Transforming growth factor-β1 up-regulates connexin43 expression in osteocytes via canonical Smad-dependent signaling pathway

Abstract

Connexin 43 (Cx43)-mediated gap junctional intercellular communication (GJIC) has been shown to be important in regulating multiple functions of bone cells. Transforming growth factor-β1 (TGF-β1) exhibited controversial effects on the expression of Cx43 in different cell types. To date, the effect of TGF-β1 on the Cx43 expression of osteocytes is still unknown. In the present study, we detected the expression of TGF-β1 in osteocytes and bone tissue, and then used recombinant mouse TGF-β1 to elucidate its effect on gap junctions (GJs) of osteocytes. Our data indicated that TGF-β1 up-regulated both mRNA and protein expression of Cx43 in osteocytes. Together with down-regulation of Cx43 expression after being treated with TGF-β type I receptor inhibitor Repsox, we deduced that TGF-β1 can positively regulate Cx43 expression in osteocytes. Thus we next focussed on the downstream signals of TGF-β and found that TGF-β1-mediated smads, Smad3 and Smad4, to translocate into nucleus. These translocated signal proteins bind to the promoter of Gja1 which was responsible for the changed expression of Cx43. The present study provides evidence that TGF-β1 can enhance GJIC between osteocytes through up-regulating Cx43 expression and the underlying mechanism involved in the activation of Smad-dependent pathway.

Keywords: Connexin 43; Gap junction; Osteocyte; Smad3; Smad4; TGF-β1.

Figure 1. Gene profile of TGF-β superfamily and receptors of TGF-β superfamily in mouse osteocyte and bone tissue

(A,B) Quantitative real-time PCR screen showing higher mRNA expression of TGF-β1 amongst TGF-β superfamily in both osteocyte cell line (A, upper) and bone tissue (B, lower). The results were based on the three independent experiments (n=3). (C,D) Quantitative real-time PCR screen showing mRNA levels of receptors of TGF-β superfamily in both osteocyte cell line (C) and bone tissue (D). The results were based on the three independent experiments (n=3).

Figure 2. The TGF-β1 promotes cell–cell GJ of osteocytes

(A) Representative images of scratch wound closure assay observed by phase-contrast microscopy at 20× magnification. The results were based on three independent experiments (n=3). (B) CCK-8 assay showing increased cell proliferation of osteocytes induced by increased dosages of TGF-β1. The results were based on the three independent experiments (n=3). *Significant difference was respected to the control group. ng denotes ng/ml. (C) Cell morphology assay showing the increase in dendritic processes in osteocytes induced by 5 ng/ml TGF-β1 by phase-contrast microscopy at 40× magnification. The results were based on three independent experiments (n=3). (D) Quantitation was done to confirm the increase in dendritic processes induced by TGF-β1. *Significant difference was respected to the control group (P<0.05). (E) The dye transfer (DT) assay showing the increased GJ formation in osteocytes induced by TGF-β1 (5 ng/ml) by CLSM at 40× magnification. The dye transfer images were collected at 10 min after Lucifer Yellow staining. The results were based on three independent experiments (n=3). (F) Quantitation was done to confirm the increase in GJ formation in osteocytes induced by TGF-β1. *Significant difference was respected to the control group (P<0.05). (G) The SL/DT assay further showing a cell density-dependent increase in GJs in osteocytes induced by TGF-β1 (5 ng/ml); 1×, 2×, and 3× represent multiplied cell densities. The images were collected at 7 min after Lucifer Yellow staining along the scrape. The boxed area further showed the different transmission speeds between the control and TGF-β1 group. The purple arrows showed transmission direction after Lucifer Yellow loading. (H) Statistical analysis showing the changes inf transmission speeds after Lucifer Yellow loading between the control and TGF-β1 group (5 ng/ml). Data are presented as mean ± S.E.M. (n=3). *P<0.05.

Figure 3. The TGF-β1 promotes GJ through the increase in Cx43

(A) mRNA expression of Cx43 by qPCR after being treated with different concentrations of TGF-β1. The results were based on three independent experiments (n=3). **P<0.01. (B) Western blot analysis of Cx43 upon exposure to TGF-β1 for 24 and 36 h. Quantitation was performed to confirm the protein changes (n=3). *P<0.05; **P<0.01. (C) Representative IF staining by CLSM showing the elevated expression of Cx43 and potential GJ in osteocytes in response to TGF-β1 (cytoskeleton, green; Cx43, red; nucleus, blue). The results were based on three independent experiments (n=3). White arrows showing that Cx43 distribution was along the dendritic processes. (D) The total fluorescent qualification was performed to show the expression changes of Cx43 induced by TGF-β1 (5 ng/ml). The results were based on three independent experiments (n=3). *P<0.05. (E) The qualification further indicated the changes of GJ numbers induced by TGF-β1 (5 ng/ml). *Significant difference was respected to the control group (P<0.05).

Figure 4. Repsox, an inhibitor of TGF-β type I receptor, reduces the expression of Cx43

(A) qPCR showing down-regulated mRNA expression of Cx43 after being treated with Repsox. The results were based on the three independent experiments (n=3). **P<0.01. (B) Western blot assay showing the expressions of Cx43 upon exposure to Repsox for 24 h. Quantitation was performed to confirm the protein changes (n=3). **P<0.01. (C) Representative IF staining by CLSM showing the decreased Cx43 in osteocytes in response to Repsox (cytoskeleton, green; Cx43, red; nucleus, blue). The results were based on three independent experiments (n=3). (D) Quantitation of GJ number by ImageJ was performed to confirm the changes after Repsox treatment (n=3). *P<0.05.

Figure 5. TGF-β1 mediates Cx43 via transducing signals, Smad3 and Smad4, by direct binding signals to the promoter of Cx43 gene

(A) Western blot assay showing the expression of Smad3, p-Smad3, and Smad4 upon exposure to TGF-β1 for 24 and 36 h. The results were based on three independent experiments (n=3). (B) Quantitations were performed to confirm the protein changes in (A) (n=3). *P<0.05, **P<0.01. (C,D) Representative IF staining by CLSM showing the nuclear translocation of Smad3 (upper) and Smad4 (lower) in osteocytes in response to TGF-β1 (cytoskeleton, green; Samd3 and Smad4, red; nucleus, blue). The results were based on three independent experiments (n=3). (E) Bioinformatics showing the binding sites of nuclear translocated-Smad3 and -Smad4 to the promoter of Cx43 gene (Gja1, GenBank name). Smad3 showed three potential binding sites in the promoter of Gja1, and the sites of sequences were located at 1434–1424, 827–817, and 672–622 kb before TSS of Gja1. Smad4 showed two binding sites in the promoter of Gja1, and the sites of sequences were located at 828–817 and 672–621 kb before TSS of Gja1.

Figure 6. Repsox represses translocation of Smad3 and Smad4 into nucleus resulting in attenuation of the effect of TGF-β1 on Cx43

(A) Western blot assay showing the expression of Smad3, p-Smad3, and Smad4 upon exposure to Repsox for 24 h. The results were based on three independent experiments (n=3). (B) Quantitation was performed to confirm the protein changes in (A) (n=3). **P<0.01. (C,D) Representative IF staining by CLSM showing reduced translocation of Smad3 (C) and Smad4 (D) in osteocytes after pretreatment with 50 μM Repsox for 6 h and then treated with 5 ng/ml TGF-β1 for 24 h compared with the group only treated with TGF-β1 (cytoskeleton, green; Smad3 and Smad4, red; nucleus, blue). The results were based on the three independent experiments (n=3). (E) Representative IF staining by CLSM showing that Repsox attenuated the Cx43 expression even in TGF-β1-treated osteocytes. (F) Quantitation further showing the changes in dendritic processes of osteocytes pre-incubated with Repsox after treatments of TGF-β1 (n=3). **P<0.01.

Figure 7. The schematic diagram elucidating the Smad-dependent pathway and non-Smad-dependent pathway involved in the changes of Cx43 expression induced by TGF-β1

The non-Smad-dependent pathway is predicted to be involved the changes of Cx43 but not in the present study.