Cross-fertilization between hydrogel development and additive manufacturing as emerging strategy towards adipose tissue engineering

Student thesis: Doctoral Thesis

Abstract

The goal of the present PhD included the development of photo-crosslinkable hydrogel building blocks which can be 3D printed into patient-specific or injectable scaffolds using AM techniques for adipose TE applications. In this respect, different photo-crosslinkable gelatins, recombinant collagens and κ-carrageenans were developed and processed via either extrusion-based or two-photon polymerization 3D printing. Over the years, animal-derived gelatin has been frequently applied for biomedical applications and can thus be considered the gold standard within the field of (adipose) TE. However, there are increasing concerns associated with the use of materials derived from animal sources for applications in modern healthcare. The drawbacks include issues with product reproducibility due to batch-to-batch variations and the risk of pathogen transmittance to occur including prions. Therefore, the potential of a recombinant protein based on collagen I (RCPhC1) to function as a mimic for animal-derived gelatin was investigated. In the present work, both RCPhC1 and gelatin were modified using MA to introduce photo-crosslinkable methacrylamide functionalities. RCPhC1-MA and Gel-MA with a DS of 90% and 97% respectively were developed. Subsequently, the modified materials were crosslinked into 2D hydrogel films upon UV irradiation in the presence of Li-TPO-L as photo-initiator. The hydrogels developed were physico-chemically characterized by gel fraction determination, swelling experiments and mechanical tests as well as in vitro biological assays and cell encapsulation experiments. The results indicated that the physical crosslinking behaviour of RCPhC1-MA was affected by the presence of the photo-crosslinkable functionalities which hampered the gel formation. The gel fraction results together with the number of reacted functionalities indicated that for both materials the UV-induced crosslinking was successful and that stable hydrogel films were formed. In addition, the mechanical properties of the hydrogels were in the same range (6000 Pa) for a polymer concentration of 10 w/v%. However, the water uptake capacity of the crosslinked RCPhC1-MA hydrogels was lower compared to the Gel-MA films due to its lower hydrophilicity. The in vitro biological evaluation indicated that both RCPhC1-MA and Gel-MA are biocompatible towards ASCs and MSCs. In addition, the cell encapsulation experiments showed that the encapsulated ASCs survived the encapsulation process and started to proliferate within the hydrogels which is an indication that RCPhC1-MA is an excellent ECM mimic with a comparable stem cells response to Gel-MA. Therefore, it can be concluded that the properties of RCPhC1-MA are comparable with those of the gold standard Gel-MA after crosslinking. RCPhC1-MA is thus an attractive synthetic alternative for animal-derived Gel-MA to be applied for different biomedical applications including adipose TE. In a next part of this PhD, a protein and a sulphated polysaccharide were combined in order to create a superior mimic of the ECM of native tissue. More specifically, gelatin and κ-carrageenan were selected as hydrogel building blocks due to their close resemblance to collagen and the glycosaminoglycans respectively which are the main components of the ECM. Both biopolymers were successfully modified with MA to obtain a DS of 97% and 15% for Gel-MA and Car-MA respectively. Subsequently, the materials were photo-crosslinked into hydrogel films composed of either Gel-MA, Car-MA or a combination of both biopolymers, i.e. hydrogel blend. The results indicated that chemically and dual crosslinked hydrogel films could be developed which were characterized by gel fractions exceeding 70%. The mechanical properties could be tuned by varying the Car-MA and/or Gel-MA concentration as well as the crosslinking method applied (i.e. chemical versus dual crosslinking) which also resulted in tunable swelling characteristics. The hydrogel crosslink density results obtained via the rubber elasticity theory showed that the Gel-MA hydrogels possessed a smaller mesh size and molecular weight between the crosslinks and accordingly, a higher crosslink density and higher mechanical properties as compared to Car-MA. The in vitro biocompatibility assays indicated that the dual crosslinked hydrogel blends supported cell adhesion and proliferation of the seeded ASCs. Furthermore, the hydrogel blend composed of 10 w/v% Gel-MA and 5 w/v% Car-MA was more biocompatible (i.e. higher cell adhesion and proliferation) compared to Gel-MA and Car-MA hydrogels separately. It can thus be concluded that the hydrogel blends developed outperformed the previously established gold standard Gel-MA rendering them promising candidates for (adipose) TE purposes. Thereafter, the potential of the photo-crosslinkable hydrogel materials to be used as biomaterial ink for extrusion-based 3D printing was investigated. This printing technique allows the fabrication of scaffolds that can be designed according to the patient’s needs. First, the influence of the pore size on the adipogenic differentiation and spatial distribution of MSCs was studied. Gel-MA scaffolds with pore sizes in the range of 200 – 600 μm were printed and seeded with MSCs. The results indicated that the MSCs differentiated robustly into the adipogenic lineage equally well in scaffolds of all pore sizes. However, the spatial distribution and the cellular infiltration varied with the pore sizes indicating that the scaffolds with pore sizes exceeding 500 μm support simultaneously differentiation and infiltration. Previously, the cell viability results showed that the hydrogel blend composed of Gel-MA and Car-MA outperformed the one-component hydrogels. However, the question remained whether these findings were also applicable in a 3D environment. Therefore, 3D scaffolds composed of the hydrogel blend were printed in order to evaluate to what extent an ideal 3D environment for the differentiation of ASCs into the adipogenic lineage can be created. In addition, the effect of the pore size on the adipogenic differentiation potential of stem cells was taken into account and thus the pore size of the hydrogel blend scaffolds developed exceeded 500 μm. The physico-chemical characteristics including swelling and mechanical properties, the in vitro biocompatibility in the presence of ASCs and the adipogenic differentiation ability of the hydrogel scaffolds were benchmarked to the characteristic properties of Gel-MA scaffolds. The results indicated that both the hydrogel blend and the Gel-MA scaffolds remained stable over time (21 days), were able to absorb large amounts of water and exhibited mechanical properties comparable to those of native breast tissue (2 kPa) indicating that the scaffolds are able to mimic the physico-chemical characteristics of the natural ECM of native adipose tissue. In addition, the results of the in vitro live/dead and proliferation assays showed a similar cell viability (> 90%) for the hydrogel blend and Gel-MA scaffolds. Although, the hydrogel blend scaffolds were able to support the differentiation of the ASCs into the adipogenic lineage, their differentiation potential was lower compared to the Gel-MA scaffolds. Consequently, a combination of a protein, Gel-MA and a polysaccharide, Car-MA, to create a superior mimic of the ECM did not contribute to an added value towards adipose TE. Therefore, the focus of this PhD work was shifted to the development of other gelatin derivatives which crosslink via step growth polymerization to potentially outperform the chain growth crosslinked Gel-MA. In this respect, norbornene-functionalized gelatin and thiolated gelatin with a DS of 53% and 72% respectively were developed which can be crosslinked via thiol-ene photo-click chemistry. Gel-SH was chosen over the conventional thiolated crosslinkers including DTT, PEG4SH and thiolated hyaluronic acid due to its cell-interactivity and the low risk of phase separation to occur which is a problem when high molecular weight crosslinkers are applied. Subsequently, Gel-NB combined with Gel-SH was applied as biomaterial ink to fabricate 3D constructs with a pore size exceeding 500 μm via extrusion-based 3D printing. In this study, the scaffolds developed were also benchmarked to Gel-MA scaffolds to gain quantitative insight in the scaffold’s performance. The results showed that the Gel-NB/SH scaffolds remained stable for at least 21 days, were able to mimic the aqueous environment of the ECM and exhibited comparable mechanical properties of native breast tissue (2 kPa). In addition, the scaffolds remained biodegradable after UV-induced crosslinking. Furthermore, the in vitro biocompatibility assays showed a similar cell viability (> 90%) for the Gel-NB/SH scaffolds in comparison to the Gel-MA scaffolds. However, the adipogenic differentiation potential of the developed Gel-NB/SH scaffolds proved to be better compared to the Gel-MA scaffolds. Therefore, it can be concluded that the developed Gel-NB/SH can be considered a promising candidate for adipose TE. However, it should be pointed out that the differentiation of mesenchymal stem cells into the adipogenic lineage is not the only aspect to take into account when engineering functional tissue. Indeed, angiogenesis/vascularization also plays a key role in the survival of tissue. Therefore, the influence of the pore size on MSC paracrine activity was examined and whether this was determined by local cell interactions in the scaffold environment. Gel-MA was utilized as model compound to produce 3D scaffolds via extrusion-based 3D printing. The results indicated that the pore size can guide the distribution of MSCs and regulate cell-cell and cell-matrix interactions at the gene-level. The angiogenic paracrine activity of the MSCs was enhanced in the medium pore size (220 μm) scaffolds in which the cells formed aggregates but also interacted with the polymer struts. The enhanced paracrine activity obtained for the medium pore size scaffolds also improved the endothelial cell migration and tube formation capacity which are key factors for angiogenesis. One of the material requirements for extrusion-based 3D printing is that the materials should have a certain viscosity (30 – 6.107 Pa.s) to be able to continuously extrude strands according to the predefined CAD model. However, the introduced photo-crosslinkable methacrylamide functionalities onto the backbone of RCPhC1-MA affected the physical crosslinking behaviour of the protein. Therefore, modified RCPhC1 is not an ideal candidate for extrusion-based 3D printing. However, this viscosity requirement is not applicable to two-photon polymerization rending the hydrogel suitable for processing via 2PP. RCPhC1 was modified with different photo-crosslinkable moieties including methacrylamide, norbornene and thiol functionalities resulting in a DS of 93%, 83% and 73% for RCPhC1-MA, RCPhC1-NB and RCPhC1-SH respectively. First, the hydrogel precursors were UV crosslinked into hydrogel films and physico-chemically characterized. Subsequently, their 2PP processing potential was evaluated to investigate the effect of chain growth versus step growth polymerization on the post-production swelling properties and thus the CAD-CAM mimicry. In addition, the ability of the RCPhC1 material to be applied as a bioink was also explored using ASCs. The results showed that a similar gel fraction (> 90%) was obtained for RCPhC1-MA and RCPhC1-NB/SH indicating that stable hydrogels are formed. Higher swelling and higher mechanical properties were obtained for RCPhC1-NB/SH hydrogels due to the formation of an orthogonal and homogeneous network. In addition, a clear difference in 2PP processing was observed between both RCPhC1 derivatives. Less swelling-related deformation and thus a superior CAD-CAM mimicry was obtained for RCPhC1-NB/SH. Furthermore, the swelling properties were also influenced by the biopolymer concentration, the writing speed and the irradiation dose applied. In addition, RCPhC1-NB/SH could be processed in the presence of ASCs-GFP which survived the encapsulation process and were able to proliferate inside the printed structures. It can thus be concluded that the RCPhC1-NB/SH hydrogels developed are processable via 2PP and can be used as bioink to produce injectable scaffolds. In a last part of this work, the design of injectable scaffolds was optimized from a cell perspective for adipose TE purposes. Not only material properties including composition and stiffness but also 3D geometrical features such as surface curvature regulate the migration and the differentiation of mesenchymal stem cells. Therefore, the influence of round versus hexagonally/pentagonally-shaped pores, the pore size, the scaffold diameter and the edge radius on the cell behaviour of ASCs were evaluated using OrmoComp®-based 2PP-printed scaffolds as a commercially available model compound. The results indicated that the cells were not in favour of the hexagonally/pentagonally shaped pores due to the sharp edges of the polygons. Therefore, new CAD models with round pores were designed and the ASCs seeded onto these scaffolds were able to produce fibronectin and collagen which are the required components for tissue formation. However, no clear trend in fibronectin nor collagen signal could be observed for the different pore sizes, scaffolds diameters or edge radii. In addition, the ASCs were able to differentiate into the adipogenic lineage. The results indicated a more pronounced lipid formation per volume in the 150 μm diameter scaffold with round edges independent of the pore size. However, these experiments should be repeated to further confirm these conclusions and to monitor statistically significant trends.
Date of Award7 Feb 2020
Original languageEnglish

Cite this

'