Polycaprolactone (PCL) fiber mats with different surface modifications were functionalized with a chitosan nanogel coating to attach the growth factor human bone morphogenetic protein 2 (BMP‐2). Three different hydrophilic surface modifications were compared with regard to the binding and in vitro release of BMP‐2. The type of surface modification and the specific surface area derived from the fiber thickness had an important influence on the degree of protein loading. Coating the PCL fibers with polydopamine resulted in the binding of the largest BMP‐2 quantity per surface area. However, most of the binding was irreversible over the investigated period of time, causing a low release in vitro. PCL fiber mats with a chitosan‐graft‐PCL coating and an additional alginate layer, as well as PCL fiber mats with an air plasma surface modification boundless BMP‐2, but the immobilized protein could almost completely be released. With polydopamine and plasma modifications as well as with unmodified PCL, high amounts of BMP‐2 could also be attached directly to the surface. Integration of BMP‐2 into the chitosan nanogel functionalization considerably increased binding on all hydrophilized surfaces and resulted in a sustained release with an initial burst release of BMP‐2 without detectable loss of bioactivity in vitro.

No abstract

Purpose: There is a plethora of studies on recombinant human bone morphogenetic protein-2 (rhBMP-2) application and delivery systems, but surprisingly few reports address the biophysical properties of the protein which are of crucial importance to develop effective delivery systems or to solve general problems related to rhBMP-2 production, purification, analysis and application. Methods:The solubility, stability and bioactivity of rhBMP-2 obtained by renaturation of E. coli derived inclusion bodies was assessed at different pH and in different buffer systems using (dynamic) light scattering and thermal shift assays as well as intrinsic fluorescence measurements and luciferase based bioassays.Results: rhBMP-2 is poorly soluble at physiological pH and higher. The presence of divalent anions further decreases the solubility even under acidic conditions. Thermal stability analyses revealed that rhBMP-2 precipitates are more stable compared to the soluble protein. Moreover, correctly folded rhBMP-2 is also bioactive as precipitated protein and precipitates readily dissolve under appropriate buffer conditions. Once properly formed rhBMP-2 also retains biological activity after temporary exposure to high concentrations of chaotropic denaturants. However, care should be taken to discriminate bioactive rhBMP-2 precipitates from misfolded rhBMP-2 aggregates, e.g. resolvability in MES buffer (pH 5) and a discrete peak in thermoshift experiments are mandatory for correctly folded rhBMP-2.Conclusions: Our analysis revealed that E. coli derived rhBMP-2 precipitates are not only bioactive but are also more stable compared to the soluble dimeric molecules. Knowledge about these unusual properties will be helpful to design improved delivery systems requiring lower amounts of rhBMP-2 in clinical applications.

Recombinant human bone morphogenetic protein-2 (rhBMP-2), a cystine-knot containing disulfide linked homodimer, was produced in form of inclusion bodies (IBs) using E. coli BL21 (DE3) and SHuffle T7 Express. Non-reducing SDS-PAGE analysis revealed that rhBMP-2 was present within IBs in both strains as monomer and in form of disulfide-linked dimers. Purified dimeric disulfide-linked rhBMP-2 was obtained from IBs by two different methods. The first method involved classical solubilisation using strong denaturants and subsequent refolding. The second method involved mild extraction without refolding. Both rhBMP-2 dimer variants were purified by Heparin-affinity chromatography. Mildly extracted rhBMP-2 was further purified by size-exclusion chromatography. The resulting dimeric rhBMP-2 variants were studied regarding bioactivity and folding status. In contrast to the rhBMP-2 dimer obtained by classical refolding it was shown that the disulfide-linked dimer obtained by mild extraction was not correctly folded e.g. the hydrophobic core not correctly formed. Moreover, refolded dimeric rhBMP-2 was bioactive and the mildly extracted dimeric rhBMP-2 did not show any bioactivity. Disulfide-bond analysis revealed that the intricate disulfide-bond pattern of the complex cystine-knot scaffold was not present in the IB embedded disulfide-linked rhBMP-2 dimer but was formed later during classical refolding of the reduced protein under appropriate redox conditions.