β‐Adrenergic Stimulation Induces Histone Deacetylase 5 (HDAC5) Nuclear Accumulation in Cardiomyocytes by B55α‐PP2A‐Mediated Dephosphorylation

Background Class IIa histone deacetylase (HDAC) isoforms such as HDAC5 are critical signal‐responsive repressors of maladaptive cardiomyocyte hypertrophy, through nuclear interactions with transcription factors including myocyte enhancer factor‐2. β‐Adrenoceptor (β‐AR) stimulation, a signal of fundamental importance in regulating cardiac function, has been proposed to induce both phosphorylation‐independent nuclear export and phosphorylation‐dependent nuclear accumulation of cardiomyocyte HDAC5. The relative importance of phosphorylation at Ser259/Ser498 versus Ser279 in HDAC5 regulation is also controversial. We aimed to determine the impact of β‐AR stimulation on the phosphorylation, localization, and function of cardiomyocyte HDAC5 and delineate underlying molecular mechanisms. Methods and Results A novel 3‐dimensional confocal microscopy method that objectively quantifies the whole‐cell nuclear/cytoplasmic distribution of green fluorescent protein tagged HDAC5 revealed the β‐AR agonist isoproterenol to induce β1‐AR‐mediated and protein kinase A‐dependent HDAC5 nuclear accumulation in adult rat cardiomyocytes, which was accompanied by dephosphorylation at Ser259/279/498. Mutation of Ser259/Ser498 to Ala promoted HDAC5 nuclear accumulation and myocyte enhancer factor‐2 inhibition, whereas Ser279 ablation had no such effect and did not block isoproterenol‐induced nuclear accumulation. Inhibition of the Ser/Thr phosphatase PP2A blocked isoproterenol‐induced HDAC5 dephosphorylation. Co‐immunoprecipitation revealed a specific interaction of HDAC5 with the PP2A targeting subunit B55α, as well as catalytic and scaffolding subunits, which increased >3‐fold with isoproterenol. Knockdown of B55α in neonatal cardiomyocytes attenuated isoproterenol‐induced HDAC5 dephosphorylation. Conclusions β‐AR stimulation induces HDAC5 nuclear accumulation in cardiomyocytes by a mechanism that is protein kinase A‐dependent but requires B55α‐PP2A‐mediated dephosphorylation of Ser259/Ser498 rather than protein kinase A‐mediated phosphorylation of Ser279.


MEF2 reporter assay
ARVM expressing the GFP-HDAC5 variants or GFP alone and transduced with the MEF2-luciferase reporter (see above) were washed with cold PBS and lysed for 3 minutes on ice in PBS containing 1% Triton X-100, EDTA-free protease inhibitor cocktail and phosphatase inhibitor cocktail 3. The protein concentration of cell lysates was determined by Bradford assay using Protein Assay Dye Reagent (Bio-Rad, 500-0006). 25 µg total protein was incubated with 70 µL ONE-Glo luciferase substrate (Promega, E6110) and the luminescence measured using a GloMax 20/20 Luminometer (Promega) after 10 minutes. Luminescence measurements were normalized to an internal control (ARVM expressing the MEF2-luciferase reporter in the absence of GFP-HDAC5 or GFP) to allow data to be pooled from two independent experiments. Lysates were run on SDS-PAGE gels and the resulting Western blots probed with an anti-GFP antibody to verify that the GFP-HDAC5 variants were expressed to similar levels.

Subcellular fractionation
ARVM were fractionated using a method adapted from Snabaitis et al. 3 Briefly, ARVM were lysed in lysis buffer (50 mM Tris-HCl pH 7.5, 5 mM EGTA, 2 mM EDTA, 100 mM NaF, 1% Triton X-100, EDTA-free protease inhibitor tablet) on ice for 5 minutes then centrifuged at 14,000 g at 4°C for 30 minutes. The supernatant (containing cytosolic proteins such as GAPDH; "soluble fraction") was collected, and the pellet (containing nuclear proteins such as histone 2B; "insoluble fraction") was resuspended in Laemmli buffer. Samples were resolved on 4-20% SDS-PAGE gels for subsequent Western blotting.

Co-immunoprecipitation of GFP-HDAC5 with PP2A subunits
ARVM were washed in cold PBS and lysed in a high salt lysis buffer (20 mM HEPES-KOH pH 7.4, 250 mM NaCl, 0.11 M KOAc, 2 mM MgCl 2 , 1 µM ZnCl 2 , 1 µM CaCl 2 , 0.5% Triton X-100, 0.1% Tween-20, 500 units/mL Benzonase nuclease (Sigma E1014), EDTA-free protease inhibitor tablet, phosphatase inhibitor cocktail 3) on ice for 3 minutes, then at room temperature for 10 minutes to activate the DNase, as previously described. 4 Cells were scraped and samples vortexed before being incubated on ice for a further 10 minutes. Whole cell lysates were centrifuged at 8,000 g at 4°C for 10 minutes, and the NaCl concentration of the resulting supernatant adjusted to 150 mM with NaCl-free lysis buffer. 750 µg of each sample was incubated with uncoupled agarose beads (Chromotek bab-20) for 1 hour at 4°C. The beads were pelleted and the cleared supernatants immunoprecipitated at 4°C overnight using a GFP-Trap A kit (Chromotek gta-20). The next day, immunoprecipitates were washed and resuspended in Laemmli buffer for SDS-PAGE, as previously described. 5

Detection of endogenous HDAC5-B55α complex
NRVM were isolated from 1-2 day old Sprague Dawley rat pups as previously described 6 and plated in DMEM containing 10% foetal bovine serum and penicillin/streptomycin at a density of 3.5x10 6 cells per 10 cm cell culture dish. The day after plating, the media was replaced with maintenance media (4:1 DMEM:M199) and the cells cultured for a further 24 hours prior to treatment with 1 µM ISO or vehicle for 60 minutes. Cells were washed in cold PBS and lysed in lysis buffer (30 mM HEPES, 150 mM NaCl, 2 mM MgCl 2 , 2% Triton X-100, 50 mM NaF, 0.2 mM Na 3 VO 4 , 1 mM PMSF, 5 µM Pepstatin A) on ice for 10 minutes. Lysates were centrifuged at 8,000 g at 4°C for 10 minutes and 500 µg of the resulting supernatant incubated with Protein A Sepharose CL-4B beads (GE Healthcare, 17-0780-01) for 1 hour at 4°C. The beads were pelleted and the cleared supernatants incubated with HDAC5 antibody (Cell Signaling 20458; 1:70) or water (no antibody control) with gentle rocking at 4°C overnight. The next day, Protein A beads were added and the samples rocked at 4°C for 1 hour. The beads were washed four times in lysis buffer and then resuspended in Laemmli buffer for SDS-PAGE.

siRNA knockdown of B55α in NRVM
NRVM were isolated and plated in 6-well cell culture plates at a density of 350,000 cells/well as described above. The day after plating, NRVM were transduced with WT GFP-HDAC5 adenovirus and incubated in serum-free DMEM overnight. NRVM were transfected with 10 nM PPP2R2A siRNA (Dharmacon, L-047957-00-0005) or a non-targeting siRNA pool (Dharmacon, D-001810-10-05) using Lipofectamine RNAiMAX Transfection Reagent (Invitrogen, 13778-150), according to the manufacturer's instructions. Transfected cells were maintained in serum-free DMEM at 37°C, 5% CO 2 for 48 hours prior to stimulation with 1 µM ISO or vehicle for 60 minutes and subsequent lysis for Western blotting.

Quantification of GFP-HDAC5 nucleo-cytoplasmic shuttling
ARVM were isolated, plated in laminated 35 mm imaging dishes (Ibidi, 81156) and transduced with adenoviruses as described above. The day after plating, ARVM were incubated with 0.1 µM Cell Tracker Orange CMRA Dye (Molecular Probes, C34551) for 15 minutes at 37°C, 5% CO 2 and then incubated in fresh modified M199 media overnight. On the day of imaging, cells were imaged using a Nikon Ti-E (inverted) microscope equipped with a Yokogawa spinning disk and a Neo 5.5 sCMOS camera (Andor). A 60x/1.40 NA Plan Apo λ oil objective was used. Images were acquired using NIS Elements AR 4.2 software. Cells were maintained at 37°C, 5% CO 2 throughout the experiment via a CO 2 chamber and a temperature-regulated Perspex box which housed the microscope stage and turret.
The Cell Tracker signal was used to select ARVM for imaging to avoid unnecessary excitation of the GFP fluorophores and to prevent possible bias during the selection of cells based on the basal distribution of GFP-HDAC5. ARVM were not selected for imaging if they satisfied one or more of the following exclusion criteria: 1) cell is not adhered to the bottom of the dish; 2) cell is twitching/moving; 3) cell is balled up or is starting to ball up at one or both ends; 4) cell is overlapping another cell or is close to floating cells (as this interfered with thresholding of the Cell Tracker Orange signal for generation of a binary layer corresponding to the whole cell). After the XY coordinates and perfect focusing system (PFS) settings for 12-15 cells had been set, z-stacks spanning 36 µm (1.5 µm steps) were acquired using the 488 nm laser line to capture the baseline GFP-HDAC5 signal for each cell. Cells were treated with vehicle control or 10 nM ISO and GFP z-stacks repeated 15 minutes post-treatment. Thirty-five minutes post-treatment, 1 µM DRAQ5 (Biostatus, DR50050) was added to the cells to label the nuclei. Ten minutes after the addition of DRAQ5 (i.e. 45 minutes after the addition of ISO/vehicle), each cell was imaged using the 488 nm, 561 nm and 640 nm laser lines to excite the GFP, Cell Tracker and DRAQ5 fluorophores, respectively.
All imaged cells were quantified using NIS Elements AR 4.2 software unless they met one of the following exclusion criteria: 1) cell does not express GFP-HDAC5, 2) z-stack does not cover entire cell volume at one or more time points; 3) cell is rotated in XY plane in one or more time points (rotation is complicated to correct in volume analysis); 4) cell shrinks over the time course of the experiment (as this interfered with the application of binary layers across time points); 5) cell disappears from field of view over the course of the experiment; 6) the GFP signal is saturated in one or more time points; 7) GFP-HDAC5 is clearly mislocalized at baseline (see Fig. S9). Using thresholding, the Cell Tracker and DRAQ5 signals from the final time point were used to generate binary volumes corresponding to the whole cell and nuclei, respectively. Subtraction of the nuclear binary volume from the whole cell binary volume generated a third binary volume corresponding to the cytoplasm. The binary volumes were then used to determine average GFP fluorescence intensities in the nuclear, cytoplasmic and whole cell volumes at each time point.

Quantification of fluorescent puncta
The number of puncta was quantified using Fiji imaging analysis software. 7 For each cell, the background was subtracted using a rolling ball radius of 50.0 pixels. A single z-section that approximately bisected at least one of the nuclei was selected from the baseline GFP z-stack and the threshold adjusted by altering the minimum intensity value until there was a light 'salt and pepper' effect within the cell (see Fig. S8A). The number of puncta in the resulting binary image was quantified by counting the number of 0.1-10 µm 2 objects using the 'Analyze Particles' function. Any artifacts (e.g. GFP fluorescence in the nuclei) were subtracted from the total object count. The same conditions were then used to quantify the number of puncta in the corresponding z-section from the 45 minute GFP z-stack. The entire analysis was performed blinded to treatment group and z-stacks from the 45 minute time point were not viewed by the user until the analysis was complete to eliminate possible bias in application of the threshold to the baseline zsection image. increased TnI phosphorylation, whereas equivalent concentrations of CPT had no effect. C, 10 nM ISO and 500 µM BNZ reduced the phosphorylation of HDAC5 at all three sites, whereas 500 µM CPT had no effect.  Representative Western blots and quantitative data from four independent experiments. In each experiment, the B55α signal in CON and ISO fractions was normalised to the signal in the respective inputs. No significant differences were detected by unpaired t-test. The GAPDH and histone 2B blots are the same as those that appear in Fig. 2D as the same Western blot was probed for B55α. Figure S7. ISO had no effect on S279 and S498 phosphorylation in NRVM. NRVM expressing GFP-HDAC5 were transfected with scrambled (Scr) or PPP2R2A (B55α) siRNAs for 48 hours prior to treatment with 1 µM ISO or CON for 60 minutes. Grouped data from four independent experiments. No significant differences were detected by two-way ANOVA.