The lectin functions of this family of proteins are the most widely studied. Indeed, galectins bind to β-galactosides through their CRD. For instance, considering its canonical ligand lactose, the C4’ and C6’ hydroxyls of the galactose and C2 and C3 of glucose are primarily responsible for the hydrogen-bond interactions with conserved residues of CRD in galectin-3 (117) and galectin-1 (118). Basis of the molecular glycan-protein interactions has also been described for other galectins (119, 120). The fine specificity of galectins for different oligosaccharides stems from residues surrounding this main binding site. Consequently, each galectin has a different glycan-binding preference contributing to its specific biological activities (121). The first described galectin inhibitors are molecules capable of binding to the CRD and preventing further ligand binding. Galectin inhibitors based on these competitive interactions consist of chemically modified mono or disaccharides structured around galactose (58, 122–125), lactose (58, 125–127), thiodigalactose (TDG) (34, 128–132), talose (133, 134) and lactulose (135). One of the first tempts to use this type of inhibitor in cancer consisted of administering a β-D-lactosyl-steroid. This treatment significantly increased the survival of mice grafted with lymphoma and glioblastoma cells (47, 136). Moreover, this compound increases the anti-tumor cytotoxic effects of cisplatin in mice (47).

Several chemical modifications of glycans have been developed to improve these molecules’ inhibitory properties. For example, introducing a sulfur atom into the glycoside linkage in TDG makes the molecule more resistant to glycosidases (137). The in vivo anti-tumor properties of some of these compounds were challenged in pre-clinical studies. For instance, TDG administration reduces pulmonary metastasis in murine breast and colon cancer models (48). TDG promotes immune infiltration, reduces angiogenesis, and protects cells against oxidative stress (49). The most advanced TDG in clinical studies is TD139 (also named as GB0139), developed by Galecto Biotech (Copenhagen, Denmark). TD139 recognizes galectin-3 CDR with high affinity (Kd 68 nM) (138). However, its absolute selectivity for galectin-3 is relative since it also binds to galectin-1 CDR (Kd 220 nM) and other galectins with lower affinities (138). This compound was initially evaluated in pre-clinical models of lung fibrosis (50, 51). Interestingly, TD139 was also evaluated in a clinical trial as a potential therapeutic for idiopathic pulmonary fibrosis (NCT02257177; www.clinicaltrials.gov [accessed November 24, 2022]; Table 2 ) (139).

More recently, a series of monosaccharide galectin-3 inhibitors with high affinities and good selectivity over other galectins have been described (140). From this series, GB1107 (3,4-dichlorophenyl 3-deoxy-3-[4(3,4,5-trifluorophenyl)-1H-1,2,3- triazol-1-yl]-1-thio-α-D-galactopyranoside) from Galecto Biotech; has good affinity (Kd 37 nM) and bind to the CRD of galectin-3. Both, TD139 and GB1107 are membrane-permeable small molecules (141). GB1107 is characterized by good biodisponibility upon oral administration and low clearance (52). It was demonstrated that the oral administration of GB1107 reduced human and mouse lung adenocarcinoma growth and blocked metastasis in murine models (52). Mechanistically, treatment with GB1107 promotes tumor M1 macrophage polarization and CD8(+) T-cell infiltration (52). Moreover, GB1107 potentiated the effects of a PD-L1 immune checkpoint inhibitor to increase expression of cytotoxic (IFNgamma, granzyme B, perforin-1, Fas ligand) and apoptotic (cleaved caspase-3) effector molecules (52, 53). In addition, GB1107 and cetuximab displayed a synergistic inhibitory effect on the growth of oral squamous cell carcinoma (54). Phase I studies with GB1211 (which shares a chemical template with GB1107) have been completed (NCT03809052, Table 2 ), and Galecto Biotech initiated safety and efficacy clinical studies with GB1211 combined with atezolizumab in the treatment of non-small-cell lung cancer (NCT05240131, Table 2 ).

Finally, it should be mentioned that chemical modifications of galactosides and their evaluation as galectin inhibitors in cancer are an intense field of research. First, synthetic glycoamines evidenced anti-tumor activity (55, 56, 142, 143). Indeed, lactulose-L-leucine mimics cancer-associated Thomsen-Friedenreich glycoantigen and binds to galectin-3. At a molecular level, it was demonstrated that this compound binds to the CRD of galectins-1 and -3 with higher affinity than lactose and TDG (135). In a murine breast cancer model, the administration of lactulose-L-leucine (and fructosyl-D-leucine) inhibited spontaneous metastasis in nude mice (56). The same group demonstrated the beneficial effects of lactulose-L-leucine in controlling and preventing prostate cancer metastasis to the bone (55). Other inhibitory molecules arising from chemical modifications of galactosides can also be cited (122, 144–146); however, they do not reach the level of in vivo evaluation.

To improve galectin inhibitors’ properties, inspiration was found in the clustering nature of galectin glycan interactions. Indeed, the synthesis of multivalent glyco-clusters with improved galectin inhibitory potential has been reported (147–152). Interestingly, cell aggregation can either be inhibited or enhanced depending on the number of lactose groups in functionalized dendrimers (153). Unfortunately, no evaluation of their in vivo biological effects in pre-clinical models was yet reported. Another strategy based on the same conceptual framework tested dendrimers obtained by galactose conjugation to the porphyrin derivatives (154). In this case, a photodynamic anti-tumor therapy was successfully reported in a pre-clinical in vivo bladder cancer model (57).

Pectins are another group of galectin-binding, inhibitory compounds. Natural pectins are large and heterogeneous polysaccharides found in plants which constitutes fiber components of our diet. Pectins have molecular weights ranging from 60 to 130 kDa and are constituted by three main polysaccharides: homogalacturonan (HG), rhamnogalacturonan-I (RG-I), and substituted galacturonans (GS) (155, 156). Pectins must be modified by pH and heat to gain solubility and biological effects. Indeed, hydrolysis induces galactoside exposure, and now, modified pectins bind galectins (157, 158). Contrary to what was often supposed, some experimental data prompt the existence of non-conventional sites of pectin binding in galectins (159–162). On the contrary, other in vitro data support that pectin-mediated biological effects are (or partially are) mediated by glycans (163–166). Adding complexity to the field, modified pectins are generally administered orally. Nevertheless, pectins are not digestible in the human intestinal tract, and their modifications are believed to increase their absorbability (167, 168). Moreover, it has been postulated that products of pectin fermentation by the human microbiota should contribute to their systemic in vivo biological effects (169). It should also be mentioned that pectins induce galectin-independent biological effects (170, 171). Altogether, these arguments indicate that more basic research is needed to clarify the fine mechanisms through which pectins induce their biological effects.

In this context, one of the most studied galectin inhibitors is the modified citrus pectin (MCP), which is obtained by partial hydrolysis of citrus pectin. In vitro studies demonstrated that MCP binds galectin-3 through galactoside residues (59, 62). Functionally, MCP inhibits galectin-3 binding to endothelial cells, and more importantly, the adhesion of breast tumors to endothelial cells (61). In addition, MCP treatment induces important metabolic changes in tumor-associated macrophages, which impacts on tumor growth and metastasis (172, 173). Interestingly, these MCP biological effects are carbohydrate dependent (59). In vivo administration of MCP inhibits melanoma (59), thyroid (60), breast and colon tumor growth, angiogenesis and metastasis (61, 173), and spontaneous metastasis in a rat prostate cancer model (62). Due to the high chemical variability of dietary MCP supplements on the market, more defined MCP variants have been described: PectaSol-C, GCS-100, GM-CT-01 and GR-MD-02. PectaSol-C has a molecular weight ranging from 5-10 kDa with 5% of monogalacturonic acid content (174). In vitro studies demonstrated the potential interest of PectaSol-C MCP in prostate (174, 175), breast (175) and ovarian cancers (176, 177), particularly if used combined with other therapies (175, 177). Interestingly, phase II pilot studies demonstrated the tolerability and encouraging biological results obtained by the use of this inhibitor in prostate patients (108, 109) (NCT01681823, Table 2 ).

GCS-100 is a complex polysaccharide prepared from modified citrus pectin. Mechanistically, GCS-100 detaches galectin-3 from CD4+ and CD8+ tumor-infiltrating lymphocytes, boosts cytotoxicity and restores IFN-gamma secretion (63). Similar effects were obtained by using N-acetyllactosamine, suggesting GCS-100 effects are carbohydrate-dependent (63). Interestingly, GCS-100 induces tumor rejection only when associated with vaccination in pre-clinical model of mastocytoma secretion (63), implying GCS-100 modulates the tumor immune attack. Altogether, these promising results prompt La Jolla Pharmaceuticals to launch GCS-100-based clinical trials. Following a phase I dose escalation safety study in patients with refractory solid tumors (178), a phase II study was completed in patients with chronic lymphocytic leukemia (179) (NCT00514696, Table 2 ). In these exploratory trials, GCS-100 was well tolerated, and 25% of patients showed a partial response (179). In addition, the use of GCS-100 has also been evaluated in chronic kidney disease (Phase I NCT01717248 and phase IIa NCT01843790, Table 2 ). In 2015, La Jolla Pharmaceuticals announced that they were discontinuing the development of GCS-100 after the Food and Drug Administration (FDA) required a more complex characterization of the compound to advance into late-stage development (NCT00776802 and NCT00609817, Table 2 ).

Another pectin-derived polysaccharide able to inhibit galectins is GM-CT-01 or DAVANAT^®. This polysaccharide is extracted from guar seeds and subjected to controlled partial chemical degradation (developed by Galectin Therapeutics, formerly Pro-Pharmaceuticals). A backbone of the galactomannan is composed of (1→4)-linked β-D-mannopyranosyl units, to which single α-D-galactopyranosyl is attached by (1→6)-linkage (64). The average repeating unit of GM-CT-01 consists of seventeen β-D-Man residues and ten α-D-Gal residues (Man/Gal ratio is 1.7), and an average polymeric molecule contains approximately 12 of such repeating units (for the average molecular weight of 51,000 Da). In vitro, GM-CT-01 boosts the cytotoxic properties of CD8(+) tumor-infiltrating lymphocytes and their ability to produce IFN-gamma (180). Indeed, this pectin prevents glycosylated cytokines (IFNγ between others) be captured by galectin-3 and therefore allowing the chemokine gradient needed to attract lymphocytes towards the tumor (181). Pre-clinical studies in mice defined GM-CT-01 non-toxic doses (alone or combined with other chemotherapies) (64). Moreover, such studies demonstrated GM-CT-01 beneficial effects in colon cancer models (64). Interestingly, a phase I clinical trial was completed in cancer patients with advanced solid tumors by administration of DAVANAT^® combined with 5-fluorouracil treatment (NCT00054977, Table 2 ). Combinatory treatment was well-tolerated. While phase II trials were announced, these trials were never initiated, having a “withdrawn/terminated status” in www.clinicaltrials.gov (NCT00388700, NCT00110721, NCT00386516, Table 2 ). In addition, melanoma peptide vaccination plus GM-CT-01 was evaluated in melanoma (NCT01723813). This clinical trial was “terminated due to end of validity of peptide vaccine” with no reported results.

Finally, GR-MD-02 (belapectin) is a galactoarabino-rhamnogalacturonan-rich polysaccharide obtained through chemical processing from apple pectin (developed by Galectin Therapeutics, Norcross, Georgia, USA). GR-MD-02 is a galectin-3 inhibitor which synergizes with anti-OX40 treatment to promote tumor regression and increases survival of tumor-bearing mice (65). This occurs through a CD8(+) T cell-dependent mechanism, reducing the immunosuppresion mediated by myeloid-derived suppressor and regulatory Foxp3(+)CD4(+)T cells (65). GR-MD-02 administration induced a significant reduction of liver fibrosis in experimental models of non-alcoholic steatohepatitis (182, 183). GR-MD-02 is being evaluated in melanoma, squamous head and neck, and non-small cell lung cancer patients combined with the negative immune checkpoint inhibitors pembrolizumab (anti-PD-1, NCT02575404, and NCT04987996, this last suspended) and ipilimumab (anti-CTLA-4, NCT02117362; Table 2 ). No results are available yet from those clinical studies. Interestingly, this compound has also been evaluated in clinical trials for non-alcoholic steatohepatitis, portal hypertension, and advanced liver fibrosis (NCT01899859 and NCT02462967, Table 2 ). In this case, GR-MD-02 was safe but not associated with significantly ameliorating hepatic disease (184).

An area under intense investigation tries to achieve formulations with improved pharmacokinetic properties for this type of carbohydrate-based inhibitors. This is the case of lactose-, galactose- or pectins-complexed nanoparticles (185–187). Apart from improving the pharmacokinetic properties of the inhibitor, these nanoparticles can also serve as delivery carriers of cytotoxic drugs toward the tumor (66, 188, 189). Moreover, attempts are being made with nanoparticle modifications to improve selective targeting of the tumor (or tumor-associated stroma) (67, 190).

Interestingly, non-carbohydrate inhibitors for galectins have also been proposed. First, the anti-tumor properties of several synthetic heterocyclic compounds able to bind galectin-1 have been evaluated. Molecular docking experiments described fine interactions between these molecules and the CRD domain of galectin-1 (191–194). Moreover, in vitro results indicate these compounds have anti-tumor cytotoxic properties (191–194). However, in vivo anti-tumor pre-clinical evaluations of such compounds remain to be performed. Second, bacteriophage display library systems for interaction screening allowed the discovery of galectin-binding peptides. For instance, a Thomsen-Friedenreich antigen-specific peptide (P-30) able to bind galectin-3 has been described (195). This peptide modulates breast and prostate tumor homotypic aggregation and tumor cell adhesion to the endothelium (195). Using similar technological approaches, stapled-peptides ligands binding galectin-3 were described (196). These peptides bind to the CRD of galectin-3 and the best one has an intermediate affinity (Kd 0.45 μM) (196). However, no functional studies have been reported for these peptides. As already mentioned, formulations with improved pharmacokinetics are being evaluated. In this context, nanoparticles combining carbohydrates (inhibitor) and peptides (addressers) have been described, a strategy that significantly improves their biodistribution and the biological effects (67).

Finally, genetic engineering methods are used to inhibit the glycan-dependent functions of galectins. For instance, a dominant negative mutant formed by the last 143 carboxyl-terminal amino acid residues and lacking the N-terminal domain of galectin-3 (named Gal-3C) has been described. This Gal-3C molecule preserves the CRD but lacks cooperative binding and crosslinking properties of the wild-type galectin-3 (197). Indeed, it is hypothesized that the administration of an excess of soluble Gal-3C competes with endogenous galectin-3 for carbohydrate binding sites (76). In this context, Gal-3C reduces angiogenesis by abrogating extracellular galectin-3 interaction with αvβ3 integrin through its carbohydrate recognition domain (198). Interestingly, Gal-3C inhibits CXCL12-induced leukocyte migration in (non-cancer) inflammatory conditions (199). Gal-3C also inhibits tumor cell motility and invasion (75, 200). Hence, Gal-3C alone or combined with other chemotherapies can reduce ovarian, breast cancer, and multiple myeloma growth and drug resistance (75, 76, 200). Interestingly, Gal-3C can be used in vivo without toxic effects (76); this treatment ameliorates heart failure after myocardial infarction (77). Galectin-9 mutants have also been described. Indeed, mutations in galectin-9 CRD abolish its binding to the negative checkpoint Tim-3; this interaction occurs via the carbohydrates (201). Dominant negative mutants can also interfere with nuclear partners in a glycan-dependent manner. This is the case of the interactions between galectin-1 and Foxp3. This transcription factor functions as a master controller of regulatory T cells (Treg). Moreover, the interaction between galectin-1 and Foxp3 controls a panoply of genes and functions in breast cancer cells (202). Consequently, galectin-1 mutants that lack the N-terminus and do not bind Foxp3 can be used to inhibit breast tumor proliferative and invasive properties (202). These results show that negative dominants could be interesting tools to inhibit galectins.

Some inhibitors do not directly interact with the CRD of galectins, but their inhibitory effects are still glycan-dependent. Indeed, these molecules function as allosteric inhibitors, interacting outside the CRD but inducing changes in this region, thereby inhibiting glycan binding and biological effects. For instance, in vivo galectin-1 inhibition through the administration of lactose-conjugated purpurinimide photosensitizers reduced the growth of radiation-induced fibrosarcoma (58). Molecular modeling analysis indicated that this compound does not interfere with the CRD (203). Similar photodynamic strategies with galactose-bound porphyrin demonstrated anti-tumor effects in bladder cancers (57). In this case, galectin-1 inhibition generates oxidative stress and apoptosis of tumor cells over-expressing this lectin (57).

However, allosteric inhibition can also be performed using non-carbohydrate molecules. Based on the significant role of galectins in the interaction between tumor and endothelial cells during tumorigenesis, a cytokine-like peptide named anginex was described as a potent anti-angiogenic tool (68). This biological effect is mediated through galectin-1 binding (69), although this peptide also binds other galectins (204). The anti-tumor effects of anginex were demonstrated in several experimental cancer models (26, 32, 43, 44). Anginex’s angiostatic beta-sheet-forming structure inspired the design of the 6DBF7, a peptidomimetic that also interacts with galectin-1 (70, 71). This 6DBF7 molecule inhibits glycan binding of galectin-1 in a noncompetitive, allosteric manner (71). Based on these studies, other potent analogs (DB16 and DB21) have also been described (71). These peptides inhibit angiogenesis and tumor growth significantly better than 6DBF7 or anginex (71). To overcome the susceptibility of these peptides to hydrolysis by proteases, Dings et al. designed a non-peptidic topomimetic of anginex and 6DBF7 based on a calixarene scaffold. Indeed, calix[4]arene compound 0118/OTX008/PTX008 binds to galectin-1 at a site away from the lectin’s carbohydrate binding site, thereby attenuating lactose binding to the lectin (205). It should be mentioned that the specificity of this compound is relative since it also binds to galectin-3, albeit more weakly (206). Pharmacokinetics and anti-tumor activity of OTX008 alone or combined with other treatments were evaluated in melanoma, glioblastoma, thyroid and ovarian carcinoma (40, 43, 72, 73, 207). A phase I study of OXT008 in patients with advanced solid tumors was reported (NCT01724320, Table 2 ). Unfortunately, this study is listed with an “unknown recruitment status”; no updates have been posted since 2012. Chemical modifications of PTX008 were also described; it is interesting to mention the PTX013 compound. This compound is more potent as a cytotoxic tumor agent than the parenteral PTX008. This higher inhibitory potency of PTX013 was demonstrated both in vitro (head and neck, breast, ovarian, renal, lung, and prostate cancer lines, several of them radiation resistant), and importantly in vivo (melanoma) (74).

Galectin inhibition can also be achieved using specific neutralizing monoclonal antibodies (mAb). It must be noted that, for this strategy, mechanisms of galectin inhibition (competition or allosteric inhibition) depends on each antibody. In the case of galectin-1, one of these antibodies (Gal-1-mAb3) has been characterized, and the epitope recognized by this mAb localizes outside the CRD although it is still capable of inhibiting N-acetyllactosamine-galectin-1 interaction (208). This antibody recognizes specifically galectin-1 with high affinity (EC50 = 523nM). This neutralizing antibody reproduces the anti-angiogenic and immunopotentiating activities observed with other types of inhibitors (208, 209). In particular, blockade of galectin-1 (Clone 25C1; Novo Castra) significantly reduces the in vitro inhibitory effects of human and mouse CD4+CD25+ Treg cells (210). Moreover, another anti-galectin-1 neutralizing mAb ameliorates the negative immune checkpoint (PD1) response in irradiated mice carrying oral cancer cells (45).

In the case of galectin-3, earlier studies described mAbs recognizing non-CRD domains but causing a profound modulation of its lectin activities (211). On the other hand, a galectin-3-specific mAb (14D11) competes with lactose for the carbohydrate-binding pocket of galectin-3 (81). This antibody inhibits invasion of Mucin-16-expressing cancer cells, prolonging overall survival in animal tumor models (81). However, inhibition of galectin-3 also impacts the tumor stroma cells. Indeed, the use of an anti-galectin-3 mAb (B2C10) promotes IFN-γ secretion by in vitro stimulated CD8+ tumor-infiltrating T lymphocytes (63).

The scientific interest in developing anti-galectin-9 mAb is major since this protein participates in various mechanisms of immune escape by tumors: control of T cell survival (212), T cell effector exhaustion and differentiation (82, 201, 213, 214), lymphocyte migration towards the tumor via an endothelial cell reprogramming (45), Treg function (215–220), regulation of antigen presentation (221–223), and myeloid suppressive cells (224). Confirmation of these functions by the use of blocking antibodies is becoming very frequent. Such is the case of two antibodies (clones 292-13 and 292-18A) reacting with high affinity with the N-CRD of human galectin-9; their use protects T cells from galectin-9 mediated cell death and promotes tumor-cell killing by T cells (225). The same group, but using a commercial anti-galectin-9 mAb (RG9-1 from InVivoMAb), demonstrated prevention of CD8+T cell exhaustion and near complete Treg depletion when this mAb is combined with anti-GITR (glucocorticoid-induced tumor necrosis factor receptor-related protein)-specific antibody (82). Two other anti-galectin-9 mAb have also been reported (Gal-Nab1 and Gal-Nab2). In this case, antibodies recognize an epitope comprising 213-224 amino-acid sequence with high affinity (in the order of nM) (226). Again, these antibodies protect T cells from galectin-9-mediated cell death (226). An anti-galectin-9 was combined with anti-Tim-3 mAb to improve taxane-based chemotherapy in breast cancer (83). Apart from the direct effects on adaptive immunity, blockade of galectin-9 by antibodies potentiates immune attack in pancreatic carcinoma through modulation of macrophage function (84). Nevertheless, galectin-9 blockade by antibodies also acts directly on tumor cells. Indeed, leukemia stem cells secrete galectin-9, which through the interaction with Tim-3 constitutes an autocrine loop critical for leukemic self-renewal and development (85). Indeed, galectin-9 neutralization is a potent way to prevent the reconstitution and the self-renewal of human acute myeloid leukemia cells in a xenogeneic transplantation model (85). Finally, an anti-galectin-9 mAb (Lyt-200) is currently under clinical investigation in phase I/II trial for its safety and efficacy in patients with relapsed/refractory metastatic solid tumors (NCT04666688, Table 2 ). In this clinical trial, Lyt-200 is evaluated alone and in combination with chemotherapy or anti-PD-1.

The use of neutralizing antibodies to block other galectin members in cancer is more incipient, and in most cases, polyclonal antibodies are evaluated. For instance, neutralizing surface-bound galectin-4 in human colorectal cancer induces significant transcriptional changes and chemokines production in tumor cells (227).

While neutralizing antibodies carry several benefits over small inhibitory carbohydrate molecules, they also have several drawbacks. Some of the concerns are related to their selectivities and biodistributions. Antibodies inhibit extra-cellular galectins, and lack restricted biodistribution in the body. These characteristics imply that antibody-mediated inhibition of galectins could act as partial inhibitors (lack of intracellular effects), and do not discriminate between non-transformed and transformed cells resulting in adverse effects. More studies are needed to fully understand the effects induced by galectin-neutralizing antibodies and their potential transfer to the clinic.

Finally, nucleotide-based molecules are a different family of galectin inhibitors. In this sense, a single-stranded DNA aptamer (AP-74 M-545) has been described as an antagonist of galectin-1 (86). This aptamer shows higher affinity (KD = 3.7 nM) and specificity than the previous inhibitors. Administration of this compound induces in vivo anti-tumor effects through activation of the immune system. Indeed, this aptamer prevents T cells from apoptosis and restores T cell-mediated immunity (86). This study did not evaluate aptamer dependence on glycans, so this point remains to be clarified.