Research Article: Structural Insights into Calcium-Bound S100P and the V Domain of the RAGE Complex

Date Published: August 1, 2014

Publisher: Public Library of Science

Author(s): Srinivasa R. Penumutchu, Ruey-Hwang Chou, Chin Yu, Andrea Motta.


The S100P protein is a member of the S100 family of calcium-binding proteins and possesses both intracellular and extracellular functions. Extracellular S100P binds to the cell surface receptor for advanced glycation end products (RAGE) and activates its downstream signaling cascade to meditate tumor growth, drug resistance and metastasis. Preventing the formation of this S100P-RAGE complex is an effective strategy to treat various disease conditions. Despite its importance, the detailed structural characterization of the S100P-RAGE complex has not yet been reported. In this study, we report that S100P preferentially binds to the V domain of RAGE. Furthermore, we characterized the interactions between the RAGE V domain and Ca2+-bound S100P using various biophysical techniques, including isothermal titration calorimetry (ITC), fluorescence spectroscopy, multidimensional NMR spectroscopy, functional assays and site-directed mutagenesis. The entropy-driven binding between the V domain of RAGE and Ca+2-bound S100P was found to lie in the micromolar range (Kd of ∼6 µM). NMR data-driven HADDOCK modeling revealed the putative sites that interact to yield a proposed heterotetrameric model of the S100P-RAGE V domain complex. Our study on the spatial structural information of the proposed protein-protein complex has pharmaceutical relevance and will significantly contribute toward drug development for the prevention of RAGE-related multifarious diseases.

Partial Text

The receptor for advanced glycation end products (RAGE) is a cell surface signaling receptor and a member of the immunoglobulin superfamily [1], [2]. RAGE is composed of an N-terminal variable-type (V) domain, two distinct C-type Ig-like domains (C1 and C2), a transmembrane helix domain (TMH) and a highly charged cytoplasmic tail [3]. The V-type domain is generally involved in ligand binding, and the highly charged cytoplasmic tail is associated with the activation of intracellular signal transduction pathways [4]. This signaling receptor is involved in a wide range of inflammation-related pathological states, such as vascular diseases, diabetes, neurodegeneration and cancer [5], [6], [7], [8]. The activation of RAGE and the signal transduction that follows is also dependent on the cell type and ligand concentration [9], [10]. Understanding RAGE signaling is invaluable for the prevention of various diseases. RAGE can interact with a variety of ligands, including advanced glycation end products (AGE) [1], [2], DNA [11], amphoterin (HMGB1) [12], β-amyloid [13] and S100 family proteins [14], [15]. RAGE ligation and its subsequent activation play a role in multiple signaling cascades, such as the MAPK, JNK and Cdc42/Rac pathways, and activate the transcription factors AP-1 and NF-κB [16], [17], [18]. Previous studies have suggested the possibility of RAGE TMH dimerization during signal transduction [19], [20], [21]. The homodimerization of RAGE is an important step for receptor activation during ligand binding and, thus, for the induction of various signaling cascades [22], [23]. The ligation of RAGE by its targets, such as S100B and AGEs, leads to the enhanced formation of RAGE homodimers and is also associated with amplified signal transduction and transcriptional activation [24].

S100 proteins are calcium-binding proteins that represent a subfamily of EF-hand proteins. The members of this family share a common topology that consists of two calcium binding helix-loop-helix structural domains that mediate calcium-dependent signal transduction. These S100 proteins engage in a large number of intracellular and extracellular functions [86]. The binding of calcium to the EF-hand motif of S100 proteins induces conformational changes that facilitate the interaction of the hydrophobic interfaces on opposite sides of the homodimer with target proteins, thereby mediating their activity [42]. Previous studies strongly indicate that S100P can be secreted, acts through RAGE in an autocrine manner and plays a significant role in the development and progression of various cancers [87]. The binding of S100P to the RAGE V domain results in RAGE homodimerization and the activation of its cytoplasmic domain for autophosphorylation, leading to the activation of the ERK and MAPK pathways to mediate cell proliferation and survival (Figure 10). This finding indicates additional significance of the V domain in RAGE homodimerization and signal transduction. Therefore, the characterization of the molecular interactions between S100P and the V domain of RAGE provides an opportunity to precisely identify the role of crucial residues at the interface of both proteins. The interactions between the RAGE V domain and S100P were characterized using various biophysical techniques, including isothermal titration calorimetry (ITC), fluorescence spectroscopy, multidimensional NMR spectroscopy and site-directed mutagenesis. The dissociation constant (Kd) estimated from ITC and fluorescence spectroscopy is approximately 6.0 µM, suggesting a moderately strong interaction. Our ITC results also suggested the presence of two identical and independent binding sites between two RAGE V domains and a calcium-bound S100P homodimer, supporting the hypothesis that hydrophobic interfaces are buried between the proteins upon complex formation. Additional NMR titration experiments identified the putative binding interface for the S100P-RAGE V domain complex. Mapping of the binding interface residues indicated the nature of the binding interfaces of S100P and the RAGE V domain. Overall, our findings reveal the extracellular role of the S100P homodimer, which symmetrically interacts with two RAGE V molecules. Based on these conclusions, we modeled the S100P-RAGE V heterotetrameric complex using HADDOCK, which revealed the putative interfaces between S100P and the RAGE V domain. An analysis of the total interface area for the HADDOCK-calculated S100P-RAGE V domain complex is approximately 4738 Å2 per two interfaces, with a ratio of the hydrophobic to hydrophilic areas of 58∶42. Interestingly, the RAGE V domain residues at the S100P binding interface, particularly R48, K52, R98 and R104, are similar to those that interact with S100B, S100A6 and AGE. Overall, we can infer that the binding interface of the RAGE V domain is well conserved for its interaction with S100P and is similar to that for other known RAGE V domain binding partners, as has been observed in the RAGE-AGE complex, the S100B-RAGE VC1 model complex, and the S100A6-RAGE V model complex [44]. Our docking results and mutagenesis study indicated that hydrophobic residues, such as F44 in the central linker region and Y88 and F89 in helix-4, and polar residues, including E5 and D13 in helix-1′, in S100P permitted the unique recognition of the RAGE V domain. Overall, our results demonstrated the ability of the S100P homodimer to form two symmetrical binding interfaces from helix 4 and the linker region of one monomer and helix 1′ from the other monomer to interact with RAGE. S100A12 was previously shown to bind to the C1C2 domain of RAGE in a similar fashion using two symmetrical hydrophobic patches from helix-2, loop-2 and helix-4 of one monomer and helix-1′ of a second monomer [88]. Furthermore, the differences in the S100P binding interface for RAGE V domain complex formation can be ascertained from the reported interactions between various S100 proteins and the RAGE V domain. It has been reported that the binding interface between S100B and the RAGE V domain consists of H42, V52, N62, D62, D69, F70 and A78, most of which are located on helix-4, helix-3, loop-1 and loop-3 [22]. For S100A11, the V domain-binding residues are located in helix-2 and helix-4 [89]. The model of the S100A6-RAGE V domain complex suggests that the RAGE V domain-binding surface of S100A6 is predominantly distributed over loop 1, loop 3 and helix 4 [75]. These differences in the binding interfaces between various S100 proteins and RAGE suggest that RAGE recognizes its S100 ligands based on their net charge, their polarity and the hydrophobic nature of the interface. The modeled structure of the S100P-RAGE V domain complex is useful for the improvement of current drug antagonists and may aid in the design of improved antagonists to disrupt the interaction between S100P and the V domain of RAGE. We identified pentamidine as a potential small molecule antagonist that disrupts this interaction using NMR and ITC analysis, HADDOCK modeling and mitogenic assays. Our present findings highlight the significance of cell proliferation induced by the interaction of S100P with RAGE. This study will also aid in the design of improved therapeutics to antagonize the S100P-RAGE interaction to prevent RAGE-dependent diseases.