| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1,
,,2
* The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA;
The Penn Cardiovascular Institute, Philadelphia, Pennsylvania, USA; and
The University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, USA
2Correspondence: The Children’s Hospital of Philadelphia, 34th St. and Civic Center Blvd., Ste. 8527 Main, Philadelphia, PA 19104, USA. E-mail: pgruber{at}mail.med.upenn.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, cell death, and vascular permeability in vivo and in vitro, thus providing potential mechanisms to explain reduced vascular leak and myocardial injury. In vitro, siRNA knockdown experiments implicate HDAC4 as a mediator of the effects in ischemic cardiac myocytes. These results demonstrate that HDAC inhibitors alter the response to ischemic injury in the heart and reduce infarct size, suggesting novel therapeutic approaches for acute coronary syndromes.—Granger, A., Abdullah, I., Huebner, F., Stout, A., Wang, T., Huebner, T., Epstein, J. A., Gruber, P. J. Histone deacetylase inhibition reduces myocardial ischemia-reperfusion injury in mice.
Key Words: heart • hypoxia inducible factor 1 • cardiomyocytes
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. With optimal medical management, the 5-yr survival of patients diagnosed with chronic postinfarction heart failure is only 50%. Some interventions have focused on myocardial protection during ischemic episodes, including the use of glucose and insulin to provide necessary metabolic substrates to ischemic tissues. However, new understandings of ischemia-induced epigenetic changes suggest additional targeted approaches. For example, in the brain, ischemia promotes deacetylation of core histones and dimethylation of histone H3 at lysine-9 (H3-K9) over the mu opioid receptor-1 promoter, a signature of epigenetic gene silencing (3)
. In addition, in the isolated kidney, ischemia-reperfusion (I/R) results in a significant change in the ratio of methylated to unmethylated cytosines at the C3 promoter (4)
.
The regulation of gene expression by histone modifications is well established (5
6
7
8)
. Acetylation of core histones H3 and H4 plays a prominent role in the regulation of transcriptional activity. The dynamics of histone acetylation result from a balance of two biochemical activities: histone acetylation by histone acetyl transferases (HATs) and histone deacetylation by histone deacetylases (HDACs). These covalent modifications provide marks for interactions with coactivator and repressor proteins to regulate transcription (9
10
11)
. In addition to histones, growing evidence suggests that HDACs and HATs have additional targets, including transcription factors (12)
.
Specific HDAC family members play essential roles in cardiac physiology. Loss of Hdac2 prevents reexpression of fetal genes and attenuates cardiac hypertrophy in hearts exposed to hypertrophic stimuli (13)
. HDAC4 and HDAC5 inhibit myogenesis by directly binding and repressing myogenic transcription factors of the myocyte enhancer factor-2 (MEF-2) family (14
, 15)
. Mice lacking either HDAC5 or HDAC9 display increased sensitivity to cardiac stress signals and cardiac hypertrophy, with double-inactivation leading to embryonic lethality or early postnatal death from multiple cardiac abnormalities (16
, 17)
.
HDAC inhibition has recently been shown to reduce ischemic injury in the brain (18
, 19)
. For example, in a middle cerebral artery occlusion model of brain ischemia, administration of the HDAC inhibitor (HDACI) trichostatin A (TSA) resulted in a 48% decrease in injury volume in treated animals compared to nontreated controls. In addition, HDACIs have recently been shown to elicit a preconditioning effect in an isolated perfused heart model associated with phosphorylation of p38, as well as with improved cardiac remodeling in a chronic LAD ligation rat model (20
, 21)
. Given the documented role that HDACs play in cardiac physiology, we sought to determine the effects of HDAC inhibition on acute I/R injury in the heart, identify the HDAC isoforms involved, and uncover the transcriptional programs that mediate the effect.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Reagents
The antibody to HIF1A was kindly provided by C. Simon (University of Pennsylvania, Philadelphia, PA, USA) The antibody to myosin light chain-2 ventricular (MLC-2v), was kindly provided by Steve Kubalak (Medical University of South Carolina, Charleston, SC, USA). Antibodies to acetylated histones H3 and H4 and β-tubulin were from Upstate (Charlottesville, VA, USA). Antibodies to histone H3, TFII-I, and cleaved caspase 3 were from Cell Signaling Technology (Danvers, MA, USA). Nullscript, Scriptaid, and TSA were from Biomol. HDAC null mice were a generous gift of Eric Olson (Department of Molecular Biology, University of Texas Southwestern, Dallas, TX, USA).
Isolation of mouse neonatal cardiac myocytes and cell culture conditions
Cardiomyocyte cultures were prepared from hearts of 1-day-old CD-1 mouse pups (Charles River, Wilmington, MA, USA). Ventricular tissues were predigested overnight in 2 x 10–5 mmol/L trypsin-Hank’s balanced salt solution without Ca2+ (HBSS)at 4°C under constant shaking. Cells were dissociated by 5 rounds of 10-min digestions with purified type II collagenase (Worthington, Lakewood, NJ, USA) in HBSS at 37°C. Fibroblasts were removed from the cell suspension by two rounds of 1 h differential plating at 37°C in Dulbecco modified Eagle medium (DMEM)/M199 (4:1 ratio) medium containing 10% horse serum, 5% fetal bovine serum, penicillin (100 U/ml), streptomycin (100 µg/ml), HEPES (25 mmol/L), and glutamine (2 mmol/L) in the presence of 5% CO2. Myocytes were cultured on fibronectin-coated tissue culture plates at an initial density of 105 cells/cm2. Medium was exchanged to a 1.5% reduced serum medium containing 1% horse serum and 0.5% fetal bovine serum after 36 h. Cells were pretreated or not for 1 h with TSA (100 nmol/L), Nullscript (6 nmol/L), or Scriptaid (6 nmol/L).
When indicated, hypoxic treatment was carried out in a calibrated incubator with 1% O2, 5% CO2, and 94% N2 for 5 h.
Cell viability assay
Murine neonatal cardiac myocytes were plated in Permanox Labtek chamber slides (Nunc, Rochester, NY, USA). After pretreatment with Nullscript or Scriptaid for 1 h under standard conditions, cells were incubated in normoxia or hypoxia for 5 h, and nonviable cell numbers were determined with ethidium homodimer-1 (EthD-1; 5 µmol/L) from the Live/Dead Viability/Cytotoxicity Kit (Molecular Probes/Invitrogen, Carlsbad, CA, USA), according to the manufacturer’s protocol. EthD-1 enters cells with damaged membranes and is excluded by the intact plasma membrane of live cells. After 30 min, cover glass was applied with Vectashield mounting solution containing DAPI (Vector Laboratories, Burlingame, CA, USA).
Microarray analysis
Neonatal mouse cardiac myocytes were cultured in the presence of Nullscript or Scriptaid under normoxic and hypoxic conditions for 4 h. Total RNA was isolated from the cells by using TRIzol reagent (Gibco/Life Technologies, Inc., Gaithersburg, MD, USA). RNA from cells cultured with Scriptaid was compared with mRNA from cells cultured with Nullscript, under normoxia or hypoxia, using the mouse MOE430A v2.0 array (Affymetrix, Santa Clara, CA, USA). A 2-fold change was considered a threshold for differences in gene expression with values of P < 0.05. Three microarrays were used for each condition. Affymetrix GCOS 1.4 was used to quantify probe intensities; default values provided by Affymetrix were applied to all analysis parameters. Border pixels were removed, and the average intensity of pixels within the 75th percentile was computed for each probe. The average of the lowest 2% of probe intensities occurring in each of 16 microarray sectors was set as background and subtracted from all features in that sector. Probe pairs were scored positive or negative for detection of the targeted sequence by comparing signals from the perfect match and mismatch probe features. The number of probe pairs meeting the default discrimination threshold (tau=0.015) was used to assign a call of absent, present, or marginal for each assayed gene, and a P value was calculated to reflect confidence in the detection call. Probe intensities were then normalized and summarized to a signal value for each transcript using the GC-RMA algorithm as implemented in ArrayAssist Lite v3.4 (Stratagene, La Jolla, CA, USA). Spotfire DecisionSite (Spotfire, Somerville, MA, USA) was used for all microarray analyses.
Quantitative reverse trascriptase-polymerase chain reaction (RT-PCR) analysis
Total RNA was extracted from isolated cardiac myocytes or whole hearts with TRIzol (Invitrogen). Reverse transcription was carried out with 1 µg total RNA using the Superscript First-Strand Synthesis System for RT-PCR (Invitrogen), according to the manufacturer’s instructions. Ten nanograms of cDNA was subjected to real-time PCR using Taqman gene expression assays (Applied Biosystems, Foster City, CA, USA) to assess the expression level of Vegfa (Mm00437304_m1), Hdac2 (Mm01193631_m1), Hdac3 (Mm01258403_gH), Hdac4 (Mm01299566_m1), Hdac5 (Mm00515917_m1), Hdac6 (Mm00515945_m1), Hdac9 (Mm00458456_m1), Egln1 (Mm00459770_m1), Egln2 (Mm00519067_m1), and Egln3 (Mm00472200_m1). β2-microglobulin (Mm00437762_m1) was used as an internal control.
siRNA transfection
siRNA (Ambion, Foster City, CA, USA) directed against Hdac2 (ID: 158016), Hdac3 (ID: 158916), Hdac4 (ID: 166635), Hdac5 (ID: 158919), Hdac6 (ID: 158922), Hdac8 (custom), Hdac9 (ID: 174326), or scrambled siRNA (ID: AM4611) were transfected with Lipofectamine 2000 (Invitrogen), according to the manufacturer’s instructions. After 48 h, knockdown efficiency was assessed by quantitative RT-PCR analysis, using the primers listed in the preceding section.
Immunoblot
Primary neonatal murine cardiac myocytes from harvest at P1 were lysed for 15 min at 4°C in a lysis buffer containing 25 mmol/L Tris, pH 7.5; 300 mmol/L NaCl; 1% Triton X-100; and protease inhibitor Complete (Roche, Palo Alto, CA, USA). Cellular debris was removed by centrifugation. Cytoplasmic and nuclear extracts were prepared using the NE-PER kit for nuclear fractionation (Pierce, Rockford, IL, USA), according to the manufacturer’s instructions. Protein concentration was determined by bicinchoninic acid (BCA) methods, and an equal amount (20 µg) was loaded on SDS NuPAGE 4–12% Bis-Tris gels (Invitrogen) and electrotransferred to polyvinylidene fluoride membranes (Amersham Pharmacia Biosciences, Piscataway, NJ, USA). Membranes were blocked with 5% nonfat dry milk in TBS and incubated with the primary antibody overnight at 4°C. Antigen-antibody complexes were visualized by chemiluminescence (Pierce) after incubating the membrane for 1 h with the appropriate secondary antibody conjugated to horseradish peroxidase (Amersham).
Immunohistochemistry
Murine neonatal cardiac myocytes were maintained at normoxia (21% O2) or hypoxia (1% O2) for 5 h. One group was pretreated with Nullscript or Scriptaid for 1 h. Cells were then fixed with 4% paraformaldehyde in PBS for 15 min and permeabilized with 0.2% Triton X-100 in PBS for 10 min. Subsequently, cells were preincubated with PBS containing 2% bovine serum albumin and incubated at 4°C overnight in a humidified chamber with rabbit antibodies directed against either cardiac MYL2 (MLC-2v, 1:200) or cleaved caspase 3 (1:100). After 5 washes, cells were incubated with 1:2000 Alexa Fluor 488 conjugated goat anti-rabbit immunoglobulin G (IgG) (Molecular Probes) and Phalloidin Alexa 568 (1:200) to visualize actin filaments. After 3 washes, cover glasses were applied with Vectashield mounting solution containing DAPI (Vector Laboratories) or TOPRO (Invitrogen) for confocal analysis on a Leica TPS confocal system (Leica Microsystems, Wetzlar, Germany).
Vascular permeability assay
Vascular permeability assay was performed as described previously, with some modifications (22)
. CD-1 mice were anesthetized with ketamine (100 µg/g, i.p.) plus xylazine (10 µg/g, i.p.). Evans blue (20 mg/g) was injected through the jugular vein over 10 s. Ten minutes after Evans blue injection, mice were sacrificed, and hearts were perfused with 5 ml PBS via the aorta. The left ventricle was harvested and dried in a SpeedVac (Thermo Scientific, Milford, MA, USA) for 4 h. Tissue was homogenized in 1 ml of formamide (Sigma, St. Louis, MO, USA), and the Evans blue was extracted at 70°C for 24 h. After centrifugation (13,000 rpm, 30 min, 4°C), 200 µl of the supernatant were analyzed by measuring absorbance at 620 nm.
Enzymatic activity assays
Fluorometric HDAC activity assay (Upstate) and colorimetric HAT activity assay (Biovision, Mountain View, CA, USA) were performed according to the manufacturer’s instructions. Results are reported as fluorescent units (FU).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) (23)
. After undergoing 45-minutes of LAD ischemia, followed by 48-hour reperfusion, control sham-operated or ischemic hearts were treated with either vector (DMSO) or the HDACI TSA (Fig. 1A
). Hearts were harvested 48 h after reperfusion, left ventricular (LV) protein extracts were prepared, and HDAC activity was measured via a fluorometric assay. LV HDAC activity nearly doubled (411±14 to 700 ±9 FU; P=0.0002) in response to I/R injury. When treated with TSA, HDAC activity decreased to baseline (403±88 FU; P=0.03) (Fig. 1B
). No significant changes in HAT activity were observed with I/R injury and HDACI treatment (Supplemental Fig. 1). Next, we asked whether myocardial infarct area could be reduced by treatment with HDACIs. We used an identical LAD I/R protocol: mice were sacrificed and hearts were excised and processed for TTC quantification of infarct area and fluorescent microsphere analysis of area at risk (Fig. 1C, D
). Controls presented consistent infarct sizes with respect to area at risk (36.0±4.3%), whereas the single-dose pretreatment protocol (T1) with the HDACI TSA resulted in a dramatic, 48.3% reduction in infarct size (18.6±3.6%; P=0.015). Notably, even treatment 45 min after the injury via the T2 protocol resulted in significantly smaller infarct sizes (16.0±4.3%; P=0.013), a reduction of 55.6%. However, treatment 12 h after injury with the T3 protocol resulted in no significant decrease in infarct size (33.0±5.3%; P=0.674), suggesting that there is a treatment window after which I/R injury becomes resistant to epigenetic manipulations (Fig. 1D
). In addition, treatment with the chemically distinct HDACI Scriptaid resulted in a nearly identical effect when compared to Nullscript, with a 46.8% reduction in infarct size (21.2±3.3%; P=0.035). These results strongly suggest that in murine models, HDACIs can reverse the induction of ischemia-induced HDAC activity in vivo and reduce myocardial infarct size by more than 50%. Therefore, we sought an in vitro system of ischemic cardiomyocyte injury to dissect the mechanisms that mediate the protective effect of HDACIs on myocardial ischemia.
|
Hypoxia induces HDAC activity in isolated cardiac myocytes
To determine whether hypoxia induced similar alterations of HDAC activity in vitro, neonatal mouse ventricular myocytes were harvested from postnatal day 1 (P1) CD-1 pups and cultured under normoxic (21% O2) conditions. Immunohistochemical staining with anti-MYL2 (MLC-2v) and phalloidin demonstrated normal sarcomeric structure and histological characteristics (Fig. 2A
). Our in vitro hypoxic protocol consisted of a 5-h period of hypoxia (1% O2) followed by 90 min of reoxygenation (21% O2) (Fig. 2B
). Cell lysates analyzed by Western blot with anti-acetylated histone H3 or H4 demonstrated a strong decrease in acetylated histones H3 and H4 with hypoxia that recovered on reoxygenation (Fig. 2C
). We hypothesized that this event might be caused by either an increase in histone deacetylase activity, a decrease in histone acetyl transferase (HAT) activity, or a combination of the two. TSA completely blocked the decrease in both histone H3 and H4 acetylation with hypoxia (Fig. 2D
). Treatment with other HDACIs (e.g., Scriptaid), but not controls (DMSO or Nullscript), had identical effects (data not shown). We also examined HDAC and HAT enzymatic activities of hypoxic myocytes identifying an increase in HDAC activity (Supplemental Fig. 2A), whereas there was no change in HAT activity compared to normoxic cells (Supplemental Fig. 2B). Thus, in our system, the primary mechanism of alterations in histone acetylation after ischemia in cardiac myocytes is due to an increase in HDAC activity.
|
Inhibition of HDACs promotes cell survival during hypoxia in isolated cardiac myocytes
Given the pleiotropic effects seen with HDACIs in noncardiac systems, we examined mechanisms that might modulate this effect, first determining whether HDACIs affected hypoxia-induced apoptotic cell death (8
, 11
, 24)
. Under hypoxic or normoxic conditions, we examined cardiac myocytes by immunohistochemistry for cleaved caspase 3 (Fig. 3A
). Consistent with other studies, we found cardiac myocytes resistant to hypoxia-induced apoptosis (25
, 26)
. However, despite the low level of apoptosis, HDACIs reduced hypoxic-induced apoptosis in cardiac myocytes by 59% (P=0.01) (Fig. 3B
).
|
Because only a low level of apoptotic activity was observed in our in vitro cardiac myocyte I/R system, we tested whether HDACIs were protective of nonapoptotic cell death. Using identical experimental conditions, we determined overall cell death by ethidium staining (Figs. 3C-E
). In contrast to minimal apoptosis activation, hypoxia induced a 5.5-fold increase in overall cell death in Nullscript-treated cells, whereas treatment with Scriptaid resulted in a hypoxia-induced increase of only 2.1-fold, a 62% reduction (P=0.01). It is worth noting that in normoxic conditions, HDAC inhibition produced an increase in cell death, consistent with a prior report in rat neonatal cardiac myocytes (1.5-fold, P=0.04) (27)
. These experiments demonstrate that environmental oxygen tension plays a critical role in the response of myocytes to HDACIs in that HDAC inhibition under normoxia induces cell death, whereas under hypoxia, they protect cardiac myocytes from cell death.
Inhibition of HDAC activity prevents hypoxia-induced increases in HIF1A and its target genes in isolated cardiac myocytes
Given the reduced damage of the heart in vivo and in vitro, we used microarray-based gene expression profiling to investigate the transcriptional consequences of HDAC inhibition in ischemic cardiac myocytes. Using our in vitro cardiac myocyte I/R protocol, we isolated and RNA purified cells, and gene expression was analyzed by Affymetrix Mouse Genome 430 2.0 Array, as described in detail in the Methods section (Fig. 4
and Supplemental Table 1). We performed a 2-way ANOVA yielding three false discovery rates and controlled P values, the first for each gene’s general responsiveness to hypoxia (comparing normoxia Nullscript [NN] to hypoxia Nullscript [HN]), the second for a gene’s general responsiveness to hypoxia with the HDACI Scriptaid treatment (comparing normoxia Scriptaid [NS] to hypoxia Scriptaid [HS]), and the third for a gene’s differential response to hypoxia depending on Scriptaid treatment (comparing hypoxia Nullscript [HN] to hypoxia Scriptaid [HS]). We identified genes that were significant for this third P value and further filtered them to yield a set that displayed a less than twofold change between NN and NS and a >2-fold change between NN and HN. Finally, we wanted to identify which genes were returned to their normoxic state and thus chose as a final criterion those probe sets with 
25% return toward NN values under the HS condition.
|
This analysis revealed a series of coregulated genes that describe the ability of HDACIs to generally revert a subset of hypoxia-induced genes back toward their prehypoxic state. A heat map describing the expression profiles of 4 conditions (NN, HN, NS, HS) is illustrated in Fig. 4A
. Here, 117 probe sets describe the set of commonly regulated genes that are not significantly changed by the administration of Scriptaid under normoxia (Fig. 4A
, columns 1 and 2), are induced by hypoxia (column 3), and are repressed back toward their prehypoxic state by the administration of the HDACI Scriptaid (column 4). Predictably, we found numerous cellular functions were affected, many of which have been previously implicated in HIF1A signaling. Gene profiles are reported for each individual gene in Fig. 4B
. Here, Vegfa (red) and Egln3 (green), known targets of HIF1A, were identified as two members of this commonly regulated gene set. Further analysis of this coregulated group of genes revealed that many were previously identified as regulated by HIF1A (Supplemental Table 1). However, there was also a set of coregulated genes lacking known association with HIF1A signaling, suggesting either a novel group of HIF1A regulated genes in cardiac myocytes or a group of genes regulated in a non-HIF1A-dependent manner. Visualization of the entire set of genes filtered for the response to hypoxia with and without Scriptaid treatment (Fig. 4C
) demonstrates a relatively small number of genes significantly activated by hypoxia and differentially regulated by HDACIs. Indeed, far more genes were differentially regulated by HDACIs alone (2562) than by hypoxia (375), demonstrating the complexity that underlies this response. Western blot analysis of hypoxic cardiac myocytes revealed a significant HIF1A protein accumulation in the nucleus of cardiac myocytes under hypoxic conditions that was prevented by Scriptaid (Fig. 4D
). Interestingly, pretreatment of hypoxic myocytes with the HDACI Scriptaid repressed this induction to baseline levels with no significant change at the mRNA levels (data not shown), extending observations in other cell types (28
, 29)
.
Microarray analysis revealed HDACI modification of both death pathways and HIF1A-dependent pathways. In addition, we were surprised to note that HDACI-mediated suppression of Vegfa, a well-characterized HIF1A target, correlated with cell survival. Using quantitative RT-PCR, we confirmed the microarray data for several genes, including Vegfa, Egln3, and Egln2 (Fig. 5
), and Glut1 (data not shown). We measured Vegfa expression in isolated cardiac myocytes in an in vitro I/R assay and found that Vegfa was indeed strongly induced with hypoxia, with maintenance of mRNA levels during the reoxygenation (Fig. 5A, B
). No significant changes were observed with a pretreatment with DMSO or Nullscript. As previously identified in our microarray experiments, this increase was prevented by the administration of either TSA (Fig. 5A
) or Scriptaid (Fig. 5B
). Progressive HDAC inhibition resulted in a dose-dependent reduction of Vegfa expression, an effect that was more pronounced under hypoxic conditions (Fig. 5C
). The same pattern was observed for Egln3, another HIF1A target gene responsible for HIF1A degradation (Fig. 5D
). To exclude the possibility of a general repressive effect of HDACIs on gene expression, we measured the expression of other members of the EglN family, Egln1 and Egln2 (Fig. 5E
). A similar repressive trend was observed for Egln1, consistent with previous studies identifying Egln1 as an HIF1A target. As expected, Scriptaid also prevented the hypoxic induction of Egln1. In contrast, Egln2 expression was unaffected by the HDACI treatment and/or the oxygen level. Together, these data suggest that HDAC activity is specifically required for positive transcriptional regulation under hypoxia, and HDACIs prevent this response in a number of genes by down-regulation of HIF1A.
|
Hdac4 is an important mediator of the hypoxic response in isolated cardiac myocytes
Because multiple Hdac isoforms play important roles in cardiac physiology, we examined the potential role of individual Hdac isoforms for their contribution to the hypoxic response (13
, 16
, 17)
. Of the Hdac isoforms expressed in the heart (Hdac2–6 and 9), we did not detect any significant changes in mRNA expression during hypoxia, either in the presence or absence of HDACI (Fig. 6A
). Given the strong role that Hdac9 plays in cardiac physiology, we examined the hypoxic induction of Vegfa and Egln3 mRNAs in cardiac myocytes isolated from Hdac9-deficient pups (P1). Hdac9 deficiency in null mice did not prevent hypoxia-induced expression of Vegfa and Egln3, suggesting that other Hdac isoforms mediate the response (data not shown).
|
RNA interference was used to deplete endogenous HDACs in neonatal cardiac myocytes, and knockdown efficiency was determined by quantitative RT-PCR with primers for individual Hdacs. Scrambled siRNA controls did not perturb the hypoxic response after 48 h of transfection. Individual, Hdac isoform-specific knockdown resulted in no changes in the expression of other HDAC isozymes (data not shown). Only individualknockdown of Hdac4 under hypoxic conditions fully recapitulated the effect of Scriptaid on the expression of Vegfa (Fig. 6B
) and Egln3 (Fig. 6C
). We also noticed a small, but significant, decrease of Egln3 expression with Hdac5 knockdown, whereas Vegfa expression remained unchanged, suggesting a potential role for HDAC5 in the hypoxic response. Taken together, these data suggest that in the context of our neonatal murine cardiomyocyte hypoxic assay system for the HIF1A targets Vegfa and Egln3, HDAC4 is the primary mediator of the HDACI-modulated repression of ischemic activation.
Single-dose i.p. administration of HDACI reduces I/R-induced Vegfa induction and vascular permeability in vivo
Although multiple mechanisms likely contribute to ischemic protection by HDACI in the intact heart, we asked whether there was a functional role for VEGFA in the process. Although VEGFA is an angiogenic factor that has been proposed as a treatment for myocardial infarction, its potent ability to stimulate vascular leak may have profound negative consequences (30)
. Thus, we hypothesized that a reduction in Vegfa activation may result in a decrease in vascular permeability, potentially reducing myocardial injury. We utilized our standard in vivo model of cardiac I/R to examine whether HDAC inhibition would be able to inhibit I/R-inducible Vegfa expression in the intact heart. Consistent with our in vitro experiments, Vegfa expression was significantly increased with I/R in vivo. Mirroring in vitro results, HDACIs completely blocked the ischemia-induced increase in Vegfa (Fig. 7A
). Hearts from mice undergoing the I/R protocol treated with either vehicle (DMSO) or HDACI (TSA) were infused with Evans Blue and analyzed for vascular permeability (31)
. Ischemia resulted in a 3-fold increase in vascular permeability, subsequently halved by treatment with HDACI (Fig. 7B
). These results strongly suggest that in murine models, HDACIs can 1) reverse the induction of ischemia-induced HDAC activity in vivo, 2) return ischemia-induced Vegfa induction to nonischemic levels, 3) prevent ischemia-induced vascular permeability, and 4) reduce myocardial infarct size by over 50%.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
HDACs play pleiotropic roles in transcriptional regulation and cellular physiology (8)
. Although HDACIs have been shown to be proapoptotic in a number of cell types, their effects are highly context dependent. A striking example of this functional duality comes from studies of erythropoiesis, in which HDAC inhibition simultaneously negatively regulates IL-3-mediated growth of one class of erythroid precursors, while playing an important role in EPO-mediated differentiation and survival of other erythroid precursors (24)
. Indeed, in rat neonatal cardiac myocytes, under normoxic culture conditions, HDACIs are mildly proapoptotic (27)
. Yet, their role in I/R injury in the intact heart was previously unexplored.
One contributing mechanism for the profound ischemic protection with HDACIs is a reduction in ischemia-induced myocardial cell death. Importantly, environmental oxygen tension was critical to the response of the cell to HDACIs, such that HDACIs were protective of myocyte cell death only under hypoxic conditions. Our results support a significant reduction in nonapoptotic cell death by HDACIs, and to a lesser extent in apoptotic cell death. Despite the fact that cardiac myocytes are largely resistant to apoptosis, gene profiling demonstrated an HDACI-mediated reduction in Bnip3 and Bnip3l activation, suggesting a potential mechanism for a small reduction of apoptotic cell death. However, the majority of HDACI-mediated protection was due to a nonapoptotic mechanism, the mechanism of which remains a target for future study.
Both hypoxia and HDAC inhibition affect a vast array of pathways. Indeed, our comprehensive profiling of gene sets derived from normoxic and hypoxic cardiac myocytes treated with and without HDACI revealed important findings. First, under our selection criteria, far more genes were activated or suppressed in response to HDAC inhibition (2562) than by hypoxia alone (375). Second, a striking number of hypoxia-inducible genes were differentially regulated by HDACIs (124). Not surprisingly, a number of these differentially regulated genes were HIF1A targets. Consistent with this observation, our results demonstrate that HDAC inhibition prevents hypoxia-inducible stabilization of HIF1A protein in isolated murine cardiac myocytes. In addition, a number of downstream HIF1A hypoxia-inducible targets are similarly down-regulated, including Vegfa, Egln1, and Egln3. However, we also identified a number of HDACI-regulated, hypoxia-inducible genes with no known association to HIF1A, suggesting that these genes either are novel HIF1A targets or are regulated in a non-HIF1A-dependent yet hypoxia-inducible fashion.
HIF1A plays a critical role in the cellular response to hypoxia and tissue ischemia and is regulated by associated proteins and covalent modifications. In response to hypoxia, HIF1A accumulates and directly induces the transcriptional activity of a number of genes, including genes that control angiogenesis (such as Vegfa) and energy metabolism [for example, the glucose transporter 1 (GLUT1/Scl2a1), 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 (PFKFB3/Pfkfb3)] (33)
. In the presence of oxygen, HIF proteins are hydroxylated on conserved prolyl residues by members of the egg-laying-defective nine (EglN) prolyl hydroxylases. The enzymatic activity of HDACs results in protein deacetylation, usually repressing the transcription of associated loci. In a direct transcriptional model, the application of HDACIs would, in general, release this HDAC-mediated repression, leading to an increase in histone acetylation and subsequent gene activation. Therefore, the ability of HDACIs to repress the ischemia-mediated induction of HIF1A and its target genes was, at first, counterintuitive. However, recent studies show that class II HDACs associate with HIF1A and modulate its stability (34
, 35)
. For example, HDAC7 is able to increase transcriptional activity of HIF1A through the formation of a complex with HIF1A, HDAC7, and p300 (34)
. In addition, HDACIs repress HIF1A independently of von Hippel-Lindau tumor suppressor, suggesting that acetylation may regulate HIF function by targeting HIF1A:cofactor complexes, rather than by direct acetylation of HIF1A (36)
. However, the molecular link between HDACs and HIF1A remains controversial.
Because HDACI protected cardiac myocytes from ischemic injury both in vitro and in vivo and regulated the response of a coordinate set of genes in vitro, we asked whether an individual HDAC might be responsible for this response. We identified HDAC4 as the critical HDAC isoform likely mediating the protective response. Hdac4 null mice die perinatally with abnormal chondrocyte hypertrophy but lacking any identifiable cardiac defects (37)
. Nevertheless, recent studies suggest that HDAC4 may play a role in cardiac physiology. Cytosolic accumulation of HDAC4 is required for cardiomyocyte hypertrophy in response to calcium/calmodulin-dependent kinase II (CamKII) signaling (38)
. Phosphorylation of HDAC4 by CaMKII induces its nuclear export and leads to derepression of HDAC target genes. Our report, to our knowledge, is the first to show that HDAC4 is also involved in cellular response to cardiac ischemia, consistent with a recent study reporting that HDAC4 plays a role in mediating hypoxic response in chondrocytes (39)
. Knockdown of Hdac4 with siRNA increases HIF1A acetylation, and its subsequent degradation via the proteasome pathway in VHL-deficient cell line (35)
. In a series of preliminary experiments using a Gal4-Trap, we have demonstrated a functional interaction between HIF1A and HDAC4 in HEK293T cells, resulting in inhibition of transcription, suggesting the associated increases in hypoxia induced HDAC activity and transcriptional activation in cardiac myocytes might be due to either recruitment of a coactivator, as has been previously shown for HIF1A, HDAC7, and p300 (34)
or HDAC4-mediated stabilization of HIF1A protein level as previously shown in VHL-deficient cell line (35)
. Although we did not examine additional post-translational modifications of HIF1A induced by class II HDACs, others, such as sumoylation, might also be considered. Furthermore, other HDAC family members might be important in the mediation of non-HIF1A-protective mechanisms.
The association of reduced Vegfa expression, reduced vascular permeability, and reduced infarct size is provocative. VEGFA is an important, multifunctional molecule and is directly regulated by HIF1A. With acute exposure, VEGFA is a potent stimulator of vascular permeability, whereas with prolonged exposure, it induces angiogenesis. Endothelial cells exposed to VEGFA develop fenestrations, caveolae, and vesiculo-vacuolar organelles (40
, 41)
. Type 2 VEGF receptor (VEGFR2) associates with VE-cadherin that rapidly dissociates in response to VEGFA (42)
. VEGFA also inhibits gap junction communication via connexin 43 phosphorylation and Src activation (43
, 44)
. Finally, VEGFA disrupts tight junctions by altering the phosphorylation of zonula occludin 1 and occludin (45)
. Consistent with our in vitro studies, Vegfa expression was significantly increased with I/R injury. Mirroring the in vitro results, HDACIs completely blocked the acute, ischemia-induced increase in Vegfa. Although VEGFA is an angiogenic factor that has been proposed as chronic treatment for myocardial infarction, its potent ability to stimulate vascular leak in the acute setting may have profound negative consequences (30)
. Although initially counterintuitive, we hypothesized that the reduction in Vegfa activation may result in a decrease in vascular permeability providing myocardial protection. The association of HDACI-mediated reduction in ischemic injury and HDACI-mediated reductions in Vegfa expression/vascular permeability is compelling and suggests further studies that may directly prove the association of acute inhibition of Vegfa expression with cardiac protection, while preserving the chronic, potential beneficial effects of VEGFA-mediated angiogenesis. Although the complete set of mechanisms that mediate the protective effect of HDACIs remains to be completely delineated, we clearly demonstrate that HDAC inhibition can reduce the size of myocardial infarction due to I/R injury by half. Importantly, treatment is possible even postinjury. Our results have important implications for the potential application of these compounds in the context of acute coronary syndromes such that the administration of HDACIs would only be predicted to have a beneficial effect in the context of active ischemia.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication March 3, 2008. Accepted for publication May 22, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and increases transcriptional activity. J. Biol. Chem. 279,41966-41974. Cancer Res. 66,8814-8821. J. Biol. Chem. 281,13612-13619This article has been cited by other articles:
|
|
 |
|
  E. W. Bush and T. A. McKinsey Protein Acetylation in the Cardiorenal Axis: The Promise of Histone Deacetylase Inhibitors Circ. Res., February 5, 2010; 106(2): 272 - 284. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Ling and L. Groop Epigenetics: A Molecular Link Between Environmental Factors and Type 2 Diabetes Diabetes, December 1, 2009; 58(12): 2718 - 2725. [Full Text] [PDF]
|
|
|
|
 |
|
 D. Pons, F. R. de Vries, P. J. van den Elsen, B. T. Heijmans, P. H.A. Quax, and J. W. Jukema Epigenetic histone acetylation modifiers in vascular remodelling: new targets for therapy in cardiovascular disease Eur. Heart J., February 1, 2009; 30(3): 266 - 277. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
HN and HN 
