Evolving Mechanistic Insights into Galectin Functions

Abstract

Galectins are an evolutionarily ancient family of glycan-binding proteins (GBPs) and are found in all animals. Although they were discovered over 30 years ago, ideas about their biological functions continue to evolve. Current evidence indicates that galectins, which are the only known GBPs that occur free in the cytoplasm and extracellularly, are involved in a variety of intracellular and extracellular pathways contributing to homeostasis, cellular turnover, cell adhesion, and immunity. Here we review evolving insights into galectin biology from a historical perspective and explore current evidence regarding biological roles of galectins.

1 Introduction

Although the importance of cell surface carbohydrates in normal cellular physiology remained elusive for many years, recent studies demonstrate that these highly complex macromolecules possess diverse roles in many cellular processes [1]. In addition to directly impacting glycoprotein function, complex carbohydrate structures serve as ligands for glycan binding proteins (GBPs), which enable cell surface glycan-dependent signaling [2–4]. Recent studies demonstrate that GBP–cell surface carbohydrate interactions play key roles in a variety of processes ranging from cellular turnover to innate immunity. Given the plasticity of immune cells, which not only enables differentiation of cellular function but also results in distinct alterations in cell surface glycosylation [5], GBP–glycan interactions appear to be especially important in the regulation of immunity.

Among GBPs with immunoregulatory activities, galectin family members appear to regulate a wide variety of immunological processes in addition to serving as factors directly involved in microbial killing [6–9]. In this review, we attempt to link original descriptions concerning the immunological activities of galectins to new mechanistic insights into their regulatory activities within immunity and beyond. In doing so, we do not seek to provide an exhaustive review regarding the biological activities of the entire galectin family, especially given the rapidly expanding nature of this field. However, we seek to highlight salient features of galectins that illustrate the broad regulatory capacity of likely one of the most pleiotropic protein families described.

2 Discovery of Galectins

Galectins were identified within a few years after the pioneering work in the 1960s of Ashwell and Morrell who discovered the asialoglycoprotein receptor, the first GBP described in vertebrates [10, 11]. Although some of the physiological functions of that receptor have only recently been elucidated [2], these early studies suggested the potential existence of other mammalian GBPs. In 1975, the first vertebrate galectin (an ortholog of mammalian galectin-1) was discovered in the electric organ of the electric eel Electrophorus electricus by Teichberg and colleagues [12]. This was followed shortly thereafter by isolation of galectins in the laboratories of Barondes and Kornfeld, who reported similar proteins in avian and mammalian sources, respectively [12–14]. There are now 15 described members of the galectin family (Gal-1 through −15) encoded in mammals (11 are known in humans), and all range in subunit size from 14 to 39 kDa [15] (Fig. 1). They are found in all metazoans and are the only known GBPs in animals that occur both in the cytoplasm, where they are synthesized on free polyribosomes, and on the plasma membrane and extracellular matrix [15].

Fig. 1.

The galectin family of β-galactoside binding proteins. Galectins are classified into three distinct groups based on their quaternary structure: prototypical, chimeric, and tandem repeat. Prototypical: Gal-1, Gal-2, Gal-7, Gal-10, Gal-13, and Gal-14. Chimeric: Gal-3. Tandem repeat: Gal-4, Gal-8, Gal-9, and Gal-12

3 Galectin-1: Regulator of Adaptive Immunity

Although the discovery of this ancient galectin family of GBPs was important in confirming the existence of multiple mammalian GBPs, the physiological functions of galectins were more difficult to define. Early studies suggested that Gal-1 might regulate the development of muscle, the first mammalian organ from which the protein was isolated [14], including maintenance of the neuromuscular junction [16–18]. Teichberg and colleagues examined whether administration of Gal-1 might affect the pathological sequelae associated with neuromuscular junction pathology. They employed an animal model of myasthenia gravis induced by autoantibody formation against the acetylcholine receptor. Consistent with the potential involvement of Gal-1 in the neuromuscular junction, exogenously added Gal-1 appeared to cause a significant enhancement of muscle function. However, the apparent amelioration of disease actually resulted from the ability of Gal-1 to suppress the autoimmunity needed to generate a myasthenia gravis model [16–18]. Thus, an indirect result of these experiments was the first evidence for what continues to be one of the more intriguing properties of Gal-1, namely, its ability to significantly suppress immune function [7, 12].

4 Gal-1 Regulation of T Cells

While several studies suggested that Gal-1 might regulate adaptive immunity [6, 7, 19–25], it was not until nearly a decade later that studies began to provide a mechanistic insight into Gal-1-mediated immunosuppression. Early studies suggested that Gal-1 may actually induce lymphocyte proliferation, which suggested to the authors that Gal-1 may enhance the development of suppressor T cells [7]. Gal-1 also appeared to mediate adhesion of thymocytes to epithelial cells, which supported a possible role for Gal-1 in the regulation of T cell development [26, 27]. However, it was a seminal paper by Baum and colleagues that suggested that Gal-1 might directly impact T cell viability by inducing apoptosis, that provided the most substantial mechanistic insight into the immunomodulatory activities of Gal-1 [28]. That study first reported that Gal-1 could induce apoptosis of primary activated T cells and several T cell lines, including MOLT-4 and ARR cells. Thus, this study suggested that Gal-1 might regulate adaptive immunity through directly inducing apoptotic death of effector T cells [28].

In addition to directly inducing cell death in activated T cells, subsequent studies suggested that Gal-1 might also serve as a key regulator of a variety of other T cell functions (Fig. 2). For example, Gal-1 induces robust IL-10 production in both CD4+ and CD8+ T cells while inhibiting IFN-γ formation, which suggests that Gal-1 may reduce adaptive immune responses by altering T cell cytokine production [29, 30]. Consistent with this, adoptive transfer of CD4+ T cells from Gal-1 treated mice, which displayed similar cytokine profiles observed following in vitro incubation with Gal-1, protected mice from uveitis with the same efficiency as injection of Gal-1 alone [23]. In addition, injection of Gal-1 into IL-10 null mice fails to convey the immunoprotective properties of Gal-1, strongly suggesting a role for IL-10 and possibly other cytokines, in mediating the immunosuppressive activities of this protein [25]. Similarly, Gal-1-Ig chimera constructs can induce significant IL-10 by T cells, providing a useful therapeutic approach to enhancing Gal-1-mediated immunosuppression [31, 32]. Regulatory T cells (Tregs) may also utilize Gal-1 to induce tolerance, as Tregs from Gal-1 null mice exhibit an impaired capacity to suppress T cell activation [33]. Furthermore, Gal-1 itself appears to facilitate the formation of induced Tregs (iTregs) by upregulating FOXP3 expression in peripheral activated T cells, strongly suggesting a key role for Gal-1 in Treg effector function and development [34]. Perturbation of the cytokine milieu by Gal-1 may indirectly impact T cell viability, as distinct T cell populations often rely on specific cytokines to maintain viability in vivo [35, 36].

Fig. 2.

Distinct galectin family members possess the capacity to differentially regulate lymphocytes. Some of the earliest studies suggested that galectins appear to possess immunomodulatory activity by directly impacting T cell viability. Subsequent studies suggested that galectin might not only regulate T cell viability but also regulate cytokine secretion and activation. Several galectins might also play a variety of roles in the development, activation and differentiation of B cells. Similar galectin-mediated regulation may play a role in NK and NKT cell activation and viability. Representative galectin-regulated activities are shown. Red arrows indicate an activity that the respective galectin increases, while blue arrows signify galectin-induced decreases in the accompanying activity

In addition to directly altering T cell cytokine production, several studies suggest that Gal-1 might also inhibit T cell activation, both in vitro and following injection during T cell activation in vivo, providing an additional mechanism whereby Gal-1 might inhibit adaptive immunity [6, 19–24, 37]. Recent studies also demonstrate an in vivo role for Gal-1 in T cell development. TCR transgenic Gal-1 null mice appear to possess altered central selection, which favors the generation of CD8αα T cells, among other alterations in T cell behavior, providing further evidence that Gal-1 affects multiple aspects of T cell biology [38]. In contrast to the ability of Gal-1 to inhibit activation and induce apoptosis of activated T cells, Gal-1 might actually sustain naive T cell survival in peripheral tissue [39]. In addition, recent studies suggest that under certain circumstances Gal-1 may actually facilitate T cell activation in the absence or presence of cognate ligand on antigen presenting cells [40] (Fig. 2).

5 T Cell Regulation: A General Theme of the Galectin Family

In the wake of early studies demonstrating that Gal-1 possessed potent immunoregulatory activities, many studies began to examine the potential role of other galectin family members in the regulation of T cell viability and function (Fig. 2). For example, subsequent studies demonstrated that Gal-3, Gal-8, and Gal-9 regulate thymocyte and T cell viability [41–45]. While different galectin family members appear to signal apoptosis in T cells, individual galectin-mediated regulation of T cell viability likely occurs through engagement of distinct counter receptors. For example, Gal-1 fails to inhibit Gal-3 induced Ca ²⁺ flux or apoptosis in primary activated T cells, although both Gal-1 and Gal-3 recognize their respective receptors with similar affinity [30], strongly suggesting engagement of distinct ligands [46]. Similarly, recent work demonstrated that Gal-9, but not Gal-1 or Gal-3, specifically induces apoptosis in CD4+ TH1 cells through engagement of Tim-3 [42], although Tim-3 fails to relay Gal-9 signaling in several T leukemic cell lines [47].

Similar to Gal-1, other galectin family members appear to regulate other aspects of T cell biology in addition to T cell viability. For example, Gal-3 appears to regulate T cell activation and the induction of T cell anergy by restricting lateral mobilization of the TCR into the immunological synapse or by dissociating CD8 from the TCR, respectively [48, 49]. Similar to Gal-1, Gal-2 induces IL-10 production, along with IL-5 and TGF-β, while also decreasing IFN-γ and IL-2, which can significantly impact T cell activation and differentiation [50]. Gal-10, originally known as Charcot–Leyden protein, appears to also be a key player in proper Treg function [51]. In contrast, several studies suggest that Gal-4 may actually exhibit pro-inflammatory activity, inducing IL-6 production and expansion of CD4+ T cells [52], likely following specific colitis-associated changes in CD4+ T cell glycosylation [53]. Gal-3 may also regulate CD4+ T cell differentiation by reducing IL-5 secretion [54]. Interestingly, in contrast to Gal-8-induced apoptosis in activated T cells [55], similar to Gal-1, recent studies suggest that Gal-8 may likewise enhance T cell proliferation [40] (Fig. 2).

6 Galectin Regulation of B Cells

Galectins also appear to serve as key regulators of B cells (Fig. 2). Similar to the ability of Gal-1-mediated thymocyte engagement of thymic epithelial cells, work by Schiff and colleagues demonstrated that Gal-1 can likewise facilitate developing B cell interactions with key stromal constituents within the bone marrow [56]. Subsequent studies demonstrated that Gal-1 mediates B cell stromal contacts by facilitating interactions between the surrogate light chain, α5β1, α4β1, and α4β7 integrins, and ADAM15/fibronectin [57, 58]. These interactions ultimately result in relocalization of the B cell receptor to the B cell-stromal interface to form an immunological synapse that induces critical signals necessary for B cell development [57, 59]. Stromal cells expressing Gal-1 appear to reflect a subset distinct from those expressing IL-7 [60], suggesting a unique role for Gal-1 in B cell maturation. Consistent with this, Gal-1 KO mice experience impaired pre-BII-cell development following marrow challenge [59].

In addition to facilitating B cell development, galectins may also play an important role in B cell activation, antibody secretion and differentiation into plasma cells. Gal-1 null mice generate reduced antibody levels following immunization with T cell dependent antigens [61]. Although this could in part reflect Gal-1-mediated regulation of T cell activation and differentiation, Gal-1 KO mice also display reduced antibody formation following challenge with T cell-independent antigens [61], strongly suggesting a direct role for Gal-1 in B cell activation and antibody secretion. Consistent with this, ectopic expression of Gal-1 within B cells facilitates B cell activation and antibody secretion in vitro, likely through engagement of cell surface glycans [62]. Endogenous Gal-1 also appears to be significantly upregulated during B cell differentiation into plasma cells, likely through a BLIMP1-dependent mechanism [61, 62], strongly suggesting a role for B cell-derived Gal-1 in plasma cell biology. Indeed, Gal-1 KO plasma cells display enhanced sensitivity to apoptosis [61], suggesting that Gal-1 may in part maintain antibody secretion by sustaining plasma cell survival. In addition to Gal-1, recent studies suggest that Gal-8 may also play a key role in B cell differentiation [63]. Knockdown of Gal-8 in Gal-1 KO B cells results in further inhibition of plasma cell differentiation and antibody production [63]. Additional studies suggest that Gal-3 may also impact B cell differentiation and survival [64].

7 Galectins and Natural Killer (NK)/Natural Killer T (NKT) Cells

Following embryo implantation, unique NK cell subsets within the decidua may utilize Gal-1 as an effector molecule in the maintenance of fetal–maternal tolerance [65, 66]. NK cells in this setting secrete Gal-1 that, in turn, appears to engage infiltrating T cells, inducing their apoptosis [67]. In contrast, Gal-3 appears to serve as a neoplastic-derived inhibitor of NK function by inhibiting NKG2D-mediated NK cell activation [68, 69], providing a potential mechanism of inhibiting antitumor immunity. In contrast to the inhibitory activity of Gal-3 on NK cells, Gal-9 induces NK cell activation through ligation of Tim3, possibly enhancing antiviral immunity in the setting of HIV infection [70]. However, Gal-9 appears to exert the opposite effect on NKT cells [71], by directly inducing apoptotic cell death [72]. Gal-9 may also facilitate NKT cell activation indirectly by inducing IL-15-dependent NKT cell activation by neighboring macrophages [72]. Similar to Gal-1, alterations in Gal-9-mediated NK and NKT cell function may contribute to the pathogenesis of preeclampsia [73] (Fig. 2).

8 Galectins and DCs/Macrophages

Given the regulatory network responsible for lymphocyte activation and the intimate connection between innate immunity and the development of an adaptive immune response [74], it is not surprising that galectin family members also appear to regulate a variety of antigen presenting cells (APCs) (Fig. 3). Early studies suggested that individual galectin family members might induce macrophage chemotaxis, macrophage activation, and monocyte to macrophage differentiation [75–78]. In addition, galectins appear to directly enhance macrophage phagocytosis [79, 80]. While all of these features would be predicted to enhance overall antigen uptake and presentation, more recent studies suggest that several galectin family members may also directly impact the ability of APCs to activate antigen specific T cells. For example, Gal-3 appears to regulate immunological synapse formation by engaging beta1,6 N-acetylglucosaminyltransferase V (Mgat5)-dependent N glycans on various glycoproteins within the immunological synapse [48]. Gal-3-mediated interactions appear to reduce synapse formation and therefore inhibit T cell activation [48]. Consistent with this, Mgat5 KO recipients display heightened T cell activation and an increased propensity to develop autoimmunity [48]. In contrast, Gal-9 appears to enhance T cell activation in the setting of autoimmunity [81]. Similar results suggest that Gal-1 and Gal-8 may enhance T cell activation by facilitating T cell interactions with APCs in the presence of cognate antigen [40].

Fig. 3.

Galectins differentially regulate monocytes/macrophages and dendritic cells. Different galectin family members appear to possess the ability to directly impact the activity of monocytes/macrophages and dendritic cells with significant implications not only on generalized inflammation but also on the ability of these cells to engage cells involved in adaptive immunity. Several galectins also appear to directly regulate T cell-antigen presenting cell interactions with implications on subsequent T cell activation and differentiation. Representative galectin-regulated activities are shown. Red arrows indicate an activity that the respective galectin increases, while blue arrows signify galectin-induced decreases in the accompanying activity

In addition to potentially regulating the formation and stability of the immunological synapse, individual galectins may also differentially impact the immunological outcome of APC-T cell interactions by altering APC cytokine profiles following activation. Incubation of DCs in the presence of Gal-1 results in the formation of tolerogenic DCs that subsequently induce IL-10-expressing T cells following TCR ligation [82]. Transfer of Gal-1-treated DCs similarly results in tolerance induction in a variety of settings. For example, transfer of Gal-1-treated DCs reduces IFNγ and increases IL-10, ultimately resulting in enhanced maintenance of fetal–maternal tolerance following maternal stress [82]. Similarly, Gal-1 reduces cell surface MHC class II levels following IFNγ treatment, yet differentially regulates FcγR1 expression depending on the presence or absence of other cytokines [83]. Gal-9 appears to also possess the capacity to induce tolerogenic APC function. Gal-9 suppresses LPS-induced TNF secretion by macrophages and transfer of DCs from Gal-9-treated mice attenuates acute lung injury [84]. These findings do not appear to be limited to murine models as Gal-1 inhibits DC production of IFNγ in patients with psoriasis [85] (Fig. 3).

In contrast to the ability of Gal-1 and Gal-9 to induce tolerogenic activity in APCs, Gal-3 appears to engage pro-inflammatory programs in various APC populations (Fig. 3). For example, in vitro studies demonstrate that Gal-3 induces macrophage activation and significant TNFα secretion [86]. Consistent with this, Gal-3 KO mice display reduced macrophage activation and inflammatory sequelae associated with liver injury [87]. Similarly, DCs isolated from Gal-3 KO mice immunized with MOG in the setting of experimental autoimmune encephalomyelitis likewise exhibit increased IL-10 production and decreased IFNγ [88]. Importantly, similar changes in DC activation and cytokine secretion can be observed in Gal-3 KO mice in the setting of reperfusion injury, stroke and Con-A-induced hepatitis [20, 89, 90], strongly suggesting that Gal-3 enhances APC activation and pro-inflammatory cytokine secretion in a variety of settings [91]. Although incubation of recombinant Gal-3 can regulate macrophage function, transfer of macrophages isolated from Gal-3 KO mice into WT recipients results in a similar outcome [91], strongly suggesting that Gal-3-mediated regulation can be a cell intrinsic phenomenon. Similar to Gal-3, recent results suggest that in certain settings, Gal-9 may also promote DC maturation with subsequent enhancement of NK cell activation [92].

9 Galectin Regulation of Granulocytes/Mast Cells

The apparent ability of Gal-1 to induce apoptotic cell death in T cells, coupled with growing evidence that Gal-1 may be involved in many aspects of immune regulation, strongly suggested that Gal-1 might be a general regulator of leukocyte activation, trafficking, and viability (Fig. 4). Consistent with this, early studies demonstrated that Gal-1 can induce neutrophil NADPH activation and enhance superoxide production, possibly through interactions with CD66a and CD66b [93, 94]. In addition to facilitating neutrophil activation, different galectin family members also appear to regulate neutrophil extravasation and induce differential effects on chemotaxis, depending on the activation state of the cells [131–134]. Similar to what had been observed in activated T cells [28], several studies suggest that galectins may also aid in neutrophil turnover. For example, recent results suggested that Gal-1 induces exposure of PS, a common ligand responsible for triggering cellular removal in cells undergoing apoptosis, in activated neutrophils and the promyelocytic cell line, HL60 [30, 95, 96]. However, in contrast to the effects of Gal-1 on T cells, induction of PS exposure in activated neutrophils by Gal-1 appeared to occur in the conspicuous absence of cell death [30, 95–97]. Despite the inability of Gal-1 to induce apoptosis in activated neutrophils, Gal-1-induced PS exposure sensitized cells to phagocytic removal [95, 97]. This form of removal may be unique to neutrophils [98] as it prepares cells for removal without inducing cell death, a process recently termed “preaparesis” [30]. As late apoptosis is often accompanied by loss of membrane integrity, this form of cellular turnover may facilitate the maintenance of membrane integrity in the setting of acute inflammation until appropriate phagocytic removal occurs. The ability to induce PS exposure in the absence of apoptosis in neutrophils does not appear to be limited to Gal-1, as subsequent studies demonstrated that Gal-2, Gal-3, Gal-4, and Gal-8 also possess a similar ability to induce apoptosis-independent expression of PS [96, 99, 100], although distinct signaling pathways appear to be engaged [96].

Fig. 4.

Galectins regulate granulocyte activation and turnover. Galectin family members appear to possess the ability to differentially activate various granulocytes, including neutrophils, eosinophils, and mast cells. In addition, galectins appear to regulate the turnover and chemotaxis of various granulocyte populations. Representative galectin-regulated activities are shown. Red arrows indicate an activity that the respective galectin increases, while blue arrows signify galectin-induced decreases in the accompanying activity

In addition to potentially regulating neutrophils, early studies suggested that galectins may also regulate other granulocyte populations. Galectin-10, the first galectin actually described, but originally known as Charcot-Leyden protein [101, 102], accumulates as crystals in vivo under allergic or parasitic conditions associated with hypereosinophilia [102]. Although the exact stimulus and purpose of copious Gal-10 accumulation in this setting remains to be elucidated, these results suggest that Gal-10 may serve a regulatory or effector role in eosinophil function. Originally described as one of several T cell-derived factors capable of regulating eosinophils [103], Gal-9 secreted by activated T cells appears to drive eosinophil chemotaxis [104]. Subsequent studies suggested that Gal-9 may not only facilitate eosinophil migration, but may also regulate extravasation and viability [105–107].

In addition to regulating neutrophils and eosinophils, some of the earliest work examining the potential immunomodulatory activity of galectins involved mast cell biology. Pioneering work by Fu-Tong Liu’s group demonstrated that a unique protein, the epsilon binding protein, recognized IgE. Subsequent studies identified epsilon binding protein as Gal-3 [108, 109]. Gal-3 not only appeared by bind IgE [110] but also recognized the Fcε receptor (FcεR) on mast cells [111], suggesting a general regulatory role for Gal-3 in mast cell function. Indeed, Gal-3 appears to directly engage the FcεR and induce mast cell degranulation [111]. Consistent with this, Gal-3 KO mast cells display significantly impaired degranulation [112]. In addition, recent studies suggest that Gal-3 may actually regulate mast cell viability [113]. In contrast to Gal-3, Gal-1 may induce the opposite effect on mast cells, as injection of Gal-1 appears to inhibit mast cell degranulation following inflammatory challenge in vivo [114] (Fig. 4).

10 Galectins Regulate Hemostasis, Tissue Repair, and Angiogenesis

Given the breadth of activities attributed to individual galectin family members in the regulation of immunity and the intimate association of immunity with coagulation [115], it appears that galectins also evolved important roles in hemostasis (Fig. 5). Several studies demonstrated that Gal-1 directly triggers platelet activation and aggregation in vitro, likely through interactions with αIIβ3 integrins [116]. In addition, Gal-1 appears to enhance platelet sensitivity to activation by other agonists [117]. Although the addition of exogenous Gal-1 can induce platelet activation, co-incubation of platelets with a Gal-1 blocking antibody significantly inhibits ADP-induced platelet aggregation [117], strongly suggesting a role for platelet-derived Gal-1 in platelet activation. Consistent with these in vitro results, Gal-1 KO mice experience increased bleeding time [116], suggesting that Gal-1 plays an important role in hemostasis in vivo. Galectin-mediated platelet activation does not appear to be limited to Gal-1, as Gal-8 similarly activates platelets [118]. However, in contrast to Gal-1, Gal-8 signals through a GP1β-dependent pathway, although a similar downstream Erk pathway becomes activated following engagement by either galectin [116, 118]. In addition to regulating platelet function, Gal-8 appears to regulate Factor V levels by enhancing its endocytic recycling in megakaryocytes [119]. In contrast, Gal-1 and Gal-3 appear to negatively regulate von Willebrand factor (vWF) interactions with platelets through direct interactions with vWF N glycans [120].

Fig. 5.

Galectins regulate hemostasis, angiogenesis and tissue repair. Various members of the galectin family regulate megakaryocyte activity, hemostasis, angiogenesis, epithelial migration, and general tissue repair following injury. Representative galectin-regulated activities are shown. Red arrows indicate an activity that the respective galectin increases, while blue arrows signify galectin-induced decreases in the accompanying activity. Plt = platelet, vWF = von Willebrand Factor

In addition to regulating hemostasis, different galectin family members appear to play a critical role in tissue repair. Work by Panjwani and colleagues demonstrated that Gal-7 in particular plays a critical role in epithelial repair, as Gal-7 induces the cell migration and proliferation necessary for wound closure in an in vitro wound healing model [121, 122] (Fig. 5). Consistent with this, Gal-7 KO mice displayed reduced epithelial repair following wounding in vivo [123]. Gal-3 likewise appears to play a critical role in this process, while Gal-1 failed to exhibit similar activity [121]. In addition to epithelial repair, early studies suggested that Gal-1 might facilitate muscle development and repair by inducing myotube fusion and fibroblast to myocyte transdifferentiation [124, 125]. Gal-1 displays a unique striated localization pattern within muscle and muscle injury results in a significant increase in Gal-1 expression [126, 127]. Gal-1 KO mice exhibit impaired muscle regeneration following BaCl₂-induced muscle injury [124]. Furthermore, knockdown of Gal-1 in zebrafish resulted in impaired muscle formation, suggesting a role in embryonic muscle development [128].

An important component of injury repair is the development of blood vessels, in which endothelial cells play a critical role [129]. Endothelial cells express several galectin family members, including Gal-1, Gal-3, and Gal-9 and activation increases galectin expression [130]. While enhanced galectin expression may regulate endothelial interactions with leukocytes and subsequent leukocyte extravasation [131–134], several galectins appear to directly impact endothelial cell-mediated regulation of angiogenesis [135, 136]. Incubation of endothelial cells with Gal-3 results in microtube development [137], possibly through NG2, a driver of angiogenesis [138], by enhancing NG2 interactions with α3β1 integrins [138]. Several studies suggest that Gal-1 may also facilitate angiogenesis, as knockdown of Gal-1 reduces endothelial migration and expansion. Unlike Gal-3, Gal-1-mediated regulation of angiogenesis appears to occur through interactions with VEGFR1 and VEGFR2 [139]. Disruption of this interaction and subsequent loss of VEGFR2 signaling likely contributes to the pathogenesis of preeclampsia, as Gal-1 KO mice exhibit lower levels of desidual vascularization and an increased frequency of preeclampsia [140]. Consistent with this, individuals with preeclampsia display reduced levels of Gal-1 expression [140]. Similar studies suggested that Gal-8 might induce endothelial cell migration and angiogenesis through interactions with CD166 [141]. In contrast, a recent study suggested that Gal-1 might engage CD146 to induce endothelial cell apoptosis [142].

11 Intracellular Roles of Galectins

Although the extracellular carbohydrate binding activities of galectins became their defining feature [143], galectin synthesis occurs within the cytosol, where individual galectin family members often accumulate prior to export through a noncanonical Golgi-independent pathway [144–146]. Given their location within the cytosol, several studies have implicated various galectin family members as key players in the regulation of a variety of intracellular processes [147] (Fig. 6). For example, Gal-3 binds GSK-β and β-catenin and can facilitate GSK-β-mediated signaling and transcriptional upregulation of downstream targets, such as fascin-1 [148, 149]. Phosphorylation of Gal-3 appears to modulate Gal-3-mediated regulation of cellular signaling and survival, possibly through altering potential interactions with K-Ras-GTP [150, 151]. Additional studies suggest that similar intracellular phosphorylation events may also regulate Gal-3-mediated extracellular activities by impacting the ability of Gal-3 to recognize carbohydrate ligands [152]. Intracellular Gal-3 can also regulate signaling pathways involved in TRAIL, Fas, and TNF-induced apoptosis [153–155]. As Gal-3 contains the bcl-2 NWGR motif, several studies suggest that Gal-3-mediated regulation of cell survival may occur through a bcl-2 pathway [155, 156]. In contrast, intracellular Gal-7 appears to induce keratinocyte apoptosis following UVB-induced damage [157]. Gal-1 also appears to directly regulate cellular signaling by modulating Ras localization and signaling [158]. Gal-12 plays a critical role in adipocyte signaling [159], which likely contributes to the increased lipolysis and reduced adiposity displayed by Gal-12 KO mice [160]. While studies suggest that several members of the galectin family, such as Gal-3 and Gal-4, likely regulate membrane trafficking through carbohydrate recognition in distinct membrane microdomains outside the cell [161, 162], recent studies suggest that Gal-8 may recognize extracellular carbohydrate ligands exposed during vesicular damage following bacterial infection intracellularly [163]. Gal-8-mediated recognition of damaged vesicles appears to target these contents for autophagy, providing a unique form of intracellular antimicrobial defense [163].

Fig. 6.

Galectin family members regulate various intracellular processes. While the earliest studies on galectins demonstrated that these proteins recognize extracellular carbohydrate ligands, subsequent studies have demonstrated that galectins possess significant roles in the regulation of a variety of fundamental intracellular activities, both in the cytoplasm and the nucleus. Representative galectin-regulated activities are shown. Red arrows indicate an activity that the respective galectin increases, while blue arrows signify galectin-induced decreases in the accompanying activity

In addition to regulating signaling pathways in the cytosol, galectins appear to modulate nuclear activities within the cell (Fig. 6). For example, some of the earliest work suggesting an intracellular role for galectins demonstrated that Gal-1 and Gal-3 might regulate a process as fundamental as RNA splicing [164]. Using nuclear extracts isolated from HeLa cells to examine RNA processing, depletion of Gal-1 or Gal-3 significantly inhibited RNA splicing, while reconstitution of only one of these galectins would largely restore this activity, strongly suggesting potential compensatory activity [164]. Subsequent studies demonstrated that Gal-1 and Gal-3 localize not only in the nucleus but also co-localize and directly interact with known factors involved in RNA splicing [165–167]. Consistent with their potential role in nuclear activities, including transcriptional regulation [168], nuclear import of galectin family members appears to be a highly regulated process [169, 170]. Inhibition of Gal-3 nuclear trafficking, by preventing serine phosphorylation via site directed mutagenesis, appears to significantly impact its ability to regulate cell survival [171].

12 Galectins and Neoplastic Disease

As galectins appear to play key regulatory roles in general immunity, hemostasis, vascular biology, cell signaling and viability, it is not surprising that altered galectin expression may impact neoplastic transformation and progression. Given the fundamental contribution of cell growth dysregulation in neoplastic transformation, it is fitting that some of the earliest studies suggesting that galectins may facilitate neoplastic transformation examined the potential impact of intracellular galectins on neoplastic cell survival. For example, seminal studies by Liu, Raz and others demonstrated that upregulation of Gal-3 can significantly reduce cellular sensitivity to apoptotic stimuli [150, 151, 153–155]. In contrast, neoplastic transformation of keratinized epithelial cells appears to result in downregulation of Gal-7 [157], consistent with its ability to induce apoptosis following cellular injury [172]. Furthermore, dysregulation of other galectin pathways also appears to likewise impact cellular signaling, with a significant effect on cell growth and neoplastic progression [173]. In addition to directly regulating neoplastic transformation at the cellular level, galectin family members may also facilitate overall tumor growth and dissemination by enhancing tumor angiogenesis [174, 175]. Several galectins, in particular Gal-3, appear to engage circulating neoplastic cells and enhance their extravasation, directly facilitating metastasis [176].

Given the role of immunosurveillance in the inhibition of neoplastic disease [177], transformed cells may also utilize galectins to inhibit immunological rejection of neoplastic lesions. Early studies demonstrated that overexpression of Gal-1 by neoplastic cells likely enhances neoplastic cell survival by directly inhibiting antitumor immunity. For example, overexpression of Gal-1 in a variety of settings, including neuroblastoma and breast cancer, appears to impair CD4+ T cells, CD8+ T cells and DC function, strongly suggesting a role for Gal-1 in the suppression of antitumor immunity [178, 179]. Consistent with this, knockdown of Gal-1 in neoplastic cells prior to transfer in vivo results in considerable expansion and enhanced function of CD4+ and CD8+ T cells, decreased levels of Tregs and enhanced DC function [178, 179]. Gal-1-mediated regulation of antitumor immunity may similarly impact neoplastic hematological disease. Increased expression of Gal-1 in patients with Hodgkin’s lymphoma correlates with a poorer prognosis and may likewise reflect significant modulation of antitumor immunity [34, 180]. In contrast, other galectins may also actually facilitate antitumor immunity. For example, Gal-9-mediated alterations in macrophage activation appear to result in significant NK cell activation, thereby enhancing NK cell-mediated antitumor immunity [92].

13 The Biochemistry of Galectins

Sensitivity to oxidative inactivation

In addition to the ability of galectins to recognize carbohydrates, the unique sensitivity of several galectin family members to oxidative inactivation represents one of their earliest recognized features. Indeed, the recognition of galectin sensitivity to redox environments provided the first name of this class of GBPs: the S-type, or thiol-dependent, lectins [181–185]. Some of the first studies on Gal-1 demonstrated that in the absence of reducing conditions Gal-1 gradually oxidizes [182], which results in profound conformational changes that preclude dimerization and recognition of carbohydrate ligands [186–188] (Fig. 7). However, redox regulation is not limited to Gal-1, as several other galectins, including Gal-2 and Gal-7, appeared to also be sensitive to oxidative inactivation.

Fig. 7.

Galectin activity is regulated by oxidation and proteolysis. Galectins, first called S-type lectins secondary to the requirement of several galectins for reduced thiols to maintain carbohydrate recognition activity, can form intramolecular and intermolecular disulfide bridges that often result in significant conformational changes that preclude carbohydrate recognition. As monomers appear to be a key intermediate in oxidative inactivation and carbohydrates can drive dimerization, ligand appears to reduce oxidative inactivation by facilitating dimer formation (K_d¹ < K_d²). [hi] = higher concentrations of Gal-1. [low] = lower concentrations of Gal-1. In contrast, several galectins, especially tandem repeat and chimeric galectins, rely on linker peptide bound carbohydrate recognition domains or N-terminal collagen-like domains to facilitate dimerization/oligomerization. Cleavage of intervening peptides that connect oligomerization domains to functional carbohydrate recognition domains can render carbohydrate recognition domains monomeric and therefore incapable of generating molecular lattices typically thought to be required for optimal galectin-mediated signaling

While Gal-1 oxidation does occur following removal of reducing agents, several factors appear to regulate this process. Gal-1 engagement of ligand stabilizes dimer formation and reduces Gal-1 sensitivity to oxidative inactivation [12, 182, 188] (Fig. 7). As a result, the unique sensitivity of Gal-1 to oxidative inactivation may be critical in regulating the spatial and temporal activity of this and related family members. Rapid oxidation of Gal-1 following tissue damage likely reduces the ability of Gal-1 and other galectins to inhibit neutrophils and other leukocytes immediately following injury, allowing these cells to neutralize potential pathogens and remove necrotic debris. However, as neutrophils and other leukocytes encroach on viable tissue surrounding an area of inflammation, leukocyte mediated-damage may release reduced, and therefore active, Gal-1, allowing engagement of leukocyte ligands and signaling of leukocyte turnover [127, 188]. Furthermore, while the redox environment can directly impact Gal-1 function, several studies suggest that the redox status of the cells themselves may also alter cellular sensitivity to Gal-1 signaling. For example, while Gal-1 typically induces apoptosis in activated T cells in the presence of reducing conditions [28], similar alterations in viability may be less apparent in the absence of reducing agents [30, 96, 97]. As a result, the outcome of Gal-1 engagement of cellular ligands appears to reflect the integration of a variety of signals that determine whether cellular death occurs. It remains possible that reducing conditions in vitro may also serve as a surrogate for other unknown factors in vivo that similarly regulate cellular fate following Gal-1 engagement. While Gal-1 oxidation results in loss of carbohydrate binding activity, oxidized Gal-1 itself appears to possess distinct biological activity. Several studies demonstrate that oxidized Gal-1 can facilitate neuronal regeneration [189, 190], suggesting that oxidation may not only limit Gal-1 carbohydrate binding activity, but also result in acquisition of alternative biological properties. In this way, Gal-1, and perhaps other galectins behave as morpheeins, proteins that possess the unique ability to regulate diverse biological processes depending on the conformational state of the protein [191].

Although oxidative inactivation likely plays a critical role in the regulation of many prototypical galectins, additional regulatory mechanisms likely govern the activity of other galectin family members. For example, chimeric Gal-3 and tandem repeat galectins, unlike prototypical galectins, possess unique domains or linker peptides that facilitate oligomerization or functional bivalency. Sequences between the collagen-like “oligomerization” domain and the carbohydrate recognition domain of Gal-3, can be cleaved by several proteases, including matrix metalloproteinases 2 and 9 and collagenase 3 [192, 193]. As Gal-3-mediated signaling and galectin signaling in general often requires clustering and lattice formation of counter receptors, collagen domain removal-induced loss of Gal-3 oligomerization likely contributes to the inability of the CRD of Gal-3 alone to signal cells [194]. While recent studies suggest that the individual domains of tandem repeat galectins may retain unique biological activity, many of the putative functions of tandem repeat galectins also require intact full length proteins for optimal signaling [100, 195], suggesting that proteolytic cleavage of the linker peptide responsible for tethering carbohydrate recognition domains would likewise inhibit this subgroup’s function [196] (Fig. 7).

Carbohydrate recognition

Perhaps the most unique and defining feature of galectins lies in their ability to regulate cell behavior through the recognition of highly modifiable carbohydrate structures [5]. Although it should be noted that subtle, but fundamental differences in carbohydrate binding preferences were noted in elegant studies utilizing only a few defined carbohydrates [197, 198], these early studies were limited by the number of available test glycans [197, 198]. Using frontal affinity chromatography coupled with a large library of target compounds, Hirabayashi and colleagues demonstrated that individual galectin family members possess unique specificity for many different carbohydrates [199]. Additional studies, using a wide variety of different modalities in addition to frontal affinity chromatography, including isothermal calorimetry, fluorescence polarization, and more recently, glycan microarrays [99, 199–201], strongly suggest that individual galectin family members exhibit unique and overlapping carbohydrate binding specificity that in addition to differences in quaternary structure [99, 100, 202], likely contributes to the distinct and overlapping biological activities displayed by this protein family.

Although galectins recognize the common disaccharide motif Galβ1–4GlcNAc, numerous studies suggest that many galectin family members prefer polymers of this structure in the form of poly-N-acetyllactosamine (polyLacNAc), when expressed on the cell surface. However, the mode and mechanisms of polyLacNAc interactions appear to vary among different galectin family members, resulting in a differential impact of polyLacNAc modification on galectin recognition [100, 203]. For example, while Gal-1 and Gal-2 recognize the terminal N-acetyllactosamine (LacNAc) determinant [204], these prototypical galectins prefer the motif expressed on extended glycans, either as N-glycans or polyLacNAc glycans bearing this terminal LacNAc structure [203–205]. As Gal-1 and Gal-2 exists as homodimers, yet only recognize the terminal LacNAc motif, this preference likely only occurs following glycan immobilization, where extended glycans may make cross-linking interactions more favorable. Consistent with this, in solution-based assays, Gal-1 and Gal-2 do not display the same preference for extended ligands [203, 205]. Not only may homodimerization influence carbohydrate binding specificity, but binding of ligand itself can actually enhance Gal-1 dimerization [188]. Similar results demonstrate that ligand may also influence the quaternary state of other galectins, such as Gal-3 [202]. In contrast to Gal-1 and Gal-2, Gal-3 and Gal-8, which do not exist as rigid homodimers, recognize internal LacNAc motifs within extended polyLacNAc [100, 203]. As these two galectin family members possess oligomeric carbohydrate recognition domains organized through flexible linker peptides, conformational constraints that provide the specificity of Gal-1 and Gal-2 toward extended glycans terminating in LacNAc following immobilization may not be relevant. Consistent with this, Gal-3 displays a preference for polyLacNAc repeats in solid phase and solution-based assays [203].

Distinct recognition of polyLacNAc results in a differential impact of polyLacNAc sialylation, a highly regulatable glycan modification, on the binding and signaling of these galectin family members [22]. Sialylation of the terminal galactose of LacNAc commonly occurs in two distinct linkages. The α2–3 sialylation linkage reflects addition of the sialic acid to the 3-OH of galactose, while α2–6 sialylation refers to the attachment of sialic acid to the 6-OH of galactose. As Gal-1 and Gal-2 recognize the terminal LacNAc, sialylation can significantly and differentially impact glycan recognition [206]. For example, although Gal-1 recognizes α2–3 sialylated glycans, it fails to recognize glycans following α2–6 sialylation [22, 203]. In contrast, Gal-2 fails to recognize glycans following sialylation with either linkage [203]. Gal-3 binds internal LacNAc within polyLacNAc, and sialylation of polyLacNAc with either linkage fails to significantly alter Gal-3 binding and signaling [22, 203], although in different cellular contexts sialylation may also impact Gal-3 binding and signaling [206]. Unlike Gal-1, Gal-2, and Gal-3, Gal-4, Gal-8, and Gal-9 possesses two unique carbohydrate recognition domains. Intriguingly, the N-terminal domain of Gal-8 recognizes sialic acid, while the C-terminal domain prefers non-sialylated glycans [207]. In contrast, the individual domains of Gal-4 and Gal-9 do not appear to possess drastic differences in glycan binding. Recent studies suggest that Gal-8 may form dimers and signal leukocytes entirely through the C-terminal domain [100], suggesting that tandem repeat galectins may in general form higher order quaternary structures that may be important in signaling. Additional studies suggest that Gal-9 can also associate with Gal-3 and Gal-8, suggesting not only that galectins may form higher order homotypic oligomers, but that complex hetero-oligomers may also occur [208]. Given the ability of galectins to form highly ordered lattices on cell surfaces and the potential impact of galectin-induced lattice formation in galectin signaling [194], these results suggest that in addition to distinct carbohydrate recognition preferences, differences in quaternary structure also likely impact the potential outcome of galectin engagement of a cell. Alterations in galectin binding to glycoconjugate ligands may not only regulate cellular signaling, but could also significantly impact galectin-mediated regulation of the sorting of cell surface glycoproteins [209–211]. In addition to common glycoprotein modifications, galectins appear to also possess affinity for glycolipids, including GM1 [212]. Given the proclivity of GM1 for unique lipid microdomains [213], galectin–glycolipid interactions may directly impact cellular signaling [214].

14 Galectins in Innate Immunity

As galectin family members recognize a wide variety of highly modifiable carbohydrate structures, it is not surprising that individual galectin family members appear to be implicated in nearly every biological process described. However, recent studies provide insight into what may be one of the first biological roles of galectin family members—their ability to recognize pathogens [215]. Similar to the ability of mannose-binding protein, complement and other soluble immune factors [216, 217], Gal-3 recognizes a variety of microbes, including Neisseria gonorrhoeae, Candida Albicans, Toxoplasma gondii, Leishmania major, Schistosoma mansoni, and Trypanosoma cruzi [8, 215, 218–222], while Gal-9 also appears to recognize Leishmania major [223, 224]. Similarly, several studies suggest that Gal-1 binds Nipah, Influenza, and HIV viruses [225, 223] [226, 227] and that other galectins may specifically recognize unique microbial species [9].

Galectin-mediated interactions with pathogens appear to result in a wide variety of consequences, from enhanced phagocytosis to direct killing (Fig. 8). For example, Gal-3 recognition of Candida albicans induces direct microbial killing [8]. In contrast, Gal-3 binding of Toxoplasma gondii glycosylphosphatidylinositols facilitates parasite recognition and TNF production by macrophages [218]. In addition to recognizing glycans unique to pathogens, recent results suggest that several galectins may be able to protect individuals from microbes that express self-like antigens. For example, Gal-4 and Gal-8 recognize and kill microbes expressing the blood group B antigen, providing a potential mechanism of protection against blood group B expressing pathogens in blood group B positive individuals [9]. Recent studies utilizing microbial glycan microarrays suggest that Gal-4 and Gal-8 recognition of self-like microbial determinants may not be limited to blood group antigens, but that galectins may recognize a variety of microbes, both gram negative and gram positive, that share the unique feature of generating mammalian-like carbohydrate determinants [228, 229]. These results suggest that galectins may provide a broad form of innate immunity against molecular mimicry. Several studies also suggest that Gal-1 binds and inhibits entry of Nipah and Influenza viruses [225, 223]. Thus, galectin recognition of pathogens appears to provide innate immunity through a variety of mechanisms.

Fig. 8.

Galectins recognize a diverse range of pathogens. While the immunoregulatory roles of galectins likely represent some of their most well known functions, the ability of galectins to recognize a diverse range of pathogens may reflect some of their earliest evolutionary activities. Galectin recognition of pathogens can result in opsonization or direct microbial killing. In contrast, pathogens may utilize galectins to facilitate attachment and invasion. Representative galectin-regulated activities are shown. Red arrows indicate an activity that the respective galectin increases, while blue arrows signify galectin-induced decreases in the accompanying activity

In contrast to innate immune activity of galectins, several pathogens appear to have subverted galectin function and use these proteins as a mechanism of host attachment and invasion (Fig. 8). For example, key work by Sato and colleagues demonstrated that Gal-1 may actually facilitate HIV entry during infection [226, 227], likely through direct glycan interactions between gp120 and CD4 [230]. Gal-1 also appears to facilitate Trichomonas vaginalis infection by enhancing attachment to the vaginal epithelium [231], while Gal-3 appears to similarly aid in the colonization of Neisseria meningitidis [222]. Galectin facilitation of pathogen survival does not appear to be limited to mammals, as a Gal-9 orthologue, PpGalec, found in the midgut of the sandfly Phlebotomus papatasi, mediates attachment and facilitates survival of Leishmania major in this key vector [232].

15 Summary

Although the discovery of galectins over 30 years ago stemmed from interest in understanding the roles of carbohydrates in fundamental biological processes, this finding ultimately uncovered an entire family of potent regulatory proteins [12–14]. Given their nearly ubiquitous expression and ability to bind highly modifiable carbohydrate ligands expressed on nearly every cell, in addition to a variety of other intracellular and extracellular partners, these proteins appear to be uniquely poised to regulate a wide variety of biological activities. Consistent with this, individual members of the galectin family have been implicated in various aspects of nearly every biological process described. As a result, these proteins may not only be some of the most ancient lectins known, but given their history throughout evolution, also appear to have been a unique evolutionary substrate in many different biological systems. Future studies will undoubtedly continue to unravel additional biological functions of this complex and intriguing protein family.

Abbreviations

APC

Antigen presenting cell

DC

Dendritic cell

DTT

Dithiothreitol

εBP

Epsilon binding protein

ER

Endoplasmic reticulum

FITC

Fluorescein isothiocyanate

Gal-1

Galectin-1

Gal-2

Galectin-2

Gal-3

Galectin-3

Gal-4

Galectin-4

Gal-7

Galectin-7

Gal-8

Galectin-8

Gal-9

Galectin-9

Gal-10

Galectin-10

Gal-12

Galectin-12

GBP

Glycan binding protein

IFN-γ

Interferon gamma

IgE

Immunoglobulin E

IL-10

Interleukin-10

IL-5

Interleukin-5

IL-6

Interleukin 6

KO

Knockout

LacNAc

N-acetyllactosamine

LPS

Lipopolysaccharide

MOG

Myelin oligodendrocyte protein

NK cell

Natural Killer Cell

NKT cell

Natural Killer T cell

PKC

Protein kinase C

PLC-γ

Phospholipase C-gamma

polyLacNAc

Poly-N-acetyllactosamine

PS

Phosphatidylserine

TCR

T cell receptor

TGF-β

Transforming growth factor beta

TH1

T helper cells type 1

TH2

T helper cells type 2

TNF

Tumor necrosis actor

Tregs

Regulatory T cells

vWF

von Willebrand Factor