Fig 1.
ECM stiffness regulates the cellular phenotype of human mesenchymal stem cells (hMSCs) and the localization and transcriptional activity of TAZ.
(A) In indicated hydrogels, cell adhesion and morphology were visualized by light microscopy. Bright field images were taken 24hr after seeding. (B) ECM stiffness controls focal adhesion complex formation and TAZ localization. hMSCs were immunostained with an anti-vinculin antibody to detect focal adhesions (green fluorescence signal) 24 h after seeding. TAZ localization was visualized as a red fluorescence signal. DAPI was used to stain the cell nucleus. (C) The expression of TAZ target genes, including CTGF and CYR61, were analyzed by qRT-PCR using the cells in panel (A). Target gene expression was normalized to the GAPDH expression. Data is shown as fold induction. Asterisks indicate statistical significance (***p < 0.005, t-test). (D) The luciferase reporter gene CTGF-luc was introduced into hMSCs. After 16 h, the transfected cells were plated on 1.37 or 4.47 kPa hydrogels. After 24 h, luciferase reporter gene activity was analyzed. The pGL3-basic luciferase reporter gene, which has no promoter for transcription, was used as a negative control. A Renilla luciferase-expressing vector was used as a transfection control. Luciferase activity was normalized to Renilla luciferase activity and is expressed as relative fold induction. (***p < 0.005, t-test). (E) hMSCs were seeded on a 1.37 or 4.47 kPa hydrogel, and twenty four hours after seeding, total RNAs were isolated and qRT-PCR analysis was assessed to see the expression of TAZ gene. Gene expression was normalized to GAPDH.
Fig 2.
ECM stiffness stimulates osteoblast differentiation and represses adipocyte differentiation.
(A) A stiff hydrogel stimulates osteogenic differentiation. hMSCs were seeded on a 1.37 or 4.47 kPa hydrogel, and osteogenic differentiation was induced 24 h after seeding. Cells were stained for alkaline phosphatase at 9 days after differentiation and Von Kossa staining was assessed to see mineralization at 21 days after differentiation. (B) qRT-PCR analysis of the osteoblast marker genes DLX5, MSX2, osteocalcin, and RUNX2 of the cells used in panel (A). Expression of the marker genes was normalized to GAPDH expression. Data are presented as fold induction. (*p < 0.05, ***p < 0.005, t-test) (C) The luciferase reporter gene construct 6OSE2-luc was introduced into hMSCs along with a Runx2-expressing vector, which provides similar amounts of Runx2 protein in assay condition. After 16 h, the transfected cells were plated on a 1.37 or 4.47 kPa hydrogel. After 24 h, luciferase reporter gene activity was analyzed. The pGL3-basic luciferase reporter gene construct, which has no promoter for transcription, was used as a negative control. A Renilla luciferase-expressing vector was used as a transfection control. Luciferase activity was normalized to Renilla luciferase activity and is expressed as relative fold induction. (***p < 0.005, t-test) (D) hMSCs were seeded on a 1.37 or 4.47 kPa hydrogel, and adipogenic differentiation was induced 24 h after seeding. Oil Red O staining was assessed to see lipid droplet at 9 days after differentiation (E) The expression of the adipogenic marker genes adiponectin, aP2, and C/EBPα were analyzed by qRT-PCR in adipocyte-differentiated hMSCs. Adipogenic differentiation was assessed according to the protocol described in the Materials and Methods. Target gene expression was normalized to GAPDH. Data are expressed as relative fold induction. (*p < 0.05, ***p < 0.005, t-test)
Fig 3.
A stiff ECM activates the ERK and JNK signaling pathway.
(A) hMSCs were plated on a 1.37 or 4.47 kPa hydrogel, and after 12 h, cell lysates were prepared and analyzed by immunoblotting. To assess the activity of ERK and JNK, phosphorylated ERK (p-ERK) and phosphorylated JNK (p-JNK) antibodies were used, respectively. As a control, total ERK and JNK protein was analyzed. The levels of phosphorylated ERK and JNK were increased in hMSCs cultured on the stiff matrix. GAPDH was used as a loading control. (B) Immunocytochemistry of p-ERK and p-JNK in hMSCs on a 1.37 or 4.47 kPa hydrogel. Cells in panel (A) were fixed and subjected to immunocytochemical staining with a p-ERK or p-JNK antibody. The signals for p-ERK or p-JNK were green fluorescence. DAPI was used for nuclear staining. (C) Fluorescence signals in panel (B) was quantified by image J software and corrected total cell fluorescence was calculated by fluorescence signal with elimination of background signal. AU is arbitrary unit.
Fig 4.
Stiffness-regulated ERK/JNK activity is crucial for TAZ target gene expression and osteogenesis at the transcriptional level.
(A) hMSCs cultured on the 4.47 kPa hydrogel were treated with a MEK inhibitor (U0126, 10 μM) or a JNK inhibitor (SP600125, 10 μM). After 12 h, total RNA was prepared, and CTGF and CYR61 expression was assessed by qRT-PCR. The results showed that CTGF and CYR61 were downregulated following inhibition of ERK or JNK in cells on the stiff hydrogel. (B) The CTGF-luc reporter gene construct or the control pGL3-basic vector was transfected into hMSCs, and after 16 h, the transfected cells were plated on 4.47 kPa hydrogels. After 24 h, luciferase reporter gene activity was analyzed. To inhibit ERK or JNK, cell were pretreated with 10 μM U0126 or 10 μM SP600125, respectively, 12 h before reporter gene analysis. A Renilla luciferase-expressing vector was used as a transfection control. Luciferase activity was normalized to Renilla luciferase activity. (C) hMSCs were differentiated into osteoblasts for 6 days in the presence of 10 μM U0126 or 10 μM SP600125. DMSO was used as the vehicle control. The expression of osteoblastic marker genes, including DLX5, MSX2, osteocalcin, and RUNX2, were analyzed by qRT-PCR. Gene expression was normalized to GAPDH. The results show that the expression of the osteogenic marker genes in cells on a 4.47 kPa hydrogel was significantly suppressed by ERK or JNK inhibition. (D) hMSCs were transfected with 6OSE2-luc or pGL3-basic (control) along with a Renilla luciferase-expressing construct. Then, the cells were plated on a 4.47 kPa hydrogel, and after 24 h, luciferase reporter gene activity was analyzed. To inhibit ERK or JNK, cells were pretreated with 10 μM U0126 or 10 μM SP600125, respectively, 12 h before the reporter gene assay. Reporter gene luciferase activity was normalized to Renilla luciferase activity. (***p < 0.005, t-test)
Fig 5.
Inhibition of the ERK or JNK signaling pathway induces TAZ cytoplasmic localization on stiff hydrogels.
(A) hMSCs on a 4.47 kPa hydrogel were treated with 10 μM U0126 or 10 μM SP600125. After 12 h of treatment, cells were subjected to immunostaining with an anti-TAZ antibody. The red fluorescence signal shows the location of TAZ, and DAPI was used to stain the nuclei. The results show that even in cells on a stiff matrix, TAZ is localized evenly to the cytoplasm and nucleus following ERK or JNK inhibition. (B) Approximately 100 cells in panel (A) were counted, and TAZ localization was analyzed in these cells. The counting procedure was done using the Image J program. The number of cells that showed an even cytoplasmic-nuclear or cytoplasm-dominant TAZ localization was higher in the presence of an ERK or JNK inhibitor than in the absence of either inhibitor. (C) Cell lysates in panel (A) were prepared, and the activity of the Hippo signaling pathway components LATS and MST kinase was analyzed by immunoblotting. The phosphorylation status of LATS and MST kinase was analyzed with p-LATS1 and p-MST1/2 antibodies, respectively. Total LATS1 and MST2 levels were detected as a loading control.
Fig 6.
Inhibition of ROCK/F-actin represses the transcriptional activity of TAZ in hMSCs.
(A) hMSCs cultured on a 4.47kPa hydrogel were treated with a ROCK inhibitor (Y27632, 50 μM) or an F-actin inhibitor (latrunculin A, 0.5 μM). After 12 h of treatment, total RNA was prepared, and CTGF and CYR61 expression was assessed by qRT-PCR. (B) The CTGF-luc reporter gene construct or the pGL3-basic control vector was transfected into hMSCs, and after 16 h, the transfected cells were plated on 4.47 kPa hydrogels. After 24 h, luciferase reporter gene activity was analyzed. To inhibit ROCK or F-actin, cells were pretreated with 50 μM Y27632 or 0.5 μM latrunculin A 12 h before reporter gene analysis. A Renilla luciferase-expressing vector was used as a transfection control. Luciferase activity was normalized to Renilla luciferase activity. (C) hMSCs on 4.47 kPa hydrogels were differentiated into osteoblasts for 6 days in the presence of 50 μM Y27632 or 0.5 μM latrunculin A. DMSO was used as the vehicle control. The expression of osteoblastic marker genes, including DLX5, MSX2, osteocalcin, and RUNX2, were analyzed by qRT-PCR. Target gene expression was normalized to GAPDH. (D) hMSCs were transfected with 6OSE2-luc or pGL3-basic (control) along with a Renilla luciferase-expressing construct. Then, the cells were plated on a 4.47 kPa hydrogel. After 24 h, luciferase reporter gene activity was analyzed. To inhibit ROCK or F-actin, the cells were pretreated with 50 μM Y27632 or 0.5 μM latrunculin A 12 h before the reporter gene assay. Luciferase activity was normalized to Renilla luciferase activity. (***p < 0.005, t-test). (E) Total RNAs of hMSCs in panel (A) were prepared and qRT-PCR was assessed to analyze the transcription of TAZ. Gene expression was normalized to GAPDH.
Fig 7.
Upon exposure to a stiff ECM environment, the Rho signaling pathway is activated, and ERK or JNK signaling is induced. Activated ERK or JNK promotes the nuclear localization of TAZ and activates the expression of TAZ target genes, including RUNX2-mediated osteoblastic marker genes.
Table 1.
Primers for qRT-PCR.