(314h) Controlling Covalent Versus Non-Covalent Patterning of Active Biomolecules within Collagen Glycosaminoglycan Scaffolds

Collagen-GAG (CG) scaffolds are a well-studied class of biomaterial used in regenerative medicine. Biomolecule (i.e. growth factors, proteins) incorporation into these materials via soluble supplementation strategies have been suggested to augment scaffold bioactivity for a range of in vitro and in vivo applications.  Insoluble patterning of these same molecules offers the potential to control their spatial distribution, control the kinetics of factor release, and overcome many of the translational shortcomings of soluble supplementation (i.e. rapid factor loss via diffusion). In the context of developing spatially- and temporally-controlled patterning techniques for a range of instructive biomolecules within collagen GAG scaffolds to improve angiogenic potential, we are first investigating the balance between covalent and non-covalent presentation of proteins within the scaffold.

The charged nature of collagen-GAG scaffolds suggested that they may form a significant fraction of non-covalent bonds with biomolecules and proteins in solution; this interaction may be suitable for biomolecules that do not function in the bound form as they are capable of transient dissociation. However, some biomolecules can be as effective if not more when covalently bound. Further, covalent bonds allow the biomolecules to stay active and prevent dissociation. In these instances, other methods of immobilization should be considered; for this we focused on assessing biomolecule immobilization using a common carbodiimide (EDC) crosslinking reagent previously used to alter CG scaffold crosslinking density and mechanics. For these studies we used Con A biotin and Con A biotin-Alexafluor 637 conjugate as our model proteins and small biomolecules. We assessed the influence of EDC concentration and exposure time on covalent factor attachment, as well as the use of pre-exposure blocking (i.e. sucrose) and post-exposure scavenging (i.e. sucrose) agents to alter non-specific binding (‘fouling’) and transient release.

Non-covalent and covalent binding were evaluated via fluorescence and compared to the total biomolecule content expores to the scaffold surface.  

While pre-exposure blocking and rigorous post-exposure washing steps reduced this fouling, we further hypothesized surface charges associated with the CG content were a major player in mediating non-specific attachment. However, we did not observe fouling as a function of modulating CG scaffold GAG content, using relatively less sulfated hyaluronic acid versus our standard chondroitin sulfate GAG. We additionally investigated the role of solution pH and addition of sucrose to the biomolecule suspension to reduce charge-associated non-specific biomolecule attachment.

We observed non-covalent binding PBS solutions at a pH of 5 (48%) or 9 (51%) were significantly higher at p<.05 than those at pH 7 (34%) or solutions with sucrose (25%).  The data suggest that there is a significant difference between sucrose additional after protein (16%) and sucrose before protein (27%, p=0.001) as well as between sucrose after protein and the scaffolds treated with PBS (27%, p<.001). Further, a significant main effect of time (p =0.009) was found for exposure times of 5 min and 20 min with percent of non-covalent binding of 20% and 26% respectively. At a small sample size, n =9, the only significant difference at p<.001 found for covalent binding by EDAC cross-linking was found between pH 7 and 5. This suggests that the sucrose is facilitating non-covalently bound biomolecule extraction from the scaffold, making it an attractive option for modulating non-specific biomolecule attachment to the scaffold surface. Ongoing work is now examining the use of covalent vs. non-covalent immobilization of estrogen to the scaffold surface as a method for impacting human endometrial epithelial cell bioactivity (proliferation, VEGF production) in order to improve scaffold angiogenic potential.