Suppression of Histone Deacetylases Worsens Right Ventricular Dysfunction after

Pulmonary Artery Banding in Rats

Harm J. Bogaard MD PhD1,3, Shiro Mizuno MD PhD1, Ayser A. Al Hussaini MD1, Stefano Toldo

PhD2, Antonio Abbate MD PhD2, Donatas Kraskauskas DVM1, Michael Kasper PhD4, Ramesh

Natarajan PhD1 and Norbert F. Voelkel MD1

Divisions of Pulmonary & Critical Care1 and Cardiology2, Dept of Medicine and Victoria Johnson

Center for Lung Research, Virginia Commonwealth University, 1220 East Broad Street,

Richmond, VA 23298; 3Dept of Pulmonary Medicine, VU University Medical Center, Amsterdam,

the Netherlands; 4Institute for Anatomy, Gustav Carus University, Dresden, Germany

Authors’ contributions: Conception and design: HJB, SM, RN, NFV; Analysis and

interpretation: all; Drafting manuscript: HJB, RN, NFV.

Corresponding Author/ Reprints: Norbert Voelkel, Director of the Victoria Johnson Center for

Obstructive Lung Disease Research, Virginia Commonwealth University, 1220 East Broad

Street, Richmond, VA 23298. Telephone: 804-6283334 and Fax: 804-6280325

Support: Victoria Johnson Center for Obstructive Lung Disease Research

Running head: HDAC inhibition worsens right heart failure

Discriptor number: 9.2 Animal Models of Pulmonary Hypertension; 17.6 Pulmonary

Hypertension: Experimental; 17.7 Right Ventricle: Function and Dysfunction

Word count: 4189

At a glance commentary: Scientific knowledge on the subject: HDAC inhibition has

cardioprotective effects in experimental left heart failure, but its effects on right heart adaptation

to pressure overload are unknown. What this study adds to the field: Despite their positive

effects in the pressure overloaded left heart, HDAC inhibitors worsen right heart dysfunction and

remodeling after pulmonary artery banding in rats, which may be related to suppression of

angiogenesis.

Page 1 of 39 AJRCCM Articles in Press. Published on February 4, 2011 as doi:10.1164/rccm.201007-1106OC

Copyright (C) 2011 by the American Thoracic Society.

This article has an online data supplement, which is accessible from this issue's table of content

online at www.atsjournals.org.

Abstract

Rationale: Inhibitors of histone deacetylases (HDACs) reduce pressure-overload induced left

ventricular hypertrophy and dysfunction, but their effects on right ventricular (RV) adaptation to

pressure overload are unknown.

Objective: Determine the effect of the broad-spectrum HDAC inhibitors trichostatin A (TSA) and

valproic acid (VPA) on RV function and remodeling after pulmonary artery banding (PAB) in

rats.

Methods: Chronic progressive RV pressure-overload was induced in rats by PAB. After

establishment of adaptive RV hypertrophy 4 weeks after surgery, rats were treated for two

weeks with vehicle, TSA or VPA. RV function and remodeling were determined using

echocardiography, invasive hemodynamic measurements, immunohistochemistry and

molecular analyses after two weeks of HDAC inhibition. The effects of TSA were determined on

the expression of pro-angiogenic and pro-hypertrophic genes in human myocardial fibroblasts

and microvascular endothelial cells.

Measurements and main results: TSA treatment did not prevent the development of RV

hypertrophy and was associated with RV dysfunction, capillary rarefaction, fibrosis and

increased rates of myocardial cell death. Similar results were obtained with the structurally

unrelated HDAC inhibitor VPA. With TSA treatment, a reduction was found in expression of

VEGF and angiopoietin-1, which proteins are involved in vascular adaptation to pressure-

overload. TSA dose-dependently suppressed VEGF, eNOS and angiopoietin-1 expression in

cultured myocardial endothelial cells, which effects were mimicked by selective gene silencing

of several class I and I HDACs.

Page 2 of 39

Conclusions: HDAC inhibition is associated with dysfunction and worsened remodeling of the

pressure-overloaded RV. The detrimental effects of HDAC inhibition on the pressure overloaded

RV may come about via anti-angiogenic and/or pro-apoptotic effects.

Introduction

A novel concept of preventing and reversing left ventricular (LV) dysfunction and maladaptive

remodeling is that of interfering with the epigenetic control of gene transcription using histone

deacetylase (HDAC) inhibitors1. Recent experimental work has indicated that HDAC inhibitors

repress the development of LV hypertrophy and preserve LV systolic function in the setting of

chronic pressure overload due to transverse aortic constriction (TAC)2-4 and after myocardial

damage induced by ischemia/reperfusion5 or adriamycin6. It has been suggested that HDAC

inhibitors limit LV hypertrophy by enhancing the activity of GSK-3, which is a constitutive

repressor of hypertrophic pathways4. HDAC inhibition is also associated with repression of

angiogenesis7 and induction of apoptosis8;9 and these effects -although undesirable in the

human heart- have led to great interest in the use of HDAC inhibitors for the treatment of

cancer.

Whereas the pathobiology of left sided heart failure has been systematically explored,

very little is known about the cellular and molecular mechanisms which determine the

development of right ventricular (RV) failure10. For example, it is unclear whether a hypertrophic

response is of comparable importance in RV and LV adaptation to pressure overload. Under

normal conditions, the RV wall thickness is only one fifth of that of the LV; the stress imposed on

the RV in patients with pulmonary arterial hypertension (PAH, with doubling or tripling of the

afterload) may require for chronic adaptation a considerably larger degree of hypertrophy than

the stress imposed by systemic hypertension or, experimentally, after TAC (approximately a

50% increase in afterload). Importantly, it was recently shown by Kreymborg et al. that the

transcriptional control of the pressure-overloaded RV is different from that of the pressure-

Page 3 of 39

overloaded LV11. At present, there are no known interventions that specifically support the

adapting RV in situations of pulmonary arterial hypertension or in pulmonary hypertension

related to left heart disease. In both situations, the development of RV failure is an important

determinant of survival12;13. The functional, structural and developmental differences which exist

between the RV and LV raise the question whether a treatment which effectively prevents the

progression of LV failure, is of benefit in RV failure. We have recently shown that adrenergic

receptor blockade has cardioprotective effects in the setting of experimental PAH in rats14. Here

we show that altering the RV response to established pressure overload with the broad-

spectrum HDAC inhibitors trichostatin A (TSA) and valproic acid (VPA) –an intervention which is

known to have beneficial effects on the pressure overloaded LV- results in RV failure in rats

subjected to pulmonary arterial banding (PAB). TSA treatment after PAB and established RV

hypertrophy did not decrease RV hypertrophy, did not prevent fetal gene reactivation and

caused maladaptive fibrotic RV remodeling. The induction of RV failure by HDAC inhibitors

following the compensatory adaptation to pressure overload was associated with decreased

expression of the angiogenesis factors VEGF and Angiopoietin 1 in the RV myocardium and

likewise in cultured cardiac microvascular endothelial cells.. Some of the results of the studies in

this manuscript have been previously reported in the form of an abstract15.

Page 4 of 39

Methods

Pulmonary artery banding

Surgical banding of the pulmonary artery (PAB) was performed in male Sprague-Dawley rats as

described previously16. Via a left thoracotomy in rats weighing 180-200 g, a silk suture was tied

tightly around an 18-gauge needle alongside the pulmonary artery. After subsequent rapid

removal of the needle, a fixed constricted opening was created in the lumen equal to the

diameter of the needle. Whereas the initial constriction was relatively mild, the combination of a

fixed banding around the pulmonary artery and animal growth resulted in a progressive increase

in RV systolic pressure and a pressure gradient of about 50 mmHg after 6 weeks (supplement),

yet, as reported previously by us16 and others17;18 does not result in RV failure. TSA treatment

(450µg/kg i.p., 5 times per week) was initiated 4 weeks after surgery –at a time when a robust

RV hypertrophy had been established -until the day of tissue harvest. The investigation

conforms with the Guide for the Care and Use of Laboratory Animals published by the US

National Institutes of Health (NIH Publication No. 85-23, revised 1996) and all protocols were

approved by the VCU Institutional Animal Care and Use Committee (protocol AM10157).

Echocardiography

Doppler echocardiography was performed using the Vevo770 imaging system (VisualSonics,

Toronto, Canada) 6 weeks after PAB and prior to invasive hemodynamic assessments. Light

anesthesia with ketamine/xylazine was used to obtain two-dimensional, M-mode and Doppler

imaging in both long axis (four-chamber) and short-axis views, using a 30-MHz probe.

Measurements were made of the RV inner diameter in diastole (RVID; long axis), tricuspid

annular plane systolic excursion (TAPSE; long axis), RV free wall thickness in diastole and

systole (short axis) and septum thickness in diastole and systole (short axis).

Page 5 of 39

Hemodynamic measurements

Six weeks after surgical banding of the pulmonary artery and two weeks after the first dose of

TSA, hemodynamic measurements were made using a 4.5 mm conductance catheter (Millar

Instruments, Houston, TX) and the Powerlab data acquisition system (AD Instruments, Colorado

Springs, CO). The rats were anesthetized with isoflurane, intubated and placed in a supine

position. After a median sternotomy the RV outflow tract was punctured with a 23G needle and

the catheter was introduced ante grade to measure RV and pulmonary artery pressures. In

separate groups, using the same anesthesia, intubation and catheterization techniques, cardiac

output was measured by thermodilution. Saline (±12○C) was injected via the jugular catheter

(advanced into the right atrium) and the change in temperature is measured in the aorta using a

thermocouple (advanced via the carotid catheter). Data analysis was performed using

GraphPad and PVAN software (AD Instruments, Colorado Springs, CO).

Gene expression studies

Rat RV, LV and lung were snap frozen in liquid nitrogen. The FastRNA® Pro Green Kit (MPBio)

was used to isolate total RNA from heart tissue. Using the FastPrep®- 24 instrument, 25mg of

tissue was homogenized by Lysing Matrix D in impact-resistant 2ml tubes. Total RNA released

into the proprietary, protective RNApro™ Solution was extracted with chloroform and

precipitated using ethanol. Total RNA (1 µg) was reverse transcribed into complimentary DNA

(cDNA) using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). First

strand cDNA was diluted and RT-QPCR performed using Power SYBR® Green PCR Master

Mix (Applied Biosystems) along with murine specific primers. Cycling parameters were: 95˚C,

10 min and 45 cycles of 95˚C, 15 sec, 60˚C, 1 min. A dissociation profile was generated after

each run to verify specificity of amplification. All PCR assays were performed in triplicate. No

template controls and no reverse transcriptase controls were included. Automated gene

expression analysis was performed using the Comparative Quantitation module of MxPro

Page 6 of 39

QPCR Software (Stratagene) to compare the levels of a target gene in test samples relative to a

sample of reference (calibrator from untreated cells).

Protein expression studies

Western blots were performed using standard procedures and antibodies commercially

available from Santa Cruz Biotechnology, Santa Cruz CA (VEGF: #sc-152, Collagen I-A1: #sc-

25974, Akt: #sc-8312, P-Akt: #sc-7985, Ang1: #sc-6320) and Cell Signaling, Danver MA

(Cleaved caspase-3: #9661, HIF-1α: #3716).

Immunohistochemistry

Masson’s Trichrome stain was used to assess the degree of fibrosis in cardiac sections.

Fibrosis was quantified on digitized images, on which blue stained tissue areas are expressed

as percentage of the total surface area. Anti-CD31 and Caveolin-1 antibodies were used to stain

and quantify the vascular density in the RV in tissue sections with cardiomyocytes that were cut

longitudinally. Cell death within the RV myocardium was identified with terminal

deoxynucleotidyl transferase-mediated dUTP nick-end labeling (Tunel stain; DNA

fragmentation, Apoptag, Chemicon, CA). The number of Tunel positive cardiomyocytes was

divided by the total number of cardiomyocytes per field at 400× under a light microscope

Cell culture

Human cardiac microvascular endothelial cells (HCMVECs) and endothelial growth medium

were purchased from Lonza (Walkersville MD). Human cardiac fibroblasts (HCFs) and fibroblast

growth medium were purchased from Cell Applications (San Diego CA). The cells were cultured

in 175-cm2 tissue culture flasks in a cell-culture incubator (37˚C, 5% CO2, and 95% air) and

used at the sixth passage (HCMVECs) or fourth passage (HCFs) after trypsinization in all the

experiments. HCMVECs and HCFs were seeded in 6 well culture plate or 10 cm culture dish

Page 7 of 39

and cultured until reaching to the confluent. After reaching to the confluent, mediums were

changed and various concentrations of TSA (0.01 to10 µM) were added to the dishes and

cultured for 24 hours. After incubation, the cells were harvested and used for RT-PCR and

Western blot analysis. To obtain more specific information regarding the role of individual

HDACs in angiogenic gene expression, HCMVECs were transfected with small interfering RNAs

(siRNAs) specific for human HDAC1-7 (Invitrogen, Carlsbad, CA) using previously published

protocols19. The efficiency of gene knock-down was verified by quantitative PCR, showing that

siRNA transfection resulted in at least 80% reduction HDAC1-7 mRNA expression. The

following sequences were used: control siRNA sense: CCUAGAACCUAAGACCCUU,

antisense: AAGGGUCUUAGGUUCUAGG; HDAC1 siRNA sense:

GCUCCAUCCGUCCAGAUAA, antisense: UUAUCUGGACGGAUGGAGC; HDAC2 siRNA

sense: CCAAUG AGUUGCCAUAUAA, antisense: UUAUAUGGCAACUCAUUGG; HDAC3

siRNA sense: CCAAGAGUCUUAAUGCCUU, antisense: AAGGCAUUAAGACUCUUGG;

HDAC4 siRNA sense: CCACGAGCACAUCAAGCAA, antisense:

UUGCUUGAUGUGCUCGUGG; HDAC5 siRNA sense: GGACUUCUCUGCACAGCAU,

antisense: AUGCUGUGCAGAGAAGUCC; HDAC6 siRNA sense:

GCAAUGGAAGAAGACCUAA, antisense: UUAGGUCUUCUUCCAUUGC; HDAC7 siRNA

sense: GCACCCAGCAAACCUUCUA, antisense: UAGAAGGUUUGCUGGGUGC.

Statistics

Differences between groups were assessed with ANOVA (parametric) and Kruskall-Wallis (non-

parametric) tests; Bonferroni’s (parametric) and Dunn’s (non-parametric) post-hoc tests were

used to assess for significant differences between pairs of groups. P-values < 0.05 were

considered significant. For reasons of clarity, all data are reported as means ± SEM, unless

specified otherwise, even if the differences between groups were tested with a non-parametric

Page 8 of 39

test that makes no use of means and standard deviations. 6-8 rats were used per group, unless

specified otherwise.

Page 9 of 39

Results

HDAC inhibition with trichostatin A does not prevent RV hypertrophy and is associated

with increased RV dysfunction

Control animals treated with TSA had a normal increase in body weight with aging, but exhibited

a mild degree of RV hypertrophy without signs of RV failure (Figures 1-2). TSA treated rats also

showed some degree of lung emphysema. TSA treatment of PAB rats was associated with a

trend towards lower RV systolic pressures. RV function was maintained after PAB in vehicle

treated rats, as we and others have shown previously16-18. However, PAB rats treated with TSA

showed a decreased cardiac output and overt signs of RV failure on cardiac ultrasound

(pericardial fluid, systolic paradox movement of the interventricular septum and increased RV

dilatation, see Figure 2). In contrast to the data obtained with the TAC model, TSA treatment did

not decrease the degree of RV hypertrophy (Figure 1) in response to pressure overload.

Adaptive myocardial hypertrophy (such as occurs with exercise) is characterized by increased

signaling to cardiomyocytes via insulin-like growth factor (IGF)-1, and IGF-1 seems to be at

least in part derived from cardiac fibroblasts20. IGF-1 stimulation of cardiomyocytes is followed

by activation (phosphorylation) of Akt, a key pro-growth and pro-survival factor in many tissues

including the heart10. In addition to the effects of IGF-1 signaling on Akt, Akt phosphorylation is

affected by many other growth factors. Whereas RV IGF-1 mRNA expression was increased

after PAB in vehicle treated rats (without significant changes in the IGF-1 receptor, data not

shown), this response was absent in TSA treated PAB rats (Figure 3). pAkt/Akt ratios were

lower in TSA treated than vehicle treated PAB rats. The structurally unrelated broad spectrum

HDAC inhibitor valproic acid (VPA) had similar -but less severe- effects on RV function and

histology in PAB rats as TSA (see online supplement).

Page 10 of 39

HDAC inhibition after PAB is associated with a greater degree of RV fibrosis and

capillary rarefaction

Cardiac fibrosis21;22 and capillary rarefaction23 can contribute to the development of LV failure in

response to pressure overload. Both histological features are present in the RV of rats with

severe angioproliferative pulmonary hypertension, but not after PAB16. TSA treatment increased

the degree of fibrosis assessed in trichrome stained RV tissue sections in PAB, but not in

control rats (Figure 4A-E). Similarly, TSA treatment induced significant capillary rarefaction after

PAB but did not induce rarefaction in control rats (Figure 4F-J and Figure 2, online supplement).

Tunel staining showed no or only occasional cell death in the RV of control rats, TSA treated

non-banded rats or vehicle treated PAB rats. In contrast, TSA treatment of PAB rats was

associated with increased rates of cell death in the RV (Figure 4K-O). The increases in fibrosis

and rates of cell death in the RVs of TSA treated PAB rats were paralleled by an increased

protein expression of collagen 1A1 and activated caspase 3 (Figure 5). Capillary rarefaction in

TSA treated PAB rats was associated with decreased protein and mRNA expression of VEGF

and Ang-1 (Figure 5; no significant changes were observed in VEGF-R1 and VEGF-R2 mRNA

expression, data not shown). Importantly, the decreased VEGF gene expression following TSA

treatment was associated with increased nuclear HIF-1α expression both in controls and in PAB

rats (see online supplement), indicating a transcriptional uncoupling of HIF-1α from the target

gene VEGF.

Fetal gene expression with TSA treatment

TAC and PAB are both associated with a re-activation of fetal genes, such as β-myosin heavy

chain (MHC; at the expense of α-MHC) and atrial natriuretic peptide (ANP)16. HDAC inhibition

has been shown to be associated with an attenuation of fetal gene reactivation in the pressure

overloaded LV24. Remarkably, TSA did not repress fetal gene reactivation in the pressure

overloaded RV (see online supplement). As expected, TSA treatment increased the ratio of α- to

Page 11 of 39

β-MHC in the normal LV (see online supplement) and RV. As reported previously24, TSA

decreased the mRNA expression of atrial natriuretic peptide (ANP) in the normal LV.

Unexpectedly, TSA increased ANP expression in the normal RV. The increased ANP

expression with TSA was associated with an increased expression of the transcription factor

Ptix2, which a major determinant of right-left asymmetry in the heart25 and one of the controllers

of ANP transcription26.

Effects of HDAC inhibition on human myocardial microvascular endothelial cells

(HCMVECs) and human cardiac fibroblasts (HCFs)

In cardiomyocytes in vitro, HDAC inhibition suppresses agonist-induced gene expression of β-

MHC and ANP24, and protects hypoxic cells from apoptosis (whereas paradoxically, HDAC

inhibition seems to increase apoptotic rates in normoxic cardiomyocytes)5. Effects of HDAC

inhibition on other cell populations in the heart are unknown. Because TSA treatment caused a

striking capillary rarefaction in the PAB RV and HDAC inhibitors have well-known anti-

angiogenic actions7, we determined the effect of TSA treatment on HMVECs and found a

significant reduction in VEGF, eNOS and Ang-1 gene and protein expression (see Figure 6). In

silencing experiments using siRNAs directed against individual class I and class II HDACs, it

appeared that the suppression of angiogenic gene expression by the broad-spectrum HDAC

inhibitors TSA and VPA was not the result of the inhibition of any single HDAC (see Figure 7).

Rather, silencing of several class I and class II HDACs resulted in the suppression of VEGF,

eNOS and Angiopoietin-1 gene expression in HMVECs.

TSA induced IGF-1 expression in HCFs in a dose-dependent fashion (see online supplement),

which deems it unlikely that the reduction in IGF-1 expression in whole RV lysates of banded

rats was directly due to TSA treatment.

Page 12 of 39

Discussion

HDACs are transcriptional repressors that promote nucleosomal condensation and have

recently emerged as important controllers of the LV hypertrophic response and its

accompanying fetal gene reactivation24. When there is no transcription, DNA is wrapped around

histone octameres in nucleosomes, which are the basic units of chromatin. The highly compact

structure that is formed by interacting nucleosomes limits access of transcriptional enzymes to

genomic DNA, thereby repressing gene expression27. Acetylation of histones by histone acetyl

transferases (HATs) relaxes the nucleosomal structures, thereby facilitating gene expression;

HDACs, when activated, have the opposite effect28. HDAC inhibitors suppress pressure and

agonist-dependent cardiac hypertrophy and prevent fetal gene reactivation2-4;24, perhaps

because class I HDACs (e.g. HDAC2) repress constitutively active inhibitors of hypertrophic

pathways such as GSK-34. On the other hand, mice lacking HDAC2 demonstrate hyperplasia

and apoptosis of cardiomyocytes and obliteration of the RV cavity29. Here we provide

experimental results that indicate that HDAC inhibitors adversely affect the pressure-overloaded

RV. Whereas chronic treatment with TSA did not influence RVSP, RV weight or cardiac output

in otherwise not stressed control animals, chronic TSA treatment in PAB animals caused a

switch from compensated hypertrophy to RV failure. The evidence for TSA-induced RV failure in

PAB animals is a decreased cardiac output, increased RV dilatation and a worsening of

myocardial fibrosis and capillary loss. The importance of our findings lies in the categorical

difference between the results of HDAC inhibition in the pressure-overloaded LV and RV: TSA

treatment in TAC mice reduced LV hypertrophy and improved function3;4, whereas TSA

treatment in PAB rats did not reduce established RV hypertrophy, reduced RV expression of

angiogenic growth factors and caused the RV to fail. The differential response to pressure

overload is perhaps understandable in view of the recent findings by Kreymborg et al. Using a

micro array approach, these authors showed that the transcriptional adaptation to pressure

Page 13 of 39

overload is very different in the RV after pulmonary artery banding than in the LV after aortic

constriction11.

Effects of HDAC inhibition on hypertrophy, fibrosis, capillary rarefaction and apoptosis

Whereas TSA reduced hypertrophy, collagen synthesis, fibrosis and fetal gene reactivation after

TAC2-4;24, we report opposite effects of TSA treatment in the pressure-overload RV. These

findings underscore the fact that with their potential impact on the transcription of a large

number of genes in different cell types, broad-spectrum HDAC inhibitors ultimately have effects

on organ function and structure which are difficult to predict and which are context-dependent.

Remarkably, TSA treatment had different effects on ANP expression in both unstressed cardiac

chambers: TSA decreased ANP expression in the control LV (as reported previously24), but

increased ANP expression in the control RV. This difference was paralleled by a LV-RV

differential response in expression of Pitx2, an important controller of the ANP gene and RV

morphogenesis26. This result could indicate intrinsic differences in transcriptional control

between both cardiac chambers, both at baseline (i.e. un-banded) and under stress (after PAB).

In fact, the increase in ANP and IGF-1 transcription that we found at baseline in TSA treated

rats resembled a stress response. It is possible that this was the physiological result of induction

of pulmonary emphysema, but the increase in IGF-1 gene transcription which we found with

TSA treatment of cultured cardiac fibroblasts may suggest that TSA can in certain

circumstances evoke a transcriptional stress response. It has recently been shown that IGF-1

production by myocardial fibroblasts is essential for LV adaptation to pressure overload20. The

small degree of RV hypertrophy which we found in TSA treated un-banded rats, could be the

direct result of IGF-1 up-regulation in cardiac fibroblasts. Absence of this response in the LV

could simply be due to fact that the context of TSA treatment in the LV is entirely different:

normal LV cardiomyocytes are relatively hypertrophic when compared to normal RV

cardiomyocytes and normal LV cardiomyocytes have a higher ANP transcription than RV

Page 14 of 39

cardiomyocytes to begin with. Unfortunately, there is to our knowledge no available method to

culture cardiac chamber specific neonatal cardiomyocytes: such a cell system could explore

chamber-specific differences in epigenetic control of gene transcription. Importantly, the degree

of emphysema induced by TSA is insufficient to explain the development of RV failure after

PAB: we recently showed that PAB rats treated with SU5416 have a comparable degree of

pulmonary emphysema, but normal RV function16.

The anti-angiogenic effects of HDAC inhibitors -described previously in cancer cell

lines7- may have been responsible for the extensive capillary rarefaction seen in our

experiments. Here we report for the first time that TSA represses gene transcription of eNOS,

VEGF and Ang-1 in HCMVECs. The gene silencing experiments suggested that these effects

were the combined result of inhibition of several, predominantly class II, HDACs. HDAC6

appeared as an important regulator of VEGF expression in human cardiac microvascular

endothelial cells. The results from the siRNA experiments underline the fact that the end-result

of treatment with broad-spectrum HDAC inhibitors is difficult to predict, which is most likely due

to interaction between, and partial redundancy of HDACs. In contrast to our findings of

extensive RV capillary rarefaction with TSA treatment of PAB rats, there were no effects

reported of HDAC inhibition on LV capillary density after TAC2-4. It is possible that an angiogenic

response is much more important for RV adaptation to PAB, than for LV adaptation to TAC. As

shown here, PAB is associated with a doubling of RV mass, whereas TAC is usually associated

with only a 20-50% increase in LV mass. RV-LV differences in vascular adaptation to pressure

overload, could be responsible for the differences in the effects of HDAC inhibition after PAB

and TAC. Anti-angiogenic effects of HDAC inhibition are usually ascribed to a repression of HIF-

1α expression7, which we did not observe in the TSA treated banded RV. We speculate that the

RV dilatation and capillary rarefaction which occurs in the banded RV, contributed to the

development of myocardial ischemia and hypoxic stabilization of HIF-1α expression. We have

reported a disconnect of VEGF gene transcription and HIF-1α protein stabilization in the failing

Page 15 of 39

RV in experimental pulmonary hypertension16. HIF-1α expression was increased by TSA in the

RV of control rats, which may point to either RV ischemia or hypoxia due to TSA-related

changes of the lung structure. Hypoxia by itself is not sufficient to induce failure of the banded

RV16 and could therefore not account for the detrimental effect of TSA treatment in PAB rats.

An increase in the number of Tunel positive cells was seen in the failing RV of TSA

treated PAB. HDAC inhibition has been shown to be associated with apoptotic and autophagic

cell death of cancer cells8 and TSA has been shown to induce apoptosis of normoxic (but not

ischemic) cardiomyocytes5. As is true for the effect of TSA on capillary rarefaction, our data do

not answer the question whether TSA-related apoptosis was directly responsible for the

detrimental effects of HDAC inhibition on the adapting RV. However, data obtained with cultured

heart microvascular endothelial cells confirm the anti-angiogenic activity of TSA, suggesting that

TSA had also exerted an antiangiogenic action in vivo in the pressure-overloaded RV.

Recently, Cho et al30 reported that valproic acid prevented the development of RV

hypertrophy in young rats where PAB had induced RV failure. The data reported by this group

differ in several aspects from the data reported here. Cho et al. describe the development of

severe RV failure in vehicle treated PAB rats, as evidenced by an increased inferior vena cava

diameter, marked RV hypertrophy (RV/LV+S increasing up to 0.80), a reduced gain in body

weight and the development of RV fibrosis30. Unfortunately, no hemodynamic data were

provided, but it seems that the technique used by this group differed from the more conventional

PAB technique, which is known to elicit a compensatory RV hypertrophic response, a limited

degree of RV fibrosis and no evidence of RV failure16-18. The different findings by Cho et al.

may have been related to the use of younger rats (three weeks at the time of surgery) or,

perhaps, to a more severe PA constriction at the time of surgery (rendering a model of acute RV

pressure overload, not chronic progressive pressure overload as induced by our technique). In

addition, Cho et al. initiated treatment directly after surgery, whereas we chose to wait until after

the generation of adaptive hypertrophy (4 weeks after surgery). We felt that this approach would

Page 16 of 39

have more resemblance to the clinical situation of TSA treatment of patients with established

pulmonary hypertension.

Limitations

Global HDAC inhibition experiments –such as those based on TSA administration– need to be

compared with those of targeted cardiac HDAC gene experiments as performed by Montgomery

et al.29. In the latter study it was shown that HDAC1 and HDAC2 redundantly regulate cardiac

growth and that deletion of either gene alone was insufficient to reduce LV hypertrophy of the

TAC-stressed hearts. The same authors showed in another study, that HDAC3 is important for

the maintenance of cardiac energy metabolism in mice.31 Because of the technical challenges

that are associated with performing PAB in mice and because of the limited possibilities for

transgenic manipulations in rats, we were restricted to make use of pharmacological

interventions. As outlined above, we did not initiate TSA treatment immediately after surgery,

but 4 weeks later. Although this may explain part of the discrepancy between our results and the

reported effects of TSA after TAC, we feel that a scenario of late treatment is likely more

pertinent to the clinical situation of patients with already established severe PAH.

Clinical implications

One important implication of our findings is that concerns are being raised regarding the

development of anti-angiogenic and pro-apoptotic drugs for the treatment of severe

angioproliferative pulmonary hypertension. Although such treatments may make intuitive

sense32, their consequences for the stressed RV are unpredictable. The second implication is

that although HDAC inhibition may seem attractive for the treatment of LV systolic failure, the

effects of HDAC inhibition on the RV are again unpredictable. Congestive heart failure almost

always involves both ventricles and there is a distinct possibility that what works for the failing

LV, may not work for the failing RV.

Page 17 of 39

Reference List

1. Bush, E. W. and T. A. McKinsey. 2010. Protein acetylation in the cardiorenal

axis: the promise of histone deacetylase inhibitors. Circ Res. 106:272-284.

2. Kee, H. J., I. S. Sohn, K. I. Nam, J. E. Park, Y. R. Qian, Z. Yin, Y. Ahn, M. H.

Jeong, Y. J. Bang, N. Kim, J. K. Kim, K. K. Kim, J. A. Epstein, and H. Kook.

2006. Inhibition of histone deacetylation blocks cardiac hypertrophy induced by

angiotensin II infusion and aortic banding. Circulation. 113:51-59.

3. Kong, Y., P. Tannous, G. Lu, K. Berenji, B. A. Rothermel, E. N. Olson, and J. A.

Hill. 2006. Suppression of class I and II histone deacetylases blunts pressure-

overload cardiac hypertrophy. Circulation 113:2579-2588.

4. Trivedi, C. M., Y. Luo, Z. Yin, M. Zhang, W. Zhu, T. Wang, T. Floss, M.

Goettlicher, P. R. Noppinger, W. Wurst, V. A. Ferrari, C. S. Abrams, P. J. Gruber,

and J. A. Epstein. 2007. Hdac2 regulates the cardiac hypertrophic response by

modulating Gsk3 beta activity. Nat.Med. 13:324-331.

5. Granger, A., I. Abdullah, F. Huebner, A. Stout, T. Wang, T. Huebner, J. A.

Epstein, and P. J. Gruber. 2008. Histone deacetylase inhibition reduces

myocardial ischemia-reperfusion injury in mice. FASEB J. 22:3549-3560.

6. Daosukho, C., Y. Chen, T. Noel, P. Sompol, R. Nithipongvanitch, J. M. Velez, T.

D. Oberley, and D. K. St Clair. 2007. Phenylbutyrate, a histone deacetylase

Page 18 of 39

inhibitor, protects against Adriamycin-induced cardiac injury. Free

Radic.Biol.Med. 42:1818-1825.

7. Ellis, L., H. Hammers, and R. Pili. 2009. Targeting tumor angiogenesis with

histone deacetylase inhibitors. Cancer Lett. 280:145-153.

8. Shao, Y., Z. Gao, P. A. Marks, and X. Jiang. 2004. Apoptotic and autophagic cell

death induced by histone deacetylase inhibitors. Proc.Natl.Acad.Sci.U.S.A

101:18030-18035.

9. Emanuele, S., M. Lauricella, and G. Tesoriere. 2008. Histone deacetylase

inhibitors: apoptotic effects and clinical implications (Review). Int.J.Oncol.

33:637-646.

10. Bogaard, H. J., K. Abe, A. Vonk-Noordegraaf, and N. F. Voelkel. 2009. The right

ventricle under pressure; cellular and molecular mechanisms of right heart failure

in pulmonary hypertension. Chest 135:794-804.

11. Kreymborg, K., S. Uchida, P. Gellert, A. Schneider, T. Boettger, R. Voswinckel,

A. Wietelmann, M. Szibor, N. Weissmann, A. H. Ghofrani, R. Schermuly, D.

Schranz, W. Seeger, and T. Braun. 2010. Identification of right heart-enriched

genes in a murine model of chronic outflow tract obstruction. J.Mol.Cell Cardiol.

49:598-605.

12. D'Alonzo, G. E., R. J. Barst, S. M. Ayres, E. H. Bergofsky, B. H. Brundage, K. M.

Detre, A. P. Fishman, R. M. Goldring, B. M. Groves, J. T. Kernis, and . 1991.

Page 19 of 39

Survival in patients with primary pulmonary hypertension. Results from a national

prospective registry. Ann.Intern.Med. 115:343-349.

13. Ghio, S., A. Gavazzi, C. Campana, C. Inserra, C. Klersy, R. Sebastiani, E.

Arbustini, F. Recusani, and L. Tavazzi. 2001. Independent and additive

prognostic value of right ventricular systolic function and pulmonary artery

pressure in patients with chronic heart failure. J Am Coll Cardiol 37:183-188.

14. Bogaard, H. J., Natarajan, R., Mizuno, S., Abbate, A., Chang, P, Chau, V, Hoke,

N, Kraskauskas, D., Kasper, M., Salloum, F., and Voelkel, N. F. Adrenergic

receptor blockade reverses right heart remodeling and dysfunction in pulmonary

hypertensive rats. Am J Respir Crit Care Med 182, 652-660. 9-1-2010.

15. Bogaard, H. J., R. Natarajan, D. Kraskauskas, L. Smithson, and N. F. Voelkel.

2009. Suppression of Histone Deacetylases Worsens Right Ventricular

Hypertrophy and Induces Failure after Pulmonary Artery Banding in Rats.

Am.J.Respir.Crit.Care Med. 179:A4142.

16. Bogaard, H. J., R. Natarajan, S. C. Henderson, C. S. Long, D. Kraskauskas, L.

Smithson, R. Ockaili, J. M. McCord, and N. F. Voelkel. 2009. Chronic pulmonary

artery pressure elevation is insufficient to explain right heart failure. Circulation.

120:1951-1960.

17. Faber, M. J., M. Dalinghaus, I. M. Lankhuizen, K. Bezstarosti, A. J. Verhoeven,

D. J. Duncker, W. A. Helbing, and J. M. Lamers. 2007. Time dependent changes

Page 20 of 39

in cytoplasmic proteins of the right ventricle during prolonged pressure overload.

J Mol Cell Cardiol 43:197-209.

18. Faber, M. J., M. Dalinghaus, I. M. Lankhuizen, P. Steendijk, W. C. Hop, R. G.

Schoemaker, D. J. Duncker, J. M. J. Lamers, and W. A. Helbing. 2006. Right and

left ventricular function after chronic pulmonary artery banding in rats assessed

with biventricular pressure-volume loops. Am J Physiol Heart Circ Physiol

291:H1580-H1586.

19. Mizuno, S., H. J. Bogaard, N. F. Voelkel, Y. Umeda, M. Kadowaki, S. Ameshima,

I. Miyamori, and T. Ishizaki. 2009. Hypoxia regulates human lung fibroblast

proliferation via p53-dependent and -independent pathways. Respir.Res. 10:17.

20. Takeda, N., I. Manabe, Y. Uchino, K. Eguchi, S. Matsumoto, S. Nishimura, T.

Shindo, M. Sano, K. Otsu, P. Snider, S. J. Conway, and R. Nagai. 2010. Cardiac

fibroblasts are essential for the adaptive response of the murine heart to

pressure overload. J Clin.Invest. 120:254-265.

21. Khan, R. and R. Sheppard. 2006. Fibrosis in heart disease: understanding the

role of transforming growth factor-beta in cardiomyopathy, valvular disease and

arrhythmia. Immunology. 118:10-24.

22. Weber, K. T., S. Janicki J., S. G. Shroff, R. Pick, R. M. Chen, and R. I. Bashey.

1988. Collagen remodeling of the pressure-overloaded, hypertrophied nonhuman

primate myocardium. Circ Res 62:757-765.

Page 21 of 39

23. Sano, M., T. Minamino, H. Toko, H. Miyauchi, M. Orimo, Y. Qin, H. Akazawa, K.

Tateno, Y. Kayama, M. Harada, I. Shimizu, T. Asahara, H. Hamada, S. Tomita, J.

D. Molkentin, Y. Zou, and I. Komuro. 2007. p53-induced inhibition of Hif-1 causes

cardiac dysfunction during pressure overload. Nature. 446:444-448.

24. Antos, C. L., T. A. McKinsey, M. Dreitz, L. M. Hollingsworth, C. L. Zhang, K.

Schreiber, H. Rindt, R. J. Gorczynski, and E. N. Olson. 2003. Dose-dependent

blockade to cardiomyocyte hypertrophy by histone deacetylase inhibitors.

J.Biol.Chem. 278:28930-28937.

25. Mommersteeg, M. T., N. A. Brown, O. W. Prall, V. C. de Gier-de, R. P. Harvey,

A. F. Moorman, and V. M. Christoffels. 2007. Pitx2c and Nkx2-5 are required for

the formation and identity of the pulmonary myocardium. Circ Res. 101:902-909.

26. Houweling, A. C., M. M. van Borren, A. F. M. Moorman, and V. M. Christoffels.

2005. Expression and regulation of the atrial natriuretic factor encoding gene

Nppa during development and disease. Cardiovascular Research 67:583-593.

27. Fischle, W., Y. Wang, and C. D. Allis. 2003. Histone and chromatin cross-talk.

Curr.Opin.Cell Biol. 15:172-183.

28. Backs, J. and E. N. Olson. 2006. Control of cardiac growth by histone

acetylation/deacetylation. Circ.Res. 98:15-24.

29. Montgomery, R. L., C. A. Davis, M. J. Potthoff, M. Haberland, J. Fielitz, X. Qi, J.

A. Hill, J. A. Richardson, and E. N. Olson. 2007. Histone deacetylases 1 and 2

Page 22 of 39

redundantly regulate cardiac morphogenesis, growth, and contractility. Genes

Dev. 21:1790-1802.

30. Cho, Y. K., G. H. Eom, H. J. Kee, H. S. Kim, W. Y. Choi, K. I. Nam, J. S. Ma, and

H. Kook. 2010. Sodium valproate, a histone deacetylase inhibitor, but not

captopril, prevents right ventricular hypertrophy in rats. Circ.J. 74:760-770.

31. Montgomery, R. L., M. J. Potthoff, M. Haberland, X. Qi, S. Matsuzaki, K. M.

Humphries, J. A. Richardson, R. Bassel-Duby, and E. N. Olson. 2008.

Maintenance of cardiac energy metabolism by histone deacetylase 3 in mice.

J.Clin.Invest 118:3588-3597.

32. Rai, P. R., C. D. Cool, J. A. King, T. Stevens, N. Burns, R. A. Winn, M. Kasper,

and N. F. Voelkel. 2008. The Cancer Paradigm of Severe Angioproliferative

Pulmonary Hypertension. Am J Respir Crit Care Med 178:558-564.

Page 23 of 39

Figure Legends

Figure 1 Effects of pulmonary artery banding (PAB) and trichostatin-A (TSA) treatment on

hemodynamics and hypertrophy. PAB leads to an increase in RV systolic pressure (RVSP,

panel A), a predictable degree of RV hypertrophy for a given degree of pressure overload

(panels B and C; triangles for vehicle treated and circles for TSA treated PAB rats). TSA

treatment leads to a small degree of RV hypertrophy in control rats and to disproportional RV

hypertrophy (panels B and C for whole RV weights and panels E and F for cardiomyocyte cross

sections) in PAB rats. There were no differences between groups in LV + septal (S) weights

(panel D).

Figure 2 Cardiac ultrasound shows that PAB is associated with an increased right ventricular

inner diameter in diastole (RVID, panel A) when compared to control rats and a maintained

geometry of the heart. TSA treatment has no effects on RV function of normal rats, but is

associated with exaggerated rotation of the heart, RV dilatation, flattening of the interventricular

septum (dotted line) and a decreased tricuspid annular plane systolic excursion (TAPSE, panel

B) in PAB rats. Cardiac output measured by thermodilution was maintained after PAB, but

decreased with TSA treatment in PAB animals (panel C). Average heart rates in all groups were

the same.

Figure 3 PAB leads to increased gene transcription of insulin-like growth factor (IGF)-1 (panel

A) and activation (phosphorylation) of Akt (panels B and C; please note different order of

experimental conditions in Western Blot). TSA treatment is associated with increased

expression of IGF-1 and Akt phosphorylation in control rats, but paradoxically, with repression of

IGF-1 and Akt phophorylation in PAB rats.

Page 24 of 39

Figure 4 RV capillary density (panels A-D; capillaries stained with anti-Caveolin-1 antibodies) is

not affected by either PAB or TSA alone, but is reduced in TSA treated PAB rats (quantification

in panel E). Similarly, PAB and TSA alone are not associated with RV fibrosis (panels F-J;

Trichrome stain) or myocardial cell death (panels K-O; Tunel stain), in contrast to TSA treatment

of PAB rats.

Figure 5 Western blots of RV whole cell lysates (panels A and B; please note different order of

experimental conditions when compared to densitometries) show an increased protein

expression of cleaved caspase-3 (densitometry in panel C) and collagen-I A1 (ColIA1;

densitometry in panel E), and a decreased protein expression of VEGF (densitometry in panel

D) and angiopoietin (Ang)-1 (densitometry in panel F) in TSA treated PAB rats. qPCR does not

show changes in VEGF gene expression in any of the experimental conditions (panel G),

whereas Ang-1 mRNA expression is reduced by PAB and TSA, alone or in combination (panel

H).

Figure 6 TSA dose-dependently represses Ang-1 (panel A), eNOS (panel B), and VEGF gene

expression (panel C) in human cardiac microvascular endothelial cells. Accordingly, Ang-1,

eNOS and VEGF protein expression in human cardiac microvascular endothelial cells is

reduced with TSA treatment (panel D, Western blot and panel E, densitometry).

Figure 7 Transfection of human cardiac microvascular endothelial cells with siRNAs against

class I (black bars) and II (grey bars) HDACs shows that Ang-1 expression is suppressed

mainly by HDACs 3 (class I) and 7 (class II; panel A), that eNOS expression is predominantly

suppressed by HDACs 1 (class I) and 4 (class II; panel B) and that VEGF expression is

suppressed by all class II HDACs (most importantly HDAC6), but none of the class I HDACs

Page 25 of 39

(panel C). Gene silencing of the class I HDACs 2 and 3 activates VEGF and eNOS mRNA

expression, respectively. a: p<0.001; b: p<0.01.

Page 26 of 39

A

C

B

D

F E

Control TSA

PAB PAB-TSA

Figure 1

Page 27 of 39

A B C

Figure 2

Control TSA

PAB PAB-TSA

Page 28 of 39

Figure 3

A B

C

Page 29 of 39

Control

PAB-TSA

PAB PAB PAB

TSA TSA TSA

Control Control

PAB-TSA PAB-TSA

F A K

B

C

D

G

H

A

I

L

M

N

E J O

Figure 4 Page 30 of 39

Figure 5

A B

C D

E F

G H

Page 31 of 39

Figure 6

A B

C D

eNOS

Ang-1

VEGF

β-actin

Control TSA 0.3 µM

E

Page 32 of 39

Figure 7

A B C

Page 33 of 39

Online Data Supplement

Suppression of Histone Deacetylases Worsens Right Ventricular Dysfunction after

Pulmonary Artery Banding in Rats

Harm J. Bogaard MD PhD1,3, Shiro Mizuno MD PhD1, Ayser A. Al Hussaini MD1, Stefano Toldo

PhD2, Antonio Abbate MD PhD2, Donatas Kraskauskas DVM1, Michael Kasper PhD4, Ramesh

Natarajan PhD1 and Norbert F. Voelkel MD1

Figure Legends

Figure 1 Like TSA treatment, valproic acid (VPA) treatment leads to increased right ventricular

(RV) dilatation (RVID is RV inner diameter in diastole, panel A) and a trend towards more

fibrosis after pulmonary artery banding (PAB, panel D). In contrast to TSA, VPA has no effect on

RV hypertrophy (panel B) or capillary density (panel C).

Figure 2 TSA treatment is associated with increased nuclear stabilization of HIF-1α protein in

control and PAB rats (panels A and B). TSA treatment inhibits the increase in number of

capillaries per cardiomyocyte which is normally found after PAB (panel C)

Figure 3 PAB is associated with fetal gene re-expression: a decreased mRNA expression of α-

myosin heavy chain (MHC; panel A) and an increased expression β-MHC (panel B), ANP (panel

C) and Pitx2 (panel D). TSA treatment increases the α- to β-MHC ratio, and ANP and Pitx2

expression in control rats, but does not affect fetal gene expression in the banded RV.

Figure 4 In the unstressed LV, TSA treatment suppresses the expression of the fetal genes β-

MHC and ANP, and activates the expression of α-MHC.

Figure 5 TSA enhances IGF-1 gene (panel A) and protein expression (panel B) in human

cardiac fibroblasts.

Page 34 of 39

Figure 1

C

A B

D

Page 35 of 39

Figure 2

A B

C

Page 36 of 39

Figure 3

A

D C

B

Page 37 of 39

Figure 4

Page 38 of 39

Figure 5

A

B

β-actin

IGF-1

Control TSA 0.3µM

Page 39 of 39