Introduction

Biofabrication offers a range of techniques and methods to build constructs with desired shapes and internal gradients of composition and cell types. Although achieving desired shapes and composition is fundamental, tissue properties also depend on the internal architecture at the microscopic level. For example, cartilage mechanical and biological properties critically depend on the orientation of collagen (col) fibrils and their zonal organization. Although biofabrication methods to control microscopic architecture are relatively mature, they often require sophisticated techniques and are low in throughput; additionally, their integration with the most accessible biofabrication methods is complex. Here, we present two techniques to control the spatial arrangement of collagen fibrils and of porosity within composite constructs based on hyaluronic acid (HA) and col. Col fibrils orientation was controlled via extrusion-based 3D printing after embedding within a continuous viscoelastic matrix based on crosslinked HA. Cell-free and cell laden constructs were prepared and characterized to determine the influence of this controlled microscopic anisotropy on cell behavior. Chondrogenic properties of the composites were assessed for a range of HA/col compositions by evaluating human bone marrow derived mesenchymal stromal cell (hMSC) differentiation to the chondrogenic lineage.

Content

Materials and Methods

The tyramine derivative of HA (THA) was prepared by functionalizing the carboxy groups of HA via amide bond formation. THA was combined with soluble acidic col at the following THA:col concentrations: 12.5:2.5; 12.5:1.7; 16.7:2.5; 12.5:1.7 mg/ml. The mixture was treated with hydrogen peroxide 18 to 22 ppm in presence of horseradish peroxidase [1] for enzymatic crosslinking of THA, and at the same time pH was shifted to neutrality at 37°C for inducing col fibrillogenesis. Shear thinning and viscoelastic properties of THA, col and composites were characterized via rheometry (AntonPaar MCR302) measuring the viscosity as a function of the shear rate, with an amplitude sweep and with an elastic recovery test performed alternating intervals of high and low strain to assess the kinetic recovery after deformation. The THA-col composite was printed (3D Discovery, RegenHU) controlling col fiber alignment with the shear forces acting during extrusion. Col orientation was visualized via Second Harmonic Generation (SHG) and immunostaining/confocal microscopy. Images were processed with image J (NIH) to quantify fiber orientation. hMSCs (passage 3) were preformed into aggregates by seeding into microwell plates (Aggrewell 400, Stemcell Technology) according to the manufacturer’s protocol using chondrogenic media (DMEM HG, 50 µg/ml ascorbic acid 2-phosphate, 100 nM dexamethasone, 1% v/v ITS premix, 1% v/v non-essential amino acids (MEM NEAA, Gibco), 10 ng/ml transforming growth factor beta (TGF-β1, Fitzgerald), 1% v/v Pen-Strep and cultured for 3 days. The obtained aggregates contained approximately 250 hMSCs and were embedded within the bioink to evaluate cell migration and orientation, assessed by F-actin and DAPI staining (Fig 1 D). In vitro chondrogenic behavior of hMSC embedded in casted THA-col (chondrogenic media containing 10 ng/ml TGF-β1) was evaluated via histology, gene expression (COL1A1, COL2A1, COL10A1, RunX2, SOX9, Aggrecan ACAN) and proteoglycan quantification. Chondrogenic differentiation was compared to hMSC pellets as a positive control.

Results

The gel nature of collagen and crosslinked HA prevents efficacious mixing. To obtain a homogeneous dispersion of col fibrils within the crosslinked HA gel, a method was developed starting from soluble acidic col and an non-crosslinked THA [2]. The liquid nature of these precursors allowed thorough mixing, followed by simultaneous col fibril formation and THA crosslinking avoiding mutual interference to achieve uniform distribution of col fibers within the HA-based viscoelastic matrix as seen in Fig 1 B.

Shear-induced fiber alignment along the printing direction was shown via immunofluorescence, SHG (Figure 1 A, B) and by immunohistochemistry (not shown). As expected, fiber alignment increased for decreasing nozzle diameter (Fig 1 C, D). The shear thinning and elastic recovery of THA [1] were preserved in the THA-col composite at the investigated mixing ratios. THA-col showed 2-fold increase in storage modulus compared to THA.

The presence of col fibers had an evident impact on cell attachment and migration. Actin filament staining showed cytoskeleton alignment along the fiber orientation, which in turn was determined by the printing direction (Fig 1C). Production of proteoglycan rich extracellular matrix during chondrogenesis was observed after 3 weeks of hMSC in vitro culture embedded in THA-col (figure 2B). At 3 weeks, chondrogenic associated genes COL2A1 (>100-fold), ACAN (>100-fold) and SOX9 were upregulated in all samples. Compared to hMSC pellets, hypertrophic markers (COL1A1, COL10A1, RunX2) were less upregulated in the THA-col hydrogels with higher ratio of SOX9/RunX2 in both THA-col samples (~10-fold) compared to no change in the pellet group (~1-fold).

Discussion

In this work, we introduce a method to 3D bioprint col with controlled orientation embedded within a viscoelastic HA matrix. Introducing control over the microscopic architecture of matrix components is key to recapitulate complex tissue structure and morphology. One potential limitation of the approach here presented is the use of the shear forces for alignment, limiting the space of design possibilities. Also, it remains to be determined how this controlled orientation influences matrix deposition. Overall, the possibility of printing matrix components with control over microscopic alignment brings biofabrication one step closer to capturing the complexity in animal tissues.

References

[1] Petta D C et al. Biofabrication. 2018 Sep 25;10(4):044104; PMID: 30188324

[2] Schwab et al. Mater Today Bio. 2020 Jun 1;7:100058; PMID: 3261318

Acknowledgments

This work is part of the osteochondral defect collaborative research program supported by the AO foundation. The Graubünden Innovationsstiftung is acknowledged for its financial support.