doi:10.1016/j.carres.2007.02.034
Copyright © 2007 Elsevier Ltd All rights reserved.
The conformation of the C-glycosyl analogue of N-acetyl-lactosamine in the free state and bound to a toxic plant agglutinin and human adhesion/growth-regulatory galectin-1
Víctor García-Apariciod, Matthieu Sollogouba, b, Yves Blériota, b, Virginie Collioua, b, Sabine Andréc, Juan L. Asensiod, F. Javier Cañadad, Hans-Joachim Gabiusc, Pierre Sinaÿa, b and Jesús Jiménez-Barberod,
, 
bUniversité Pierre et Marie Curie-Paris 6, Institut de Chimie Moléculaire (FR 2769), 75005 Paris, France
cInstitut für Physiologische Chemie, Tierärztliche Fakultät, Ludwig-Maximilians-Universität, Veterinärstr. 13, 80539 München, Germany
dEcole Normale Supérieure, Département de Chimie, UMR CNRS 8642, 24, rue Lhomond, 75231 Paris Cedex 05, France
Received 16 January 2007;
revised 19 February 2007;
accepted 21 February 2007.
Available online 4 March 2007.
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
Abstract
The conformational behavior of the C-glycoside analogue of N-acetyl-lactosamine, β-C-Gal-(1→4)-β-GlcNAc-OMe, 1, has been studied using a combination of molecular mechanics calculations and NMR spectroscopy (J and NOE data). It is shown that the C-disaccharide populates three distinctive conformational families in solution, the major one being the anti-ψ conformation. Of note, this conformation is only marginally populated for the O-disaccharide. Due to its conspicuous role in the regulation of adhesion, growth and tissue invasion of tumors and its avid binding to N-acetyl-lactosamine human, galectin-1 was tested as a receptor. This endogenous lectin recognizes a local minimum of 1, the syn-ΦΨ conformer, and thus a conformational selection process is correlated with the molecular recognition event.
Graphical abstract
Galectin-1 selects a local minimum of the conformational distribution of the C-glycosyl analogue of N-acetyl lactosamine.
Keywords: C-Glycosides; Conformational analysis; Glycomimetics; Molecular recognition; NMR
Figure 1. Positions of local minima for 1, the respective probability distribution, and key interproton distances (between 2.0 and 2.5 Å) for the Φ/Ψ torsion angles. The global minimum, C, is located within the anti-Ψ region.
Figure 2. Stereoscopic representation of the different minima, A (180/0), B (36/−72), C (54/18) and D (54/−180), of the C-glycosyl analogue of N-acetyllactosamine (1).
Figure 3. 1D-NOESY spectra (400 ms mixing time) of 1 at 500 MHz and 300 K in D2O, after selective inversion of H1Gal (A) and H2Gal (B). (C) Enlargement of the key area of the 2D-T-ROESY spectrum (400 ms mixing time) for clarity.
Figure 4a. The 500 MHz 1H NMR spectrum of 1 in the free state (bottom) and in the presence of a 5% molar ratio of VAA (top trace). Ligand–protein ratio is 20:1 and the temperature 298 K. Ligand concentration is 2 mM.
Figure 4b. The 500 MHz STD spectrum after 2 s of on- and off-resonance saturation time (top trace) of 1 in the presence of VAA at a 50:1 ligand–protein molar ratio. Ligand concentration is 2 mM. The major epitope corresponds to the Gal residue. The regular 1D spectrum of 1 is shown below for comparison.
Figure 4c. Representation of the decay of magnetization in an inversion recovery experiments for the estimation of selective T1 values for H1Gal of 1 in the free state and for a sample containing a 20:1 molar ratio of 1:VAA. Ligand concentration is 2 mM. The value for the bound state is considerably shortened indicating binding of 1 to VAA.
Figure 4d. STD Values measured for 1 in the presence of VAA (50:1 ligand–protein molar ratio). Ligand concentration is 2 mM.
Figure 5a. The 1H NMR spectrum of 1 in the free state (bottom) and in the presence of a 5% molar ratio of h-Gal 1 (top trace). Ligand–protein ratio is 20:1. Ligand concentration is 2 mM.
Figure 5b. Representation of the decay of magnetization in an inversion recovery experiments for the estimation of selective T1 values for H1Gal of 1 in the free state and for a sample containing a 20:1 molar ratio of 1:h-Gal-1. Ligand concentration is 2 mM. The value for the bound state is considerably shortened indicating binding of 1 to h-Gal-1.
Figure 6. Top: NOESY spectrum (mixing time 700 ms) for free 1 in D2O solution. Cross peaks have different sign to diagonal peaks. Middle: TRNOESY spectrum (mixing time 200 ms) for a 20:1 molar ratio of 1: hGal-1 (0.1 mM of lectin). Cross-peaks have the same sign as diagonal peaks. No H1Gal–H3GlcNAc contact is observed, indicating that the anti-Ψ conformer is not bound by this lectin. Bottom: Expansion of the key region of the TRNOESY spectrum. Preferential recognition of the syn-Ψ conformer is evident due to the presence of a strong H1Gal-H4 GlcNAc cross-peak.
Table 1.
1H and 13C chemical shifts (δ, ppm) and key coupling constants (Hz) for 1 (D2O, pH 7.0), at 500 MHz and 300 K

Table 2.
Torsion angle values, relative steric energies and relative population densities of the major conformational low-energy positions of 1 according to mm3* calculations

Energy values are given in kJ mol−1.
Table 3.
J Values (Hz) for the local conformational minima and corresponding Φ, Ψ values for 1

The values were deduced by applying the generalized Karplus equation proposed by Altona to the derived molecular geometries derived from mm3* calculations.
Table 4.
Key interproton distances (Å) for the different conformers, the ones corresponding to putative strong NOEs being bolded

The protons belonging to the Gal residue are primed.
Table 5.
Average distances and calculated NOEs (estimated by applying a full matrix relaxation approach to the ensemble averaged distances computed from the probability distribution map shown in Fig. 1) in comparison with the observed NOEs for compound 1

The calculations are shown for a NOE mixing time of 600 ms. The protons belonging to the Gal residue are primed.