Synthesis of Bivalent and Monovalent Sugar Ligands, their Interfacial and Solution Phase Lectin Bindng Studies

Abstract

Carbohydrate-protein interactions are responsible for several biological functions. While these interactions maintain high levels of specificities, the binding strength of individual carbohydrate-protein recognitions are weak, with dissociation constants (Kd) ~10-3-10-6 M. In order to increase the binding strengths meaningful to physiological functions, multivalent, clustered patches of carbohydrate ligands are required. Synthetic glycoclusters contribute in a significant manner to understand the fine details of the weak carbohydrate-protein interactions. The extent of clustering of the ligands, spatial, topological orientations and the nature of the scaffolds are prominent issues to address the carbohydrate-protein interactions in general. Chapter 1 of the Thesis presents a summary of the synthetic cluster glycosides, mechanisms and energetics of their interactions with lectins. The presence of several ligands within the molecular scaffold is not sufficient, rather there exists a critical demand on the spatial disposition of the individual ligands in the multivalent ligand system to achieve enhanced binding affinities. In order to assess the multivalent effects, influence of linkers and the spatial disposition of the ligands, a systematic study was undertaken, involving a series of the most minimal of the multivalent sugar ligand system, namely, the bivalent sugar ligands. In a programme, it was desired to study the bivalent and monovalent sugar ligand-lectin interactions in a two-dimensional membrane model system. An appropriate model system was the Langmuir monolayer formations of the sugar ligands and their recognitions of the lectins at the interface. A series of bivalent and monovalent glycolipids were thus designed and synthesized. Molecular structure of the ligands utilized to study the lectins binding behavior at the air-water interface are presented in Figure 1. The sugar density dependent lectin binding at the air-water interface caused by the glycolipids was studied in detail. Prior to lectin binding studies, the monolayer behavior of the glycolipids (GL), non-sugars (NS) and their mixtures were assessed. It was observed that the apparent molecular areas of the mixed monolayers increased with increasing percentage of the glycolipid in the mixed monolayer. Interactions of the glycolipid mixed monolayers with lectin were assessed at a constant surface pressure of 10 mN/m. The adsorption kinetics of the lectin concanavalin A (Con A) with the mixed monolayers was monitored by the surface area variation (ΔA) as a function of time.

The detailed studies showed a maximum increase in ΔA of 10% of the bivalent glycolipids in the mixed monolayer and a ΔA of 20% of the monovalent glycolipids (Figure 2). With both bivalent and monovalent glycolipids, change in the area per molecule had decreased progressively with higher percentage of the glycolipids in the monolayers. On the other hand, with ethylene glycol spacers, the relative responses and the amount of bound lectin increased.

Figure 2. Ligand-lectin interactions at the air-water interface as a function of the percentage of (a) bivalent glycolipids and (b) monovalent glycolipids in the mixed monolayers. To verify the specificity of these interactions, the mannopyranoside non-specific lectin, namely, wheat germ agglutinin (WGA) was tested and there were no deviations in the ΔA for various ratios of the sugar–non-sugar mixed monolayers. The study established that (i) maximal binding of the lectin to the bivalent glycolipids occurred at lower sugar densities at the interface than that for the monovalent glycolipids and (ii) the surface presenting sparsely populated sugar residues are efficient for a lectin binding. Chapter 2 presents the details of synthesis and ligand-lectin interactions at the air-water interface, relevant in the two-dimensional membrane model system. A study of the multivalent effects originating through glycolipid micelles and their lectin interactions was undertaken in another programme. The kinetic studies of the glycolipid micelles-lectin interactions were conducted with the aid of surface plasmon resonance (SPR) technique. Prior to the SPR studies, the critical micellar concentration (CMC), aggregation number and the hydrodynamic diameter of each glycolipid (GL-1 to GL-6, Figure 1) micelles were determined. The glycolipid micelles were used as the analytes on a Con A immobilized surface. The sensorgrams obtained for the interaction of the various glycolipid micelles with Con A are presented in Figure 3.

Figure 3. Sensorgrams obtained for the binding of various glycolipids micelles to a Con A immobilized surface, at a constant glycolipid concentration of 250 µM. The kinetic studies of the interactions were performed and the analysis showed that the bivalent analyte model provided a better fitting for the interaction sensorgrams. The analysis revealed that the ka1/kd1 values remained largely uniform for all the glycolipids, whereas the ka2/kd2 values were about two orders of magnitude larger for the bivalent glycolipid (GL-4 to GL-6) micelle-lectin interactions than for the monovalent series (GL-1 to GL-3) (Table 1). From the SPR studies, it emerged that the additional sugar unit in the bivalent glycolipid micelles provided a favorable complexation between the sugar ligand and the lectin. Further, the glycolipid micelles mediated layer-by-layer Con A multilayer formation was also studied by SPR and atomic force microscopy (AFM) methods. Chapter 3 provides the SPR studies of glycolipid micelles-lectin interactions.

A study of the monomolecular recognitions of the mono- and bivalent sugar ligands 1-8 (Figure 4) to a lectin was undertaken subsequently. The kinetic studies of the bivalent vs monovalent ligands during lectin binding were conducted by employing the SPR technique, for which the sugar ligands 1-6 were used as the analytes on a lectin coated sensor surface. Figure 4. Structures of the mono- and bivalent sugar ligands 1-8 and the NS derivative. The following observations were made from the SPR analysis. (i) Within the mono- and bivalent series, the response units increased in the series 1–3 and 4–6; (ii) the equilibrium responses were attained within 105 seconds in the monovalent ligands and (iii) the association response gradually increased for the bivalent ligands 5 and 6 and reached an equilibrium after ~3 min. An important outcome of the kinetic studies was the identification of ka and kd for the monomolecular interactions, that were distinctly different for the bivalent ligands. Specifically, the ka was significantly faster and kd was slower for bivalent sugar ligands, in comparison to the monovalent sugar ligands (Table 2). Table 2. SPR derived kinetic parameters for the interactions of sugar ligand to a Con A immobilized surface at 25 oC.

Isothermal titration calorimetry (ITC) studies were also conducted, in order to correlate the functional valencies and the thermodynamic parameters. The studies were conducted at ligand concentrations much below their CMCs. The general observations from the ITC studies were that the binding site saturations were slower for the monovalent sugar ligands, in comparison to the bivalent sugar ligands. It was observed that the binding affinities of bivalent ligands 5 and 6 enhanced ~5 times higher than the monovalent ligands 2 and 3 (Table 3). The effective linker length, which allowed the sugar ligands to be functionally active, was determined to be ~15 Å and this separation was necessary for the intermolecular cross-linking formation. The dynamic light scattering (DLS) study of the bivalent ligands 5 or 6-lectin complexes showed the presence of intermolecular cross-linked complexes that existed in solution from the initial stages of the binding process. Upon realizing the nanometric diameters of the sugar ligand-lectin complex, an attempt was undertaken to visualize the complexes by transmission electron micoscopy (TEM). In TEM, 4-Con A complex exhibited particle sizes in the range of 5-10 nm, matching nearly the size of the lectin alone. On the other hand, 5–Con A and 6–Con A complexes provided sizes varying between 20¬150 nm. These particle sizes corresponded to similar aggregate sizes derived from the DLS studies. Chaper 4 describes the kinetic, thermodynamic, DLS and TEM studies of sugar ligand-lectin intearctions. Table 3. Binding stoichiometries and thermodynamic parameters of the sugar ligand-Con A interactions at 25 oC.a Ligand n Ka (x 10 -4) ΔG ΔH TΔS 1 0.91 9.14 ( ± 0.75) -6.76 -3.39 3.37 2 1.01 5.76 (± 0.80) -6.49 -3.98 2.51 3 1.09 7.06 (± 1.23) -6.61 - 3.01 3.60 4 1.10 5.75 (± 0.27) -6.49 - 6.39 0.10 5 0.50 20.6 (± 1.7) -7.59 - 12.80 -5.21 6 0.47 37. 4 (± 2. 4) -7.61 -11.54 -3.93 7 1.03 0.86 (± 0.06) -5.36 -7.9 -2.62 8 1.05 2.48 (± 0.12) -5.99 -6.3 -0.32 MeαMan 1.04 0.79 (± 0.04) -5.27 -7.83 -2.56

Ka is in the unit of M-1; ΔG, ΔH and TΔS are in the units of kcal mol-1. Errors in ΔG are ~1-4%. Errors in ΔH are in the range of 1-8%. Errors in TΔS are in the range of 1-6 %. A study was undertaken further to assess the kinetic interactions of the tumor-associated T-antigen with a lectin. Synthesis of amine-tethered T-antigen and lactose derivatives (Figure 5) were accomplished and an assessment of their kinetic interactions with lectin peanut agglutinin (PNA) was conducted. Figure 5. Structures of the amine-tethered T-antigen, lactose and mannose derivatives. The lectin PNA was used as the analyte onto the sugar ligand immobilized surfaces. It was found that the interaction with T-antigen showed higher response units than the lactose derivative (Figure 6). The kinetic studies of PNA with immobilized T-antigen and the lactose derivatives demonstrated that the binding followed a bivalent analyte model of the interaction. The T-antigen derivative interacted with the lectin and relatively faster association (ka) and a slower dissociation (kd) were observed, in comparison to the lactose derivative. The ratio of second binding kinetic constants (ka2/kd2) was observed higher than the first binding kinetic constants (ka1/kd1). Further, the ITC studies were conducted, in order to provide the thermodynamic parameters governing the lectin-T-antigen interactions. The combined approach of SPR and ITC studies showed that the T-antigen derivative exhibited a higher binding affinity to PNA than the lactose derivative. Chapter 5 presents synthesis of the T-antigen and lactose derivatives and studies of their lectin interactions.

In summary, the thesis provides a detailed insight into the kinetic and thermodynamic parameters of the bivalent sugar ligand-lectin interactions, in comparison to the monovalent sugar ligands. Langmuir monolayer formation of the sugar ligands and the assessment of their lectin binding at the air-water interface demonstrated that the surface presenting sparsely populated sugar residues are efficient for a lectin binding. The kinetic studies of various glycolipid micelles-lectin interactions showed that the additional sugar unit in the bivalent glycolipid micelles provided a favorable complexation between the sugar ligand and the lectin. The detailed monomolecular kinetic studies showed that the bivalent sugar ligands underwent a faster association (kon) and a slower dissociation (koff) of the ligand-lectin complexes. The ITC studies on sugar ligand-lectin interactions led to identify not only the thermodynamic parameters, but also the influence of the hydrophobic alkyl units and the linker moieties. The DLS and TEM characterizations of sugar ligand-lectin complexes showed the status of the complexation, sizes and the morphologies. The studies were extended further to tumor associated T-antigen-lectin interactions. Overall, the Thesis establishes the most minimal multivalent sugar ligands, namely, the bivalent sugar ligand-letin interactions. The studies presented in the Thesis should be useful to design multivalent sugar ligands for highly avid lectin interactions and also to raise possibilities for the construction of defined lectin oligomers, facilitated through the multivalent sugar ligand-lectin cross-linking interactions.