Encoding Cell‐Surface Receptor Function Through Glycosylation: Lessons From CD44

Abstract

Cell surface receptors are traditionally conceptualized as discrete molecular entities whose functional diversity arises primarily from gene expression and protein sequence variation. However, this view overlooks the extensive heterogeneity generated by post‐translational modifications, particularly glycosylation, which can profoundly alter receptor structure, dynamics, and interactions. Increasing evidence suggests that glycosylation does not simply fine‐tune receptor activity but rather defines distinct functional receptor states by modulating ligand recognition, nanoscale organization, interaction networks, and signaling competence. Among glycosylated receptors, CD44 provides a particularly informative model due to its extensive alternative splicing, dense glycosylation, and well‐characterized interactions with the extracellular matrix. CD44 exhibits remarkable structural and functional plasticity and regulates diverse biological processes ranging from cell adhesion and migration to inflammation and cancer progression. Importantly, glycosylation influences CD44 interactions with its principal ligand, hyaluronan, as well as its ability to cluster, engage coreceptors, and initiate downstream signaling pathways. In this review, we discuss how glycosylation regulates CD44 function across multiple levels, from molecular recognition to membrane organization and signaling. Using CD44 as a paradigmatic example, we explore how glycosylation may shape distinct aspects of receptor function, underscoring its importance as a fundamental determinant of receptor biology and its relevance for therapeutic targeting.

CD44 function is not solely determined by its protein sequence, it is dynamically encoded by its glycosylation state. Glycan remodeling generates distinct functional receptor states that regulate ligand recognition, receptor organization, lectin interactions, and signaling. Glycosylation results in an information‐rich layer that controls receptor function and provides new insights into CD44 biology in health and disease.

1. Introduction

Cell‐surface receptors mediate the exchange of information between cells and their extracellular environment, controlling processes as diverse as adhesion, migration, proliferation, differentiation, and immune responses. Traditionally, receptor functions have been interpreted primarily through gene expression, protein sequence, and ligand availability, leading to the view that receptor activities are largely encoded by the genome. However, this perspective does not fully account for the remarkable molecular heterogeneity displayed by many membrane proteins. Among the mechanisms that generate receptor diversity, glycosylation stands out as one of the most abundant, dynamic, and structurally complex post‐translational modifications in biology [1]. Unlike proteins, glycans are not directly template‐encoded but arise from tightly regulated biosynthetic pathways that integrate genetic, metabolic, and environmental inputs [2]. As a result, glycan structures exhibit extensive heterogeneity that varies across tissues, developmental stages, and pathological conditions [3]. This diversity introduces an additional layer of molecular information capable of profoundly influencing protein structure, dynamics, and molecular recognition.

Within this framework, glycans are increasingly viewed as integral components of receptor architecture that indeed can regulate receptor behavior at multiple levels of biological organization, including ligand accessibility, conformational dynamics, membrane localization, receptor clustering, and interactions with coreceptors, lectins, and extracellular matrix (ECM) components. Through these mechanisms, glycosylation contributes to the generation of distinct functional receptor states whose biological outputs may differ substantially despite sharing an identical protein backbone. This concept is particularly relevant for receptors operating within the highly crowded and dynamic environment of the cell surface. Receptor activity is increasingly recognized as an emergent property arising not only from direct ligand recognition but also from the organization of complex interaction networks involving proteins, glycans, lectins, lipids, and ECM components. Understanding how glycosylation influences these networks is therefore essential for deciphering the molecular basis of cell‐surface communication.

Among glycosylated receptors, CD44 provides an especially informative model system. CD44 combines extensive alternative splicing with dense N‐ and O‐glycosylation, generating a broad repertoire of structurally and functionally distinct receptor forms [4]. Best known as the principal receptor for hyaluronan, CD44 participates in diverse biological processes including cell adhesion, migration, inflammation, tissue remodeling, and cancer progression. Importantly, many of these functions are highly sensitive to glycosylation, which regulates not only ligand recognition but also receptor organization, lectin–mediated interactions, endocytic trafficking, and signaling competence.

In this Review, we use CD44 as a paradigmatic example to explore how glycosylation encodes receptor functional diversity. We discuss how glycans regulate CD44 function across multiple organizational levels, from molecular recognition and receptor architecture to membrane organization and signaling outputs. Through this example, we propose that glycosylation constitutes an information‐rich layer that encodes receptor functional states. Beyond CD44, these concepts provide a broader framework for understanding how glycosylation shapes receptor biology in physiological and pathological contexts.

2. CD44 as a Glycosylated Receptor System

The remarkable functional diversity of CD44 arises from the combination of multiple layers of molecular heterogeneity. Alternative splicing, post‐translational modifications, and dynamic extracellular interactions collectively generate a broad repertoire of receptor forms that differ in their structural and functional properties. Among these mechanisms, glycosylation occupies a particularly prominent role because it directly reshapes the molecular landscape through which CD44 engages ligands, lectins, coreceptors, and ECM components.

2.1. Structural Determinants of CD44 Diversity

CD44 is a type I transmembrane glycoprotein composed of four principal structural regions: an N‐terminal extracellular domain (ectodomain), a membrane‐proximal stem region, a single‐pass transmembrane helix, and a short cytoplasmic tail [4]. Together, these domains provide the structural framework that enables CD44 to integrate extracellular recognition with intracellular signaling and cytoskeletal interactions (Figure 1).

FIGURE 1.

CD44 domains organization. (A) Schematic view of CD44 domains, including the hyaluronic acid binding domain (HABD), stem region, transmembrane domain and cytoplasmic tail. (B) X‐ray crystallographic structure of HABD (PDB ID: 1POZ). (C) CD44 gene. Exons 1–5 and 16–20 are constitutively expressed while exons 6–15 undergo alternative splicing.

The ectodomain contains the hyaluronan‐binding domain (HABD), a conserved link module responsible for the recognition of hyaluronan (HA), the principal ligand of CD44. Structural studies using X‐ray crystallography, NMR spectroscopy, and molecular modeling have revealed an elongated binding surface enriched in positively charged residues that mediates interaction with the negatively charged HA polymer (Figure 1B) [5]. Importantly, these studies also highlighted the conformational plasticity of the HABD, suggesting that CD44 recognition involves a dynamic ensemble of binding‐competent states rather than a single rigid structure. Notably, most available structural information derives from nonglycosylated forms of the HABD, emphasizing the need to consider these structural models within the context of the extensively glycosylated receptor found at the cell surface.

Following the globular HABD, CD44 contains a membrane‐proximal stem region that displays a remarkable structural diversity through extensive alternative splicing. The human CD44 gene contains 20 exons, of which exons 1–5 and 16–20 are constitutively expressed, forming the standard isoform (CD44s). In contrast, exons 6–15 (commonly referred to as variant exons v1–v10) can be differentially included, generating multiple variant isoforms collectively known as CD44v (Figure 1C) [6, 7, 8, 9, 10]. Although the theoretical number of possible splice combinations is extremely large, only a subset of isoforms appears to be biologically abundant, exhibiting tissue‐specific and context‐dependent expression patterns [11, 12]. The variable segments are enriched in serine and threonine residues and are predicted to adopt highly flexible and intrinsically disordered conformations [13, 14]. Rather than simply extending the extracellular domain, these variable segments are thought to modulate the spatial positioning, accessibility, and interaction landscape of the HABD relative to the cell surface [6]. Through these effects, alternative splicing can alter the repertoire of extracellular interactions available to CD44 and influence its participation in signaling complexes. Importantly, alternative splicing is frequently dysregulated in pathological conditions, particularly cancer, where distinct CD44 variants have been associated with altered cell adhesion, migration, invasion, stemness, and therapeutic resistance [10, 12]. A major challenge in understanding CD44 functional diversity is that the consequences of alternative splicing extend beyond changes in protein sequence. By introducing regions with distinct glycosylation potential, splice variants create extracellular domains that differ not only in protein sequence but also in their capacity to acquire specific glycan modifications. By modifying the length, composition, and glycosylation potential of the extracellular domain, splice variants create distinct molecular scaffolds that support different interaction networks [15]. Consequently, protein sequence and glycosylation cannot be considered independently; rather, they cooperate to generate the structural diversity that underlies CD44 functional specialization.

2.2. Glycosylation Landscapes of CD44

Although several post‐translational modifications have been described for CD44, including phosphorylation [16, 17] and glycosaminoglycan attachment [18], glycosylation represents by far the most abundant and functionally influential modification of its extracellular domain (Figure 2). The extensive glycosylation of CD44 introduces an additional layer of structural diversity that complements alternative splicing and profoundly influences receptor functions [4, 8, 19, 20].

FIGURE 2.

CD44 glycosylation. (A) Schematic representation of CD44 and its post‐translational modifications. N‐site, N‐glycosylation site; O‐site, O‐glycosylation site; P‐site, phosphorylation site; GAG, glycosaminoglycan attachment site. (B) Protein sequence of CD44, highlighting experimentally validated N‐glycosylation sites (in bold) and NetOGlyc4.0 predicted O‐glycosylation sites (in bold and italic).

CD44 is extensively modified by both N‐linked and O‐linked glycans. In the standard isoform, five conserved N‐glycosylation sites (Asn25, Asn57, Asn100, Asn110, and Asn120) are located within or adjacent to the hyaluronan‐binding domain. Mass spectrometry analysis indicated high occupancy of all five N‐glycosylation sites, predominantly carrying complex‐type N‐glycans, except for N120, which is mainly occupied with oligomannose glycans [21]. The spatial positioning of these glycans places them in close proximity to the ligand–binding interface, where they can influence receptor accessibility, local electrostatics, folding, oligomerization, conformational dynamics, and intermolecular interactions, for instance, with hyaluronan [22]. Site‐specific glycoproteomic analyses have revealed that these positions are occupied by structurally diverse glycans whose composition varies according to cell type and physiological state [12].

O‐glycosylation provides an additional source of heterogeneity. While O‐glycans can be found throughout the extracellular domain, they are particularly abundant within the alternatively spliced variable region, which contains numerous serine and threonine residues capable of supporting dense mucin‐like glycosylation. As a result, variant isoforms frequently display substantially greater glycan content than CD44s, generating extended and highly hydrated extracellular architectures that may alter receptor accessibility and interaction potential [23].

Importantly, glycosylation is not static. The composition, branching, terminal modifications, and site occupancy of CD44 glycans are dynamically regulated by the cellular glycosylation machinery, which is highly sensitive to developmental, inflammatory, and oncogenic signals [24]. Consequently, a single CD44 isoform may exist as multiple glycoforms that may differ substantially in their structural and functional properties. Advanced glycoproteomic approaches are beginning to reveal the extraordinary complexity of CD44 glycoforms [12]. Alterations in N‐glycan branching, terminal sialylation, fucosylation, and O‐glycan truncation have been associated with distinct cellular phenotypes and disease states, particularly in cancer [4].

Similar principles of glycosylation‐dependent structural diversification have been described in other cell‐surface receptors, including mucins, integrins, and receptor tyrosine kinases (RTKs). In mucins, O‐glycans define an extended architecture that governs molecular accessibility [25], whereas in integrins and RTKs, N‐glycosylation modulates conformational equilibria, clustering, and ligand interactions [26], all positioning glycans as key regulators of receptor activation, organization, and signaling output.

The following sections examine how glycosylation influences CD44 function across multiple levels of biological organization, from molecular recognition and receptor clustering to signaling outputs and disease‐associated phenotypes. While the effects of CD44 glycosylation in ligand recognition have been clearly demonstrated, direct evidence linking its impact on receptor clustering and signaling remains limited.

3. From Molecular Recognition to Signaling Outputs: Glycosylation‐Dependent Functional States of CD44

3.1. Glycan‐Controlled Ligand Recognition: Hyaluronan Binding

CD44 is the principal receptor for hyaluronan (HA), a linear and highly negatively charged polysaccharide that constitutes a major component of the ECM [27]. This interaction underlies key cellular processes such as cell adhesion, migration, proliferation, and signaling [28] and stands as one of the most characterized examples of glycosylation‐dependent regulation of receptor function. Early functional studies established that not all CD44‐expressing cells bind hyaluronan, as they can exist in three different HA‐binding states: inactive, inducible, and constitutively active [29], and interconversion between states can be driven by glycosylation remodeling [30, 31, 32, 33, 34]. Together, these findings indicate that ligand binding is not solely determined by receptor expression level but also by dynamic structural regulation of the ectodomain.

Subsequent structural and computational analyses provided detailed insights into the molecular basis of CD44‐HA recognition. X‐ray crystallography, NMR, and site‐directed mutagenesis studies of the nonglycosylated HABD revealed an extended positively charged binding surface engaging mainly residues Arg41, Tyr42, Arg78, and Tyr79 [35, 36, 37]. These studies also highlighted that CD44 recognition involves broader conformational rearrangements and/or multiple binding surfaces. More recently, atomistic molecular dynamics simulations further supported this view by proposing distinct canonical and noncanonical binding configurations that differ in affinity and accessibility [38]. According to this model, N‐glycosylation emerges as a primary regulator of ligand binding.

The five conserved N‐glycosylation sites (Asn25, Asn57, Asn100, Asn110, and Asn120) are located within or adjacent to the HABD, positioning glycans near the HA–binding interface (Figure 3). Functional studies demonstrated that inhibition of N‐glycan biosynthesis or enzymatic removal of N‐glycans or their terminal sialic acids can activate CD44 for HA binding in otherwise nonbinding cells [22, 29, 39, 40, 41], while site‐specific mutagenesis identified N25 and N120 as critical regulators of receptor activation [29]. Mechanistically, N‐glycosylation could regulate HA binding through both steric and electrostatic effects. Molecular dynamics simulations predicted that complex‐type N‐glycans attached to N25, N100, and N110 collectively form an interlocking glycan shield over the canonical HA‐binding groove, thereby restricting ligand access to this high‐affinity binding site [42]. Further simulations with glycosylated HABD showed that this shielding effect becomes more pronounced as glycan size and branching complexity increase, physically restricting ligand access to this high‐affinity (sub‐μM) binding interface. In contrast, smaller N‐glycans, such as simply the first GlcNAc residue, lack both the size and charge necessary to impair HA binding, and instead, they could even facilitate the binding by providing additional interactions. Overall, N‐glycosylation does not simply reduce overall affinity but instead selectively redistributes the receptor toward distinct ligand‐binding topologies.

FIGURE 3.

Crystallographic structure of murine CD44‐HABD in complex with HA₈ (PBD ID: 2JCR). N‐glycosylation sites (purple) and key HA binding residues (orange) are highlighted.

Terminal glycan modifications further refine this regulation. High levels of sialylation on N‐glycans have been shown to reduce HA binding [29, 30, 41] and are proposed to amplify this steric effect through both increased glycan size and enhanced negative charge density. Molecular dynamics simulations provided a mechanistic explanation for this inhibitory effect. Sialic acid residues can engage key positively charged amino acids within the HABD, such as Arg41 and Arg154, thereby sequestering residues required for productive ligand interaction and generating electrostatic competition with the negatively charged HA polymer [43]. In this context, electrostatic effects appear to play a dominant role, with steric hindrance arising as a secondary consequence of glycan–mediated interactions. Interestingly, not all N‐glycosylation sites exert identical effects: while N25 directly modulates ligand recognition, N120 has been implicated in regulating CD44 oligomerization, suggesting an additional layer of control through modulation of receptor clustering and avidity [35, 44].

O‐glycosylation introduces an additional layer of complexity. Alternative splice variants containing serine/threonine‐rich extracellular segments frequently display an altered HA‐binding behavior compared with the standard isoform [31]. While early studies associated extensive O‐glycosylation with reduced HA binding, more recent evidence indicates that the effects of O‐glycans are highly context‐dependent. In breast cancer stem cells, the expression of α2,3‐sialylated core 2 O‐linked glycans promotes CD44 interaction with HA, facilitating downstream signaling associated with stem‐like properties [45]. In contrast, truncation of CD44 O‐glycans in gastric cancer cells notably increases HA binding capacity as well as alters CD44 extracellular shedding dynamics [46]. Together, these studies indicate that O‐glycosylation can either constrain or facilitate ligand recognition depending on glycan composition, structure, and cellular context.

Similar glycosylation‐dependent modulation of ligand recognition has been described in other receptor systems. In integrins, N‐glycosylation regulates ligand binding, receptor stability, and cell‐ECM adhesion, thereby tuning receptor activity and downstream signaling outputs [26, 47, 48]. Likewise, in RTKs such as EGFR, glycosylation influences receptor conformation and accessibility, impacting ligand‐induced dimerization and activation [49, 50]. In cytokine receptors, glycosylation can further modulate ligand affinity and receptor assembly, highlighting the broader role of glycans in fine‐tuning protein–ligand interactions [51].

Taken together, these observations establish glycosylation as a central determinant of CD44 ligand recognition. By modulating accessibility, electrostatics, conformational dynamics, and receptor activation, glycans regulate the transition between distinct HA‐binding states, which likely could impact on the organization of CD44 at the plasma membrane.

3.2. From Ligand Recognition to Receptor Organization

Ligand recognition represents only the first step in CD44 activation. Once hyaluronan (HA) engagement occurs, receptor function becomes increasingly dependent on the spatial organization of CD44 within the plasma membrane. In this context, glycosylation‐dependent regulation of ligand binding may extend beyond molecular recognition and influence the assembly of higher‐order receptor structures that ultimately determine signaling competence and cellular responses.

Accumulating evidence indicates that HA binding promotes the lateral redistribution and clustering of CD44 at the cell surface [41, 42]. This process is strongly influenced by the glycosylation state of the receptor since glycan‐dependent control of ligand accessibility determines which receptor populations can participate in HA‐mediated assembly. Consequently, glycosylation not only regulates whether HA binding occurs but also may shape the organization of receptor populations that emerge following ligand engagement. The properties of HA itself further contribute to this organizational process. High‐molecular‐weight hyaluronan (HMW‐HA), which predominates in healthy tissues, behaves as a multivalent scaffold capable of simultaneously engaging multiple CD44 molecules [52, 53, 54]. Through this mechanism, HMW‐HA promotes receptor clustering and the formation of stable membrane assemblies. These clustered structures display increased avidity relative to individual receptor–ligand interactions, transforming transient binding events into persistent adhesive platforms. In contrast, low‐molecular‐weight HA (LMW‐HA) fragments and HA oligosaccharides exhibit reduced valency and are less efficient at promoting receptor cross‐linking, as shown by surface plasmon resonance measurements [54]. Indeed, several studies have demonstrated that smaller HA fragments can disrupt preexisting CD44 clusters and favor receptor declustering. Thus, the molecular size distribution of HA directly influences receptor organization and provides an additional mechanism through which ECM composition regulates cellular behavior [52, 55].

These distinct organizational states have important functional consequences. HMW‐HA‐induced clustering is generally associated with tissue homeostasis, stable adhesion, cell survival, and anti‐inflammatory responses. Conversely, LMW‐HA fragments, which accumulate during inflammation, tissue injury, and tumor progression, favor more dynamic receptor arrangements and trigger rapid adhesive shifts associated with migratory and invasive phenotypes [52, 56, 57, 58]. The biological effects of HA therefore emerge not only from receptor occupancy but also from the specific membrane architecture generated following ligand engagement.

From a biophysical perspective, CD44 clustering represents a transition from molecular recognition to collective receptor behavior. Glycosylation regulates access to ligand‐binding states, while HA multivalency drives the assembly of receptor ensembles with distinct spatial and functional properties. Although direct experimental evidence remained limited, through this interplay, extracellular glycans and ECM components may have cooperated to shape receptor organization at the nanoscale. Importantly, receptor clustering is unlikely to represent the final stage of regulation. Rather, these organized receptor assemblies create platforms capable of engaging additional molecular partners, including lectins, coreceptors, and signaling proteins.

3.3. Glycan–Dependent Interaction Networks

The organizational states generated by hyaluronan engagement represent only one layer of CD44 regulation. In addition to protein‐mediated ligand recognition, CD44 also participates in a diverse array of glycan–mediated interactions through direct engagement with endogenous lectins. These interactions constitute a fundamentally distinct mode of molecular recognition in which binding specificity is encoded not by the protein backbone itself but by the composition, structure, and spatial presentation of glycans. Consequently, glycosylation does not only regulate receptor accessibility but actively creates new interaction interfaces that expand the functional repertoire of CD44. A defining feature of lectin–glycan interactions is their intrinsically low affinity, which is typically compensated by multivalent presentation of glycan ligands [59]. CD44 is particularly well suited for this mode of recognition because its extensive N‐ and O‐glycosylation provides a highly multivalent platform capable of engaging multiple carbohydrate‐binding proteins simultaneously. Through these interactions, CD44 can participate in dynamic molecular networks that influence receptor clustering, membrane organization, trafficking, and signaling (Figure 4).

FIGURE 4.

Schematic representation of CD44 interaction with lectins. Preferred binding epitopes, β‐galactosides, sialyl‐Lewis X and sialic acid for galectins, selectins and Siglecs, respectively, are depicted.

Among the endogenous lectins that are known to interact with CD44, galectins represent the best‐characterized family. Galectins recognize β‐galactoside‐containing glycans and are increasingly viewed as key organizers of the cell surface. Rather than acting as simple binding partners, galectins can crosslink glycosylated receptors into higher‐order assemblies, thereby translating glycan composition into membrane architecture and signaling behavior [60]. Galectin‐1 provides an illustrative example of this principle. Binding of Galectin‐1 to N‐glycosylated CD44 promotes the formation of multivalent interaction networks involving additional glycoproteins, including integrins and adhesion receptors [61]. Because Galectin‐1 exists in a monomer‐dimer equilibrium, changes in its oligomerization state can influence its capacity to bridge receptor populations and reorganize membrane assemblies. In this context, glycosylation‐dependent Galectin‐1 recognition contributes to the integration of CD44 within broader adhesion and signaling networks. Galectin‐3 offers an even more direct link between glycan recognition and membrane organization [62]. Owing to its unique ability to form higher‐order oligomers, Galectin‐3 can generate extensive glycan‐dependent lattices at the cell surface. Interactions between Galectin‐3 and branched N‐glycans on CD44 promote receptor clustering and increase receptor residency at the plasma membrane. Moreover, cooperation between Galectin‐3 and glycosphingolipids has been shown to drive membrane curvature and clathrin‐independent endocytosis of CD44 [63], linking glycan recognition to receptor trafficking and turnover. These observations highlight how glycan–mediated interactions can reshape membrane architecture and regulate receptor availability. Other galectins further illustrate the diversity of glycan‐dependent regulatory mechanisms. Galectin‐8 can bridge multiple glycosylated ligands with different specificities [64], facilitating the assembly of extracellular multiprotein complexes that modulate receptor engagement and inflammatory responses [65]. Galectin‐9 has emerged as a particularly important regulator of CD44 organization. Galectin‐9 has also been identified to interact with CD44 in a glycan‐dependent manner across multiple immune cell types, where it promotes the formation of functional receptor complexes at the cell surface [66]. As a tandem‐repeat galectin, Galectin‐9 can simultaneously engage multiple glycan epitopes, promoting CD44 cross‐linking and nanoscale clustering, as recently demonstrated by super‐resolution fluorescence microscopy in dendritic cells [67]. Beyond structural organization, Galectin‐9 regulates the formation of signaling platforms that couple CD44 to intracellular pathways, including the activation of AKT‐, ERK‐, and RhoA‐dependent signaling cascades. Through this mechanism, Galectin‐9 controls cytoskeletal remodeling, cell polarity, and migratory behavior. In regulatory T cells (Treg), clustering of CD44 driven by Galectin‐9 was reported to enhance the stability and suppressive functions of Treg cells by amplifying the TGF‐β (transforming growth factor‐β) signaling [68].

Beyond galectins, glycosylation can confer entirely new biological functions to CD44. A striking example is provided by the HCELL glycoform, in which specific sialofucosylated glycans convert CD44 into a functional ligand for E‐selectin [69, 70]. In this case, glycosylation effectively rewires receptor identity, enabling CD44 to mediate selectin‐dependent adhesion and cellular homing [71]. Such glycoform‐specific interactions play important roles in leukocyte trafficking, stem cell migration, and metastatic dissemination, illustrating how glycan remodeling can redirect receptor function toward entirely different biological pathways.

Interactions with sialic acid‐binding immunoglobulin‐like lectins (Siglecs) provide an additional level of regulation. Although the contribution of Siglecs to CD44 biology remains less well understood, increasing evidence suggests that highly sialylated CD44 glycoforms may engage Siglec‐mediated regulatory circuits within inflammatory and tumor microenvironments [4, 72, 73]. In addition, CD44 has been identified as a ligand for Siglec‐15 [74], and this interaction stabilizes receptor expression and promotes cell migration by limiting CD44 lysosomal‐mediated degradation [75]. These findings suggest that Siglecs’ recognition may contribute to the spatial organization and persistence of specific CD44 functional states.

Collectively, these observations reveal that glycosylation expands CD44 biology far beyond its interaction with hyaluronan. By generating recognition motifs for galectins, selectins, Siglecs, and other glycan‐binding proteins, glycans create a complex network of molecular interactions that influences receptor organization, trafficking, and cellular communication. Therefore, the signaling properties of CD44 therefore could emerge not only from its capacity to bind ligands but also from its ability to engage glycan–mediated interaction networks that organize and coordinate receptor function at the cell surface.

3.4. Emergence of Signaling Competence

The formation of glycosylation–dependent interaction networks raises a central question: how are these organizational states translated into cellular responses? Increasing evidence suggests that glycosylation regulates the assembly, stability, and signaling competence of membrane‐associated receptor complexes, thereby influencing how extracellular information is processed and transmitted into the cell.

CD44 functions as a multifunctional signaling platform by coordinating the assembly of membrane‐associated signaling complexes through interactions with RTKs, cytoskeletal adaptors, and membrane‐organizing proteins. Specifically, CD44 was early described as a cooperator of receptors such as EGFR and MET, facilitating the activation of downstream pathways including Ras/MAPK, PI3K/AKT, and Rho GTPase signaling cascades [28, 76, 77].

This signaling repertoire further extends to pathways associated with inflammation, transcriptional reprogramming, and cellular plasticity, including STAT3, NF‐κB, YAP/TAZ, and β‐catenin signaling, as well as broader stemness‐associated transcriptional programs. Collectively, these observations position CD44 as an integrator of signaling networks governing proliferation, migration, survival, and phenotypic plasticity [77, 78, 79] (Figure 5).

FIGURE 5.

Major signaling networks associated with CD44‐organized membrane signaling complexes. Importantly, CD44 is not depicted as a direct activator of these pathways but rather as an organizer of membrane‐associated signaling networks whose activity depends on receptor context and complex assembly.

Adhesion and ECM sensing, together with receptor organization at the plasma membrane, are key determinants of CD44 signaling behavior during neoplastic progression. These processes regulate cellular migration, tissue invasion, metastatic dissemination, and the dynamic interaction between tumor cells and their surrounding microenvironment, while also contributing to epithelial‐to‐mesenchymal transition (EMT), mechanotransduction responses, and cancer stem‐like phenotypes [78].

However, the contribution of glycans to these CD44‐associated signaling effects and their implications for human disease still remain to be completely understood. Rather than directly regulating pathway activation, glycans are likely influencing the structural context within which signaling occurs. By regulating ligand accessibility, receptor clustering, lectin engagement, membrane residency, and endocytic trafficking, glycans shape the probability that signaling‐competent complexes will form and persist at the cell surface [4]. In particular, lectin‐mediated interactions, such as galectin‐driven lattice formation, provide a mechanistic link between glycosylation and the spatial organization of signaling platforms. As previously described, increased N‐glycan branching can enhance galectin‐mediated lattice formation and receptor retention at the plasma membrane, whereas terminal sialylation can modulate both ligand accessibility and lectin recognition [80]. Likewise, sialofucosylated epitopes such as sialyl Lewis X generate specialized glycoforms capable of engaging selectins and promoting migratory behaviors [4, 81]. These examples demonstrate how glycosylation can alter receptor interaction networks and adhesive properties, although direct relation between specific glycoforms and defined signaling outcomes remains elusive. Through these mechanisms, CD44 glycosylation may influence the formation and stabilization of complexes containing RTKs and other membrane‐associated proteins, shaping the duration, spatial distribution, and persistence of signaling responses linked to PI3K/AKT‐, MAPK‐, and Src‐related pathways [4, 80].

The studies discussed throughout this review indicate that the effects of glycosylation on CD44 functionality emerge from the integration of multiple organizational levels, including ligand recognition, receptor clustering, lectin engagement, membrane residency, and signaling complex formation. In this context, distinct glycoforms can be viewed as distinct functional receptor states whose nature may change in response to developmental, inflammatory, metabolic, and pathological cues. Although substantial progress has been made in characterizing signaling differences among CD44 isoforms, considerably less attention has been devoted to understanding how dynamic glycosylation changes within a given isoform reshape signaling behavior. Defining how specific glycan structures influence the formation, stability, and activity of signaling complexes therefore remains a major challenge for the field. Addressing this question will be essential for understanding how glycosylation contributes to the functional diversity of CD44 in both physiological and pathological contexts.

4. Functional Receptor States in Disease and Therapeutic Intervention

The concept that glycosylation encodes receptor functional states has important implications for understanding disease. Alterations in glycosylation pathways are among the most common molecular changes associated with cancer, chronic inflammation, fibrosis, and immune dysfunction [82]. Consequently, pathological conditions frequently remodel the distribution of receptor states present at the cell surface, shifting receptors toward alternative organizational and signaling landscapes. CD44 provides a particularly compelling example of this principle. Its extensive alternative splicing, dynamic glycosylation, and diverse interaction network generate a broad spectrum of receptor states whose abundance and functional properties are profoundly altered during disease progression [83]. Understanding how these states emerge and contribute to pathology may provide new opportunities for therapeutic intervention.

4.1. Disease‐Associated Remodeling of CD44 Functional States

CD44 plays a central role in tumor progression by coordinating ECM sensing, cell adhesion, signaling, and cellular plasticity. Increased CD44 expression together with hyaluronan accumulation is a hallmark of many tumor microenvironments, supporting proliferation, migration, and metastatic dissemination. Moreover, cancer provides one of the clearest examples of glycosylation‐dependent remodeling of CD44 function.

For example, glycosylation‐dependent changes in CD44 binding affinity toward HA, fibronectin, and immune receptors such as Siglec‐15 can either activate or suppress oncogenic signaling pathways, ultimately influencing tumor cell survival, invasion, and metastatic dissemination [4]. In addition, tumor‐associated alterations in O‐glycan biosynthesis frequently occur during cancer progression and strongly influence CD44‐mediated signaling, EMT, stemness acquisition, and metastatic dissemination. Consequently, CD44 O‐glycosylation has emerged as both a prognostic biomarker and a potential target for precision oncology therapies [12, 84, 85]. Similarly, aberrant fucosylation of CD44 variants, particularly CD44v6, enhances the activation of pro‐survival signaling pathways such as PI3K/AKT/mTOR, thereby promoting tissue invasion, cancer stemness, and resistance to chemotherapy‐induced apoptosis. Elevated levels of fucosylated CD44 are associated with poor overall survival and increased metastatic potential, suggesting that tumor‐associated glycosylation converts CD44 into a potent oncogenic signaling platform [86, 87].

Glycosylation‐dependent interactions with immune lectins further contribute to tumor progression. The interaction between CD44 and Siglec‐15 stabilizes CD44 expression, thereby sustaining oncogenic signaling and promoting immune evasion. Together, these findings support a model in which tumor‐associated glycosylation converts CD44 into a dynamic signaling platform that integrates extracellular cues with oncogenic responses [75]. Despite increasing evidence supporting the importance of CD44 glycosylation in cancer biology, the glycosylation landscape of individual CD44 isoforms remains poorly characterized. In particular, the specific glycosylation patterns associated with CD44s and CD44v isoforms across different tumor types are still largely unknown, representing a major limitation in understanding how CD44 glycoforms contribute to tumor heterogeneity and disease progression [88].

A similar framework may apply to chronic inflammatory and fibrotic disorders. Inflammatory environments are characterized by extensive remodeling of both ECM composition and cellular glycosylation pathways, creating conditions that favor alternative CD44 functional states. Through its roles in leukocyte trafficking, ECM sensing, and tissue remodeling, CD44 contributes to the regulation of immune responses and fibrosis progression [89, 90]. Although the specific glycoforms involved remain incompletely characterized, increasing evidence suggests that glycosylation‐dependent changes in receptor organization and signaling may play important roles in these processes.

Viewed from this perspective, disease‐associated changes in CD44 function are not solely determined by receptor abundance. Rather, pathological conditions remodel the distribution of isoforms and glycosylation‐dependent receptor states, altering the balance between adhesive, migratory, inflammatory, and signaling‐competent populations.

4.2. Targeting CD44 Functional States

Traditional targeting strategies have largely focused on receptor expression levels or the presence of specific splice variants. However, increasing evidence suggests that therapeutic efficacy may ultimately depend on the ability to selectively target disease‐associated receptor states rather than CD44 as a whole.

Among CD44‐targeted approaches, chimeric antigen receptor (CAR)‐based therapies have gained significant attention. CD44v6 has emerged as a particularly promising target due to its restricted expression in healthy tissues and strong association with tumor aggressiveness, invasion, and therapy resistance. Preclinical studies demonstrated that CD44v6‐directed CAR‐T cells selectively recognize and eliminate tumor cells in a manner directly correlated with CD44v6 expression levels [91]. Similarly, CD44v6‐directed CAR‐T cells effectively infiltrate and control tumor growth in multiple solid tumor models, including ovarian and lung carcinomas, supporting the translational potential of CD44‐targeted adoptive immunotherapy [92]. Similarly, CD44‐targeted CAR‐NK cells exhibit potent cytotoxic activity, highlighting their potential as safer and effective alternatives to CAR‐T therapies [93]. Advanced strategies incorporating multiantigen targeting, such as dual CD44/CD133 CAR constructs, further improve therapeutic efficacy by reducing antigen escape and enhancing elimination of cancer stem cell populations [94].

Monoclonal antibodies targeting CD44 provide another therapeutic approach by blocking ligand binding, inducing receptor shedding, disrupting receptor clustering, and modulating immune cell recruitment [95]. Anti‐CD44 antibodies have demonstrated efficacy in both cancer and inflammatory diseases. In inflammatory arthritis, therapeutic anti‐CD44 antibodies reduce granulocyte accumulation within inflamed joints, suggesting that leukocyte depletion and impaired leukocyte trafficking significantly contribute to their anti‐inflammatory effects [96]. Similarly, anti‐CD44 monoclonal antibody treatment suppresses adipose tissue inflammation, improves insulin resistance, reduces hepatic steatosis, and lowers blood glucose levels in obesity‐associated metabolic dysfunction, suggesting that CD44 may also represent a therapeutic target in metabolic inflammation and type 2 diabetes [97]. In hematological malignancies, anti‐CD44 monoclonal antibodies suppress AML cell proliferation through inhibition of the PI3K/Akt/mTOR pathway and reduced protein synthesis, supporting the use of CD44 blockade as a potential therapeutic strategy in leukemia treatment [98]. Additional studies demonstrated that anti‐CD44 antibodies inhibit ovarian cancer stem‐like cells by suppressing proliferation and promoting apoptosis, suggesting that CD44 targeting may enhance the efficacy of conventional therapies while reducing tumor recurrence [99]. The clinical translation of anti‐CD44 therapies has already been explored. In a first‐in‐human phase I clinical trial, the humanized anti‐CD44 monoclonal antibody RG7356 was well tolerated in heavily pretreated patients with advanced CD44‐positive solid tumors, demonstrating the feasibility of CD44‐directed therapeutic intervention in humans [100]. Importantly, many of these therapeutic strategies likely function not only through receptor elimination but also through disruption of receptor organization and signaling competence. Antibody‐mediated interference with ligand binding, receptor clustering, or lectin engagement may alter the stability of disease‐associated receptor states and thereby modify downstream signaling behavior. Viewed in this way, therapeutic targeting can be interpreted as an attempt to reshape receptor‐state distributions rather than simply reduce receptor abundance. This conceptual shift may prove particularly important for highly dynamic receptors such as CD44, whose biological activity depends strongly on organizational and glycosylation‐dependent properties that are not captured by expression measurements alone.

4.3. Glycoform‐Specific Therapeutics: Toward Precision Glycobiology

A major limitation of current CD44‐targeted therapies is their inability to discriminate between physiological and pathological receptor states. Because canonical CD44 is widely expressed in healthy tissues, strategies based solely on receptor expression often face challenges related to specificity and toxicity. An emerging alternative is the selective targeting of disease‐associated glycoforms. Aberrant glycosylation generates unique molecular epitopes that may distinguish pathological receptor states from their physiological counterparts. Cancer‐associated O‐glycoforms, hypersialylated variants, aberrantly fucosylated species, and glycoforms generated through the combined action of alternative splicing and glycosylation represent particularly attractive candidates for selective therapeutic intervention (Figure 6A). Integrating glycoproteomic data, Aasted et al. identified a cancer‐specific epitope within the CD44v6 isoform and developed a monoclonal antibody (4C8) targeting a Tn‐glycosylated site [101]. The antibody showed high affinity and strict specificities for Tn‐modified epitopes, selectively recognizing cancer cells while sparing normal tissues despite CD44v6 expression. Immunohistochemical analysis demonstrated tumor‐selective staining across multiple carcinoma types. When reformatted as a CAR, 4C8‐directed T cells exhibited potent and Tn‐dependent cytotoxicity in vitro, reduced tumor growth and improved survival in vivo, and selectively eliminated malignant cells in human organotypic skin model without affecting normal keratinocytes. This work highlights the potential of targeting aberrant glycosylation to achieve improved specificity and safety in solid tumor immunotherapy (Figure 6B).

FIGURE 6.

Targeting CD44 glycoforms to improve cancer specificity. (A) Schematic representation of CD44 glycosylation diversity highlighting the distinction between physiological and cancer‐associated glycoforms. (B) Glycan‐based therapeutic targeting of CD44.

Recent advances in glycoproteomics, glycoproteogenomics, spatial glycomics, and glycoengineering are beginning to provide the tools required to identify and characterize these disease‐associated receptor states at high resolution. These technologies may ultimately enable the development of antibodies, cellular therapies, and targeted delivery systems that recognize glycosylation‐defined receptor states with unprecedented specificity. In parallel, CD44‐targeted nanotherapeutic systems are being developed for selective drug delivery. Functionalized nanocarriers exploiting CD44‐mediated internalization have demonstrated improved tumor targeting, enhanced drug accumulation, and reduced systemic toxicity in several solid tumors, including nonsmall cell lung cancer [102]. The long‐term objective extends beyond targeting CD44 itself. By identifying and selectively manipulating pathological receptor states, future therapeutic strategies may exploit glycosylation as a source of molecular precision. Such approaches would represent a shift from targeting receptors as static molecular entities toward targeting the dynamic functional states that govern their biological activity. CD44 provides an ideal model system for exploring this emerging paradigm of precision glycobiology.

5. Summary and Outlook

CD44 exemplifies how receptor function cannot be fully understood from protein sequence, expression levels, or ligand availability alone. Rather, its biological activity emerges from the integration of alternative splicing, glycosylation, membrane organization, and extracellular interactions. Through the interplay of these mechanisms, CD44 exists not as a single molecular entity but as a dynamic ensemble of structurally and functionally distinct receptor states. The studies discussed in this Review demonstrate that glycosylation influences CD44 at multiple levels. Strong evidence supports the role of CD44 glycosylation in regulating ligand accessibility and molecular recognition and the formation of receptor interaction networks. Together, these findings indicate that glycosylation‐dependent mechanisms contribute to modulating receptor clustering and membrane compartmentalization and ultimately shape signaling competence and cellular behavior. Therefore, glycan effects may extend beyond the regulation of individual binding events and contribute to the assembly and stability of complex signaling networks at the cell surface. Although CD44 provides a particularly tractable model system, the principles emerging from its study are likely applicable to many classes of cell‐surface receptors, including integrins, RTKs, cytokine receptors, and immune receptors.

A major challenge for the coming years will be to establish direct links between specific glycan structures, receptor organization, and signaling outputs in living systems. Advances in glycoproteomics, glycoproteogenomics, glycoengineering, structural biology, molecular simulations, and high‐resolution imaging are beginning to provide the tools necessary to address this challenge. Integrating these complementary approaches should enable the identification and characterization of receptor states with unprecedented molecular detail and reveal how glycosylation contributes to cellular heterogeneity in both physiological and pathological settings.

From a translational perspective, disease‐associated glycoforms represent an emerging class of therapeutic targets. Rather than targeting receptors solely on the basis of expression, future therapeutic strategies may exploit glycosylation‐defined receptor states to selectively modulate pathological signaling while preserving physiological functions. Such approaches could open new opportunities for precision medicine in cancer, inflammation, fibrosis, and immune disorders.

Ultimately, the lessons learned from CD44 point to a broader principle of cell biology: glycosylation is not solely a post‐translational modification that modulates receptor activity. Instead, glycosylation emerges as an information‐rich layer that encodes receptor functional states and shapes cellular responses to the extracellular environment. Deciphering how glycan structures organize receptor behavior will be essential for a comprehensive understanding of cell‐surface communication and may reveal fundamentally new principles governing molecular recognition and signaling in biology.

Funding

This study was supported by Agencia Estatal de Investigación (PID2023‐150779OA‐I00 and PID2021‐1237810B‐C21).

Conflicts of Interest

The authors declare no conflicts of interest.

Biographies

María Payá‐García received her B.Sc. in Chemistry and M.Sc. in Organic Chemistry from the University of Valencia (Spain). In 2024, she joined the Chemical Glycobiology Laboratory at CIC bioGUNE, where she is currently conducting her Ph.D. under the supervision of Prof. Jesús Jiménez‐Barbero and Dr. Ana Gimeno. Her doctoral thesis focuses on glycoprotein–lectin interactions, integrating NMR spectroscopy, molecular biology, and biophysical approaches to explore the role of glycosylation in molecular recognition and cancer‐related processes.

José Manuel Rojas‐Marcos Hernaez obtained his B.Sc. in Chemistry from the University of Seville and subsequently completed an M.Sc. in Organic Chemistry at the University of Santiago de Compostela. He is currently pursuing his Ph.D. in the Chemical Glycobiology Laboratory at CIC bioGUNE, under the supervision of Dr. Ana Gimeno and Prof. Jesús Jiménez‐Barbero. His research focuses on the study of protein–glycan interactions and the molecular recognition processes underlying glycoprotein function, with a particular emphasis on the application of NMR spectroscopy to glycobiology.

Iris A. Bermejo studied chemistry at the University of La Rioja (Spain), where she received her Ph.D. in 2018, funded by a Ph.D. fellowship from the Spanish Association Against Cancer (AECC), working on cancer immunotherapy and the synthesis of glycomimetic‐based cancer vaccines. In 2020, she was awarded a Marie Skłodowska‐Curie Action (MSCA) for a postdoctoral stay at the University of Vienna (Austria). Since 2023, she has been a researcher at CIC bioGUNE, supported by a second MSCA. Her research focuses on glycans as regulators of biomedical processes, particularly in human coronaviruses, and on glycan–protein interactions using a multidisciplinary approach.

Diego Del Balzo is a biochemist and cell biologist with experience in intracellular trafficking, host–pathogen interactions, immunity, gene editing, and glycobiology. He obtained the Ph.D. in Biological Chemistry from IMBECU–CONICET, Argentina, in 2024, where he studied intracellular transport during Chlamydia trachomatis infection and evaluated vaccine formulations. During the Ph.D., Diego completed research internships at Harvard Medical School, Gustave Roussy Cancer Campus, and the Institut de Psychiatrie et Neurosciences de Paris. Currently Diego is postdoctoral researcher in the Chemical Glycobiology Lab at CIC bioGUNE, Spain, where he generates genetically edited cell libraries to investigate and manipulate cellular glycosylation. His research combines molecular and cellular biology with gene editing, structural, and chemical biology approaches.

Ana Ardá is Ikerbasque Associate Professor and Associate Principal Investigator at CICbioGUNE. Her research focuses on molecular recognition phenomena, particularly on elucidating the molecular basis of glycan‐mediated biological events involved in immune regulation and host–pathogen interactions. She studied chemistry and obtained her Ph.D. from the University of A Coruña (Spain). She subsequently joined the Utrecht University as a postdoctoral researcher. She then moved to Madrid to join the group of Prof. Jiménez‐Barbero at the Center for Biological Research (CIB‐CSIC). In 2014, she moved to CIC bioGUNE and in 2015. she was awarded a tenure‐track position through the Ramón y Cajal Programme of the Spanish Ministry of Science. In 2022, she was promoted to Ikerbasque Research Associate, and in 2025 to Associate Professor.

Luca Unione is an Ikerbasque Associate Professor and Associate Principal Investigator at CICbioGUNE, leading research on glycan‐receptor interactions in infectious diseases. Trained as a chemist in Italy, he earned his Ph.D. in pharmacy in Madrid. His career includes translational pharmaceutical research and an HFSP Fellowship at Utrecht University to develop chemo–enzymatic tools to study glycans and viral infections. He joined CICbioGUNE in 2022 and received an ERC Starting Grant in 2023 to investigate glycans in infectious diseases at atomic resolution. In 2025, he was awarded a Ramón y Cajal Fellowship. He is visiting professor at the University of Deusto.

Jesús Jiménez‐Barbero is Ikerbasque Professor and Scientific Director of CIC bioGUNE since 2014. After his PhD (Chemistry, Madrid), he was postdoc in Zurich, Mill Hill, and Pittsburgh before joining CSIC as research professor and later bioGUNE. A recognized leader in chemical glycobiology, his group develops and applies NMR methods for studying glycan–protein recognition. He has coauthored 700 publications, delivered 350 invited lectures, supervised 40 PhDs, and received the Whistler Award, RSEQ Medal, and Spain's National Research Prize, among others.

Ana Gimeno is an emerging scientist at CIC bioGUNE (Spain), holding both Ikerbasque and Ramón y Cajal Research Fellow positions. She earned her PhD at the University of Valencia (Spain), completed postdoctoral research at CIC bioGUNE (Spain) and Utrecht University (Netherlands), and returned to CIC bioGUNE to establish her independent research program. Her work develops chemical biology tools to address key biomedical challenges, combining NMR spectroscopy, molecular biology, and biophysical methods to uncover how glycosylation influences cancer progression and molecular recognition. Her work focuses on glycan–lectin interactions, glycoconjugate structure, and therapeutic glycoproteins.

Dedicated to Professor Manuel Martin‐Lomas on the occasion of this 85th anniversary.

Data Availability Statement

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