Rho kinase-dependent activation of SOX9 in chondrocytes

Abstract

Objective: The transcription factor SOX9 directly regulates the expression of the major proteoglycans and collagens comprising the cartilage extracellular matrix. The DNA binding activity and cellular localization of SOX9 is controlled through posttranslational modifications, including phosphorylation. The activity of Rho kinase (ROCK) has profound effects on the actin cytoskeleton, and these effects are instrumental in determining the phenotype and differentiation of chondrocytes. However, the mechanisms linking ROCK to altered chondrocyte gene expression remain unknown. The purpose of the present study was to test for a direct interaction between ROCK and SOX9.

Methods: Human SW1353 chondrosarcoma cells were transfected with constructs coding for RhoA, ROCK, Lim kinase, and SOX9. The interaction between ROCK and SOX9 was tested on purified proteins, and was verified within a cellular context using induced overexpression and activation of the Rho pathway. The effects of SOX9 transcriptional activation were quantified with a luciferase reporter plasmid containing SOX9 binding sites from the COL2A1 enhancer element.

Results: SOX9 was found to contain a consensus phosphorylation site for ROCK. In vitro, ROCK directly phosphorylated SOX9 at Ser(181), and the overexpression of ROCK or the activation of the RhoA pathway in SW1353 chondrosarcoma cells increased SOX9(Ser181) phosphorylation. ROCK caused a dose-dependent increase in the transcription of a SOX9-luciferase reporter construct, and increased phosphorylation and nuclear accumulation of SOX9 protein in response to transforming growth factor beta treatment and mechanical compression.

Conclusion: These results demonstrate a new interaction that directly links ROCK to increased cartilage matrix production via activation of SOX9 in response to mechanical and growth factor stimulation.

Figure 1

A, Sox9 contains a ROCK consensus sequence. Amino-acid alignments of Sox9 with several known ROCK substrates. The consensus serine and threonine phosphorylation residues are highlighted in grey, and the commonly required basic residues at positions -2 and -3 are boldface. B, ROCK directly interacts with and phosphorylates Sox9^Ser181 in vitro. FLAG-Sox9 expressed in SW1353 chondrosarcoma cells and purified by immunoprecipitation with anti-FLAG serves as a substrate for phosphorylation by recombinant human ROCK in vitro. Phosphoryla-tion was measured by western blotting with an anti-phospho-Sox9^Ser181 antibody, and the blot was stripped and re-probed with anti-total Sox9 to show even protein loading.

Figure 2

A, ROCK phosphorylates Sox9 in intact cells. Inducible ROCK:ER and ROCK(KD):ER were cotransfected with Sox9 into SW1353 chondrosarcoma cells, and ROCK was activated for 2 hours by the addition of Tamoxifen. The amount of phosphorylated Sox9 present in the cell lysate was assayed by western blotting with an anti-phospho-Sox9^Ser181 antibody. Total protein in each lane was constant. Results shown are typical of 3 independent experiments. The inset graphically shows image densitometry, error bars represent standard deviations. B, Sox9 phosphorylation is downstream of RhoA activation and independent of Lim kinase. RhoA was co-transfected with Sox9 and the amount of phosphorylated Sox9^Ser181 was assayed by western blotting of total cell lysate 24 hours after transfection. Sox9^Ser181 phosphorylation was significantly increased only when RhoA was co-transfected (p<0.05). As negative controls, Sox9 plasmid was co-transfected with empty pCDNA3.1 vector (EV). Co-transfection with LIM kinase plasmid did not affect Sox9 phosphorylation (p=.618). Results shown are typical of 4 independent experiments. The inset graphically shows image densitometry, error bars represent stan-dard deviations.

Figure 3. ROCK causes a dose-dependent increase in Sox9 transcriptional activity

Co-transfection of various amounts of ROCK:ER or a kinase-dead mutant ROCK(KD):ER with a col2a1-Luciferase reporter containing Sox9 binding sites into SW1353 chondrosarcoma cells in monolayer culture. After 48 hours cells were assayed for luciferase activity. The total amount of plasmid was held constant by the addition of pCDNA3.1 empty vector, and transfection efficiency was controlled for by assaying renilla activity in each sample. The amount of ROCK:ER or ROCK(KD):ER plasmid transfected is indicated in the X-axis legend. Error bars represent standard deviation of 3 replicate transfections within a typical experiment, * indicates p<0.05 compared to control.

Figure 4

A, ROCK-dependent Sox9 phosphorylation in TGF-ß-treated SW1353 chondrosarcoma cells. The levels of endogenous Sox9 phosphorylation were assayed by immunoblot in SW1353 cells treated 20 hours with 10ng/ml TGF-ß₁. 5μM hydroxyfasudil was added 2 hours prior to and during the TGF-β₁ treatment to inhibit ROCK activity. The blot was stripped and re-probed with anti-GAPDH to demonstrate even protein loading. Results shown are typical of 4 independent experiments. B, ROCK-dependent increase in nuclear Sox9 in TGF-ß₁ treated chondrocytes. Passaged human chondrocytes were grown in alginate suspension culture and treated with TGF-ß₁, ROCK inhibitor, or a combination. Cells were fixed and endogenous Sox9 was localized by immunofluorescence microscopy. The amount of nuclear Sox9 was quantified by colocalization with the ToPro3 nuclear dye, shown in arbitrary units on the Y-axis. Error bars indicate the standard deviation of Sox9 nuclear intensity, n=8 cells for each group, * indicates p<0.01 compared to control. C, Images of an untreated control chondrocyte (left) and a TGF-ß-treated chondrocyte (right) demonstrating increased nuclear localization of Sox9. Nuclei were detected with ToPro3 dye, and Sox9 was detected by anti-Sox9 followed by Alexa488 secondary antibody. Images were taken with a Zeiss LSM510 inverted confocal microscope and the entire selection represents an area of 26um square.

Figure 5

A, ROCK-dependent increase in nuclear Sox9 after 2 hours of dynamic compression. Alginate-embedded normal human articular chondrocytes were subjected to 2 hours of continuous cyclic dynamic compression from 5–15% strain at 0.5Hz. Cells were fixed, released from alginate, and endogenous Sox9 was quantified by immunofluorescence microscopy. The amount of nuclear Sox9 was quantified by colocalization with the ToPro3 nuclear dye, shown in arbitrary units on the Y-axis. Error bars indicate the 95% confidence interval of the mean, n>300 cells for each group, * indicates p<0.01 compared to control. B, Increased Sox9 phosphorylation and total protein levels by overnight dynamic compression. Agarose-embedded chondrocytes were subjected to 16–18 hours of dynamic compression (DC), or free-swell uncompressed controls (FS). After dynamic compression, the gels were immediately homogenized in SDS-PAGE sample buffer and analyzed by western blot for the amount of phosphorylated Sox9^Ser181 with an anti-phospho-Sox9^Ser181 antibody. Blots were stripped and re-probed for total Sox9 protein levels. The insets graphically show the image densitometry of the immunoblot, error bars indicate the standard deviation from 3 donors.

Figure 6. Pathway Schematic

The established pathway of Rho signaling is through ROCK and LIM Kinase to inactivate cofilin, an actin-severing protein. This results in an overall increase in F-actin, which through a series of unknown events leads to changes in matrix gene expression. Based on the observations in this manuscript we present a revised model in which ROCK directly phosphorylates Sox9^Ser181, causing increased nuclear localization and enhanced matrix gene expression. Thus, while actin remodeling also occurs upon stimulation of the revised pathway, this is a parallel event to the direct activation of Sox9 by ROCK.