Type I collagen deposition via osteoinduction ameliorates YAP/TAZ activity in 3D floating culture clumps of mesenchymal stem cell/extracellular matrix complexes

Abstract

Background: Three-dimensional (3D) floating culture clumps of mesenchymal stem cell (MSC)/extracellular matrix (ECM) complexes (C-MSCs) consist of cells and self-produced ECM. Previous studies have demonstrated that C-MSCs can be transplanted into bony lesions without an artificial scaffold to induce bone regeneration. Moreover, osteoinductive medium (OIM)-treated C-MSCs (OIM-C-MSCs) have shown rapid and increased new bone formation in vivo. To apply OIM-C-MSCs for novel bone regenerative cell therapy, their cellular properties at the molecular level must be elucidated. The transcriptional co-activators yes-associated protein/transcriptional co-activator with PDZ-binding motif (YAP/TAZ) have been recognized as key players in the mechanotransduction cascade, controlling cell lineage commitment in MSCs. It is plausible that 3D C-MSCs/OIM-C-MSCs cultured in floating conditions could provide distinct microenvironments compared to conventional 2D culture systems and thereby induce unique mechanotransduction cascades. Therefore, this study investigated the YAP/TAZ activity in 3D-cultured C-MSCs/OIM-C-MSCs in floating conditions.

Methods: Human bone marrow-derived MSCs were cultured in growth medium supplemented with ascorbic acid. To obtain C-MSCs, confluent cells that had formed on the cellular sheet were scratched using a micropipette tip and were then torn off. The sheet was rolled to make round clumps of cells. Then, YAP/TAZ activity, filamentous actin (F-actin) integrity, collagen type I (COL1) production, and the differentiation potency in 3D floating culture C-MSCs/OIM-C-MSCs were analyzed.

Results: C-MSCs cultured in floating conditions lost their actin cytoskeleton to downregulate YAP/TAZ activity, which directed cells to undergo adipogenesis/chondrogenesis. OIM treatment induced abundant COL1 deposition, which facilitated Intβ1-dependent actin fiber formation and YAP/TAZ activity to elevate the expression levels of osteogenic master transcriptional factor runt-related transcription factor 2 (RUNX2) mRNA in C-MSCs. Importantly, elevation of YAP/TAZ activity via OIM was associated with COL1 deposition and F-actin integrity, suggesting a positive feedback loop in OIM-C-MSCs.

Conclusion: These findings suggest that OIM-C-MSCs, which form a unique microenvironment that maintains high YAP/TAZ activity, can serve as better candidates for bone regenerative cell therapy than C-MSCs.

Keywords: 3D culture; C-MSCs; F-actin; Mechanotransduction; YAP/TAZ.

Fig. 1

YAP and TAZ are downregulated in 3D-cultured C-MSCs. a Schematic image of C-MSC culture. Bar = 5 mm. b–g C-MSCs were generated and maintained in growth medium (GM) for the indicated culture periods. b Upper panels indicate macroscopic images of C-MSCs. Lower panels show confocal immunofluorescence images of COL1 and nuclei in C-MSCs. Bar = 100 μm. c Confocal immunofluorescence images of YAP/TAZ (green), F-actin (red), and nuclei (blue) in C-MSCs. Bar = 20 μm. White boxes show enlarged images. d The graph summarizes the distribution of YAP/TAZ patterns. The patterns were classified as mainly nuclear (N > C), diffuse (N = C) and mainly cytoplasmic or undetectable (N < C or undetectable). e Immunoblotting for YAP/TAZ in C-MSCs. f, h Real-time PCR for YAP and TAZ (f) or YAP/TAZ target genes (h) in C-MSCs. Data were normalized to the values on day 0. Values represent means ± S.D. of three cultures (**p < 0.01). g Phos-tag-SDS-PAGE was used to show the phosphorylation levels of YAP/TAZ. Arrowheads indicate the position of unphosphorylated YAP/TAZ. Levels of total YAP/TAZ were analyzed by immunoblotting. All graphs and images are representative of three independent experiments

Fig. 2

Three-dimensional floating culture C-MSCs can be directed towards adipogenesis/chondrogenesis but not osteogenesis due to reduced YAP/TAZ activity. a–d C-MSCs were cultured in a OIM, b AIM, c CIM, or d dual medium for 5 days. Differentiation marker gene expression levels were analyzed by real-time PCR. Data were normalized to the values of C-MSCs maintained in GM. e–g C-MSCs transfected with a constitutively active TAZ mutant (TAZS89A) or control plasmid were maintained in e OIM, f AIM, or g CIM for 5 days. Differentiation marker gene expression levels were analyzed by real-time PCR. Data were normalized to the values of C-MSCs transfected with control vector (cont). Values represent means ± S.D. of three cultures (**p < 0.01). All graphs are representative of three independent experiments

Fig. 3

Osteoinductive medium facilitates F-actin formation and YAP/TAZ activity in C-MSCs. a, b C-MSCs were cultured in OIM for the indicated culture period. a Confocal immunofluorescence images of COL1 and nuclei. Bar = 100 μm. b Confocal immunofluorescence images of YAP/TAZ (green), F-actin (red), and nuclei (blue) in C-MSCs. Bar = 20 μm. White boxes show enlarged images. c The graph summarizes the distribution of YAP/TAZ localization patterns. The patterns were classified as mainly nuclear (N > C), diffuse (N = C), or mainly cytoplasmic or undetectable (N < C or undetectable). d–f C-MSCs: C-MSCs cultured in GM for 5 days. OIM-C-MSCs: C-MSCs cultured in OIM for 5 days. d Immunoblotting for YAP/TAZ. A rabbit anti-YAP/TAZ (D24E4) mAb (Cell Signaling) was used to detect both the YAP and TAZ proteins (middle panel). To show YAP expression more clearly, we also used a rabbit anti-YAP (D8H1X) mAb (upper panel). e Real-time PCR of YAP/TAZ target genes. Data were normalized to the values of C-MSCs. Values represent means ± S.D. of three cultures (**p < 0.01). f C-MSCs transfected with negative control or YAP and TAZ siRNAs were cultured in OIM for 5 days. RUNX2 expression levels were analyzed by real-time PCR. Data were normalized to the values of OIM-C-MSCs transfected with negative control siRNA. Values represent means ± S.D. of three cultures (**p < 0.01). All graphs and images are representative of three independent experiments

Fig. 4

The effects of OIM components on COL1 production, F-actin integrity, and YAP/TAZ activity in C-MSCs. a–c C-MSCs were cultured with GM (none) in the presence of β-glycerophosphate (Gly), ascorbic acid (AA), and/or dexamethasone (Dex) (Gly+AA, Gly+Dex, AA+Dex, and Gly+AA+Dex (equivalent to OIM)) for 5 days. a Confocal immunofluorescence images of COL1 (green) and nuclei (blue) in C-MSCs. Bar = 100 μm. b Confocal immunofluorescence images of YAP/TAZ (green), F-actin (red), and nuclei (blue) in C-MSCs. Bar = 20 μm. c The graph summarizes the distribution of YAP/TAZ patterns. The patterns were classified as mainly nuclear (N > C), diffuse (N = C), and mainly cytoplasmic or undetectable (N < C or undetectable)

Fig. 5

OIM-induced YAP/TAZ activity in C-MSCs is due to actin cytoskeletal tension caused by Intβ1-ROCK signaling. a–j C-MSCs: C-MSCs cultured with GM for 5 days. OIM-C-MSCs: C-MSCs cultured with OIM for 5 days. a–e C-MSCs transfected with negative control or Intβ1 siRNAs were cultured with GM or OIM for 5 days. a Confocal immunofluorescence images of YAP/TAZ (green), F-actin (red), and nuclei (blue) in OIM-C-MSCs. Bar = 20 μm. White boxes show enlarged images. b The graph summarizes the distribution of YAP/TAZ localization patterns. The patterns were classified as mainly nuclear (N > C), diffuse (N = C), and mainly cytoplasmic or undetectable (N < C or undetectable). c Immunoblotting for YAP/TAZ in OIM-C-MSCs. A rabbit anti-YAP/TAZ (D24E4) mAb (Cell Signaling) was used to detect both the YAP and TAZ proteins (middle panel). To show YAP expression more clearly, we also used a rabbit anti-YAP (D8H1X) mAb (upper panel). d Real-time PCR analysis of YAP/TAZ target genes. Data were normalized to the values of control siRNA-transfected C-MSCs. Values represent means ± S.D. of three cultures (**p < 0.01). e RUNX2 expression levels in OIM-C-MSCs were analyzed by real-time PCR. Data were normalized to the values of OIM-C-MSCs transfected with negative control siRNA. Values represent means ± S.D. of three cultures (**p < 0.01). All graphs and images are representative of three independent experiments. f–j C-MSCs or OIM-C-MSCs were cultured with or without a ROCK inhibitor (Y27632, 50 μM), the non-muscle myosin inhibitor blebbistatin (Blebbist., 50 μM), or an appropriate concentration of DMSO for 5 days. f Confocal immunofluorescence images of YAP/TAZ (green), F-actin (red), and nuclei (blue) in OIM-C-MSCs. Bar = 20 μm. White boxes show enlarged images. g The graph summarizes the distribution of YAP/TAZ localization patterns. The patterns were classified as mainly nuclear (N > C), diffuse (N = C), and mainly cytoplasmic or undetectable (N < C or undetectable). h Immunoblotting for YAP/TAZ in OIM-C-MSCs. A rabbit anti-YAP/TAZ (D24E4) mAb (Cell Signaling) was used to detect both the YAP and TAZ proteins (middle panel). To show YAP expression more clearly, we also used a rabbit anti-YAP (D8H1X) mAb (upper panel). i Real-time PCR analysis of YAP/TAZ target genes. Data were normalized to the values of C-MSCs treated with DMSO. Values represent means ± S.D. of three cultures (**p < 0.01). j RUNX2 expression levels in OIM-C-MSCs were analyzed by real-time PCR. Data were normalized to the values of OIM-C-MSCs treated with DMSO. Values represent means ± S.D. of three cultures (**p < 0.01). All graphs and images are representative of three independent experiments

Fig. 6

YAP/TAZ activity is associated with COL1 synthesis and F-actin formation, generating a positive feedback loop in OIM-C-MSCs. (a–m) C-MSCs: C-MSCs cultured with GM for 5 days. OIM-C-MSCs: OIM-C-MSCs cultured with OIM for 5 days. a–e C-MSCs transfected with negative control, YAP and TAZ, or Intβ1 siRNAs were cultured with OIM for 5 days. a, d COL1A1 mRNA expression levels were analyzed by real-time PCR. Data were normalized to the values of OIM-C-MSCs transfected with negative control siRNA. Values represent means ± S.D. of three cultures (**p < 0.01). b, e Confocal immunofluorescence images show COL1 (green) and nuclei (blue) in OIM-C-MSCs. Bar = 100 μm. c Confocal immunofluorescence images of YAP/TAZ (green), F-actin (red), and nuclei (blue) in OIM-C-MSCs. Bar = 20 μm. f, g OIM-C-MSCs were cultured with or without a ROCK inhibitor (Y27632, 50 μM), the non-muscle myosin inhibitor blebbistatin (Blebbist., 50 μM), or an appropriate concentration of DMSO for 5 days. f COL1A1 mRNA expression levels were analyzed by real-time PCR. Data were normalized to the values of C-MSCs treated with DMSO. Values represent means ± S.D. of three cultures (**p < 0.01). g Confocal immunofluorescence images show COL1 (green) and nuclei (blue) in C-MSCs. Bar = 100 μm. h–l C-MSCs transfected with a constitutively active TAZ mutant (TAZS89A) or control plasmid were maintained in GM for 5 days. h COL1A1 mRNA expression levels were analyzed by real-time PCR. Data were normalized to the values of C-MSCs transfected with control vector (cont). Values represent means ± S.D. of three cultures (**p < 0.01). i Confocal immunofluorescence images show COL1 (green) and nuclei (blue) in C-MSCs. Bar = 100 μm. j Confocal immunofluorescence images of YAP/TAZ/TAZS89A (green), F-actin (red), and nuclei (blue) in C-MSCs. Bar = 20 μm. k Immunoblotting for YAP/TAZ in OIM-C-MSCs. A rabbit anti-YAP/TAZ (D24E4) mAb (Cell Signaling) was used to detect YAP, TAZ, and TAZS89A mutant proteins (middle panel). To show YAP expression more clearly, a rabbit anti-YAP (D8H1X) mAb was also used (upper panel). l Real-time PCR of YAP, TAZ, or YAP/TAZ target genes. Data were normalized to the values of C-MSCs transfected with a control vector (cont). Values represent means ± S.D. of three cultures (**p < 0.01)

Fig. 7

Schematic summary of mechanotransduction cascade in C-MSCs and OIM-C-MSCs