Many physiological and pathophysiological processes in cells or tissues are involving mechanical stretch, which is inducing gap junction gene expression and cytokine TGF beta changes. However, the underlying mechanisms of gap junction gene expression changes remain unknown. Here, we showed that the expression of Cx43 at mRNA and protein level in Human umbilical-vein endothelial cells (HUVECs) is significantly increased after 24 h stretch stimulation, and TGF beta1, but not TGF beta2 expression is also upregulated. Administration of TGF betal into HUVECs without stretch also induced upregulation of Cx43 mRNA and protein expression. While simultaneously administration of TGF beta1 with SB431542, a specific inhibitor of TGF beta1 receptor, blocked the Cx43 protein upregulation by TGF beta1. Further, the increase of Cx43 protein expression under stretch condition can be partially blocked by SB431542; moreover, it also can be partially blocked by simultaneously administration of anti-TGF beta1 monoclonal neutralization antibody. Importantly, the stretch induced upregulation of Cx43 can be blocked by administration of actin and microtubule inhibitors, while NEDD4, a key element in mediating Cx43 protein ubiquitination and degradation, is not changed under stretch condition. Therefore, we conclude that upregulation of Cx43 expression under 24 h stretch condition is mediated by TGF beta1 via TGF beta1 receptor signaling pathway, and it also involves the actin and microtubule cytoskeletal network.
Cells or tissues in normal physiological or pathophysiological conditions undergo stretch to try to maintain the homeostasis. In blood vessel smooth muscle cells, mechanical stretch is a physiological feature due to blood pressure and blood flow. After meal, the stomach is expanded and underwent mechanical stretch for fulfilling its normal digestive function1.
Gap junctions are intercellular channels that allow direct cell to cell exchange of ions and small molecules with molecular weight less than 1000 Dalton such as water, ATP, calcium, CAMP, CGMP, IP3 etc. Gap junctional intercellular communications plays pivotal roles in maintaining physiological function of cells and tissues, which include cell differentiation, cell proliferation, cell survival, apoptosis, cell development and tissue homeostasis2. There are twenty-one connexin genes in human genome, among these connexins, connexin43 (Cx43) is the most widely distributed gap junction protein in different cells3.
Transforming growth factor beta1 (TGF beta1) is a secreted polypeptide cytokine and it belongs to the member of transforming growth factor superfamily. TGF beta1 signaling pathway is involved in many important cellular processes such as cell development, cell proliferation and cell differentiation in variety of cell types4, which are same as Cx43.
There are relevant literatures showed that either TGF beta1 or Cx43 were increased in different cell types under different stretch condition. For example, cyclic stretching showed selectively up-regulation of TGF beta1 but not TGF beta2 in cultured rat mesangial cells5; and mechanical stretch increased TGF beta1 expression in vascular smooth muscle cells6, in fetal hu-man intestinal smooth muscle cells, cardiomyocytes and in hepatic stellate cells7–9. In cultured neonatal rat cardiomyocytes, cyclic mechanical stretch showed a significant up-relation of Cx43 mRNA and protein under 4 to 16 hours stretch10. In human trabecular meshwork cells, it has been reported that Cx43 protein expression also increased under mechanical stretch condition11. Recently, some studies showed that administration of exogenous TGF beta1 into cultured cells can enhance Cx43 mRNA or protein expression12,13.
Taking into consideration of the same functions played by Cx43 and TGF beta1, it seems possible that TGF beta1 and Cx43 are functional related in some cell types. To date, however, there is no evidence for stretch induced Cx43 changes is mediated by TGF beta1 and the effect of TGF beta1 on Cx43 expression under stretch condition has not yet been investigated. Therefore, in the current study we try to evaluate Cx43 and TGF beta1 expression under stretch condition, blocking TGF beta1 signaling pathway on Cx43 expression and temporal effect of administration TGF beta1 on Cx43 expression in cultured Human umbilical-vein endothelial cells (HUVECs). In addition, Cx43 has been demonstrated to be associated with actin and microtubule cyto-skeleton network14–16, and Cx43 has also been shown to be heavily regulated by an E3 ubiquitination enzyme, name NEDD417. Therefore, we also investigated the effect of blocking actin and microtubule cytoskeletal network, as well as the contribution of NEDD4 on mechanical stretch induced Cx43 expression changes.
Materials and reagents.
Rabbit polyclonal anti-Cx43 antibody (Cat. No 71-0700) was purchased from Invitrogen, N-terminal monoclonal anti-Cx43 antibody and monoclonal anti-NEDD4 antibody were obtained from BD Bioscience Inc. Polyclonal anti-GAPDH, anti-beta Tubulin were acquired from Sigma Aldrich, anti-Cx37 antibody, SB431542, monoclonal antibodies for TGF beta1 and TGF beta2, Latrunculin A and Nocodazole were obtained from Invitrogen, Tocris, BD Bioscience Inc., and Sigma-Aldrich, respectively.
HUVECs were purchases from American Tissue and Cell Culture (ATCC) and cultured in a 75 ml flask in F-12K cell culture medium containing 10% FBS, 1% penicillin and 0.03 mg/ml endothelial cell growth supplement (ECGS) according to the instructions from the manufacturer. After the cells reached about 90% confluency, they were transferred to three collagen I coated Bio flex 6-well plates (Flex cell International Corporation, Burlington, NC, USA) in F-12 K medium containing serum, after culturing the HUVECs for three days and the cells con-fluency reached about 80%, the cell culture medium changed to serum free. After five hours stabilization, the cells were undergoing 12 or 24 hours of mechanical stretch. For mRNA studies, samples were collected from both 12 and 24 hours of mechanical stretch condition, for protein studies, samples were collected only from 24 hours of mechanical stretch condition. Control cells plated the same condition in collagen I coated Bio flex 6-well plates without being subjected to mechanical stretch.
For mechanical stretch, after cultured cells reaching around 80% confluence, HUVEC cells were switched to serum-free DMEM for five hours, followed by cyclic mechanical stretch for 24 hours with the following parameters: (10% stretching, 1 cycles/second) using the FX-5000 Tension System controlled by the computer (Flexcell International Corporation).
For administration of actin and microtubule inhibitors, the cells were incubated with 50uM of actin cytoskeleton inhibitor Latrunculin A or microtubule inhibitor Nocodazole or Latrunculin A and Nocodazole combined together (all dissolved in dimthylsulfoxide, DMSO), which has been demonstrated previously by other groups that using the same dosage of Nocodazole or Latrunculin A can significantly block actin and microtubule cytoskeleton. Moreover, administration of DMSO was used as a control, then undergoing mechanical stretch treatment. All experiments were performed in duplicate and repeated up to three times.
After 24 hours mechanical stretch, the cell culture serum free medium from control and stretched plates were collected and were subjected to measurement of TGF beta1 and TGF beta2 concentration using Elisa kits with standard protocol provided by the manufacture (R & D systems).
Total RNA extraction, RT-PCR and real time PCR.
The procedure for RNA extraction, RT-PCR and real time PCR were used and described previously and complied with the manufacturer’s protocol . Briefly, cells were collected after 12 or 24 h mechanical stretch, total RNA was extracted using trizol reagents (Invitrogen) and RNA concentrations were determined using a Nanodrop machine. Reverse transcription was conducted using SuperScript® III reverse transcriptase (RT) and oligo (dT)20 primer. The PCR conditions for annealing is dependent on the primers. Single band PCR product were verified by electrophoresis in a 2% agarose gel stained with ethidium bromide. Real time PCR was performed using RT² SYBR Green Fluor qPCR Master mix and the standard protocol.
The following primers were used in this study.
GAPDH Sense: 5’-AATCCCATCACCATCTTCCAGGAG-3’
GAPDH antisense: 5’-CACCCTGTTGCTGTAGCCAAATTC-3’
Cx37 Sense: 5’-TCAGCACACCCACCCTGGTCT-3’
Cx37 antisense: 5’-GGATGCGCAGGCGACCATCTT-3’
Cx43 sense: 5’- GGTCTGAGTGCCTGAACTTGCCT-3
Cx43 antisense: 5’-AGCCACACCTTCCCTCCAGCA-3’
TGF-beta2 sense primer: 5′- GAAGACCCCACATCTCCTGCTA-3’
TGF-beta2 antisense primer: 5′- AGCAATAGGCCGCATCCAA-3’
Human Cx43 expression vector construction:
The procedure for constructing human Cx43 expression vector was conducted as previously described18. Oligonucleotide primers were obtained from Integrated DNA technologies (IDT) Inc. (Coralville, Iowa), transfection reagents were purchased from Life Technologies Inc. Primers chosen for RT-PCR were based on the human Cx43 sequence (NCBI Reference Sequence: NM_000165.5). These were: full length Cx43 sense primer: 5’-ATGGGTGACTGGAGCGCCTTAG-3’; full length Cx43 antisense primer: 5’-CTAGATCTCCAGGTCATCAGG-3’.
For amplification of human Cx43 coding cDNA, PCR was conducted in 20 µL solution containing 2 µL 10 × PCR buffer; 0.8 µL 50 mM MgCl2; 200 µM dNTP; 100 ng sense and antisense primers;1 unit Taq DNA polymerase; and 1 µL template cDNA. For PCR, DNA was denatured at 95°C for 3 min, and then subjected to 30 cycles at 95°C for 60 s, 55°C for 60 s and 72°C for 90 s; this was followed by a final extension at 72°C for 10 min for T-A cloning. PCR products were separated by electrophoresis in 1% agarose gel, stained with ethidium bromide and purified using a gel purification kit (Qiagen Inc, Mississauga, ON, Canada). PCR products were subcloned into PCR 2.1 vector, digested with BamHI and ApaI, and ligated into pcDNA3.1 expression vector using T4 DNA ligase according to the manufacturer’s instructions. Recombinant plasmids were extracted, orientation was verified with KpnI and BstX1 digestion, and at least two recombinant plasmids were sequenced using T7 universal primer and specific Cx43 primer for confirmation of sequence.
HeLa cells (American Type Culture Collection, Rockville, MD, USA) were grown in Dulbecco’s modified Eagle’s medium with low glucose and supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin. For transient transfection, HeLa cells with 80% confluence were transfected for 48 h with pcDNA3.1 vector, full length Cx43-pcDNA3 or truncated Cx43-pcDNA3.1 plasmids using LipofectAMINE 3000 reagent as described previously18,19, and the cells were harvested followed by Western blotting assay for validating anti-Cx43 antibody specificty.
The procedure for immunofluorescence labeling was used as we previously described18,20. Briefly, mechanical stretched and non-stretched HU-VECs grown in 6 well Bio flex plates were fixed with 2% cold paraformaldehyde for 10 min, and then were washed with PBS. For immunolabelling, slides were incubated in 50 mM Tris-HCl, pH 7.4, containing 1.5% sodium chloride (TBS) and 0.3% Triton X-100 (TBSTr) and 5% NGS for 24 h at 4 ℃ with Cx43 polyclonal antibody at 1:1000, they were then washed for 1 h in TBSTr and incubated for 1.5 h at room temperature simultaneously with Alexa Fluor 488-conjugated goat anti-rabbit IgG at 1:1500 dilution (Molecular Probes Eugene, Oregon). Following incubation with secondary antibodies, slides were sequentially washed in TBSTr for 20 min, in 50 mM Tris-HCl buffer, pH 7.4 for 30 min, and then coverslipped using anti-fade medium. To test for inappropriate cross-reactions between primary and secondary antibodies or between different secondary antibodies, control procedures included omission of one of the primary antibodies with inclusion of each of the secondary antibodies. Confocal immunofluorescence images were gathered on an Olympus confocal microscope using the Fluoview program. Some confocal images are presented as z-stacks of six to ten scans at z scanning intervals of 0.5 µm.
Western blotting. The procedure for western blotting was described previously18,21. Briefly, Cells in mechanical stretch and non-stretch condition were harvested in an IP buffer (20 mM Tris-HCl, pH 8.0, 140 mM NaCl, 1% Triton X-100, 10% glycerol, 1 mM EGTA, 1.5 mM MgCl2, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, and 5 µg/ml each of leupeptin, pep-statin A, and aprotinin) and sonicated. Homogenates were centrifuged at 20,000 x g for 20 min at 4°C, and the supernatants were taken for protein determination using Bradford reagent (Bio-Rad Laboratories). Proteins containing 5% β-mercaptoethanol were boiled for 5 min and were separated by SDS-PAGE (5 µg of protein per lane) using 10% or 12.5% gel followed by transblotting to polyvinylidene difluoride membranes (Bio-Rad Laboratories) in standard Tris-glycine transfer buffer, pH 8.3, containing 0.5% SDS. Membranes were blocked for 2 h at room temperature in TBSTw (10 mM Tris-HCl, pH 8.0, 150 mM NaCl, and 0.2% Tween 20) containing 5% nonfat milk powder, rinsed briefly in TBSTw, and incubated overnight at 4°C with polyclonal anti-Cx43 primary antibody diluted at 1: 1000 in PBS containing 1% nonfat milk or polyclonal anti-alpha-GAPDH antibody diluted at 1:1000 (Novus Biotechnological Inc) in TBSTw containing 1% nonfat milk powder. Membranes were washed four times in TBSTw for 40 min, incubated with horseradish peroxidase-conjugated donkey anti-rabbit IgG or anti-mouse IgG diluted 1:3000 to 1:5000 (Sigma-Aldrich) in TBSTw containing 1% nonfat milk powder, washed four times in TBSTw for 40 min, and re-solved using an Odyssey imagers. The optical density (OD) of specific band was also acquired.
Statistical analysis. All data were expressed as mean ± SEM. The one-way ANOVA test was employed to examine the statistically significant differences between groups. Significant differences were set at * (P<0.05), ** (P<0.01) and *** (P<0.001).
Here, we used HUVECs to monitor the changes of Cx43 mRNA and protein expression level under mechanical stretch, which occurs in many physiological and pathological conditions. Even if we used the Cx43 antibody previously with the same catalog number, the different lot number may generate different results, therefore, we first characterized the Cx43 antibody we employed, we constructed Cx43 expression vector and transiently transfected full-length Cx43 expression plasmid into Cx43 cell-cell communication deficient HeLa cells using lepofectamine transfection reagents as we previously described, an empty vector was used as a control. After two days of transfection, the cell lysates were harvested and subjected to SDS-PAGE electrophoresis and western blotting analysis. As shown in Fig. 1C, a protein band migrating at predicted location is present in the lysates of Cx43 transiently transfected HeLa cells, but not in the lysates of empty vector transfected HeLa cells, then the membrane was stained with Coomassie blue to verify the protein loading at both lanes, indicating the specificity of this anti-Cx43 antibody we utilized. Further, we used this validated anti-Cx43 antibody and checked the Cx43 protein expression using immunofluorescence labeling in HUVECs. As shown in Figure 1A and 1B, the punctate labeling of Cx43 protein was much robust labeled under 24 h stretch condition compared with non-stretch condition.
The immunofluorescence labeling of Cx43 showed typical Cx43 cellular localization, but the more accurate quantification of Cx43 at mRNA and protein level needs real time qPCR and western blotting techniques. Subsequently, we investigated Cx43 and Cx37 expression in mRNA and protein level under 12 or 24 h stretch condition. As shown in Figure 2A-C, Western blot results showed that Cx43 protein was significantly increased after 24 h stretch, while there was no change of Cx37 protein level in stretch and non-stretch condition. After verification of the specificity of the PCR primers using regular RT-PCR, real time qPCR was used to determine Cx43 and Cx37 mRNA expression under stretch condition, as shown in Figure 2D,E, Cx43 mRNA was and significantly increased after 12 or 24 h stretch; however, for Cx37 mRNA expression, there was no statistically significant changes under stretch and non-stretch condition. It appears that stretch can increase Cx43 mRNA and protein level, but no changes in Cx37 mRNA and protein level.
Next step, we carried out detailed analysis of TGFbeta1 and TGFbeta2 mRNA level under 12 and 24 h stretch condition using real time qPCR and Elisa techniques. As shown in Figure 3A and 3B, the TGFbeta1 mRNA level was increased with statistically significance under 12 or 24 h stretch condition, however, there was no change of TGF beta2 mRNA expression level under the same condition. From this analysis, it was clearly shown that TGF beta1 but not TGF beta2 mRNA level was increased under stretch condition. We also checked TGF beta1 and TGF beta2 protein con-centration from serum free medium under 12 and 24 h stretch condition using Elisa assay. As shown in Figure 3C and 3D, the TGF beta1 protein concentration in cell culture medium was significantly higher compared with cell culture medium under no stretch condition. However, for TGF beta2 concentration, there was no statistically difference between stretch and no stretch condition. From the Elisa assay, we concluded that there was an increased TGF beta1 protein concentration in cell culture medium under stretch stimulation.
3.3. Administration of TGF beta1 upregulates Cx43 mRNA and protein expression in HU-VECs and the effect of TGF beta1 on Cx43 can be partially inhibited by TGF beta1 receptor inhibitor SB431542.
Taking into consideration that administration of TGF beta1 exogenously into some cells can stimulate Cx43 mRNA and protein expression, while in some different cells [11, 12, 15, 16], ad-ministration of TGF beta1 can down-regulate Cx43 expression [17, 18]. Therefore, it is importance to investigate whether administration of TGF beta1 can increase or decrease Cx43 mRNA or protein expression in HUVECs. Real time Q RT-PCR results indicated TGF beta1 can up-regulate Cx43 mRNA expression, which was shown in Figure 4A. We further investigated whether Cx43 protein expression level was also upregulated. As shown in Figure 4B and 4C, administration of TGF beta1 (10 ng/ml) into HUVECs for 24 h can significantly increase and accumulate Cx43 labeling, which was evident of robust labeling of punctate Cx43. Western blot results showed strong Cx43 detection after TGF beta1 treatment, which is statistically increased compared with control cells. Furthermore, we used a specific TGF beta1 receptor inhibitor (SB431542) to check whether it can block the upregulation of Cx43 expression induced by TGF beta1, as shown in Figure 4D, the Cx43 protein expression was significantly inhibited after simultaneously treated cells with TGF beta1 (10 ng/ml) and SB431542 (5 uM).
In summary, these data demonstrated that TGF beta1 can increase Cx43 mRNA and protein expression in HUVECs, it appears the effect of TGF beta1 on Cx43 expression is via TGF beta1 receptor.
3.4. Upregulation of Cx43 under stretch condition can be partially blocked by SB431542, an inhibitor of TGF beta1 receptor.
In order to test whether the TGF beta1 receptor inhibitor (SB431542) play a role in regulating Cx43 protein expression in HUVECs under stretch condition, we tested Cx43 protein expression after simultaneously treated cells with SB431542 (5 uM) under 24 h stretch condition. As shown in Figure 5A and 5B, the results showed the upregulation of Cx43 protein level was significantly reduced in HUVECs co-cultured with SB431542 (5 uM) under 24 h stretch condition compared with HUVECs co-cultured with DMSO, a reagent used for dissolving SC431542 under 24 h stretch only. Similarly, the immunofluorescence labeling of Cx43 was also reduced after administration of SB431542 (5 uM) under stretch condition. This results clearly indicated that administration of TGF beta1 receptor inhibitor, SB431542 to HUVECs under stretch condition can partially block the effect of stretch induced Cx43 protein expression.
3.5. Upregulation of Cx43 under stretch condition can be partially blocked by TGF beta1 monoclonal neutralization antibody.
It was well known that administration of monoclonal anti-TGF beta1 antibody can block the TGF beta1 biological activity, therefore we checked the Cx43 expression after administration of TGF beta1 monoclonal neutralization antibody into HUVECs under 24 h stretch condition. As shown in Figure 6A and 6B, Cx43 protein level was significantly reduced in HUVECs co-cultured with anti-TGF beta1 monoclonal neutralization antibody (5 g/ml) under stretch condition com-pared with HUVECs under stretch condition only, with results obtained using immunofluorescence labeling and western blot assay. Both the western blot and immunofluorescence data indicated that administration of TGF beta1 monoclonal neutralization antibody to HUVECs under stretch condition can partially block the upregulation of Cx43 expression under 24 h stretch condition.
3.6. Upregulation of Cx43 under stretch condition can be partially blocked by administration of actin and microtubule cytoskeleton inhibitors
Because Cx43 associates with actin and microtubule cytoskeleton, so we further in-vestigated whether the upregulation of Cx43 under stretch involves actin and microtubule tubulin cytoskeleton. Immunoblot result showed that the increment of Cx43 protein expression under mechanical stretch was partially inhibited in stretched HUVECs after administering actin cyto-skeleton inhibitor Latrunculin A (1 uM) or microtubule inhibitor Nocodazole (1 uM), and after administrating of Latrunculin A (1 uM) and Nocodazole (1 uM) simultaneously, it appears that stretch induced the upregulation of Cx43 was further dramatically inhibited (P<0.0001, Figure 7). These data indicated that both actin and microtubule cytoskeleton were involved in the upregulation of Cx43 expression under mechanical stretch condition.
It is well established that gap junctions composed of Cx43 are highly dynamic structures, and the turnover rate of Cx43 formed gap junctions are heavily regulated by its repaid protein ubiquitination or degradation with a half-life around 3 hours. To date, NEDD4 (neuronal prescusor cell-expressed developmentally downregulated 4), an E3 ubiquitin ligase, has been shown to interact with Cx43 and promote Cx43 protein degradation via endocytosis and lysosomal sorting of Cx43. Because previous work has shown that there is endogenous NEDD4 protein expression in HUVEC cells, so we also investgatedwhether there is any change in NEDD4 protein expression under stertch condition. If there is changes in NEDD4 protein expression, we can not exclude the the possibility that changes in Cx43 proetin expression under stretch condition may relate to changes in NEDD4 mediated Cx43 degradation and loss. As shown in Figure 8, 5 ug of cell lysates of HUVECs from streched and no-stretched cells were loaded and separated on SDS-PAGE gels, then transbloted to PVDF membranes, followed by probing with monoclonal anti-NEDD4 and polyclonal anti-beta-Tubulin antibodies simultaneously and relevant secodary antibodies, WB results showed that a 110 kDa perdicted high molecular weight band of NEDD4 protein was probed with green color (Figure 8A, upper band), while a 55 kDa predicted lower molecular weight band of beta-Tubulin protein were probed with red color (Figure 8A, lower band), the same membrane was also exposed in black and white background (Figure 8B) and showing the presence of high molecular weight NEDD4 band, as well as lower molecular beta-Tubulin band, indicating the specificity of these antibodies employed. The statistic results showed that there was no significant difference in NEDD4 protein expression in cell lysates of HUVECs between stretach and no-stretch groups (Figure 8C), indicating NEDD4 E3 ubiquitin ligase, one povital enzyme that modulate Cx43 degradation and loss, may not play an important role in the upregulation of Cx43 expression under mechnical stretch condition.
The results from current experiments are five-fold: First, we showed that the expression of one gap junction gene, Cx43 at mRNA and protein level in HUVECs is significantly increased after stretch stimulation, and TGF beta1, but not TGF beta2 expression is also upregulated using qRT-PCR and Elisa assay. Second, administration of TGF beta l into HUVECs without stretch also induced upregulation of Cx43 mRNA and protein expression. While simultaneously administration of TGF beta1 with SB431542, a specific inhibitor of TGF beta1 receptor, blocking the Cx43 protein up-regulation. Third, the increase of Cx43 protein expression under stretch condition can be partially blocked by utilizing TGF beta1 receptor inhibitor SB431542 and it also can be partially blocked by simultaneously administration of anti-TGF beta1 monoclonal neutralization antibody. Fourth, the upregulation of Cx43 under stretch condition can be partially blocked by administration of actin and microtubule cytoskeleton inhibitors. Last, there is no changes in NEDD4 protein under stretch condition, indicating the E3 ubiquitin ligase NEDD4 may not involve in the Cx43 upregulation. Therefore, we conclude that upregulation of Cx43 expression under 24 h stretch condition is mediated by TGF beta1 via TGF beta1 receptor signaling pathway, further, this process involves actin and microtubule cytoskeletal network, but no involvement of NEDD4 mediated Cx43 degradation pathway.
Cx43 is one of the most ubiquitously distributed gap junction protein in the body, it plays a pivotal role in the regulation of human homeostasis, which includes cell differentiation, cell growth, cell proliferation, passing apoptosis signaling, etc. Cx43 protein is highly regulated at the transcriptional level and also can undergo post-translational modification such as phosphorylation and ubiquitination, gap junctions including Cx43 are considered as dynamic structures with a very short turnover or a short half-life of several hours2. Cx43 is highly regulated by mechanical stretch, which can change Cx43 expression at mRNA and protein levels. Previously literature showed that stretch and shear stress can increase Cx43 mRNA and protein expression at different cell types and different stretch systems22, our results showed that stretch can upregulate both of TGF beta1 and Cx43 expression, but no statistically changes of TGF beta2 and Cx37 in HUVECs. Our results of upregulation TGF beta1 under 24 h stretch condition is in line with previous works5–9. Mechanical stretch can change the TGF beta level, and previous work also showed that mechanical stretch increased TGF beta1 mRNA level by northern blot or using qRT-PCR method, and changed TGF beta1 protein by Elisa method, furthermore, these data showed that TGF beta2 mRNA or protein is not significantly changed5–9. Our results are consistent with these previous works. However, previous works were only investigated stretch on Cx43 expression, or stretch on TGF beta expression, but not both. To the best of our knowledge, our work demonstrates the possible mechanism of stretch induced upregulation of Cx43 pression, which is mediated by TGF beta1.
TGF beta signaling pathway plays a pivotal role in the regulation of Cx43 gene and protein expression. TGF beta signaling pathway includes canonical pathway via Smad2/3 signaling, and non-canonical signaling pathway via P38 signaling4. Previous works by other groups showed that TGF beta1 is an important regulator or Cx43 mRNA and protein level in different cell lines or tissues arranging from mouse, rat to human12,13,23,24. It showed that the TGF beta1 may regulate Cx43 expression in a broad manner with diversity of cells.
TGF beta1 can regulate Cx43 gene expression via phosphorylation of SMAD2/3 pathway in human granulosa cells12 or via phosphorylation of P38 MAPK pathway in normal murine mammary gland epithelial cells13. While previously work also showed that mechanical stretch can both increase SMAD2/3 phosphorylation23 in mesenchymal stem cells and P38 phosphorylation in human saphenous vein24. The specific TGF beta1 receptor inhibitor SB431542 can inhibit SMAD2 phosphorylation. In our model system, it raises the possibility that TGF beta1 upregulation under stretch condition may increase SMAD2 phosphorylation therefore modulating of Cx43 gene expression, but we cannot exclude the possibility that stretch also may enhance P38 phosphorylation and subsequently upregulate Cx43 expression.
Cellular cytoskeleton plays an important role in maintaining the homeostasis of cells and tissues, and many physiological activities of cells such as cell motility, migration need the participation of the cellular cytoskeleton, the mechanical stretch is not an exception because stretch can change actin cytoskeleton25–27. The regulation of mechanical stretch and changes in Cx43 expression by cellular cytoskeleton is not surprising, because Cx43 has been reported to be associated with both actin and microtubule cytoskeleton28,29, and microinjection of actin antibody can block GJIC in cultured astrocytes30; similarly, blocking actin cytoskeleton causes stop trafficking Cx43 to the plasma membrane, and reduced plasma membrane Cx43 as well as GJIC after destroying microtubule network by Latrunculin A4,31. Our work showed that mechanical stretch induced the upregulation of Cx43 can be partially blocked by actin and microtubule cytoskeleton inhibitors, and when utilizing both of actin and microtubule cytoskeleton inhibitors simultaneously, it showed further reduced Cx43 expression. These data indicated that actin cytoskeleton as well as microtubule cytoskeleton were involved in the regulation of Cx43 expression under mechanical stretch condition.
The dynamic process that governs Cx43 removal and disposal via ubiquitination-dependent and ubiquitination-independent pathway is well established17. For the ubiquitin degradation pathway, it involves one key E3 ubiquitin ligase, NEDD4 and other participants, such as ubiquitin, ubiquitin-activating enzyme (E1) and ubiquitin-conjugating enzyme (E2). NEDD4 interacts with Cx43 through the WW domain of NEDD4 and a PY consensus motif at the C-terminal Cx43. If there is change in NEDD4 expression, it will affect the association of NEDD4 and Cx43 and subsequently, has an effect on Cx43 degradation, as well as its expression. Due to there is no changes in NEDD4 protein expression in stretch condition, so we conclude that NEDD4 is not involved in the Cx43 expression changes under stretch condition. However, in human studies, it has been found that a single-nucleotide polymorphisms in NEDD4L may cause partial loss function on NEDD4L and therefore, reducing its function on inhibiting some epithelial sodium channels expression, and subsequently causing some high blood pressure symptoms in patients32. A recent report also demonstrated that NEDD4 protein expression level was significantly increased in breast cancer samples33, considering the fact that some reports showed a decreased expression of Cx43 in breast cancer samples, it may indicate the reduced Cx43 expression may be related to increased NEDD4 expression, because inhibiting NEDD4 expression induces changes in its interacting protein degradation has been reported34.
It is noteworthy to show that previous works by others showed the involvement of TGF beta1 in upregulation of Cx43 expression12,13,35,36. However, some works also showed that administration of TGF beta1 can down-regulate Cx43 expression at mRNA or protein level37,38, this discrepancy may be related to different cells with different response or different time point dependent action.
It is also noteworthy to mention that Cx43 can regulate TGF beta signaling pathway via its competition with SMAD2 for binding to microtubules24,39. In our cell culture system, whether upregulated Cx43 can play a role in modulating TGF beta1 upregulation remains unknown.
Further, there are additional growth factors other than TGF beta1 that can regulate Cx43 expression, such as epidermal growth factor (EGF), and Vascular endothelial growth factor (VEGF)40,41. In addition, Cx43 expression can be regulated at transcriptional level or posttranslational level such as changing degradation. Our results demonstrated that stretch can upregulate Cx43 mRNA in HUVECs, therefore it appears that TGF beta1 regulate Cx43 expression is related to its regulation at transcriptional level, however, whether changes in Cx43 degradation under stretch condition is also an important contributor of Cx43 protein upregulation need further investigation.
In summary, our current study demonstrate that mechanical stretch upregulates both Cx43 and TGF beta1 expression in HUVECs, the upregulation of Cx43 expression under stretch condition can be partially blocked by TGF beta1 receptor inhibitor SB431542 or by specific anti-TGF beta1 monoclonal neutralization antibody, indicating the involvement of TGF beta1 signally pathway in the regulation of Cx43 expression in HUVECs under stretch condition. More importantly, we showed that mechanical stretch induced the upregulation of Cx43 can be blocked by administration of actin and microtubule cytoskeleton inhibitors, indicating that actin and microtubule cytoskeletal networks are involved in this process, while NEDD4, the key enzyme that regulates Cx43 degradation pathway, is not changed, indicating NEED4 is not involved in the upregulation of Cx43 under stretch condition. Therefore, we conclude that there are functional relationships among Cx43, TGF beta1, actin and microtubule cytoskeletal networks under stretch condition.
Author contributions statement
YS: acquisition of data and drafting the manuscript. XL: conception and design of study. JY: revising the manuscript critically for important intellectual content.
The authors declare no conflict of interest.