Exploring the conformational dynamics of the CRD4: a model for receptor lysosomal activity shifts in search of 'sweet’ and ‘sour’ states

Abstract The macrophage mannose receptor (RMM) is a crucial component of the immune system involved in immune responses, inflammation resolution, and tissue remodeling. When RMM is activated by a specific ligand, it undergoes internalization, forming an endosome that matures into a lysosome. Within the lysosome, structural changes in RMM facilitate the dissociation of ligands for further processing. However, the precise details of these structural changes are not well understood. In this study, we used molecular dynamics simulations to investigate the conformational dynamics of a specific region called CRD4 in RMM. Our simulations explored different conditions, including pH variations and the presence of Ca2+ ions. By analyzing the simulation data, we found that conformational changes primarily occur in loop regions, while the secondary structure remains stable. The binding site of CRD4, essential for ligand interaction, is located on the protein surface between two specific loop regions. Ligand binding is stabilized by three important amino acids. Interestingly, the interaction patterns differ between monosaccharide and disaccharide ligands. These findings improve our understanding of CRD4's dynamics and how it recognizes ligands. They provide insights into the structure of CRD4 and its role in ligand dissociation within lysosomes. The study also highlights the significance of loop regions in functional dynamics and interactions. Further research is needed to fully uncover the complete structure of CRD4, understand ligand binding modes, and explore the influence of environmental factors. This study lays the foundation for future investigations targeting carbohydrate-protein interactions and the development of therapeutics based on RMM's unique properties. Communicated by Ramaswamy H. Sarma


Introduction
The macrophage mannose receptor (RMM) is a C-type lectin that plays a crucial role in both innate and adaptive immune responses, as well as in inflammation resolution and tissue remodeling.The normal physiological function of RMM involves its activation by a specific ligand, followed by internalization facilitated by the actin cytoskeleton.This internalization process leads to the formation of an endosome, which subsequently matures into a lysosome through fusion events.The lysosome undergoes changes in its environment, such as enzymatic pH reduction and protein glycosylation, which impact the behavior of RMM.These environmental alterations result in the dissociation of the receptor-ligand complex, allowing the ligands to be processed by digestive enzymes (Hu et al., 2019) (Figure 1).The exact conformational changes responsible for this dissociative effect remain unclear.Once the ligands are released, the receptor can either be recycled back to the cell surface or be digested along with its ligands.This lysosome-mediated transition from the active "sweet" conformation of RMM to the inactive "sour" conformation directly influences the macrophage's phagocytic function in immune responses and tissue remodelling (Figure 1).
The structural details of the macrophage mannose receptor (RMM) are not yet fully understood.Two important publications have discussed the different conformational states of RMM: Feinberg et al., 2021 proposed an extended conformation for the sweet flavor, while (Hu et al., 2019) described a compact sour flavor where the N-terminal region is positioned differently within the receptor.However, the overall structure of RMM modeled by AlphaFold is limited due to the presence of numerous loops, making it difficult to identify distinct poles for the N-and C-terminals.RMM stands out among other C-type lectins as it possesses multiple recognition domains for different ligands.It has a total of 1,308 amino acid residues and can be divided into three main regions: a N-terminal cysteine-rich domain, a fibronectin IIlike domain, and eight consecutive carbohydrate recognition domains (CRDs).Among these domains, CRD4 shows significant affinity for carbohydrates such as mannose and fucose, while the others exhibit minimal or no affinity.Extensive research has focused on studying CRD4, resulting in the publication of its three-dimensional structure when bound to various ligands (Hu et al., 2019).
Although the precise structural characteristics of RMM remain elusive, the studies provide valuable insights into the flavors and functional regions of the receptor.Further investigations are needed to better understand the conformational changes associated with ligand dissociation and to uncover the complete three-dimensional structure of RMM.

Structure preparation
The CRD4 structure was generated using the Maestro edition tool (PDB:7L65).To ensure accuracy and completeness, the structure underwent additional refinement steps.Hydrogen atoms were added, and any missing side chains were fixed using the Biopolymer tool of Sybyl-X 2.0, developed by TRIPOS Inc.Furthermore, PROPKA (Sastry et al., 2013) was employed to prepare the CRD4 structure at pH levels of 5.0 and 7.0 (Rostkowski et al., 2011).These steps were taken to enhance the reliability and suitability of the CRD4 structure for further analysis.

Molecular dynamics simulations
The prepared CRD4 complex (APO-apostructure under pH ¼ 7,0; ACID-apostructure under pH ¼ 5,0; CAAUX-CRD4-Ca II complex under pH ¼ 7,0; 1CHO-CRD4-a-methylmannose complex under pH ¼ 7,0; 2CHO-CRD4-a-man-1,2-a-mannose complex under pH ¼ 7,0) were simulated through Molecular Dynamics (MD).The engine used was the software Desmond (Bowers et al., 2006) with the OPLS-2005 force-field (Banks et al., 2005).The system was constructed with a protein-ligand complex, a predefined solvent water model (TIP3P; XXX (Jorgensen et al., 1983)) for our system we used approximately 50.000 molecules of water, charge neutralizing counterions (Na þ (2) or Cl -(3)), and periodic boundary conditions (PBC).Concerning the parameters of Ca 2þ , Cl -, and Na þ , we used the OPLS-2005 force field for parametrization (Jorgensen et al., 1996).The system box was constructed considering at least the distance of 13 Å from the protein atoms to the box edges.Short-range coulombic interactions were set at a cut-off value of 9.0 Å and the integration time step calculated every 2 fs, while long-range coulombic interactions were estimated using the Smooth Particle Mesh Ewald (PME) method (Jorgensen et al., 1996).Each system was subjected to at least 1 ms and all trajectory and interaction data are available on the Zenodo repository (DOI: 10.5281/zenodo.7961575).
Atomic interactions, distances, and secondary structure elements were determined using the Simulation Event Analysis pipeline as implemented in Maestro 2022v.4(Schr€ odinger LCC).The criteria for protein-ligand H-bond are 2.5 A˚ distance between the donor and acceptor atoms (D-H���A); �120 � angle between the donor-hydrogen-acceptor atoms (D-H���A); and �90 � angle between the hydrogenacceptor-bonded atoms (H���A-X).The protein-water and water-ligand H-bonds were considered when 2.8 A˚ (D-H���A); �110 � (D-H���A); and �90 � (H���A-X).Non-specific hydrophobic interactions are defined by the presence of a hydrophobic side chain within 3.6 A˚ of the ligand's aromatic or aliphatic carbons.p-p interactions are recorded when two aromatic groups are stacked face-to-face or face-to-edge and within 4.5 A˚ of distance.The MD trajectories were visualized in PyMol v.2.2.3 (Schr€ odinger LCC, New York, NY, USA).The distance and angles analyses were used to assess the movement of structures (Bao et al., 2023;Chen et al., 2015).The distance of selected regions was calculated by the center of mass distance using the script (trj_asl_distance.py)available on Schrodinger package 2019v.4.

Principal component analysis and Essential dynamics
Principal Component Analysis (PCA) was employed to investigate the primary characteristics of CRD4 complexes and the apostructure.The backbone of the domains was extracted and centered on the initial frame of the apostructure simulation using the "trj_selection_dl.py"and "trj_center.py"scripts from the Schrodinger package version 2019v.4.Subsequently, the simulations were aligned and merged utilizing the "gmx_ trjconv" tool from GROMACS 2023 (Abraham et al., 2015).To calculate the covariance matrix, the "gmx_covar" program was utilized, and the resulting eigenvectors were analyzed through the "gmx_anaeig" tool.Essential dynamics analysis (Amadei et al., 1993) was performed using the "trj_essential_dynamics.py" script from the Schr€ odinger pipeline.During these calculations, the trajectories were aligned using the "trj_align.py"script.The mode vectors were generated using PyMol v.2.2.3 (Schr€ odinger LCC, New York, NY, USA).

Results and discussion
In this study, we present a ground-breaking investigation utilizing microsecond molecular dynamics (MD) simulations to explore the conformational dynamics of CRD4.Each system under analysis was subjected to a minimum of 1 ls of simulation time, distributed across no fewer than five independent replicas (each spanning 200 ns).Our objective was to compare various configurations, including the apostructure at pH 7 and pH 5, as well as the with Ca 2þ as an auxiliary ion, aiming to elucidate the induced changes triggered by alterations in pH and the presence of Ca 2þ .Remarkably, our simulation data amounted to approximately 5 ms worth of invaluable insights.
Through this extensive MD exploration, we endeavoured to unlock the intricate symphony of molecular events governing CRD4's behaviour.By meticulously examining the system's response to varying environmental conditions, we have unveiled novel and transformative details regarding the conformational dynamics of CRD4.This study not only enhances our fundamental understanding of this critical receptor but also provides a platform for further investigations into its functional implications and potential therapeutic applications.

Structural insights into CRD4
CRD4, spanning from residue 655-778, constitutes a compact region characterized by two a-helices positioned on opposite sides of the peptide and flanked by seven b-sheets, as depicted in Figure 2A and B. The crucial mannose-binding site (MBS) within CRD4 is comprised of five key residues, three of which are carboxylic acids: E741, N743, E745, N765, and D766.The integrity of this site is further fortified by the essential Ca 2þ ion, aptly referred to as the main calcium ion, which forms coordination bonds with these residues.Notably, the MBS exhibits a highly accessible surface, facilitating robust interactions with ligands.Specifically, the 3and 4-OH groups of carbohydrates participate in calcium chelation, culminating in the formation of the active ligandreceptor complex.It is important to emphasize that the successful carbohydrate-calcium interaction necessitates the distinctive anomeric positions of both ligand hydroxyl groups, as exemplified by D-mannose and L-fucose.Additionally, the hexose ring assumes a quasi-parallel orientation relative to the neighbouring Y469 within the MBS (Feinberg et al., 2000(Feinberg et al., , 2021;;Mullin et al., 1997).These findings shed light on the structural intricacies underlying the ligand recognition and binding properties of CRD4, providing a solid foundation for further investigations in this realm.
An auxiliary calcium ion has been proposed to be associated with the active form of CRD4 apoprotein as well as its ligand-receptor complexes (Mullin et al., 1997).However, initial crystallographic experiments on CRD4 (PDB: 1EGG, 1EGI; (Feinberg et al., 2000)) did not reveal any electronic density corresponding to this ion.Directed mutagenesis studies of this domain, coupled with affinity-exclusion chromatography and proteolysis investigations, have indicated that CRD4 exhibits a second order K cat , suggesting a 2:1 calcium/CRD4 ratio for optimal protein activity.Recently, Feinberg et al. (2021) elucidated the three-dimensional structures of CRD4mannoside complexes (PDB: 7JUH, 7L64), in which a secondary calcium ion was observed to be chelated by S689, N691, and E951 at a distance from the MBS.However, the exact role of this secondary ion in CRD4 activity and ligand complexation remains poorly understood.
It was reported (Hu et al., 2019;Mullin et al., 1997) that have explored the sweet-sour conformational changes in RMM and how they can affect CRD4 ligand accessibility under different conditions.Although these studies present conflicting findings, both RMM models suggest that CRD4 potentially interacts with other regions of RMM in the sweet conformation, thereby potentially enhancing the receptor's overall affinity for carbohydrates.However, the specific conformational changes occurring within CRD4 and its neighbouring regions, if any, that account for the low ligand affinity observed in the sour conformation have yet to be proposed.Given that these two distinct conformational states exhibit different patterns of ligand recognition, it is crucial to understand the structural features required for ligands to bind to both conformations to ensure proper ligand release within lysosomes.In this study, we aim to investigate the impact of acidic conditions on CRD4 and its implications for drug design studies targeting RMM (Hu et al., 2019).

Insights from PCA analysis and dynamics of loop and secondary structure elements
The application of Principal Component Analysis (PCA) revealed that the combined simulations of CRD4 predominantly exhibit conformational changes represented by the first two principal components, which account for 27.78% and 20.07% of the total variance, respectively.Furthermore, the motion distribution between the Apo state at pH 7.0 and the state with Ca 2þ auxiliary showed similar patterns to that of the Apo state at pH 5.0 (Figure 3A).Based on these findings, we propose that the MBS residues display limited movement and maintain chelation with Ca I throughout the entire simulation.Supporting our observations, available crystallographic data (PDB: 7L68 (Feinberg et al., 2021)) indicate that the MBS is expected to adopt a folded conformation at neutral pH even in the absence of ligands.Moreover, previous studies (Feinberg et al., 2000) have reported that evidence suggests the persistent chelation of Ca I, even under extreme conditions such as pH 7-8.0.This suggests that the chelation of Ca I is primarily regulated by the environmental concentration of Ca 2þ .
The results obtained from PCA provide valuable insights into the conformational changes observed in our simulations.Our analysis reveals that these conformational changes predominantly occur in the loops, specifically the N-terminal loop and the loop spanning residues 726-752 (Figure 3B).These loop regions exhibit pronounced motions, suggesting their involvement in the dynamics of the system.
Interestingly, the secondary structure elements (SSE) regions, which include alpha helices and beta strands, display minimal movement throughout the simulations (Figure 3C and D).Based on the PCA results, we observed that a specific loop (715-735) exhibited greater stability in an acidic environment compared to the apostructure.In contrast, another loop (735-755) displayed increased flexibility in the acidic environment compared to the apostructure.These findings suggest that the region spanning amino acids 715-755 undergoes conformational changes that are influenced by pH levels.This observation is supported by the low root mean square fluctuation (RMSF) values of these residues, indicating that the structural integrity of the SSE regions remains relatively stable.The pronounced motions observed in the loop regions suggest their importance in mediating functional dynamics or interactions in the system.These flexible loop regions often play crucial roles in protein function, such as ligand binding, substrate recognition, or conformational changes necessary for enzymatic activity (Bu et al., 2013;Shen et al., 2021).
The limited movement and stability of the SSE regions suggest their contribution to maintaining the overall structure and stability of the protein.These regions typically form the core of the protein structure, providing a stable scaffold for the dynamic regions to perform their functional roles (Bu et al., 2013).Overall, the findings indicate that the observed conformational changes primarily occur in the loop regions, while the SSE regions remain relatively stable.
Our results suggest that CRD4 cannot reliably predict the sweet-sour transition observed in the receptor when under acidic pH.As such, drug discovery based exclusively on its structure may lead to compounds that bind so tightly to this domain and cannot be released for lysosomal digestion.

Characterizing the interactions of mannose: insights into the binding profile
The CRD4 binding site, positioned on the protein surface between the N-terminal loop and the loop spanning residues 726-752, plays a pivotal role in the interaction with mannose (Feinberg et al., 2021).Our investigation ratifies that the binding of the ligand is primarily stabilized by main Ca chelation and by direct hydrogen bond (HB) with N765, engaging with both mono (1CHO) and disaccharide (2CHO) molecules, facilitating the formation of stable molecular complexes.Interestingly, our analysis indicates a disparity in HB interactions between 1CHO and 2CHO with remaining key amino acids.On the other hand, 1CHO does not form strong hydrogen bonds with N745 and E751, in contrast to 2CHO (Figure 4A and B).This suggests a differential binding mechanism between the two ligands, where 1CHO interacts with these amino acids through alternative means.
Further investigation reveals that 1CHO predominantly interacts via hydrogen bonds (98%), whereas 2CHO predominantly interacts through water bridges (71%).These contrasting interaction patterns shed light on the specific molecular recognition mechanisms employed by CRD4 for these ligands (Feinberg et al., 2021).These findings provide valuable insights into the molecular basis of CRD4's interaction with mannose ligands.Noticeably, mannosidic ligands interact with CRD4 exclusively through 3-and 4-OH, as other hydroxyls remain solvated.Our study adds to the growing body of literature on carbohydrate-protein interactions and expands our understanding of the molecular basis of ligand recognition by CRD4.These findings have implications for further investigations into the functional significance of CRD4 in biological systems and can potentially inform the design of therapeutic interventions targeting carbohydrateprotein interactions.Further studies are warranted to explore the precise structural determinants and functional consequences of the observed differential ligand binding modes.Investigating the dynamics of the CRD4 catalytic site in the presence of ligands and understanding the role of other amino acid residues in ligand recognition will contribute to a comprehensive understanding of CRD4's functional role and potential applications in various biological contexts.

Conclusion
In this research, our focus was to investigate the behavior of a specific component of the macrophage mannose receptor known as carbohydrate recognition domain 4 (CRD4) using molecular dynamics simulations.Our primary objective was to gain a deeper understanding of the structure of CRD4 and its role in the release and dissociation of ligands within lysosomes.These processes are crucial for the immune response, resolution of inflammation, and tissue remodeling.Throughout our simulations, we made noteworthy observations regarding the flexible loop regions of CRD4, which play a vital role in its functional dynamics and interactions.Conversely, the more stable elements of CRD4, such as alpha helices and beta strands, remained relatively unaltered, serving as a stable framework for the dynamic regions.
Based on our findings, it appears that CRD4 cannot reliably predict the receptor's transition from perceiving sweetness to sensing sourness under acidic pH conditions.Our results provide valuable insights into the structural aspects of CRD4 and enhance our understanding of its functional behavior in relevant physiological contexts.We also examined CRD4 MBS, located between two specific loop regions.This site is crucial for ligand binding and is stabilized by main Ca chelation and by HB with N765.Interestingly, we observed different interaction patterns between monosaccharide and disaccharide ligands, with monosaccharides forming weaker hydrogen bonds with certain amino acids compared to disaccharides.Our study enhances our understanding of CRD4's conformational dynamics and how it recognizes ligands.It provides valuable insights into CRD4's structure and its role in releasing and dissociating ligands within lysosomes.These findings have implications for further investigations into CRD4's importance in biological systems and may guide the development of therapies targeting carbohydrate-protein interactions.
Future research should focus on exploring the complete three-dimensional structure of the macrophage mannose receptor, studying how environmental factors influence CRD4's dynamics, and understanding the specific structural factors and functional consequences of the observed ligand binding These investigations will contribute to a comprehensive understanding of CRD4's function and its potential applications in various biological contexts.

Figure 2 .
Figure 2. CRD4 structural features.A -Residue sequence with secondary structure elements highlighted in the 3D structure, coloured accordingly, B -highlight on the two a-helixes in red and seven b-sheets in blue; C -MBS residue organization and 2D schematic mannose-CRD4 interaction.

Figure 3 .
Figure 3. CRD4 PCA principal component analysis.(A) Distributions over the two significant PCs (PC1 and PC2) separated for each simulated system apostructure (pH 7) is highlighted in pink; apostructure (acid, pH 5) in brown and Ca 2þ auxiliary in green.(B) Extreme motions from the PC1 displayed over the CRD4 tertiary structure represented by arrows.(C) RMSF and SSE for apostructures in pH 7 (apo) and 5 (acid).(D) RMSF and SSE for apostructures (apo) and with Ca 2þ auxiliary (caaux).

Figure 4 .
Figure 4. Comprehensive interactions of CRD4 with Mono (1CHO) and disaccharide (2CHO): 2D structural representations, interaction frequencies, and snapshot frames.2D structures representations of ligands mono (1CHO; A); disaccharide (2CHO; B) and frequencies of interactions.Snapshot frame representation of tryad site (C).Inter-residue profile distance on tryad site on CRD4 in pH 7 and 5, in presence of 1CHO, 2CHO and Ca 2þ (D).CRD4 residues are colored according to the types of atoms in the interacting amino acid residues with the protein carbon, light grey; nitrogen, blue; oxygen, red.