Structural insights reveal a recognition feature for tailoring hydrocarbon stapled-peptides against the eukaryotic translation initiation factor 4E protein

Abstract

Stapled-peptides have emerged as an exciting class of molecules which can modulate protein-protein interactions. We have used a structure-guided approach to rationally develop a set of hydrocarbon stapled-peptides with high binding affinities and residence times against the oncogenic eukaryotic translation initiation factor 4E (eIF4E) protein. Crystal structures of these peptides in complex with eIF4E show that they form specific interactions with a region on the protein-binding interface of eIF4E which is distinct from the other well-established canonical interactions. This recognition element is a major molecular determinant underlying the improved binding kinetics of these peptides with eIF4E. The interactions were further exploited by designing features in the peptides to attenuate disorder and increase helicity which collectively resulted in the generation of a distinct class of hydrocarbon stapled-peptides targeting eIF4E. This study details new insights into the molecular basis of stapled-peptide: eIF4E interactions and their exploitation to enhance promising lead molecules for the development of stapled-peptide compounds for oncology.

Fig. 1. sTIP-04 peptide. Crystal structure of hydrocarbon stapled sTIP-04 peptide (¹KKRYSR*QLL*L¹²) in complex with eIF4E (PDB ID: ; 4BEA). The protein eIF4E is shown in surface (gray) and the backbone of the peptide in ribbon (green) representations respectively. The side-chain of the peptide residues are explicitly shown in stick representation and labeled. The hydrocarbon linker is highlighted in orange color. This depiction is followed in the rest of the figures unless specified. All the molecular graphics figures were created using PyMol molecular visualization software (Schrödinger).

Fig. 2. Canonical binding and interactions. Crystal structures of (A) sTIP-05 (PDB ID: 5ZJY), (B) sTIP-07 (PDB ID: ; 5ZJZ), (C) sTIP-08 (PDB ID: ; 5ZK9), (D) sTIP-09 (PDB ID: ; 5ZML), (E) sTIP-10 (PDB ID: ; 5ZK5) and (F) sTIP-14 (PDB ID: ; 5ZK7) hydrocarbon stapled-peptides in complex with eIF4E underlining the common binding mode and conserved interactions across all the structures. The backbone of the peptides is shown in ribbon (green), the protein in surface (gray) and the hydrocarbon linker is explicitly shown in stick (orange) representation. The backbone stereochemistry of the hydrocarbon linker in sTIP-05 is (R,R) whereas all the other peptides are in the (S,S) configuration. The residues involved in hydrogen-bond and salt-bridge interactions are indicated. The pocket where the conserved leucine residue (L9) docks onto the protein is specified by an arrow. The residue numbering for the protein is done as per the native eIF4E protein sequence (Uniprot ID: P06730).

Fig. 3. Untapped patch and its engagement. Crystal structures of (A) sTIP-05 (PDB ID: 5ZJY), (B) sTIP-07 (PDB ID: ; 5ZJZ), (C) sTIP-08 (PDB ID: ; 5ZK9), (D) sTIP-09 (PDB ID: ; 5ZML), (E) sTIP-10 (PDB ID: ; 5ZK5) and (F) sTIP-14 (PDB ID: ; 5ZK7) hydrocarbon stapled-peptides in complex with eIF4E highlighting the untapped patch on the protein and its engagement by different peptides. The residues forming the patch on eIF4E are emphasized with a different colour combination as compared to the rest of the protein. The side-chain of the residues from the peptide that interact with the patch are explicitly shown and the hydrogen-bond wherever formed is indicated. Residue “&13” is the resolved “Lys” moiety of the modified “Lys(ButPhI)” amino acid in sTIP-08.

Fig. 4. Residue-wise binding energy contribution. The average binding energy and standard deviation is computed from the ensemble of structures generated from the MD simulations of (A) sTIP-05, (B) sTIP-07, (C) sTIP-08, (D) sTIP-09, (E) sTIP-10 and (F) sTIP-14 hydrocarbon stapled-peptides in complex with eIF4E. The amino acid sequence of the respective peptides is indicated in the plot. The non-natural amino acids forming the hydrocarbon linker are represented by “*”. The calculation was done using the Molecular Mechanics/Generalized Born Surface Area (MM/GBSA) method by following the same procedure and parameters as described previously.

Fig. 5. Physicochemical property of the N-terminal. (A) The canonical bound-state structure of the 12mer hydrocarbon stapled-peptide represented by sTIP-05. The disordered N-terminal region (absence of a defined secondary structure) and the ordered C-terminal segment are orthogonal to each other and highlighted to indicate the “Reverse L-shaped conformation”. (B) B-Factor and root mean square fluctuation (Rmsf) values of the CA atoms of the peptides. Only residues 1–12 across all the peptides are compared. The original B-factors values of sTIP-05 and sTIP-14 were respectively multiplied and divided by a factor of two for comparative analysis with other peptides. The rmsf (shown in inset) is computed with reference to the energy minimized structure of the respective peptides. The cylindrical and rectangular sketch below the plot represents the ordered helical and disordered random states respectively. (C) Crystal structure of sTIP-05 in complex with eIF4E. The protein is represented in electrostatic surface and the side-chain of the residues in the N-terminal region of the peptide are shown explicitly. The electrostatic potential surface was created using the APBS plugin through the PyMol molecular visualization software (Schrödinger). A colour gradient from blue to red represents the range of surface potential kT/e values from strongly positive (+5.0) to strongly negative (–5.0). (D) Isoforms of human eIF4G and 4EBP proteins, their Uniprot ID and the respective 12mer peptide segments that interact with eIF4E. The residues across these peptides that are structurally equivalent to the N-terminal region in sTIP-05 are emphasized in blue colour.

Fig. 6. Modelled complex and residue-wise binding energy. (A) Modelled complex structure of sTIP-15 and eIF4E. The conserved canonical interactions, the new detected patch on the protein surface and its engagement by the C-terminal residue of the peptide are highlighted. Residue “R12” in the peptide is not shown for clarity. (B) The average binding energy and standard deviation of the hydrocarbon stapled-peptide residues computed from the ensemble of structures generated from MD simulations of sTIP-15 and eIF4E complex. The non-natural amino acids forming the hydrocarbon linker is represented by “*”. The computation was done using MM/GBSA method as described previously.

Fig. 7. Terminal modulation and structural conformation of hydrocarbon stapled-peptide. Superimposition of the bound-state structures of sTIP-10 and sTIP-14 which highlight the distinct variations in the terminal regions of the peptide and the positions of the hydrocarbon linker. The backbone of K1 residue in sTIP-10 is modelled for comparison since it is not resolved in the crystal structure. The residues forming the patch on eIF4E are emphasized with a different colour combination as compared to the rest of the protein. The cylindrical and rectangular sketch represents the ordered helical and floppy disordered states respectively.