Date Published: April 15, 2019
Author(s): Daniel A. Foyt, Dheraj K. Taheem, Silvia A. Ferreira, Michael D.A. Norman, Jonna Petzold, Gavin Jell, Agamemnon E. Grigoriadis, Eileen Gentleman.
Tissue engineering strategies often aim to direct tissue formation by mimicking conditions progenitor cells experience within native tissues. For example, to create cartilage in vitro, researchers often aim to replicate the biochemical and mechanical milieu cells experience during cartilage formation in the developing limb bud. This includes stimulating progenitors with TGF-β1/3, culturing under hypoxic conditions, and regulating mechanosensory pathways using biomaterials that control substrate stiffness and/or cell shape. However, as progenitors differentiate down the chondrogenic lineage, the pathways that regulate their responses to mechanotransduction, hypoxia and TGF-β may not act independently, but rather also impact one another, influencing overall cell response. Here, to better understand hypoxia’s influence on mechanoregulatory-mediated chondrogenesis, we cultured human marrow stromal/mesenchymal stem cells (hMSC) on soft (0.167 kPa) or stiff (49.6 kPa) polyacrylamide hydrogels in chondrogenic medium containing TGF-β3. We then compared cell morphology, phosphorylated myosin light chain 2 staining, and chondrogenic gene expression under normoxic and hypoxic conditions, in the presence and absence of pharmacological inhibition of cytoskeletal tension. We show that on soft compared to stiff substrates, hypoxia prompts hMSC to adopt more spread morphologies, assemble in compact mesenchymal condensation-like colonies, and upregulate NCAM expression, and that inhibition of cytoskeletal tension negates hypoxia-mediated upregulation of molecular markers of chondrogenesis, including COL2A1 and SOX9. Taken together, our findings support a role for hypoxia in regulating hMSC morphology, cytoskeletal tension and chondrogenesis, and that hypoxia’s effects are modulated, at least in part, by mechanosensitive pathways. Our insights into how hypoxia impacts mechanoregulation of chondrogenesis in hMSC may improve strategies to develop tissue engineered cartilage.
Cartilage tissue engineering strategies often aim to drive progenitor cell differentiation by replicating the local environment of the native tissue, including by regulating oxygen concentration and mechanical stiffness. However, the pathways that regulate cellular responses to mechanotransduction and hypoxia may not act independently, but rather also impact one another. Here, we show that on soft, but not stiff surfaces, hypoxia impacts human MSC (hMSC) morphology and colony formation, and inhibition of cytoskeletal tension negates the hypoxia-mediated upregulation of molecular markers of chondrogenesis. These observations suggest that hypoxia’s effects during hMSC chondrogenesis are modulated, at least in part, by mechanosensitive pathways, and may impact strategies to develop scaffolds for cartilage tissue engineering, as hypoxia’s chondrogenic effects may be enhanced on soft materials.
Osteoarthritis (OA) is one of the leading causes of disability worldwide and constitutes a significant individual and socioeconomic burden . Indeed, the costs of OA in the USA, Canada, UK, France and Australia may account for between 1 and 2.5% of these countries’ gross domestic products . One goal in the field of cartilage tissue engineering (TE) is to repair cartilage lesions before they progress to OA. To coax progenitor cells to differentiate appropriately and produce cartilage, many TE strategies aim to create scaffolds that mimic characteristics of the native tissue that progenitor cells are exposed to as the tissue is formed , , . For cartilage, which has a very poor capacity for self-repair in the adult, much of what is known about the conditions that foster cartilage formation come from developmental studies, including the study of limb bud development and endochondral ossification in the axial skeleton. Cartilage TE strategies that mimic both the biochemical milieu as well as the mechanical environment that native cells experience during these developmental processes may yield engineered constructs that can repair damaged adult tissues , , , .
hMSC are known to respond to culture on soft and stiff surfaces by altering their cytoskeletal arrangements , . Here, we first aimed to confirm that under normoxic conditions in the presence of chondrogenic medium containing TGF-β3, substrate stiffness did indeed regulate hMSC spread area and circularity. To create culture substrates that cells would perceive as either akin to the soft matrix in the developing limb bud  or as significantly stiffer, we formed PA surfaces with varying concentrations of acrylamide and bis-acrylamide and measured their Young’s moduli using AFM-based microindentation (Supplementary Fig. 3). After coating with fibronectin, we then cultured hMSC for 24 h on the PA hydrogels and found that on stiff surfaces with a Young’s modulus of 49.6 kPa, they adopted spread morphologies and actin staining with fluorophore-conjugated phalloidin confirmed that they produced defined stress fibres (Fig. 1A). However, on soft surfaces with a Young’s modulus of 0.167 kPa, hMSC appeared round and stress fibres were not evident (Fig. 1B). Quantification showed that hMSC on stiff hydrogels had significantly larger cell areas (Fig. 1C) and that cells exhibited significantly lower levels of circularity compared to hMSC cultured on soft hydrogels (Fig. 1D).Fig. 1Substrate stiffness impacts cell morphology and YAP nuclear localisation in hMSC cultured in chondrogenic medium. (A + B) Actin immunodetection by phalloidin with DAPI counterstain after 24 h culture on stiff (A) and soft (B) substrates. Representative images of 4 independent repeats shown. (C + D) Quantification of cell area (C) and circularity (D) based on phalloidin staining. Values plotted represent the area/circularity of a single cell, with values from 4 independent repeats plotted and the mean values represented by the red line. A perfect circle has a circularity of 1. (E) Quantification of nuclear YAP on hydrogels of each stiffness. Each value plotted represents the percentage of a single DAPI-marked nucleus that is occupied by YAP. Values are from 3 independent repeats with the red horizontal lines representing the mean. (F–I) YAP immunodetection with DAPI counterstain after 24 h on stiff (F + G) and soft (H + I) substrates. Representative images of 3 independent repeats shown. Statistical analysis: *p < 0.05. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Here, we confirmed that hMSC respond to culture on soft surfaces by adopting round morphologies and localising YAP diffusely in their cytoplasm, and that these effects are not grossly impacted in chondrogenic medium containing TGF-β3. We then showed that pharmacological inhibition of ROCK, which ablates positive staining for pMLC2, prompts hMSC on soft surfaces to adopt more spread morphologies. This is contrast to hMSC behaviour on stiff surfaces where treatment with Y-27632 only subtly affected cell morphology. These data suggest that on soft substrates, hMSC morphology is governed, at least in part, by ROCK, and confirm that chondrogenic induction with TGF-β3 does not impact hMSC’s expected behaviour under these conditions . Our findings show that on soft substrates hypoxia prompts cells to adopt more spread morphologies, assemble in compact mesenchymal condensation-like colonies, and upregulate NCAM expression and that these processes are, at least in part, dependent on cytoskeletal tension. Understanding the factors that regulate chondrogenic differentiation of hMSC may inform on strategies to repair acute cartilage defects using TE approaches. Indeed, our findings suggest that soft TE scaffolds that mimic the soft conditions that progenitor cells experience during native tissue formation may be more conducive for driving hypoxia-mediated chondrogenesis than stiffer scaffolds. Source: http://doi.org/10.1016/j.actbio.2019.03.002