Re-engineered stromal cell-derived factor-1α and the future of translatable angiogenic polypeptide design

Abstract

Smaller engineered analogs of angiogenic cytokines may provide translational advantages, including enhanced stability and function, ease of synthesis, lower cost, and, most important, the potential for modulated delivery via engineered biomaterials. In order to create such a peptide, computational molecular modeling and design was employed to engineer a minimized, highly efficient polypeptide analog of the stromal cell-derived factor-1α (SDF) molecule. After removal of the large, central β-sheet region, a designed diproline linker connected the native N-terminus (responsible for receptor activation and binding) and C-terminus (responsible for extracellular stabilization). This yielded energetic and conformational advantages resulting in a small, low-molecular-weight engineered SDF polypeptide analog (ESA) that was shown to have angiogenic activity comparable to or better than that of recombinant human SDF both in vitro and in a murine model of ischemic heart failure.

Figure 1

Crystallographic structures of SDF(Dealwis et al. 1998) and designed model structure ESA. The different regions of the structure are colored as, N terminal (green), central region (yellow) and C terminal (magenta). The central β-sheet region (yellow) is replaced by a diproline linker in ESA. The corresponding amino acid sequences of SDF-1α and ESA are also depicted, where the different regions are colored accordingly.

Figure 2

Top and side view of experimental SDF analog peptides utilizing one proline (a), two proline (b), and three proline (c) residues to link the N and C terminus. The images depict a composite of the 50 most energetically stable conformations of each peptide sequence. The peptide with the two proline linker (b) adopts a more uniform tertiary profile than the others and recovers the perpendicular orientation between the N- and C- termini found in native SDF.

Figure 3

Space-filling representation of the ESA and SDF-1α structures. The bulky side chain of phenylalanine 14 (F14^ESA) is interacting with glutamic acid 15 (E15^ESA), histidine 17 (H17^ESA) and to some extent with leucine 20 (L20^ESA or residue 55 in native SDF-1α). The interactions of F14^SDF, E15^SDF, H17^SDF and L55^SDF are absent in crystallographic structure of SDF-1α. These new interactions form a small cluster-like structure which, when coupled with the diproline spacer (yellow in ESA), may help to provide the necessary conformational stability and rigidity found in native SDF.

Figure 4

Homology modeling of the CXCR4 receptor and docking of SDF and ESA yields detailed characterization of the atomic interactions of the systems CXCR4/SDF and CXCR4/ESA. This figure shows a superposition of docked SDF (gold) and ESA (green) in the receptor complex, where equivalent positions of both ligands contact the surface of CXCR4. As proposed in previous studies, the N-terminal region of both ligands is bound to the receptor in a relatively shallow groove. In the model, ESA positions the C-terminal helix in a manner that superimposes its location on SDF.