Differential Solvent DEEP-STD NMR and MD Simulations Enable the Determinants of the Molecular Recognition of Heparin Oligosaccharides by Antithrombin to Be Disentangled

Citation: Parafioriti, M.; Elli, S.; Muñoz-García, J.C.; Ramírez-Cárdenas, J.; Yates, E.A.; Angulo, J.; Guerrini, M. Differential Solvent DEEP-STD NMR and MD Simulations Enable the Determinants of the Molecular Recognition of Heparin Oligosaccharides by Antithrombin to Be Disentangled. Int. J. Mol. Sci. 2024, 25, 4669. https://doi.org/10.3390/ ijms25094669 Academic Editor: Cristina Airoldi Received: 28 February 2024 Revised: 16 April 2024 Accepted: 20 April 2024 Published: 25 April 2024 Copyright: © 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). International Journal of Molecular Sciences Article Differential Solvent DEEP-STD NMR and MD Simulations Enable the Determinants of the Molecular Recognition of Heparin Oligosaccharides by Antithrombin to Be Disentangled Michela Parafioriti 1, Stefano Elli 1, Juan C. Muñoz-García2, Jonathan Ramírez-Cárdenas 2, Edwin A. Yates 3,4, Jesús Angulo 2,* and Marco Guerrini 1,* 1Istituto di Ricerche Chimiche e Biochimiche “G. Ronzoni”, Via Giuseppe Colombo 81, 20133 Milano, Italy; [email protected] (M.P.); [email protected] (S.E.) 2Instituto de Investigationes Químicas (IIQ)-Consejo Superior de Investigaciones Científicas (CSIC), Avenida Americo Vespucio 49, 41092 Sevilla, Spain; [email protected] (J.C.M.-G.); [email protected] (J.R.-C.) 3Department of Biochemistry and Systems Biology, Institute of Systems, Molecular and Integrative Biology, University of Liverpool, Liverpool L697ZB, UK; [email protected] 4Centre for Glycoscience, Keele University, Newcastle-Under-Lyme ST55BG, UK *Correspondence: [email protected] (J.A.); [email protected] (M.G.) Abstract: The interaction of heparin with antithrombin (AT) involves a specific sequence corresponding to the pentasaccharide GlcNAc/NS6S-GlcA-GlcNS3S6S-IdoA2S-GlcNS6S (AGA*IA). Recent studies have revealed that two AGA*IA-containing hexasaccharides, which differ in the sulfation degree of the iduronic acid unit, exhibit similar binding to AT, albeit with different affinities. However, the lack of experimental data concerning the molecular contacts between these ligands and the amino acids within the protein-binding site prevents a detailed description of the complexes. Differential epitope mapping (DEEP)-STD NMR, in combination with MD simulations, enables the experimental observation and comparison of two heparin pentasaccharides interacting with AT, revealing slightly different bound orientations and distinct affinities of both glycans for AT. We demonstrate the effectiveness of the differential solvent DEEP-STD NMR approach in determining the presence of polar residues in the recognition sites of glycosaminoglycan-binding proteins. Keywords: ligand-based NMR; STD NMR; DEEP-STD NMR; MD simulations; radial distribution function; molecular recognition; glycosaminoglycan–protein interactions 1. Introduction Heparin is a long linear negatively charged polysaccharide belonging to the glycosaminoglycan (GAG) family [ 1 ]. It is composed of repeating disaccharide units comprising uronic acid ( β -D-glucuronic acid or α -L-iduronic acid) (1-4)-linked to α -D-glucosamine. The uronic acids can be 2-O-sulfated, while the glucosamines are N-acetylated or N-sulfated, 6-O-sulfated and, more rarely, 3-O-sulfated [ 2 ]. In solution, the glucosamine (GlcN) and glucuronic acid (GlcA) residues adopt a 4 C 1 chair conformation, while the iduronic acid (IdoA) residues exist in equilibrium between the 4 C 1 chair, 1 C 4 chair and 2 S 0 skew-boat forms [ 3 – 5 ]. The population of each IdoA conformer depends on the degree of sulfation of the iduronic acid and the adjacent glucosamines [ 6 ]. Endogenous heparin is located within mast cell granules, where it interacts with histamine, proteases and inflammatory mediators, controlling their storage, retention and activation. Exogenous heparin, derived from porcine and bovine tissues, is used as an anticoagulant drug, due to its ability to bind and regulate the activity of several factors involved in the blood clotting cascade [ 7 ]. Specifically, heparin binds and activates antithrombin (AT), accelerating its inhibitory effect on the enzymes of the coagulation system, such as factor IIa (thrombin) and factor Xa. Int. J. Mol. Sci. 2024,25, 4669. https://doi.org/10.3390/ijms25094669 https://www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2024,25, 4669 2 of 19 This interaction is crucial in the prevention and treatment of thromboembolic disorders, avoiding the excessive formation of blood clots [8–10]. In this context, the study of GAG–protein complexes is fundamental for understanding the molecular mechanisms underlying GAG–protein interactions and, therefore, designing innovative glycomimetic drugs [11,12]. Ligand-based nuclear magnetic resonance (NMR) experiments, including transferred nuclear Overhauser effect spectroscopy (tr-NOESY) and saturation transfer difference (STD) NMR, allow definition of the bioactive conformation and the epitope mapping of the ligand; however, they provide no information about the protein-binding site [ 13 – 16 ]. A novel multifrequency and multisolvent STD NMR approach, known as differential epitope mapping (DEEP)-STD NMR, was developed by Angulo and co-workers to gain details on the nature of the protein residues (aliphatic, aromatic or polar) involved in the molecular recognition process [ 17 ]. This new approach exploits the differences in the ligand epitope mapping that arise from acquiring a pair of STD NMR experiments under two different conditions, such as different frequencies or solvents. The multifrequency STD NMR, conducted by irradiating the aliphatic and aromatic residues of the protein, permits determination of the aliphatic or aromatic nature of the amino acid side chains in contact with the ligand. Ligand protons close to directly irradiated protein residues receive a higher degree of saturation, showing a relative increase in their STD intensities. Multisolvent STD NMR, performed in both D 2 O and H 2 O, allows identification of polar residues in the protein-binding site. In D 2 O, the exchangeable protons of polar residues are replaced by deuterium, which is unable to transfer proton saturation; in H 2 O, these protons may allow an additional transfer of saturation, depending on their exchange rate with the bulk water [ 18 ]. If ligand protons are in proximity to slow exchanging polar residues, they receive an extra saturation in the STD NMR experiments conducted in H 2 O, showing a relative increase in their STD intensities. Theoretical methods, such as atomistic molecular dynamics (MDs) simulations using state-of-the-art force-fields and the 3D structure of GAG–protein complexes obtained by X-ray crystallography or cryogenic electron microscopy (cryo-EM), are crucial to reconstruct the details of the protein residues and intermolecular forces that drive GAG–protein interactions [ 19 , 20 ]. Early MD simulations of a complex between a specific heparin pentasaccharide sequence and AT were run by Grootenhuis et al. (1991) and by Verli et al. (2005), spanning simulation times of less than 1 ns [ 21 , 22 ]. For the first time, they described the contribution of the amino acids of the protein recognition site to ligand binding, identifying that electrostatic forces drive the interaction and sketching the mechanism of allosteric activation. Later, Balogh et al. (2020) applied Gaussian accelerated MD simulations for a deeper analysis of the mechanism of allosteric activation [ 23 ]. Sarkar et al. (2016) underlined the solvation energy as one of the key aspects of GAG–protein affinity [24]. Our group applied sub-microsecond MD simulations to unravel the specificity of the interaction between heparin oligosaccharides and AT, successfully complementing the NMR interaction evidence [ 25 , 26 ]. Comprehensive reviews introduce and guide readers on how MD simulations are used to determine local and global structural and conformational properties of GAGs in their unbound and protein-bound states. Almond et al. (2018) gives a wide perspective of theoretical methods for the structural and dynamic aspects of GAGs, underlining the multiscale modelling from mono- to polysaccharides and the timescale from nano- to microseconds [ 27 ]. While microsecond MD simulation is a generally desirable target in GAG–protein complex descriptions, some molecular aspects, characterized by short relaxation periods, can also be estimated using sub-microsecond MD simulations. This was verified by the interglycosidic conformations of heparin oligosaccharides in the bound state with FGF-1 by Bojarski et al. (2019) [28]. Several studies have been performed by our group to elucidate the intricate molecular recognition processes underlying the interactions between GAGs and their respective receptors, with a specific focus on the interaction between heparin and AT. Within this context, our investigations revealed a noteworthy phenomenon wherein the IdoA2S unit of Int. J. Mol. Sci. 2024,25, 4669 3 of 19 the heparin pentasaccharide, a region specifically recognized by AT, undergoes a significant conformational shift toward the 2 S 0 skew-boat conformation upon binding [ 29 ]. This conformational change was found to be independent of the sulfation state of the iduronate unit and the specific position of the pentasaccharide along the heparin chain [ 25 , 30 ]. Despite the valuable insights gained from these studies, the absence of experimental data concerning the molecular contacts between the ligand and the amino acids within the protein-binding site has prompted the exploration of advanced techniques. In this regard, the application of DEEP-STD NMR emerges as a promising avenue. In this work, we explored the ability of DEEP-STD NMR, carried out in different solvents (D 2 O and H 2 O), to identify the presence of polar residues in the heparin-binding site of AT, which is mainly constituted by arginine (R) and lysine (K) residues (K11, R13, R46, R47, K114, K125 and R129) [ 31 – 33 ]. It is known that AT binds to heparin recognizing a specific pentasaccharide sequence [D-GlcNS6S α (1-4) D-GlcA β (1-4) D-GlcNS3S6S α (1-4) L-IdoA2S α (1-4) D-GlcNS6S α (1-4)] present in about 30% of its chains [ 34 , 35 ], even if some degeneracy of this sequence has been observed in limited cases [ 25 , 26 ]. Therefore, two structurally related heparin pentasaccharides (1) and (2) were selected for comparison as ligand probes (Figure 1). Int. J. Mol. Sci. 2024, 25, x FOR PEER REVIEW 3 of 20 Within this context, our investigations revealed a noteworthy phenomenon wherein the IdoA2S unit of the heparin pentasaccharide, a region specifically recognized by AT, undergoes a significant conformational shift toward the 2 S 0 skew-boat conformation upon binding [29]. This conformational change was found to be independent of the sulfation state of the iduronate unit and the specific position of the pentasaccharide along the heparin chain [25,30]. Despite the valuable insights gained from these studies, the absence of experimental data concerning the molecular contacts between the ligand and the amino acids within the protein-binding site has prompted the exploration of advanced techniques. In this regard, the application of DEEP-STD NMR emerges as a promising avenue. In this work, we explored the ability of DEEP-STD NMR, carried out in different solvents (D 2 O and H 2 O), to identify the presence of polar residues in the heparin-binding site of AT, which is mainly constituted by arginine (R) and lysine (K) residues (K11, R13, R46, R47, K114, K125 and R129) [31–33]. It is known that AT binds to heparin recognizing a specific pentasaccharide sequence [D-GlcNS6S α (1-4) D-GlcA β (1-4) D-GlcNS3S6S α (1- 4) L-IdoA2S α (1-4) D-GlcNS6S α (1-4)] present in about 30% of its chains [34,35], even if some degeneracy of this sequence has been observed in limited cases [25,26]. Therefore, two structurally related heparin pentasaccharides (1) and (2) were selected for comparison as ligand probes (Figure 1). Figure 1. Chemical structure of the heparin pentasaccharides selected as AT binders. The ligand units are depicted in their prevalent conformation in the unbound state: the GlcN and GlcA units adopt the 4 C 1 form, while the IdoA2S and IdoA units are in equilibrium between the 1 C 4 and 2 S 0 forms. The structure of pentasaccharide (2) differs from that of pentasaccharide (1), which corresponds to the heparin sequence that is specifically recognized by AT, only through the absence of the 2-O-sulfate group in the iduronate unit (Figure 1). Owing to the absence of the 2-O-sulfate group in the iduronic acid unit, pentasaccharide (2) possesses a lower affinity than pentasaccharide (1) [36–38]. The binding of pentasaccharide (1) to AT is welldescribed: in the first stage, the stiff non-reducing trisaccharide of pentasaccharide (1), attracted by the positively charged surface of the AT-binding site, induces a protein conformational change (AT activation) that allows, in the second stage, interaction with the flexible reducing disaccharide [39–41]. This phase is also assisted by the conformational flexibility of the iduronic acid unit, which adopts a pure 2 S 0 skew-boat conformation upon binding, enhancing the contacts. Since the reducing disaccharide of Figure 1. Chemical structure of the heparin pentasaccharides selected as AT binders. The ligand units are depicted in their prevalent conformation in the unbound state: the GlcN and GlcA units adopt the 4C1form, while the IdoA2S and IdoA units are in equilibrium between the 1C4and 2S0forms. The structure of pentasaccharide (2) differs from that of pentasaccharide (1), which corresponds to the heparin sequence that is specifically recognized by AT, only through the absence of the 2-O-sulfate group in the iduronate unit (Figure 1). Owing to the absence of the 2-O-sulfate group in the iduronic acid unit, pentasaccharide (2) possesses a lower affinity than pentasaccharide (1) [ 36 – 38 ]. The binding of pentasaccharide (1) to AT is well-described: in the first stage, the stiff non-reducing trisaccharide of pentasaccharide (1), attracted by the positively charged surface of the AT-binding site, induces a protein conformational change (AT activation) that allows, in the second stage, interaction with the flexible reducing disaccharide [ 39 – 41 ]. This phase is also assisted by the conformational flexibility of the iduronic acid unit, which adopts a pure 2 S 0 skew-boat conformation upon binding, enhancing the contacts. Since the reducing disaccharide of pentasaccharide (1) interacts with AT in an area of the binding pocket with the highest density of arginine residues, we expected to observe a relative increase in the STD intensities of the ligand protons belonging to this moiety in the STD NMR experiment conducted in H 2 O. The same behavior was also hypothesized for pentasaccharide (2), whose binding mode should be Int. J. Mol. Sci. 2024,25, 4669 4 of 19 similar, albeit not exactly analogous, to that of pentasaccharide (1). Additionally, differences in the conformations of pentasaccharides (1) and (2) in the AT-bound state were investigated by tr-NOESY experiments combined with docking and explicit solvent MD simulations covering 1 µs. Finally, the methodology based on the reduced relaxation matrix (RedMat) was carried out to simulate the theoretical binding epitopes and to compare them with the corresponding experimental ones in order to validate the MD simulation trajectories of the two complexes [42]. 2. Results 2.1. Comparison of the Interactions of Pentasaccharides (1) and (2) with AT Using STD and DEEP-STD NMR For each complex, two samples, one in heavy water buffer (D 2 O) and one in standard aqueous buffer (H 2 O), were prepared to perform differential solvent DEEP-STD NMR analysis. The STD NMR spectra were recorded at 0.5 s saturation time (Figures S1–S4). The absolute and relative STD percentages were calculated as described in the experimental section (Tables S1 and S2), and the binding epitope maps were obtained in both D 2 O and H2O conditions (Figure 2). Int. J. Mol. Sci. 2024, 25, x FOR PEER REVIEW 4 of 20 pentasaccharide (1) interacts with AT in an area of the binding pocket with the highest density of arginine residues, we expected to observe a relative increase in the STD intensities of the ligand protons belonging to this moiety in the STD NMR experiment conducted in H 2 O. The same behavior was also hypothesized for pentasaccharide (2), whose binding mode should be similar, albeit not exactly analogous, to that of pentasaccharide (1). Additionally, differences in the conformations of pentasaccharides (1) and (2) in the AT-bound state were investigated by tr-NOESY experiments combined with docking and explicit solvent MD simulations covering 1 µs. Finally, the methodology based on the reduced relaxation matrix (RedMat) was carried out to simulate the theoretical binding epitopes and to compare them with the corresponding experimental ones in order to validate the MD simulation trajectories of the two complexes [42]. 2. Results 2.1. Comparison of the Interactions of Pentasaccharides (1) and (2) with AT Using STD and DEEP-STD NMR For each complex, two samples, one in heavy water buffer (D 2 O) and one in standard aqueous buffer (H 2 O), were prepared to perform differential solvent DEEP-STD NMR analysis. The STD NMR spectra were recorded at 0.5 s saturation time (Figures S1–S4). The absolute and relative STD percentages were calculated as described in the experimental section (Tables S1 and S2), and the binding epitope maps were obtained in both D 2 O and H 2 O conditions (Figure 2). Figure 2. Ligand epitope mapping of pentasaccharides (1) and (2) for their interaction with AT. Panels (a,b): Binding epitope maps of pentasaccharide (1) in D 2 O and H 2 O. Panels (c,d): Binding epitope maps of pentasaccharide (2) in D 2 O and H 2 O. Figure 2. Ligand epitope mapping of pentasaccharides (1) and (2) for their interaction with AT. Panels (a,b): Binding epitope maps of pentasaccharide (1) in D 2 O and H 2 O. Panels (c,d): Binding epitope maps of pentasaccharide (2) in D2O and H2O. For pentasaccharide (1), the epitope mapping in D 2 O showed that H2C, H6 ′′ C, H2B, H2A, H3A and OCH 3 (A) receive the strongest saturation (80–100%), while H1E, H2E, H5E, H1D, H1C, H1B and H1A are slightly less affected by the saturation transfer (60–80%) (Table S1 and Figure 2a). The same analysis in H 2 O found that all the mentioned protons Int. J. Mol. Sci. 2024,25, 4669 5 of 19 present an STD intensity in the range of 60–80%, except for H2C, whose STD value is 100% (Table S1 and Figure 2b). Overall, considering the experimental error, the STD NMR results suggested a consistent binding mode of pentasaccharide (1) in both solvents. However, two significant observations emerged from the binding epitope maps (Figure 2a,b). Firstly, the STD NMR analysis in D 2 O revealed remarkable interactions between the reducing disaccharide and the protein. This indicates that STD NMR provides valuable insights into the final conformation of the bound ligand after the two steps of the recognition process described in the literature, leading to close contacts with the reducing disaccharide [ 39 , 40 ]. Secondly, the relative STDs were lower in H 2 O than in D 2 O, suggesting that the presence of exchangeable polar protons amplifies the relaxation-based decrease in the saturation transfer to the ligand protons (saturation leakage). The epitope map of pentasaccharide (2) in D 2 O showed that the STD intensity is strong (80–100%) for H2E, H2C and H2A, medium (60–80%) for H1E, H5E, H1D, H1C and H6 ′′ C and low (40–60%) for H4B and OCH 3 (A) (Table S2 and Figure 2c). The same evaluation in H 2 O displayed strong STD enhancements for H1E, H2C and H2A, medium STD effects for H2E and H5E and low STD percentages for H1D, H1C, H6 ′′ C, H4B and OCH 3 (A) (Table S2 and Figure 2d). Because of the superimposition between the signals of the anomeric protons of IdoA(B) and GlcNS6S(A), these could not be integrated separately; nevertheless, the STD NMR spectra indicate that they are involved in the interaction (Figures S3 and S4). In the case of pentasaccharide (2), the binding epitope map in D 2 O showed higher relative STDs for the non-reducing trisaccharide, which might indicate that the second step of binding is not as efficient and, therefore, AT does not hold the reducing disaccharide toward its surface as tightly as pentasaccharide (1). This is in agreement with the lower affinity of pentasaccharide (2) and supports the contention that the absence of the 2-O-sulfate group in the iduronic acid unit of pentasaccharide (2) confers the higher mobility of the IdoA(B)- GlcNS6S(A) disaccharide at the reducing end [ 36 ]. In fact, the comparison of the binding epitopes between the pentasaccharide (1)-AT and pentasaccharide (2)-AT complexes in D 2 O revealed that OCH 3 (A) receives significantly lower saturation in the pentasaccharide (2)-AT complex. Again, the impact of exchangeable polar protons in the H 2 O sample globally reduces the relative STDs of the ligand, as in the case of pentasaccharide (1). An additional difference between the binding epitope maps of both pentasaccharides was observed for H6 ′′ C: the lower STD enhancement characterizing this group in the pentasaccharide (2)-AT system suggests a weaker contact between the 6-O-sulfate group of GlcNS3S6S(C) and R46 (as it will be shown later in the text). Next, the application of the differential solvent DEEP-STD NMR led to an interesting observation from the comparison of the total levels of saturation (i.e., the sum of the absolute STD intensities) achieved in D 2 O with that recorded in H 2 O. Notably, this analysis revealed that the saturation transferred to pentasaccharide (1) was lower in D 2 O than in H 2 O (15.16% < 19.73%), whereas, on the contrary, the saturation received by pentasaccharide (2) was higher in D 2 O than in H 2 O (19.50% > 15.94%) (Tables S1 and S2). This difference correlates with the distinct affinity of these pentasaccharides for AT. Indeed, high affinity determines a long residence time of the ligand within the protein pocket, which leads to greater protection of the protein polar protons interacting with the ligand from their exchange with the bulk water. This enhances their contribution to the saturation transfer to the ligand, increasing the total saturation available for the transfer to the ligand protons. In this case, the protection is higher when pentasaccharide (1) is bound to AT, due to its higher affinity for the protein. In this way, the results establish that the differential solvent DEEP-STD NMR data comparison of similar ligands to a given protein receptor is useful to explain differences in affinity. While it is challenging to directly correlate changes in absolute STD intensities with differences in affinity, as the observation of STD signals can be influenced by varying kinetics, detecting an increase in the total saturation acquired by a ligand in H 2 O in comparison to D 2 O would suggest a high affinity interaction (within the range of the STD NMR observation). In contrast, ligands with medium or low affinity (within the Int. J. Mol. Sci. 2024,25, 4669 6 of 19 range of the STD NMR observation) are expected to show a decrease in the total saturation when changing from D2O to H2O. From the differential solvent DEEP-STD NMR experiments, we carried out an analysis of the DEEP-STD factors ( ∆ STDs). To compare equally the DEEP-STD NMR data collected from the analyzed systems, only the STD NMR signals in common between the two complexes were selected. Interestingly, the two complexes showed significantly different patterns in the DEEP-STD NMR data (Table S3 and Figure 3). Int. J. Mol. Sci. 2024, 25, x FOR PEER REVIEW 6 of 20 directly correlate changes in absolute STD intensities with differences in affinity, as the observation of STD signals can be influenced by varying kinetics, detecting an increase in the total saturation acquired by a ligand in H 2 O in comparison to D 2 O would suggest a high affinity interaction (within the range of the STD NMR observation). In contrast, ligands with medium or low affinity (within the range of the STD NMR observation) are expected to show a decrease in the total saturation when changing from D 2 O to H 2 O. From the differential solvent DEEP-STD NMR experiments, we carried out an analysis of the DEEP-STD factors (∆STDs). To compare equally the DEEP-STD NMR data collected from the analyzed systems, only the STD NMR signals in common between the two complexes were selected. Interestingly, the two complexes showed significantly different patterns in the DEEP-STD NMR data (Table S3 and Figure 3). Figure 3. Differential epitope mapping (D 2 O/H 2 O) of both pentasaccharides (1) and (2) in complex with AT. ΔSTD histograms: protons with the strongest positive ΔSTDs are in orange; protons with the highest negative ΔSTDs are in blue. The pentasaccharide (1)-AT complex displayed very small and probably not significant DEEP-STD factors (Table S3 and Figure 3). In contrast, the differential epitope map of the pentasaccharide (2)-AT complex exhibited stronger negative and positive ∆STDs for H1E and H1D, respectively (Table S3 and Figure 3). The relative decrease in the ∆STD characterizing H1E in D 2 O suggests proximity to slow exchanging polar residues (i.e., arginine), while the relative increase in the ∆STD belonging to H1D in D 2 O indicates that this ligand proton is close to fast exchanging polar residues (i.e., lysine) and/or nonexchangeable residues. The similarity of the binding epitopes, despite displaying significantly different DEEP-STD NMR patterns, was intriguing, prompting the use of MD simulations to deepen our understanding of the observed differences. 2.2. Comparative Analysis of the Conformations of Pentasaccharides (1) and (2) in the Unbound and AT-Bound States by 3 J H-H Coupling Constants and NOESY/tr-NOESY Experiments Three-bond proton–proton coupling constants ( 3 J H-H ) of pentasaccharides (1) and (2) were measured to define the conformation of the sugar rings in the free state (Tables S4 and S5). The analysis of the 3 J H-H values showed that the glucosamines [GlcNS6S(E, C, A)] and the glucuronic acid [GlcA(D)] in both pentasaccharides adopt a 4 C 1 chair conformation, while the iduronic acid in pentasaccharide (1) [IdoA2S(B)] exists in equilibrium between the 1 C 4 chair (36%) and 2 S 0 skew-boat (64%) conformations, according to Ferro et al. (1990) [6,43]. Unfortunately, the overlapping of the IdoA(B) signals did not permit the same analysis in pentasaccharide (2) to be performed. The NOE results confirmed the conformational data obtained from the 3 J H-H analysis for both pentasaccharides in the unbound state. NOESY and tr-NOESY experiments were carried out to describe the intra-residue and interglycosidic conformational changes in the two glycans after binding (Figures S5 and S6). A significant increase in the intensity of the cross-peaks of the tr-NOESY spectra (in the presence of the protein) compared to the NOESY spectra (free ligands) was observed, providing proof of interaction (Tables S6– S10). The conformation of the iduronic acid in pentasaccharides (1) and (2) in the free state and upon AT-binding was investigated by comparing the NOE and tr-NOE data. Since Figure 3. Differential epitope mapping (D2O/H2O) of both pentasaccharides (1) and (2) in complex with AT. ∆ STD histograms: protons with the strongest positive ∆ STDs are in orange; protons with the highest negative ∆STDs are in blue. The pentasaccharide (1)-AT complex displayed very small and probably not significant DEEP-STD factors (Table S3 and Figure 3). In contrast, the differential epitope map of the pentasaccharide (2)-AT complex exhibited stronger negative and positive ∆ STDs for H1E and H1D, respectively (Table S3 and Figure 3). The relative decrease in the ∆ STD characterizing H1E in D 2 O suggests proximity to slow exchanging polar residues (i.e., arginine), while the relative increase in the ∆ STD belonging to H1D in D 2 O indicates that this ligand proton is close to fast exchanging polar residues (i.e., lysine) and/or non-exchangeable residues. The similarity of the binding epitopes, despite displaying significantly different DEEP-STD NMR patterns, was intriguing, prompting the use of MD simulations to deepen our understanding of the observed differences. 2.2. Comparative Analysis of the Conformations of Pentasaccharides (1) and (2) in the Unbound and AT-Bound States by 3JH-H Coupling Constants and NOESY/tr-NOESY Experiments Three-bond proton–proton coupling constants ( 3 J H-H ) of pentasaccharides (1) and (2) were measured to define the conformation of the sugar rings in the free state (Tables S4 and S5) . The analysis of the 3 J H-H values showed that the glucosamines [GlcNS6S(E, C, A)] and the glucuronic acid [GlcA(D)] in both pentasaccharides adopt a 4 C 1 chair conformation, while the iduronic acid in pentasaccharide (1) [IdoA2S(B)] exists in equilibrium between the 1 C 4 chair (36%) and 2 S 0 skew-boat (64%) conformations, according to Ferro et al. (1990) [ 6 , 43 ]. Unfortunately, the overlapping of the IdoA(B) signals did not permit the same analysis in pentasaccharide (2) to be performed. The NOE results confirmed the conformational data obtained from the 3 J H-H analysis for both pentasaccharides in the unbound state. NOESY and tr-NOESY experiments were carried out to describe the intra-residue and interglycosidic conformational changes in the two glycans after binding (Figures S5 and S6). A significant increase in the intensity of the cross-peaks of the tr- NOESY spectra (in the presence of the protein) compared to the NOESY spectra (free ligands) was observed, providing proof of interaction (Tables S6–S10). The conformation of the iduronic acid in pentasaccharides (1) and (2) in the free state and upon AT-binding was investigated by comparing the NOE and tr-NOE data. Since the iduronic acid exhibits distinct H5-H2 distances in the 1 C 4 chair and 2 S 0 skew-boat conformations (4.0 Å and 2.4 Å, respectively), the ratio between the H5-H2 and H5-H4 NOEs permits estimation of its prevalent form. Unfortunately, the partial overlapping of the H5B-H4B cross-peak in the NOESY and tr-NOESY spectra of pentasaccharide (1) and pentasaccharide (1)-AT complex, respectively, with the H5B-H3B signal (generated by spin diffusion) did not Int. J. Mol. Sci. 2024,25, 4669 7 of 19 allow the conformer populations of IdoA2S(B) in the unbound and bound states to be confirmed. Nevertheless, it is well described that the iduronic acid unit belonging to the pentasaccharide (1) structure approaches a pure 2 S 0 form upon AT-binding [ 6 , 29 , 44 ]. The interaction between pentasaccharide (2) and AT changes the H5B-H2B/H5B-H4B NOE ratios of IdoA(B) from 0.3 to 0.7, revealing that the conformation of this moiety, which prevalently adopts the 1 C 4 chair conformation in the free state, is only partially shifted toward the AT-binding 2 S 0 skew-boat form in the bound state (Table S8). This result supports the reduced involvement of the reducing end moiety [IdoA(B)-GlcNS6S(A)] of pentasaccharide (2) in the binding to AT observed by STD NMR. The comparison of the interglycosidic NOEs and tr-NOEs, defined for both glycans in the unbound and bound states, underlined that the GlcA(D)-GlcNS3S6S(C) linkage is the only part of the glycosidic backbone that undergoes a significant conformational change upon AT-binding (the H1D- H6′′C/H1D-H4C NOE ratios change from 1.6–1.1 to 1.1–0.5) (Tables S9 and S10). 2.3. Comparison of the Interactions of Pentasaccharides (1) and (2) with AT by Molecular Docking The orientation that the two ligands assume in the protein-binding site was confirmed by molecular docking. The evaluation of the obtained poses was carried out considering the binding energy (GlideScore) and the root-mean square distance (RMSD) from the position of the reference pentasaccharide co-crystallized with AT (PDB ID: 1AZX) [ 39 ]. For the pentasaccharide (1)-AT complex, few poses were identified. In particular, the best pose, characterized by a GlideScore of − 9.0 Kcal/mol and RMSD of 2.0 Å, is very similar to the X-ray orientation (PDB ID: 1AZX). For the pentasaccharide (2)-AT complex, the best pose is defined by a GlideScore of − 9.5 Kcal/mol and RMSD of 3.0 Å. In contrast to the pentasaccharide (1)-AT system, several poses were collected when pentasaccharide (2) was docked on AT, some of them characterized by shifts in the glucosamine at the reducing end [GlcNS6S(A)]. In general, the docking results confirmed what was already known about the binding mode of heparin oligosaccharides to AT: the pentasaccharide sequence, here represented by the structures (1) and (2), specifically binds AT in the active site formed by helix-D, assuming a single orientation where its reducing end points toward the bent of the helix. Interestingly, the difference in the binding affinity between the two pentasaccharides can be correlated to the few poses obtained for pentasaccharide (1), implying the tighter binding of this ligand to the protein, against the higher number of structures collected for pentasaccharide (2), indicating greater mobility of this compound in the binding pocket. As docking calculations deal with rigid structures of the protein receptor, we next decided to carry out MD simulations of the complexes. 2.4. Investigation of the Structural Features of the Pentasaccharide (1)-AT and Pentasaccharide (2)-AT Complexes by MD Simulations The initial 3D structures of the pentasaccharide (1)-AT and pentasaccharide (2)-AT complexes were built based on the pentasaccharide-AT complex obtained by X-ray diffraction (PDB ID: 1AZX). The latter was found to be comparable to the docking poses characterized by the best GlideScore values. Both complexes were simulated for 1 µ s in explicit solvent condition, sampling the isothermal-isobaric (NPT) ensemble. Figures 4and 5depict two snapshots extracted from the MD simulation trajectories at 600 ns. The analysis of the RMSD showed that pentasaccharide (1) remains stable in the AT-binding pocket for the entire simulation time, while pentasaccharide (2) in the AT- bound state presents higher mobility that correlates to its lower affinity for the protein, even if no drift indicating early unbinding of this glycan was observed over the timescale (Figure S7). The root-mean square fluctuation (RMSF) analysis allowed the flexibility observed for pentasaccharide (2) within the AT-binding site to be correlated with the fluctuations of GlcNS6S(A) at its reducing end. Specifically, the comparison of the RMSF values calculated for the ligand units revealed that GlcNS6S(A) exhibits higher fluctuations in pentasaccharide (2) compared to its respective unit in pentasaccharide (1) and all other units (Figure S8a). Furthermore, a similar evaluation conducted on key amino acids Int. J. Mol. Sci. 2024,25, 4669 8 of 19 of the AT active site displayed higher mobility of the R13 and R129 side chains when pentasaccharide (1) binds to AT (Figure S8b). Int. J. Mol. Sci. 2024, 25, x FOR PEER REVIEW 8 of 20 Figure 4. Selected geometry of the pentasaccharide (1)-AT complex (600 ns simulation time). Pentasaccharide (1) is represented by purple, red, blue and yellow tubes corresponding to carbon, oxygen, nitrogen and sulfur atoms; AT is reported as grey ribbons; arginine (R) residues are shown as fuchsia ribbons and thin fuchsia and blue tubes representing carbon and nitrogen atoms; lysine (K) residues are displayed as yellow ribbons and thin yellow and blue tubes indicating carbon and nitrogen atoms. Letters from E to A indicate the residues of pentasaccharide (1) as in Figure 1. Helix- A and Helix-D in AT are labelled. Figure 5. Selected geometry of the pentasaccharide (2)-AT complex (600 ns simulation time). Pentasaccharide (2) is represented by pink, red, blue and yellow tubes corresponding to carbon, oxygen, nitrogen and sulfur atoms; AT is reported as grey ribbons; arginine (R) residues are shown as fuchsia ribbons and thin fuchsia and blue tubes representing carbon and nitrogen atoms; lysine (K) residues are displayed as yellow ribbons and thin yellow and blue tubes indicating carbon and nitrogen atoms. Letters from E to A indicate the residues of pentasaccharide (2) as in Figure 1. Helix- A and Helix-D in AT are labelled. The analysis of the RMSD showed that pentasaccharide (1) remains stable in the AT- binding pocket for the entire simulation time, while pentasaccharide (2) in the AT-bound state presents higher mobility that correlates to its lower affinity for the protein, even if no drift indicating early unbinding of this glycan was observed over the timescale (Figure S7). The root-mean square fluctuation (RMSF) analysis allowed the flexibility observed for pentasaccharide (2) within the AT-binding site to be correlated with the fluctuations of GlcNS6S(A) at its reducing end. Specifically, the comparison of the RMSF values calculated for the ligand units revealed that GlcNS6S(A) exhibits higher fluctuations in pentasaccharide (2) compared to its respective unit in pentasaccharide (1) and all other units (Figure S8a). Furthermore, a similar evaluation conducted on key amino acids of the AT active site displayed higher mobility of the R13 and R129 side chains when pentasaccharide (1) binds to AT (Figure S8b). Figure 4. Selected geometry of the pentasaccharide (1)-AT complex (600 ns simulation time). Pentasaccharide (1) is represented by purple, red, blue and yellow tubes corresponding to carbon, oxygen, nitrogen and sulfur atoms; AT is reported as grey ribbons; arginine (R) residues are shown as fuchsia ribbons and thin fuchsia and blue tubes representing carbon and nitrogen atoms; lysine (K) residues are displayed as yellow ribbons and thin yellow and blue tubes indicating carbon and nitrogen atoms. Letters from E to A indicate the residues of pentasaccharide (1) as in Figure 1. Helix-A and Helix-D in AT are labelled. Int. J. Mol. Sci. 2024, 25, x FOR PEER REVIEW 8 of 20 Figure 4. Selected geometry of the pentasaccharide (1)-AT complex (600 ns simulation time). Pentasaccharide (1) is represented by purple, red, blue and yellow tubes corresponding to carbon, oxygen, nitrogen and sulfur atoms; AT is reported as grey ribbons; arginine (R) residues are shown as fuchsia ribbons and thin fuchsia and blue tubes representing carbon and nitrogen atoms; lysine (K) residues are displayed as yellow ribbons and thin yellow and blue tubes indicating carbon and nitrogen atoms. Letters from E to A indicate the residues of pentasaccharide (1) as in Figure 1. Helix- A and Helix-D in AT are labelled. Figure 5. Selected geometry of the pentasaccharide (2)-AT complex (600 ns simulation time). Pentasaccharide (2) is represented by pink, red, blue and yellow tubes corresponding to carbon, oxygen, nitrogen and sulfur atoms; AT is reported as grey ribbons; arginine (R) residues are shown as fuchsia ribbons and thin fuchsia and blue tubes representing carbon and nitrogen atoms; lysine (K) residues are displayed as yellow ribbons and thin yellow and blue tubes indicating carbon and nitrogen atoms. Letters from E to A indicate the residues of pentasaccharide (2) as in Figure 1. Helix- A and Helix-D in AT are labelled. The analysis of the RMSD showed that pentasaccharide (1) remains stable in the AT- binding pocket for the entire simulation time, while pentasaccharide (2) in the AT-bound state presents higher mobility that correlates to its lower affinity for the protein, even if no drift indicating early unbinding of this glycan was observed over the timescale (Figure S7). The root-mean square fluctuation (RMSF) analysis allowed the flexibility observed for pentasaccharide (2) within the AT-binding site to be correlated with the fluctuations of GlcNS6S(A) at its reducing end. Specifically, the comparison of the RMSF values calculated for the ligand units revealed that GlcNS6S(A) exhibits higher fluctuations in pentasaccharide (2) compared to its respective unit in pentasaccharide (1) and all other units (Figure S8a). Furthermore, a similar evaluation conducted on key amino acids of the AT active site displayed higher mobility of the R13 and R129 side chains when pentasaccharide (1) binds to AT (Figure S8b). Figure 5. Selected geometry of the pentasaccharide (2)-AT complex (600 ns simulation time). Pentasaccharide (2) is represented by pink, red, blue and yellow tubes corresponding to carbon, oxygen, nitrogen and sulfur atoms; AT is reported as grey ribbons; arginine (R) residues are shown as fuchsia ribbons and thin fuchsia and blue tubes representing carbon and nitrogen atoms; lysine (K) residues are displayed as yellow ribbons and thin yellow and blue tubes indicating carbon and nitrogen atoms. Letters from E to A indicate the residues of pentasaccharide (2) as in Figure 1. Helix-A and Helix-D in AT are labelled. The known flexibility of the iduronate moiety in the AT-bound state was investigated plotting the distance between the H5 and H2 protons of IdoA2S(B) and IdoA(B) (Figure S9) [ 6 , 45 – 48 ]. This analysis revealed that IdoA2S(B) in the pentasaccharide (1)-AT complex presents a higher percentage of the 2 S 0 form than IdoA(B) in the pentasaccharide (2)-AT complex, in agreement with what was observed by NMR. The conformations of the glycosidic linkages of pentasaccharides (1) and (2) in the AT-bound state were described by Ramachandran plots, in which a color gradient density map allowed the most populated states to be localized (Figure 6). Int. J. Mol. Sci. 2024,25, 4669 9 of 19 Int. J. Mol. Sci. 2024, 25, x FOR PEER REVIEW 9 of 20 The known flexibility of the iduronate moiety in the AT-bound state was investigated plotting the distance between the H5 and H2 protons of IdoA2S(B) and IdoA(B) (Figure S9) [6,45–48]. This analysis revealed that IdoA2S(B) in the pentasaccharide (1)-AT complex presents a higher percentage of the 2 S 0 form than IdoA(B) in the pentasaccharide (2)-AT complex, in agreement with what was observed by NMR. The conformations of the glycosidic linkages of pentasaccharides (1) and (2) in the AT-bound state were described by Ramachandran plots, in which a color gradient density map allowed the most populated states to be localized (Figure 6). Figure 6. Ramachandran plots of the glycosidic dihedral angles φ i /ψ i of pentasaccharides (1) and (2) in the AT-bound state. A density color map is superposed on each Ramachandran plot; the color gradient (blue to red) is proportional to the density of the sampled states φ i /ψ i and qualitatively predicts the preferred conformation of each glycosidic linkage. The conformational analysis revealed that pentasaccharide (1) is characterized by a single populated state for each dihedral angle, while pentasaccharide (2) exhibits greater conformational mobility at the reducing end (Figures 6, 7 and S10). Figure 7. Pentasaccharide (1) [panel (a)] and pentasaccharide (2) [panel (b)] interacting with AT. For each pentasaccharide, three structures were selected from the corresponding MD simulation. These snapshots were sampled considering the most populated states of each glycosidic dihedral angle φ i /ψ i . Pentasaccharide (1) is represented by purple, red, blue and yellow tubes corresponding to carbon, oxygen, nitrogen and sulfur atoms; pentasaccharide (2) is depicted as pink, red, blue and yellow tubes indicating carbon, oxygen, nitrogen and sulfur atoms; AT is reported as gray ribbon. Specifically, it unveiled that the glycosidic bond between IdoA(B) and GlcNS6S(A) in pentasaccharide (2) presents two populated φ i /ψ i states. The most populated states for Figure 6. Ramachandran plots of the glycosidic dihedral angles φi / ψi of pentasaccharides (1) and (2) in the AT-bound state. A density color map is superposed on each Ramachandran plot; the color gradient (blue to red) is proportional to the density of the sampled states φi / ψi and qualitatively predicts the preferred conformation of each glycosidic linkage. The conformational analysis revealed that pentasaccharide (1) is characterized by a single populated state for each dihedral angle, while pentasaccharide (2) exhibits greater conformational mobility at the reducing end (Figures 6,7and S10). Int. J. Mol. Sci. 2024, 25, x FOR PEER REVIEW 9 of 20 The known flexibility of the iduronate moiety in the AT-bound state was investigated plotting the distance between the H5 and H2 protons of IdoA2S(B) and IdoA(B) (Figure S9) [6,45–48]. This analysis revealed that IdoA2S(B) in the pentasaccharide (1)-AT complex presents a higher percentage of the 2 S 0 form than IdoA(B) in the pentasaccharide (2)-AT complex, in agreement with what was observed by NMR. The conformations of the glycosidic linkages of pentasaccharides (1) and (2) in the AT-bound state were described by Ramachandran plots, in which a color gradient density map allowed the most populated states to be localized (Figure 6). Figure 6. Ramachandran plots of the glycosidic dihedral angles φ i /ψ i of pentasaccharides (1) and (2) in the AT-bound state. A density color map is superposed on each Ramachandran plot; the color gradient (blue to red) is proportional to the density of the sampled states φ i /ψ i and qualitatively predicts the preferred conformation of each glycosidic linkage. The conformational analysis revealed that pentasaccharide (1) is characterized by a single populated state for each dihedral angle, while pentasaccharide (2) exhibits greater conformational mobility at the reducing end (Figures 6, 7 and S10). Figure 7. Pentasaccharide (1) [panel (a)] and pentasaccharide (2) [panel (b)] interacting with AT. For each pentasaccharide, three structures were selected from the corresponding MD simulation. These snapshots were sampled considering the most populated states of each glycosidic dihedral angle φ i /ψ i . Pentasaccharide (1) is represented by purple, red, blue and yellow tubes corresponding to carbon, oxygen, nitrogen and sulfur atoms; pentasaccharide (2) is depicted as pink, red, blue and yellow tubes indicating carbon, oxygen, nitrogen and sulfur atoms; AT is reported as gray ribbon. Specifically, it unveiled that the glycosidic bond between IdoA(B) and GlcNS6S(A) in pentasaccharide (2) presents two populated φ i /ψ i states. The most populated states for Figure 7. Pentasaccharide (1) [panel (a)] and pentasaccharide (2) [panel (b)] interacting with AT. For each pentasaccharide, three structures were selected from the corresponding MD simulation. These snapshots were sampled considering the most populated states of each glycosidic dihedral angle φi / ψi . Pentasaccharide (1) is represented by purple, red, blue and yellow tubes corresponding to carbon, oxygen, nitrogen and sulfur atoms; pentasaccharide (2) is depicted as pink, red, blue and yellow tubes indicating carbon, oxygen, nitrogen and sulfur atoms; AT is reported as gray ribbon. Specifically, it unveiled that the glycosidic bond between IdoA(B) and GlcNS6S(A) in pentasaccharide (2) presents two populated φi / ψi states. The most populated states for each dihedral angle of the two systems and the relative population percentages of the dihedral angle between IdoA(B) and GlcNS6S(A) are reported in Table 1. MD poses based on these interglycosidic dihedral angles illustrate different orientations of the reducing end, consistent with the observed increased mobility of this substructure within the pentasaccharide (2)-AT complex, as confirmed by the RMSD and RMSF spectra (Figure S10). The dynamic ensembles derived from MD simulations were also subjected to comprehensive clustering analysis to discern groups of structures sharing similar conformational features. In the case of pentasaccharide (1), the clustering analysis unveiled a single dominant cluster, indicating a homogeneous conformational ensemble (Figure S11a,b). Int. J. Mol. Sci. 2024,25, 4669 16 of 19 pentasaccharide (2)-AT complexes were simulated for 1 µ s, sampling the NPT ensemble without any restraint. The first 0.5 µ s were considered as the equilibration stage, while the last 0.5 µ s were defined as the production stage. The conformational equilibration stage was established by plotting vs. time the RMSD of the ligand in comparison to its initial position (time 0) and the interglycosidic dihedral angles, φi (t)/ ψi (t), until a stationary oscillatory behavior was observed for all variables (Figures S7, S24 and S25). In all cases, periodic boundary conditions and the particle mesh Ewald method were applied. A cut-off of 10 Å was used for all non-bonded interactions. A Langevin thermostat with a collision frequency of 3 ps −1 and a Berendsen barostat with a relaxation time of 2 ps were used. The SHAKE algorithm was employed to restrain all bonds involving hydrogen, allowing a timestep of 2 fs. The RMSD, RMSF and clustering analyses were carried out with CPPTRAJ [ 55 ] included in Ambertool 14 package [ 54 ]. Clustering of trajectories was performed using the DBSCAN algorithm with the minimum number of points allowed for the cluster set to 25 and epsilon set to 0.5 Å [56]. The Ramachandran plots and the density color maps were generated using the statistical software R 4.3.1 (R Core Team, 2013) and the hexbin package [57,58]. The Poisson Boltzmann free energy of binding ( ∆Gbind PB ) was estimated applying the molecular mechanics Poisson Boltzmann surface area (MMPBSA) method (MMPBSA.py application) [ 59 , 60 ]. For each protein–ligand complex, the ∆Gbind PB was determined as average on the production stage of the MD simulation (between 0.5 and 1 µ s) with a sampling frequency of 2 ns and a sample size of 5000 poses covering a range of 500 ns. The standard error of the mean (SEM) of ∆Gbind PB was calculated as σ/√N , where σ represents the estimated standard deviation and Nindicates the number of samples. The profile of the local concentration of the water molecules surrounding selected atoms (or a set of atoms) of protein residues, ligand units or protons was investigated using the radial distribution function [g(r)]: g(r) = lim dr→0"V 4πNpair ·p(r) r2dr #(3) where the variable rrepresents the distance between the reference atom (or set of atoms) and the oxygen atom of each water molecule, g(r)dr reflects the concentration of water molecules in a spherical layer of infinitesimal thickness (between rand r + dr) centered on the reference atom, Vis the volume of the simulation cell, Npair denotes the number of possible atom pairs [61–64]. 4.5. Model Validation by RedMat The RedMat analysis was employed to calculate the theoretical binding epitope from the MD trajectories of the pentasaccharide (1)-AT and pentasaccharide (2)-AT complexes [ 42 ]. For the RedMat calculation, the following parameters were set: the NMR spectrometer frequency at 600 MHz, the complex rotational correlation time at 34.956 ns, the concentrations of ligand and protein at 2000 µ M and 50 µ M, respectively, and the cut-off distance at 10 Å. The agreement between the theoretical and experimental binding epitopes was evaluated using the R-factor (R-NOE): R-NOE =v u u u u t ∑STDexp,k−STDcalc,k2 ∑STDexp,k2(4) where STD exp,k is the experimental STD value for a proton kand STD calc,k is the STD value simulated using the RedMat algorithm. The experimental binding epitope was determined by employing the smallest saturation time still producing integrable signals (0.5 s), so that the experimental STD values could be considered not significantly affected by differential Int. J. Mol. Sci. 2024,25, 4669 17 of 19 proton relaxation rates without the need of determining STD intensities at the saturation time of 0 (STD0). Supplementary Materials: The following supporting information can be downloaded at https: //www.mdpi.com/article/10.3390/ijms25094669/s1. Author Contributions: Conceptualization, J.A. and M.G.; methodology, S.E., J.C.M.-G., J.A. and M.G.; formal analysis, M.P., S.E., J.C.M.-G. and J.R.-C.; investigation, M.P., S.E., J.C.M.-G. and J.R.-C.; data curation, M.P., S.E., J.C.M.-G. and J.R.-C.; writing—original draft preparation, M.P.; writing—review and editing, M.P., S.E., J.C.M.-G., J.R.-C., E.A.Y., J.A. and M.G.; supervision, J.A. and M.G. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by G. Ronzoni Foundation and the Ministerio de Ciencia e Innovación through the grants AEI/10.13039/501100011033/(PID2019-109395GB-I00 and PID2022- 142879NB-I00, both to J.A.), co-funded by the European Regional Development Fund (ERDF) “A way of making Europe”. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Data supporting the results of the study are available from the corresponding authors upon request. Acknowledgments: We acknowledge the 7C Consortium for NMR Research in Biotechnology and Material Science (Milan) and the NMR facility at the Biomolecular Interactions Platform (cicCartuja, Seville). Conflicts of Interest: The authors declare no conflicts of interest. References 1. Lever, R.; Mulloy, B.; Page, C.P. (Eds.) Heparin—A Century of Progress; Springer Berlin Heidelberg: Berlin/Heidelberg, Germany, 2012; Volume 207. 2. Lindahl, U.; Bäckström, G.; Thunberg, L.; Leder, I.G. Evidence for a 3-O-sulfated D-glucosamine residue in the antithrombinbinding sequence of heparin. Proc. Natl. Acad. Sci. USA 1980,77, 6551–6555. [CrossRef] [PubMed] 3. Casu, B.; Choay, J.; Ferro, D.R.; Gatti, G.; Jacquinet, J.C.; Petitou, M.; Provasoli, A.; Ragazzi, M.; Sinay, P.; Torri, G. Controversial glycosaminoglycan conformations. Nature 1986,322, 215–216. [CrossRef] 4. Sanderson, P.N.; Huckerby, T.N.; Nieduszynski, I.A. Conformational equilibria of α -L-iduronate residues in disaccharides derived from heparin. Biochem. J. 1987,243, 175–181. [CrossRef] [PubMed] 5. Mulloy, B.; Forster, M.J.; Jones, C.; Davies, D.B. N.m.r. and molecular-modelling studies of the solution conformation of heparin. Biochem. J. 1993,293, 849–858. [CrossRef] 6. Ferro, D.R.; Provasoli, A.; Ragazzi, M.; Casu, B.; Torri, G.; Bossennec, V.; Perly, B.; Sinaÿ, P.; Petitou, M.; Choay, J. Conformer populations of L-iduronic acid residues in glycosaminoglycan sequences. Carbohydr. Res. 1990,195, 157–167. [CrossRef] 7. Björk, I.; Lindahl, U. Mechanism of the anticoagulant action of heparin. Mol. Cell. Biochem. 1982,48, 161–182. [CrossRef] [PubMed] 8. Choay, J.; Petitou, M.; Lormeau, J.C.; Sinaÿ, P.; Casu, B.; Gatti, G. Structure-activity relationship in heparin: A synthetic pentasaccharide with high affinity for antithrombin III and eliciting high anti-factor Xa activity. Biochem. Biophys. Res. Commun. 1983,116, 492–499. [CrossRef] 9. Olson, S.T.; Björk, I.; Sheffer, R.; Craig, P.A.; Shore, J.D.; Choay, J. Role of the antithrombin-binding pentasaccharide in heparin acceleration of antithrombin-proteinase reactions. Resolution of the antithrombin conformational change contribution to heparin rate enhancement. J. Biol. Chem. 1992,267, 12528–12538. [CrossRef] 10. Olson, S.T.; Richard, B.; Izaguirre, G.; Schedin-Weiss, S.; Gettins, P.G.W. Molecular mechanisms of antithrombin–heparin regulation of blood clotting proteinases. A paradigm for understanding proteinase regulation by serpin family protein proteinase inhibitors. Biochimie 2010,92, 1587–1596. [CrossRef] 11. Lindahl, U.; Hook, M. Glycosaminoglycans and Their Binding to Biological Macromolecules. Annu. Rev. Biochem. 1978,47, 385–417. [CrossRef] 12. Jackson, R.L.; Busch, S.J.; Cardin, A.D. Glycosaminoglycans: Molecular properties, protein interactions, and role in physiological processes. Physiol. Rev. 1991,71, 481–539. [CrossRef] [PubMed] 13. Campbell, A.P.; Sykes, B.D. The Two-Dimensional Transferred Nuclear Overhauser Effect: Theory and Practice. Annu. Rev. Biophys. Biomol. Struct. 1993,22, 99–122. [CrossRef] [PubMed] 14. Mayer, M.; Meyer, B. Characterization of Ligand Binding by Saturation Transfer Difference NMR Spectroscopy. Angew. Chem. Int. Ed. Engl. 1999,38, 1784–1788. [CrossRef] Int. J. Mol. Sci. 2024,25, 4669 18 of 19 15. Mayer, M.; Meyer, B. Group Epitope Mapping by Saturation Transfer Difference NMR To Identify Segments of a Ligand in Direct Contact with a Protein Receptor. J. Am. Chem. Soc. 2001,123, 6108–6117. [CrossRef] [PubMed] 16. Meyer, B.; Peters, T. NMR Spectroscopy Techniques for Screening and Identifying Ligand Binding to Protein Receptors. Angew Chem. Int. Ed. Engl. 2003,42, 864–890. [CrossRef] [PubMed] 17. Monaco, S.; Tailford, L.E.; Juge, N.; Angulo, J. Differential Epitope Mapping by STD NMR Spectroscopy To Reveal the Nature of Protein-Ligand Contacts. Angew. Chem. Int. Ed. Engl. 2017,56, 15289–15293. [CrossRef] 18. Liepinsh, E.; Otting, G. Proton exchange rates from amino acid side chains—Implications for image contrast. Magn. Reson. Med. 1996,35, 30–42. [CrossRef] 19. Kirschner, K.N.; Yongye, A.B.; Tschampel, S.M.; González-Outeiriño, J.; Daniels, C.R.; Foley, B.L.; Woods, R.J. GLYCAM06: A generalizable biomolecular force field. Carbohydrates. J. Comput. Chem. 2008,29, 622–655. [CrossRef] 20. Singh, A.; Tessier, M.B.; Pederson, K.; Wang, X.; Venot, A.P.; Boons, G.J.; Prestegard, J.H.; Woods, R.J. Extension and validation of the GLYCAM force field parameters for modeling glycosaminoglycans. Can. J. Chem. 2016,94, 927–935. [CrossRef] 21. Grootenhuis, P.D.; Boeckel, V. Constructing a molecular model of the interaction between Antithrombin-III and a potent synthetic heparin Analogue. J. Am. Chem. Soc. 1991,113, 2743–2747. [CrossRef] 22. Verli, H.; Guimarães, J.A. Insights into the induced fit mechanism in antithrombin–heparin interaction using molecular dynamics simulations. J. Mol. Graph. Model. 2005,24, 203–212. [CrossRef] [PubMed] 23. Balogh, G.; Komáromi, I.; Bereczky, Z. The mechanism of high affinity pentasaccharide binding to antithrombin, insights from Gaussian accelerated molecular dynamics simulations. J. Biomol. Struct. Dyn. 2020,38, 4718–4732. [CrossRef] [PubMed] 24. Sarkar, A.; Yu, W.; Desai, U.R.; MacKerell, A.D.; Mosier, P.D. Estimating glycosaminoglycan–protein interaction affinity: Water dominates the specific antithrombin–heparin interaction. Glycobiology 2016,26, 1041–1047. [CrossRef] [PubMed] 25. Stancanelli, E.; Elli, S.; Hsieh, P.; Liu, J.; Guerrini, M. Recognition and Conformational Properties of an Alternative Antithrombin Binding Sequence Obtained by Chemoenzymatic Synthesis. ChemBioChem 2018,19, 1178–1188. [CrossRef] [PubMed] 26. Elli, S.; Stancanelli, E.; Wang, Z.; Petitou, M.; Liu, J.; Guerrini, M. Degeneracy of the Antithrombin Binding Sequence in Heparin: 2-O-Sulfated Iduronic Acid Can Replace the Critical Glucuronic Acid. Chem. Eur. J. 2020,26, 11814–11818. [CrossRef] [PubMed] 27. Almond, A. Multiscale modeling of glycosaminoglycan structure and dynamics: Current methods and challenges. Curr. Opin. Struct. Biol. 2018,50, 58–64. [CrossRef] [PubMed] 28. Bojarski, K.K.; Sieradzan, A.K.; Samsonov, S.A. Molecular dynamics insights into protein-glycosaminoglycan systems from microsecond-scale simulations. Biopolymers 2019,110, e23252. [CrossRef] [PubMed] 29. Hricovíni, M.; Guerrini, M.; Bisio, A.; Torri, G.; Petitou, M.; Casu, B. Conformation of heparin pentasaccharide bound to antithrombin III. Biochem. J. 2001,359, 265–272. [CrossRef] [PubMed] 30. Guerrini, M.; Mourier, P.A.J.; Torri, G.; Viskov, C. Antithrombin-binding oligosaccharides: Structural diversities in a unique function? Glycoconj J. 2014,31, 409–416. [CrossRef] 31. Caldwell, E.E.O.; Nadkarni, V.D.; Fromm, J.R.; Linhardt, R.J.; Weiler, J.M. Importance of specific amino acids in protein binding sites for heparin and heparan sulfate. Int. J. Biochem. Cell Biol. 1996,28, 203–216. [CrossRef] 32. Fromm, J.R.; Hileman, R.E.; Caldwell, E.E.O.; Weiler, J.M.; Linhardt, R.J. Pattern and Spacing of Basic Amino Acids in Heparin Binding Sites. Arch. Biochem. Biophys. 1997,343, 92–100. [CrossRef] [PubMed] 33. Lindahl, U.; Thunberg, L.; Bäckström, G.; Riesenfeld, J.; Nordling, K.; Björk, I. Extension and structural variability of the antithrombin-binding sequence in heparin. J. Biol. Chem. 1984,259, 12368–12376. [CrossRef] 34. Craig, P.A.; Olson, S.T.; Shore, J.D. Transient kinetics of heparin-catalyzed protease inactivation by antithrombin III. Characterization of assembly, product formation, and heparin dissociation steps in the factor Xa reaction. J. Biol. Chem. 1989,264, 5452–5461. [CrossRef] [PubMed] 35. Petitou, M.; Casu, B.; Lindahl, U. 1976–1983, a critical period in the history of heparin: The discovery of the antithrombin binding site. Biochimie 2003,85, 83–89. [CrossRef] [PubMed] 36. Petitou, M.; Lormeau, J.; Choay, J. Interaction of heparin and antithrombin III. Eur. J. Biochem. 1988,176, 637–640. [CrossRef] 37. Petitou, M.; van Boeckel, C.A.A. A Synthetic Antithrombin III Binding Pentasaccharide Is Now a Drug! What Comes Next? Angew. Chem. Int. Ed. Engl. 2004,43, 3118–3133. [CrossRef] [PubMed] 38. Jin, L.; Pedersen, L.C. Anticoagulant heparan sulfate: Structural specificity and biosynthesis. Appl. Microbiol. Biotechnol. 2007,74, 263–272. 39. Jin, L.; Abrahams, J.P.; Skinner, R.; Petitou, M.; Pike, R.N.; Carrell, R.W. The anticoagulant activation of antithrombin byheparin. Proc. Natl. Acad. Sci. USA 1997,94, 14683–14688. [CrossRef] 40. Desai, U.R.; Petitou, M.; Björk, I.; Olson, S.T. Mechanism of Heparin Activation of Antithrombin. J. Biol. Chem. 1998,273, 7478–7487. [CrossRef] 41. Petitou, M.; Barzu, T.; Herault, J.-P.; Herbert, J.-M. A unique trisaccharide sequence in heparin mediates the early step of antithrombin III activation. Glycobiology 1997,7, 323–327. [CrossRef] 42. Nepravishta, R.; Walpole, S.; Hicks, T.; Muñoz-García, J.C.; Angulo, J. Fast Validation of Static and Dynamic 3D Models of Weak Protein-Ligand Complexes from STD NMR Spectroscopy. ChemRxiv 2022. [CrossRef] 43. Mikhailov, D.; Mayo, K.H.; Vlahov, I.R.; Toida, T.; Pervin, A.; Linhardt, R.J. NMR solution conformation of heparin-derived tetrasaccharide. Biochem. J. 1996,318, 93–102. [CrossRef] [PubMed] Int. J. Mol. Sci. 2024,25, 4669 19 of 19 44. Torri, G.; Casu, B.; Gatti, G.; Petitou, M.; Choay, J.; Jacquinet, J.C.; Sinaÿ, P. Mono- and bidimensional 500 MHz 1H-NMR spectra of a synthetic pentasaccharide corresponding to the binding sequence of heparin to antithrombin-III: Evidence for conformational peculiarity of the sulfated iduronate residue. Biochem. Biophys. Res. Commun. 1985,128, 134–140. [CrossRef] [PubMed] 45. Ferro, D.R.; Provasoli, A.; Ragazzi, M.; Torri, G.; Casu, B.; Gatti, G.; Jacquinet, J.C.; Sinay, P.; Petitou, M.; Choay, J. Evidence for conformational equilibrium of the sulfated L-iduronate residue in heparin and in synthetic heparin mono- and oligo-saccharides: NMR and force-field studies. J. Am. Chem. Soc. 1986,108, 6773–6778. [CrossRef] 46. Muñoz-García, J.C.; López-Prados, J.; Angulo, J.; Díaz-Contreras, I.; Reichardt, N.; de Paz, J.L.; Martín-Lomas, M.; Nieto, P.M. Effect of the substituents of the neighboring ring in the conformational equilibrium of iduronate in heparin-like trisaccharides. Chemistry 2012,18, 16319–16331. [CrossRef] [PubMed] 47. Hsieh, P.H.; Thieker, D.F.; Guerrini, M.; Woods, R.J.; Liu, J. Uncovering the Relationship between Sulphation Patterns and Conformation of Iduronic Acid in Heparan Sulphate. Sci. Rep. 2016,6, 29602. [CrossRef] [PubMed] 48. Das, S.K.; Mallet, J.M.; Esnault, J.; Driguez, P.A.; Duchaussoy, P.; Sizun, P.; Hérault, J.P.; Herbert, J.M.; Petitou, M.; Sinaÿ, P. Synthesis of Conformationally Locked Carbohydrates: A Skew-Boat Conformation of L-Iduronic Acid Governs the Antithrombotic Activity of Heparin. Angew. Chem. Int. Ed. Engl. 2001,40, 1670–1673. [CrossRef] 49. Mosier, P.D.; Krishnasamy, C.; Kellogg, G.E.; Desai, U.R. On the Specificity of Heparin/Heparan Sulfate Binding to Proteins. Anion-Binding Sites on Antithrombin and Thrombin Are Fundamentally Different. PLoS ONE 2012,7, e48632. [CrossRef] [PubMed] 50. Elli, S.; Raffaini, G.; Guerrini, M.; Kosakovsky Pond, S.; Matrosovich, M. Molecular modeling and phylogenetic analyses highlight the role of amino acid 347 of the N1 subtype neuraminidase in influenza virus host range and interspecies adaptation. Front. Microbiol. 2023,14, 1309156. [CrossRef] 51. Petitou, M.; Jaurand, G.; Derrien, M.; Duchaussoy, P.; Choay, J. A new, highly potent, heparin-like pentasaccharide fragment containing a glucose residue instead of a glucosamine. Bioorg. Med. Chem. Lett. 1991,1, 95–98. [CrossRef] 52. Friesner, R.A.; Banks, J.L.; Murphy, R.B.; Halgren, T.A.; Klicic, J.J.; Mainz, D.T.; Repasky, M.P.; Knoll, E.H.; Shelley, M.; Perry, J.K.; et al. Glide: A New Approach for Rapid, Accurate Docking and Scoring. 1. Method and Assessment of Docking Accuracy. J. Med. Chem. 2004,47, 1739–1749. [CrossRef] [PubMed] 53. Halgren, T.A.; Murphy, R.B.; Friesner, R.A.; Beard, H.S.; Frye, L.L.; Pollard, W.T.; Banks, J.L. Glide: A New Approach for Rapid, Accurate Docking and Scoring. 2. Enrichment Factors in Database Screening. J. Med. Chem. 2004,47, 1750–1759. [CrossRef] [PubMed] 54. Case, D.A.; Babin, V.; Berryman, J.T.; Betz, R.M.; Cai, Q.; Cerutti, D.S.; Cheatham, T.E., III.; Darden, T.A.; Duke, R.E.; Gohlke, H.; et al. AMBER 14; University of California: San Francisco, CA, USA, 2014. 55. Roe, D.R.; Cheatham, T.E., III. PTRAJ and CPPTRAJ: Software for Processing and Analysis of Molecular Dynamics Trajectory Data. J. Chem. Theory Comput. 2013,9, 3084–3095. [CrossRef] 56. Ester, M.; Kriegel, H.P.; Sander, J.; Xu, X. A density-based algorithm for discovering clusters in large spatial databases with noise. Knowl. Discov. Data Min. 1996,96, 226–231. 57. R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria, 2013. 58. Azzalini, A.; Menardi, G.; Rosolin, T. Cluster: Cluster Analysis via Nonparametric Density Estimation, version 1.0-0; Universitádi Padova: Padova, Italy, 2012. 59. Weis, A.; Katebzadeh, K.; Soderhjelm, P.; Nilsson, I.; Ryde, U. Ligand affinities predicted with the MM/PBSA method: Dependence on the simulation method and the force field. J. Med. Chem. 2006,49, 6596–6606. [CrossRef] [PubMed] 60. Miller, B.R., 3rd; McGee, T.D., Jr.; Swails, J.M.; Homeyer, N.; Gohlke, H.; Roitberg, A.E. MMPBSA.Py: An efficient program for end-state free energy calculations. J. Chem. Theory Comput. 2012,8, 3314–3321. [CrossRef] [PubMed] 61. Levine, B.G.; Stone, J.E.; Kohlmeyer, A. Fast analysis of molecular dynamics trajectories with graphics processing units—Radial distribution function histogramming. J. Comput. Phys. 2011,230, 3556–3569. [CrossRef] [PubMed] 62. Humphrey, W.; Dalke, A.; Schulten, K. VMD: Visual molecular dynamics. J. Mol. Graph. 1996,14, 33–38. [CrossRef] 63. Raffaini, G.; Elli, S.; Ganazzoli, F. Computer simulation of bulk mechanical properties and surface hydration of biomaterials. J. Biomed. Mater. Res. A 2006,77A, 618–626. [CrossRef] 64. Raffaini, G.; Ganazzoli, F. Adsorption of charged albumin subdomains on a graphite surface. J. Biomed. Mater. Res. A 2006,76A, 638–645. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.