Structural insights into glycan recognition by plant immune receptors: a comparative perspective with animal innate immune systems

SUMMARY

Pattern recognition receptors (PRRs) function across diverse eukaryotic organisms as a powerful surveillance system to perceive danger signals and to trigger specific adaptive responses. This complex receptor network constitutes the first layer of perception of the innate immune system of plants and mammals. PRRs specifically detect conserved “non‐self” microbe‐associated molecular patterns (MAMPs) derived from pathogens, or “self”‐derived damage‐associated molecular patterns (DAMPs) released or synthesized by the host upon tissue damage or infection. MAMP/DAMP recognition by their specific PRRs thereby activate complex immune signaling pathways. In organisms lacking adaptive immunity, that is, those without specialized immune cells, innate immunity plays a central role, and PRRs are often overrepresented in their genomes. In recent years, many carbohydrate‐based MAMPs from microbial outer layers and DAMPs from plant cell walls have been characterized, and their putative plant receptors identified. However, the structural basis of the recognition of these glycans by the extracellular ectodomains (ECDs) of plant PRRs remains poorly characterized, especially when compared with the broader knowledge available for glycan perception by carbohydrate recognition domains (CRDs) of mammalians receptors. In this review, we focus on the crystallized PRR glycan–ligand pairs both in plants and animals and explore whether these binding mechanisms might be conserved across kingdoms. Given the significant potential of glycan‐derived DAMPs/MAMPs as elicitors of disease resistance activation, we emphasize the need for further mechanistic and structural studies to clarify how ECD‐PRRs engage carbohydrate ligands. We also highlight the importance of such structural knowledge to guide the use of PRRs/oligosaccharide pairs as sustainable alternatives for crop protection.

Significance Statement

Understanding how plant pattern recognition receptors (PRRs) bind carbohydrate ligands is essential for elucidating conserved immune mechanisms across eukaryotes. While mammalian glycan perception is structurally well understood, plant PRR–oligosaccharide interactions remain poorly characterized. By integrating structural insights from both systems, this review highlights potentially conserved glycan‐recognition strategies and underscores the potential of glycan‐derived elicitors and PRR counterparts as sustainable tools to strengthen crop immunity and disease resistance.

INTRODUCTION

Eukaryotic organisms are constantly challenged by a wide range of microorganisms and pathogenic agents that exploit host resources to propagate. To counter these threats, eukaryotes have evolved highly specialized immune mechanisms that discriminate between non‐self and self‐molecules to trigger rapid defensive responses and preserve cellular integrity and survival. In mammals, immune defense is composed of two distinct systems: innate and adaptive immunity (Warrick et al., 2025). The latter relies on lymphocytes, which generate highly specific defense molecules, such as antibodies (Ponnachan et al., 2025). In contrast, plants and invertebrates lack specialized immune cells and therefore rely exclusively on innate immunity, a constitutive network of genetically encoded defense mechanisms triggered upon pathogen detection (Pradeu et al., 2024).

In 1989, Charles Janeway introduced the concept of a cell‐intrinsic surveillance system in animals, mediated by pattern recognition receptors (PRRs) (Janeway JR., 1989). These receptors detect conserved microbe‐associated molecular patterns (MAMPs), present in bacteria, fungi, protozoans, viruses, insects, and helminths, and activate immune responses (Medzhitov & Janeway JR., 1997). Among the best‐characterized animal PRRs are toll‐like receptors (TLRs) and nucleotide‐binding oligomerization domain (NOD)‐like receptors (NLRs), which can detect MAMPs extracellularly and intracellularly, respectively (Carpenter & O'neill, 2024). The discovery and characterization of these receptors established a unifying framework for innate immunity across taxa, later reinforced by the identification of similar mechanisms in plants through the characterization of the leucine‐rich repeat (LRR) receptor kinase (RK) flagellin‐sensing 2 (FLS2), the receptor for the flagellin‐derived peptide flg22 in Arabidopsis thaliana (Gomez‐Gomez & Boller, 2000). In addition to PRRs, plants also possess nucleotide‐binding LRR proteins, NLRs (formerly R genes), which are specialized in detecting pathogen‐derived effectors inside the host cell leading to the activation of a second layer of immunity named Effector Triggered Immunity (ETI) (Van Wersch et al., 2020).

Remarkably, organisms lacking adaptive immunity exhibit a broader repertoire of innate immune receptors. For instance, the sea urchin genome contains 222 TLRs and more than 200 NLRs (Hibino et al., 2006). Similarly, Arabidopsis possesses more than 660 genes coding NLRs and over 610 genes encoding plasma membrane‐located receptors, among which more than 60 have been established as PRRs to date, highlighting the potential of many others to also function as PRRs (Del Hierro et al., 2021; Van De Weyer et al., 2019). In contrast, the human genome encodes over 65 canonical PRRs and 23 NLRs (Chen et al., 2025). This expansion would ensure sufficient structural diversity to detect the wide range of potentially harmful agents they encounter. Beyond sensing non‐self MAMPs, PRRs also function as danger sensors by recognizing damage‐associated molecular patterns (DAMPs), host‐derived signals released by tissue damage or synthesized to amplify immune responses (Carpenter & O'neill, 2024; Gust et al., 2017; Tanaka & Heil, 2021). The first animal DAMP identified was the nuclear protein high mobility group box 1 (HMGB1), which, upon release into the extracellular space, acts as a potent pro‐inflammatory mediator (Andersson et al., 2000; Scaffidi et al., 2002). In plants, the first DAMPs identified were oligogalacturonides (OGs), fragments of homogalacturonan from cell wall pectins released upon their breakdown during pathogen attack or tissue damage (Hahn et al., 1981), which elicit robust defense responses and enhance resistance to some pathogens (Degli Esposti et al., 2025).

Box 1. Bullet‐point summary of the main points.

• Glycan molecules, derived from microorganisms and hosts, trigger immune responses in different organisms.

• A plethora of plant extracellular and membrane‐bound receptors might specifically recognize carbohydrate‐based danger signaling molecules

• Crystal structures of PRR extracellular domains bound to their carbohydrate ligands reveal different mechanisms of glycan recognition.

• Structural comparisons highlight both conserved and unique features of glycan perception between plants and mammals.

• Despite occasional shared motifs (e.g., LysM), plants and animals seem to have evolved distinct solutions for glycan perception, reflecting divergent evolutionary pressures.

Recognition of a specific MAMP/DAMP by its counterpart PRR triggers a transient activation of downstream signaling events, leading to pattern‐triggered immunity (PTI). These may include ion fluxes through the plasma membrane, cytoplasmic Ca⁺² burst, mitogen activated protein kinase (MAPK) phosphorylation cascades, production of effector molecules, such as reactive oxygen species (ROS), and the synthesis of antimicrobial peptides and immune‐amplifying molecules, resulting in a complex and efficient cellular defensive response (Carpenter & O'neill, 2024; Couto & Zipfel, 2016; Rzemieniewski & Stegmann, 2022; Snoeck et al., 2025). However, specialized pathogens secrete virulence factors that suppress PRR‐mediated PTI responses. These effectors are detected by intracellular NLRs, triggering ETI (Remick et al., 2023; Yu et al., 2024). This immune signaling leads to the formation of supramolecular complexes, inflammasomes in animals and resistosomes in plants, that amplify immune responses and enhance both PTI and ETI (Coll et al., 2022; Wang et al., 2019). Notably, this second layer of immunity is not activated by symbionts or other non‐pathogenic microbes, allowing the host to fine‐tune its defenses according to the threat (Pradeu et al., 2024; Remick et al., 2023).

During the early stages of host–pathogen interactions, the pathogen's surface, composed of proteins, complex carbohydrates, and/or lipopolysaccharides (LPS), directly contacts the extracellular matrix (ECM) from animal cells or the plant cell wall. At this interface, carbohydrates are major structural components of both host and microbial surfaces, and accordingly many identified DAMPs/MAMPs are of glycan nature. Notably, among the more than 610 genes encoding membrane‐bound receptors in the Arabidopsis genome (Liu et al., 2024), over half encode proteins with extracellular domains (ECDs) that could potentially bind glycans based on their ECD predicted structures (Del Hierro et al., 2021). Similarly, in humans, three of the six types of canonical PRR families described are proposed to recognize glycoconjugate ligands (Chen et al., 2025).

In recent years, a significant progress has been made in the identification of novel carbohydrate‐based DAMPs and MAMPs recognized by the plant immune system (Chaube et al., 2022; Dewangan et al., 2023; Fernández‐Calvo et al., 2024; Khokhani et al., 2021; Mélida et al., 2018; Mélida et al., 2020; Pring et al., 2023; Rebaque et al., 2021; Rebaque et al., 2023; Sun, Xiao, et al., 2025; Wanke et al., 2021). However, much less is known about the molecular mechanisms underlying the specific recognition of these glycans by PRRs, particularly the structural features of their ECDs and the structural basis of ligand binding.

In this review, we specifically focus on glycan recognition by PRRs. We examine the diversity of glycan molecules that act as danger signals and immune activators in plants and animals, with a particular emphasis on the receptors whose structures have been resolved in complex with their ligands. By comparing crystallized PRRs from both kingdoms, we aim to highlight commonalities and differences in the structural determinants that govern glycan binding by these receptors (Table 1, Box 1).

Table 1.

Comparative glycan‐recognition strategies in plant and animal PRRs

	Domain/architecture/motif	Representative plant PRR	Glycan‐based ligand	Binding evidence	Representative animal PRR	Glycan‐based ligand	Binding evidence	Binding features
Canonical glycan‐binding folds	Lysin motif (LysM): small carbohydrate‐binding module found in diverse proteins that specifically recognizes N‐acetylglucosamine‐containing polymers	AtCERK1	(GlcNAc)5	PDB: 4EBZ	LYSMD3	(GlcNAc)5–7, chitin particles and β‐glucan laminarin	ELISA and co‐IP with chitin beads	Extended shallow grooves accommodate consecutive linear GlcNAc. Recognition relies on hydrogen bond networks engaging the N‐acetyl group and ring hydroxyls
		OsCEBiP	(GlcNAc)4	PDB: 5JCE
		LjCERK6	(GlcNAc)5	PDB: 9H3B
		OsCERK1	(GlcNAc)6	PDB: 7VS7
			Mixed‐linked β‐1,3/1,4‐glucans	Microscale thermophoresis (MST)
		AtLYK5	(GlcNAc)8	Isothermal titration calorimetry (ITC)
		ljNFR1	Lipo‐chitooligosaccharides	MST	DLysMD3/4	PGN	in vitro affinity‐binding assays
	C‐type lectin: carbohydrate recognition domain (CRD) that contain EPN or QPD domains and binds sugars in a calcium‐dependent manner	nd	nd	nd	DECTIN‐1	β‐glucans (curdlan)	PDB:2CL8	Ca2+‐mediated coordination of mannose hydroxyl groups and stabilizing hydrogen bond interactions
					DECTIN‐2	α‐mannans	PDB:5VYB
					MINCLE	Trehalose dimycolate	PDB:4KZV
					DC‐SIGN	GlcNAc2Man3	PDB:1K9I
						Manose	PDB1SL5
					Langerin	GlcNAc	PDB:3P5F
					Manose Receptor	methyl GlcNAc	PDB:7 L66
	Galectin‐type lectin: contain a conserved CRD with a β‐sandwich fold that binds β‐galactoside sugars without metal ions	nd	nd	nd	Galectin‐1	LacNAc	PDB:4Q26	Compact pre‐formed pocket that binds N‐acetyl–containing motifs through hydrogen bonds and polar contacts. The rest of the glycan remains solvent exposed
					Galectin‐3	LacNAc	PDB:1KJL
					Galectin‐7	Galβ‐1,3‐GlcNAc	PDB:4XBQ
					Galectin‐9	(LacNAC)3	PDB:2ZHM
	G‐type lectin fold: Also known as Galanthus nivalis agglutinin (GNA)‐type domain that contains a β‐prism I fold	OsLecRK1	Mixed‐linked β‐1,3/1,4‐glucans	MST	nd	nd	nd	nd
Composite architectures for glycan recognition	LRR‐malectin domain: a LRR domain followed by a carbohydrate‐binding domain originally identified in the endoplasmic reticulum that recognizes specific glucose‐containing oligosaccharides	IGP1	CEL3	PDB: 9HHX/9WT6	nd	nd	nd	Pre‐formed pocket accommodates linear β‐1,4‐D‐glucans. Binding is stabilized by aromatic CH–π stacking
			CEL4	PDB: 9WT7
			CEL5	ITC
	Malectin‐like LRR domain: a β‐sandwich malectin domain followed by a LRR domain	MLOP1	Polygalacturonic acid	Dot‐blot and protein binding assays	nd	nd	nd	nd
	Cysteine‐rich Epidermal Growth Factor (EGF) motif: a conserved cysteine‐rich domain characterized by disulfide bonds that stabilize a compact fold	WAK1‐WAK5	Demethylesterified OG10‐15	ELISA	nd	nd	nd	nd
Generalist scaffolds involved in glycan sensing	V‐set Ig domain: immunoglobulin‐like domain that adopts a β‐sandwich fold that mediates molecular recognition	nd	nd	nd	Siglec‐2	Sialic acid in α2‐6 linkage	PDB: 5VKM	Deep pre‐formed pocket with extended shallow glycan‐binding surface. Recognition by aromatic CH–π stacking and H‐bonds
	Leucine‐Rich Repeat (LRR) domain: a protein module composed of tandem leucine‐rich repeats that fold into a curved solenoid structure	ARMs	Rhamnogalacturonan‐I fragments	MST	NOD2	Muramyl dipeptide (MDP)	self‐assembled MDP monolayers	nd
						(PGN)
Non‐classical groove‐mediated recognition	Fibrinogen‐Related Domain (FReD): globular domain structurally related to fibrinogen that mediates ligand binding	nd	nd	nd	FBCD1	(GlcNAc)2	PDB: 6ZQX	Compact cleft binds GlcNAc. Recognition mediated by hydrogen bonds and polar contacts to the N‐acetyl group
					M‐Ficolin	GlcNac or GalNAc	PDB: 2JHK
					L‐Ficolin	(β‐1,3‐D‐Glc)4	PDB: 2J0Y
	PGRP Fold: characterized by α‐helices, β‐strands, and a peptidoglycan‐binding groove, often with a Zn2+‐binding site and associated amidase activity in some variants	nd	nd	nd	PGRP‐IαC	Muramyl pentapeptide	PDB: 2APH	Deep pre‐formed cleft integrating glycan and peptide stems, pocket‐based discrimination

GLYCAN‐BASED SIGNALS IN PLANT IMMUNITY

The plant–microbe interface is a dynamic and complex environment, where polysaccharides from microbial outer layers and plant cell walls are constantly remodeled (Delmer et al., 2024; Munzert & Engelsdorf, 2025; Pinto et al., 2025). Enzymatic degradation by plant‐ or microbe‐derived enzymes of these polysaccharides releases glycan‐based bioactive signals that may modulate immune responses (Pinto et al., 2025). Among pathogen‐derived MAMPs, oligosaccharides from chitin (β‐1,4‐D‐N‐acetylglucosamine, GlcNAc), an abundant polysaccharide in fungal cell walls, insects, and nematodes' exoskeletons, are potent inducers of plant defense responses and can also activate mammalian innate immunity (Cao et al., 2014; Chaudhari et al., 2011; Fuchs et al., 2018; Liu et al., 2016). Partially deacetylated chitin fragments, known as chitosan, similarly trigger robust immune responses and are widely used as plant immune stimulants (Stanley‐Raja et al., 2021). Other microbial glycan‐based MAMPs include bacterial peptidoglycan (PGN) and diverse β‐glucans, such as short, non‐branched β‐1,2‐glucans from bacteria, β‐1,3‐glucans, and branched β‐1,6/β‐1,3‐glucans from fungi and oomycetes, and β‐1,4 or mixed‐linked β‐1,4/β‐1,3‐glucans from oomycetes (Fuertes‐Rabanal et al., 2024; Kelly et al., 2023; Mélida et al., 2018; Rebaque et al., 2021; Wanke et al., 2020; Wanke et al., 2023). Notably, PGN, β‐1,2‐glucans, β‐1,3, and β‐1,6/β‐1,3‐glucans also stimulate innate immunity in animals, illustrating functional conservation of glycan recognition across kingdoms (Chen et al., 2025).

Plants also perceive a wide diversity of endogenous glycan fragments as DAMPs. Recent reviews increasingly recognize plant cell walls, with their remarkable structural diversity, as a major reservoir of DAMPs (De Lorenzo & Cervone, 2022; Molina, Jorda, et al., 2024; Molina, Sanchez‐Vallet, et al., 2024; Pinto et al., 2025; Sun, Xiao, et al., 2025). Examples include cellulose‐derived β‐1,4‐D‐glucose oligomers (cellobiose, CEL2; cellotriose, CEL3; cellotetraose, CEL4) (Aziz et al., 2007; Johnson et al., 2018; Martín‐Dacal et al., 2023; Souza et al., 2017), hemicellulose‐derived oligosaccharides containing xylose, arabinose, fucose and/or mannose (Claverie et al., 2018; Dewangan et al., 2023; Fernández‐Calvo et al., 2024; Mélida et al., 2020; Zang et al., 2019), and pectin‐derived OGs, released from homogalacturonan (α‐1,4 D‐galacturonic acid (GalA); Benedetti et al., 2015, Voxeur et al., 2019). In this context, pectin resembles hyaluronic acid (HA), an acidic polysaccharide of the animal ECM that similarly generates glycan‐derived DAMPs (Termeer et al., 2002). Additional bioactive glycans include structural oligosaccharides from marine algal cell wall polymers (e.g., agar, alginate, fucoidan, carrageenan), and pustulan (β‐1,6‐D‐glucose) from lichens (Aitouguinane et al., 2023; Li et al., 2019; Wang et al., 2023), as well as plant storage oligosaccharides, such as fructans, or maltose‐derived oligosaccharides (Fernández‐Calvo et al., 2024; Versluys et al., 2016).

Interestingly, some immunogenic glycans are present in both plants and microbes and can act as either DAMPs or MAMPs. These include unbranched mixed‐linked glucans (MLGs; β‐1,3/β‐1,4‐D‐glucans), found in grasses, bryophytes, algae, and oomycetes (Fry et al., 2008; Perez‐Mendoza et al., 2015; Pettolino et al., 2009; Rebaque et al., 2021; Rebaque et al., 2023); β‐1,6‐glucans (algae, fungi, lichens) (Wang et al., 2017); and linear β‐1,3‐glucans (plant callose, fungal cell walls) (Chen & Kim, 2009; Klarzynski et al., 2000; Mélida et al., 2018).

However, not all glycan fragments are immunostimulatory, as bacterial cyclic β‐1,2‐glucans and fungal α‐1,3‐glucan may suppress plant immunity (Fujikawa et al., 2012; Rigano et al., 2007). Moreover, short‐chain OGs (e.g., OGs with a degree of polymerization, DP, of DP2‐DP7) may suppress immune signaling, whereas longer fragments enhance plant defense (Moerschbacher et al., 1999, Xiao et al., 2024). The molecular mechanism underlying this striking size‐dependent dual activity of cell wall‐derived polymers remains poorly understood and requires further investigation (see Box 2). Remarkably, plants actively modulate glycan‐based signals to maintain immune homeostasis. Enzymes of the FAD‐binding berberine bridge enzyme‐like (AtBBE‐like) superfamily, including a cellodextrin oxidase (CELLOX) and four oligogalacturonide oxidases (OGOXs), oxidize cellulose and OGs fragments, respectively, inactivating them as cell wall‐derived DAMPs by preventing their recognition by their PRRs, thereby limiting prolonged signaling and hyperimmunity (Benedetti et al., 2018; Degli Esposti et al., 2025; Jiménez‐Sandoval et al., 2025; Locci et al., 2019; Pontiggia et al., 2020). These modifications also highlight a limitation of the PRR surveillance system, since DAMP modifications, such as oxidation of oligosaccharides by microbial enzymes, can reduce immunogenicity, potentially allowing pathogens to evade plant immune detection and dampen DAMP‐triggered immunity (see Box 2; Giovannoni et al., 2025, Sun, Xiao, et al., 2025, Turella et al., 2025).

Box 2. Open questions.

• How can structural analyses of PRR ectodomains help to understand the binding specificities for novel glycan DAMP/MAMPs?

• Are the different glycan‐binding mechanisms described in mammals also present in plants?

• What structural features define the immunogenic activity of a glycan ligand in plants?

• How is the availability of glycan DAMPs/MAMPs regulated in planta during development and pathogen infection?

• To what extent do PRR complexes cooperate or compete in perceiving glycans at the plant cell surface?

• Can structural insights into PRR–glycan interactions be exploited to design novel elicitors for crop protection?

GLYCANS PERCEIVED BY ANIMAL INNATE IMMUNITY

In animal cells, carbohydrates represent a highly diverse class of biomolecules. Their structural complexity, shaped by monosaccharide composition, glycosidic linkages, branching, and modifications, generates distinct arrays on cell surfaces, in the ECM, and within glycoconjugates (Taylor & Drickamer, 2011). Glycosaminoglycans (GAGs), such as HA, heparan sulfate (HS), and chondroitin sulfate (CS), are usually bound to core proteins to generate proteoglycans (e.g., biglycan and decorin), which function as key structural scaffolds in connective tissues, regulating cell adhesion, development and differentiation (Linhardt & Toida, 2004, Taylor & Drickamer, 2011, Diehl et al., 2021). Upon tissue injury or necrosis, fragments of ECM‐derived glycans and proteoglycans (including biglycan, decorin, HS, and low molecular weight HA fragments generated by hyaluronidases) act as potent DAMPs that stimulate innate immunity responses and inflammation (Babelova et al., 2009; Chen et al., 2024; Gray et al., 2022; O'callaghan et al., 2015; Schaefer et al., 2005; Termeer et al., 2002).

Beyond the ECM, carbohydrates decorate membrane glycoproteins and glycolipids (with N‐ and O‐linked glycans) whose branching and terminal modifications (e.g., sialylation, fucosylation) act as molecular signatures recognized by mammalian lectins to regulate communication and immune homeostasis (Pinho et al., 2023). During stress, apoptosis, or infection, these glycan arrays are often cleaved or remodeled exposing unusual carbohydrate motifs. For example, desialylated glycans expose galactose residues recognized by mammal PRRs (Drickamer & Taylor, 2015), and oxidized or truncated glycan motifs on glycoproteins can engage scavenger receptors, triggering clearance and inflammation (Harris et al., 2012). Notably, aberrant β‐1,6‐GlcNAc‐branched N‐glycans on tumor cells are detected by the PRR DENDRITIC CELL (DC)‐ASSOCIATED C‐TYPE LECTIN‐1 (Dectin‐1), triggering immune signaling and cytotoxicity (Chiba et al., 2014). Of note, many immunogenic glycans contain GlcNAc, a monosaccharide widely present in structural polymers, such as chitin, PGN, and GAGs. Upon degradation, GlcNAc‐rich oligomers are generated and efficiently sensed by PRRs, making them recurrent triggers of innate immune responses in plants and animals (Cao et al., 2014; Fuchs et al., 2018; Girardin et al., 2003; He et al., 2021; Termeer et al., 2002; Willmann et al., 2011).

Similar to plants, animal innate immunity also detects many microbial carbohydrate‐based MAMPs, such as PGN, chitin, and mannans from fungal cell walls, mannosylated N‐glycans from viruses, and trehalose dimycolate from mycobacteria (Fuchs et al., 2018; He et al., 2021; Ma et al., 2024; Zhou & Cobb, 2021). A particularly illustrative case refers to fungal β‐glucans, abundant in yeast and filamentous fungal cell walls. Their β‐1,3‐linked backbones with β‐1,6 branches resemble ECM‐derived DAMPs and are sensed by Dectin‐1 and other lectins triggering antifungal immunity (Adachi et al., 2004; Brown et al., 2007). By engaging the same pathways as self‐derived fragments, microbial β‐glucans exemplify how glycan recognition serves as a convergent mechanism to sense perturbations in tissue integrity, regardless of their origin (Varki, 2011).

PRRs INVOLVED IN PERCEPTION MECHANISMS OF IMMUNOGENIC GLYCANS

Plant PRRs belong to the plasma membrane‐bound receptors that constitute one of the largest protein families in eukaryotes, reflecting their central role in plant adaptation (Bender & Zipfel, 2023). Their extracellular domains display a wide structural diversity, that enables perception of peptides, proteins, nucleotides, carbohydrates, and lipids. Therefore, based on the structural features of their ECDs, PRRs are classified into several subgroups, whose evolution, structure and function have been extensively reviewed in recent publications (Li et al., 2024; Liu et al., 2024; Molina, Jorda, et al., 2024; Snoeck et al., 2025). Plant PRRs involved in glycan recognition contain the following types of ECDs: lysin motif (LysM), lectins, cysteine‐rich epidermal growth factor (EGF) motif (present in wall‐associated kinases, WAKs, and WAKs‐like), LRRs, LRR‐malectin and malectin‐like LRR (Table 1; Bellande et al., 2017, Molina, Jorda, et al., 2024, Snoeck et al., 2025). However, evidence for direct binding of glycans by these ECDs has been reported in just a few cases. The LysM receptors are among the best‐characterized PRRs in plants, as they mediate the perception of diverse glycans including, chitin oligomers ((GlcNAc)₆–₈), β‐1,3‐D‐glucans, MLGs (e.g. MLG43), PGN and LPS, in different species, such as Arabidopsis thaliana, Oryza sativa, Vitis vinifera, Medicago truncatula, or Lotus japonicus (Dai et al., 2023; Desaki, Kouzai, et al., 2018; Desaki, Miyata, et al., 2018; Yang et al., 2021). Crystal structures of different LysM‐ECDs bound to chitin oligomers, such as Arabidopsis CHITIN ELICITOR RECEPTOR KINASE 1 (CERK1), OsCERK1, CHITIN ELICITOR BINDING PROTEIN (OsCEBiP), and LjCERK6, have provided high‐resolution insights into polymer binding and receptor activation (Bozsoki et al., 2017; Liu et al., 2016; Liu, Liu, et al., 2012; Xu et al., 2023). In rice, binding of OsCERK1 and the LECTIN RECEPTOR KINASE (OsLecRK1) to mixed‐linked β‐1,3/1,4‐glucans has been demonstrated using microscale thermophoresis (MST) assays (Table 1, Dai et al., 2023, Yang et al., 2021). A second structurally supported group of glycan‐binding PRRs are the recently characterized Arabidopsis family of LRR‐malectin IMPAIRED IN GLYCAN PERCEPTION, (IGP1, IGP3, and IGP4), which is relevant for the perception and activation of immune responses triggered by fragments of cellulose, MLGs, and oligosaccharides containing β‐1,4‐D‐(xylose)_n, like xylotetraose (Xyl4) and 3³‐α‐L‐arabinofuranosyl‐xylotetraose (XA³XX; Tseng et al., 2022, Martín‐Dacal et al., 2023, Fernández‐Calvo et al., 2024). Direct binding of cellulose‐derived oligomers (CEL3 and CEL5), but not for MLG43, by IGP1‐ECD has been demonstrated by isothermal titration calorimetry (ITC) (Martín‐Dacal et al., 2023). Notably, the crystal structure of the ECD‐IGP1 bound to CEL3 has recently revealed the structural basis of this recognition (Jiménez‐Sandoval et al., 2025; Sun, Wei, et al., 2025). Additionally, the LRR‐RKs termed AWARENESS of RAMNOGALACTURONAN‐I MAINTENANCES (ARMs) have been shown to bind rhamnogalacturonan‐I fragments (Lee et al., 2025). WAK family members were long considered prime candidates for OG receptors, because early biochemical studies showed that WAK1‐ECD binds demethylesterified OG_10‐15 in vitro, likely through the EGF‐like domain (Brutus et al., 2010; Decreux et al., 2006; Decreux & Messiaen, 2005). However, a CRISPR‐Cas9 deletion of the entire WAK1‐WAK5 cluster retains full OG_10‐15 responsiveness, indicating that WAKs are not essential for OG perception. To date, no structural evidence supports direct glycan binding by WAKs, while still being strong candidates for modulating glycan‐triggered immunity (Amsbury, 2020; Herold et al., 2024). Recently, a novel receptor, MALECTIN‐LIKE DOMAIN‐CONTAINING LRR‐RLK INVOLVED IN OLIGOGALACTURONIDE PERCEPTION 1, MLOP1, that mediates OG‐induced defense responses has been identified. Biochemical assays have shown that MLOP1 ECD binds polygalacturonic acid in vitro and mutants impaired in this receptor are more susceptible to Pseudomonas syringae (Liu et al., 2025).

In animals, the perception of DAMP/MAMPs by innate immunity systems is mediated by six canonical PRR families: membrane‐bound CLRs and TLRs, and cytosolic NLRs, cyclic GMP‐AMP (cGAMP) synthase (cGAS), absent in melanoma 2 (AIM2)‐like receptors (ALRs) and retinoic acid‐inducible gene i (RIG‐I)‐like receptors (RLRs). The cGAS and AIM2 families are cytoplasmic DNA sensors, while RLRs are primarily involved in sensing viral RNA and RNA replication intermediates (Chen et al., 2025). In contrast, CLRs, TLRs, and NLRs are main families participating in glycan recognition (Table 1). CLRs family includes over 15 members relevant to immunity, specialized in recognition of microbial glycans and DAMPs, with substantial structural data available. For example, Dectin‐1 binds β‐glucans (Brown et al., 2007), Dectin‐2 recognizes high‐mannose glycans and fungal α‐mannans (Feinberg et al., 2021), MACROPHAGE‐INDUCIBLE C‐TYPE LECTIN (Mincle) detects glycolipids like trehalose dimycolate (Ishikawa et al., 2009), and Langerin binds GlcNAc (Feinberg et al., 2011). Although some members of the TLR family are involved in the perception of glycolipids or fungal polysaccharides, there are no crystal structures of ECD TLRs recognizing glycans as primary epitopes. The TLR1–TLR2 structures in complex with a synthetic lipopeptide (PDB 2Z7X) have shown that recognition relies on the lipid chains, rather than by the sugar group (Jin et al., 2007).

The human genome encodes 23 NLRs, including NOD2, which recognize bacterial PGN through its LRR domain (Girardin et al., 2003). Critical residues for binding are known from mutagenesis and docking studies; however, the complete experimental structure of these proteins bound to their ligands is still lacking (Lauro et al., 2017; Tanabe et al., 2004). In addition, several non‐canonical PRR families also contribute to the detection of carbohydrate‐based patterns and play important roles in modulating immune responses (Table 1; Chen et al., 2025). Examples of these non‐canonical PRR families include: (i) sialic acid‐binding immunoglobulin‐type lectins (Siglecs), which detect altered sialylation patterns on stressed or dying cells, acting as checkpoints between tolerance and inflammation (Smith & Bertozzi, 2021); (ii) Galectins that binds β‐galactosides present in microbial glycoproteins and glycolipids (Chan et al., 2018; Liu & Stowell, 2023) and (iii) fibrinogen‐related domain (FReD)‐containing proteins that include ficolins, or the receptor FIBRINOGEN C DOMAIN‐CONTAINING 1 (FIBCD1), that recognize microbial carbohydrates (Garlatti, Belloy, et al., 2007; Garlatti, Martin, et al., 2007; Williams et al., 2024). Notably, recent conceptual frameworks have expanded the definition of PRRs in animal immunity to include inhibitory PRRs (iPRRs), such as certain Siglecs (e.g., Siglec2), which recognize MAMPs and DAMPs but transmitting inhibitory signals to prevent immune overactivation and ensure a balanced response to danger signals (Chen et al., 2025; Rumpret et al., 2022). These findings underscore the importance of proper regulation of immune and inflammatory responses.

Perception of MAMPs and DAMPs by ECD‐PRRs triggers the formation of higher order membrane‐bound protein complexes, including homo and heterodimers and interactions with co‐receptors or other regulatory proteins, leading to the assembly of large supramolecular signaling complexes that in animal innate immunity are referred to as signalosomes (Bryant, 2024; Saijo et al., 2018; Snoeck et al., 2025). In plants, most of the best‐characterized plant ECD‐PRR/ligand pairs correspond to LRR‐peptide systems, such as Arabidopsis FLAGELLIN‐SENSING 2 (FLS2) binding the bacterial peptide flg22 (Sun et al., 2013) and the receptor kinase MALE DISCOVERER 1‐INTERACTING RECEPTOR‐LIKE KINASE 2 (MIK2) binding the SERINE‐RICH ENDOGENOUS PEPTIDES (SCOOPs) (Zhai et al., 2024). Notably, FLS2 and MIK2 have been crystallized with the co‐receptor BRI1‐ASSOCIATED KINASE 1 (BAK1), highlighting co‐receptor relevance in LRR‐RK signaling (Sun et al., 2013; Wu et al., 2024). In contrast, Arabidopsis responses to non‐branched β‐1,3‐glucan oligosaccharides, long‐chain β‐glucans, or chitin occur independently of BAK1 (Mélida et al., 2018; Wanke et al., 2020). The identification of the malectin‐like/LRR‐RK, IMPAIRED OOMYCETE SUSCEPTIBILITY 1 (IOS1) as a novel regulatory protein involved in FLS2, EFR, and CERK1‐mediated signaling revealed that plant PRR signaling requires additional components (Yeh et al., 2016). Therefore, further studies should define the complete composition of these supramolecular signaling complexes at the plasma membrane that modulate synergistic signaling and immunity (see Box 2).

STRUCTURAL BASES FOR GLYCAN RECOGNITION

A major advantage in studying animal PRRs is the abundance of high‐resolution structural data from crystallography, NMR, and cryo‐EM, which provides detailed insights into ligand recognition, domain architecture, and binding specificity. This structural knowledge forms a foundation for understanding how innate immunity interprets glycan signatures across diverse receptor families. Unlike their animal counterparts, plant PRRs involved in glycan sensing remain relatively underexplored at the structural level, as only LysM and LRR‐malectin containing PRRs have been structurally resolved bound to their glycan‐ligands (Bozsoki et al., 2017; Jiménez‐Sandoval et al., 2025; Liu et al., 2016; Liu, Liu, et al., 2012; Sun, Wei, et al., 2025; Xu et al., 2023). The following sections provide a detailed analysis of solved crystal PRR/glycan complexes in animals and plants, focusing on their glycan‐binding strategies and structural mechanisms underlying recognition.

RECOGNITION OF N‐ACETYLATED GLYCANS

GlcNAc‐based ligands such as chitin oligomers, chitooligosaccharides (COs), lipo‐chitooligosaccharides (LCOs), and related N‐acetylated motifs provide one of the clearest illustrations of the structural convergence between plants and animals. In plants, LysM‐type receptors constitute a highly specialized platform for polymer sensing, optimized to interpret the physical architecture of extended β‐1,4‐linked GlcNAc chains. AtCERK1 exemplifies the structural basis of chitin‐derived oligosaccharides perception. The crystal of the ECD bound to (GlcNAc)₅ (PDB: 4EBZ) shows that its ECD contains three LysM subdomains (LysM1‐LysM3; Figure 1a), each of which adopts a conserved βααβ fold (see Glossary), forming a clover‐shaped architecture. All LysM domains have a similar structure with two loops exposed to the solvent. Of these, LysM2 forms the main chitin‐binding site. The binding groove, formed by loops L1 and L2, accommodates GlcNAc oligomers and is characterized by its broad, extended, and hydrophilic nature (Figure 1a, Le et al., 2014). The chitin chain, due to its β‐1,4‐bonds, adopts a fully extended conformation with alternating approximately 180° rotations along its length, allowing it to fit closely into the binding groove of the LysM2 domain. The N‐acetyl groups are critical for binding specificity, forming hydrogen bonds and van der Waals interactions with conserved ECD residues (Figure 1b). Such interactions are not possible with glucose or other non‐acetylated sugars, explaining the groove's selectivity to chitin and the lack of binding to chitosan (de‐acetylated) oligomers by AtCERK1 (Ye et al., 2020). Since both ends of the chitin fragment extend toward the solvent, longer oligomers such as (GlcNAc)₈ promote CERK1 homodimerization, driving immune signaling. Notably, the Ala138His mutation significantly weakens ligand binding and impairs dimerization, highlighting the essential role of ligand binding in dimerization (Liu, Liu, et al., 2012).

Figure 1.

Unique structures deposited for plant PRRs bound to glycoligands.

(a) General structure of the AtCERK1 ectodomain bound to chitin (PDB: 4EBZ). The cloverleaf conformation common to all LysM receptors can be seen. The LysM1 domain is shown in light magenta, LysM2 in aquamarine, and LysM3 in slate blue. The L1 and L2 glycan‐binding loops are colored green within LysM2.

(b) AtCERK1 binding site for chitin oligomer (PDB: 4EBZ). The carbonyl oxygen of the N‐acetyl group in the fourth GlcNAc monomer forms a hydrogen bond with the nitrogen of the main chain of Glu114, while its methyl group interacts with the side chains of Met127, Gln131, and Ala138. The second GlcNAc inserts its N‐acetyl group into a different pocket, forming additional specific hydrogen bonds with Glu110 and Ile141, while the third and fourth GlcNAcs are stabilized by nearby residues such as Asn140 and Ile141. It should be noted that the bond was with (GlcNAc)₅, although only four of the five monomers could be resolved.

(c) OsCERK1 binding site for chitin oligomer (PDB: 7VS7). Hydrogen bonds involving Gly116, Thr120, Ala143, Asn145, and Ile146 stabilize the ligand. In addition, the side chains of Arg115, Asn145, and Ile146 provide van der Waals interactions that reinforce the ligand–receptor association.

(d) OsCEBiP binding site for chitin oligomer (PDB: 5JCE). It should be noted that the bond was with (GlcNAc)₄, although only three of the four monomers could be resolved. Hydrogen bonds between Asn142 and Thr155 and certain hydroxyl groups of the ligand can be observed.

(e) LjCERK6/LYS6 binding site for chitin oligomer (PDB: 9H3B). The hydrogen bonds involving the main chains of Gly44, Ser45, Leu47, Val79, and Ala81 are particularly significant, along with the hydrogen bond formed by the side chain of Asp77. In all cases, the protein surface is shown in green and the glycoligands as light gray sticks. Interacting residues are represented in cyan, with their side chains as sticks and their main chains as lines. Hydrogen bonds are shown as yellow dashed lines. Molecular structures were prepared using PyMOL 3.1.6 (Schrödinger, LLC), while chemical schemes and compound structures were drawn with ChemDraw Professional 12.0 (Revvity Signals Software, Inc.).

CERK1 forms a pre‐assembled complex with the LysM RECEPTOR‐LIKE KINASE 5 (LYK5), crucial for high‐affinity chitin perception in Arabidopsis, with LYK5 displaying nanomolar binding affinity for chitin oligomers (Cao et al., 2014). Although no experimental structure is available, modeling and mutational analyses suggest in LYK5 a conserved LysM‐based binding interface similar to CERK1. LYK5's high affinity complements CERK1's signaling capacity, forming a functional receptor complex. Additionally, LYK5 interacts with the pseudokinase LYK4, which homodimerizes and heterodimerizes with LYK5 even in the absence of chitin. Upon elicitation, they form a CERK1‐LYK5‐LYK4 tripartite LysM complex (Xue et al., 2019). This mode of recognition of chitin‐derived MAMPs reflects a broader principle in plant glycan sensing: LysM receptors engage extended surfaces along the polymer, so the overall shape, length, and rigidity of the chain directly influence receptor assembly and activation. In plants, long GlcNAc polymers can bridge two LysM receptors simultaneously, meaning that the ligand itself provides the multivalency required for receptor dimerization. This topological recognition strategy appears to be conserved across angiosperms.

In rice, chitin recognition relies on a heteromeric complex formed by the LysM receptor‐like protein OsCEBiP and the LysM receptor kinase OsCERK1, both of which employ a similar strategy for ligand binding through their ECDs. Crystal structures of their ECDs (in the apo‐form and bound to GlcNAc oligomers) have been resolved. OsCERK1 (PDB: 7VS7) interacts with chitin oligomers, specifically (GlcNAc)₄ and (GlcNAc)₆, through hydrogen bonds involving backbone and side chain atoms, complemented by van der Waals interactions (Figure 1c, Xu et al., 2023). Similarly, OsCEBiP (PDB: 5JCE) contains three LysM domains and binds chitin primarily through its LysM2 domain, forming interactions that stabilize the oligomer in a rigid conformation (Figure 1d). Ligand binding promotes OsCEBiP homodimerization and recruits OsCERK1, whose kinase activity triggers downstream immune signaling (Liu et al., 2016). Although structural data are still lacking for many species, biochemical and mutational analyses suggest that chitooligosaccharides (COs) binding follows a conserved Lys‐mediated recognition mode. Wheat TaCEBiP and TaLYK5 exhibit high‐affinity binding to (GlcNAc)₆, whereas TaCERK1 contributes primarily to signaling despite weak affinity (Liu et al., 2023). Grapevine VvLYK5‐1 functions as a high‐affinity chitin receptor that recruits VvLYK1‐1 upon ligand perception, mirroring the CERK1‐LYK5 activation paradigm (Roudaire et al., 2023). Across these systems, plants repeatedly deploy heteromeric LysM assemblies in which one subunit optimizes glycan engagement, and another provides kinase activity, reinforcing the principle that ligand structure orchestrates receptor complex formation.

Legumes further illustrate the evolutionary flexibility of the LysM scaffold. Receptors such as LysM RECEPTOR KINASE 6, LjLYS6, whose holo‐form structure (ECD bound to (GlcNAc)₅) has recently been resolved (PDB: 9H3B, Figure 1e), and MtCERK1 retain the canonical chitin‐binding residues and mediate immune responses (Bozsoki et al., 2017; Feng et al., 2019). In contrast, paralogs such as NOD FACTOR RECEPTORs, NFR1/NFR5 in Lotus and NOD FACTOR PRECEPTION, NFP/LYK3 in Medicago have undergone structural modifications that enable them to recognize symbiotic rhizobia‐derived LCOs. Domain‐swap and mutational analyses show that ligand class specificity maps to discrete surfaces within LysM1, where single amino acid substitutions can switch recognition from immune COs to symbiotic LCOs (Bouchiba et al., 2022; Bozsoki et al., 2020). LysM2 may contribute to kinetic stabilization by engaging the acyl chain of LCOs, while LysM3 fine‐tunes affinity through hydrophobic groove interactions (Gysel et al., 2021). These findings reveal how plants have repurposed a polymer‐sensing fold to discriminate between structurally related glycans with distinct biological outcomes, maintaining parallel pathways for immunity and symbiosis. All these data across species indicate that LysM receptors operate as polymer‐responsive modules whose activation depends on how the ligand's topology engages and organizes receptor complexes.

The recent identification of the LysM receptor LYSIN MOTIF DOMAIN 3 (LysMD3) in human epithelial cells suggests that the well established mechanism of chitin perception found in plants might operate in animal immunity too. Enzyme‐Linked Immunosorbent Assays (ELISAs) have shown that the LYSMD3 ECD binds chitin oligomers; however, there is still no evidence of homo‐ or heterodimer formation (He et al., 2021).

Interestingly, animals possess other GlcNAc recognition mechanisms that follow a fundamentally different architectural logic. The clearest example is FIBCD1, a membrane‐anchored receptor that contributes to immune modulation in the intestinal and respiratory epithelia (Schlosser et al., 2009). The crystal structure bound to GlcNAc (PDB 6ZQR) and (GlcNAc)₂ (PDB 6ZQX) shows that its ectodomain forms a homotetramer of FReDs, each containing a conserved S1 pocket that binds N‐acetylated sugars through a combination of hydrogen bonding and hydrophobic interactions. Aromatic residues in the ectodomain create a hydrophobic environment that accommodates the acetyl group of GlcNAc, while a cis‐peptide bond and an adjacent calcium‐binding site stabilize the binding geometry (Figure 2a), a configuration shared with other FReD‐containing proteins (Shrive et al., 2014; Williams et al., 2024). Notably, FIBCD1 has an adjacent S1 (2) pocket that, together with S1, allows accommodation of acetylated and sulfated glycans, including chitin oligomers and GAGs (Williams et al., 2024).

Figure 2.

Examples of deposited structures of animal PRRs bound to carbohydrate‐based ligands.

(a) FIBCD1 binds the ligand N‐acetyl group in the S1 site through hydrogen bonds with Cys414, His415, and Tyr431, with the ligand further stabilized within a hydrophobic aromatic pocket (PDB 6ZQR).

(b) L‐ficolin binding site for β‐1,3‐D‐glucan oligomer (PDB 2J0Y). Panel shows the interaction at the S subsite, where β‐1,3‐glucan is bound via CH–π interaction and by H‐bonding with Asp133, Thr136, and Gln223.

(c) PGRP‐IαC binding site for muramyl pentapeptide (MpP) (PDB 2APH). PGRP‐IαC engages the peptide ligand tightly within a pocket, forming hydrogen bonds with Thr209, Tyr242, Arg235, Tyr266, Asn269, and Gly267, contributing to specific recognition of Lys‐type PGNs.

(d) DC‐SIGN bound to GlcNAc₂Man₃ (PDB 1K9I). Subtle substitutions in the binding pocket, such as replacing a hydrophobic residue with a polar one, can dramatically alter specificity. Recognition involves hydrogen bonds with Ser308, Glu354, Ser360, Gly361, and Asp367, while Phe313 and Val351 contribute van der Waals interactions. The different types of intermolecular interaction are shown as dashed lines: yellow for hydrogen bonds and green for aromatic stacking.

Other soluble FReD‐containing proteins, in contrast, target chemically related motifs such as N‐acetylated sugars, but do not act as dedicated chitin receptors. For example, human M‐ficolin (Ficolin‐1) forms a trimer in which each monomer contributes to a central S1 site that binds N‐acetylated sugars such as GlcNAc and N‐acetylgalactosamine (GalNAc, Garlatti, Martin, et al., 2007). This pocket is relatively small, relying on a hydrophobic and aromatic‐rich microenvironment to accommodate the acetyl group while hydrogen bonds stabilize the sugar carbonyl. Its limited depth and local flexibility confer broad ligand promiscuity, allowing recognition of larger or structurally diverse glycans, including sialylated motifs.

By contrast, the CLR MACROPHAGE GALACTOSE LECTIN (MGL) employs a different binding architecture. Its carbohydrate recognition domain contains a Ca²⁺‐dependent binding pocket shaped by the conserved motif QPD, which imposes strong geometric and stereochemical constraints that confer high specificity toward GalNAc glycan residues. MGL aligns residues to engage C4‐oriented hydroxyls and uses aromatic side chains to shape a complementary pocket, while a histidine stabilizes GalNAc acetamido recognition, enabling binding to tumor‐associated glycans (Gabba et al., 2021). In a similar manner, the MANNOSE RECEPTOR CD206 (MR) engages GlcNAc through its C‐type lectin domain, recognizing N‐acetylated glycans via multivalent interactions (Feinberg et al., 2021). Last, galectins represent another major mode of GlcNAc‐containing glycan recognition, employing a β‐sandwich CRD architecture (see Glossary) and diverse oligomerization strategies. Some members of this family are well studied structurally, as multiple crystals both apo and ligand‐bound have been obtained. Galectin‐1, a prototypical β‐galactoside‐binding lectin, engages disaccharides such as N‐acetylated lactose (LacNAc) using a combination of hydrogen bonds and aromatic stacking (see Glossary; Bianchet et al., 2000). Chimera‐type galectin‐3 combines a canonical CRD with a disordered N‐terminal tail that drives oligomerization; its open pocket, centered on aromatic stacking with galactose and hydrogen bonding with GlcNAc, accommodates internal LacNAc units and even sialylated ligands (Seetharaman et al., 1998, Bum‐Erdene et al., 2013). In this context, the term “aromatic stacking” refers specifically to CH–π interactions, which are the only aromatic‐type contacts possible for carbohydrates because monosaccharides lack aromatic rings. Galectin‐7 fine‐tunes specificity through structural adjustments: shortening of loop L4 and the formation of a Glu‐centered salt bridge triad reduce water coordination within the binding site, shifting preference toward Galβ1‐3GlcNAc (Hsieh et al., 2015; Liang et al., 2022). Galectin‐9 adds further asymmetry, with its N‐terminal CRD preferring linear poly‐LacNAc chains and its C‐terminal CRD recognizing branched or sialylated motifs through a flexible arginine‐mediated mechanism (Nagae et al., 2008; Yoshida et al., 2010).

Across these systems, it seems that LysM receptors are specialized to bind long GlcNAc oligomers and depend on the ligand's physical architecture to organize receptor complexes (based mainly on plant PRRs data). While animal CRLs, galectins, and FReD‐containing proteins rely on compact, chemically defined pockets that typically interact with one or two sugar units at a time. Because their ligands are short, multivalency is generated by receptor oligomerization (tetramers in FIBCD1, trimers in ficolins, dimers or higher order assemblies in galectins) rather than by ligand extension, reflecting a genuine difference in recognition logic in comparison with LysM receptors.

RECOGNITION OF Β‐GLUCANS

In contrast to chitin recognition that relies on the rigidity and linearity of β‐1,4‐linked GlcNAc chains, glucan receptors must discriminate among chemically identical glucose units arranged in linkages that vary subtly in orientation, curvature, flexibility, and branching. This makes glucan recognition by receptors a problem of ligands topology. The recently described cellulose‐derived oligosaccharide LRR‐malectin IGP1 receptor exemplifies this recognition strategy (Jiménez‐Sandoval et al., 2025; Sun, Wei, et al., 2025). The crystal structure of IGP1 bound to CEL3 revealed that this DAMP is accommodated within the LRR domain of the ECD. IGP1/CORK1 binds cello‐oligomers such as CEL3 (PDB 9HHX/9WT6) and CEL4 (PDB 9WT7) through a combination of aromatic CH‐π interactions (see Glossary) and hydrogen bonds, anchoring specific glucose units within a pocket that is essentially pre‐formed in the apo state (Jiménez‐Sandoval et al., 2025, Sun, Wei, et al., 2025). The malectin‐like domain of IGP1‐ECD primarily provides a rigid structural scaffold and limited polar contacts, whereas its LRR domain contributes to the aromatic clamp (see Glossary), formed by Trp 146 and Tyr 196, and the key polar interactions that engage and stabilize the glucan backbone and accommodate the oligosaccharide without requiring induced fit. The minimal structural rearrangement upon ligand binding indicates that the receptor is prepared for rapid detection of cellulose‐derived fragments, mirroring the principle seen in LysM receptors: specificity is embedded in the static architecture of the ectodomain, and the polymer's geometry dictates receptor engagement. However, the binding mechanism is notably different from that of LysM receptors. While chitin binding can be bidirectional in the binding groove, CEL3 entry into the IGP1 binding pocket is exclusively unidirectional, allowing only the binding of intact anomeric carbons (Jiménez‐Sandoval et al., 2025; Sun, Wei, et al., 2025). In fact, oxidized CEL3 does not bind to the IGP1 pocket, demonstrating that DAMPs modifications hamper their eliciting activity. Additionally, the malectin domain contains two conserved cysteines that have been proposed to function as redox sensors modulating ligand binding affinity (Sun, Wei, et al., 2025). However, mutation of one of these cysteines to alanine (C592A) has no effect on the binding affinity of CEL3 to the IGP1 ECD; therefore, further studies are required to clarify the role of these residues (Jiménez‐Sandoval et al., 2025). Importantly, IGP1 provides the first structural insight into plant glycan receptors outside the LysM family, expanding the understanding of how plants perceive chemically simple but topologically diverse oligosaccharides.

MLGs ligands recognized by plants introduce further complexity, as their β‐1,3/1,4 patterns generate local curvature and flexibility distinct from linear cello‐oligomers. In rice, OsCERK1 contributes to the perception of MLG43 (Yang et al., 2021) through an unknown mechanism. In parallel, OsLecRK1, a G‐type lectin receptor kinase, has also been described to bind multiple MLGs but shows no detectable affinity for CEL4, demonstrating a finely tuned specificity that distinguishes between subtle variations in linkage pattern (Dai et al., 2023). The absence of binding in the paralog OsLecRK2 highlights how small differences in lectin‐domain architecture can dramatically alter ligand selectivity. Given that OsCERK1 and OsCEBiP also participate in MLG perception in rice, OsLecRK1 may form heteromeric complexes with these receptors, suggesting that plants might combine distinct extracellular modules to interpret glucan shape. However, structural characterization of LecRK ectodomains and OsCER1/OsCEBiP bound to MLGs will be essential to clarify the molecular basis of their specificity.

Plants also perceive a broader spectrum of β‐glucans through receptors that are only partially characterized. Linear β‐1,2‐D‐glucans appear to be recognized by an as‐yet‐unidentified PRR, as functional studies suggest the existence of a dedicated sensing mechanism (Fuertes‐Rabanal et al., 2024). Linear β‐1,3‐D‐glucans, or laminarins, engage CERK1, consistent with its broader role in β‐glucan perception, Finally, laminarins and β‐1,6‐D‐glucans also trigger CERK1‐dependent responses, indicating that this receptor participate in the perception of diverse glucans beyond its canonical ligands (Chaube et al., 2022; Mélida et al., 2018; Wanke et al., 2020). Interestingly, human LysMD3 ECD, involved in chitin perception, binds also linear β‐1,3 and branched β‐1,3/ β‐1,6‐glucans (He et al., 2021). Although structural data for these interactions remain limited, these observations suggest that plants and animals employ modular receptor scaffolds that may interpret subtle variations in β‐glucan linkage and branching patterns.

Other animal glucan receptors employ different architectures but converge on similar chemical principles for glucan‐based ligand recognition. The CLR Dectin‐1 recognizes β‐1,3/1,6‐glucans through hydrogen bonding and CH–π interactions mediated by aromatic residues, similarly to IGP1‐CEL3 binding, but uses flexible loops that adapt to the curvature of the ligand (Brown et al., 2007; Kalia et al., 2021; Malamud & Brown, 2024). Interestingly, many animal PRRs, including L‐ficolin and Mincle, employ calcium ions to stabilize the binding geometry and properly orient the glucan within the pockets. L‐ficolin (ficolin‐2) binds β‐1,3‐glucans via multiple discrete sites, stabilizing the ligand through CH–π interactions through a Trp residue, hydrogen bonds with Asp133, Thr136, and Gln223, and polar contacts (PDB 2J0Y) (Figure 2b) (Garlatti, Belloy, et al., 2007; Laffly et al., 2014). Mincle binds trehalose‐based glycolipids by combining sugar‐headgroup recognition with a hydrophobic groove that accommodates lipid tails (Feinberg et al., 2016; Ishizuka et al., 2023). With the exception of Dectin‐1, these additional receptors rely on flexible loops and induced‐fit mechanisms that allow them to accommodate short glucan motifs, reflecting a recognition strategy centered on chemically defined pockets rather than polymer topology. Dectin‐1 is distinct not because it lacks flexible loops; it does use them to adapt to β‐1,3/β‐1,6‐glucan curvature, but because it can undergo ligand‐dependent oligomerization when engaging long β‐glucan chains, a behavior reminiscent of polymer‐driven clustering described for plant LysM receptors such as AtCERK1. This ligand‐length–dependent assembly has not been reported for the other animal receptors discussed here. Thus, while plant receptors described to date tend to interpret glucan topology through pre‐organized surfaces and modular ectodomains, animal receptors more commonly rely on flexible loops and pocket‐based architectures that accommodate short glucan fragments through induced fit. Both systems exploit aromatic stacking and hydrogen bonding, as well as LysM‐based mechanisms.

RECOGNITION OF PEPTIDOGLYCANS

Both kingdoms possess receptors that detect the conserved disaccharide backbone, GlcNAc‐MurNAc (N‐acetylmuramic acid). In plants, PGN perception is mainly mediated by LysM‐containing receptors that function as cooperative, ligand‐responsive assemblies. In Arabidopsis, the tripartite system composed of LYSIN MOTIF 1 (LYM1), LYM3, and CERK1 illustrates this principle with exceptional clarity. LYM1 and LYM3 are GPI‐anchored proteins that directly bind insoluble PGN fragments, whereas CERK1 lacks direct ligand‐binding capacity but is indispensable for signal transduction (Willmann et al., 2011). Biochemical evidence indicates that specificity resides primarily in the recognition of the glycan backbone rather than the peptide stems, as modifications to the peptide moieties do not alter binding or signaling, and free peptides do not compete for receptor engagement. This suggests that plant LysM receptors detect PGN through conserved sugar motifs that remain invariant across bacterial taxa, although the structural basis for this selectivity remains unresolved (Mesnage et al., 2014). Rice presents a more versatile PGN perception system, reflecting an evolutionary trajectory toward multiligand detection. LysM‐CONTAINING PROTEIN receptors LYP4 and LYP6 exhibit dual specificity, binding both bacterial PGN and fungal chitin (Liu, Li, et al., 2012). Biochemical assays show that these receptors recognize PGN oligomers with high specificity, and cross‐desensitization experiments reveal that PGN and chitin responses compete, indicating that both ligands engage overlapping receptor complexes. Structurally, LYP4 and LYP6 exist as pre‐associated heterooligomers at the plasma membrane, which disassemble upon ligand binding to recruit the receptor‐like kinase OsCERK1. Although OsCERK1 does not directly bind PGN, its extracellular LysM domains facilitate receptor heteromerization, while its intracellular kinase domain initiates MAPK cascades (Ao et al., 2014). This dynamic assembly mechanism couples ligand detection to receptor reorganization, mirroring the modular, topology‐driven logic seen in chitin and glucan perception.

In animals, a similar mechanism based on LysM perception has also been described, as the orthologs of human LysMD3 in Drosophila LysMD3/4 can bind insoluble PGN by in vitro affinity‐binding assays, but not chitin (Snee et al., 2023). However, animals possess other PGN recognition strategies mediated by the NLR NOD2 and peptidoglycan recognition proteins (PGRPs) (Girardin et al., 2003; Guan et al., 2004). NOD2 is a cytoplasmic LRR receptor that binds muramyl dipeptide (MDP; N‐acetylmuramyl‐L‐alanyl‐D‐isoglutamine). Recognition is mainly mediated by peptide interaction; however, the MurNAc moiety is indispensable for high‐affinity recognition as it provides additional hydrogen bonds and Van der Waals interactions (Girardin et al., 2003). Additionally, human PGRPs share a conserved fold built around a central β‐sheet flanked by α‐helices and stabilized by disulfide bonds and an N‐terminal segment, forming a deep, pre‐organized cleft that accommodates both the glycan backbone and the peptide stems (Guan et al., 2004). This pocket integrates multiple structural features of PGN within a single binding site, allowing the receptor to discriminate among chemotypes based on the geometry of the disaccharide and the orientation of the peptide moieties. Structural studies of different PGRPs reveal variations on a common architectural framework. PGRP‐S binds muramyl tripeptides and glycan fragments but lacks the catalytic zinc ion (Figure 2c) and therefore functions as a non‐enzymatic sensor (Guan et al., 2004). Both catalytic and non‐catalytic PGRPs, including PGRP‐IαC and PGRP‐IβC, accommodate the disaccharide core and the peptide stem within the same binding cleft. PGRP‐IαC shows strong specificity for Lys‐type peptidoglycan (Guan et al., 2006), whereas PGRP‐IβC induces subtle conformational changes in the MurNAc‐linked D‐lactyl group, a distortion that may interfere with bacterial cross‐linking (Cho et al., 2007). Despite these functional differences, all PGRPs use a pre‐formed, rigid binding cleft rather than relying on ligand‐induced reorganization of multimeric receptor complexes.

Together, these systems illustrate two distinct evolutionary solutions to the same biochemical problem, perceiving PGN. LysM modular receptor networks in which ligand binding reorganizes multimeric complexes and autonomous receptors, such as PGRPs with deep binding clefts that integrate glycan and peptide information within a single fold. Both strategies detect conserved PGN motifs with high specificity, but the architectural logic underlying recognition and signaling is fundamentally different.

RECOGNITION OF OTHER CARBOHYDRATE‐BASED DAMP AND MAMPs

Structural data insights from the rhizobial exopolysaccharide (EPS) receptor LjEPR3 in legumes further illustrate the adaptability of plant glycan‐binding modules. Although no ligand‐bound structure is available, the apo‐form ECD structure of this LysM receptor reveals how a LysM‐related fold has adapted to accommodate exclusively the branched architectures of EPS (Wong et al., 2020). The LjEPR3 ECD is composed of three modules: an N‐terminal LysM1 with a novel βαββ fold, an atypical βαβ LysM2 (lacking part of the canonical LysM α‐helix) that has not previously been reported in carbohydrate‐binding proteins, and a conventional βααβ LysM3. This unusual surface does not engage ligands via the AtCERK1‐style N‐acetyl interactions, and accordingly, EPR3 does not bind chitin or LCOs. Instead, it accommodates chemically diverse bacterial EPS with affinities in the micromolar range, independent of O‐acetylation (Wong et al., 2020). These innovations highlight how plant receptors repurpose conserved folds to detect structurally distinct ligands. Interestingly, the composition of EPS resembles animals GAG, as both contain many acidic, amino, and sulfated sugars.

To detect fragments released from the ECM, animals rely on compact, chemically sculpted pockets rather than broad surfaces. This logic is exemplified by C‐type lectin receptors, whose Ca²⁺‐dependent CRDs use EPN or QPD motifs and loop plasticity to tune specificity. DC‐SIGN and DC‐SIGNR bind high‐mannose (Figure 2d) and fucosylated glycans through Ca²⁺‐coordinated hydrogen bonds and aromatic contacts (Feinberg et al., 2001; Guo et al., 2004), whereas Langerin adapts the same fold to recognize sulfated ligands via a lysine‐mediated salt bridge (Chatwell et al., 2008; Feinberg et al., 2013; Hanske et al., 2017). Other lectins such as Dectin‐2 and DCIR stabilize internal glycan residues through secondary sites that reshape loop conformations, illustrating how subtle architectural variations within the CRD scaffold diversify ligand recognition (Feinberg et al., 2017; Nagae et al., 2016). Finally, the MANNOSE RECEPTOR CD206 (MR) binds mannose via a Ca²⁺‐dependent pre‐formed binding site stabilized by extensive hydrogen bonding (Feinberg et al., 2021).

Siglecs provide yet another variation on pocket‐based recognition, generating a canonical sialic acid‐binding pocket. Their V‐set Ig domains (see Glossary) bind sialylated glycans through a conserved arginine‐carboxylate interaction, while surrounding loops define linkage and motif specificity (Smith & Bertozzi, 2021). Rigid architectures such as Siglec‐2 enforce strict α2‐6 linkage preference (Ereno‐Orbea et al., 2017; Ereno‐Orbea et al., 2021), whereas flexible loops in Siglec‐7 expand the pocket to accommodate bulkier α2‐8 disialylated structures (Attrill et al., 2006). Siglec‐8 achieves high affinity through a compact aromatic pocket that enhances CH–π and hydrophobic interactions (Lenza et al., 2023).

The structural basis of the recognition of other plant cell wall‐derived DAMPs, such as OGs, xyloglucan‐derived fragments (e.g., XA³XX) and mannan‐derived oligosaccharide (Chen et al., 2024; Molina, Sanchez‐Vallet, et al., 2024; Ranf, 2017) by plant receptors, is not well characterized, though several type of receptors have been involved in the PTI responses mediated by these DAMPs. The IGP family has emerged as a core group of receptors required also for the perception of β‐1,4‐xylotetraose (XYL4) and XA³XX. Cross‐elicitation experiments show that XYL4 and XA³XX share recognition features with CEL3 but differ from chitin, and their perception is largely independent of the CERK1‐LYK4‐LYK5 module (Fernández‐Calvo et al., 2024). Additionally, xyloglucans such as heptamaloxyloglucan, and mannans (β‐1,4‐D‐(Man)) also trigger PTI via an as‐yet‐unidentified PRRs, with functional assays supporting receptor engagement despite a lack of structural information (Claverie et al., 2018).

CONVERGENT AND DIVERGENT MECHANISMS OF GLYCAN PERCEPTION IN PLANT AND MAMMALS: FUTURE DIRECTIONS

Glycan perception is a fundamental aspect of innate immunity across eukaryotes. Despite the evolutionary distance between plants and mammals, structural studies have revealed a conservation in some of the strategies used to recognize glycans. In both kingdoms, glycan binding relies on similar types of ectodomains and interactions, including hydrogen bonds and van der Waals contacts. Several binding pockets display similar chemical and spatial features, particularly the presence of aromatic residues that stabilize sugar interactions through π‐stacking or CH–π‐stacking (see Glossary) with the glucan rings. This functional convergence can be observed between the LRR‐malectin ECD of IGP1 and animal families such as CLRs or ficolins, which, despite lacking direct homology, share some similarities in the architecture of their binding site that seems to confer specificity and affinity toward distinct carbohydrate‐based ligands.

However, with the notable exception of LysM domains, most glycan‐binding motifs used by innate immune receptors in plants and animals are distinct, reflecting independent and divergent evolutionary solutions to sugar perception (Table 1). In addition, plant receptors uniquely integrate their extracellular carbohydrate‐binding domains with intracellular kinase domains, whereas animal glycan‐binding proteins typically lack such receptor‐kinase architecture, leading to different signaling mechanisms. Moreover, the distribution of receptor families in eukaryote genomes highlights different evolutionary trajectories. In plants, LysM domains are highly represented, playing a relevant role in MAMPs perception, while, these domains are scarcely found in animals, suggesting the independent evolution of chitin and oligosaccharides recognition. Conversely, mammals possess families such as SIGLECs, which can sense glycosylation changes on glycoproteins and lipids (e.g., sialylation, branching) (Drickamer & Taylor, 2015), whereas no equivalent mechanism has yet been documented in plants. The wealth of structural information available from mammalian glycan‐binding receptors provides a valuable framework for predicting and characterizing binding mechanisms in plants, where structural data remains comparatively scarce. Importantly, the same PRRs may detect both MAMPs and DAMPs, depending on the origin of the glycan. For instance, in mammals, DECTIN‐1 and related lectins recognize β‐1,3‐glucans with β‐1,6 branches from fungal cell walls or ECM degradation (Brown et al., 2007), and in plants LysM receptors (e.g. CERK1) are required for PTI activation mediated by MLGs from the plant cell wall (DAMPs) or microbial pathogens (MAMPs) (Martín‐Dacal et al., 2023; Rebaque et al., 2021). This dual recognition underscores that PRRs are tuned to detect conserved structural motifs, rather than to discriminate their origin.

PRRs typically operate in complexes with co‐receptors to stabilize glycan binding and initiate downstream signaling. Current models, both in animals and plants, support the existence of supramolecular complexes involving PRRs, co‐receptors, numerous regulatory proteins, and downstream signaling components involved in glycan‐based DAMP/MAMP perception and immune activation (Anderluh et al., 2021; Snoeck et al., 2025). These higher order, membrane‐bound signaling complexes mediate interactions that signal transduction. Understanding the key players that regulate the timing of the immune responses is essential to elucidate the mechanisms of signal transduction and amplification (see Box 2). Therefore, identifying the molecular basis of ligand‐dependent supramolecular complex assembly is crucial for the rational design of elicitors and glycan‐based immunomodulators. Moreover, potential crosstalk between different receptor families should also be considered in future studies.

Advances in structural biology, such as those associated with new developments on experimental X‐ray crystallography, NMR spectroscopy, cryogenic electron microscopy or tomography (cryo‐EM, cryo‐ET) techniques, together with the breathtaking achievements in structural modeling provided by artificial intelligence (AI)‐based methods, provide a solid ground for further structural studies. The existence of reliable models predicted by different AlphaFold versions (Abramson et al., 2024; Jumper et al., 2021) for more than 200 million proteins in UniProt available in the AlphaFold Protein Structure Database (updated October 2025 with UniProt release 2025_03; Tunyasuvunakool et al., 2021, Fleming et al., 2025) opens a virtually unlimited field of exploration for structural investigations. Nonetheless, AlphaFold predictions must be interpreted with caution. While domain‐level accuracy is often high, the relative orientation of domains in multidomain PRRs and the geometry of large complexes can be inaccurate due to flexible linkers and missing long‐range restraints. Structural hypotheses of glycan‐binding pockets derived from AI‐based models should therefore be validated through experimental data (site‐directed mutagenesis) and integrative modeling (docking and molecular dynamics) (Snoeck et al., 2025). Even so, the combination of AI‐driven structural predictions with state‐of‐the‐art experimental and computational approaches now provides an unprecedented framework for dissecting glycan perception and PRR signaling. This integrative perspective positions the field to uncover the structural basis of glycan recognition in plant PRRs and to accelerate the rational use of glycan/PRR pairs and design novel glycan‐based immunomodulators for sustainable crop protection. Further structural characterization of glycan‐receptor pairs, together with functional validation of newly predicted glycan interactions obtained in silico, is urgently required. Yet, despite this momentum, many fundamental questions remain unresolved, including how plants integrate signals from multiple glycan receptors, how ligand specificity emerges from flexible and modular ectodomains, and how PRR complexes dynamically assemble at the cell surface, highlighting that glycan perception continues to be one of the most intriguing frontiers in plant immunity (see Box 2).

GLOSSARY

βααβ fold

A secondary structure arrangement consisting of two α‐helices packing against a short two‐strand β‐sheet that stabilizes the domain and creates a shallow helix‐sheet groove that can scaffold ligand (N‐acetylglucosamine‐containing carbohydrates) binding. Together, they form a compact structural module commonly found in carbohydrate‐binding domains such as LysM.

β‐sandwich

A protein fold in which two β‐sheets are packed against each other like a “sandwich,” forming a stable scaffold. This architecture is common in carbohydrate‐binding domains and immune receptors, providing structural stability and a platform for ligand interaction.

Aromatic stacking

A non‐covalent interaction in which aromatic residues, such as tryptophan, tyrosine, or phenylalanine interact with planar ligands, often through CH–π stacking with the C–H groups of carbohydrate rings, thereby stabilizing ligand binding.

CH–π interactions

Non‐covalent interactions between a C–H bond of a ligand (commonly from a sugar) and the electron π‐system of an aromatic residue. These interactions enhance specificity and stability in protein–carbohydrate complexes.

Aromatic clamp

A pair or cluster of aromatic residues that stabilize carbohydrate ligands via π‐stacking or CH–π interactions. Aromatic clamps often anchor sugar rings in a defined orientation for recognition and signaling.

V‐set Ig domains

Immunoglobulin‐like domain that adopts a β‐sandwich fold that mediates molecular recognition.

AUTHOR CONTRIBUTIONS

GVP: Writing, figure design and editing (Figures 1 and 2, and Table 1). MGA: Review and editing, funding acquisition. AM: Original draft preparation, review and editing, funding acquisition. MAT: Original draft preparation, writing, review and editing. LJ: Original draft preparation, writing, review and editing, funding acquisition.

CONFLICT OF INTEREST

The authors have no conflict of interest to declare.

DATA AVAILABILITY STATEMENT

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.