In silico investigation of Alsin RLD conformational dynamics and phosphoinositides binding mechanism

Abstract

Alsin is a protein known for its major role in neuronal homeostasis and whose mutation is associated with early-onset neurodegenerative diseases. It has been shown that its relocalization from the cytoplasm to the cell membrane is crucial to induce early endosomes maturation. In particular, evidences suggest that the N-terminal regulator of chromosome condensation 1 like domain (RLD) is necessary for membrane association thanks to its affinity to phosphoinositides, membrane lipids involved in the regulation of several signaling processes. Interestingly, this domain showed affinity towards phosphatidylinositol 3-phosphate [PI(3)P], which is highly expressed in endosomes membrane. However, Alsin structure has not been experimentally resolved yet and molecular mechanisms associated with its biological functions are mostly unknown. In this work, Alsin RLD has been investigated through computational molecular modeling techniques to analyze its conformational dynamics and obtain a representative 3D model of this domain. Moreover, a putative phosphoinositide binding site has been proposed and PI(3)P interaction mechanism studied. Results highlight the substantial conformational stability of Alsin RLD secondary structure and suggest the role of one highly flexible region in the phosphoinositides selectivity of this domain.

Fig 1. Alsin RLD structural refinement.

(A) RLD^AF model is characterized by three loops protruding from the propeller, i.e. L1 (green), L2 (blue), and L3 (red). Chain A and B are colored in grey and white, respectively. (B) Refined Alsin RLD model. Residues are colored according to their RMSF to highlight in red the most flexible regions. For visualization purposes, the highly flexible N-terminal, aa. 220–225, and aa. 520–525 are not shown. (C) Secondary structure probability for each residue compared to the initial structure of RLD^AF. The region with increased turn probability after the dynamics is highlighted.

Fig 2. Identification of a putative PIP-binding site in Alsin RLD.

(A) Visual rendering of the investigated binding pocket. The positive and hydrophobic surface patches are shown in blue and green, respectively, while residues forming the binding site are colored in red. (B) Binding pose of PI(3)P at the end of the dynamics, where interactions between two atoms are represented by dashed lines. The strength of an interaction is visually represented by the length of the cylinder. (C) Superimposition of RLD query residues (light grey) and the corresponding amino acids of the best hit found through ASSAM (dark grey). In the figure, residue numbers are reported according to the scheme N_query/N_hit.

Fig 3. Comparison of RLD-PIPs interaction.

(A) Decomposition of PI(3)P and PI(3,4,5)P₃ binding free energy, where error bars represent standard deviations. (B) Visual rendering of PI(3)P binding energy decomposition. (C) Visual rendering of PI(3,4,5)P₃ binding energy decomposition. (D) Comparison of RMSF values in absence and presence of PIPs. The shaded areas represent the standard deviation of the RMSF between the simulations. Loop L1, L2, and L3 are highlighted within the RMSF profile.

Fig 4. Analysis of L3 conformational dynamics.

L3 dynamics has been described as RMSD from its position in the refined RLD model and distance from the investigated binding site. The probability distribution in the case of RLD alone, PI(3)P-bound RLD, and PI(3,4,5)P₃-bound RLD is shown in black, blue, and red, respectively. Representative configurations of L3 are shown highlighting its distance from the binding pocket in red.