Statistic Copolymers Working as Growth Factor-Binding Mimics of Fibronectin

Abstract

Growth factors (GFs) play important roles in biological system and are widely used in tissue regeneration. However, their application is greatly hindered by short in vivo lifetime of GFs. GFs are bound to fibronectin dynamically in the extracellular matrix, which inspired the authors to mimic the GF binding domain of fibronectin and design GF-binding amphiphilic copolymers bearing positive charges. The optimal amino acid polymer can bind to a variety of representative GFs, such as bone morphogenetic protein-2 (BMP-2) and TGF-β1 from the transforming growth factor-β superfamily, PDGF-AA and PDGF-BB from the platelet-derived growth factor family, FGF-10 and FGF-21 from the fibroblast growth factor family, epidermal growth factor from the EGF family and hepatocyte growth factor from the plasminogen-related growth factor family, with binding affinities up to the nanomolar level. 3D scaffolds immobilized with the optimal copolymer enable sustained release of loaded BMP-2 without burst release and significantly enhances the in vivo function of BMP-2 for bone formation. This strategy opens new avenues in designing GF-binding copolymers as synthetic mimics of fibronectin for diverse applications.

Keywords: fibronectin mimicking; growth factor binding; statistic copolymers.

Figure 1

Fibronectin (FN) mimicking amino acid copolymer (AA copolymer) displaying growth factor (GF) binding function. a) FN is a ubiquitous ECM protein and possesses a binding domain, FNIII 12–14, toward diverse classes of GFs. Blue color represents the positive charge, red color refers to the negative charge and white color represents hydrophobic. b) The scheme for the synthesis of 28 FN mimicking amphiphilic AA copolymers. c) High throughput screening for GFs binding AA copolymers using an immunofluorescence assay. d) Heat map summarizing GF adsorption evaluation for AA polymer library (mean values from n = 2 replicates for each polymer). e‐h) GF binding ability to variable AA copolymers as evaluated from GFs adsorption study for transforming growth factor‐β (TGF‐β) superfamily (e), platelet‐derived growth factor (PDGF) family (f), fibroblast growth factor (FGF) family (g), and other families (h) (*p < 0.05, **p < 0.01, n = 5). Mean ± SD are shown.

Figure 2

Growth factors binding affinity of the selected AA copolymer (Lys_0.4Nle_0.6) using surface plasmon resonance (SPR) analysis. a–d) SPR chip was functionalized with the selected AA copolymer (Lys_0.4Nle_0.6), and each protein was flown over the chip at indicated concentrations. e) GFs and Lys_0.4Nle_0.6 binding kinetics values determined from the experimental curve fits. ¹ND means bovine serum albumin (BSA) has negligible binding to Lys_0.4Nle_0.6 as is not detectable.

Figure 3

a) The amino acid sequence of the Lys‐Nle peptide. b) BMP‐2 binding affinity of the selected Lys‐Nle peptide using SPR analysis. BMP‐2 and Lys‐Nle peptide binding kinetics values determined from the experimental curve fits. c) 3D structure of Lys‐Nle peptide. d) The 3D binding model between BMP‐2 and Lys‐Nle peptide. The backbone of chain A of BMP‐2 is depicted as dark green cartoon, the backbone of chain B of BMP‐2 is depicted as green cartoon. The residues in BMP‐2 are colored in cyan. Lys‐Nle peptide is depicted as yellow cartoon. The residues in Lys‐Nle peptide are colored in yellow. e) The surface binding model between BMP‐2 and Lys‐Nle peptide. The surface of chain A of BMP‐2 is colored in dark green, the surface of chain B of BMP‐2 is colored in green. Lys‐Nle peptide is depicted as yellow cartoon. The lightblue dashes represent hydrogen bond interaction. There are salt bridges between the groups marked by the fog group. f) The 3D binding model between BMP‐2 and FNIII 12–14. The backbone of chain A of BMP‐2 is depicted as dark green cartoon, the backbone of chain B of BMP‐2 is depicted as green cartoon. The residues in BMP‐2 are colored in cyan. The backbone of FNIII 12–14 is depicted as pink cartoon. The residues in FNIII 12–14 are colored in pink. g) The surface binding model between BMP2 and FNIII 12–14. The surface of chain A of BMP‐2 is colored in dark green, the surface of chain B of BMP‐2 is colored in green and the surface of FNIII 12–14 is colored in pink. The lightblue dashes represent hydrogen bond interaction. There are salt bridges between the groups marked by the fog group.

Figure 4

Fabrication and characterization of AA copolymers‐immobilized 3D scaffolds. a) Covalently immobilize the selected AA copolymer to gelatin scaffold (Gel) via a one‐step coupling between the amine groups in Lys_0.4Nle_0.6 on the carboxyl groups on the surface of Gel. b) High‐resolution X‐ray photoelectron spectroscopy characterization on modified scaffolds revealed the successful immobilization of Lys_0.4Nle_0.6 onto the Gel scaffolds. Black line represents experimental spectrum; blue line represents fitting spectrum; red/green/azure line represents peak decomposition. c) Scanning electron microscope characterization on the morphology of Gel and Gel‐Lys_0.4Nle_0.6 scaffolds. d) Cell density in bare Gel and Gel‐Lys_0.4Nle_0.6 scaffolds (**p < 0.01, n = 3). Representative images are shown. Mean ± SD are shown.

Figure 5

Delivering BMP‐2 within AA copolymers‐immobilized 3D scaffolds induces the osteogenic differentiation in vitro and in vivo. a) BMP‐2 loaded Gel scaffolds. b) Release behavior of BMP‐2 from Gel scaffolds were quantified by ELISA. c) Expression of osteogenic genes marker (corresponding to runt‐related transcription factor 2 (Runx2), type I collagen (Col I), and osteocalcin (OCN)) within C2C12 cells cultured in variable scaffolds for 3 days and 7 days, using cells cultured on the tissue culture plate as the control. d) Western blot analysis on protein expression of Runx2, Col I and OCN within C2C12 cells cultured in variable scaffolds for 3 days and 7 days, using cells cultured on the tissue culture plate as the control. e) The scheme for ectopic bone formation experiments. f) Harvested implants after 2 and 4 weeks were evaluated by bone wet weight. g) 3D SRµCT of obtained ectopic bones and corresponding quantified data of bone volume fraction after implants were harvested 2 and 4 weeks post implantation. h) Histological evaluation of harvested ectopic bone sections by Masson's trichrome staining (F: fibrous tissue, TB: trabecular bone) (*p < 0.05, **p < 0.01, n = 3). Representative images are shown. Mean ± SD are shown.