive effects that progressively enhanced resistance [66]. Critically, the line with all four resistance QTLs (AA-4RQ) exhibited consistently strong resistance with minimal environmental modulation, achieving average lesion areas of ≤1%—comparable to the durable resistant donor cultivar Owarihatamochi [66]. This comprehensive study provided important evidence that pyramiding QTL alleles, each potentially controlling different response mechanisms to M. oryzae, confers strong, non-race-specific, and environmentally stable resistance, thereby constituting a durable defense system that avoids an evolutionary "arms race" with the pathogen [66].

Table 3: Comparison of Individual vs. Pyramided QTL Effects on Rice Blast Resistance

Genotype	Average Lesion Area (%)	Coefficient of Variation Across Environments	Resistance Stability
Aichiasahi (Recurrent Parent)	20-40%	High	Highly susceptible
Single QTL Lines	5-15%	≥15%	Environmentally sensitive
Two-QTL Pyramids	3-8%	Moderate reduction	Improved stability
Three-QTL Pyramids	1-4%	Further reduced	More consistent
Four-QTL Pyramid (AA-4RQ)	≤1%	Lowest observed	Highly stable, comparable to donor

Resistance Signaling Pathways in NBS-Mediated Immunity

The effectiveness of pyramided NBS-LRR genes depends on their integration into the plant's complex immune signaling network. Understanding these pathways is essential for designing optimal gene combinations that activate complementary defense mechanisms. The core signaling pathways involved in NBS-mediated immunity include:

Figure 2: Signaling pathways in NBS-LRR-mediated plant immunity

The NBS-LRR genes function within the Effector-Triggered Immunity (ETI) system, which is activated when specific pathogen effectors are recognized by their corresponding NBS-LRR receptors [63]. This recognition triggers a complex signaling cascade that often involves MAPK signaling pathways, plant hormone signal transduction pathways (particularly salicylic acid and ethylene), and leads to the activation of defense responses including the hypersensitive response and systemic acquired resistance [63]. Different NBS-LRR genes may utilize distinct signaling components; for example, the Pi34 blast resistance QTL in rice showed sensitivity to salicylic acid application, while the pi21 QTL did not respond to ethylene biosynthesis antagonists [66].

This pathway diversity has important implications for pyramiding strategies. Combining NBS-LRR genes that activate different signaling pathways can create a more robust and comprehensive defense system that is less vulnerable to pathogen suppression. Additionally, connecting NBS-LRR genes with upstream pathogen recognition and downstream defense execution creates an integrated immune network that provides multiple layers of protection against diverse pathogens.

Deployment Strategies for Durable Resistance

Comparative Analysis of Deployment Approaches

The durability of pyramided resistance depends not only on the specific gene combinations but also on how these improved varieties are deployed in agricultural systems. Mathematical modeling has been used to evaluate the long-term effectiveness of different deployment strategies, particularly under scenarios where virulence genes in pathogen populations may have no fitness costs [65]. Three primary deployment strategies have been analyzed:

Sequential Deployment involves using single-gene resistant varieties one after another, where the second variety is introduced when resistance of the first is overcome. Modeling has shown that this approach provides the shortest useful life of resistance genes among the strategies evaluated, particularly when the fraction of resistant host area is small [65].

Simultaneous Deployment involves cultivating multiple single-gene resistant varieties at the same time in a region. This approach extends the durability compared to sequential deployment but still allows for relatively rapid adaptation in pathogen populations when virulence genes carry no fitness costs [65].

Pyramiding Deployment involves stacking multiple resistance genes in a single variety. Modeling consistently identifies this as the most durable solution, as it requires pathogens to simultaneously accumulate multiple virulence mutations to successfully infect the host plant [65]. Field observations have confirmed many successes with this approach [65].

Integrated Resistance Management

For optimal durability, gene pyramiding should be integrated with other resistance management strategies. These include maintaining genetic diversity in cropping systems, incorporating partial resistance genes that may exert less selection pressure on pathogen populations, and implementing appropriate agricultural practices that reduce disease pressure. Additionally, monitoring pathogen populations for virulence shifts remains essential for proactive management of resistance genes.

Research suggests that a "mixed strategy" combining pyramided varieties with single-gene varieties in the landscape may help reduce selection pressure and extend the useful life of resistance genes [65]. However, careful consideration must be given to potential negative interactions, as deploying pyramided varieties together with single-gene varieties containing the same resistance genes could potentially compromise the durability of the pyramid if not properly managed [65].

The pyramiding of NBS-LRR genes represents a powerful strategy for developing crop varieties with durable, broad-spectrum disease resistance. By stacking multiple genes with complementary functions and signaling pathways, breeders can create robust defense systems that are difficult for pathogens to overcome through simple evolutionary adaptations. The success of this approach is clearly demonstrated in multiple case studies, particularly in rice, where pyramids of three to seven genes/QTLs have provided strong resistance to complex disease pressures [67] [66].

Future advancements in NBS gene pyramiding will likely be driven by several emerging technologies. Gene editing approaches such as CRISPR-Cas systems offer opportunities for precise manipulation of NBS-LRR genes, potentially allowing for custom design of resistance specificities and targeted improvement of existing genes [68]. Advanced genomic selection techniques will enable more efficient identification of optimal gene combinations based on comprehensive understanding of NBS-LRR gene networks and their evolutionary dynamics. Furthermore, synthetic biology approaches may permit the engineering of novel resistance genes with expanded recognition specificities, providing new genetic resources for pyramiding programs.

As these technologies mature, the strategic pyramiding of NBS-LRR genes will continue to be a cornerstone of crop improvement programs worldwide, contributing significantly to global food security by reducing yield losses to important plant diseases while promoting sustainable agricultural practices through reduced dependence on chemical pesticides.

Overcoming Challenges in NBS Gene Annotation and Functional Characterization

The identification and characterization of nucleotide-binding site (NBS) domain genes are fundamental to understanding plant immunity. These genes, particularly those encoding NBS-leucine-rich repeat (NLR) proteins, constitute the largest class of plant disease resistance (R) genes and play a central role in effector-triggered immunity (ETI) [28] [70]. However, accurate annotation of these genes remains a significant bioinformatic challenge due to their complex genomic architecture. NBS genes are often fragmented in genome assemblies, frequently reside in complex, rapidly evolving clusters, and exhibit substantial structural diversity across plant species [71] [5] [14]. These challenges are particularly pronounced in non-model organisms, including many medicinal plants, where genomic resources may be limited [11] [71]. This technical guide outlines robust experimental and computational strategies to address these annotation complexities, enabling more accurate characterization of NBS gene families in plant immunity research.

Genomic Architecture of NBS-LRR Genes

NBS-LRR genes encode modular proteins characterized by a conserved nucleotide-binding site (NBS) domain and C-terminal leucine-rich repeats (LRRs). Based on their N-terminal domains, they are classified into several subfamilies: TIR-NBS-LRR (TNL) with a Toll/interleukin-1 receptor domain, CC-NBS-LRR (CNL) with a coiled-coil domain, and RNL with an RPW8 domain [28] [70]. The central NB-ARC domain (nucleotide-binding adaptor shared by APAF-1, R proteins, and CED-4) functions as a molecular switch, while the LRR domain is responsible for pathogen recognition specificity [70] [72].

This inherent modularity and the presence of repeated motifs create substantial annotation challenges. The LRR domains, in particular, consist of multiple short, repeating units that are difficult to resolve with short-read sequencing technologies and complicate gene model prediction [28] [14].

• Gene Fragmentation: In genome assemblies, NBS genes are often split across multiple contigs or scaffolds. This occurs due to their large size (typically 3-5 kb for coding regions, but often larger with introns), complex exon-intron structures, and the presence of repetitive regions that complicate sequence assembly [71] [14].

• Complex Gene Clusters: NBS-LRR genes are frequently organized in tandem arrays and complex clusters within plant genomes. For example, bread wheat (Triticum aestivum) contains approximately 460 documented R genes, many clustered in specific genomic regions [28]. These clusters arise from frequent duplication and recombination events, leading to groups of highly similar paralogs that are difficult to disentangle in genome assemblies [5] [14].

• Structural Diversity and Atypical Genes: Beyond the typical NLR structure, many atypical configurations exist, including integrated domains (IDs) that function as decoys for pathogen effectors, and "sensor-helper" NLR pairs that require interaction between partner proteins for function [70]. These integrated domains can include WRKY, kinase, heavy metal-associated (HMA), and zinc-finger domains, further increasing annotation complexity [70].

• Transposable Element Associations: Miniature inverted-repeat transposable elements (MITEs) are often associated with NBS genes and can contribute to their evolution and regulation. In plant genomes, MITEs show a 20,000-fold variation in copy numbers between species and preferentially insert near genes, where they can influence expression and contribute to genome diversity [73].

Table 1: Common Challenges in Annotating NBS Domain Genes

Challenge	Impact on Annotation	Example from Literature
Gene Fragmentation	Incomplete gene models; missing domains	In Vernicia fordii, only 62 of 196 identified NBS genes contained complete N-terminal and LRR domains [11] [14]
Tandem Clusters	Difficulty distinguishing paralogs; misassembly	Salvia miltiorrhiza genome shows non-random, clustered distribution of NBS genes across chromosomes [11]
Domain Diversity	Failure to detect atypical architectures	Identification of NBS genes with novel domain fusions (TIR-NBS-TIR-Cupin_1, TIR-NBS-Prenyltransf) [5]
Species-Specific Expansions	Varying repertoire sizes complicate pipelines	TNL subfamily completely absent in monocots (O. sativa, Z. mays) and some eudicots (S. miltiorrhiza, V. fordii) [11] [14]

Computational Strategies for Enhanced Annotation

Domain-Based Bioinformatics Pipelines

Traditional domain-based approaches remain foundational for NBS gene identification. These methods utilize conserved protein motifs and domain architectures with tools such as:

• InterProScan: For comprehensive protein domain identification [28]
• HMMER: With custom hidden Markov models (HMMs) for NBS domain detection (e.g., PF00931) [11] [5] [14]
• PfamScan: For domain architecture analysis using the Pfam-A HMM library [5]

Specialized pipelines like DRAGO2/3, RGAugury, RRGPredictor, NLR-Annotator, and NLRtracker have been developed specifically for resistance gene annotation [28]. These tools scan genomes or proteomes for known domain combinations and structural motifs characteristic of resistance proteins.

Machine Learning and Deep Learning Approaches

Recent advances in machine learning (ML) and deep learning (DL) have significantly improved R-protein prediction. These methods can identify patterns and features that may be missed by traditional domain-based approaches:

• Sequence-based classifiers: Trained on known R-protein sequences to identify novel candidates [28]
• Structural prediction algorithms: Utilizing AlphaFold2 and RoseTTAFold for protein structure prediction [28]
• Ensemble methods: Combining multiple algorithms to improve prediction accuracy [28]

ML approaches are particularly valuable for identifying genes with atypical architectures or those that have diverged significantly from canonical sequences. However, challenges remain, including limited data quality, class imbalance in training datasets, and insufficient model interpretability [28].

Comparative Genomics and Orthology Analysis

Comparative approaches across multiple species provide powerful constraints for gene annotation. Key methods include:

• OrthoFinder: For identifying orthogroups across species [5]
• DIAMOND: For fast sequence similarity searches [5]
• MCL clustering algorithm: For gene family clustering [5]

These tools help distinguish recent lineage-specific expansions from conserved NBS genes, informing annotation quality. For example, a study analyzing 34 plant species identified 168 distinct classes of NBS domain architecture, including both classical and species-specific structural patterns [5].

Figure 1: Integrated Computational Annotation Workflow for NBS Genes

Experimental Validation and Functional Characterization

Transcriptomic Validation

RNA sequencing provides critical evidence to support computational predictions and resolve fragmented gene models. Key approaches include:

• Time-course experiments: Monitoring expression during pathogen infection [11] [5]
• Tissue-specific expression profiling: Identifying expression patterns across different organs [5]
• Stress-responsive expression analysis: Assessing induction under various biotic and abiotic stresses [11] [5]

For example, expression profiling of NBS genes in Gossypium hirsutum under cotton leaf curl disease (CLCuD) revealed distinct patterns between susceptible and tolerant accessions, helping to validate putative resistance genes [5].

Functional Characterization Methods

Several experimental approaches are essential for confirming the function of annotated NBS genes:

• Virus-Induced Gene Silencing (VIGS): A powerful tool for rapid functional assessment. For instance, VIGS of GaNBS (OG2) in resistant cotton demonstrated its role in virus defense [5]. Similarly, VIGS of Vm019719 in Vernicia montana confirmed its function in Fusarium wilt resistance [14].

• Transgenic Complementation: Introducing candidate genes into susceptible genotypes. The identification and transfer of the YPR1 gene from wild rice (Oryza rufipogon) into susceptible cultivars conferred resistance to multiple Xanthomonas oryzae strains [72].

• CRISPR/Cas9 Gene Editing: For validation through knockout studies. Knockout of YPR1 in common wild rice resulted in increased susceptibility to most Xoo strains, confirming its functional role in immunity [72].

Table 2: Experimental Protocols for Functional Validation of NBS Genes

Method	Key Steps	Applications in NBS Gene Research
VIGS	1. Clone gene fragment into VIGS vector2. Transform into Agrobacterium3. Infiltrate plants4. Challenge with pathogen5. Assess disease symptoms and gene expression	Functional assessment of Vm019719 in tung tree Fusarium wilt resistance [14]; Validation of GaNBS in cotton leaf curl disease response [5]
Transgenic Overexpression	1. Clone full-length CDS into expression vector2. Transform into susceptible host3. Select and regenerate transformants4. Inoculate with pathogen5. Evaluate resistance spectrum	Confirmation of YPR1 function in bacterial blight resistance in rice [72]
CRISPR/Cas9 Knockout	1. Design sgRNAs targeting gene of interest2. Clone into CRISPR/Cas9 vector3. Transform into plant material4. Screen for edited lines5. Phenotype under infection	Validation of YPR1 necessity for immunity in wild rice [72]

The Scientist's Toolkit: Research Reagent Solutions

Table 3: Essential Research Reagents for NBS Gene Characterization

Reagent/Tool	Function	Example Application
HMMER Software	Identification of NBS domains using hidden Markov models	Genome-wide identification of 196 NBS-LRR genes in Salvia miltiorrhiza [11]
OrthoFinder Package	Orthogroup inference and comparative genomics	Analysis of 12,820 NBS-domain-containing genes across 34 plant species [5]
VIGS Vectors	Transient gene silencing for functional validation	Silencing of Vm019719 to confirm role in Fusarium wilt resistance in Vernicia montana [14]
CRISPR/Cas9 Systems	Targeted gene knockout for functional analysis	Knockout of YPR1 in common wild rice to validate role in bacterial blight resistance [72]
P-MITE Database	Identification of miniature inverted-repeat transposable elements	Analysis of MITE impact on genome structure and gene regulation [73]
11-methylnonadecanoyl-CoA	11-methylnonadecanoyl-CoA, MF:C41H74N7O17P3S, MW:1062.1 g/mol	Chemical Reagent

Case Study: Annotation Pipeline in Medicinal Plants

Medicinal plants present particular challenges for NBS gene annotation due to complex genomes and limited genomic resources. A recent study on Salvia miltiorrhiza (Danshen) demonstrates an effective annotation pipeline:

• Genome Sequencing and Assembly: Use of long-read sequencing (PacBio SMRT or ONT) combined with Hi-C scaffolding to achieve chromosome-level assembly [71].

• NBS Gene Identification: Application of HMMER with HMM profiles from InterPro to identify 196 NBS-domain-containing genes [11].

• Domain Architecture Analysis: Classification into typical and atypical NBS-LRRs based on domain integrity, revealing 62 typical NLRs with complete N-terminal and LRR domains [11].

• Phylogenetic Analysis: Construction of phylogenetic trees with known NLRs from model plants to classify genes into CNL, TNL, and RNL subfamilies [11].

• Expression Validation: RNA-seq analysis under various stress conditions to confirm expression and identify candidates with potential immune functions [11].

This pipeline revealed a marked reduction in TNL and RNL subfamily members in Salvia species compared to other angiosperms, demonstrating how robust annotation can reveal evolutionary patterns in NBS gene families [11].

Figure 2: Specialized Annotation Pipeline for Complex Plant Genomes

Future Directions and Emerging Technologies

Advancements in multiple technologies are poised to further address annotation challenges:

• Telomere-to-telomere (T2T) genomes: Complete gapless assemblies resolve complex regions that traditionally fragmented NBS gene models. As of 2025, 11 medicinal plants have T2T assemblies, with contig N50 values reaching 35.87 Mb [71].

• Long-read transcriptomics: Full-length isoform sequencing helps validate complex gene models and alternative splicing in NBS genes [71].

• Single-cell sequencing: Enables characterization of cell-type-specific expression of NBS genes during pathogen infection [28].

• Integrated ML and evolutionary models: Combining machine learning with evolutionary analyses to predict functional residues and pathogen recognition specificities [28].

• Pan-genome analyses: Capturing the full diversity of NBS genes across multiple individuals and varieties of a species, revealing presence-absence variation and structural polymorphisms [28] [5].

As these technologies mature, they will progressively overcome the current limitations in annotating fragmented genes and complex gene clusters, accelerating the discovery and functional characterization of NBS domain genes in plant immunity.

The study of nucleotide-binding site (NBS) domain genes represents a critical frontier in plant immunity research, as these genes encode the largest class of intracellular immune receptors that confer disease resistance against diverse pathogens. However, research in this field faces a fundamental bioinformatics challenge: low sequence homology among these genes across plant species. Traditional alignment-based methods, which rely on detectable sequence similarity, often fail to identify evolutionarily related NBS-encoding genes when sequence conservation drops below a critical threshold, typically around 25-30% sequence identity [74]. This limitation substantially impedes the comprehensive identification and characterization of these crucial immune receptors across the plant kingdom.

The NBS gene superfamily exhibits remarkable structural and sequence diversity driven by continuous evolutionary arms races with rapidly evolving pathogens. Studies across land plants have identified thousands of NBS-domain-containing genes classified into numerous structural variants, with both classical (NBS, NBS-LRR, TIR-NBS, TIR-NBS-LRR) and species-specific architectural patterns [5]. This diversity, while biologically essential for effective immune recognition, creates significant obstacles for conventional homology-based annotation methods. As plant immunity research increasingly shifts toward precision breeding and engineering of disease resistance traits, overcoming these bioinformatics limitations becomes paramount for unlocking the full potential of NBS gene families in crop improvement [3] [75].

Limitations of Traditional Alignment-Based Methods

Fundamental Constraints in Detection Sensitivity

Alignment-based methods, including BLAST, HMMER, and related homology search tools, operate on the principle of identifying statistically significant sequence similarities between query and database sequences. These tools have formed the backbone of NBS gene identification pipelines for decades, with researchers using conserved domain models (e.g., PF00931 for the NB-ARC domain) to scan plant genomes for potential resistance genes [16] [18]. However, these approaches encounter substantial limitations when applied to the highly diversified NBS gene family.

The core issue stems from their dependence on sequence conservation. When sequence identity falls below approximately 25%, traditional methods struggle to distinguish true homologous relationships from random background matches [74]. This is particularly problematic for NBS genes, which often exhibit pronounced sequence divergence even within the same plant species due to their role in adapting to recognize rapidly evolving pathogen effectors. Research on NBS-LRR genes in medicinal plants like Salvia miltiorrhiza has revealed a marked reduction in specific subfamilies (TNL and RNL) compared to model plants, highlighting the taxonomic functional constraints that further complicate cross-species comparisons [11].

Practical Consequences for NBS Gene Research

The limitations of alignment-based methods manifest in several critical aspects of plant immunity research:

• Incomplete gene annotation: Studies consistently reveal that automated annotation pipelines miss substantial portions of NBS gene repertoires. For example, genome-wide analyses of Nicotiana species identified hundreds of NBS genes that were previously unannotated or misannotated [18]. The unique genomic structure of R-gene clusters, with their numerous similar sequences, often leads to fragmented annotations and assembly issues [43].

• Taxonomic functionality restrictions: The practical transfer of NLR resistance genes between distantly related species is often hampered by restricted taxonomic functionality (RTF). For instance, the pepper Bs2 gene, which confers resistance to bacterial spot disease in tomato, fails to function in Arabidopsis despite recognizing the same effector [3]. This functional limitation mirrors the computational challenges in identifying homologous relationships across taxonomic boundaries.

• Bias toward characterized clades: Alignment-based searches preferentially identify NBS genes with close similarity to previously characterized sequences, creating a systematic bias against novel or highly divergent subfamilies. This is evident in the underrepresentation of certain NBS subtypes across multiple plant species [11] [5].

Table 1: Comparison of NBS-LRR Gene Identification Results Using Different Methods

Plant Species	NBS Genes Identified (Alignment-Based)	Additional Genes Potentially Missed	Reference
Salvia miltiorrhiza	196 (62 typical NLRs)	TNL and RNL subfamilies markedly reduced	[11]
Nicotiana benthamiana	156	Limited TNL-type (only 5 identified)	[16]
Nicotiana tabacum	603	45.5% contain only NBS domain (possible fragments)	[18]
Various land plants (34 species)	12,820	Species-specific structural patterns potentially underannotated	[5]

Advanced Computational Solutions for Remote Homology Detection

Deep Learning Approaches for Structure-Aware Sequence Analysis

Recent advances in deep learning have yielded powerful new tools that address the fundamental limitations of alignment-based methods. These approaches leverage protein language models and structural prediction to detect homologous relationships that evade traditional sequence comparison techniques.

TM-Vec represents a groundbreaking innovation in this domain. This twin neural network model is trained to approximate TM-scores (a metric of structural similarity) directly from sequence pairs without requiring intermediate structure computation [74]. Unlike traditional methods that become unreliable below 25% sequence identity, TM-Vec maintains accurate structural similarity predictions (median error = 0.026) even for sequence pairs with less than 0.1% sequence identity. The method creates structure-aware vector embeddings for protein sequences, enabling efficient database indexing and rapid identification of structurally similar proteins through nearest-neighbor searches in the embedding space [74].

Complementing TM-Vec, DeepBLAST performs structural alignments using only sequence information by identifying structurally homologous regions between proteins. It employs a differentiable Needleman-Wunsch algorithm trained on proteins with known structures to predict structural alignments that rival structure-based alignment methods in accuracy [74]. This capability is particularly valuable for NBS gene analysis, where structural conservation often persists despite extensive sequence divergence.

Specialized Tools for Plant Resistance Gene Identification

The development of PRGminer represents a domain-specific application of deep learning tailored to plant immunity research. This tool implements a two-phase prediction framework: Phase I classifies input protein sequences as resistance genes or non-resistance genes, while Phase II categorizes predicted R-genes into specific structural classes (CNL, TNL, RLK, etc.) [43].

PRGminer demonstrates remarkable performance, achieving 98.75% accuracy in k-fold testing and 95.72% accuracy on independent validation for R-gene identification [43]. By leveraging dipeptide composition and deep learning architecture rather than sequence alignment, PRGminer effectively identifies resistance genes that lack close sequence homologs in databases. This capability is particularly valuable for mining NBS genes from wild plant species and crop relatives where traditional annotation methods face significant challenges due to the absence of closely related reference sequences [43].

Table 2: Comparison of Advanced Methods for Remote Homology Detection

Method	Approach	Key Advantages	Performance Metrics	Applications in NBS Research
TM-Vec	Twin neural network predicting TM-scores from sequences	Works at <0.1% sequence identity; enables scalable database search	r=0.97 vs TM-align; median error=0.026 at low sequence identity	Structural similarity search across plant immune receptors
DeepBLAST	Differentiable structural alignment from sequences	Identifies structurally homologous regions without 3D structures	Outperforms sequence alignment; similar to structure-based methods	Mapping functional domains in divergent NBS genes
PRGminer	Deep learning classification of R-genes	Domain-aware without relying on sequence alignment	98.75% accuracy (k-fold); 95.72% (independent testing)	Genome-wide annotation of NBS genes across plant taxa

Experimental Protocols for Validation and Functional Characterization

Genome-Wide Identification and Classification Pipeline

The comprehensive characterization of NBS gene families requires integrated experimental protocols that combine advanced computational prediction with empirical validation:

Step 1: Sequence-Based Identification

• Extract protein sequences from the target plant genome
• Perform HMMER search using NB-ARC domain model (PF00931) with E-value < 1e-20 [16] [18]
• Confirm domain architecture using SMART, CDD, and Pfam databases
• Retain only genes containing associated NBS domains for further analysis

Step 2: Structural Classification and Phylogenetics

• Classify identified NBS genes into structural categories (CNL, TNL, RNL, etc.) based on domain composition [11] [18]
• Perform multiple sequence alignment using MUSCLE or MAFFT with default parameters
• Construct phylogenetic trees using Maximum Likelihood method (e.g., FastTreeMP) with 1000 bootstrap replicates [5]
• Analyze expansion/contraction of NBS subfamilies across target species

Step 3: Expression Profiling Under Stress Conditions

• Collect RNA-seq data from tissues under biotic and abiotic stresses
• Process raw sequencing data with Trimmomatic for quality control
• Map cleaned reads to reference genome using HISAT2
• Quantify transcript expression with Cufflinks/Cuffdiff using FPKM normalization
• Identify differentially expressed NBS genes during pathogen challenge [18]

Functional Validation Through Genetic Approaches

Virus-Induced Gene Silencing (VIGS) Protocol:

• Select target NBS genes showing differential expression during pathogen infection
• Clone 300-500 bp gene-specific fragments into TRV-based VIGS vectors
• Inoculate plants with Agrobacterium tumefaciens carrying VIGS constructs
• Monitor gene silencing efficiency 2-3 weeks post-inoculation via qRT-PCR
• Challenge silenced plants with target pathogens and assess disease symptoms
• Quantify pathogen biomass to determine effect on resistance [5]

Interfamily Transfer Validation:

• Identify sensor NLRs with recognized effectors conserved across pathogen species
• Co-transform recipient plants with sensor NLR and cognate helper NLR genes (e.g., NRC family)
• Validate protein expression via Western blotting
• Challenge transgenic lines with corresponding pathogens
• Assess for effector-dependent cell death and resistance phenotypes [3]

The following diagram illustrates the integrated experimental workflow for NBS gene identification and validation:

Table 3: Key Research Reagent Solutions for NBS Gene Studies

Reagent/Resource	Function	Application Example	Reference
HMMER Suite	Hidden Markov Model search for domain identification	Identifying NB-ARC domains (PF00931) in plant genomes	[16] [18]
PRGminer Web Server	Deep learning-based R-gene prediction and classification	High-throughput annotation of NBS genes in newly sequenced species	[43]
TM-Vec Database	Structural similarity search from sequence information	Identifying structurally similar NBS proteins across taxonomic boundaries	[74]
VIGS Vectors (TRV-based)	Transient gene silencing in plants	Functional validation of candidate NBS genes in resistant varieties	[5]
OrthoFinder	Orthogroup inference and comparative genomics	Evolutionary analysis of NBS gene families across land plants	[5]
NRC Helper NLRs	Conserved signaling components for sensor NLRs	Enabling interfamily transfer of resistance specificity	[3]

The study of NBS domain genes in plant immunity has entered a transformative phase where advanced computational methods are overcoming the long-standing limitations of alignment-based homology detection. By integrating structure-aware deep learning tools like TM-Vec and DeepBLAST with domain-specific classifiers like PRGminer, researchers can now identify and characterize NBS genes that previously evaded detection due to low sequence homology.

These computational advances, combined with robust experimental validation frameworks, are accelerating the discovery of novel resistance genes from diverse plant species. The successful interfamily transfer of sensor and helper NLR pairs, as demonstrated in engineering resistance against bacterial leaf streak in rice, highlights the practical applications of these approaches for crop improvement [3]. As these methods continue to mature and integrate with emerging technologies like protein structure prediction and single-cell genomics, they promise to unlock the full diversity of the plant immune repertoire, providing powerful new tools for sustainable agriculture and crop protection.

Resolving Subfamily Classification Ambiguities for Non-Canonical and Atypical NBS Proteins

Within the realm of plant immunity research, nucleotide-binding site (NBS) domain genes encode intracellular immune receptors that play a pivotal role in effector-triggered immunity (ETI). These receptors, commonly known as NLRs (Nucleotide-binding, Leucine-rich Repeat proteins), constitute one of the most diverse and rapidly evolving gene families in plants [76]. The canonical NLR classification system categorizes these proteins based on their N-terminal domains into TNL (Toll/Interleukin-1 Receptor), CNL (Coiled-Coil), and RNL (RPW8) subfamilies [77]. However, this system fails to adequately accommodate the substantial portion of NBS-encoding genes that exhibit non-canonical, truncated, or atypical architectures, creating significant ambiguities in subfamily classification [11] [16].

The prevalence of these atypical NBS proteins is substantial. Recent genomic studies have revealed that in species such as Salvia miltiorrhiza, 196 identified NBS-LRR genes included only 62 with complete N-terminal and LRR domains [11]. Similarly, in Nicotiana benthamiana, among 156 identified NBS-LRR homologs, only 30 were typical TNL or CNL types, while the remainder represented various atypical forms [16]. This prevalence underscores the critical need for refined classification frameworks that can accurately categorize the full spectrum of NBS protein architectures, thereby enabling more precise functional characterization within plant immunity research.

Current Classification Frameworks and Their Limitations

Traditional NLR Classification System

The conventional classification of plant NLR proteins is primarily based on their domain architecture, with particular emphasis on the N-terminal domain. This system establishes three major classes:

• TNL (TIR-NBS-LRR): Characterized by an N-terminal TIR domain involved in signal transduction [77].
• CNL (CC-NBS-LRR): Featuring an N-terminal coiled-coil domain that facilitates protein-protein interactions [77].
• RNL (RPW8-NBS-LRR): Containing an N-terminal RPW8 domain that mediates broad-spectrum resistance [16].

These canonical NLRs function as molecular switches within plant immunity, existing in an inactive ADP-bound state and transitioning to an active ATP-bound state upon pathogen perception [76]. This activation triggers immune signaling, often accompanied by a hypersensitive response to limit pathogen spread [76].

The Spectrum of Non-Canonical and Atypical NBS Proteins

Beyond these canonical architectures, plants possess a diverse array of non-canonical NBS proteins that defy straightforward classification. These atypical forms include:

• Truncated Variants: Proteins lacking complete domain structures, classified as N (NBS-only), TN (TIR-NBS), CN (CC-NBS), and NL (NBS-LRR) types [11] [16].
• Integrated Domain (ID) Variants: NLR proteins fused with additional, non-canonical domains that often function as pathogen effector baits or decoys [78] [79].
• Lineage-Specific Architectures: Species-specific structural patterns, such as the TIR-NBS-TIR-Cupin1-Cupin1 domain fusion identified in broad comparative analyses [5].

The functional significance of these atypical NBS proteins is increasingly recognized. For instance, truncated TN proteins in Arabidopsis thaliana, such as TN13, have been demonstrated to interact with and contribute to the immune signaling of full-length CNL proteins like RPS5 [80]. Similarly, integrated domains in NLR proteins, such as the WRKY domain in RRS1-R and heavy metal-associated (HMA) domains in rice RGA5 and Pik-1, enable specific pathogen recognition by mimicking authentic effector targets [78] [79].

Table 1: Classification of NBS Protein Types Based on Domain Architecture

Category	Subtype	Domain Architecture	Functional Role	Example
Canonical NLRs	TNL	TIR-NBS-LRR	Pathogen recognition & signaling	Arabidopsis RPS4
	CNL	CC-NBS-LRR	Pathogen recognition & signaling	Arabidopsis RPM1
	RNL	RPW8-NBS-LRR	Defense signal transduction	Arabidopsis ADR1
Atypical/Truncated	N	NBS-only	Regulatory/Adapter functions	Various species
	TN	TIR-NBS	Sensor/regulatory roles	Arabidopsis TN13
	CN	CC-NBS	Sensor/regulatory roles	Various species
	NL	NBS-LRR	Impaired recognition capability	Various species
NLR-IDs	Integrated Decoys	NLR with additional domains	Effector recognition as molecular baits	RRS1 (WRKY), Pik-1 (HMA)

Methodological Framework for Resolving Classification Ambiguities

Genomic Identification and Annotation Pipeline

A robust, reproducible pipeline for identifying and classifying NBS proteins requires integrated bioinformatic approaches:

Step 1: Domain Identification Initiate with Hidden Markov Model (HMM) searches using the NB-ARC domain (PF00931) from the Pfam database with an expectation value (E-value) cutoff of <1*10⁻²⁰ [16] [5]. Follow with comprehensive domain annotation using multiple databases including Pfam, SMART, and Conserved Domain Database (CDD) to identify all associated domains [16].

Step 2: Architecture Classification Categorize proteins based on presence/absence of TIR, CC, RPW8, LRR, and additional integrated domains. Employ tools like COILS for coiled-coil prediction and MEME for motif discovery to identify conserved sequence patterns [16] [77].

Step 3: Phylogenetic Analysis Construct phylogenetic trees using maximum likelihood methods with bootstrap validation (typically 1000 replicates) to establish evolutionary relationships and validate classification [11] [16].

Step 4: Orthogroup Mapping Perform comparative analysis across multiple species using OrthoFinder or similar tools to identify conserved and lineage-specific NBS gene families [5].

This integrated approach enables systematic resolution of classification ambiguities, particularly for proteins with non-canonical architectures.

Structural and Motif-Based Classification Enhancement

Complementary to domain-based classification, conserved motif analysis provides additional resolution for categorizing atypical NBS proteins. Research across multiple plant species has identified six core motifs within the NBS domain: P-loop, RNBS-A, kinase-2, RNBS-B, RNBS-C, and GLPL, which are essential for ATP/GTP binding and resistance signaling [77]. These motifs exhibit subfamily-specific conservation patterns that can help resolve ambiguous classifications:

• The P-loop motif (Walker A motif) facilitates nucleotide binding and is conserved across most NBS proteins [77].
• The RNBS-B and RNBS-C motifs show distinct sequence signatures between TNL and CNL subfamilies, even in truncated forms [77].
• The GLPL motif, located near the end of the NBS domain, contributes to nucleotide-binding pocket formation and signaling activation [77].

MEME suite analysis with motif counts set to 10 and width lengths ranging from 6 to 50 amino acids effectively identifies these conserved patterns in both typical and atypical NBS proteins [16].

Table 2: Experimental Approaches for NBS Protein Classification and Functional Validation

Method	Key Parameters	Application in Classification	Technical Considerations
HMMER Search	E-value <1*10⁻²⁰, NB-ARC domain (PF00931)	Initial identification of NBS-domain containing proteins	Adjust E-value based on genome size and quality
Phylogenetic Analysis	Maximum likelihood, 1000 bootstrap replicates	Evolutionary relationship inference and subfamily assignment	Use conserved NBS domain sequences for alignment
OrthoFinder	DIAMOND for sequence similarity, MCL for clustering	Cross-species orthogroup mapping and conserved gene family identification	Helps distinguish lineage-specific innovations
MEME Motif Analysis	10 motifs, width 6-50 amino acids	Identification of conserved subdomain structures	Can reveal functional motifs in truncated forms
Protein-Protein Interaction	Yeast two-hybrid, co-immunoprecipitation	Functional validation of regulatory/sensor relationships	Critical for characterizing atypical NBS proteins

Visualizing Classification Pathways and Relationships

The following diagram illustrates the integrated workflow for resolving NBS protein classification ambiguities, incorporating both bioinformatic and experimental approaches:

Table 3: Essential Research Reagents for NBS Protein Classification and Functional Studies

Reagent/Resource	Specific Examples	Function/Application	Technical Notes
HMM Profiles	NB-ARC (PF00931) from Pfam	Identification of NBS-domain containing proteins	Foundation for genomic screening
Domain Databases	Pfam, SMART, CDD	Comprehensive domain architecture analysis	Multi-database approach increases accuracy
Motif Discovery Tools	MEME Suite	Identification of conserved subdomains and motifs	Set to 10 motifs, width 6-50 aa for NBS proteins
Phylogenetic Software	MEGA7, OrthoFinder	Evolutionary relationship inference	Use maximum likelihood with bootstrap validation
VIGS Vectors	Tobacco Rattle Virus (TRV)-based vectors	Functional validation through gene silencing	Essential for characterizing NBS gene function
Expression Vectors	Yeast two-hybrid systems, Co-IP compatible vectors	Protein-protein interaction studies	Critical for validating regulatory relationships
Primer Sets	Degenerate primers for NBS domain amplification	Isolation of resistance gene analogs (RGAs)	Designed based on conserved NBS motifs

Case Studies in Classification Resolution

Truncated TN Proteins in Arabidopsis Immunity

The functional characterization of Arabidopsis TN13 exemplifies the classification challenges and resolutions for atypical NBS proteins. Initially categorized simply as a TIR-NBS (TN) protein, detailed investigation revealed its specific functional role: TN13 interacts with the CC and NBS domains of the full-length CNL protein RPS5, contributing to RPS5-mediated immunity against Pseudomonas syringae carrying the AvrPphB effector [80]. This functional partnership illustrates how atypical NBS proteins can operate as regulatory components within NLR immune networks, necessitating a classification that captures both structural features and functional partnerships.

NLR-IDs with Integrated Decoy Domains

The discovery of NLRs with integrated domains (NLR-IDs) has fundamentally expanded NBS protein classification paradigms. Well-characterized examples include:

• RRS1-R in Arabidopsis: Contains a C-terminal WRKY domain that functions as a decoy for bacterial effectors (PopP2, AvrRps4) that typically target WRKY transcription factors [78] [79].
• RGA5 and Pik-1 in rice: Integrate heavy metal-associated (HMA) domains that directly bind fungal effectors (AVR-Pia, AVR1-CO39, AVR-Pik) from Magnaporthe oryzae [78] [79].

These NLR-IDs challenge traditional classification systems by incorporating non-canonical domains that serve as effector baits, supporting the "integrated decoy" model where these domains mimic authentic pathogen targets [78]. The systematic identification of 265 unique NLR integrated domains across 40 plant species confirms this as a widespread evolutionary strategy for expanding pathogen recognition capacity [79].

Lineage-Specific Subfamily Expansions and Contractions

Comparative genomic analyses reveal substantial lineage-specific variation in NBS subfamily distributions that further complicates classification. In Salvia miltiorrhiza, among 62 typical NLRs, 61 belong to the CNL subfamily with only one RNL member and complete absence of TNLs [11]. Similarly, pepper (Capsicum annuum) exhibits dramatic dominance of nTNL genes (248) over TNLs (4) [77]. These lineage-specific patterns reflect distinct evolutionary trajectories and highlight the necessity for classification frameworks that accommodate taxonomic context rather than relying solely on domain architecture.

Resolving classification ambiguities for non-canonical and atypical NBS proteins requires an integrated approach that combines multiple bioinformatic methods with experimental validation. The proposed framework incorporates domain architecture, phylogenetic relationships, conserved motifs, and functional interactions to create a more nuanced classification system. This refined approach accurately captures the functional diversity of NBS proteins beyond canonical NLRs, enabling more precise characterization of their roles in plant immunity. As structural studies continue to reveal mechanisms of NLR activation and signaling [13], and genomic analyses uncover ever-greater diversity in NBS protein architectures [5] [79], classification systems must remain adaptable to incorporate new insights into the complex landscape of plant immune receptors.

Managing Data Quality and Class Imbalance in Machine Learning Models for R-Gene Prediction

Plant resistance genes (R-genes), particularly those encoding nucleotide-binding site (NBS) domain proteins, constitute a fundamental component of the plant immune system, enabling detection of diverse pathogens through effector-triggered immunity (ETI) [12] [81]. The NBS-leucine rich repeat (LRR) class of proteins functions as intracellular immune receptors that recognize pathogen effector molecules, initiating robust defense signaling cascades [28] [5]. Accurate computational prediction of these genes from plant genomes represents a critical research area for accelerating crop improvement and enhancing food security.

Traditional methods for R-gene identification have relied primarily on alignment-based tools and domain search algorithms, which often fail with sequences exhibiting low homology [43] [28]. Machine learning (ML) and deep learning (DL) approaches have emerged as powerful alternatives, capable of recognizing complex patterns beyond simple sequence homology. However, these methods face significant challenges in data quality and severe class imbalance, as R-genes typically represent a very small fraction of the total gene repertoire in plant genomes [43] [5]. This technical guide examines these challenges within the context of plant immunity research and presents comprehensive computational strategies for developing robust R-gene prediction models.

The NBS-LRR Gene Family in Plant Immunity

Structural Diversity and Classification

NBS-LRR genes encode modular proteins characterized by three fundamental domains: a variable N-terminal domain, a central nucleotide-binding adaptor (NB-ARC) domain, and C-terminal leucine-rich repeats (LRRs) [28] [5]. This gene family is classified into major subclasses based on N-terminal domain architecture:

• TIR-NBS-LRR (TNL): Contains a Toll/interleukin-1 receptor-like domain
• CC-NBS-LRR (CNL): Features a coiled-coil domain at the N-terminus
• RPW8-NBS-LRR (RNL): Possesses a Resistance to Powdery Mildew 8 domain [5]

Plant genomes exhibit remarkable diversity in their NBS-LRR repertoires. Recent studies have identified 12,820 NBS-domain-containing genes across 34 plant species, classifying them into 168 distinct domain architecture classes [5]. This extensive diversification reflects an evolutionary arms race with rapidly evolving pathogens, but simultaneously complicates comprehensive computational identification.

Functional Mechanisms in Pathogen Recognition

NBS-LRR proteins function as essential components of the plant immune system through multiple detection mechanisms. Direct recognition occurs when R proteins physically interact with pathogen effectors, as demonstrated by the rice Pi-ta protein binding to the fungal effector AVR-Pita [12]. In contrast, indirect recognition follows the "guard hypothesis," where R proteins monitor host cellular components that are modified by pathogen effectors [12] [28]. The Arabidopsis RPM1 protein, for instance, detects Pseudomonas syringae effectors AvrRpm1 and AvrB through their modification of the host protein RIN4 [12].

Activation of NBS-LRR proteins triggers conformational changes that promote nucleotide exchange (ADP to ATP), initiating downstream defense signaling cascades culminating in the hypersensitive response (HR) and systemic acquired resistance (SAR) [12] [81]. This sophisticated immune recognition system provides a genetic basis for disease resistance breeding programs, highlighting the critical importance of accurate R-gene identification.

Data Quality Challenges in R-Gene Prediction

Genomic Annotation Complexities

The unique genomic architecture of R-genes presents substantial challenges for data quality in ML pipelines. Several factors contribute to these difficulties:

• Gene clustering and duplication: R-genes are frequently organized in complex clusters of closely related sequences, leading to assembly and annotation difficulties [43]. Tandem duplications create regions of high sequence similarity that challenge genome assembly algorithms, often resulting in fragmented or incomplete gene models [5].

• Low expression levels: Many R-genes exhibit constitutively low transcript abundance, making them difficult to validate through RNA-Seq evidence [43]. This limitation reduces the effectiveness of transcriptome-based annotation methods.

• Misannotation as repetitive elements: The repetitive nature of LRR domains often causes misclassification as transposable elements or other repetitive sequences during automated annotation [43].

These technical challenges frequently result in incomplete, fragmented, or missing R-gene annotations in public databases, directly impacting the quality of training datasets for machine learning models.

Dataset Curation and Feature Engineering

High-quality training data is essential for developing accurate prediction models. The PRGminer tool exemplifies rigorous dataset construction, incorporating protein sequences from multiple public databases including Phytozome, Ensemble Plants, and NCBI [43]. Feature representation strategies significantly impact model performance, with dipeptide composition demonstrating particularly effective representation for R-gene prediction in PRGminer, achieving 98.75% accuracy in k-fold validation [43].

Table 1: Performance Comparison of R-Gene Prediction Tools

Tool/Method	Approach	Key Features	Reported Accuracy	Strengths
PRGminer	Deep Learning	Dipeptide composition, Two-phase classification	95.72% (Independent testing)	High MCC (0.91), Webserver available
Domain-based pipelines	HMMER, InterProScan	Conserved domain detection	Varies by tool	Interpretable, Biological basis
Traditional ML	SVM, Random Forest	Multiple feature representations	Not specified	Works with small datasets
NLR-Annotator	Domain-based	Specific NLR identification	Not specified	Specialized for NLR class

The Class Imbalance Problem in R-Gene Prediction

Biological Origins of Imbalance

Class imbalance in R-gene prediction stems from fundamental biological constraints. NBS-LRR genes typically represent less than 2% of the total gene repertoire in most plant genomes, creating a natural imbalance where R-genes constitute the minority class [5]. For example, comprehensive analyses have identified approximately 2012 NBS-encoding genes in wheat, representing a small fraction of the total genome [5]. This imbalance is further exacerbated by:

• Varying repertoire sizes: NLR family sizes range from approximately 25 in bryophytes to thousands in some angiosperms [5]
• Species-specific expansions: Lineage-specific adaptations lead to dramatic differences in R-gene numbers between species
• Annotation inconsistency: Incomplete annotations systematically reduce positive examples in training datasets

Consequences for Model Performance

Severely imbalanced datasets negatively impact model training and evaluation through multiple mechanisms:

• Misleading accuracy metrics: Models that always predict "non-R-gene" can achieve high accuracy while completely failing to identify the positive class [82] [83]. For instance, a model achieving 99% accuracy would be useless if it missed all true R-genes.

• Majority class bias: Standard training procedures optimize overall accuracy, disproportionately weighting the majority class and resulting in poor minority class performance [82] [84]. The algorithm becomes biased toward predicting the majority class due to its higher prevalence in the training data.

• Insufficient minority representation: Small batch sizes during training may contain no examples of the minority class, preventing effective learning of R-gene characteristics [84].

These challenges necessitate specialized approaches for both model training and evaluation to develop useful R-gene prediction systems.

Technical Solutions for Class Imbalance

Data-Level Strategies

Data-level approaches modify training set composition to address class imbalance:

• Random undersampling: Reduces majority class examples by randomly removing instances until a more balanced distribution is achieved [82]. This approach increases the probability that training batches contain sufficient minority examples, but risks discarding potentially useful majority class information.

• Informed undersampling: Techniques such as Tomek Links remove majority class examples near minority class instances to clean decision boundaries [82]. Cluster-based undersampling groups majority examples and samples from each cluster, preserving distribution characteristics.

• Strategic downsampling: Google's ML guidelines recommend downsampling the majority class while simultaneously upweighting the downsampled examples in the loss function [84]. This approach separates learning feature characteristics from learning class distribution, requiring experimentation with different rebalancing ratios.

Table 2: Comparison of Imbalance Handling Techniques

Technique	Mechanism	Advantages	Limitations	Applicability to R-gene Prediction
Random Undersampling	Reduces majority class	Simple, fast training	Loss of information	Moderate (with sufficient data)
Tomek Links	Removes ambiguous examples	Cleaner decision boundaries	Does not create new examples	High (for boundary refinement)
Downsampling + Upweighting	Adjusts loss function	Faster convergence, better model	Requires hyperparameter tuning	High (recommended approach)
Ensemble Methods	Combines multiple models	Robust performance	Computational complexity	Moderate (with resources)

Algorithmic and Evaluation Approaches

Algorithm-level strategies modify the learning process to address imbalance without resampling:

• Cost-sensitive learning: Assigns higher misclassification costs to minority class errors, directly incorporating imbalance awareness into the objective function [82]. This approach effectively makes false negatives more costly than false positives.

• Threshold adjustment: Modifies the default classification threshold (typically 0.5) to favor minority class prediction, trading off precision and recall based on application requirements [83].

• Appropriate evaluation metrics: Replaces accuracy with metrics that better capture minority class performance:

  • Recall (Sensitivity): Measures the proportion of actual R-genes correctly identified [83] [85]
  • Precision: Quantifies the proportion of predicted R-genes that are true R-genes [83] [85]
  • F1-score: Harmonic mean of precision and recall, providing a balanced metric [83] [85]
  • Matthews Correlation Coefficient (MCC): Correlation coefficient between observed and predicted classifications, particularly informative for imbalanced datasets [43]

The PRGminer tool demonstrates effective application of these principles, reporting both high accuracy (95.72%) and MCC (0.91) on independent testing, indicating robust performance despite class imbalance [43].

Experimental Design and Protocols

Genome-Wide R-Gene Identification Pipeline

A comprehensive protocol for R-gene identification involves multiple computational stages:

Step 1: Data Collection and Preprocessing

• Obtain high-quality genome assemblies and annotation files from Phytozome, Ensemble Plants, or NCBI [43] [5]
• Extract protein coding sequences and translate to amino acid sequences
• Remove low-quality sequences and potential contaminants

Step 2: Domain Identification and Feature Extraction

• Use HMMER or PfamScan with default e-value (1.1e-50) and Pfam-A_hmm model to identify NB-ARC domains [5]
• Extract additional domains (TIR, CC, LRR, RPW8) using InterProScan or similar tools
• Calculate feature representations (dipeptide composition, physiochemical properties)

Step 3: Model Training with Imbalance Handling

• Implement strategic downsampling of majority class (non-R-genes)
• Apply upweighting to downsampled class in loss function (e.g., multiply loss by downsampling factor)
• Train deep learning architecture with k-fold cross-validation
• Optimize hyperparameters including downsampling ratio and class weights

Step 4: Evaluation and Validation

• Calculate comprehensive metrics including precision, recall, F1-score, and MCC
• Perform independent testing on held-out datasets
• Validate predictions against known R-genes from databases such as PRGdb or Annotated NBS-LRR Genes [28]

Orthogroup Analysis for Evolutionary Studies

For evolutionary analyses across multiple species:

• Perform orthogroup clustering using OrthoFinder v2.5.1 with DIAMOND for sequence similarity and MCL for clustering [5]
• Identify core orthogroups (conserved across species) and lineage-specific expansions
• Construct phylogenetic trees using maximum likelihood methods (FastTreeMP) with 1000 bootstrap replicates [5]
• Analyze expression patterns across tissues and stress conditions using RNA-seq data (FPKM values)

Visualization of Key Methodologies

R-Gene Prediction Workflow

Growth-Defense Tradeoff Signaling Network

Table 3: Research Reagent Solutions for R-Gene Studies

Resource Category	Specific Tools/Databases	Function	Application in R-gene Research
Genomic Databases	Phytozome, Ensemble Plants, NCBI	Source of genome sequences and annotations	Training data for ML models [43] [5]
Domain Prediction Tools	HMMER, PfamScan, InterProScan	Identification of conserved domains	Feature extraction for classification [43] [5]
R-gene Specific Databases	PRGdb, ANNA, NLR Atlas	Curated collections of resistance genes	Benchmarking and validation [28]
Orthology Analysis	OrthoFinder, DIAMOND, MCL	Evolutionary relationship inference	Identification of conserved R-gene families [5]
Expression Databases	IPF, CottonFGD, Cottongen	Tissue-specific and stress-induced expression	Functional validation of predictions [5]
ML Frameworks	TensorFlow, PyTorch, scikit-learn	Model development and training	Implementation of prediction algorithms [43]

Effective management of data quality and class imbalance is essential for developing accurate machine learning models in R-gene prediction. The integration of sophisticated imbalance handling techniques with biologically informed feature engineering has enabled tools like PRGminer to achieve impressive performance metrics, with accuracy exceeding 95% and MCC values above 0.9 [43]. These computational advances are particularly significant given the biological importance of NBS-LRR genes in plant immunity and their potential applications in crop improvement.

Future research directions should focus on enhancing model interpretability, integrating multi-omics data sources, and developing specialized architectures for rare R-gene subclass identification. The continued expansion of curated R-gene databases and standardized benchmarking datasets will further accelerate progress in this critical field. As machine learning methodologies mature and biological datasets expand, computational prediction of resistance genes will play an increasingly vital role in enabling sustainable agriculture through targeted genetic improvement of crop disease resistance.

The nucleotide-binding site (NBS) domain genes encode a critical class of plant immune receptors that form the backbone of the effector-triggered immunity (ETI) system. These proteins, typically featuring a conserved NBS domain coupled with a leucine-rich repeat (LRR) region, constitute the largest family of plant resistance (R) genes, with approximately 80% of cloned R genes belonging to this family [11] [86] [28]. Despite their crucial role in pathogen recognition and defense activation, functional studies of NBS-LRR genes face significant methodological challenges, primarily stemming from extensive gene redundancy within plant genomes and the frequently low basal expression of individual family members. This technical guide addresses these experimental bottlenecks by presenting optimized approaches for the accurate characterization of NBS gene function, enabling researchers to advance plant immunity research and disease-resistance breeding programs.

Systematic Characterization of NBS Gene Families

Comprehensive Identification and Classification

Before embarking on functional studies, a thorough genome-wide identification of NBS-LRR genes is essential. The standard methodology employs Hidden Markov Model (HMM) profiling using domain models (e.g., PF00931 from PFAM) to identify candidate genes, followed by validation through conserved domain databases (CDD) and structural analysis [11] [18] [16].

Table 1: NBS-LRR Family Size Across Plant Species

Plant Species	Total NBS Genes	Typical NLRs	CNL	TNL	RNL	Reference
Salvia miltiorrhiza	196	62	61	0	1	[11]
Nicotiana tabacum	603	-	-	-	-	[18]
Nicotiana benthamiana	156	53	25	5	4*	[16]
Vernicia fordii	90	24	12	0	-	[14]
Vernicia montana	149	24	9	3	-	[14]
Arabidopsis thaliana	207	-	-	-	-	[11]

Note: *RPW8-containing genes in N. benthamiana

The classification system for NBS-LRR genes is based on domain architecture:

• Typical NBS-LRRs: Contain complete N-terminal, NBS, and LRR domains
  • TNL: Toll/Interleukin-1 receptor (TIR) domain at N-terminus
  • CNL: Coiled-coil (CC) domain at N-terminus
  • RNL: Resistance to powdery mildew 8 (RPW8) domain at N-terminus
• Atypical NBS-LRRs: Lack complete domains (N, TN, CN, NL types) [11] [16]

Expression Pattern Analysis

Basal expression profiling under normal growth conditions typically reveals low expression levels for most NBS-LRR genes, with selective induction occurring during pathogen challenge. Integration of transcriptome data with promoter analysis has revealed an abundance of cis-acting elements related to plant hormones and abiotic stress in NBS gene promoters, providing insights into their regulatory mechanisms [11] [86]. In Salvia miltiorrhiza, expression pattern analysis demonstrated a close association between specific SmNBS-LRRs and secondary metabolism, suggesting interconnected defense and metabolic pathways [11].

Figure 1: Comprehensive Workflow for NBS Gene Functional Studies

Advanced Strategies to Overcome Experimental Challenges

Addressing Gene Redundancy

The high degree of sequence similarity and functional redundancy among NBS-LRR genes within clusters necessitates specialized approaches for accurate functional characterization.

NBS Profiling for High-Resolution Genotyping

The NBS profiling technique employs targeted amplification of NBS domains using primers complementary to conserved motifs (P-loop, Kinase-2, and GLPL), followed by high-throughput sequencing. This method enables researchers to generate a compendium of NBS sequence tags that capture the diversity of R gene alleles across multiple genotypes [87]. In potato, this approach identified 587 distinct NBS domains across 91 genomes, detecting an average of 26 nucleotide polymorphisms per locus [87].

Table 2: Research Reagent Solutions for NBS Gene Studies

Reagent/Technique	Application	Key Features	Experimental Considerations
HMMER with PF00931	Domain identification	Identifies NBS domains with E-values < 1*10⁻²⁰	Requires subsequent validation with CDD [18] [16]
NBS Profiling Primers	Targeted amplification	Amplifies hypervariable regions flanking conserved motifs	16 primers sufficient for comprehensive coverage [87]
Virus-Induced Gene Silencing (VIGS)	Functional validation	Enables transient gene silencing in plants	Critical for testing essential genes [14]
Synthetic NBS Libraries	Comparative genomics	Enables cross-species evolutionary analysis	Requires high-quality genome assemblies [11]

Phylogenetic Analysis for Functional Prediction

Constructing detailed phylogenetic trees integrating NBS-LRR genes from multiple species allows researchers to identify orthologous relationships and make functional predictions based on clustering with characterized R genes. For example, phylogenetic analysis in Salvia miltiorrhiza revealed that SmNBS55 and SmNBS56 clustered with the well-characterized A. thaliana resistance protein RPM1, suggesting similar roles in pathogen recognition [11].

Overcoming Low Expression Challenges

The characteristically low basal expression of NBS-LRR genes necessitates specialized methodologies for detecting and measuring their expression and function.

Promoter Analysis and Induction Strategies

Comprehensive promoter analysis has revealed that NBS genes contain abundant cis-acting elements related to plant hormones and abiotic stress, providing a roadmap for experimental induction [11] [86]. Designing induction experiments based on these elements—using appropriate hormones (e.g., jasmonic acid, salicylic acid) or stress conditions—can significantly enhance expression levels to detectable ranges.

Sensitive Expression Detection Methods

When conventional expression analysis methods fail due to low transcript levels, implementing nested PCR approaches and RNA-seq with deep sequencing provides the sensitivity required for accurate detection. The integration of multiple transcriptome datasets under various stress conditions significantly enhances the detection probability for lowly expressed NBS-LRR genes [11].

Experimental Protocols for Functional Validation

Virus-Induced Gene Silencing (VIGS) Protocol

VIGS has emerged as a powerful technique for functional characterization of NBS-LRR genes, particularly for essential genes where stable knockout mutants would be lethal.

Step-by-Step Protocol:

• Target Sequence Selection: Identify a 200-300 bp gene-specific region with minimal similarity to other NBS-LRR genes to ensure silencing specificity
• Vector Construction: Clone the target sequence into appropriate VIGS vectors (e.g., TRV-based vectors)
• Plant Infiltration: Inoculate young plants (approximately 2-week-old seedlings) using agrobacterium-mediated delivery
• Silencing Validation: Confirm gene silencing efficiency using qRT-PCR 2-3 weeks post-infiltration
• Phenotypic Assessment: Challenge silenced plants with target pathogens and evaluate disease susceptibility compared to controls [14]

Application Example: In tung trees, VIGS-mediated silencing of Vm019719 in resistant Vernicia montana converted the phenotype to susceptible, confirming its role in Fusarium wilt resistance [14].

Heterologous Expression Systems

For NBS-LRR genes with persistent low expression in native systems, heterologous expression in model plants provides an effective alternative for functional analysis.

Step-by-Step Protocol:

• Candidate Gene Selection: Identify candidate NBS-LRR genes with predicted specific functions based on phylogenetic analysis
• Expression Vector Construction: Clone full-length coding sequences into plant expression vectors under strong constitutive promoters
• Plant Transformation: Introduce constructs into model plants (e.g., Arabidopsis thaliana, Nicotiana benthamiana)
• Pathogen Challenge: Assess transformed lines for enhanced resistance to specific pathogens
• Molecular Analysis: Examine hypersensitive response and defense marker gene expression [18] [28]

Application Example: Heterologous expression of a maize NBS-LRR gene in Arabidopsis thaliana improved resistance to Pseudomonas syringae, demonstrating conserved function across species [18].

Figure 2: Experimental Strategies to Overcome Key Challenges

Integration with Breeding Applications

The ultimate validation of NBS-LRR gene function comes from their application in disease-resistance breeding programs. The identification of key NBS-LRR genes through the described methodologies enables the development of perfect markers for marker-assisted selection. In the tung tree system, the resistant allele Vm019719 from V. montana contains a functional W-box element in its promoter that is bound by VmWRKY64, while the susceptible allele Vf11G0978 from V. fordii has a deletion in this element, explaining the differential resistance [14]. Such precise molecular understanding enables the development of co-dominant markers for breeding programs.

The systematic characterization of NBS-LRR genes in various crops has revealed substantial variation in subfamily composition and expansion patterns. For example, comparative analysis across Salvia species revealed a marked reduction in TNL and RNL subfamily members compared to other angiosperms [11]. Similarly, monocotyledonous species such as rice, wheat, and maize have completely lost TNL and RNL subfamilies [11] [86]. This evolutionary perspective informs researchers about the expected NBS-LRR repertoire in their species of interest and guides experimental design.

Overcoming the challenges of redundancy and low expression in NBS gene families requires an integrated approach combining comprehensive bioinformatic characterization with sophisticated experimental methodologies. The strategies outlined in this guide—including NBS profiling for high-resolution genotyping, phylogenetic analysis for functional prediction, promoter analysis for expression modulation, and VIGS for functional validation—provide researchers with a robust toolkit for elucidating NBS-LRR gene functions. As genome sequencing technologies continue to advance and functional genomic tools become more sophisticated, the pace of NBS gene characterization will accelerate, enabling more effective deployment of these critical immune receptors in crop improvement programs aimed at enhancing sustainable agricultural production.

The nucleotide-binding site (NBS) domain genes, particularly those encoding NBS-leucine-rich repeat (NBS-LRR) proteins, constitute the largest and most critical class of plant resistance (R) genes, serving as essential intracellular immune receptors in plant defense systems [88] [11]. These genes enable plants to detect pathogen-secreted effectors and activate robust defense mechanisms through effector-triggered immunity (ETI), often accompanied by a hypersensitive response that limits pathogen spread [16] [11]. The NBS domain functions as a molecular switch by binding and hydrolyzing ATP/GTP, while the LRR domain provides specificity for pathogen recognition [11] [89].

Despite their critical biological function, NBS-LRR genes present substantial challenges for genomic studies due to their characteristic tandem duplication in clusters, extensive sequence similarity between paralogs, and exceptionally high repetitive content [90] [32]. Conventional genome assembly pipelines frequently collapse these regions or produce fragmented representations, leading to incomplete R-gene repertoires [90]. This technical limitation significantly impeders the identification of agronomically valuable resistance genes for crop improvement. This guide examines advanced techniques that are overcoming these challenges to enable complete and accurate resolution of NBS-LRR regions in plant genomes.

Technical Challenges in NBS-LRR Region Assembly

Intrinsic Genomic Features Complicating Assembly

The structural organization of NBS-LRR genes creates inherent obstacles for conventional sequencing and assembly approaches. These challenges primarily stem from several key characteristics:

• Clustered Genomic Organization: NBS-LRR genes typically reside in complex clusters of tandemly duplicated genes, though they can also appear as single genes dispersed throughout the genome [90]. This arrangement promotes frequent non-allelic homologous recombination, driving rapid evolution and generating significant sequence diversity that complicates assembly [32].

• Repetitive Nature and Low Expression: The repetitive architecture of these regions often causes assembly algorithms to collapse similar sequences, leading to missing or fragmented annotations [90]. Additionally, many NBS-LRR genes exhibit low constitutive expression levels, providing insufficient transcriptomic evidence to support accurate gene model prediction [90].

• Annotation Pipeline Deficiencies: Standard automated gene prediction tools frequently misannotate NBS-LRR loci due to their similarity to transposable elements and complex exon-intron structures [90]. The common practice of repeat masking prior to genome annotation further exacerbates this issue by inadvertently removing legitimate R-genes from consideration [90].

Consequences of Incomplete Assembly

Incomplete resolution of NBS-LRR regions has direct implications for plant immunity research and breeding:

Table 1: Impact of NBS-LRR Assembly Quality on Research Outcomes

Assembly Quality	Gene Repertoire	Variant Discovery	Breeding Applications
Fragmented Assembly	Incomplete R-gene catalog; missing alleles	Limited structural variation information	Overlooked valuable resistance traits
Complete Telomere-to-Telomere	Comprehensive R-gene inventory	Full spectrum of SVs and polymorphisms	Informed selection of resistance genes

Advanced Genome Assembly Strategies

Telomere-to-Telomere (T2T) Assembly Approaches

Recent breakthroughs in sequencing technologies have enabled the production of complete telomere-to-telomere genome assemblies that fully resolve repetitive regions, including NBS-LRR clusters:

• Hybrid Sequencing Strategies: The integration of multiple long-read technologies leverages their complementary strengths. The PacBio HiFi platform generates highly accurate long reads (typically 15-20 kb), while Oxford Nanopore Technology (ONT) produces ultra-long reads exceeding 100 kb, spanning even the most extensive repeats [91]. This combination has successfully resolved the 14.51 Gbp hexaploid bread wheat genome, including all centromeres and telomeres [91].

• Multi-Platform Data Integration: Effective T2T assemblies combine PacBio HiFi, ONT ultra-long reads, chromosome conformation capture (Hi-C), and optical mapping data [92] [91]. For the African wild rice (Oryza longistaminata), this approach yielded a 331-Mb T2T genome assembly with all 24 telomeres and 12 centromeres resolved, dramatically improving the representation of repetitive regions [92].

The following diagram illustrates the integrated workflow for achieving T2T assemblies:

Specialized Computational Methods for R-Gene Identification

Beyond improved sequencing, specialized computational methods have been developed specifically for resolving complex R-gene regions:

• Homology-based R-gene Prediction (HRP): This innovative approach addresses limitations in conventional domain searches by implementing a two-level homology search strategy. The method first identifies an initial set of R-genes in automated gene predictions, then uses these as queries for full-length homology searches against the entire genome assembly [90]. This method identified 363 NB-LRR genes in the tomato genome, outperforming previous approaches that had found only 326 genes [90].

• Domain-Focused Annotation Pipelines: Customized pipelines incorporate Hidden Markov Models (HMMs) specific to NBS domains (PF00931) combined with manual curation to improve gene model accuracy [32]. These approaches typically apply stringent E-value thresholds (e.g., < 1×10⁻²⁰) followed by manual verification of domain integrity to distinguish functional genes from pseudogenes [32].

Table 2: Comparison of NBS-LRR Identification Methods

Method	Principle	Advantages	Limitations
Protein Domain Search (PDS)	Identifies genes containing NBS domains in annotated gene sets	Simple implementation; standardized workflow	Misses fragmented genes; affected by repeat masking
Homology-based R-gene Prediction (HRP)	Uses known R-genes as queries for genome-wide similarity search	Recovers full-length genes missed by annotation; better handles complex loci	Requires high-quality reference R-genes; computationally intensive
Manual Curation (RenSeq)	Combines domain search with experimental validation and expert review	Highest accuracy; resolves complex gene models	Time-consuming; not scalable for multiple genomes

Experimental Protocols for NBS-LRR Characterization

Comprehensive Genome-Wide Identification Pipeline

The following protocol provides a robust framework for identifying and characterizing NBS-LRR genes in plant genomes:

Step 1: Initial Candidate Identification

• Retrieve the Hidden Markov Model for the NBS (NB-ARC) domain (PF00931) from the Pfam database
• Perform HMMER search (HMMER v3 suite) against all predicted proteins in the target genome
• Apply initial E-value threshold (e.g., < 0.01) to select candidate sequences
• Verify the presence of intact NBS domains using Pfam, SMART, and NCBI CDD tools

Step 2: Domain Architecture Classification

• Identify associated N-terminal domains (TIR, CC, RPW8) using hmmpfam with respective HMMs (TIR: PF01582, RPW8: PF05659)
• Detect coiled-coil domains using Paircoil2 with P-score cutoff of 0.03 [32]
• Classify genes into structural categories (TNL, CNL, RNL, TN, CN, NL, N) based on domain composition
• Identify LRR domains using multiple models (PF00560, PF07723, PF07725, PF12799) to capture variation

Step 3: Phylogenetic and Genomic Distribution Analysis

• Extract NB-ARC domain regions (typically ~250 amino acids after P-loop) from full-length sequences
• Perform multiple sequence alignment using ClustalW or MAFFT with default parameters
• Construct maximum likelihood phylogenetic trees (MEGA11 or FastTree) with 1000 bootstrap replicates
• Map gene positions to chromosomes and identify clustering patterns (genes within 200 kb considered clustered)

Expression Profiling Under Pathogen Challenge

To connect genomic findings to biological function, evaluate NBS-LRR expression patterns following pathogen infection:

• Transcriptome Sequencing: Collect tissue samples at multiple time points (e.g., 0, 6, 12, 24, 48 hours) post-inoculation with target pathogens
• Generate RNA-seq libraries (Illumina platform, ≥30 million reads per sample) from inoculated and control tissues
• Quantify expression values (FPKM or TPM) for all NBS-LRR genes and identify significantly differentially expressed genes

• Validate key candidates via qRT-PCR using gene-specific primers

• Functional Validation:

  • Implement Virus-Induced Gene Silencing (VIGS) to knock down candidate NBS-LRR genes
  • Assess changes in disease susceptibility compared to control plants
  • Quantify pathogen biomass using species-specific qPCR assays
  • Document hypersensitive response symptoms and other defense phenotypes

Table 3: Key Research Reagents and Computational Tools for NBS-LRR Genomics

Resource Category	Specific Tools/Reagents	Application Purpose	Key Features
Sequencing Technologies	PacBio HiFi, ONT Ultra-long	Generate long reads spanning repetitive regions	>20 kb reads with high accuracy; >100 kb reads with lower accuracy
Assembly Pipelines	SPART, HiCanu, hifiasm	Construct contiguous assemblies from long reads	Integration of multiple data types; specialized repeat handling
Domain Databases	Pfam, SMART, CDD	Identify NBS and associated domains	Curated HMM profiles; domain boundary prediction
Specialized R-gene Tools	HRP, RGAugury, NLR-annotator	Comprehensive R-gene identification	Homology-based prediction; genome-wide annotation
Expression Databases	IPF, CottonFGD, Phytozome	Access tissue-specific and stress-induced expression	Pre-computed RNA-seq data; user-friendly query interfaces

The resolution of repetitive NBS-LRR regions in plant genomes has evolved from a persistent challenge to an achievable goal through integrated experimental and computational approaches. The combination of T2T assembly strategies employing multi-platform sequencing with specialized bioinformatics tools like HRP has dramatically improved our capacity to completely catalog the plant immune repertoire. These advances are uncovering previously hidden genetic resources for crop improvement while providing fundamental insights into plant immunity mechanisms. As these methodologies become more accessible and scalable, comprehensive NBS-LRR characterization will increasingly support the development of durable disease resistance in agricultural systems, ultimately contributing to global food security.

Benchmarking and Validating NBS Gene Function Across Plant Species

Nucleotide-binding site (NBS) domain genes constitute one of the largest superfamilies of plant resistance (R) genes, playing a critical role in effector-triggered immunity against diverse pathogens. Orthogroup analysis has emerged as a powerful computational framework for identifying evolutionarily conserved and lineage-specific NBS genes across plant taxa. This technical guide details comprehensive methodologies for orthogroup identification, classification, and evolutionary analysis of NBS genes, enabling researchers to decipher the complex evolutionary dynamics that shape plant immune systems. Through systematic comparison of orthogroups, scientists can identify core NBS genes maintained across evolutionary timescales alongside species-specific expansions that may underlie specialized resistance mechanisms, providing fundamental insights for crop improvement and disease resistance breeding.

Plant NBS genes encode intracellular immune receptors that recognize pathogen effectors and initiate robust defense responses, including the hypersensitive response [31]. The majority of these proteins belong to the NLR family, characterized by three fundamental domains: an N-terminal domain (TIR, CC, or RPW8), a central nucleotide-binding site (NBS) domain, and a C-terminal leucine-rich repeat (LRR) region [15]. Based on their N-terminal domains, NLRs are classified into three principal subfamilies: TNL (TIR-NBS-LRR), CNL (CC-NBS-LRR), and RNL (RPW8-NBS-LRR) [59].

The NBS gene family exhibits remarkable diversity in plant genomes, ranging from dozens to thousands of members across different species [15]. This extensive variation results from rapid gene birth-and-death evolution, including frequent tandem duplications, segmental duplications, and gene losses [93] [94]. Orthogroup analysis provides a systematic framework for tracing these complex evolutionary dynamics across multiple species, distinguishing conserved immune components from recent, species-specific innovations.

Computational Identification of NBS Genes

Data Collection and Preparation

The initial step in orthogroup analysis involves compiling high-quality genomic and proteomic datasets. Genome assemblies and annotation files for target species should be obtained from reputable databases such as NCBI, Phytozome, or Plaza [95]. The selection of species should represent evolutionary diversity relevant to the research objectives, potentially spanning from bryophytes to higher plants to capture deep evolutionary conservation [95].

Domain Identification and Classification

NBS gene identification relies on detecting the conserved NB-ARC domain (Pfam: PF00931) using Hidden Markov Model (HMM) profiles. The following workflow provides a robust pipeline for comprehensive NBS gene identification:

Step 1: HMMER Search

• Perform HMM search using hmmsearch from the HMMER suite with the NB-ARC domain profile (PF00931)
• Recommended E-value cutoff: 1.1e-50 [95] or 1e-10 [96] for stringent identification
• Command example: hmmsearch --cpu 4 --domtblout output_file -E 1e-10 Pfam-A.hmm protein_dataset.fasta

Step 2: Complementary BLAST Search

• Conduct BLASTP searches against reference NLR proteins
• E-value threshold: 1.0 [93] [59]
• Merge results with HMMER outputs and remove redundancies

Step 3: Domain Architecture Validation

• Validate putative NBS genes using InterProScan or NCBI's Conserved Domain Database [96] [59]
• Confirm presence of complete NBS domain with E-value ≤ 1e-5 [96]
• Classify genes into subfamilies (TNL, CNL, RNL) by identifying additional domains:
  • TIR domain (PF01582)
  • RPW8 domain (PF05659)
  • LRR domain (PF08191)
  • CC domain (detected using COILS/PCOILS with threshold 0.9) [93] [97]

Table 1: NBS Gene Identification Tools and Parameters

Tool	Purpose	Key Parameters	Reference
HMMER v3.0	Domain identification	E-value: 1e-10 to 1e-50	[95] [96]
NCBI CDD	Domain validation	E-value: ≤ 1e-5	[96] [59]
InterProScan	Domain architecture	Default parameters	[96]
COILS/PCOILS	CC domain detection	Threshold: 0.9	[93] [97]
MEME Suite	Motif discovery	Motif count: 10, Width: 6-50 aa	[93] [59]

Orthogroup Analysis Methodology

Orthogroup Inference

Orthogroup analysis clusters genes into groups of orthologs and paralogs using specialized software. The recommended workflow utilizes OrthoFinder v2.5.1 or later versions [95] [96], which provides advanced algorithms for accurate orthogroup inference.

Experimental Protocol:

• Input Preparation: Compile complete protein sequences for all identified NBS genes from each species in FASTA format
• Sequence Comparison: Use DIAMOND tool for rapid sequence similarity searches [95]
• Clustering: Apply MCL (Markov Cluster Algorithm) to group sequences into orthogroups based on sequence similarity graphs [95]
• Orthogroup Refinement: Use DendroBLAST for ortholog identification and multiple sequence alignment with MAFFT 7.0 [95]
• Phylogenetic Reconstruction: Construct gene trees using maximum likelihood algorithm in FastTreeMP with 1000 bootstrap replicates [95]

Evolutionary Analysis

Orthogroups can be categorized based on evolutionary patterns:

• Core Orthogroups: Contain genes from most species, indicating evolutionary conservation (e.g., OG0, OG1, OG2) [95]
• Species-Specific Orthogroups: Restricted to particular lineages, suggesting recent expansions (e.g., OG80, OG82) [95]
• Lineage-Specific Expansions: Show differential duplication rates across phylogeny

Table 2: Orthogroup Classification and Characteristics

Orthogroup Type	Definition	Evolutionary Significance	Examples from Literature
Core Orthogroups	Present in most species surveyed	Ancient conserved functions in immunity	OG0, OG1, OG2 [95]
Species-Specific Orthogroups	Restricted to single species or lineage	Recent adaptations to specific pathogens	OG80, OG82 in cotton [95]
Expanded Orthogroups	Experienced significant duplications in specific lineages	Response to lineage-specific pathogen pressures	Solanaceae-specific expansions [93]
Contracted Orthogroups	Experienced gene losses in specific lineages	Relaxed selection or specialization	Asparagus officinalis contraction [96]

Figure 1: Orthogroup Analysis Workflow for NBS Genes

Complementary Experimental Approaches

Expression Profiling

Transcriptomic analysis validates the functional relevance of identified NBS orthogroups. Standard approaches include:

RNA-seq Data Analysis:

• Retrieve expression data (FPKM values) from databases like IPF, CottonFGD, or NCBI BioProjects [95]
• Categorize expression patterns into tissue-specific, abiotic stress-responsive, and biotic stress-responsive profiles [95]
• Identify differentially expressed NBS genes under pathogen challenge

qRT-PCR Validation:

• Select candidate genes based on expression profiles [98]
• Design gene-specific primers
• Perform time-course experiments post-pathogen inoculation
• Analyze relative expression using reference genes

Genetic Variation Analysis

Compare genetic variation in NBS genes between resistant and susceptible genotypes:

• Identify single nucleotide polymorphisms (SNPs) and insertions/deletions (indels)
• Detect positive selection through dN/dS ratio analysis [94]
• Correlate specific variants with resistance phenotypes [95]

Functional Validation

Virus-Induced Gene Silencing (VIGS):

• Select target NBS genes from significant orthogroups
• Design VIGS constructs using TRV-based vectors
• Infect plants with recombinant virus and challenge with pathogen
• Quantify disease symptoms and pathogen load [95]

Protein Interaction Studies:

• Perform protein-ligand interaction assays to test ADP/ATP binding [95]
• Conduct yeast-two-hybrid or co-immunoprecipitation to identify interacting partners [31] [12]
• Test interactions with pathogen effector proteins [12]

The Scientist's Toolkit: Essential Research Reagents

Table 3: Key Research Reagents for NBS Gene Analysis

Reagent/Resource	Function/Application	Example Sources/References
OrthoFinder v2.5.1+	Orthogroup inference from genomic data	[95] [96]
Pfam NB-ARC HMM (PF00931)	Identification of NBS domains	[95] [96] [59]
MEME Suite	Conserved motif discovery in NBS domains	[93] [59] [97]
PlantCARE Database	cis-element analysis in promoter regions	[96]
Phytozome/NCBI Databases	Genomic sequences and annotations	[95] [93]
TRV-based VIGS vectors	Functional validation through gene silencing	[95]
RNA-seq datasets (NCBI BioProject)	Expression profiling under stress conditions	PRJNA490626, PRJNA594268 [95]

Case Studies and Applications

Cotton Leaf Curl Disease Resistance

A comprehensive study identified 12,820 NBS genes across 34 plant species, classifying them into 168 distinct architectural classes [95]. Orthogroup analysis revealed 603 orthogroups, with OG2, OG6, and OG15 showing significant upregulation in tolerant cotton accessions under cotton leaf curl disease (CLCuD) stress [95]. Genetic variation analysis between susceptible (Coker 312) and tolerant (Mac7) Gossypium hirsutum accessions identified 6,583 unique NBS gene variants in the tolerant line, highlighting potential causal polymorphisms [95]. Functional validation through VIGS silencing of GaNBS (OG2) demonstrated its critical role in virus resistance [95].

Solanaceae NBS Evolution

Comparative analysis of potato (447 NBS genes), tomato (255), and pepper (306) revealed distinct evolutionary patterns: "consistent expansion" in potato, "expansion then contraction" in tomato, and "shrinking" in pepper [93]. The current NBS repertoires were derived from approximately 150 CNL, 22 TNL, and 4 RNL ancestral genes, with species-specific tandem duplications driving most expansions [93].

Asparagus Domestication Impact

Analysis of NLR genes in Asparagus species revealed significant contraction during domestication, with wild relative A. setaceus containing 63 NLR genes compared to only 27 in cultivated A. officinalis [96]. Orthologous analysis identified 16 conserved NLR pairs between wild and cultivated species, with most showing reduced or unresponsive expression to pathogen challenge in the domesticated species, explaining its increased susceptibility [96].

Figure 2: NBS Protein Activation Through the Guard Mechanism

Orthogroup analysis provides a powerful systematic framework for deciphering the complex evolutionary history of NBS genes and identifying functionally important candidates for crop improvement. Through integrated computational and experimental approaches, researchers can distinguish evolutionarily conserved immune components from recent, adaptive innovations. The methodologies outlined in this guide enable comprehensive characterization of NBS gene diversity, evolution, and function, facilitating the discovery of genetic elements crucial for enhancing disease resistance in agricultural systems. As genomic resources continue to expand, orthogroup analysis will play an increasingly vital role in translating evolutionary insights into practical crop protection strategies.

Plant nucleotide-binding site-leucine-rich repeat (NBS-LRR or NLR) genes constitute the largest class of intracellular immune receptors, capable of recognizing pathogen-secreted effectors to trigger robust immune responses known as effector-triggered immunity (ETI). These genes account for approximately 1% of all open reading frames in both Arabidopsis thaliana and Oryza sativa (rice), representing one of the most expansive and dynamic gene families in plant genomes [11] [99]. However, the accurate annotation of NLR genes remains a substantial bioinformatic challenge that directly impacts our understanding of plant immunity mechanisms. These genes are frequently misannotated during automated proteome prediction, and standard identification tools that rely on existing annotations struggle to recover missing NLRs from genomic sequences [100]. This annotation gap is particularly pronounced in non-model species, including medicinal plants and crop wild relatives, which represent valuable reservoirs of disease resistance genes.

The implications of incomplete NLR annotation extend far beyond genomic cataloging. Without comprehensive identification of these immune receptors, researchers cannot fully elucidate plant-pathogen interactions, map resistance quantitative trait loci (QTLs), or develop strategic breeding programs for durable disease resistance. Recent studies have revealed that functional NLRs often exhibit signatures of high expression in uninfected plants across both monocot and dicot species, challenging previous assumptions about their transcriptional repression [101]. This emerging understanding, coupled with advances in computational methodologies, enables more sophisticated in silico validation approaches that combine phylogenetic analysis, cross-referencing of diverse databases, and experimental verification. This technical guide outlines established and emerging frameworks for validating NLR gene annotations, with particular emphasis on phylogenetic cross-referencing and database integration within the context of plant immunity research.

The Challenge: Systemic NLR Misannotation and Its Consequences

Prevalence and Impact of Missing NLR Genes

The problem of NLR misannotation is not trivial; studies indicate that conventional annotation pipelines may overlook a significant proportion of these crucial immune receptors. The development of NLRSeek, a genome reannotation-based pipeline for NLR identification, demonstrated striking gaps in even well-annotated model systems. In the extensively studied Arabidopsis thaliana genome, NLRSeek identified a previously unannotated NLR gene whose expression and translation were confirmed by transcriptome and ribosome-profiling data [100]. The situation is considerably more severe in non-model species with less mature genomic resources. For example, in yam species (Dioscorea spp.), NLRSeek identified 33.8%–127.5% more NLR genes than conventional methods, with 45.1% of the newly annotated NLRs exhibiting detectable expression—strong evidence that they represent functional genes previously overlooked by standard annotation approaches [100].

Structural and Evolutionary Complexities in NLR Annotation

Several biological factors contribute to the challenges in accurate NLR annotation:

• Domain diversity and architecture: NLR proteins are classified based on their N-terminal domains into several major classes, including coiled-coil (CC-NB-LRR or CNL), Toll/interleukin-1 receptor (TIR-NB-LRR or TNL), and resistance to powdery mildew 8 (RPW8-NB-LRR or RNL). These domains are frequently subject to differential expansion and contraction across plant lineages. For instance, comparative analysis of Salvia species revealed a marked reduction in TNL and RNL subfamily members, with some species completely lacking TNL subfamilies [11].
• Atypical configurations: Many NLR proteins lack complete N-terminal or LRR domains and are classified as atypical NBS-LRRs, including subtypes such as N (NBS only), TN (TIR-NBS), CN (CC-NBS), and NL (NBS-LRR). These atypical forms are particularly prone to misannotation [11].
• Genomic clustering: NLR genes are distributed non-randomly throughout plant genomes, often forming complex clusters that facilitate tandem duplication of paralogous sequences and generation of new resistance specificities. These clusters present challenges for assembly and annotation, particularly in regions with high sequence similarity [99].

Table 1: Classification of Plant NLR Genes Based on Domain Architecture

Classification	N-terminal Domain	Central Domain	C-terminal Domain	Representative Examples	Functional Role
CNL	Coiled-coil (CC)	Nucleotide-binding site (NBS)	Leucine-rich repeat (LRR)	Arabidopsis RPS2, RPM1	Effector recognition & immunity activation
TNL	TIR	NBS	LRR	Arabidopsis RPS4	Effector recognition & immunity activation
RNL	RPW8	NBS	LRR	Arabidopsis ADR1	Helper NLR for signaling amplification
N	None	NBS	None	Various	Regulatory functions
TN	TIR	NBS	None	Various	Signaling components
CN	CC	NBS	None	Various	Regulatory functions

Methodological Framework: Integrated In Silico Validation Approaches

The NLRSeek Pipeline for Comprehensive NLR Identification

The NLRSeek pipeline represents a significant advancement in NLR annotation methodology by integrating de novo detection of NLR loci at the genome level with targeted genome reannotation, systematically reconciling these results with existing annotations to produce a comprehensive set of NLR predictions [100]. This approach addresses the fundamental limitation of conventional methods that rely primarily on established proteomic data, which inherently contain gaps for rapidly evolving gene families like NLRs. The workflow employs a multi-tiered strategy that combines homology-based searching, structural feature detection, and expression evidence integration to achieve superior sensitivity while maintaining specificity.

The implementation of NLRSeek involves several critical stages, beginning with whole-genome scanning for NLR-associated structural features and domains, independent of existing gene models. This de novo detection phase identifies genomic loci that exhibit characteristics of NLR genes but may have been missed by standard annotation pipelines. Subsequently, the pipeline performs targeted reannotation of these loci, incorporating evidence from transcriptomic datasets where available. Finally, the results are reconciled with existing annotations to produce a non-redundant, comprehensive set of NLR predictions. This method has demonstrated particular efficacy for non-model species with preliminary annotations, revealing substantial numbers of previously overlooked NLR genes with supporting expression evidence [100].

Phylogenetic Cross-Referencing for Annotation Validation

Phylogenetic analysis provides a powerful orthogonal validation method for NLR annotations by establishing evolutionary relationships among putative NLR genes within and across species. This approach leverages the principle that truly orthologous genes should cluster together in phylogenetic trees based on sequence similarity, while also revealing lineage-specific expansions and contractions that characterize NLR evolution. A robust phylogenetic framework for NLR validation involves several key steps, beginning with the identification of conserved domain architecture across candidate sequences, followed by multiple sequence alignment of these domains, and culminating in tree construction using appropriate evolutionary models.

In practice, phylogenetic validation has revealed important insights into NLR evolution and annotation accuracy. For example, analysis of NLRs from Salvia miltiorrhiza alongside model plants enabled classification according to established CNL, TNL, and RNL subfamilies, while also revealing a marked reduction in TNL and RNL subfamily members within Salvia species—an evolutionary pattern that would be obscured by incomplete annotation [11]. Similarly, phylogenetic approaches have identified novel NLR pairs in wheat with simplified domain architectures, expanding our understanding of the genetic basis of disease resistance in cereals [9]. These cross-species phylogenetic comparisons not only validate annotations but also provide evolutionary context for functional characterization.

Diagram 1: Integrated in silico validation workflow for NLR gene annotation, combining automated annotation, de novo detection, phylogenetic analysis, and experimental validation.

Expression-Based Functional Filtering

Recent evidence challenges the long-held assumption that NLRs are necessarily transcriptionally repressed in uninfected plants. Analysis of known functional NLRs across multiple plant species has revealed that they frequently exhibit high steady-state expression levels in uninfected tissues, with functional NLRs significantly enriched among the most highly expressed NLR transcripts [101]. This observation provides a valuable filtering criterion for prioritizing candidate NLRs from in silico predictions. For example, in Arabidopsis thaliana, known functional NLRs are significantly enriched in the top 15% of expressed NLR transcripts compared with the lower 85%, with the most highly expressed NLR (ZAR1) displaying expression levels above the median and mean for all genes in the accession Col-0 [101].

This expression signature has practical applications in validation pipelines. Researchers can leverage RNA-seq data from uninfected tissues to rank NLR candidates by expression level, prioritizing highly expressed transcripts for functional characterization. This approach proved successful in a large-scale screen of grass NLRs, where expression level combined with high-throughput transformation identified 31 new resistance NLRs in wheat (19 against stem rust and 12 against leaf rust) [101]. The integration of expression data provides a valuable complement to phylogenetic and structural validation methods, offering evidence of transcriptional activity that supports genuine coding potential.

Experimental Validation and Integration

Cross-Referencing with Multi-Omics Data

Comprehensive NLR validation requires integration of diverse data types beyond genomic sequence. Modern pipelines increasingly incorporate transcriptomic, proteomic, and epigenomic evidence to support annotation accuracy and functional potential. Ribosome profiling data can confirm translation of predicted NLR genes, while chromatin accessibility assays provide insights into regulatory potential. The integration of multi-omics data creates a powerful framework for distinguishing functional NLR genes from pseudogenes or annotation artifacts.

The emergence of specialized databases has facilitated more sophisticated cross-referencing approaches. For example, the P-MITE database encompasses miniature inverted-repeat transposable elements (MITEs) from 41 plant genomes, which is particularly relevant given the tendency of MITEs to insert near genes and influence their expression, including NLR genes [73]. Similarly, databases of repetitive elements are invaluable for distinguishing bona fide NLR genes from transposable element fragments, a common source of false positives in NLR annotation. Integration with expression atlases and proteomic resources provides additional layers of validation, creating a multi-dimensional evidence framework for annotation confidence.

In Planta Functional Characterization

While in silico methods provide powerful screening tools, ultimate validation of NLR annotations requires functional assessment in plant systems. Traditional approaches involving stable transformation and pathogen challenge provide definitive evidence but are resource-intensive and low-throughput. Recent advances have enabled more scalable functional validation through high-throughput transformation systems coupled with efficient phenotyping platforms. For example, researchers generated a wheat transgenic array of 995 NLRs from diverse grass species to identify new resistance genes against rust pathogens, demonstrating the power of scale in functional NLR validation [101].

The functional transfer of NLR pairs across taxonomic boundaries provides additional validation of annotation accuracy. Recent studies have shown that paired NLR modules can be functionally transferred between distantly related species to confer disease resistance. For instance, co-transfer of the pepper NLRs Pik-1 and Pik-2 into rice and tomato conferred resistance to corresponding pathogens, demonstrating conserved functionality despite evolutionary distance [9]. Such cross-species complementation assays not only validate annotation accuracy but also provide insights into conserved immune signaling mechanisms across plant taxa.

Table 2: Key Experimental Reagents and Resources for NLR Validation

Resource Type	Specific Examples	Application in NLR Validation	Considerations
Reference Genomes	Arabidopsis thaliana (Col-0), Oryza sativa (Nipponbare)	Phylogenetic benchmarking, synteny analysis	Assembly quality (contig N50, LAI >10), annotation version
Specialized Databases	P-MITE, Repbase, PmiREN2.0, miRBase	Repeat masking, miRNA target prediction	Currency, species coverage, false positive rates
Bioinformatics Tools	NLRSeek, MITE-Hunter, EDTA-TIR-Learner, RepeatMasker	De novo NLR identification, repeat element annotation	Parameter optimization, computational requirements
Expression Resources	RNA-seq datasets, ribosome profiling data	Expression evidence, translational confirmation	Tissue specificity, growth conditions, replication
Validation Platforms	Wheat transgenic array, tobacco transient expression	High-throughput functional screening	Throughput, physiological relevance, pathogen compatibility
Structural Resources	AlphaFold predictions, crystallographic data	Domain boundary verification, functional residue identification	Model confidence metrics, experimental validation

Case Studies and Applications

Comparative NLR Annotation Across Plant Lineages

Large-scale comparative analyses have revealed striking variation in NLR composition and evolution across plant lineages, highlighting the importance of accurate annotation for understanding plant immunity evolution. In Salvia miltiorrhiza (Danshen), a medicinal plant, researchers identified 196 NBS-LRR genes through genome-wide analysis, but only 62 possessed complete N-terminal and LRR domains, underscoring the prevalence of atypical NLRs and the importance of domain-aware annotation approaches [11]. Phylogenetic analysis placed these NLRs within established CNL, TNL, and RNL subfamilies, while also revealing a marked reduction in TNL and RNL members compared to other angiosperms—an evolutionary pattern with potential functional implications for immune signaling in this species.

The application of advanced annotation pipelines to non-model species has uncovered previously hidden genetic resources for disease resistance breeding. In yam (Dioscorea spp.) species, the NLRSeek pipeline identified 33.8%–127.5% more NLR genes than conventional methods, with nearly half of the newly annotated NLRs showing detectable expression [100]. Subsequent analysis revealed that NLRs have undergone expansion in D. zingiberensis through tandem duplication—an evolutionary insight that was not attainable using previous NLR annotation tools. These findings demonstrate how improved in silico validation methods can reveal untapped genetic resources for engineering disease-resistant crops.

Structural Insights Guiding Annotation Accuracy

Recent advances in structural biology have provided new dimensions for NLR annotation validation by elucidating conserved structural features that define functional NLR proteins. Structural studies have revealed how NLRs assemble into oligomeric resistosomes, with ZAR1 and Sr35 forming Ca²⁺-permeable channels, and TNL resistosomes acting as NADases to generate signaling molecules [13]. These structural insights enable more sophisticated sequence-based annotation through the identification of conserved functional motifs and domain interfaces essential for NLR function.

The expanding understanding of NLR pairs and networks provides additional criteria for annotation validation. Studies have identified novel NLR pairs in wheat with simplified domain architectures, organized in head-to-head orientation [9]. Interestingly, functional analysis revealed that the head-to-head orientation is not essential for the function of these NLR pairs, as random insertion of the two genes into a susceptible wheat variety still conferred resistance. This flexibility in genetic organization has important implications for annotation, suggesting that NLRs traditionally classified as atypical due to domain truncations may in fact represent functional components of paired immune receptors.

Diagram 2: NLR protein domain architecture and signaling pathways, showing the structural components and downstream immune activation mechanisms.

Future Directions and Implementation Recommendations

Emerging Technologies and Methodologies

The field of NLR annotation and validation continues to evolve rapidly, with several emerging technologies promising to enhance accuracy and comprehensiveness. The integration of long-read sequencing technologies enables more complete genome assemblies, particularly in complex regions where NLR clusters reside. Similarly, optical mapping and chromatin conformation data help resolve tandemly duplicated NLR arrays that have traditionally challenged short-read assembly approaches. These technological advances are progressively eliminating the assembly gaps that contribute to NLR misannotation.

The application of machine learning approaches represents another promising direction for NLR annotation. As structural and functional data accumulate for NLR proteins, supervised learning models can be trained to recognize subtle sequence features that distinguish functional NLRs from pseudogenes or non-immunity related NBS-containing proteins. The integration of protein structure prediction tools like AlphaFold further enhances annotation confidence by enabling in silico validation of predicted domain boundaries and tertiary structures. These computational advances, coupled with the growing availability of plant genomic resources, promise to progressively close the NLR annotation gap across diverse plant species.

Recommendations for Implementation

Based on current best practices and emerging methodologies, researchers undertaking NLR annotation and validation should consider the following implementation recommendations:

• Employ integrated pipelines: Combine multiple complementary approaches rather than relying on single methods, integrating homology-based, de novo, and expression-aware annotation strategies.
• Leverage phylogenetic cross-referencing: Include phylogenetically diverse reference species in analyses to establish evolutionary context and identify lineage-specific innovations.
• Incorporate expression evidence: Utilize available transcriptomic data to prioritize candidates with support for expression, particularly those with high steady-state expression levels characteristic of functional NLRs.
• Validate structural features: Confirm the presence of conserved domain architectures and functional motifs essential for NLR function, using both sequence-based and predicted structural approaches.
• Implement tiered validation: Establish a multi-tiered validation framework progressing from in silico prediction to experimental confirmation, with clear criteria for advancement between stages.

The systematic implementation of these recommendations will enhance the accuracy and comprehensiveness of NLR annotations, ultimately advancing our understanding of plant immunity and creating new opportunities for crop improvement through informed manipulation of disease resistance pathways.

Accurate annotation of NLR genes through robust in silico validation methodologies represents a foundational requirement for advancing plant immunity research and breeding. The integration of phylogenetic cross-referencing, multi-omics data integration, and structural analysis creates a powerful framework for distinguishing functional NLR genes from annotation artifacts, revealing the full complement of these crucial immune receptors across diverse plant species. As these methodologies continue to evolve alongside emerging technologies and expanding genomic resources, researchers are positioned to increasingly unlock the genetic potential of NLR-mediated immunity for crop improvement and sustainable agriculture. The systematic application of these in silico validation approaches will accelerate the discovery and characterization of disease resistance genes, ultimately contributing to enhanced food security through the development of durably resistant crop varieties.

Within the broader thesis on the role of nucleotide-binding site (NBS) domain genes in plant immunity, this guide addresses the critical phase of transcriptomic validation. Nucleotide-binding site-leucine-rich repeat (NBS-LRR) genes constitute the largest family of plant disease resistance (R) genes, encoding intracellular receptors that recognize pathogen effectors and activate effector-triggered immunity (ETI) [29] [5]. While traditionally associated with biotic stress resistance, growing evidence confirms their significant involvement in abiotic stress responses, including salinity, drought, and hormone signaling [102] [63]. Transcriptomic validation bridges computational identification of NBS genes with functional characterization, providing insights into their expression patterns, regulation, and potential roles in plant stress responses. This technical guide outlines comprehensive methodologies and analytical frameworks for the transcriptomic assessment of NBS gene expression under diverse stress conditions, providing researchers with standardized approaches to validate their protective functions.

Quantitative Profiling of NBS Gene Expression Under Stress

Transcriptomic studies across multiple plant species have quantified dynamic expression patterns of NBS genes when challenged by biotic and abiotic stressors. The following tables consolidate key quantitative findings from recent investigations.

Table 1: NBS Gene Expression Under Biotic Stresses

Plant Species	Pathogen/Stressor	Number of NBS Genes Analyzed	Key Expression Findings	Reference
Dendrobium officinale	Salicylic Acid (SA) Treatment	22 NBS-LRR genes	6 genes significantly upregulated (Dof013264, Dof020566, Dof019188, Dof019191, Dof020138, Dof020707)	[63]
Nicotiana benthamiana	Pseudomonas fluorescens (PTI activation)	Genome-wide transcriptomics	10,300 differentially expressed genes from 57,139 predicted genes	[103]
Gossypium hirsutum (Cotton)	Cotton Leaf Curl Disease (CLCuD)	Multiple orthogroups	Putative upregulation of OG2, OG6, and OG15 orthogroups in tolerant plants	[5]

Table 2: NBS Gene Expression Under Abiotic Stresses

Plant Species	Abiotic Stress	Number of NBS Genes Analyzed	Key Expression Findings	Reference
Lathyrus sativus (Grass pea)	Salt stress (NaCl)	9 genes validated by qPCR	Majority showed upregulation at 50 and 200 μM NaCl; LsNBS-D18, LsNBS-D204, LsNBS-D180 showed reduced or drastic downregulation	[102] [104]
Lathyrus sativus (Grass pea)	Various stresses	274 identified NBS-LRR genes	85% of encoded genes showed high expression levels in RNA-Seq analysis	[102] [54]
Dendrobium officinale	Hormone signaling	22 NBS-LRR genes	Genes participate in plant hormone signal transduction and Ras signaling pathways	[63]

Table 3: NBS-LRR Gene Classification Across Plant Species

Plant Species	Total NBS Genes	TNL Genes	CNL Genes	RNL Genes	Reference
Lathyrus sativus (Grass pea)	274	124	150	Not specified	[102] [54]
Arabidopsis thaliana	210	Not specified	40 (CNL-type)	Not specified	[63]
Dendrobium officinale	74	0	10 (CNL-type)	Not specified	[63]
Rosaceae species (12 genomes)	2188	Variable across species	Variable across species	Variable across species	[105]

Experimental Protocols for Transcriptomic Validation

Genome-Wide Identification and Classification of NBS Genes

Protocol 1: Identification of NBS-LRR Genes from Genome Assemblies

• Data Acquisition: Obtain genomic data from public databases such as NCBI (e.g., Grass pea genotype LS007, NCBI ID: CABITX010000000) [102] [54].
• Sequence Similarity Search: Perform Local TBLASTN searches using known NBS-LRR protein sequences from related species (e.g., chickpea, apple, Brassica napus) with a sequence similarity threshold of 90% and sequence length of 600 nucleotides [102].
• Coding Region Prediction: Use TransDecoder (Release v5.5.0) or similar tools to predict potential coding regions from the identified sequences.
• Domain Verification: Screen protein sequences for conserved domains using hmmsearch from the HMMER package (v3.1b2) with the NBS domain (pfam00931) from the Pfam database. Follow with NCBI-CDD tool for conserved domain verification [102].
• Gene Structure Prediction: Use AUGUSTUS tool (version 3.3) or similar software to evaluate gene structure and predict alternative transcripts [102].
• Classification: Classify verified NBS genes into subfamilies (TNL, CNL, RNL) based on N-terminal domains (TIR, CC, or RPW8) using phylogenetic analysis with reference sequences from related species [102] [105].

Transcriptome Sequencing and Expression Analysis

Protocol 2: RNA-Sequencing for Expression Profiling Under Stress

• Experimental Design:

  • Apply biotic (pathogen inoculation) or abiotic (salt, drought, hormone) stresses to plant materials with appropriate controls.
  • Include multiple time points (e.g., 6h, 12h, 24h post-treatment) and biological replicates (minimum 3) [103].

• RNA Extraction and Quality Control:

  • Extract total RNA using commercial kits (e.g., Qiagen RNeasy Plant Mini Kit).
  • Assess RNA quality using Nanodrop (A260/A280 ratio 1.8-2.0), Qubit for quantification, and agarose gel electrophoresis to check for degradation [106].

• Library Preparation and Sequencing:

  • Use NEXTFLEX Rapid DNA-seq kit or similar for Illumina platform library preparation.
  • Fragment RNA (200-250 bp), ligate to barcoded adaptors, and PCR-amplify (4 cycles) [106].
  • Perform paired-end sequencing (e.g., 150 cycles) on Illumina platforms (HiSeq X Ten) [106].
  • For long-read sequencing, prepare libraries using ligation sequencing kit (SQK-LSK109) and sequence on GridION X5 (Oxford Nanopore Technologies) with a SpotON flow cell R9.4 [106].

• Bioinformatic Analysis:

  • Process raw reads: quality control (FastQC), adapter trimming (Trimmomatic), and alignment to reference genome (HISAT2, STAR).
  • Assemble transcripts and quantify gene expression (StringTie, Cufflinks).
  • Identify differentially expressed genes (DEGs) using packages like DESeq2 or edgeR with threshold of |log2FC| > 1 and FDR < 0.05 [63].

qPCR Validation of NBS Gene Expression

Protocol 3: Quantitative Real-Time PCR Validation

• Candidate Gene Selection: Select NBS genes showing significant differential expression in RNA-seq analysis [102].
• Reference Gene Validation: Identify and validate stable reference genes for normalization (e.g., NbUbe35, NbNQO, and NbErpA in N. benthamiana) using algorithms (geNorm, NormFinder, BestKeeper) [103].
• cDNA Synthesis: Perform reverse transcription with 1μg total RNA using High-Capacity cDNA Reverse Transcription Kit with RNase inhibitor.
• Primer Design and Validation:
  • Design gene-specific primers (amplicon size 80-150 bp) using Primer-BLAST.
  • Validate primer specificity through melting curve analysis (single peak) and ensure amplification efficiency (90-110%) using cDNA dilution series [103].
• qPCR Reaction: Set up reactions with SYBR Green Master Mix, run in technical triplicates on real-time PCR systems with cycling conditions: 95°C for 10 min, followed by 40 cycles of 95°C for 15s and 60°C for 1 min.
• Data Analysis: Calculate relative expression using the 2^(-ΔΔCt) method with stable reference genes for normalization [102].

Signaling Pathways in NBS-Mediated Stress Response

The following diagram illustrates the integrated signaling pathways through which NBS genes participate in plant stress responses, particularly in the ETI system and cross-talk with hormone signaling.

This pathway diagram illustrates how NBS-LRR proteins function as central hubs in plant stress responses. They recognize pathogen effectors directly or indirectly through cellular damage, activating ETI and the hypersensitive response [29] [63]. Transcriptomic studies reveal significant crosstalk between biotic and abiotic stress signaling, where NBS gene expression is modulated by hormone pathways (SA, JA, ABA) and influences downstream MAPK and Ras signaling cascades, ultimately regulating defense gene expression and enhancing overall stress resistance [102] [63].

The Scientist's Toolkit: Research Reagent Solutions

Table 4: Essential Research Reagents for NBS Gene Transcriptomic Studies

Reagent/Resource	Specific Examples	Function/Application	Technical Notes
Genome Databases	NCBI Genome, Phytozome, Plaza, Rosaceae.org	Source of genomic sequences for identification	Grass pea genome (NCBI ID: CABITX010000000) [102]
Domain Databases	Pfam (PF00931), NCBI-CDD, SMART	Identification of NBS, TIR, CC, LRR domains	Use HMMER with Pfam models [102] [5]
Sequencing Platforms	Illumina (HiSeq X Ten), Oxford Nanopore (GridION X5)	Whole genome and transcriptome sequencing	Hybrid assembly recommended [106]
NBS Identification Tools	HMMER (hmmsearch), BLAST, PfamScan.pl	Identification of NBS domain-containing genes	e-value threshold 1.1e-50 recommended [5]
qPCR Reference Genes	NbUbe35, NbNQO, NbErpA (N. benthamiana)	Normalization of qPCR data	Validate stability with geNorm, NormFinder [103]
Differential Expression Tools	DESeq2, edgeR, Cufflinks	Identification of differentially expressed NBS genes	Threshold: |log2FC| > 1, FDR < 0.05 [63]
Plant Growth Regulators	Salicylic Acid, Methyl Jasmonate, Abscisic Acid	Treatment to study NBS gene regulation in signaling	103 transcription factors identified upstream of NBS genes respond to these [102]

Transcriptomic validation provides crucial evidence for understanding the functional roles of NBS genes in plant stress responses. The integrated methodologies outlined in this guide—from genome-wide identification and RNA-seq analysis to qPCR validation—enable researchers to confidently characterize NBS gene expression patterns under diverse stress conditions. The consistent findings of NBS gene involvement in both biotic and abiotic stress responses across species highlights the versatility of this gene family and its potential as a target for crop improvement strategies. Future research should focus on functional validation of specific NBS genes through genetic manipulation and the exploration of their signaling networks, particularly the cross-talk between different stress response pathways. Standardization of these transcriptomic approaches will facilitate comparative analyses across plant species and accelerate the development of stress-resistant crop varieties.

This technical guide explores the functional validation of nucleotide-binding site (NBS) domain genes in plant immunity, with a specific case study on the silencing of GaNBS in cotton and its consequential impact on viral pathogen titers. The article details the experimental workflow, presents quantitative data, and provides visualization of the underlying signaling pathways. Framed within the broader context of plant immune receptor research, this work highlights the critical role of NBS-domain-containing genes in effector-triggered immunity (ETI) and demonstrates the utility of virus-induced gene silencing (VIGS) as a rapid reverse genetics tool for functional genomics in polyploid crops like cotton.

Plant immunity relies on a sophisticated two-tiered system to defend against pathogens. The first layer, pattern-triggered immunity (PTI), is initiated by cell-surface pattern recognition receptors (PRRs). The second layer, effector-triggered immunity (ETI), is primarily mediated by intracellular nucleotide-binding and leucine-rich-repeat receptors (NLRs) that contain a central Nucleotide-Binding Site (NBS) domain [10] [3]. These NBS-domain-containing genes are one of the largest and most variable gene families in plants, involved in recognizing pathogen effectors and initiating a robust immune response, often including a localized programmed cell death known as the hypersensitive response (HR) to confine pathogens [5] [3].

NLR genes are modular proteins typically comprised of three fundamental components:

• An N-terminal domain (either a Toll/Interleukin-1 Receptor (TIR) or a Coiled-Coil (CC) domain).
• A central NB-ARC (NBS) domain, which acts as a molecular switch regulated by nucleotide states (ADP/ATP).
• A C-terminal Leucine-Rich Repeat (LRR) domain responsible for effector recognition [5] [3].

The NLR family has undergone significant expansion in flowering plants, with some species harboring thousands of members, creating a diverse repertoire for pathogen recognition [5]. Cotton leaf curl disease (CLCuD), caused by begomoviruses, is a devastating disease, and NLRs are a main class of resistance genes responding to such viral infections [5]. Understanding the specific function of individual NBS genes is therefore crucial for developing durable disease resistance in crops.

VIGS: A Key Tool for Functional Genomics in Cotton

Virus-Induced Gene Silencing (VIGS) is an RNA interference-mediated reverse genetics technique that has become an effective tool for investigating gene function in plants [107]. It knocks down gene expression through post-transcriptional gene silencing (PTGS) by engineering viral vectors to contain sequences homologous to the host target gene. When infected, the plant's defense machinery targets both the virus and the corresponding endogenous mRNA for degradation [107].

In cotton, which has large genomes, polyploidy, and challenging transformation, VIGS provides a fast and cost-efficient alternative to stable transformation for validating gene function [107]. Two viral vector systems are particularly effective in cotton:

• Tobacco Rattle Virus (TRV)-based VIGS: A bilateral positive-sense single-stranded RNA virus. RNA2 is modified to insert the target gene fragment [107].
• Cotton Leaf Crumple Virus (CLCrV)-based VIGS: A bipartite single-stranded DNA geminivirus. The DNA-A component can be engineered to carry the target sequence [107].

The following diagram illustrates the core workflow of the VIGS mechanism.

Case Study: Functional Validation of GaNBS in Virus Tittering

Experimental Background and Rationale

A comprehensive study identified 12,820 NBS-domain-containing genes across 34 plant species, uncovering significant diversity and numerous orthogroups (OGs) [5]. Expression profiling in cotton under biotic stress revealed that certain orthogroups, including OG2, OG6, and OG15, were putatively upregulated. This case study focuses on the functional validation of a specific NBS gene, GaNBS, a member of OG2, to confirm its role in conferring resistance to Cotton Leaf Curl Disease (CLCuD) [5].

The study utilized contrasting Gossypium hirsutum accessions: a tolerant variety (Mac7) and a susceptible variety (Coker 312). Genetic variation analysis identified a greater number of unique variants in the NBS genes of the tolerant Mac7 (6,583 variants) compared to the susceptible Coker 312 (5,173 variants), suggesting a potential link between NBS diversity and disease resilience [5].

Detailed Experimental Protocol

Target Gene Selection and Vector Construction

• Gene Selection: The GaNBS gene (from orthogroup OG2) was selected based on transcriptomic data showing its upregulation in response to CLCuD in resistant cotton [5].
• Insert Amplification: A ~300-400 base pair fragment specific to the GaNBS coding sequence was amplified via PCR.
• Vector Assembly: The amplified fragment was cloned into a multiple cloning site of a VIGS vector, such as the TRV2 plasmid [107]. The recombinant plasmid was then transformed into Agrobacterium tumefaciens strain GV3101.

Plant Material and Agro-infiltration

• Plant Growth: Resistant cotton plants (e.g., G. arboreum or the tolerant Mac7 accession) were grown under controlled conditions until the 2-4 true leaf stage.
• Agrobacterium Culture: Transformed Agrobacterium cultures were grown overnight, pelleted, and resuspended in an induction medium (e.g., with acetosyringone) to a final OD₆₀₀ of 1.0-2.0 [107].
• Infiltration: The bacterial suspension was injected into the abaxial side of cotyledons or true leaves using a needleless syringe. Control plants were infiltrated with:
  • Empty Vector (EV) Control: TRV2 vector without an insert.
  • Positive Control: TRV2 vector containing a fragment of a marker gene like Phytoene Desaturase (PDS) or Chloroplastos Alterados 1 (CLA1), which produces a visible photobleaching phenotype [107].

Challenge Inoculation and Phenotyping

• After 2-3 weeks of silencing establishment, both GaNBS-silenced and control plants were challenge-inoculated with Cotton Leaf Curl Virus via agro-infiltration or viruliferous whiteflies (Bemisia tabaci) [5].
• Plants were monitored daily for disease symptom development (leaf curling, vein thickening, stunting) over a period of 2-4 weeks.

Molecular and Biochemical Analysis

• Silencing Efficiency Check: RNA was extracted from leaf tissue, and RT-qPCR was performed using GaNBS-specific primers to confirm the knockdown of the target gene transcript relative to control plants. An internal control gene like Ubiquitin was used for normalization.
• Virus Titer Quantification: Viral DNA accumulation was measured using:
  • qPCR: Quantifying viral genomic components with primers specific to a CLCuV gene (e.g., AC1 or AV1).
• Protein-Ligand Interaction: Computational docking and/or yeast two-hybrid assays were performed to investigate the physical interaction between the GaNBS protein and viral proteins (e.g., CLCuV Rep or CP) or nucleotides (ADP/ATP) [5].

Key Findings and Quantitative Data

The functional validation of GaNBS via VIGS yielded critical insights into its role in plant immunity.

• Silencing Confirmation: RT-qPCR analysis confirmed a significant reduction (e.g., >70%) in GaNBS transcript levels in silenced plants compared to empty vector controls [5].
• Phenotypic Impact: Plants with silenced GaNBS lost their resistant phenotype and showed significantly more severe CLCuD symptoms (leaf curling, stunting) after viral challenge compared to control plants [5].
• Impact on Virus Titers: Crucially, viral titer quantification demonstrated a significant increase in virus accumulation in GaNBS-silenced plants, directly demonstrating that GaNBS is required to limit viral replication, a process known as virus tittering [5].
• Protein Interaction: Protein-ligand interaction studies showed strong binding of the GaNBS protein (and other putative NBS proteins from the study) with ADP/ATP and with core proteins of the cotton leaf curl disease virus, suggesting a direct mechanistic role in pathogen recognition or immune signaling [5].

Table 1: Quantitative Results from GaNBS Silencing Experiment in Resistant Cotton

Parameter Measured	GaNBS-Silenced Plants	Empty Vector Control Plants	Measurement Technique
GaNBS Relative Expression	Strong decrease (e.g., <30% of control)	100% (baseline)	RT-qPCR
Disease Symptom Severity	Severe leaf curling and stunting	Mild or no symptoms	Visual phenotyping
Viral DNA Accumulation	Significant increase (e.g., 5-10 fold higher)	Low baseline level	qPCR
Protein-Ligand Interaction	Strong interaction with ADP/ATP and viral proteins	Not Applicable	Computational docking / Yeast two-hybrid

The experimental workflow and the key finding—that silencing GaNBS impairs virus tittering—are summarized in the diagram below.

The Scientist's Toolkit: Research Reagent Solutions

The following table lists essential reagents and resources for executing VIGS-based functional validation studies in cotton, as exemplified in the case study.

Table 2: Key Research Reagents for VIGS-based Functional Validation in Cotton

Reagent / Resource	Function / Purpose	Examples / Notes
VIGS Vectors	To deliver the host-derived gene fragment into plant cells and initiate silencing.	TRV-based vectors (pYL156/TRV1, pYL192/TRV2), CLCrV-based vectors [107].
Agrobacterium tumefaciens Strain	The bacterial workhorse for delivering the VIGS vector DNA into plant tissues.	GV3101, LBA4404 [107].
Marker Gene Constructs	To visually confirm the success and efficiency of the VIGS system in experimental plants.	TRV2::PDS, TRV2::CLA1 (cause photobleaching), TRV2::ANS (causes color change) [107].
Target-Specific Constructs	To silence the gene of interest and study its function.	TRV2::GaNBS, TRV2::GH1 (for abiotic stress studies) [5] [108].
qPCR / RT-qPCR Assays	To quantitatively measure silencing efficiency (mRNA knockdown) and pathogen titer.	Requires primers for target gene (e.g., GaNBS), pathogen genome, and internal reference genes (e.g., Ubiquitin, GAPDH).
Pathogen Inoculum	To challenge silenced plants and assess the functional role of the gene in resistance/susceptibility.	Purified viral clones for agro-infection, or viruliferous insect vectors (e.g., Bemisia tabaci for CLCuD) [5].

The functional validation of GaNBS using VIGS provides a compelling case study that directly links a specific NBS-domain gene to virus resistance in cotton. The results demonstrate that GaNBS plays a critical role in virus tittering, limiting the accumulation of the Cotton Leaf Curl Virus. This work underscores the power of VIGS as a rapid and effective tool for functional genomics in complex crops. Furthermore, it highlights the importance of characterizing the vast repertoire of NBS genes to fully understand the plant immune system and identify key genetic components that can be leveraged to engineer durable, broad-spectrum disease resistance in crops, aligning with the growing field of synthetic plant immunity [10] [3].

The nucleotide-binding site-leucine-rich repeat (NBS-LRR) gene family constitutes the largest and most critical class of plant disease resistance (R) proteins, serving as intracellular immune receptors that initiate effector-triggered immunity (ETI) upon pathogen recognition [11] [16]. Beyond their established role in biotic stress responses, emerging evidence suggests these genes participate in complex signaling networks that also mediate responses to abiotic stresses, including salinity [109]. This technical guide provides a comprehensive framework for employing quantitative PCR (qPCR) and mutant analysis to validate the functional role of specific NBS-LRR genes in plant responses to concurrent salt and disease stresses. Within the broader context of plant immunity research, this integrated approach enables researchers to dissect the molecular mechanisms through which NBS genes coordinate defense signaling pathways and potentially contribute to stress cross-tolerance, offering insights vital for developing improved crop varieties with enhanced dual-stress resilience.

The NBS-LRR Gene Family in Plant Immunity

Structural Classification and Functional Mechanisms

NBS-LRR proteins are characterized by a conserved nucleotide-binding site (NBS) domain and a C-terminal leucine-rich repeat (LRR) domain. The NBS domain is responsible for ATP/GTP binding and hydrolysis, acting as a molecular switch for immune activation, while the LRR domain facilitates pathogen recognition through direct or indirect interaction with pathogen-secreted effectors [11] [16]. Based on their N-terminal domains, NBS-LRR proteins are classified into several major subfamilies:

• TNL: Contains a Toll/Interleukin-1 receptor (TIR) domain
• CNL: Contains a Coiled-coil (CC) domain
• RNL: Contains a Resistance to powdery mildew 8 (RPW8) domain

Additionally, atypical NBS-LRR proteins exist that lack complete N-terminal or LRR domains, classified as TN (TIR-NBS), CN (CC-NBS), NL (NBS-LRR), or N (NBS only) types [11] [16]. These atypical forms often function as adaptors or regulators for typical NBS-LRR proteins.

Upon pathogen recognition, NBS-LRR proteins undergo conformational changes from ADP-bound (inactive) to ATP-bound (active) states, triggering downstream signaling cascades that frequently result in a hypersensitive response (HR) and programmed cell death at infection sites to restrict pathogen spread [16]. Recent studies have revealed that the traditional dichotomy between effector-triggered immunity (ETI) and pathogen-associated molecular pattern-triggered immunity (PTI) represents overlapping defense continua rather than distinct pathways, with NBS-LRR proteins playing integral roles in both processes [11].

Genomic Distribution and Variation Across Species

The composition and size of the NBS-LRR gene family exhibit remarkable diversity across plant species, reflecting adaptations to specific pathogen pressures and evolutionary histories. Table 1 summarizes the distribution of NBS-LRR genes across various plant species based on recent genome-wide analyses.

Table 1: Comparative Analysis of NBS-LRR Gene Family Across Plant Species

Plant Species	Total NBS-LRR Genes	CNL	TNL	RNL	Atypical	Reference
Salvia miltiorrhiza	196	61	2	1	132	[11]
Nicotiana benthamiana	156	25	5	-	126	[16]
Arabidopsis thaliana	207	-	-	-	-	[11]
Oryza sativa (rice)	505	-	0	0	-	[11]
Solanum tuberosum (potato)	447	-	-	-	-	[11]
Pinus taeda (loblolly pine)	311	-	~89.3%	-	-	[11]

Notably, comparative genomic analyses reveal significant lineage-specific expansions and contractions. For instance, monocot species like rice have completely lost TNL and RNL subfamilies, while gymnosperms like Pinus taeda exhibit dramatic expansion of TNL genes [11]. In medicinal plants like Salvia miltiorrhiza, there is a marked reduction in TNL and RNL subfamily members compared to model plants, with only 62 of 196 identified NBS-LRR genes possessing complete N-terminal and LRR domains [11]. This diversity underscores the importance of species-specific characterization of NBS-LRR genes before designing functional validation experiments.

Experimental Design for NBS Gene Validation

Integrated Workflow for qPCR and Mutant Analysis

A robust validation strategy combines gene expression analysis through qPCR with functional characterization using mutant plants. The following diagram illustrates the comprehensive experimental workflow:

Stress Treatment Design

Effective validation requires carefully controlled stress treatments that mimic realistic field conditions while allowing for precise molecular analysis. The res tomato mutant study demonstrates how salt stress can be applied to observe phenotypic recovery and associated gene expression changes [109]. For combined stress experiments, the following treatment structure is recommended:

• Control conditions: Optimal growth environment
• Salt stress: 150-200 mM NaCl for 5-14 days, depending on species sensitivity
• Pathogen challenge: Inoculation with relevant pathogens (e.g., Pseudomonas syringae for bacterial resistance, TMV for viral resistance)
• Combined stress: Simultaneous application of salt and pathogen treatments

The duration and intensity of stress treatments should be optimized for each plant system based on preliminary phenotypic assessments. For instance, in the res tomato mutant, 5 days of 200 mM NaCl treatment was sufficient to observe phenotypic recovery, making it an appropriate timepoint for transcriptomic analysis [109].

qPCR Experimental Protocols

RNA Extraction and Quality Control

High-quality RNA is essential for reliable qPCR results. The following protocol ensures RNA integrity and purity:

• Tissue Collection: Flash-freeze tissue samples in liquid nitrogen immediately after collection. Store at -80°C until processing.
• RNA Extraction: Use commercial kits with DNase I treatment to eliminate genomic DNA contamination.
• Quality Assessment:
  • Determine RNA concentration using spectrophotometry (NanoDrop)
  • Verify RNA integrity via agarose gel electrophoresis (clear 28S and 18S rRNA bands)
  • Confirm purity (A260/A280 ratio of 1.8-2.0, A260/A230 ratio >2.0)
• RNA Normalization: Dilute all samples to equal concentration (e.g., 100 ng/μL) for cDNA synthesis.

cDNA Synthesis and qPCR Setup

Following RNA extraction, proceed with reverse transcription and qPCR preparation:

• cDNA Synthesis:

  • Use 0.5-1 μg total RNA per reaction
  • Employ reverse transcriptase with random hexamers and oligo-dT primers
  • Include negative controls without reverse transcriptase (-RT) to detect genomic contamination

• qPCR Reaction Setup:

  • Use SYBR Green or TaqMan chemistry according to experimental requirements
  • Prepare master mixes to minimize pipetting error
  • Set up triplicate technical replicates for each biological sample
  • Include no-template controls (NTC) to detect reagent contamination

Table 2: qPCR Reaction Components for SYBR Green Assay

Component	Final Concentration	Volume per Reaction (μL)
SYBR Green Master Mix (2X)	1X	5.0
Forward Primer (10 μM)	0.5 μM	0.5
Reverse Primer (10 μM)	0.5 μM	0.5
cDNA Template	-	2.0
Nuclease-free Water	-	2.0
Total Volume	-	10.0

Primer Design and Validation

Proper primer design is critical for specific and efficient amplification:

• Design Parameters:

  • Amplicon length: 80-150 bp
  • Primer length: 18-22 nucleotides
  • Melting temperature (Tm): 58-62°C, with <1°C difference between forward and reverse primers
  • GC content: 40-60%

• Validation Steps:

  • Check specificity by BLAST analysis against the species' genome
  • Verify amplification efficiency using standard curves (90-110% efficiency acceptable)
  • Confirm single amplification product through melt curve analysis

• Reference Gene Selection:

  • Validate multiple candidate reference genes (e.g., EF1α, ACTIN, GAPDH, UBQ)
  • Test stability across all experimental conditions using algorithms like geNorm or NormFinder
  • Use minimum of two reference genes for normalization

Data Analysis and Quality Assessment

The "Dots in Boxes" visualization method provides an efficient approach for quality assessment across large qPCR datasets [110]. This method plots calculated efficiency against delta Cq (distance between the last template solution and no-template control), translating 18 wells of data per target into a single dot. Quality scores (1-5) based on curve sigmoidality and triplicate Cq tightness determine dot size, with larger dots representing higher quality data [110].

For data analysis:

• Cq Determination: Set threshold consistently in the exponential phase of all amplifications
• Calculation of ΔCq: Subtract Cq of reference gene from Cq of target gene for each sample
• Calculation of ΔΔCq: Compare ΔCq values between experimental and control groups
• Fold Change Calculation: Express results as 2^(-ΔΔCq)
• Statistical Analysis: Perform appropriate tests (e.g., t-tests, ANOVA) on ΔCq values, not fold changes

Mutant Analysis Approaches

Characterizing NBS Gene Mutants

The analysis of plant mutants with altered stress responses provides powerful insights into NBS gene function. The res tomato mutant exemplifies how phenotypic characterization under stress conditions can reveal important genetic networks [109]. Key aspects of mutant analysis include:

• Phenotypic Assessment:

  • Document growth patterns, morphological alterations, and stress recovery capabilities
  • Measure physiological parameters (chlorophyll content, photosynthetic efficiency, oxidative damage markers)
  • In the res mutant, notable characteristics included leaf chlorosis under control conditions and phenotypic normalization under salt stress [109]

• Transcriptomic Profiling:

  • Identify differentially expressed genes between mutant and wild-type plants
  • Analyze expression patterns under control and stress conditions
  • In res mutant roots, 3046 DEGs were identified under control conditions versus only 295 under salt stress, indicating transcriptional normalization coinciding with phenotypic recovery [109]

• Pathway Analysis:

  • Map DEGs to functional categories using systems like Mapman
  • Identify enriched pathways, particularly in hormone signaling, transcription factors, and stress responses
  • The res mutant showed constitutive alteration of jasmonate and ethylene pathways, with expression differences attenuated under salt stress [109]

Signaling Pathways in NBS-Mediated Stress Responses

NBS-LRR genes function within complex signaling networks that integrate responses to both biotic and abiotic stresses. The following diagram illustrates key pathways and their interactions:

This integrated signaling network explains the growth-defense tradeoff often observed in plants with constitutive NBS-LRR activation, such as the res mutant which exhibits growth inhibition under normal conditions but enhanced stress tolerance [109].

The Scientist's Toolkit: Research Reagent Solutions

Table 3: Essential Research Reagents for NBS Gene Validation Experiments

Reagent/Category	Specific Examples	Function/Application	Technical Notes
qPCR Reagents	Luna Universal qPCR Master Mix, SYBR Green kits	Sensitive detection of NBS gene expression	Select kits with high efficiency (90-110%) and robust performance across different template concentrations [110]
RNA Extraction Kits	Commercial kits with DNase I treatment	High-quality RNA isolation for transcriptomic studies	Ensure RNA Integrity Number (RIN) >8.0 for reliable results
Reverse Transcriptase	M-MLV, Superscript IV	cDNA synthesis from RNA templates	Use random hexamers and oligo-dT primers for comprehensive coverage
Primer Design Tools	Primer-BLAST, NCBI Primer Designing Tool	Specific primer design for NBS gene targets	Validate primer specificity against entire genome to avoid pseudogene amplification
Reference Genes	EF1α, ACTIN, UBQ, GAPDH	Normalization of qPCR data	Validate stability across all experimental conditions before use
Bioinformatics Tools	HMMER, Pfam, MEME, PlantCARE	Identification and characterization of NBS-LRR genes	Use HMM profile PF00931 (NB-ARC domain) for initial identification [16]
Plant Stress Inducers	NaCl, pathogen isolates (e.g., TMV, P. syringae)	Application of biotic and abiotic stresses	Optimize concentration and duration for specific plant species

Data Interpretation and Integration

Correlation of Expression Patterns with Phenotypic Outcomes

Successful validation of NBS gene function requires careful correlation of expression data with phenotypic observations. Key considerations include:

• Expression-Phenotype Relationships:

  • Determine whether NBS gene upregulation correlates with enhanced stress tolerance
  • Identify if expression patterns differ between single and combined stress conditions
  • Assess whether expression changes precede or follow phenotypic manifestations

• Subfunctionalization Analysis:

  • Evaluate whether different NBS gene clades show distinct expression patterns
  • Determine if CNL, TNL, and RNL subfamilies respond differently to salt versus pathogen stress
  • In Salvia miltiorrhiza, specific SmNBS genes clustered with well-characterized R proteins from model plants, suggesting similar functions [11]

• Cross-Talk Assessment:

  • Analyze expression of hormone pathway genes alongside NBS genes
  • Identify potential regulatory relationships between different signaling pathways
  • In the res mutant, constitutive jasmonate pathway activation influenced both development and stress response networks [109]

Technical Validation and Troubleshooting

Ensure data reliability through rigorous validation:

• qPCR Quality Metrics:

  • Efficiency: 90-110%
  • R² value for standard curves: >0.985
  • Melt curves: Single peaks indicating specific amplification
  • Cq values for NTC: Undetermined or >5 cycles beyond sample Cqs

• Common Issues and Solutions:

  • Low efficiency: Redesign primers or optimize annealing temperature
  • High variability between replicates: Check pipette calibration, mix reagents thoroughly
  • Genomic contamination: Repeat DNase I treatment or use intron-spanning primers
  • RNA degradation: Ensure proper tissue preservation and RNase-free conditions

The integrated application of qPCR and mutant analysis provides a powerful approach for validating the role of specific NBS-LRR genes in plant responses to salt and disease stresses. This technical guide outlines comprehensive methodologies—from experimental design through data interpretation—that enable researchers to establish causal relationships between NBS gene expression and stress tolerance phenotypes. The protocols and reagents detailed here facilitate robust, reproducible experiments that can advance our understanding of plant immunity mechanisms and contribute to the development of stress-resistant crops. As research in this field progresses, the continued refinement of these techniques will further elucidate the complex networks through which NBS-LRR proteins coordinate plant responses to simultaneous environmental challenges.

The nucleotide-binding site (NBS) domain genes encode the largest class of disease resistance (R) proteins in plants, serving as critical intracellular immune receptors that mediate effector-triggered immunity (ETI) [15]. These NBS-LRR (NLR) proteins recognize pathogen-secreted effector molecules and initiate robust immune responses, often accompanied by a hypersensitive response (HR) and programmed cell death at infection sites [11] [111]. The NBS gene family exhibits remarkable diversity across plant lineages, with significant expansions, contractions, and structural variations reflecting co-evolutionary arms races with diverse pathogens [15] [88]. This technical analysis examines the comparative genomics, evolutionary dynamics, and functional characteristics of NBS repertoires across three strategically important plant groups: medicinal plants, legumes, and cereals, providing insights for researchers and drug development professionals investigating plant immunity mechanisms.

Genome-Wide Diversity of NBS Genes Across Plant Lineages

Comparative Quantitative Analysis

Table 1: NBS-LRR Gene Distribution Across Plant Species

Plant Category	Species	Total NBS Genes	CNL	TNL	RNL	Atypical	Reference
Medicinal Plants	Salvia miltiorrhiza	196	61	2	1	132	[11]
Legumes	Glycine max (Soybean)	314	281	33	-	-	[112] [88]
Cereals	Secale cereale (Rye)	582	581	0	1	-	[113]
	Oryza sativa (Rice)	505	505	0	-	-	[11] [15]
	Triticum aestivum (Wheat)	~2012	~2010	0	~2	-	[88]
Other Dicots	Arabidopsis thaliana	150-207	~100	~50	~4	58	[11] [15]
	Capsicum annuum (Pepper)	252	248	4	-	-	[77]
	Solanum tuberosum (Potato)	447	-	-	-	-	[11]

Phylogenetic and Structural Diversification

The NBS gene family is characterized by two major subclasses defined by N-terminal domains: TIR-NBS-LRR (TNL) and CC-NBS-LRR (CNL), with a minor subclass containing RPW8 domains (RNL) [15] [88]. Comparative analysis reveals striking lineage-specific distribution patterns. Monocot cereals, including rye, rice, and wheat, demonstrate a complete absence of TNL genes, with CNLs dominating their NBS repertoires [113] [11] [15]. In contrast, dicot species generally maintain both TNL and CNL subfamilies, though with significant variation in relative proportions.

Medicinal plants like Salvia miltiorrhiza exhibit notable reduction in TNL and RNL subfamilies, with only 2 TNL and 1 RNL member identified among 62 typical NLRs [11]. This pattern extends across the Salvia genus, with comparative analysis of S. bowleyana, S. divinorum, S. hispanica, and S. splendens revealing complete absence of TNL subfamily members and limited RNL copies (1-2), significantly fewer than in other angiosperms like Arabidopsis thaliana and Vitis vinifera [11].

Legumes display intermediate characteristics, with soybean (Glycine max) maintaining both TNL and CNL subfamilies but with CNL predominance (281 CNL vs. 33 TNL) [112] [88]. This distribution reflects both evolutionary history and functional specialization, as TNL and CNL proteins typically activate defense signaling through different downstream pathways [15].

Genomic Organization and Evolutionary Dynamics

Chromosomal Distribution and Gene Clustering

NBS-LRR genes typically display uneven chromosomal distribution and tend to cluster in specific genomic regions, a pattern consistent across plant species. In Secale cereale, chromosome 4 contains the largest number of NBS-LRR genes, differing from patterns observed in barley and wheat B/D genomes but similar to wheat A genome, suggesting shared ancestral inheritance [113]. Pepper (Capsicum annuum) demonstrates variable NBS-LRR density across chromosomes, with chromosome 3 harboring the highest number (38 genes) while chromosomes 2 and 6 contain only 5 genes each [77].

Table 2: Cluster Analysis of NBS-LRR Genes Across Species

Species	Total NBS Genes	Clustered Genes	Percentage Clustered	Number of Clusters	Largest Cluster	Reference
Capsicum annuum	252	136	54%	47	8 genes (Chr3)	[77]
Secale cereale	582	Information not specified	Information not specified	Information not specified	Information not specified	[113]
Glycine max	314	Information not specified	Information not specified	Information not specified	Information not specified	[112]

Cluster analysis in pepper reveals that 54% of NBS-LRR genes (136 genes) form 47 physical clusters, with the largest cluster (8 genes) on chromosome 3 [77]. These clusters often contain genes from the same subfamily, though some exhibit mixed subfamily composition, suggesting functional coordination and complex evolutionary relationships [77].

Evolutionary Mechanisms and Birth-and-Death Dynamics

The expansion and diversification of NBS gene families are driven by several evolutionary mechanisms, including whole-genome duplication (WGD), segmental duplication, and tandem duplication [88]. Tandem duplications particularly contribute to the formation of gene clusters and generate novel resistance specificities through unequal crossing-over and gene conversion events [15].

Phylogenetic analysis of Secale cereale, Hordeum vulgare (barley), and Triticum urartu (diploid wheat) suggests that at least 740 NBS-LRR lineages were present in their common ancestor [113]. However, most have been inherited by only one or two species, with just 65 preserved in all three, indicating extensive lineage-specific gene loss and diversification. The S. cereale genome inherited 382 ancestral NBS-LRR lineages, 120 of which were lost in both barley and T. urartu [113].

This evolutionary pattern follows a "birth-and-death" model, where genes duplicate and then diverge through neutral evolution and natural selection, with some copies maintained while others degenerate or are deleted [15]. Diversifying selection particularly acts on solvent-exposed residues in the LRR domain, consistent with its role in pathogen recognition specificity [15].

Experimental Methodologies for NBS Gene Analysis

Genome-Wide Identification and Annotation

Protocol 1: Identification of NBS-LRR Genes from Genome Sequences

• Data Retrieval: Obtain genome sequences and annotation files from appropriate databases (e.g., NCBI, EnsemblPlants, species-specific databases).
• Domain Search: Perform HMMER search using the NB-ARC domain (Pfam: PF00931) HMM profile against protein sequences with E-value cutoff of 1.0 [113] [88].
• BLAST Refinement: Use obtained sequences as queries for BLASTp search against the proteome with E-value 1.0 [113].
• Domain Validation: Confirm NB-ARC domain presence using HMMscan against Pfam-A database (E-value 0.0001) [113].
• Additional Domain Identification: Scan sequences against NCBI Conserved Domain Database (CDD) to identify CC, TIR (PF01582), RPW8 (PF05659), and LRR domains [113] [11].
• Motif Analysis: Identify conserved motifs using MEME suite with 20 motifs as default parameters [113].
• Classification: Categorize sequences into CNL, TNL, RNL, and atypical subclasses based on domain architecture [11] [88].

CRISPR/Cas9-Mediated Diversification of NBS Genes

Protocol 2: Accelerated Diversification of Tandemly Duplicated NBS Genes

• Target Selection: Design sgRNAs targeting conserved regions within tandemly duplicated NBS gene clusters, preferably in NBS or LRR domains [112].
• Vector Construction: Clone sgRNA cassettes under plant U6 promoters and Cas9 under constitutive promoter (e.g., 35S or DaMV) into binary vector [112].
• Plant Transformation: Transform embryonic tissues using Agrobacterium-mediated transformation [112].
• Variant Screening: Identify rearrangements using:
  • Copy Number Variation: Droplet digital PCR (ddPCR) to detect large-scale rearrangements [112].
  • Paralogue Sequencing: Amplify and sequence target regions to identify chimeric genes with intact open reading frames [112].
• Phenotypic Screening: Assess novel paralogs for new disease resistance specificities using pathogen assays [112].

This approach has successfully generated novel resistance specificities in soybean NBS clusters (Rpp1L and Rps1), with rearrangement frequencies up to 58.8% in progeny of primary transformants [112].

Diagram 1: CRISPR/Cas9-mediated diversification of NBS genes. Targeted chromosome cleavage induces double-strand breaks (DSBs) repaired through various pathways, generating novel chimeric paralogs with potential new resistance specificities [112].

Functional Characterization of NBS Proteins

Protocol 3: Domain Interaction Analysis via Trans-Complementation

• Domain Constructs: Clone separate constructs encoding CC-NBS and LRR domains, or CC and NBS-LRR domains, with appropriate tags (e.g., HA epitope) [31].
• Transient Expression: Co-express domain constructs with pathogen effector in suitable system (Nicotiana benthamiana) [31].
• Hypersensitive Response Assessment: Monitor for cell death indicating functional complementation and immunity activation [31].
• Co-immunoprecipitation: Validate physical interactions between separately expressed domains with and without effector presence [31].
• Mutational Analysis: Test conserved motif mutants (e.g., P-loop) to determine nucleotide dependence of interactions [31].

This approach demonstrated that Rx protein domains can function in trans, with LRR-CC-NBS interaction disrupted by coat protein effector, suggesting sequential conformational changes during activation [31].

NBS Protein Structure and Activation Mechanisms

Domain Architecture and Functional Motifs

NBS-LRR proteins typically contain four distinct domains: variable N-terminal domain (TIR or CC), NBS domain, LRR domain, and variable C-terminal regions [15]. The NBS domain contains several conserved motifs essential for nucleotide binding and hydrolysis:

• P-loop: ATP/GTP binding phosphate binding loop [77]
• RNBS-A, RNBS-B, RNBS-C: Additional conserved motifs distinguishing TNLs and CNLs [15]
• Kinase-2: Involved in nucleotide hydrolysis [77]
• GLPL: C-terminal motif of NBS domain [77]

The LRR domain typically contains 14 repeats on average, with extensive sequence variation generating potential for over 9×10¹¹ variants in Arabidopsis alone, providing exceptional diversity for pathogen recognition [15].

Activation Models and Signaling Mechanisms

Table 3: NBS-LRR Protein Activation Models

Model	Mechanism	Supporting Evidence	Reference
Direct Recognition	NLR directly binds pathogen effector via LRR domain	Pita-AVR-Pita interaction in rice	[11]
Guard Hypothesis	NLR monitors host proteins ("guardees") modified by effectors	Multiple Arabidopsis and solanaceous systems	[15] [77]
Decoy Hypothesis	NLR interacts with host proteins that mimic effector targets but lack function	RIN4 proteins in Arabidopsis	[15]
Integrated Decoy	NLR proteins incorporate domains that mimic effector targets	RGA5 and Pik-1 rice NLRs	[15]

Diagram 2: NBS-LRR protein activation pathways. Effector perception through direct binding or guard mechanisms induces conformational changes, receptor oligomerization, and defense activation [31] [15].

Recent structural studies reveal that plant NLRs oligomerize into resistosomes upon activation, creating channels or signaling platforms that initiate immune responses [9]. For CNL proteins, oligomerization often forms calcium-permeable channels that trigger cell death, while TNL proteins frequently form NADase complexes that produce signaling molecules [9].

The Scientist's Toolkit: Essential Research Reagents and Solutions

Table 4: Key Research Reagents for NBS Gene Analysis

Reagent/Solution	Function/Application	Examples/Specifications	Reference
HMMER Suite	Identification of NBS domains in genomic sequences	Pfam NB-ARC (PF00931) HMM profile	[113] [88]
CRISPR/Cas9 System	Targeted mutagenesis and diversification of NBS clusters	SpCas9 with species-specific sgRNAs, multiple gRNA constructs	[112] [114]
Co-immunoprecipitation Reagents	Protein-protein interaction studies between NBS domains	HA-tagged domain constructs, co-IP with effector proteins	[31]
Transient Expression Systems	Functional assays of NBS protein activation	Nicotiana benthamiana leaf infiltration, pathogen effectors	[31]
OrthoFinder	Evolutionary analysis and orthogroup identification	DIAMOND for sequence similarity, MCL for clustering	[88]
MEME Suite	Conserved motif discovery in NBS domains	Identifies P-loop, RNBS, kinase motifs	[113]
ddPCR/qPCR Reagents	Copy number variation analysis in NBS clusters	Species-specific probes for paralog quantification	[112]

The comparative analysis of NBS repertoires across medicinal plants, legumes, and cereals reveals both conserved features and lineage-specific adaptations in plant immune systems. The complete absence of TNL genes in cereals and their reduction in medicinal plants contrasts with their maintenance in most dicots, suggesting different evolutionary trajectories in immune system architecture. The clustering of NBS genes in genomes and their rapid evolution through duplication and rearrangements provides plants with a versatile toolkit for pathogen recognition.

Emerging technologies, particularly CRISPR/Cas9, now enable targeted diversification of NBS genes to generate novel resistance specificities, potentially accelerating breeding programs [112] [114]. The transfer of functional NLR pairs across taxonomic boundaries demonstrates the potential for engineering broad-spectrum resistance in crop species [9]. Future research directions include comprehensive characterization of NBS genes in underrepresented medicinal species, structural analysis of NLR activation mechanisms, and development of more precise genome editing tools for tailored resistance enhancement.

Understanding the diversity, evolution, and function of NBS genes across plant lineages provides not only fundamental insights into plant-pathogen co-evolution but also practical resources for developing durable disease resistance in crop plants through molecular breeding and biotechnological approaches.

Conclusion

NBS domain genes represent a sophisticated, rapidly evolving immune arsenal that is fundamental to plant survival. Research has transitioned from foundational discovery to a detailed understanding of their genomic architecture, diversified functions, and complex regulation. The integration of advanced computational biology, including deep learning, with robust experimental validation frameworks is dramatically accelerating the pace of discovery. Future research must focus on elucidating the precise signaling mechanisms of different NBS subfamilies, understanding the fitness costs of maintaining large NBS repertoires, and exploring the cross-kingdom parallels between plant and animal innate immunity. For biomedical and clinical research, the principles of pathogen recognition and immune receptor evolution uncovered in plants offer valuable conceptual models. Furthermore, engineering these sentinel genes into crops provides a sustainable, genetic solution to enhance global food security, reducing reliance on chemical pesticides and contributing to healthier ecosystems and human populations.

References

• 1. Recent Advances in Effector-Triggered Immunity in Plants [https://pmc.ncbi.nlm.nih.gov/articles/PMC8124997/]
• 2. Plant Immune System: Understanding Pathogen ... [https://premierscience.com/pjs-24-682/]
• 3. from nucleotide-binding leucine-rich repeat receptor functional ... [https://phytopatholres.biomedcentral.com/articles/10.1186/s42483-025-00387-5]
• 4. Plant Nucleotide Binding Site–Leucine-Rich Repeat (NBS ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC3645687/]
• 5. Comparative analysis, diversification, and functional ... [https://www.nature.com/articles/s41598-024-62876-5]
• 6. Early molecular events in PAMP-triggered immunity [https://www.sciencedirect.com/science/article/pii/S1369526609000648]
• 7. - PTI synergistic signal ETI in mechanisms plant immunity [https://pubmed.ncbi.nlm.nih.gov/38470397/]
• 8. and Pti | PDF eti [https://www.slideshare.net/slideshow/manoj-ppt-4-88472658/88472658]
• 9. The expanding world of plant NLR pairs [https://www.sciencedirect.com/science/article/abs/pii/S136013852500305X]
• 10. Rewriting the code of plant immunity | Nature Biotechnology [https://www.nature.com/articles/s41587-025-02886-4]
• 11. Genome-Wide Characterization and Expression Pattern ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12469828/]
• 12. Plant NBS - LRR in pathogen sensing and host defense - PMC proteins [https://pmc.ncbi.nlm.nih.gov/articles/PMC1973153/]
• 13. NLRs in plant immunity: Structural insights and molecular ... [https://www.sciencedirect.com/science/article/pii/S2772899425000096]
• 14. Functional characterization of NBS-LRR genes reveals an ... [https://bmcbiol.biomedcentral.com/articles/10.1186/s12915-024-01836-x]
• 15. Plant NBS - LRR : adaptable guards | Genome Biology | Full Text proteins [https://genomebiology.biomedcentral.com/articles/10.1186/gb-2006-7-4-212]
• 16. Genome-wide identification and characterization of NBS ... [https://www.nature.com/articles/s41598-025-03507-5]
• 17. Genome-wide identification and characterization of NBS ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12125187/]
• 18. Genome-Wide Identification of NBS-LRR Family in Three ... [https://www.mdpi.com/2073-4425/16/6/680]
• 19. RNL helper NLRs and their importance in plant innate immunity [https://pmc.ncbi.nlm.nih.gov/articles/PMC10118176/]
• 20. Structural basis of NLR activation and innate immune ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8813719/]
• 21. Nucleotide Binding Site - an overview [https://www.sciencedirect.com/topics/biochemistry-genetics-and-molecular-biology/nucleotide-binding-site]
• 22. Structure, biochemical function, and signaling mechanism ... [https://www.sciencedirect.com/science/article/pii/S1674205222004117]
• 23. Genome-Wide Identification and Functional Evolution of NLR ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12562843/]
• 24. Genomics in plant pathogen identification and control [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2025.1661432/full]
• 25. Evolutionary Conservation and Pre-Activated Immunity of ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2025.1656786/full]
• 26. Do duplication-inducing elements 'cooperate' with genes in ... [https://bmcplantbiol.biomedcentral.com/articles/10.1186/s12870-025-07328-6]
• 27. Genome-Wide Identification of NBS-LRR Family in Three ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12192410/]
• 28. Computational Strategies for Elucidating Plant Disease ... [https://www.sciencedirect.com/science/article/abs/pii/S0885576525003790]
• 29. Plant NBS-LRR proteins: adaptable guards - PubMed Central [https://pmc.ncbi.nlm.nih.gov/articles/PMC1557992/]
• 30. Patterns of Positive Selection in the Complete NBS-LRR Gene ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC186657/]
• 31. Interaction between domains of a plant NBS–LRR protein in ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC126192/]
• 32. Identification and distribution of the NBS-LRR gene family in ... [https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-015-1554-9]
• 33. Genome-Wide In Silico Analysis of Leucine-Rich Repeat R ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC11855831/]
• 34. Evolutionary balance between LRR domain loss and young ... [https://bmcgenomics.biomedcentral.com/articles/10.1186/s12864-019-6212-1]
• 35. Multilevel evolution shapes the function of NB-LRR ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2022.1007288/full]
• 36. Tracking ancestral lineages and recent expansions of NBS ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC4991334/]
• 37. Genome-wide characterization of nitric oxide-induced NBS ... [https://bmcplantbiol.biomedcentral.com/articles/10.1186/s12870-024-05587-3]
• 38. Genome-Wide Analysis of NBS-LRR Genes From an Early ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC9140740/]
• 39. Comparative-genomic analysis reveals dynamic NLR gene ... [https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2023.1141194/full]
• 40. An angiosperm NLR Atlas reveals that NLR gene reduction ... [https://www.sciencedirect.com/science/article/pii/S1674205221003142]
• 41. The microRNA miR482 regulates NBS-LRR genes in ... [https://www.sciencedirect.com/science/article/abs/pii/S0168945224001055]
• 42. InterProScan: protein domains identifier - PubMed Central [https://pmc.ncbi.nlm.nih.gov/articles/PMC1160203/]
• 43. Frontiers | PRGminer: harnessing deep learning for the prediction of resistance genes involved in plant defense mechanisms [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2025.1411525/full]
• 44. Research Progress of Plant Nucleotide-Binding Leucine ... [https://www.mdpi.com/2311-7524/9/1/122]
• 45. Adaptation of High-Throughput Screening in Drug Discovery ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC3269696/]
• 46. High Throughput Screening - an overview [https://www.sciencedirect.com/topics/biochemistry-genetics-and-molecular-biology/high-throughput-screening]
• 47. Artificial Intelligence-Assisted Breeding for Plant Disease ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12154274/]
• 48. Tanshinone and salvianolic acid biosynthesis are ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC7016019/]
• 49. Identification and expression profiling analysis of NBS ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC6500516/]
• 50. Plant Immunity Is Regulated by Biological, Genetic, and ... [https://www.mdpi.com/2073-4395/12/11/2790]
• 51. Emerging Roles of Receptor‐Like Proteins in Plant Immunity [https://pmc.ncbi.nlm.nih.gov/articles/PMC12602163/]
• 52. A novel TIR-NBS-LRR gene regulates immune response to ... [https://www.sciencedirect.com/science/article/pii/S2214514122000733]
• 53. Evolution patterns of NBS genes in the genus Dendrobium ... [https://pubmed.ncbi.nlm.nih.gov/36376794/]
• 54. Identification, characterization, and validation of NBS ... [https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2023.1187597/full]
• 55. Genome-wide identification, characterization, and ... [https://www.sciencedirect.com/science/article/abs/pii/S0378111922009374]
• 56. Genome-wide identification, characterization, and ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC6195493/]
• 57. DRPPP: A machine learning based tool for prediction of ... [https://www.sciencedirect.com/science/article/abs/pii/S0010482516302268]
• 58. Genome-wide Identification and Evolutionary Analysis of ... [https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2021.771814/full]
• 59. Identification and Characterization of NBS Resistance ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2021.758559/full]
• 60. prPred: A Predictor to Identify Plant Resistance Proteins by ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC7859348/]
• 61. Phylogenetic, Structural, and Evolutionary Insights into Pepper... [https://pmc.ncbi.nlm.nih.gov/articles/PMC11899730/]
• 62. Unveiling the underlying complexities in breeding for ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2025.1559751/full]
• 63. patterns of Evolution in the genus Dendrobium and... NBS genes [https://bmcplantbiol.biomedcentral.com/articles/10.1186/s12870-022-03904-2]
• 64. Identification and Characterization of NBS Resistance ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8585750/]
• 65. Modelling the effect of gene deployment strategies on ... [https://www.sciencedirect.com/science/article/abs/pii/S0261219416303544]
• 66. Gene pyramiding enhances durable blast disease ... [https://www.nature.com/articles/srep07773]
• 67. Gene Pyramiding for Achieving Enhanced Resistance to ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC7711134/]
• 68. Pyramiding disease resistance genes in elite winter wheat ... [https://www.sciencedirect.com/science/article/pii/S2214514119301230]
• 69. Phylogenetic and evolutionary analysis of NBS-encoding ... [https://www.sciencedirect.com/science/article/abs/pii/S1055790306005112]
• 70. Recent advances in innovative strategies for plant disease ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2025.1586375/full]
• 71. A long road ahead to reliable and complete medicinal plant ... [https://www.nature.com/articles/s41467-025-57448-8]
• 72. Functional Analysis of the NLR Gene YPR1 from Common ... [https://www.mdpi.com/2073-4425/16/11/1321]
• 73. MITE Annotation and Landscape in 207 Plant Genomes ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12550479/]
• 74. Protein remote homology detection and structural ... [https://www.nature.com/articles/s41587-023-01917-2]
• 75. In silico prediction of variant effects: promises and limitations ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12304032/]
• 76. NLR receptors in plant immunity: making sense of the ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10561179/]
• 77. Phylogenetic, Structural, and Evolutionary Insights into ... [https://www.mdpi.com/1422-0067/26/5/1828]
• 78. Plant NLRs with Integrated Domains: Unity Makes Strength [https://pmc.ncbi.nlm.nih.gov/articles/PMC6446777/]
• 79. Comparative analysis of plant immune receptor architectures ... [https://bmcbiol.biomedcentral.com/articles/10.1186/s12915-016-0228-7]
• 80. The TIR-NBS protein TN13 associates with the CC- ... [https://pubmed.ncbi.nlm.nih.gov/33982335/]
• 81. R gene [https://en.wikipedia.org/wiki/R_gene]
• 82. Handling Imbalanced Data: Key Techniques for Better ... [https://medium.com/@dakshrathi/handling-imbalanced-data-key-techniques-for-better-machine-learning-6e33b466f8b7]
• 83. Classification: Accuracy, recall, precision, and related metrics [https://developers.google.com/machine-learning/crash-course/classification/accuracy-precision-recall]
• 84. Class-imbalanced datasets | Machine Learning [https://developers.google.com/machine-learning/crash-course/overfitting/imbalanced-datasets]
• 85. Understanding Precision, Recall, and F1 Score Metrics [https://medium.com/@piyushkashyap045/understanding-precision-recall-and-f1-score-metrics-ea219b908093]
• 86. Genome-Wide Characterization and Expression Pattern ... [https://www.mdpi.com/1422-0067/26/18/9063]
• 87. A compendium of genome-wide sequence reads from NBS ... [https://www.nature.com/articles/s41598-020-67848-z]
• 88. Comparative analysis, diversification, and functional validation of plant ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC11126693/]
• 89. Genome-Wide Identification of Nucleotide-Binding Site– ... [https://www.mdpi.com/2311-7524/10/6/634]
• 90. Prediction of NB‐LRR resistance genes based on full‐length ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC9322396/]
• 91. A telomere-to-telomere genome assembly coupled with ... [https://www.nature.com/articles/s41588-025-02137-x]
• 92. Telomere-to-telomere African wild rice (Oryza longistaminata ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC12360840/]
• 93. Distinct Patterns of Gene Gain and Loss [https://pmc.ncbi.nlm.nih.gov/articles/PMC5427506/]
• 94. Organization and molecular evolution of a disease-resistance ... [https://bmcgenomics.biomedcentral.com/articles/10.1186/1471-2164-12-240]
• 95. Comparative analysis , diversification, and functional validation of plant ... [https://www.nature.com/articles/s41598-024-62876-5?error=cookies_not_supported&code=95da165c-75ff-42d2-9bbb-aa8142196912]
• 96. Frontiers | Comparative analysis of the NLR gene family in the... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2025.1681919/full]
• 97. Reconstructing the Evolutionary History of Nucleotide-Binding Site... [https://juniperpublishers.com/ijesnr/IJESNR.MS.ID.556296.php]
• 98. -wide identification, and Genome expression gene of... analysis [https://link.springer.com/article/10.1007/s12298-024-01475-0]
• 99. Molecular characterisation and genetic mapping of candidate genes ... [https://bmcplantbiol.biomedcentral.com/articles/10.1186/1471-2229-9-62]
• 100. A reannotation–based pipeline for mining missing NLR ... [https://www.sciencedirect.com/science/article/pii/S2662173825002085]
• 101. Discovery of functional NLRs using expression level, high- ... [https://www.nature.com/articles/s41477-025-02110-w]
• 102. Identification, characterization, and validation of NBS ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC10318170/]
• 103. Transcriptome-based identification and validation of ... [https://www.nature.com/articles/s41598-018-38247-2]
• 104. Identification, characterization, and validation of NBS ... [https://pubmed.ncbi.nlm.nih.gov/37408775/]
• 105. Genome-wide analysis of NBS-LRR genes in Rosaceae ... [https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2022.1052191/full]
• 106. Whole genome sequencing reveals transcriptional and ... [https://www.nature.com/articles/s41598-025-18334-x]
• 107. Applications of Virus-Induced Gene Silencing in Cotton [https://pmc.ncbi.nlm.nih.gov/articles/PMC10819639/]
• 108. Genome-wide analysis and functional validation of the ... [https://pubmed.ncbi.nlm.nih.gov/41254446/]
• 109. Identification of key genes involved in the phenotypic alterations of res... [https://bmcplantbiol.biomedcentral.com/articles/10.1186/s12870-018-1436-9]
• 110. Dots in Boxes: Visualization of qPCR Data [https://www.neb.com/en-us/tools-and-resources/video-library/dots-in-boxes-visualization-of-qpcr-data?srsltid=AfmBOopSVR6ZwXkrT5DoczAyiKC882rHEjkggu06DzQoxVjJJC6l07_p]
• 111. Plant disease resistance protein signaling: NBS-LRR ... [https://pubmed.ncbi.nlm.nih.gov/15231261/]
• 112. Novel disease resistance gene paralogs created by ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8184530/]
• 113. Genome-wide Identification and Evolutionary Analysis of NBS ... [https://pmc.ncbi.nlm.nih.gov/articles/PMC8630680/]
• 114. The Enhancement of Plant Disease Resistance Using ... [https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2018.01245/full]