Retinoic acid protects cardiomyocytes from high glucose-induced apoptosis through inhibition of NF-κB signaling pathway

Abstract

We have previously shown that retinoic acid (RA) has protective effects on high glucose (HG)-induced cardiomyocyte apoptosis. To further elucidate the molecular mechanisms of RA effects, we determined the interaction between nuclear factor (NF)-κB and RA signaling. HG induced a sustained phosphorylation of IKK/IκBα and transcriptional activation of NF-κB in cardiomyocytes. Activated NF-κB signaling has an important role in HG-induced cardiomyocyte apoptosis and gene expression of interleukin-6 (IL-6), tumor necrosis factor (TNF)-α, and monocyte chemoattractant protein-1 (MCP-1). All-trans RA (ATRA) and LGD1069, through activation of RAR/RXR-mediated signaling, inhibited the HG-mediated effects in cardiomyocytes. The inhibitory effect of RA on NF-κB activation was mediated through inhibition of IKK/IκBα phosphorylation. ATRA and LGD1069 treatment promoted protein phosphatase 2A (PP2A) activity, which was significantly suppressed by HG stimulation. The RA effects on IKK and IκBα were blocked by okadaic acid or silencing the expression of PP2Ac-subunit, indicating that the inhibitory effect of RA on NF-κB is regulated through activation of PP2A and subsequent dephosphorylation of IKK/IκBα. Moreover, ATRA and LGD1069 reversed the decreased PP2A activity and inhibited the activation of IKK/IκBα and gene expression of MCP-1, IL-6, and TNF-α in the hearts of Zucker diabetic fatty rats. In summary, our findings suggest that the suppressed activation of PP2A contributed to sustained activation of NF-κB in HG-stimulated cardiomyocytes; and that the protective effect of RA on hyperglycemia-induced cardiomyocyte apoptosis and inflammatory responses is partially regulated through activation of PP2A and suppression of NF-κB-mediated signaling and downstream targets.

Fig. 1. High glucose induces activation of NF-κB in cardiomyocytes

A. Neonatal cultured cardiomyocytes were exposed to high glucose (HG) for different time periods, and whole cell lysate, nuclear and cytosolic proteins were extracted separately, and the nuclear translocation of NF-κB determined by Western blotting, using anti-p65 antibody. Mannitol (M, 19.5mM) was used as an osmolarity control. Equal loading was assessed with an anti-histone and anti-actin antibody, respectively. B. Cardiomyocytes were exposed to HG for 4–24 h, in the presence or absence of JSH-23, and the NF-κB luciferase reporter assay performed, as described in the Materials and Methods section. Results are expressed as the mean ± SEM, from 3 independent experiments. *p<0.05 vs control; #, p<0.05 vs HG (24 h). C. Cardiomyocytes were infected with AdIκBαDN or AdLacZ, and exposed to HG for 24 h, and the nuclear translocation of NF-κB was determined. D. After infection with AdIκBαDN or AdLacZ, HG-induced NF-κB promoter activity was analyzed. *p<0.05 vs control; #, p<0.05 vs HG.

Fig. 2. HG-induced cardiomyocyte apoptosis is regulated through NF-κB signaling

A. Cardiomyocytes were pretreated with 10–30 µM of JSH-23 or infected with 10 MOI of AdLacZ and AdIκBαDN, and exposed to HG for 24 h. The levels of cleaved (active) caspase-3 were determined by Western blot. B. Cardiomyocytes were pretreated with or without JSH-23 for 30 min, and exposed to HG for 24 h. Mannitol (M) was used as an osmolarity control. Cells were harvested and fragmented DNA isolated and separated by electrophoresis. C. Following the same treatment procedures as (A), the expression of Bcl-2 and Bax was determined. Blots were reprobed using anti-actin antibody, to verify equal loading. D. The intensity of the bands was analyzed by densitometry and the Bax/Bcl-2 ratio calculated. Data are expressed as the mean ± SEM, from 3 independent experiments. *p<0.05 vs control; #, p<0.05 vs HG or HG+LacZ.

Fig. 3. Activation of RAR and RXR signaling inhibits HG-induced activation of NF-κB

A. Cardiomyocytes were pretreated with or without 1µM of ATRA (RA) or LGD1069 (LGD) for 2 h, and exposed to HG for an additional 24 h. Mannitol (M) was used as an osmolarity control. Whole cell lysate (T), cytosolic (C) and nuclear (N) proteins were extracted and expression of NF-κB determined. Blots were reprobed using anti-actin and anti-histone antibodies to verify equal loading, respectively. B. Following the same treatment procedures as (A), the localization of NF-κB was determined by immunofluorescense staining, using anti-p65 NF-κB antibody (green). Cardiomyocyte nuclei were stained with DAPI (blue). Results shown are merged pictures, representative of three separate experiments. C. Cardiomyocytes were pretreated with or without different doses of ATRA (RA) or LGD1069 (LGD) for 2 h, exposed to HG for an additional 24 h, and an NF-κB luciferase reporter assay performed. The results are expressed as the mean ± SEM, from 3 independent experiments. *p<0.05 vs control; #, p<0.05 vs HG.

Fig. 4. RA inhibits HG-induced gene expression of pro-inflammatory cytokines

Cardiomyocytes were infected with 10 MOI of AdIκBαDN, AdLacZ (A), or pretreated with or without 1 and 5 µM of ATRA (RA1, RA5), LGD1069 (LGD1, LGD5) or 30 µM of JSH-23 (JSH) (B), and exposed to HG for 24 h. Gene expression of IL-6, TNF-α and MCP-1 was determined by real-time RT-PCR. The mRNA levels were normalized to GAPDH. Data (mean ± SEM, n=3) are expressed as a relative value compared to control. *, p<0.05, vs control; #, p<0.05, vs HG or HG + AdLacZ.

Fig. 5. RA inhibits HG-induced phosphorylation of IKK and IκBα

A. Cardiomyocytes were exposed to HG for different time periods (0–24 h), and phosphorylation of IKK and IκBα determined by Western blot, using specific anti-phospho-IKKα/β and anti-phospho-IκBα antibody, respectively. Blots were reprobed with anti-IKK and anti-IκBα antibodies, to determine the expression of IKK and IκBα, respectively. Equal loading was assessed using an anti-actin antibody. B. Cardiomyocytes were pretreated with different doses (0.01–10 µM) of ATRA (RA) or LGD1069 (LGD) for 2 h, and exposed to HG for an additional 4 h, and the phosphorylation and expression of IKK and IκBα determined. C. Densitometric quantification of p-IKK, p-IκBα and total IκBα protein bands, normalized to actin. Results are expressed as the mean ± SEM, from 3 independent experiments. *p<0.05 vs control; #, p<0.05 vs HG.

Fig. 6. Role of PP2A in HG-induced phosphorylation of IKK and IκBα

A. Cardiomyocytes were exposed to HG for different time periods (0–24 h), and tyrosine phosphorylation of PP2Ac determined, using specific anti-phospho-PP2Ac antibody. Equal loading was determined using anti-PP2Ac antibody. B. Cardiomyocytes were pretreated with okadaic acid (OA, 5–20 nM) for 1 h, exposed to HG for an additional 4 h, and phosphorylation and expression of IKK, IκBα and PP2Ac determined. Equal loading was assessed using an anti-actin antibody. C. Cardiomyocytes were pretreated with okadaic acid (OA, 5, 10 and 20 nM) for 1 h, exposed to HG for an additional 24 h, and NF-κB luciferase activity was determined. *, p<0.05 vs control. D. The expression of PP2Ac was silenced by transfecting cardiomyocytes with PP2Ac siRNA. The expression of PP2Ac (left panel) and phosphorylation of PP2Ac (right panel) was determined. Blots were reprobed with anti-actin antibody to verify equal loading. E. Cardiomyocytes were exposed to HG for different time periods, with or without silencing the expression of PP2Ac, and phosphorylation and expression of IKK and IκBα determined and quantified. The results are expressed as the mean ± SEM, from 3 independent experiments. *p<0.05 vs control.

Fig. 7. Role of PP2A in the RA-mediated inhibitory effect on IKK and IκBα

A. Cardiomyocytes were pretreated with 0.1–1 µM of ATRA (RA) or LGD1069 (LGD), exposed to HG for an additional 1 h, and the tyrosine phosphorylation of PP2Ac determined. Mannitol (M) was used as an osmolarity control. B. Cardiomyocytes were pretreated with 10 nM of okadaic acid for 1 h, treated with 1 µM of ATRA (RA) or LGD1069 (LGD) for 2h, and then exposed to HG for an additional 4 h. The phosphorylation and expression of IKK and IκBα was determined. Equal loading was determined using anti-actin antibody. C. After pretreatment with okadaic acid, ATRA and LGD1069, as noted in B, cardiomyocytes were exposed to HG for 24 h, and NF-κB Luciferase promoter activity determined. *p<0.05 vs control; #, p<0.05 vs HG; §, p<0.05 vs HG+RA, †, p<0.05 vs HG+LGD. D & E. Scrambled (Scram) or PP2Ac siRNA (PP2Ac) transfected cardiomyocytes were pretreated with 1µm of ATRA (RA) or LGD1069 (LGD), and exposed to HG for an additional 4 h (D) and 24 h (E), and phosphorylation and expression of IKK and IκBα (D) and NF-κB promoter activity (E) determined. Results are expressed as the mean ± SEM, from 3 independent experiments. *p<0.05 vs control; #, p<0.05 vs HG; §, p<0.05 vs HG+RA, †, p<0.05 vs HG+LGD.

Fig. 8. Effect of LGD1069 on NF-κB signaling in diabetic hearts from ZDF rats

A. Lean and ZDF rats, treated with or without ATRA (RA) orLGD1069 (LGD) for 2 weeks. Left ventricles were isolated and myocardium lysates prepared. Phosphorylation and expression of IKK, IκBα and PP2Ac were determined by Western blot. B. Cytosol and nuclear protein was prepared from left ventricles and the expression of NF-κB in cytosol (Cyto) and nuclear (Nu) fractions was determined by Western blot. Equal loading was verified by actin (cytosol) and histone (nuclear). C. Total RNA was isolated from left ventricles and gene expression of IL-6, TNF-α and MCP-1 determined by real-time RT-PCR. The mRNA levels were normalized to GAPDH. Data (mean± SEM, n=6) are expressed as a relative value, compared to Lean control. *, p<0.05, vs Lean; #, p<0.05, vs ZDF.