TF-containing MUC1 glycopeptides fail to entice Galectin-1 recognition of tumor-associated Thomsen-Freidenreich (TF) antigen (CD176) in solution

Abstract

Aberrant Mucin-1 (MUC1) glycosylation with the Thomsen-Friedenreich (TF) tumor-associated antigen (CD176) is a hallmark of epithelial carcinoma progression and poor patient prognosis. Recognition of TF by glycan-binding proteins, such as galectins, enables the pathological repercussions of this glycan presentation, yet the underlying binding specificities of different members of the galectin family is a matter of continual investigation. While Galectin-3 (Gal-3) recognition of TF has been well-documented at both the cellular and molecular level, Galectin-1 (Gal-1) recognition of TF has only truly been alluded to in cell-based platforms. Immunohistochemical analyses have purported Gal-1 binding to TF on MUC1 at the cell surface, however binding at the molecular level was inconclusive. We hypothesize that glycan scaffold (MUC1’s tandem repeat peptide sequence) and/or multivalency play a role in the binding recognition of TF antigen by Gal-1. In this study we have developed a method for large-scale expression of Gal-1 and its histidine-tagged analog for use in binding studies by isothermal titration calorimetry (ITC) and development of an analytical method based on AlphaScreen technology to screen for Gal-1 inhibitors. Surprisingly, neither glycan scaffold or multivalent presentation of TF antigen on the scaffold was able to entice Gal-1 recognition to the level of affinity expected for functional significance. Future evaluations of the Gal-1/TF binding interaction in order to draw connections between immunohistochemical data and analytical measurements are warranted.

Introduction

The mechanistic functionality of carbohydrates, or glycans, presented extracellularly relies on the evolutionarily orchestrated interactions that entail recognition by glycan-binding proteins. These interactions permit the processing of the “glycan-code”, an information-rich cellular programming language that, through robust structural and sequence-based complexity, provides a third tier of algorithmic diversity beyond the canonical nucleotide or amino acid programs [1–4]. Nevertheless, processing of this “glycan-code” by glycan-binding proteins can be observed in both normal and pathological contexts.

Mucin-1 (MUC1), an epithelial transmembrane glycoprotein “scaffold” for O-glycosylation, is characteristically overexpressed in a vast majority of tumor tissue types including breast, prostate, gastric, ovarian, and hepatocellular malignancies [5–8]. Coinciding with a pathological expression profile deviant from that of normal cells, MUC1 O-glycosylation in diseased tissues displays aberrant glycan patterns characterized by the short glycan structures [7, 9–13]. The abundant diversity beknown of the extracellular glycocalyx regresses towards a rather barren landscape of select glycans, such as (GalNAc or Tn) and (Gal1–3βGalNAc or TF). The Thomsen-Freidenreich (TF) tumor-associated antigen (CD176) has been observed in over 90% of human carcinomas [14]. Recognition of TF is mediated by members of the β-galactoside binding family of lectins known as galectins [15–17].

The 15 members of the human galectin family are structurally subdivided as proto-type, which consists of noncovalent homodimers (Galectin-1, -2, -5, -7, -10, -11, -13, -14, -15), tandem-repeat type, which consists of covalent homodimers tethered together by an amino acid sequence-embedded linker (Galectin-4, -6, -8, -9, -12), and chimera-type, a C-terminal carbohydrate recognition domain (CRD) with an N-terminal tail (Galectin-3) [18]. Galectin-1 (Gal-1) is one of the most well-defined galectin members with functions regulating cell growth, adhesion, and differentiation, and regions of activity demarcated by the lectin’s localization within the nucleus, translocation to the intracellular membrane, or extracellular secretion [19–22]. Of focus is the malignancy associated extracellular interactions of proto-type Gal-1, which promote cell-signaling cascades induced by receptor clustering at the cell surface, a consequence of Gal-1 glycan recognition and subsequent non-covalent cross-linking [23, 24]. In the active, reduced form, Gal-1 is known to recognize TF O-glycosylated MUC1 based on immunohistochemical analyses and cell-based binding assays [25–28]. Intriguingly, Gal-1 did not bind significantly to TF antigen presented in a glycan microarray format [29] pointing out that the specificity of Gal-1 for glycans depends not only on the structure of glycans but also the mode of their presentation. Furthermore, Bian et al. failed to observe Gal-1 recognition of the TF antigen in solution beyond millimolar affinity using isothermal titration calorimetry [16]. The mutagenesis studies of Gal-1 pinpointed His-52 as the key CRD residue precluding TF binding due to steric hindrance between His-52 and -NHAc of TF [16].

Although carbohydrate recognition by glycan-binding proteins is inherently weak for most naturally occurring glycans found on glycoproteins, factors such a lectin and/or glycan multivalency, and glycan presentation (i.e. peptide scaffold architecture) enable high-affinity binding [30–33]. As shown in our previous study, presentation of TF on a 20-mer peptide at Thr/Ser (HGVTSAPDTRPAPGSTAPPA) of the MUC1 tandem repeat enhanced Gal-3 recognition of TF nearly 10-fold [34]. A thermodynamic shift in the driving force of the binding interaction between Gal-3 and TF was also notably observed, wherein Gal-3 recognizes TF disaccharide in an enthalpy-driven manner yet recognizes MUC1-TF glycopeptides in an entropy-driven manner. By extending the role of the MUC1 scaffold on glycan recognition by Gal-3 into our present study, we aimed to establish whether Gal-1 recognition of TF is enabled, in part, by presentation on the MUC1 scaffold, its physiologically relevant carrier protein.

Utilizing isothermal titration calorimetry (ITC), the gold standard for analytical affinity measurements, we were able to derive both precise binding affinities and a complete thermodynamic profile of the binding interaction(s) in a solution-based, label-free experiment. While ITC provides an attractive information-rich approach, the technique requires substantial quantities of material that are expended at relatively low-throughput. In order to meet the considerable lectin quantities required to perform our ITC studies, a high-yield expression and purification protocol for recombinant Gal-1 was developed and optimized, followed by characterization of our lectin product via SDS-PAGE and Western Blot. A histidine-tagged Gal-1 analog was also expressed for development of an AlphaScreen competitive binding assay to screen for inhibitors of low-affinity glycan-lectin interactions, as was previously developed for Gal-3 using asialofetuin (ASF), a glycoprotein with three triantennary LacNAc presentations, as the binding partner [35]. In short, Gal-1 maintained previously reported binding and thermodynamic profiles with LacNAc, lactose, and ASF by ITC and bound to TF antigen in cell-based assay. However, Gal-1 failed to recognize TF disaccharide and Thr-TF (TF antigen) beyond millimolar affinity, and mono- and trivalent-TF presented on the MUC1 glycopeptide scaffold at the molecular level, warranting further studies into the mechanistic nature of TF recognition by Gal-1 observed at the cell surface.

Materials and methods

Reagents

Fmoc-protected amino acids and coupling reagents for peptide synthesis, 2-(6-chloro-1H-benzotriazol-1-yl)-1,1,3,3-tetramethylaminium hexafluorophosphate (HCTU) and 1-hydroxybenzotriazole (HOBt), were obtained from Sigma Aldrich. Dimethyl formamide (DMF) peptide synthesis grade and HPLC-grade acetonitrile, used for peptide purification and analysis by HPLC were purchased from Fischer Scientific. N-Acetyllactosamine, 2-Acetamido-2-deoxy-3-O-β-D-galactopyranosyl-D-galactose (Galβ1–3GalNAcα1), and Threonyl-TF (Thr-TF) were purchased from Toronto Research Chemicals. Asialofetuin (ASF) from fetal calf serum Type 1 and α-lactose agarose resin were purchased from Sigma-Aldrich. Untagged LGALS1 cDNA ORF clone for human galectin-1 was purchased from OriGene. pET-28a(+) and pET-26b(+) T7 expression vectors, Escherichia coli Rosetta(DE3)pLysS cells, and Ni-NTA His-Bind resin were purchased from Novagen. Mini-PROTEAN TGX stain-free gels, nitrocellulose transfer stacks, and Clarity Max™ Western ECL blotting substrate were purchased from Bio-Rad. Polyclonal Goat IgG anti-Galectin-1 antibody, Polyclonal Donkey IgG anti-Goat HRP-conjugated antibody, and Polyclonal Goat anti-Mouse IgG HRP-conjugated antibody were purchased from R&D Systems. Monoclonal mouse IgG THE™ His-Tag antibody was purchased from GenScript, Thomsen-Friedenreich (A78-G/A7) mouse monoclonal antibody from Santa Cruz Biotechnology, and NTA-Atto-488 from Sigma Aldrich. General reagents for microbiology, protein expression, and purification work were purchased from Fisher Scientific. AlphaScreen Histidine (nickel chelate) detection kit and solid white polystyrene microwell plates, Proxiplate-384 PLUS, were purchased from PerkinElmer. Biotin-labeled asialofetuin (biotin-ASF) was purchased from Axxora. MCF7 (ATCC® HTB-22™) human cell line was purchased from American Type Culture Collection (ATCC, Manassas, VA).

Cloning, expression and purification of Gal-1 and histidine-tagged Gal-1

The open reading frame (ORF) of an untagged LGALS1 cDNA clone for human Gal-1 was amplified by PCR utilizing a 5’ oligonucleotide primer (5’ – GCCGATCATATGGCTTGTGGTCTGGTCG – 3’) encoded with a NdeI restriction site and a 3’oligonucleotide primer 3’– GCCGATTTCGAATCAGTCAAAGGCCACACATTTG – 5’) encoded with a HindIII restriction site. The amplified fragment was ligated into a pET-28a(+) T7 expression vector for the addition of an N-terminal histidine-tag Gal-1 and a pET-26b(+) T7 expression vector for the unlabeled Gal-1 construct were used (Supplementary Fig. S1). The identity of the targeted gene was confirmed by full-length sequencing (Eurofins). Competent E. coli Rosetta(DE3)pLysS cells were transformed with positive plasmid vectors via heat shock at 42°C. Transformed cells were incubated at 37°C in Luria-Bertani (LB) broth until OD₆₀₀ measurements reached an absorbance of 0.6 was identified, at which point induction was performed using isopropyl β-D-1-thiogalactopyranoside (IPTG) for a final concentration of 0.5 mM. The culture was incubated for an additional 5 h at 30°C (Supplementary Table S1). Harvested cells were resuspended in phosphate buffer (75mM Na₂HPO₄/KH₂PO₄, 2mM EDTA, 4mM β-mercaptoethanol, pH 7.2) supplemented with Halt Protease Inhibitor Cocktail, and sonicated (Fisher Scientific Model 120 sonicator) on ice. The supernatant containing Gal-1 was applied to an α-lactose agarose resin, and in case of his-tagged Gal-1, Ni-NTA His-Bind resin was followed by α-lactose agarose, respectively (Supplementary Fig. S2). The expressed protein were purified according to our published protocol for Gal-3 and his-tagged Gal-3 [35], and further details are given in Supplementary material (Supplementary pages S3–S4). Expression yields were determined spectrophotometrically at 280 nm using the sequence-specific extinction coefficient (ExPASy ProtParam) for his-tagged-Gal-1 of 5.4 E^1%_1cm using an EPOCH micro-volume plate reader (BioTek). Purity of expressed proteins was confirmed via 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western Blot (WB). The polyclonal Goat IgG anti-Gal-1 primary antibody and a monoclonal mouse IgG THE™ anti-his tag antibody were used to detect presence of Gal-1 and his-tag. The blot was imaged with an Odyssey® FC imaging system (LI-COR) using chemiluminescence detection.

Synthesis of MUC1-TF glycopeptides

Standard Fmoc solid-phase peptide chemistry was employed to synthesize two MUC1 (glyco)peptide analogs, MUC1-TF (Thr⁹) as previously reported [34] and MUC1-TF (Thr^4,9.16) on a PS3 automated peptide synthesizer (GyrosProtein Technologies Inc.). TentaGel S RAM resin (Advanced ChemTech) was used, and amino acid couplings were done using a 4-fold excess of amino acids, HOBt, and HCTU, in the presence of 0.4 M N-methylmorpholine (NMM) in dimethylformamide (DMF). The pentafluorophenyl ester of glycoamino acid was coupled manually using a 1.5-fold excess, in the presence of N, N-diisopropylethylamine (DIPEA) (pH 8). The coupling was carried out for 16 h. Upon completion of the assembly of the peptide chain, the N-terminal Fmoc protection was retained. The azido group was converted into the desired acetamido group by treatment of the resin with thioacetic acid for 3 days with daily changes of the reagent. The Fmoc protective group was removed with 20% piperidine in DMF. The fully assembled glycopeptide chains were cleaved from the resin using a TFA/thioanisole/water mixture (95:2.5:2.5) ratio for 3 h. The cleavage solution was then precipitated in cold methyl tert-butyl ether (MTBE) to yield the crude acetylated glycopeptides. Acetyl groups were removed using 0.01 M NaOH solution for 15 min, solution was neutralized with hydrochloric acid (HCl), and lyophilized to yield the final crude MUC1-Tn glycopeptides.

(Glyco)peptide purification was performed using a 1260 Infinity reversed-phase high-performance liquid chromatography (RP-HPLC) system by Agilent Technologies. The analytical RP-HPLC uses an Aeris Peptide C18 column (150 × 4.6 mm, 3.6μm, 100Å) at 0.8 mL/min flow rate, with 0.1% TFA in water (A) and 0.1%TFA in acetonitrile (B) as the eluents. The elution gradient for analytical RP-HPLC purification was 0 to 30% B over 30 min. Preparative RP-HPLC was utilized for final peptide purification, which uses the Grace Vydac monomeric C18 column (250 × 22 mm, 10μm, 120Å) at 10 mL/min flow rate, with 0.1% TFA in water (A) and 0.1%TFA in acetonitrile (B) as the eluents. The elution gradient for preparative RP-HPLC purification was 0 to 25% over 110 min. The peptide analogs were detected at 214 nm by the UV-Vis detector (variable detector, Agilent Technologies). Purified peptides were characterized by MALDI-TOF MS by Voyager-DE™ STR (Applied Biosystems, Foster City, CA) using α-cyano-4-hydroxycinnamic acid as matrix (Supplementary Fig. S4–S7).

Isothermal titration calorimetry

Ligand and Gal-1 or histidine-tagged Gal-1, were prepared in buffer containing 20 mM phosphate, 10 mM NaCl, 10 mM β-mercaptoethanol, pH 7.2 for experimental titrations. Purified Gal-1 and histidine-tagged Gal-1 were dialyzed against buffer for approximately 72 h at 4 °C, with buffer exchange every 12 h, to remove bound lactose from purification. Calorimetric measurements were recorded using a PEAQ-ITC calorimeter (Malvern Panalytical). A solution of 36.4 μL of the ligand was titrated in aliquots of 2 μL into the calorimetric cell containing 200 μL of lectin. Injections were performed every 150 seconds, for a total of 19 injections, at 25 °C and 750 rpm stir speed. Thermodynamic analysis was performed using the MicroCal PEAQ-ITC analysis software and a fitted offset parameter was applied to each titration to account for potential background signal.

Cell binding assay with Galectin-1

MCF-7 cells were grown in DMEM supplemented with 10% FBS at 37 °C with 5% CO₂. About 0.5 × 10⁴ cells were plated in 48-well plate (Thermo Fisher Scientific, Cat#130187). 24 h after plating, cells were washed with PBS, and then fixed with 4% paraformaldehyde for 30 min. The wells were then washed with PBS, blocked with 3% BSA for 30 min at room temperature, and incubated with 100 μl his-tag Gal-1 (50 μM) or PBS (as control) overnight at 4°C. For pretreatment with TF antibody, 100 μl of A78-G/A7 antibody (1:100 dilution) was incubated for 90 min before his-tag Gal-1 was added. Wells were rinsed three times with PBS, incubated with NTA-ATTO 488 (1:500 dilution) for 3 h, and then washed again three times with PBS. Cell images were taken using Cytation 5 image reader (BioTek).

AlphaScreen assay development

Initial assay characterization and validation was carried out with his-tagged Gal-1 and biotinylated-asialofetuin. Final his-tagged Gal-1 concentration ranged from 100 to 5000 nM. Final biotin-ASF concentration ranged from 0.1 to 100 nM. The assay was performed in 25 mM HEPES, 100 mM NaCl, 0.05% Tween-20, pH 7.4, in five replicate sets. The final well volume was 15 μL in a 384-well plate. his-tagged Gal-1 (2.5 μL) and biotin-ASF (2.5 μL) were incubated for 1 h at room temperature, followed by the addition of nickel chelate acceptor beads (5 μL) and streptavidin conjugate donor beads (5 μL) to a final concentration of 20 μg/mL for each bead type. The plate was incubated for 1 h at room temperature in the absence of light. The plate was then read in counts-per-second (cps) under an AlphaScreen protocol (excitation wavelength = 680 nM, emission wavelength = 570 nM, 22 °C) on a Cytation 5 Multi-Mode Plate Reader (BioTek).

Results and Discussion

Expression and purification of Gal-1 and histidine-tagged Gal-1

Here we report the optimized protocol for the expression of recombinant Gal-1 and his-tagged Gal-1 to maximize yields for their relatively high ITC experimental consumption. The expression was carried out in E. coli Rosetta (DE3) pLysS competent cells. The IPTG concentration used for induction via the T7 RNA polymerase pathway was varied (0.5 to 4 mM) along with the length of induction (3–6 h) (Supplementary Table S1). The highest expression yields were achieved with 0.5 mM IPTG and 5 h growth post-induction for both lectins. About 24 mg of Gal-1 and 11 mg of his-tagged Gal-1 were recovered per 100 mL of culture, respectively. The crude protein extracts of Gal-1 and his-tagged Gal-1 supplemented with protease inhibitor cocktail ware purified by affinity chromatography on α-lactose agarose column or combination of Ni-NTA His·Bind and α-lactose agarose column, respectively. It was found that higher purity of his-tagged Gal-1 was achieved when the purification was done first on the Ni-NTA His·Bind resin, followed up by the α-lactose agarose column. Dialysis against phosphate sodium buffer (20 mM phosphate, 10 mM NaCl and 10 mM β-mercaptoethanol, pH 7.2) allowed for complete removal of lactose. Western blot staining with an anti-Gal-1 and anti-his-tag antibody confirmed the purified lectin products (Fig. 1).

Figure 1.

(a) SDS-PAGE and (b) Western blots WT Gal-1 and His-tagged Gal-1.

Binding of Gal-1 and histidine-tagged Gal-1 to LacNAc, Lac, and ASF

Lectin binding activity was confirmed for both Gal-1 and his-tagged Gal-1 via titrations with LacNAc and lactose. Failure to remove bound lactose from purification due to incomplete dialysis and the potential for lectin insolubility due to disulfide crosslinking and/or hydrophobic interactions are common enough to warrant control titrations with LacNAc prior to performing experimental titrations with a dialyzed and concentrated lectin sample. LacNAc, lactose, and ASF titrations with Gal-1 returned K_d values and thermodynamic parameters in agreement with previously obtained values [36–39] and are provided in Table 1, Fig. 3a, Fig. S8a, and Fig. S9. The addition of the poly-histidine tag at the N-terminus did not appear to impede lectin activity, as LacNAc, lactose, and ASF titrations with his-tagged Gal-1 were in agreement with values obtained for Gal-1 and previously reported value [40]. The thermodynamic parameters for his-tagged Gal-1 binding to monovalent and multivalent control ligands are provided in Table 2, Fig. S8b, and Fig. S10 Binding stoichiometries (n-value) obtained for all binding titrations were reflective of titrations reaching full saturation. Gal-1 was treated as a dimer for concentration input, therefore n-values for LacNAc and lactose were 2, and n-values for ASF were 0.222.

Table 1.

Summary of thermodynamics of binding of Gal-1 to ligands at 25 °C calculated using the one-set-of-sites binding model.

Ligand	[Ligand] (mM)	[cell] (μM)	Ka (x104 M−1)	−ΔG (kcal/mol)	−ΔH (kcal/mol)	−TΔS (kcal/mol)	n	Kd (μM)
LacNAc	6.0	140	1.20	5.57	9.97 (± 0.151)	4.40	2.05	83.0 (± 3.13)
Lactose	8.0	80	0.313	4.77	11.9 (± 0.906)	7.09	2.07	319 (± 15.7)
ASF	218	61	17.4	7.15	65.4 (± 1.42)	58.2	0.222	5.76 (± 0.27)
TF	6.0	50	0.026	3.30	16.7 (± 0.221)	13.4	2.00a	3800 (± 61.2)
Thr-TF	6.0	50	0.025	3.28	16.6 (± 0.269)	13.3	2.00a	3960 (± 77.8)
MUC1-TF (Thr9)	1.0/4.0	140	-	-	-	-	-	n.d.b
MUC1-TF (Thr4,9,16)	1.0/6.0	140	-	-	-	-	-	n.d.b

^an-value is fixed.

^bnot determined (n.d.) due to low or no affinity and/or insufficient generation of heat - requisite to resolve a complete set of thermodynamic binding parameters.

Figure 3.

ITC titration profile of (a) LacNAc (6.0 mM), (b) TF (6.0 mM) and (c) MUC1-TF (Thr^4,9,16) (6.0 mM) binding to Gal-1 in phosphate-buffered saline (pH 7.2) containing 20 mM phosphate, 10 mM NaCl and 10 mM β-mercaptoethanol. The top panels show the experimental ITC data and bottom panels a fit to a one-site model of the binding data using MicroCal PEAQ-ITC analysis software. Resulting values for the stoichiometry (n), binding affinity (K_a), dissociation constant (K_d), enthalpy (ΔH), and change in entropy with respect to temperature (TΔS) are given in Table 1.

Table 2.

Summary of thermodynamics of binding of his-tagged Gal-1 to ligands at 25 °C calculated using the one-set-of-sites binding model.

Ligands	[Ligand] (mM)	[His-tag-Gal-1] (μM)	Ka (x104 M−1)	−ΔG (kcal/mol)	−ΔH (kcal/mol)	−TΔS (kcal/mol)	n	Kd (μM)
LacNAc	6.0	134	1.12	5.53	9.79 (± 0.206)	4.27	2.02	89.6 (± 4.32)
Lactose	8.0	128	0.246	4.63	8.67 (± 1.43)	4.04	2.05	406 (± 56.9)
ASF	0.525	200	25.32	7.37	57.5 (± 0.708)	50.1	0.283	3.95 (± 0.184)
ASF	0.700	200	34.01	7.55	57.1 (± 0.552)	49.6	0.274	2.94 (± 0.146)

Binding of Gal-1 and histidine-tagged Gal-1 to TF antigen, Thr-TF conjugate and MUC1-TF glycopeptides

Titrations of the TF disaccharide (Gal1–3βGalNAc) and the Thr-TF antigen (Gal1–3βGalNAcα1-O-Thr) revealed binding by Gal-1 in mM range (Table 1, Fig. 3b, and Fig. S11), in agreement with data previously reported by Bian et al. [16]. While significant structural similarities between the Gal-1 and Gal-3 CRDs are evident, the L4 loop between S4 and S5 β-stands in the Gal-1 CRD creates an inner fold that conformationally restricts TF access into the Gal-1 CRD - in essence, a more narrow cavity than the Gal-3, which contains different residues in the L4 loop motif [16]. Reported modeling based on crystal structure data further pinpointed His52 as the primary precluding residue to TF binding. His52 resides on a type-II β-turn, is stabilized by H-bonds between Ala51 and Asp54 and is essential for β-Gal recognition through C2’, O2’, and O4 stacking. Although steric interactions between His52 and the N-acetyl group of the TF disaccharide supposedly impede access into the CRD, mutagenesis of His52, shown by Bian et.al [16], enabled binding of TF, albeit high μM affinity, further supporting the significance of His52 [16].

Yet, Gal-1 recognition of the TF antigen at the cell surface has been proposed based on immunohistochemical analysis and cell-based binding assays [25, 27, 28, 41]. Similarly, our recombinantly expressed Gal-1 bound to the cell surface of breast cancer MCF7 cells (Fig. 2), known to express tumor-associated MUC1 decorated with TF antigen [42]. Binding of Gal-1 to cells was partially inhibited by the TF antigen-specific monoclonal antibody (A78-G/A7, Fig. S12). A consideration must therefore be made for factors that may elicit Gal-1 binding to TF at the molecular level but have been experimentally neglected in previous studies. A first consideration can be made of the role of multivalency in glycan-lectin interactions. Avidity often enables glycan recognition to translate into cellular functionality, and therefore this apparent K_d must be considered when studying glycan-binding interactions. The polyvalent glycan presentation of cancer-associated MUC1, which can extend 200–500 nm beyond the cell surface [43], promotes a „cluster glycoside effect“, wherein the effective local glycan concentration is increased, as is seen with the apparent K_d of multivalent binding interactions. Upregulated expression of both MUC1 and Gal-1 in the pathological context does also enhance the potency of the binding interactions; in the case of Gal-1, increased lectin concentrations promote the dimerization of CRD subunits and functional noncovalent cross-linking [44, 45].

Figure 2.

MCF-7 cells treated with a) PBS as control and b) his-tag Gal-1 (50 μM). NTA-Atto 488 probe was used for fluorescence detection of his-tag on Gal-1. The scale bars indicate 1000 μm.

The scaffold architecture must also be considered irrespective of density of the ligand, as it has been shown to significantly modulate lectin recognition of attached glycans [30, 34]. In the context of MUC1, the sequence of residues and attached glycans in the peptidic scaffold inform the entropic character of the amino acid backbone through secondary structure. The location and frequency of residues capable of N- or O-glycosylation inform the valency of the construct; while the identity of the residues flanking the glycosylation site directs the local solvation structure and may or may not directly interact with the glycan and/or lectin. In all, the scaffold architecture lends specificity to the interaction and contributes to the thermodynamic driving force of the interaction.

We previously revealed a thermodynamic switch in the binding of Gal-3 to TF and MUC1 glycopeptides, wherein the binding interaction shifted to an entropy-driven mechanism for the MUC1 glycopeptides, tied to a nearly 10-fold increase in affinity [34]. Considering this shift, we prepared two TF containing glycopeptides to explore the potential contribution of the MUC1 peptidic scaffold and/or multivalency to binding by Gal-1. The heavily glycosylated extracellular domain of MUC1 is defined by extended tandem repeats of 20 amino acids (HGVTSAPDTRPAPGSTAPPA); peptides were constructed as a single 20 residue sequence. Of five possible glycosylation sites (Thr4, Ser5, Thr9, Ser15, Thr,16), we chose a single glycosylated construct at Thr9 (Fig. 4), which is located within a highly immunogenic region containing the sequence PDTR [46, 47], and a triple glycosylated construct at Thr4, Thr9, and Thr16 (Fig. 4). ITC titrations and thermodynamic analysis did not reveal evidence of either MUC1 peptide scaffold or multivalency playing a role in an affinity boost to either glycopeptide (Table 1, and Fig. 3c). On the contrary, the binding affinity (K_d), of Gal-3 for triple glycosylated MUC1 fragment (Fig. S13), was 30–35 times lower (K_d [8.4 μM]) than that of either Thr-TF or free TF disaccharide [34, 48]. Furthermore, the apparent K_d observed between Gal-1 and ASF, with a glycan valency of nine, remained in the uM range (Table 1), suggesting that the valency of our glycopeptide constructs would need to be significantly higher to observe a binding interaction in solution, confirming that Gal-3 and Gal-1 exhibit differential recognition of TF antigen presented on MUC1. Additionally, the fact that our experimental set-up, while highly sensitive, is completely solution-based may introduce an extrinsic bias that strays from the endogenous environment in which MUC1-TF and Gal-1 may interact. Noteworthy, the conversion of non-covalent association of two Gal-1’s CRDs to covalent connection, by short (Gly-Gly) linker, provided no robust evidence for cooperativity for its cognate ligands (LacNAc/Lac) [33]. The data obtained for the glycoengineered Gal-1/Gal-3 heterodimers, analyzed by the sequential model, derived estimates for the affinity of the Gal-1 CRD for TF to be in high mM range at full loading [48]. In summary, it is feasible to speculate that Gal-1 requires a cell surface density of ligand presentation above a certain threshold for high-affinity binding, that is currently not achieved with the synthetic model MUC1-glycopeptides bearing TF antigen (one or all three Thr-site occupied) at the molecular level. Considering the diversity of presence of glycoproteins and its complexity within the cellular glycocalyx, the formation of unique two- and three-dimensional cross-linked lattices seems a likely plausible explanation. Atomic force microscopy studies suggest that the non-covalent association of Gal-1’s CRD appears to be more suited for cis-cross-linking and transient trans-bridging than for establishing firm contacts [49]. In all, we have shown that TF presented with low valency on a short peptidic mimic of the native MUC1 scaffold failed to entice recognition by Gal-1 in solution at the level of affinity expected for functional significance.

Figure 4.

MUC1 glycopeptide containing 20 amino acid tandem repeat HGVTSAPDTRPAPGSTAPPA.

AlphaScreen assay development

The methodology used by Yegorova et al. [35] for the development of a competitive binding AlphaScreen assay tuned for both the discovery of inhibitors of low affinity lectin-glycan interactions and as a quantitative validation by IC₅₀, was applied for Gal-1. The experimental versatility of AlphaScreen technology is well-known [50, 51]. His-tagged Gal-1 and biotin-tagged ASF were incubated with streptavidin donor beads and nickel chelate (Ni-NTA) acceptor beads. Binding of Gal-1 to ASF would bring the respective conjugated beads into proximity for the transfer of singlet ¹O₂ and consequent acceptor bead emission, detection of which would validate binding. While experimental titrations for our thermodynamic analyses were performed in phosphate buffer containing β-mercaptoethanol, control titrations with LacNAc and ASF were additionally performed in HEPES buffer containing 0.05% Tween-20 prior to sample use in the assay to confirm activity. Dissociation constants for binding of Gal-1 to LacNAc and ASF (not given) mirrored the values obtained in phosphate buffer.

Despite AlphaScreen technology affording a highly sensitive assay format with low background signal generation, we were unable to demonstrate sufficient signal separation to qualify the assay for further application in a competitive assay format. The signal-to-background (S/B) ratio was determined using the average maximum signal divided by the average background signal (beads only). We previously reported S/B ratios in the several hundreds for the Gal-3/ASF binding format, wherein we only obtained S/B ratios in the single digits for Gal-1/ASF binding (Fig. S14). While Gal-3 affinity for LacNAc is ca 2x greater than of Gal-1, Gal-3 affinity for ASF with three triantennary LacNAc residues is ca 12x greater than of Gal-1 (Table 1 and Fig. S15). It is important to consider the variable functional significance of multivalent ligand presentation on lectin binding affinity, which is highly dependent on lectin architecture and monomer crosslinking/aggregation dynamics. Gal-3 and Gal-1 are inherently different in this regard [48]. Therefore, we can speculate that the added affinity afforded by the presentation of nine LacNAc residues on a single ASF unit was inadequate to afford high affinity binding between Gal-1 and ASF required to facilitate sufficient S/B separation for the application of this assay in a competitive format. A binding partner with sub-micromolar affinity and also capable of linker conjugation must therefore be explored to advance this particular assay format for Gal-1.

Conclusion

Gal-1 maintained previously reported binding and thermodynamic profiles with LacNAc, lactose, and ASF, yet failed to recognize TF disaccharide and Thr-TF (TF antigen) beyond the millimolar affinity at the molecular level by ITC method. Similarly, neither presentation of TF on MUC1 tandem repeat peptide scaffold, or density (trivalent presentation) played a role in increasing the binding affinity for Gal-1, warranting further studies into the mechanistic nature of TF recognition by Gal-1 observed at the cell surface. Insufficient valency of our glycopeptide constructs as well as the extrinsic influence of our solution-based method rather than a more native ligand presenting method, that allows for cross-linking, may prove to be the most upfront barriers to detecting a binding interaction. The observed differences between the Gal-1 and Gal-3 in binding to MUC1 glycopeptides may aid in design of novel high-affinity and specificity inhibitors of these two lectins. Additionally, the low micromolar affinity of Gal-1 for ASF did not create sufficient separation in signal between the binding population and the control, directing us to pursue a sub-micromolar affinity binding partner as the next-step for this assay development.