Coreceptor functions of cell surface heparan sulfate proteoglycans

Abstract

Receptor-ligand interactions play an important role in many biological processes by triggering specific cellular responses. These interactions are frequently regulated by coreceptors that facilitate, alter, or inhibit signaling. Coreceptors work in parallel with other specific and accessory molecules to coordinate receptor-ligand interactions. Cell surface heparan sulfate proteoglycans (HSPGs) function as unique coreceptors because they can bind to many ligands and receptors through their HS and core protein motifs. Cell surface HSPGs are typically expressed in abundance of the signaling receptors and, thus, are capable of mediating the initial binding of ligands to the cell surface. HSPG coreceptors do not possess kinase domains or intrinsic enzyme activities and, for the most part, binding to cell surface HSPGs does not directly stimulate intracellular signaling. Because of these features, cell surface HSPGs primarily function as coreceptors for many receptor-ligand interactions. Given that cell surface HSPGs are widely conserved, they likely serve fundamental functions to preserve basic physiological processes. Indeed, cell surface HSPGs can support specific cellular interactions with growth factors, morphogens, chemokines, extracellular matrix (ECM) components, and microbial pathogens and their secreted virulence factors. Through these interactions, HSPG coreceptors regulate cell adhesion, proliferation, migration, and differentiation, and impact the onset, progression, and outcome of pathophysiological processes, such as development, tissue repair, inflammation, infection, and tumorigenesis. This review seeks to provide an overview of the various mechanisms of how cell surface HSPGs function as coreceptors.

INTRODUCTION

Virtually all cells of eukaryotic, multicellular organisms express one or several types of heparan sulfate proteoglycans (HSPGs) on their cell surface. HSPGs consist of a core protein and one or several covalently attached heparan sulfate (HS) chains. Some HSPGs are hybrid proteoglycans carrying both HS and chondroitin sulfate (CS) chains. HS is a linear polysaccharide comprised of repeating disaccharide units with a signature repeat of (GlcUAβ1-4GlcNAcα1–4)_n (1, 2). HS is attached to and polymerized on Ser residues surrounded by a motif containing a hydrophobic amino acid adjacent to the Ser-Gly dipeptide sequence and a cluster of acidic amino acids on the C-terminal side of Ser-Gly sequence in HSPG core proteins (3, 4). The unmodified HS polysaccharides are modified in the Golgi by N-deacetylation, N-sulfation, epimerization, and several O-sulfation reactions that are catalyzed by distinct membrane-bound enzymes (5). The biosynthetic machinery of HS is evolutionarily conserved among vertebrates and invertebrates, suggesting conservation of function for the HS chains. Because HS polymerization and modification reactions do not go to completion, the biosynthetic process generates an exceptionally diverse array of HS structures, both in length and extent of modification. HS can vary in length from 20 to 150 disaccharides with cell type and core protein (6). A mere HS decasaccharide can potentially assume over 10⁶ distinct sequences, which is already in vast excess of the estimated gene products that the whole human genome can generate. This enormous structural diversity largely explains why and how HS can bind to so many proteins.

Cell surface HSPGs include the syndecan (SDC) and glypican (GPC) families with four and six members in mammals, neuropilin-1, CD44, neurexin, and betaglycan (7–11). The latter three are part-time proteoglycans as they may appear in proteoglycan form or as glycoproteins lacking glycosaminoglycans (GAGs). Several HSPGs can harbor HS or CS chains and the type of GAG that they carry can influence their coreceptor function. For example, neuropilin-1 enhances VEGFR2 signaling when it carries HS chains, but inhibits VEGFR2 signaling when it carries CS chains (12). SDCs, neuropilin-1, CD44, neurexin, and betaglycan are type I transmembrane core proteins whereas GPC core proteins are linked to the cell surface by a glycosylphosphatidylinositol (GPI) anchor.

Coreceptors in general are selective for particular receptor-ligand interactions. For example, CD4 is a coreceptor for T cell receptor interactions (13) and CCR5 is a coreceptor for human immuodeficiency virus (HIV) infection (14). HSPG coreceptors are unique in that they can bind to many HS/heparin (HP)-binding proteins through their HS chains and regulate many different types of receptor-ligand interactions. Furthermore, several cell surface HSPGs can interact with ligands independent of their HS chains. For example, betaglycan and neuropilin-1 bind to transforming growth factor (TGF)β family members and vascular endothelial growth factor (VEGF), respectively, through their core proteins. Similarly, the core protein of GPC3 binds with high affinity to sonic Hedgehog (Shh) and Indian Hedgehog (Ihh) (15, 16).

HSPG coreceptors bind soluble and insoluble ligands and screen them, place them at appropriate locations, stabilize them, or distribute them to neighboring cells (Fig. 1). HSPG coreceptors can also increase the avidity of ligands for their signaling receptors by increasing their local concentration, inducing a conformational change, or by serving as a platform for oligomerization (Fig. 1). Similarly, HSPGs can bind to receptors and transport them to appropriate locations, assist them in ligand binding, induce a conformational change, or facilitate their oligomerization (Fig. 2). Furthermore, HSPG coreceptors can regulate signaling components by recruiting them to the receptor complex, altering their stability, and influencing their intracellular trafficking patterns (Fig. 2). In addition, the transmembrane HSPG coreceptors can induce separate signal transduction events that modulate the cellular response. Through these activities, cell surface HSPGs fine tune receptor-ligand interactions and ensuing cellular responses. Insight into functions of cell surface HSPGs as coreceptors was first revealed by studies that discovered the importance of cell surface HS binding to fibroblast growth factor (FGF) 2 for high-affinity interactions with receptor tyrosine kinases (17, 18). These observations served to further galvanize interest in the role of cell surface HSPGs as coreceptors for other growth factors and other HS/HP-binding proteins.

Figure 1.

Schematic overview of coreceptor functions of cell surface heparan sulfate proteoglycans (HSPGs) I. A: interaction with soluble ligands: cell surface HSPGs capture soluble ligands, increase their local concentration, and increase their availability for signaling receptors. B: interaction with extracellular matrix (ECM) components: cell surface HSPG binding facilitates binding of ECM components to integrin receptors in a heparan sulfate (HS)-dependent manner. C: formation of ligand/receptor/HSPG complexes: cell surface HSPGs bind to both ligands and signaling receptors to form a tertiary receptor complex. D: oligomerization of ligands: cell surface HSPGs provide a HS platform with multiple binding sites for ligand oligomerization, and ligand oligomers interact more avidly with signaling receptors.

Figure 2.

Schematic overview of coreceptor functions of cell surface heparan sulfate proteoglycans (HSPGs) II. A: interaction with receptors: cell surface HSPGs can enhance signaling by directly regulating receptor distribution, stability, oligomerization state, and activity. B: trans coreceptors: cell surface HSPGs present ligands to receptors on neighboring cells thereby functioning as in trans coreceptors. C: interaction with intracellular molecules: cell surface HSPGs with cytoplasmic domains can enhance receptor signaling through recruitment of intracellular signaling molecules to the receptor complex. D: intracellular trafficking of ligands: HSPG coreceptors control ligand availability at the cell surface through endocytosis of ligands and either recycling them back to the cell surface of transporting them to lysosomes for degradation. E: formation of ligand gradients: cell surface HSPGs immobilize morphogens to form morphogen gradients during development and immobilize chemokines to form haptotactic gradients to guide the directional migration of leukocytes.

INTERACTION WITH SOLUBLE LIGANDS

Cell surface HSPGs bind to soluble ligands and enhance their activity by affecting their concentration, destination, conformation, oligomerization, or stability (7, 10, 19) (Fig. 1A). Two seminal studies published in Science (17) and Cell (18) in 1991 were the first to establish the importance of cell surface HSPG coreceptors in growth factor signaling. Before these studies, it was already known that many growth factors bind to HS/HP (20–24). It was also known that FGF2 binds to HS/HP with low- to mid-nanomolar affinity and that deletion of HS/HP-binding regions in FGF2 dramatically reduces its biological activity (25, 26). The two studies established that cell surface HSPGs function as a coreceptor for FGF2 by showing that 1) mutant CHO cells lacking HS but expressing the high affinity, receptor tyrosine kinase fibroblast growth factor receptor (FGFR)1, and CS are unable to bind FGF2; 2) FGF2 binds to sulfated HS expressed on the surface of fibroblasts and skeletal muscle cells; 3) binding to cell surface HS is required for efficient binding of FGF2 to FGFR1; and 4) cell surface HS is required for FGF2 signaling in fibroblasts and skeletal muscle cells (17, 18). Consistent with these data, FGF2 signaling measured by the activation of ERK MAP kinases is prominent in mouse embryonic fibroblasts (MEFs) expressing SDC1, but markedly reduced in Sdc1^−/− MEFs (Fig. 3A). Soluble HP, but not CS, can restore FGF2 binding to cells without sulfated GAGs at concentrations as low as 1 ng/mL (17). These findings suggest that HSPGs not only localize FGF2 to the cell surface, but HSPG binding also induces changes in FGF2 or FGFR1. Indeed, monomeric FGF2 forms oligomers when bound to HS that, in turn, dimerize FGFR1, which is required for signaling (27). Similarly, exogenous HS enhances HB-EGF signaling in a dose-dependent manner in wild-type (Wt) mouse mesenchymal stem cells (MSCs), but CS does not (Fig. 3B). HSPGs also regulate FGF signaling in lower organisms, such as Drosophila (28) and zebrafish (29). Other members of the FGF family are also regulated by cell surface HSPGs, but the structural requirement for HS binding to FGFs is apparently distinct and rather complex. For example, FGF1 and FGF2 require different HS structures for binding (30) and IdoA is required for FGF2 binding but not for FGF10 binding (31).

Figure 3.

Cell surface syndecan (SDC)1 and exogenous heparan sulfate (HS) enhance growth factor signaling in mouse embryonic fibroblasts (MEFs) and mesenchymal stem cells (MSCs). A: MEFs isolated from Wt or syndecan-1 null (Sdc1^−/−) mice on C57BL/6J were stimulated with fibroblast growth factor (FGF)2 at the indicated concentrations for 30 min at 37°C. Whole cell lysates were Western blotted for phospho-p44/42 (Thr202/Tyr204) (p-p44/42) and total p44/42 (Cell Signaling Technology, Danvers, MA). B: MSCs isolated from inguinal fat pads of Wt C57BL/6J mice were incubated with HB-EGF without or with HS or chondroitin sulfate (CS) for 30 min at 37°C. Whole cell lysates were Western blotted for p-p44/42 and total p44/42.

Similar to FGF2, Wingless (Wg), the Drosophila homolog of the proto-oncogene WNT1, binds to HP and Wg bound to cells can be released by exogenous HS (32). Wg signaling is inhibited by treatment of cells with heparinases or chlorate, indicating that sulfated HSPGs bind Wg and enhance Wg activity. Furthermore, exogenous HP can restore Wg activity in chlorate-treated cells. The difference between FGF2 and Wg signaling is that CSPGs can also serve as a coreceptor for Wg (32). However, abrogation of HS synthesis severely impairs Wg, Hedgehog (Hh), and decapentaplegic (Dpp) signaling in Drosophila (33), whereas deletion of the gene for N-deacetylase N-sulfotransferase 1 (Ndst1) impairs Shh and FGF signaling in mice (34), suggesting that proteoglycans carrying HS are the dominant coreceptors. Consistent with these findings, Dally and Dally-like protein (Dlp) are the HSPG coreceptors for Wg in Drosophila (35–37), whereas several GPCs (38–41) and SDCs can serve as coreceptors for WNTs in mammalian cells (42–44). Interestingly, loss of Dlp has a more marked effect on the accumulation of Wg than the loss of Dally (45). Dally binds to Wg via HS, whereas Dlp can bind via both HS and the protein core (46). This biphasic binding is apparently important in shielding the palmitoleate moiety of Wg (45). The human homologs of Dlp, GPC4, and GPC6, also bind to the lipid moiety of WNT, but not GPC3 and GPC5 (45), demonstrating that GPC family members have distinct coreceptor functions.

The HS-modified isoform of CD44, CD44v3, binds to hepatocyte growth factor (HGF) and concentrates HGF at the cell surface, thereby facilitating its interaction with c-Met and activation of signaling in lymphoma cells (47). However, several studies also suggest that CD44 may enhance the activation of Met and other receptor tyrosine kinases in an HS-independent manner as CD44 isoforms lacking HS chains also bind and present HGF and VEGF to their respective signaling receptors (48). Cell surface SDC1 also binds to HGF and enhances HGF signaling through Met in multiple myeloma cells (49). Interestingly, shed SDC1 also binds to HGF and facilitates its interaction with Met in multiple myeloma cells (50), suggesting that SDC1 not only localizes HGF to the cell surface but also boosts HGF activity. In hepatocellular carcinoma (HCC), HCC cell migration and motility do not respond to either canonical or noncanonical Wnt activation but markedly increase with HGF treatment (51). GPC3 binds to HGF in an HS-dependent manner and facilitates HGF-Met signaling in HCC cells, which leads to increased cell migration and motility. Importantly, blocking HS chains of GPC3 inhibits HCC spheroid formation in vitro and liver tumor growth in mice (51). Together, these studies indicate that several HSPG coreceptors positively regulate growth factor interactions to promote cancer cell motility, proliferation, and survival.

Cell surface HSPGs are also used as coreceptors by secreted microbial virulence factors. HIV envelope protein gp120 can be secreted (52) and its interaction with genital epithelial cells triggers innate immune responses that impair mucosal barrier functions (53). Soluble gp120 binds to HS, Toll-like receptor (TLR)2, and TLR4, and stimulates proinflammatory cytokine production via activation of NF-κB in HEK293 cells (53), suggesting that cell surface HSPGs expressed by epithelial cells facilitate gp120 engagement of TLRs. How this is accomplished is not known, but gp120 is activated when it undergoes a conformational change (52) and HS binds and activates TLR4 signaling in dendritic cells (54). These findings suggest that HSPGs may enhance gp120 signaling by inducing a conformational change in gp120, stimulating TLR4, or both.

Other HIV-secreted factors also bind to HSPGs and these interactions are considered to contribute to AIDS pathogenesis. Cell surface HSPGs are coreceptors for p17 matrix protein released by HIV-infected cells (55). Here, HSPG binding via sulfated HS motifs promotes p17 oligomerization and association with its signaling receptors CXCR1 and CXCR2 on leukocytes (56). Colocalization of SDC2 and CXCR2 is seen in hepatic stellate cells when treated with p17 (57), suggesting that SDC2 is one of the HSPG coreceptors for HIV p17.

The HIV transactivating factor Tat is a small, cationic protein released from HIV-infected cells and is found in blood of patients infected with HIV. Cell surface HSPGs bind to Tat, protect Tat from degradation, and induce Tat oligomerization (58–60). Tat-HSPG interaction is important (61), but not essential (62), for clathrin-dependent endocytosis of Tat, indicating involvement of other receptors for internalization (62). Indeed, Tat also binds to αvβ3 integrins and activates focal adhesion kinase (FAK) signaling (63).

Using its HSPG connection, Tat can facilitate the extravasation of lymphoid cells (64). Tat accumulates on the surface of endothelium by binding to HSPGs, and B-lymphoid cells expressing SDC1 or HIV-infected T cells endogenously expressing Tat can attach to the Tat-HSPG complex on the endothelial surface. This heterotypic lymphoid cell-endothelial cell interaction is inhibited by HS inhibition, but not by integrin inhibition, and requires the homodimerization of Tat. Furthermore, Tat-HSPG on endothelial cells promotes the transendothelial migration of PBMCs in vitro and in a zebrafish embryo model of inflammation. Together, these data indicate an interesting mechanism where Tat bound to lymphoid HSPG and Tat bound to endothelial cells dimerize to form a HSPG-Tat-Tat-HSPG quaternary complex (64). This mechanism physically links lymphoid cells to the endothelium and promotes their extravasation.

INTERACTION WITH ECM COMPONENTS

Cell surface HSPGs also function as coreceptors for ECM components (Fig. 1B). Cell surface HSPGs, especially members of the transmembrane SDC family, facilitate formation of a signaling complex that translates information from the extracellular environment to specific cellular responses. Most ECM components contain HS/HP-binding domains and cell surface HSPGs bind to these molecules in an HS-dependent manner. The list is extensive and includes fibrillar collagens (65–67), fibronectin (68–70), thrombospondin (71), laminin (72), vitronectin (73), tropoelastin (74), fibrillin (75), and tenascin (76), among others. Cell surface HSPGs generally facilitate integrin interactions with ECM components. For example, SDC4 coordinates with α5β1 integrins in fibroblasts to attach and spread on fibronectin, and promote focal adhesion formation (69). Both SDC4 and GPC1 can bind to the C-terminal HP-binding HepII domain of fibronectin, but only SDC4 can support focal adhesion formation because the ability of SDC4’s cytoplasmic domain to bind and activate PKCα is important in this process (77). SDC4, but not SDC2, is also associated with focal adhesions formed on laminin, vitronectin, or collagen I in a β1 or β3 integrin-dependent manner (78), suggesting that SDC4 is the primary cell surface HSPG coreceptor that coordinates with integrins in the formation of focal adhesions.

Cell surface HSPGs also participate as coreceptors in cell attachment to other adhesive substrates. For example, cell surface HSPGs mediate mesenchymal cell attachment to immobilized ADAM12 (a disintegrin and metalloproteinase) (79, 80). HS chains of SDC1, SDC2, and SDC4, but not GPC1, bind to the disintegrin domain of ADAM12. After binding via SDCs, cells spread, assemble stress fibers, and form focal adhesions. The postadhesion events are mediated by β1 integrins because blocking or deleting β1 integrins does not inhibit attachment, but inhibits cell spreading. These findings indicate that mesenchymal cells use SDCs for attachment and β1 integrins for signaling when interacting with immobilized ADAM12 (79).

Cell surface HSPGs also positively regulate the assembly of ECM components. Soluble fibronectin is assembled into fibronectin fibrils in a cell-mediated process that involves β1 integrins and cell surface HSPGs (81–83). In plasma, fibronectin exists as a soluble dimer, but in the ECM it is found as an insoluble fibril. Cell surface HSPGs are thought to promote the assembly of fibronectin into detergent-insoluble fibrils by bringing together soluble fibronectin by binding to the C-terminal HepII domain of fibronectin (68), and promoting fibronectin-fibronectin interactions in the N-terminal type I modules (84, 85). Indeed, CHO cells lacking HS do not assemble a fibronectin matrix (85), indicating that HSPGs are important cofactors. Furthermore, SDC2, but not SDC4, promotes fibronectin assembly in cultured mesenchymal cells (86, 87) and in zebrafish (88). SDC1 also interacts with β1 integrins via cytoplasmic domain interactions in epithelial cells (89), and deletion of SDC1 markedly impairs fibronectin assembly in corneal basement membranes (90). Together, these findings indicate that both mesenchymal and epithelial cells use SDCs to facilitate fibronectin assembly in the interstitial ECM and in basement membranes of select tissues.

Like fibronectin, laminins are modular proteins with domains that interact with both cells and various ECM components (91, 92). Laminin is a large heterotrimeric protein consisting of an α chain, β chain, and γ chain, and named based on the chain composition of the heterotrimer (e.g., LM-α3β3γ2 = LM332) (92, 93). Laminin is essential for basement assembly because the laminin lattice serves as a scaffold for the assembly and incorporation of other basement membrane components (92–94). Although laminin can self-assemble (95), the cell surface provides a nucleation surface for laminin polymerization (92, 96), and laminin binding to cell surface receptors is critical for laminin assembly (97). Cell surface tethering of laminin is mediated by several receptors, including integrins, α-dystroglycan, sulfated glycolipids, and HSPGs (92, 96). The identity of HSPGs that participate in laminin assembly is still under investigation, but SDC1, SDC2, and SDC4, but not GPC1, bind to laminin β1 chain peptide (98). Similarly, SDC2 and SDC4, but not GPC1, bind to the α3 LG4/5 domain peptide (99), and SDC2 promotes laminin assembly in zebrafish (88). Furthermore, SDC1 binds to the LG4/5 domain of laminin α3 chain (72, 100, 101) and to the N-terminal LE domain of laminin γ2 chain (102). These findings suggest that both epithelial SDC1 and mesenchymal SDC2 and SDC4 might mediate cell attachment to laminin and coordinate with other laminin receptors to assemble laminin networks in the basement membrane.

Elastin is a major fibrillar ECM component with key structural and biological functions in elastic tissues where reversible elasticity is critical, such as major arterial blood vessels, lungs, and skin. Elastin maturation involves the assembly of tropoelastin into a highly cross-linked polymer by cell surface receptors and enzymes. Cell surface HSPGs bind to tropoelastin (74), and SDC4 colocalizes with microfibrillar components of the elastic fiber, suggesting a potential role for cell surface HSPGs in elastin assembly (103). Tropoelastin also contains domains in the N-terminal region where cells attach and spread via αv and α5β1 integrins. Interestingly, cell surface HSPGs bind to a peptide sequence in an adjacent domain that contains Lys residues and induce a conformational ordering of the region (104). These data suggest that cell surface HSPGs might mediate the initial attachment of cells to tropoelastin, which is followed by αv and α5β1 integrin-dependent binding and signaling (104). Whether these HSPG and integrin interactions contribute to elastic fiber assembly is not known.

INTERACTION WITH PATHOGENS

Microbial pathogens interact with cell surface HSPGs in a manner similar to mammalian cells, except here, the objective of the interaction is to promote pathogenesis and survival in the host environment. A large number of viruses, bacteria, parasites, and fungi bind to the HS moiety of HSPGs (7, 105–107). Because there are many HS-binding pathogens, some have questioned the significance and relevance of the microbial interactions and suggested that the ability of pathogens to interact with HSPGs may be a result of cell culture adaptation (106, 108, 109). However, a counterargument is that the ability to exploit a ubiquitous molecule like cell surface HSPGs as attachment and invasion coreceptors is likely to expand cellular targets for infection, strengthen specific interactions of microbes with their primary receptors, and give a significant survival advantage to microbes in the host environment. Furthermore, HSPG-pathogen studies are potentially significant because coreceptor-based therapeutic strategies have the benefit of targeting invariant host determinants, which in principle, should not be vulnerable to the development of resistance from a rapidly mutating pathogen population.

Interaction with cell surface HSPGs is most prominent among intracellular pathogens. For example, several serotypes of adenovirus use cell surface HSPGs as the low affinity, initial binding site to promote its interaction with internalization receptors (110). HIV uses SDC3 as an attachment receptor that cooperates with DC-SIGN and CD4 to facilitate infection of leukocytes that express low levels of CD4 (111). HIV binds to the HS moiety of SDC3 via gp120 and SDC3-bound HIV can infect T cells in trans (111). Ebola virus uses T-cell immunoglobulin and mucin domain 1 (TIM-1) and several other receptors for its internalization (112), but its entry into polarized epithelial cells is inhibited by removal of cell surface HSPGs (113). Interestingly, Ebola virus strictly infects polarized epithelial cells via the basolateral route, suggesting that SDC1, abundantly expressed in these cellular compartments, is the primary coreceptor for Ebola virus.

The coronavirus disease 2019 (COVID-19) pandemic has infected more than 470 million people, killing more than 6.1 million people worldwide as of March 2022. The type I transmembrane angiotensin-converting enzyme 2 (ACE2) is the entry receptor for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of COVID-19. Several studies indicate that cell surface HSPGs function as coreceptors for SARS-CoV-2 infection (114, 115). Spike (S) glycoprotein binds to HS and ACE2 and forms a ternary complex. HP binds at long, positively charged patches on the S glycoprotein, thereby masking basic residues of both the receptor-binding domain (RBD) and the multifunctional S1/S2 site (116). HP enhances the open conformation of the RBD in S protein, suggesting that cell surface HSPGs enhance S protein binding to ACE2 (114). Cell surface HSPGs also promote SARS-CoV-2 infection in neighboring cells. Although dendritic cells (DCs) are not infected by SARS-CoV-2, DCs efficiently bind SARS-CoV-2 via HSPGs and distribute the virus to surrounding ACE2-positive cells (115).

Neutralizing antibodies against SARS-CoV-2 isolated from patients with COVID-19 inhibit SARS-CoV-2 binding to HSPGs, suggesting that targeting viral interactions with cell surface HSPGs is a viable therapeutic option. Indeed, unfractionated HP, nonanticoagulant HP, low-molecular-weight HP, 6-O-desulfated HP, and heparinases all block S protein binding and cellular infection by SARS-CoV-2 virus (114, 115, 117). HP not only interferes with SARS-CoV-2 infection by blocking S protein binding but also S protein cleavage by furin (116). Consistent with these data, several randomized clinical trials and retrospective cohort studies have demonstrated beneficial effects of HP and low-molecular-weight HP therapy in COVID-19 (118–121). Although therapeutic doses of HP do not appear to have beneficial effects on survival or cardiovascular or respiratory organ injury in critically ill patients with COVID-19 (122), HP therapy increases the probability of survival to hospital discharge with reduced use of cardiovascular or respiratory organ support in noncritically ill patients (121). These data are consistent with the findings that cell surface HSPGs promote the initial attachment and invasion of SARS-CoV-2 virus, and not the progression of COVID-19, which is primarily mediated by a dysregulated inflammatory response (123, 124).

Bacterial pathogens also exploit cell surface HSPGs as coreceptors for their attachment and invasion. Cell adhesion receptors, such as integrins, cadherins, and immunoglobulin-related cell adhesion molecules are frequently recognized by specific bacterial surface proteins. Binding of bacterial pathogens to these receptors triggers a signaling cascade that leads to bacterial internalization (125). Cell surface HSPGs also frequently interact either directly or indirectly with these cell adhesion receptors (7, 8), suggesting that HSPGs are located at sites where they can function as coreceptors for bacterial pathogens. Indeed, Borrelia burgdorferi, the causative agent of Lyme disease, attaches to cell surface HSPGs and αvβ3 and α5β1 integrins on host cells (126, 127).

Similarly, available data suggest that Listeria monocytogenes, the causative agent of listeriosis, coopts cell surface HSPGs as coreceptors for its attachment and invasion. First, internalin A (InlA) mediates L. monocytogenes entry into intestinal epithelial cells by binding to E-cadherin (128), and E-cadherin expression is associated with that of SDC1 (129). Second, L. monocytogenes internalin B (InlB) also binds to HSPGs (130–132). InlB is a virulence factor that binds to c-Met (133) and C1qR (134). C1qR primarily mediates the uptake of L. monocytogenes in professional phagocytes, whereas Met mediates bacterial internalization in nonphagocytic cells, such as hepatocytes, via clathrin-dependent endocytosis (135). Importantly, InlB binds tightly to HP (132), and HP can potentiate Met activation by InlB by 10-fold (130, 131). Structural studies indicate that the C-terminal region of InlB induces HP-mediated Met clustering, which is required for full Met activation (136). Together, these observations suggest that HSPG coreceptors promote InlA and InlB interactions during L. monocytogenes attachment and internalization.

FORMATION OF LIGAND/RECEPTOR/HSPG COMPLEXES

Cell surface HSPGs can facilitate receptor-ligand interactions by binding both ligand and receptor and forming a ternary ligand/receptor/HSPG complex (Fig. 1C). For example, HSPGs can bind to both FGF ligands and FGF receptors and promote signaling (137). In the Slit-Robo system, GPC1 and SDC bind to both Slit (ligand) and Robo (receptor) via their HS moiety to form a ternary Slit/Robo/HSPG complex and regulate axon guidance (138–140). In B lymphoma cells, SDC1 expression levels are associated with an increased capacity to migrate in response to HIV Tat and to promote lymphoma pathogenesis in AIDS (141). Here, the formation of a receptor complex where SDC1 interacts with Tat via its HS, with αvβ3 integrin via its core protein, and with pp60src via its cytoplasmic domain switches the cellular response from a CXCR4/G-protein/Rac pathway to one that activates FAK and stimulates tumor cell migration.

Several GPCs regulate the Hh signaling pathway by binding to both Hh ligands and their receptors. GPC1 functions as a coreceptor for Shh in commissural neurons where it regulates repulsive guidance cues mediated by Shh signaling (142). GPC5 also enhances Shh signaling in rhabdomyosarcoma cells to promote tumor cell proliferation by binding to both Shh and its signaling receptor Patched (143). GPC5 binds to Shh and Patched in a HS-dependent manner through highly sulfated HS domains containing a critical 2-O-sulfate motif (144). Drosophila Dlp stimulates Hh signaling and interacts with both Hh and Patched during wing development (145). GPC6 also stimulates Hh signaling by binding to both Hh and Patched, and this coreceptor activity is important for the growth of long bones and intestinal elongation during development (39, 146). Furthermore, GPC3 binds to Shh, but in an HS-independent manner (15, 16). The reason why GPC3 binds to Hh via its core protein and not HS is incompletely understood, but the low level of HS sulfation in GPC3 compared with other GPCs is thought to play a role.

Betaglycan (TGFβRIII) is able to bind with high affinity to inhibin, bone morphogenetic proteins (BMPs), and all three isoforms of TGFβ. Betaglycan does not have a kinase domain and its primary function is to facilitate binding of TGFβ ligands to the TGFβRI & II complex and to enhance SMAD2/3 signaling. Betaglycan null mice are embryonic lethal because of severe defects in various organs, including the liver, heart, kidney, thymus, and testis (147–150), underscoring the importance of betaglycan’s coreceptor activities in development. Coreceptor functions of betaglycan are particularly important for TGFβ2 because TGFβ2 alone binds very weakly to TGFβRII compared with TGFβ1 and TGFβ3 (151). Consistent with these observations, TGFβ2 null mice exhibit many of the same phenotypes seen in betaglycan null mice (152).

OLIGOMERIZATION OF LIGANDS

Because mature HS chains contain multiple bindings sites for HS/HP-binding ligands, HSPGs can promote receptor-ligand interactions by serving as a platform for the oligomerization of ligands (Fig. 1D). For example, HSPG binding induces HGF oligomerization into a more stable and active form, which facilitates c-Met receptor dimerization and activation (153). Similarly, HS chains of SDCs and GPCs promote the oligomerization of a proliferation-inducing ligand (APRIL), a TNF superfamily member cytokine. Oligomerization of APRIL is a prerequisite for activating APRIL’s signaling receptors, transmembrane activator and CAML interactor (TACI), and B cell maturation antigen (BCMA), on B and T cells, and to enhance lymphocyte activation, migration, and survival (154). Chemokine oligomerization and binding to HS chains are required for their ability to stimulate cell migration (155). The propensity to bind to cell surface HSPGs and to form higher order oligomers are essential for the in vivo activity of several chemokines (156). Although most chemokines form homodimers on HSPGs, some can form heterodimers with other chemokines (157). Interestingly, viral chemokines, whose function is to misdirect leukocyte recruitment to evade immune responses, also bind and oligomerize on HSPG surfaces (158).

INTERACTION WITH RECEPTORS

Cell surface HSPGs can enhance signaling by interacting directly with receptors and regulating their location, stability, oligomerization state, and activity (Fig. 2A). In Drosophila, Sdc and Dlp have different effects on synapse morphogenesis and function (159). Both Sdc and Dlp bind with high affinity to the protein tyrosine phosphatase leukocyte antigen-related tyrosine phosphatase (LAR), a receptor that controls both neuromuscular junction (NMJ) growth and active zone morphogenesis. In NMJs, the majority of Sdc is found either on the muscle cell surface or in the space between the presynaptic membrane and the subsynaptic reticulum, whereas Dlp is found in a subset of restricted synaptic space. Here they both bind to LAR, but Sdc binding promotes LAR function and NMJ growth, whereas Dlp inhibits LAR function and influence active zone structure and physiology. How this is accomplished is not fully understood, but these findings suggest that cell surface HSPGs can regulate receptor activity by specifying its location.

GPC3 expression is significantly increased in HCC and is considered a biomarker for HCC diagnosis and prognosis (160). GPC3 promotes hepatoma cell proliferation for the progression of HCC, mainly by activating the canonical Wnt pathway (41) and increasing the expression of c-Myc, a classical downstream target of the canonical Wnt pathway (161). GPC3 functions in this manner by interacting directly with Frizzled G-protein coupled receptors (GPCRs) to facilitate the recruitment of Wnt into the signaling complex at the cell membrane and later translocate into the cytoplasm via endocytosis (40). Expression of soluble GPC3 inhibits the canonical Wnt pathway (162), suggesting that shed GPC3 interferes with GPC3-Frizzled interactions at the cell surface. SDC3 also enhances Wnt signaling in osteoblasts by interacting directly with Frizzled 1 (44). This interaction is apparently important in bone formation because Sdc3^−/− mice exhibit low bone volume, reduced bone formation, increased bone marrow adipose tissue, increased bone fragility, and a blunted anabolic bone formation response to mechanical loading. These phenotypes were traced to reduced Wnt signaling in osteoblasts, which led to delayed osteoblast maturation, impaired osteoblast function, and increased osteoclast-mediated bone resorption.

Cell surface HSPGS also enhances signaling by facilitating the formation of receptor complexes. In bone morphogenetic protein (BMP) signaling, cell surface HSPGs do not affect BMP binding to type I receptor subunits but instead, enhance the recruitment of type II receptor subunits to BMP-type I receptor complexes to enhance BMP signaling (163). Cell surface HSPGs also directly regulate signaling by receptor for advanced glycation endproducts (RAGE), one of the receptors for HMGB1 that mediates vascular inflammation. Here, HSPGs bind to RAGE and facilitate the oligomerization of RAGE into hexamers, which is believed to play an essential role in signal transduction (164).

HSPGs AS IN TRANS CORECEPTORS

Cell surface HSPGs can also present ligands to receptors on neighboring cells thereby functioning as in trans coreceptors (Fig. 2B). The development of blood capillary structures from ES cells is strictly dependent on signaling through VEGFR-2. Here, VEGF signaling in endothelial cells is stimulated by cell surface HSPGs expressed in trans in adjacent perivascular smooth muscle cells (165), indicating an HSPG-mediated cross talk between adjacent cells. A similar transactivation mechanism is also seen during Xenopus development. In left-right patterning during early gastrulation, ectodermal SDC2 presents Vg 1, an essential TGFβ family member in Xenopus development, to mesodermal cells (166). These mechanisms demonstrate that cell surface HSPGs have a broader coreceptor activity by functioning in trans to modulate receptor-ligand interactions.

HIV also exploits HSPGs as trans coreceptors (167). All four SDCs are capable of attaching HIV to the cell surface when expressed in nonpermissive, CD4-negative cells. HIV binds to 6-O-sulfated motifs in SDC HS via gp120 (168). Because Arg298 in gp120 is critical for gp120 binding to both SDCs and the chemokine receptor CCR5, it is thought that 6-O-sulfated motifs in SDC HS mimic sulfated Tyr residues in the N-terminus of CCR5 and that HIV gp120 recognizes these similar structures (168). Importantly, SDC-bound HIV retains its infectivity for 6 days, whereas unbound virus loses its infectivity in less than a day, and HIV bound on SDCs efficiently infects CD4-positive T cells (167). Vascular endothelial cells are thought to present SDC-bound HIV to CD4-positive T cells in vivo (167). Because vascular endothelial cells express SDC2 and SDC4, these SDCs may be the primary HSPGs functioning as in trans coreceptors for HIV transmission. Together, it appears as though HIV has maximized its exploitation of SDC coreceptors. HIV uses SDCs as coreceptors for its virulence factor Tat (58–60), as in cis coreceptors for binding to macrophages and DCs (111), and as in trans coreceptors for infection of T cells (167).

INTERACTION WITH INTRACELLULAR MOLECULES

Cell surface HSPGs with cytoplasmic domains can enhance receptor signaling through regulation of intracellular signaling molecules (Fig. 2C). SDCs, for example, has a short cytoplasmic domain that contains two conserved regions (C1 and C2) separated by a variable (V) region unique to each SDC. The SDC cytoplasmic domain interacts with quite a large number of intracellular mediators, including ERM (ezrin, radixin, moesin) proteins, c-SRC, cortactin, tubulin, PKCα, and several PDZ (PSD95, Dlg, ZO1) binding proteins, among others (7, 8). Apart from GPCs, all transmembrane HSPGs contain a PDZ binding motif in their cytoplasmic domain (8), suggesting that PDZ interactions may mediate common functions of transmembrane HSPG coreceptors.

An elegant study showed that the interaction of SDC1 with the PDZ binding protein calcium/calmodulin-dependent serine kinase (CASK) is critical in assembling cholinergic neuromuscular junctions in C. elegans (169). Here, the anterograde organizer Punctin is secreted by cholinergic motor neurons in the synaptic cleft and activates postsynaptic differentiation. Punctin binds to muscle cell SDC1 and localizes SDC1 at postsynaptic sites. Punctin also concentrates and activates another transmembrane receptor, un-coordinated 40 (UNC-40)/DCC, and SDC1 acts as a coreceptor to strengthen the binding of Punctin to UNC-40 at synaptic sites. The clustering of SDC1 and UNC-40 increases the local concentration of PDZ domain-containing cytoplasmic domains of these transmembrane receptors and stimulates the intracellular recruitment of CASK and FRM-3/FARP by direct interaction with the PDZ domains of SDC1 and UNC-40, respectively. The resulting CASK-FRM-3/FARP complex then triggers synaptic clustering of nicotine-sensitive acetylcholine receptors. SDC-CASK interactions are also seen in neuronal synapses where SDC2-CASK complexes are enriched in postsynaptic density fractions of rat brain (170). These studies suggest the importance of SDC-CASK in organizing postsynaptic structures of neuronal and neuromuscular synapses. Interestingly, neurexins, major synaptic organizing proteins, are transmembrane HSPGs with PDZ binding motifs in their cytoplasmic domains. Presynaptic neurexin binds to neuroligin and LRRTM2 in postsynaptic neurons and presents pleiotrophin to LRRTM2 (11). These HS-mediated neurexin interactions are required for presynaptic differentiation, synapse development, and survival in mice (11). Furthermore, mutations in HS biosynthetic enzymes have been implicated in defects in synaptic functions and autism susceptibility (171, 172). Together, these findings suggest that HSPG coreceptor functions at the cell surface and in intracellular compartments have a strong impact on neuronal and synaptic biology.

Among cell surface HSPGs, SDC4 plays a prominent role in regulating the formation of focal adhesions (70, 173). Cells form focal adhesions when adhering to fibronectin via α5β1 integrins. This process requires support from the SDC4 coreceptor and its ability to recruit PKCα to the signaling complex and to activate this Ser/Thr kinase. Inhibition of PKCα impairs focal adhesion formation and cell migration mediated by α5β1 integrin, demonstrating the importance of SDC4-mediated PKCα activation. Furthermore, the α5β1 integrin and SDC4 complex in focal adhesions coordinate cell migration by providing separate signals that contribute to the suppression of RhoA activity during fibronectin engagement (174). Activation of the guanosine triphosphatase-activating protein p190RhoGAP is critical because it inhibits RhoA in an SRC-dependent manner. Fibronectin engagement stimulates Tyr-phosphorylation of p190RhoGAP, and this process is solely mediated by α5β1 integrin. SDC4 mediates the redistribution of Tyr-phosphorylated pool of p190RhoGAP between membrane and cytosolic fractions through activation of PKCα. By doing so, RhoA is effectively suppressed and focal adhesion formation and cell migration are promoted.

SDC4-mediated PKCα recruitment and activation is also crucial in the formation of focal adhesions when cells attach to immobilized ADAM12 (80). In contrast to cell adhesion on fibronectin, SDC4 binding to ADAM12 triggers β1 integrin-dependent cell spreading, stress fiber assembly, and focal adhesion formation. Pharmacological inhibition of PKCα completely inhibits ADAM12-induced cell spreading and a mutant form of SDC4 deficient in the capacity to activate PKCα is unable to promote cell spreading, indicating the absolute requirement for SDC4-mediated PKCα recruitment and activation in this process. Furthermore, SDC2 is unable to promote cell spreading, suggesting that the capacity to promote β1 integrin-dependent cell spreading and to form focal adhesions is specific to SDC4. However, some integrins can form focal adhesions without coreceptor support from SDC4 and also without PKCα activation (70), suggesting that coreceptor functions of SDC4 might be important in select cell types and in select cellular microenvironments.

INTRACELLULAR TRAFFICKING OF LIGANDS

Cell surface HSPGs can regulate receptor-ligand interactions through endocytosis of ligands thereby controlling the availability of ligands at the cell surface (Fig. 2D). In vascular smooth muscle cells, the presence of cell surface HSPGs extends the half-life of intracellular FGF2 and prolongs the availability of an intracellular pool of bioactive FGF-2 (175). By doing so, HSPGs are thought to function as a cellular switch between immediate and prolonged FGF-2 signaling in vascular smooth muscle cells, which might be important for the regulation of angiogenesis.

Loss of function mutations in GPC3 causes Simpson–Golabi–Behmel overgrowth syndrome (176). Corroborating the clinical features of this syndrome, GPC3 null mice display developmental overgrowth (177). GPC3 null embryos display increased Hh signaling, which positively regulates body size during development. GPC3 inhibits Hh activity in cultured MEFs by inducing Hh endocytosis and degradation, thereby limiting the availability of Hh at the cell surface (15).

Neuropilin-1 promotes α5β1 integrin-dependent endothelial cell adhesion to fibronectin and this interaction is crucial for vascular development (178). This mechanism is dependent on the SEA motif in neuropilin-1’s cytoplasmic domain. Neuropilin-1 interacts with α5β1 integrin at adhesion sites but does not directly mediate cell spreading on fibronectin. Instead, neuropilin-1 stimulates endothelial cell spreading by increasing the Rab5/GIPC1/Myo6-dependent internalization of active α5β1 integrin. The absence of this mechanism is thought to be causal in the angiogenesis defects observed in neuropilin-1 null mice.

FORMATION OF LIGAND GRADIENTS

HSPGs facilitate the formation of morphogen gradients during development (9, 179–182), and chemokine gradients during an inflammatory response to infection or tissue injury (183, 184). Cell surface HSPGs present morphogens and chemokines in trans to recipient cells (Fig. 2E). Morphogen gradients play crucial roles in the patterning and morphogenesis of the early developing embryo. Concentration gradients of morphogens are critical for establishing positional identities of differentiating cells. Several mechanisms of morphogen gradient formation have been proposed, such as restricted diffusion (185), spread of soluble morphogens through a continuous extracellular matrix (186), exosome transport (187), morphogen clustering (188), morphogen shuttling (189), and active transport of morphogens through an actin-based, filopodia-like cellular protrusion called cytoneme (190). Cell surface HSPGs are considered critical mediators in most of these mechanisms. For example, Wg does not move across a strip of cells lacking HS (191). Instead, Wg moves through the restricted diffusion mechanism by binding to cell surface HSPGs via Dally and Dlp (191). Cell surface HSPGs also control FGF gradients by regulating their diffusion in tissues. In midbrain development, binding to HSPGs is essential in the formation of a long range, posterior to anterior FGF8 gradient spanning the midbrain (192). In another study, tracking of FGF2 movement in living cells showed that the spatial organization of HS and the availability of FGF2 binding sites in HS can promote either FGF2 confinement or substantial movement (193). Other members of the FGF family are also bound by cell surface HSPGs. FGF4 and FGF8 are tethered to the cell surface by SDC1 and these interactions are critical for their local retention in the extraembryonic ectoderm (194). The extent of HS sulfation apparently regulates FGF morphogen gradients. For example, regional patterns of HS O-sulfation control FGF10 gradients formed during branching morphogenesis in lung development (195). HSPGs in the lung mesenchyme at sites of prospective budding express low sulfated HS, whereas highly sulfated HS chains are present in basement membranes of branching epithelial tubes (195).

HSPGs, especially those of the GPC family, can also facilitate intercellular transport of morphogens or intercellular communication between germ layers during development in morphogen gradient formation. In Drosophila wing development, Dally and Dlp control the movement of Hh across receiving cells in a functionally redundant manner (196). Interestingly, in regions of active Hh signaling in the Drosophila wing disk and abdominal epidermis, cells generate cytonemes (190). Cytoneme formation associates with Hh gradient establishment in space and time, and conditions that interfere with cytoneme formation reduce Hh gradient length, suggesting that cytonemes are critical for Hh gradient formation in wing and abdominal epithelia. Although the role of cell surface HSPGs in these processes is incompletely understood, cytonemes are thought to transport morphogens in exosomes (197) and SDC1 facilitates exosome biogenesis (198), suggesting that SDC1 might indirectly regulate morphogen gradient formation through its effects on morphogen packaging in cytoneme exosomes. More importantly, GPC4 aids in the formation of cytonemes that transport Wnt from endoderm to mesoderm in zebrafish embryos (199). The exact function of GPC4 in cytoneme formation remains to be defined, but without GPC4, the cytonemes are too short and too few to effectively transport Wnts from cell to cell.

Chemokines are chemotactic cytokines that comprise the most prominent group of migration guidance cues that operate in vertebrates (200, 201). Chemokines are distinguished from other cytokines by their ability to act on GPCRs. Chemokines diffuse from their site of production, and their interactions between their nonreceptor binding domains and cell surface and matrix HSPGs along the way are thought to drive the formation of linear immobilized (haptotactic) gradients that provide directional guidance for leukocyte migration (183, 184). Without this mechanism, chemokines will disperse and result in broad, nonlinear, and soluble gradients, which are less robust and stable, and rather transient in nature compared with immobilized haptotactic gradients. For many chemokines, their HS binding mutants that retain chemotactic activity in vitro have significantly reduced the capacity to induce leukocyte migration in vivo (156, 202), demonstrating that tethering by HSPGs is essential for chemokine activity. However, the chemokine gradient concept is somewhat oversimplified in several aspects. The role of HSPGs in the chemokine gradient concept largely assumes a similar level of expression, similar chemokine binding avidity, and similar effects on chemokine activity by HSPGs expressed in different tissue compartments. These assumptions are likely not true. Furthermore, except for few examples, the identity of cell surface HSPGs that tether chemokines and form chemokine gradients is mostly unknown. Moreover, HSPGs themselves are highly regulated during an inflammatory response, but the impact of these mechanisms on the regulation of chemokine gradients has not yet been rigorously examined.

Concluding Remarks

Studies during the last 30 years have clearly established the role of cell surface HSPGs as important coreceptors that facilitate receptor-ligand interactions and subsequent signaling. We mostly restricted our discussion to stimulatory functions of HSPG coreceptors, but there are many examples where cell surface HSPGs function as inhibitory coreceptors. For example, cell surface HSPGs control the basal activation of macrophages by type I interferons in atherosclerosis and obesity by sequestering IFNβ and inhibiting interferon-α/β receptor (IFNAR) signaling (203). Similarly, GPC3 suppresses Hh signaling by binding Hh and mediating Hh endocytosis and degradation (15, 16). Also, it is important to note that some transmembrane cell surface HSPGs possess the ability to directly signal, and in certain cases it is difficult to distinguish whether the primary function of HSPGs is that of a coreceptor or a signaling receptor. Furthermore, several microbial pathogens and virulence factors have been reported to use cell surface HSPGs as primary receptors (204–207). In addition, while beyond the scope of this review, cell surface HSPGs can regulate receptor-ligand interactions when shed as soluble ectodomains (107). Ectodomain shedding rapidly reduces the amount of cell surface HSPG coreceptors and releases soluble HSPG ectodomains that can regulate ligand binding in an autocrine and paracrine manner (7, 10).

Although mechanisms of cell surface HSPG coreceptors have been extensively studied in vitro and in vivo in lower organisms, comparatively little is known about their contribution in vivo in mammalian organisms. In principle, HSPG coreceptors are likely important in vivo, but not essential, since their primary function is to fine-tune receptor-ligand interactions and cellular responses. In other words, they primarily control the efficiency of an interaction and not the occurrence. Furthermore, even if a particular HSPG coreceptor function is essential, the redundancy of HSPGs may compensate for the loss of a particular HSPG and mask its importance in vivo. Consistent with these ideas, knockout (KO) mice with constitutive deletion of genes for Sdc1, Sdc2, Sdc3, Sdc4, Gpc1, Gpc2, Gpc4, and Cd44 and systemic knockdown of Gpc5 are not lethal in mice (8, 107, 208–212), and the majority of these KO mice show relatively minor phenotypes under unchallenged conditions. On the other hand, deletion of genes for neuropilin-1 (Nrp1), betaglycan (Tgfbr3), neurexin (Nrxn1-3), and Gpc6 are embryonic lethal (147, 213–215) and deletion of Gpc3 is perinatal lethal (177), indicating that these cell surface HSPGs are essential for development. Furthermore, adult Sdc1 and Cd44 KO mice show significant pathological phenotypes when challenged with inflammatory, infectious, and oncogenic agents or conditions (7, 8, 107, 216–218), suggesting that certain postdevelopmental functions of these cell surface HSPGs are specific and essential.

Obviously, cell surface HSPGs have other functions, and it is unclear if these developmental and postdevelopmental phenotypes of HSPG KO mice are solely due to their coreceptor activities. The majority of coreceptor functions of cell surface HSPGs might only be significant and relevant in select cell types and in specific steps of developmental and postdevelopmental processes. Nonetheless, studies so far suggest that coreceptor functions of cell surface HSPGs are critical in providing support to various key steps of a receptor-ligand interaction. Absence of this activity appears to tip the balance toward developmental abnormalities, disease onset and progression, and in the worst case lethality. On the other hand, a large number of pathogens apparently evolved or adapted to exploit coreceptor functions of HSPGs for their pathogenesis. These findings suggest that precise manipulations of HSPG coreceptor functions might have therapeutic value in many hereditary and acquired diseases.

GRANTS

This study is supported by the National Institutes of Health (NIH) Grants R01 HL142213, R01 HL132573, and R21 AI156284.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

This article is part of the special collection “Deciphering the Role of Proteoglycans and Glycosaminoglycans in Health and Disease.” Liliana Schaefer, MD, served as Guest Editor of this collection.

AUTHOR CONTRIBUTIONS

K.H. and P.W.P. conceived and designed research; K.H. performed experiments; K.H. and P.W.P. analyzed data; K.H. and P.W.P. interpreted results of experiments; K.H. and P.W.P. prepared figures; K.H., R.S.A., and P.W.P. drafted manuscript; R.S.A. and P.W.P. edited and revised manuscript; K.H., R.S.A., and P.W.P. approved final version of manuscript.