Stretch-induced hypertrophy activates NFkB-mediated VEGF secretion in adult cardiomyocytes

Abstract

Hypertension and myocardial infarction are associated with the onset of hypertrophy. Hypertrophy is a compensatory response mechanism to increases in mechanical load due to pressure or volume overload. It is characterized by extracellular matrix remodeling and hypertrophic growth of adult cardiomyocytes. Production of Vascular Endothelial Growth Factor (VEGF), which acts as an angiogenic factor and a modulator of cardiomyocyte function, is regulated by mechanical stretch. Mechanical stretch promotes VEGF secretion in neonatal cardiomyocytes. Whether this effect is retained in adult cells and the molecular mechanism mediating stretch-induced VEGF secretion has not been elucidated. Our objective was to investigate whether cyclic mechanical stretch induces VEGF secretion in adult cardiomyocytes and to identify the molecular mechanism mediating VEGF secretion in these cells. Isolated primary adult rat cardiomyocytes (ARCMs) were subjected to cyclic mechanical stretch at an extension level of 10% at 30 cycles/min that induces hypertrophic responses. Cyclic mechanical stretch induced a 3-fold increase in VEGF secretion in ARCMs compared to non-stretch controls. This increase in stretch-induced VEGF secretion correlated with NFkB activation. Cyclic mechanical stretch-mediated VEGF secretion was blocked by an NFkB peptide inhibitor and expression of a dominant negative mutant IkBα, but not by inhibitors of the MAPK/ERK1/2 or PI3K pathways. Chromatin immunoprecipitation assays demonstrated an interaction of NFkB with the VEGF promoter in stretched primary cardiomyocytes. Moreover, VEGF secretion is increased in the stretched myocardium during pressure overload-induced hypertrophy. These findings are the first to demonstrate that NFkB activation plays a role in mediating VEGF secretion upon cyclic mechanical stretch in adult cardiomyocytes. Signaling by NFkB initiated in response to cyclic mechanical stretch may therefore coordinate the hypertrophic response in adult cardiomyocytes. Elucidation of this novel mechanism may provide a target for developing future pharmacotherapy to treat hypertension and heart disease.

Figure 1. Cyclic mechanical stretch induces VEGF secretion in primary ARCMs.

A) Cyclic mechanical stretch for either 24 h or 48 h induces a significant increase in VEGF secretion in primary ARCMs attached to laminin compared to non-stretched controls. Stretched (10% stretch) and non-stretched control (0% stretch) cells were allowed to adhere to laminin for 24 h prior to initiating experiments. VEGF concentration in the conditioned media of non-stretched and stretched ARCMs was analyzed by ELISA. ***P<0.001; One-Way ANOVA with a Bonferroni post-test was used to determine the statistical significance of data. The values represent the average of three independent experiments. B) Isolated ARCMs remain viable in culture when attached to laminin. Cell cultures were examined for their ability to reduce MTT after 24 h or 48 h in culture. Fold induction in relative mitochondrial activity represents the amount of viable cells at each time point. ***P<0.001; One-Way ANOVA with a Bonferroni post-test was used to determine the statistical significance of data. Insert: The isolation procedure yielded a >90% pure ARCM population. Phase contrast image of isolated, primary ARCMs 24 h after binding to laminin demonstrate that ARCM differentiated morphology is maintained as indicated by the rod-shaped, branched striated cell examined at 40X magnification.

Figure 2. Cyclic mechanical stretch activates the NFkB, MAPK/ERK1/2 and PI3K pathways in ARCMs.

A) Isolated ARCMs attached to laminin were subjected to 24 h of cyclic mechanical stretch (10% stretch). Non-stretched control (0% stretch) ARCMs attached to laminin were incubated under identical conditions. Cells were then lysed, fractionated into cytosolic (C) and nuclear (N) fractions and immunoblotted with an anti-NFkB p65 antibody. In stretched ARCMs, the nuclear fraction contained significantly more of the NFkB p65 subunit compared to the non-stretched controls. Relative intensity of nuclear to cytoplasmic (N/C) fraction was determined. An antibody to the anti-TATA-binding protein (TBP) was used to determine fractionation efficiency. Cyclic mechanical stretch for 24 h induced a significant increase in MAPK/ERK1/2 and PI3K activity in isolated ARCMs attached to laminin. Isolated ARCMs were allowed to attach to laminin for 24 h and subsequently subjected to 24 h of cyclic mechanical stretch (10% stretch) or no stretch (0% stretch). Cells were lysed and immunoblotted for (B) ERK1/2 activity using an antibody against phospho-ERK1/2 (pERK) or a total ERK1/2 antibody (tERK) or for (C) PI3K activity via an antibody against phospho-AKT (pAKT) or total AKT (tAKT). Relative intensity of pERK/tERK or pAKT/tAKT was determined. *P<0.05; ***P<0.001; One-Way ANOVA with a Bonferroni post-test was used to determine the statistical significance of data. The values represent the average of three independent experiments.

Figure 3. Cyclic mechanical stretch-induced VEGF secretion is independent of the MAPK/ERK1/2 or PI3K pathways.

Inhibition of either the MAPK/ERK1/2 or PI3K signaling pathways in isolated ARCMs did not block VEGF secretion induced by cyclic mechanical stretch. Isolated ARCMs attached to laminin for 24 h were treated with the ERK1/2 kinase inhibitor (U0126; 0.5 µM) or DMSO-carrier control and subjected to (A) 24 h or 48 h of cyclic mechanical stretch (10% stretch) or no stretch (0% stretch). ELISA was used to determine the amount of VEGF secreted into the media. (B) The ERK1/2 kinase inhibitor was active in isolated primary ARCMs as pERK1/2 levels decreased in cells treated with the ERK1/2 kinase inhibitor (U0126; 0.5 µM) compared to the DMSO-carrier control. Stretch-activated VEGF secretion is independent of PI3K. ARCMs attached to laminin for 24 h were treated with the PI3K inhibitor (LY294002; 0.5 µM) or DMSO-carrier control and subjected to 24 h or 48 h (C) of cyclic mechanical stretch (10% stretch) or no stretch (0% stretch). ELISA was used to determine the amount of VEGF secreted into the media. (D) The PI3K inhibitor (LY294002; 0.5 µM) was active in isolated ARCMs as pAKT levels decreased in cells treated with the PI3K inhibitor (LY294002; 0.5 µM) compared to the DMSO-carrier control. Relative intensity of phospho-ERK1/2/total ERK1/2 (pERK/tERK) or phospho-AKT/total AKT (pAKT/tAKT) was determined. (E) ELISA was used to determine the amount of BNP secreted into the media. Treatment with the either ERK1/2 (U0126) or PI3K (LY294002) inhibitors blocked BNP secretion at 24 hours compared to DMSO-carrier control. ***P<0.001; One-Way ANOVA with a Bonferroni post-test was used to determine the statistical significance of data. n.s. = not significant. The values represent the average of three independent experiments.

Figure 4. Inhibition of the NFkB pathway blocks cyclic mechanical stretch-induced VEGF secretion in a dose-dependent manner.

Primary ARCMs cultured on laminin for 24 h were incubated with the NFkB inhibitor (SN50; 15 µg/ml) or its inactive control peptide (SN50M; 15 µg/ml), and subjected to 24 h or 48 h of cyclic mechanical stretch (10% stretch). Non-stretched control cells (0% stretch) were incubated under identical conditions. ELISA was used to analyze VEGF levels secreted into the culture media at 24 h (A) or 48 h (B). Inhibition of the NFkB pathway blocks cyclic mechanical stretch-induced VEGF secretion in a dose-dependent manner. Isolated primary ARCMs cultured on laminin were treated with NFkB inhibitor (SN50; 15 µg/ml or 30 µg/ml) and stretched (10% stretch) for 24 h (C) or 48 h (D). ELISA was used to analyze the concentration of VEGF in the media. Isolated primary ARCMs treated with the NFkB inhibitor (SN50) or its inactive peptide control (SN50M) remained viable at 24 h (E) and 48 h (F) of culture as assessed by the MTT cell viability assay. Fold induction in relative mitochondrial activity represents the number of viable cells present after each treatment. (G) The NFkB peptide inhibitor (SN50) was active in isolated ARCMS as p65 levels were reduced in the nuclear fraction upon inhibitor treatment compared to peptide control. ***P<0.001; One-Way ANOVA with a Bonferroni post-test was used to determine the statistical significance of data. n.s. = not significant. The values represent the average of three independent experiments.

Figure 5. Expression of a dominant negative mutant IkBα blocks cyclic mechanical stretch-induced VEGF secretion in a dose-dependent manner.

Primary ARCMs transduced with a recombinant adenovirus encoding an IkBα dominant negative mutant (DN IkBα) were cultured on laminin for 24 h and then subjected to 24 h of cyclic mechanical stretch (10% stretch). Non-stretched control cells (0% stretch) were incubated under identical conditions. ELISA was used to determine VEGF levels secreted into the culture media at 24 h (A). (B) Expression of the IkBα dominant negative mutant blocks cyclic mechanical stretch-induced VEGF secretion in a dose-dependent manner. ELISA was used to analyze the concentration of VEGF in the media. (C) The IkBα dominant negative mutant was expressed in isolated ARCMs as IkBα levels were significantly increased in ARCMs expressing the mutant compared to control ARCMs. (D) The IkBα dominant negative mutant was active in isolated ARCMs as p65 levels were significantly decreased in the nucleus of ARCMs expressing the mutant compared to control ARCMs. ***P<0.001; One-Way ANOVA with a Bonferroni post-test was used to determine the statistical significance of data. n.s. = not significant. The values represent the average of three independent experiments. (E) Representative ChIP assay PCR showing hypertrophic stretch increases NFkB binding to the native VEGF promoter in ARCMs. Protein-DNA complexes were immunoprecipitated with an NFkB p65 antibody followed by DNA isolation and purification and PCR. Non-immunoprecipitated chromatin was used as an “input” control (-). ARMCs subjected to 24 h stretch (10%) resulted in increased binding of NFkB to the VEGF promoter over 0% stretch.

Figure 6. VEGF expression is enhanced in the myocardium and in cardiomyocytes upon pressure overload-induced hypertrophy in vivo.

Immunostaining for VEGF using a polyclonal anti-VEGF antibody in paraffin fixed heart tissue from a pressure-overload hypertrophy model or sham control at day 10. Box =  higher magnification inset. Representative of three independent experiments.