CaMKII and PKA-dependent phosphorylation co-regulate nuclear localization of HDAC4 in adult cardiomyocytes

Figures

• 

  Fig. 1. Baseline HDAC4 localization is governed by CaMKII. a Representative images of a live adult rabbit cardiomyocyte infected with HDAC4-GFP (top left) and the endogenous localization of HDAC4 in rabbit cardiomyocytes, detected by ICC (top right). Nuclear localization was confirmed by DAPI staining. Bottom panels represent zoom-in of corresponding cell nuclei. b Average values of HDAC4 nucleo-to-cytoplasmic fluorescence ratio FNuc/FCyto in the absence or presence of specific inhibitors of cellular kinases or phosphatases. Cells were incubated with indicated drugs for 1 h at 37 °C. c Average values of HDAC4 FNuc/FCyto ratio in WT and CAMKIIδ knockout (CAMKIIδ-KO) mice. ***p < 0.0001

• 

  Fig. 2. In rabbit ventricular myocytes, CaMKII activity promotes nuclear depletion of HDAC4. a Average values of HDAC4 nucleo-to-cytoplasmic fluorescence ratio in quiescent cells incubated in Normal Tyrode solution containing 0, 0.5, 1 and 2 mM Ca2+. b Average values of HDAC4 nucleo-to-cytoplasmic fluorescence ratio (FNuc/FCyto) in cells overexpressing calmodulin (CaM) with and without preincubation with KN93. c Average values of HDAC4 nucleo-to-cytoplasmic fluorescence ratio in cells treated with 100 nM AngII for 30 min in the presence or absence of KN93. d Time course of wild type (grey) or R601F (red) HDAC4-GFP translocation measured as FNuc/FCyto in response to low (light) or high (dark) pacing frequency (n = 20–24 cells per group). e Representative confocal images of cultured adult cardiomyocyte infected with wild type or R601F HDAC4-GFP, paced at 1.5 Hz for 0, 15 or 30 min. f Average values of HDAC4 FNuc/FCyto ratio in cells overexpressing wild type or R601F HDAC4-GFP with or without preincubation with KN93. g Average values of HDAC5 FNuc/FCyto in cells treated with 100 nM AngII for 30 min in the absence or presence of KN93. ***p < 0.0001, **p < 0.001, *p < 0.01

• 

  Fig. 3. β-AR activation in rabbit myocytes induces nuclear accumulation of HDAC4 and requires S265/266. a HDAC4 sequence with selected domain and residues indicated. NLS = nuclear localization sequence, NES = nuclear export signal. b Average values of HDAC4 nucleo-to-cytoplasmic fluorescence ratio of endogenous HDAC4 or cells infected with HDAC4-GFP (polled data) and treated for 30–60 min with either 1 μM Iso or 10 μM forskolin at 37 °C. c Average values of HDAC4 nucleo-to-cytoplasmic fluorescence ratio of cells infected with HDAC4-S265/266A-GFP and treated as in (b). d Dose-dependent accumulation of HDAC4 in the nucleus upon Iso treatment. Each point represents an n = 22–26. e Time course of HDAC4 redistribution in response to Iso (black) or forskolin (grey). Each point represents an n = 20–25. Representative example (f) and quantification (g) of immunoprecipitation with a 14–3-3 antibody and a GFP antibody to detect the HDAC4–14–3-3 complex with or without forskolin treatment. ***p < 0.0001

• 

  Fig. 4. Iso- or forskolin-induced HDAC4 nuclear accumulation is PKA-dependent. Average values of HDAC4 FNuc/FCyto under PKA signaling agonists in the absence or presence of specific PKA inhibitor H-89 (a) or PKI (b). Cells were incubated with indicated drugs for 30 min at 37 °C. c Average values of HDAC4 FNuc/FCyto ratio under PKA signaling agonists with or without 30 min preincubation with specific CaMKII inhibitor KN93. d Average values of HDAC4 FNuc/FCyto under PKA signaling agonists in the absence or presence of H-89 in WT and CAMKIIδ knockout (KO) mice. ***p < 0.0001 and **p < 0.001 vs. baseline; #p < 0.01 vs. WT

• 

  Fig. 5. PKA-dependent nuclear accumulation of HDAC4 is dominant over CaMKII-dependent cytosolic accumulation. a Average values of HDAC4 FNuc/FCyto in control rabbit myocytes and myocytes after overexpression of CaMKIIδC. b Average values of HDAC4 FNuc/FCyto in the absence or presence of specific agonists of PKA and CaMKII signaling. Myocytes were incubated with indicated drugs for 30–60 min at 37 °C. c Average values of HDAC4 FNuc/FCyto in quiescent cells incubated for 1 h in Normal Tyrode solution containing 0, 2 and 4 mM [Ca2+]° in the presence or absence of Iso. d Time course of HDAC4 redistribution in response to forskolin (dark grey), LMB (light grey) or both (black). Images were taken every 20 min to assess maximal nuclear accumulation as well as the time to reach half-max. Each point represents an n = 22–27. ***p < 0.0001, **p < 0.001

• 

  Fig. 6. PKA and CaMKII have different time course of action on HDAC4 redistribution. a Original recordings of cytoplasmic [Ca2+] transients in isolated ventricular rabbit cardiomyocytes during increases of stimulation frequency from 0.2 to 1.5 Hz at baseline (grey) and in the presence of 100 nM Iso (blue). b Corresponding average values of diastolic [Ca2+] (left) and amplitude (right) of cytoplasmic [Ca2+] transients. c Time course of HDAC4-GFP translocation measured as FNuc/FCyto in response to pacing frequency at baseline (grey) and in the presence of 100 nM Iso (blue). b, c Each point represents an n = 20–24. d Time course of HDAC4-GFP translocation measured as FNuc/FCyto in response to pacing frequency at baseline (grey) and in the presence of 100 nM Iso (red). Each point represents an n = 8–10. ***p < 0.0001, **p < 0.001

• 

  Fig. 7. Co-regulation of HDAC4 localization by PKA and CaMKII is altered in heart failure. a Representative images of the endogenous localization of HDAC4 in failing rabbit cardiomyocytes, detected by ICC (top). Nuclear localization was confirmed by DAPI staining. Bottom panels represent zoom-in of corresponding cell nuclei. b Average values of HDAC4 nucleo-to-cytoplasmic fluorescence ratio (FNuc/FCyto) in the absence or presence of specific inhibitors/activators of cellular kinases without (gray) or with (blue) 10 min electrical stimulation of 0.2 or 1.5 Hz. NF—non-failing. Cells were incubated with indicated drugs for 1 h at 37 °C. c Representative images of the endogenous localization of HDAC4 in cardiomyocytes isolated from wild type (WT; top) or CaMKIIδC transgenic (CaMKIIδC-TG; bottom) mice, detected by ICC. d Average values of HDAC4 FNuc/FCyto ratio in WT and CAMKIIδC-TG mice in the absence or presence of specific inhibitors/activators of protein kinase A without (gray) or with (blue) 10 min electrical stimulation of 0.2 or 1.5 Hz. Cells were incubated with indicated drugs for 1 h at 37 °C. ***p < 0.0001. ***p < 0.001, *p < 0.05

• 

  Fig. 8. HDAC4 localization in human ventricular myocytes is regulated in similar manner. a Representative image of endogenous HDAC4 localization in cardiomyocytes isolated from non-failing human heart, detected by ICC. Bottom panel represent zoom-in of corresponding cell nuclei. b Average values of HDAC4 FNuc/FCyto ratio in non-failing (NF) human myocytes in the absence or presence of specific agonists or antagonists of PKA or CaMKII signaling. c Representative image of endogenous HDAC4 localization in cardiomyocytes isolated from failing human heart, detected by ICC. Bottom panel represent zoom-in of corresponding cell nuclei. d Average values of HDAC4 FNuc/FCyto ratio in failing human myocytes in the absence or presence of specific agonists or antagonists of PKA or CaMKII signaling. Cells were incubated with indicated drugs for 1 h at 37 °C. ***p < 0.0001, *p < 0.01

• 

  Fig. 9. HDAC4 nucleo-cytoplasmic shuttling is differentially controlled by PKA and CaMKII. Left Panel: Cytosolic, dephosphorylated HDAC4 has an NLS recognized by importin-α, leading to nuclear import of HDAC4 via the nuclear pore complex. Once in the nucleus, HDAC4 can complex with MEF2 allowing for both the deacetylation of histone tails within nucleosomes, as well as preventing HATs (like p300) from complexing with MEF2. The HDAC4-MEF2 complex plays a role in preventing reactivation of the fetal gene program by maintaining the relevant DNA in a condensed (and therefore inaccessible) chromatin conformation. Middle Panel: HDAC4 can be phosphorylated by kinases including CaMKII and PKA. CaMKII contributes to the phosphorylation of Ser 467 and 632, whereupon chaperone protein 14-3-3 recognizes the phosphoserines and binds to them. This 14-3-3 binding masks the NLS and helps expose the NES, which can bind the exportin CRM-1, leading to nuclear export. Once in the cytosol, 14-3-3 and CRM-1 can dissociate leaving phosphorylated HDAC4 which is not available to be imported unless dephosphorylated by phosphatases like PP2A. Additionally, cytosolic, dephosphorylated HDAC4 can also be…