(K-406) Controlling stiffness and porosity of three-dimensional collagen scaffolds

Approximately half the patients who experienced severe burn wounds suffer from chronic pain and itch. The underlying mechanisms of such conditions are believed to be neuropathies, yet there remains a key knowledge gap regarding the mechanisms involved. Our lab has developed an in vitro model to investigate: an innervated engineered skin substitute capable of recreating skin morphology and physiology. Studies show that neurite extension typically increases as tissue stiffness decreases; however, this does not appear to hold true at very low stiffnesses. Additionally, preliminary data showed complete inhibition of neurite ingrowth in a 3D hydrogel at 70/30 collagen type I/III composition. These results led us to believe crucial morphological factors in the extracellular matrix (ECM) dictated neurite development caused by the structural differences between the different scaffolds. In this project, we aimed to develop 3D scaffolds to capture these phenomena on innervation and cell behaviors, thus gaining better insights into tissue-engineered models. 

To fabricate the three-dimensional collagen hydrogels, acid-solubilized bovine dermal collagen I was neutralized and mixed with cell culture medium for a final concentration of 0.02% (wt./v) collagen. Under this condition, collagen fibrils assembled and formed a gel-like structure. The resulting hydrogels underwent crosslinking with 0%, 0.02%, and 0.05% (v/v) glutaraldehyde for 2 hours, then washed with PBS. To prepare a sponge, the hydrogels were frozen at -80^oC overnight and lyophilized for 24 hours (-80^oC, 0.07-0.1 mbar). Lyophilized collagen scaffolds were kept dried until analysis. Rheological analysis was performed on rehydrated collagen sponges using an HR-20 rheometer (TA instrument). The viscoelastic storage modulus of each scaffold at 10 rad/s under oscillation/frequency mode was reported for comparison. High-magnification imaging of the scaffolds was carried out using scanning electron microscopy (SEM) (FEI Apreo) under 150x and 5000x magnifications, Low-Vac mode. Post-imaging analysis was performed on ImageJ.

Statistical analysis was performed using a one-way ANOVA followed by Games-Howell post hoc test. A p-value of less than 0.05 was considered significant.

Results

The storage modulus of rehydrated collagen sponges increased as glutaraldehyde concentration increased (p = 0.004). Post-hoc test results showed no difference between non-crosslinked and 0.02% crosslinked sponges, but 0.05% crosslinked sponges were significantly stiffer than the other two. Crosslinking degree also affected matrix pore size (p < 0.001). A pairwise comparison test confirmed significant differences among sample types.

Discussions

The measured stiffness tendency was as expected. An increase in the crosslinked structure provided additional mechanical integrity, thus increasing stiffness. More importantly, the 0.05% crosslinked sponges reached 1.2 kPa mean storage modulus, resembling brain tissue stiffness at 1-3kPa.

The same explanation also applied to pore size. More inter-fibril connecting structures theoretically reduced void space within the matrix, which could be seen from the crosslinked sponges (mean pore area at 0.32 um² to 0.22 um²). However, the result showed that the non-crosslink structure had the smallest pore area (mean of 0.15 um²). This could be a side effect of lyophilization when collagen fibrils fused after all water content within the hydrogels evaporated, and there were no preformed structures to withstand the pressure inside the chamber. Though, measured pore sizes were still relatively small compared to a neuron body (approximately 20 um in diameter) and even the neurite itself. 

Experimental limitations should also be addressed. Determining pore size within a 3D structure by SEM was a biased method. Hence there might be some deviation from the actual data. Also, SEM images showed signs of sodium chloride particle formations, both on the surface and buried within the sponges. Salt particles caused uneven exposure to the surrounding area, thus causing unwanted contrast to the images and deforming the collagen structure at a certain level.  

Conclusions