Transcription factors, transcriptional coregulators, and epigenetic modulation in the control of pulmonary vascular cell phenotype: therapeutic implications for pulmonary hypertension (2015 Grover Conference series)

Abstract Abstract

Pulmonary hypertension (PH) is a complex and multifactorial disease involving genetic, epigenetic, and environmental factors. Numerous stimuli and pathological conditions facilitate severe vascular remodeling in PH by activation of a complex cascade of signaling pathways involving vascular cell proliferation, differentiation, and inflammation. Multiple signaling cascades modulate the activity of certain sequence-specific DNA-binding transcription factors (TFs) and coregulators that are critical for the transcriptional regulation of gene expression that facilitates PH-associated vascular cell phenotypes, as demonstrated by several studies summarized in this review. Past studies have largely focused on the role of the genetic component in the development of PH, while the presence of epigenetic alterations such as microRNAs, DNA methylation, histone levels, and histone deacetylases in PH is now also receiving increasing attention. Epigenetic regulation of chromatin structure is also recognized to influence gene expression in development or disease states. Therefore, a complete understanding of the mechanisms involved in altered gene expression in diseased cells is vital for the design of novel therapeutic strategies. Recent technological advances in DNA sequencing will provide a comprehensive improvement in our understanding of mechanisms involved in the development of PH. This review summarizes current concepts in TF and epigenetic control of cell phenotype in pulmonary vascular disease and discusses the current issues and possibilities in employing potential epigenetic or TF-based therapies for achieving complete reversal of PH.

Transcription factors (TFs) are sequence-specific DNA-binding proteins that control the process of transcription. They contain one or more DNA-binding domains, which help them to attach to specific sequences of DNA adjacent to genes. TFs typically regulate gene expression by binding regulatory DNA elements called enhancers, an event that recruits cofactors, the general transcriptional machinery, and Pol II (RNA polymerase II) complexes to target genes.^1-3 An active enhancer typically binds multiple TFs in a cooperative fashion and regulates transcription from core promoters, often via long-range genomic interactions that involve looping of DNA.³ In addition, TFs can bind directly to core promoter elements in proximity to transcriptional start sites to recruit transcriptional machinery and regulate gene expression.⁴

TFs are classified into two broad categories based on the functions they are involved in: (1) general TFs that are involved in basal transcriptional regulation and (2) regulatory, or gene-specific, TFs that contribute to differential gene expression. Regulatory TFs exert control over just one or a few genes and determine whether the gene is switched “on” or “off” by binding to regulatory sites, which may be located some distance from the coding region. They are crucial in ensuring that genes are expressed in the right cell at the right time and in the right amount, depending on the changing requirements of the organism.^5,6

Growing evidence indicates that TFs require other molecules (e.g., coactivators and repressors, hereafter called “coregulators”) to regulate their activities.⁷ While it was logical to initially consider that there might be distinct use of specific coregulator complexes by distinct classes of TFs or by different members of a DNA-binding TF family, what has emerged is evidence of combinatorial usage.^1,6,8 For example, many of the cofactors initially identified were based on their interactions with nuclear receptors, which have been found to play significant roles in mediating the actions of numerous classes of TFs.⁹ Conversely, specific TFs can use distinct combinations of cofactors, depending on cell type, promoter, DNA binding site, and the actions of various signaling pathways/ligands. Thus, coactivators/corepressors and sequence-specific TFs constitute distinct axes on a matrix for many potential combinatorial interactions that are used in a context-dependent manner to control functional gene expression.^8,9 Recent advances suggest that many human diseases and disorders are associated with misregulation in TFs and TF coregulators, chromatin regulators, and noncoding RNAs (ncRNAs).

Importance of TFs and TF coregulators in the pathogenesis of pulmonary hypertension

Pulmonary hypertension (PH) is a deadly disease characterized by vasoconstriction and abnormal remodeling of pulmonary vessels, leading to a progressive increase in pulmonary artery pressure (PAP), and vascular stiffness, which culminates in right ventricular (RV) failure and premature death.¹⁰ Broadly speaking, upon vascular injury, deregulation of normal cellular processes in the pulmonary vasculature mediates proliferation, differentiation, inflammation, and cell death programs, which all together contribute to the development of PH. Numerous stimuli and pathological conditions (shear stress, hypoxia, oxidative stress, infection, HIV, and others) can result in PH by activating a complex cascade of signaling pathways.^11,12 Ultimately, different signaling cascades converge on a common program targeting the activity of certain TFs.¹³ This leads to activation of a vascular gene program in the nucleus that is manifested finally as the PH vascular phenotype. An understanding of the altered gene expression in diseased cells lays the foundation for novel therapeutic strategies involving manipulation of gene expression. Targeting TFs represents one such innovation that can allow manipulation of gene activation and suppression in a specific fashion.

Strategies targeting single growth factors, cytokines, or receptor tyrosine kinases (RTKs) might exhibit limited efficacy because of the redundancy of the multiple stimuli that activate pulmonary vascular endothelial, smooth muscle, and fibroblast proliferation, as is observed in this progressive disease, which exhibits certain overlapping characteristics with cancer. One strategy to overcome these limitations might be to target multiple RTKs in parallel; however, a broad side-effect profile may be the consequence. A further novel and emerging concept is to target central downstream effector molecules that integrate the multiple signals. Along this line, targeting TFs might hold promise to be effective in interfering with multifactorial disease processes, that is, to combine strong antiproliferative, anti-inflammatory, and proapoptotic effects with a high specificity for activated vascular smooth muscle cells (SMCs), activated adventitial fibroblasts, and degenerative endothelial cells.

Brief review of TFs implicated in PH

There are numerous TFs and transcriptional coactivators that have been implicated in PH and RV dysfunction (Table 1; Fig. 1). In this review, we focus on the TFs Forkhead box O (FoxO) and CBF1/RBP-Jκ (recombination signal binding protein for immunoglobulin kappa J region), hypoxic inducible factors (HIFs), the transcriptional coactivator pyruvate kinase isozyme PKM2, the corepressor CtBP1 (a member of the C-terminal binding protein [CtBP] family), and the Twist family bHLH transcription factor 1 (TWIST1).

Table 1.

Transcription factors in PH

Gene	Full name	Role in PH	Cells/tissue samples investigated in PH	References
PPARG1	Peroxisome proliferator–activated receptor γ	Decreased expression in PH; PPARγ agonists might reverse pulmonary vascular remodeling	IPAH lungs, SuHx-PH	Ameshima et al.;14 Hansmann et al.15
HIF2A	Hypoxia-inducible factor 2α	Hif2a/Epas1 heterozygous mice were fully protected against PH and RVH; HIF2A-mediated upregulation of vasoconstrictors contributes to the development of hypoxic PH	Rodent HPH	Brusselmans et al.16
HIF1A	Hypoxia-inducible factor 1α	Normoxic HIF activation in FHR and human PAs from IPAH; partial deficiency of either HIF1α- or HIF2α-attenuated PAP and RVH	Rodent HPH, FHR-PH, IPAH PASMCs	Bonnet et al.17
NFATc2	Nuclear factor of activated T-cells, cytoplasmic, calcineurin-dependent 2	Activated in CD3+ lymphocytes; NFAT inhibition reverses MCT-induced established PAH	IPAH PASMCs MCT-PH	Bonnet et al.18
STAT3	Signal transducers and activators of transcription 3	Persistent activation of STAT3 was identified in IPAH PAECs	IPAH	Masri et al.19
HES5	Hes family bHLH transcription factor 5	High levels of HES5 in lung tissue correlated with worsening disease severity	IPAH, rodent HPH	Li et al.20
Smad8/Smad9	Mothers against decapentaplegic homolog 8	Abnormal vascular remodeling in Smad8 mutants	Transgenic	Huang et al.21
OCT4	Octamer-binding transcription factor 4	OCT4 isoforms were upregulated in PASMCs from human PAH	IPAH PASMCs	Firth et al.22
NFATc3	Nuclear factor of activated T-cells, cytoplasmic, calcineurin-dependent 3	Activated in neonatal mice exposed to chronic hypoxia; Nfatc3 KO mice do not show PA remodeling	Rodent HPH	Bierer et al.23
KLF5	Krüppel-like factor 5	KLF5 is upregulated in human and MCT-induced PAH	IPAH lungs and PASMCs	Courboulin et al.24
P53	Tumor protein p53	Gene deficiency promotes HPH and vascular remodeling	Rodent HPH	Mizuno et al.25
Lef1	Lymphoid enhancer factor 1	Upregulated in HPH in neonatal rats	Rodent HPH	Xu et al.26
PPARGC1A (PGC-1α)	Peroxisome proliferator-activated receptor gamma coactivator 1α	Upregulated mRNA expression	IPAH blood samples	Mata et al.27
NF-KB/RELA	Nuclear factor κ-B	NF-κB is activated in pulmonary macrophages, lymphocytes, endothelial cells, and PASMCs	IPAH	Price et al.28
GATA6	GATA-binding protein 6	Endothelial GATA-6 deficiency promotes PAH	SSc-PAH, IPAH	Ghatnekar et al.29
c-JUN, c-FOS	Jun proto-oncogene, FBJ murine osteosarcoma viral oncogene homolog	Higher levels of total and phosphorylated c-JUN and c-FOS in the vessel wall	IPAH lungs	Biasin et al.30
FOXO1	Forkhead box O1	FOXO1 expression is downregulated and inactivated; FOXO1 activation reverses the development of PH	IPAH, MCT-PH, SuHx-PH	Savai et al.31
KLF4	Krüppel-like factor 4	KLF4 gene and protein expression is reduced in lungs from patients with PAH; loss of endothelial KLF4 exacerbates pulmonary vessel muscularization	IPAH lungs	Shatat et al.32
MEF2	Myocyte enhancer factor 2	MEF2 transcriptional activity is impaired	IPAH/FPAH PAECs	Kim et al.33
TWIST1	Twist family bHLH transcription factor 1	EndoMT; neointimal and plexiform lesion formation; role in neomuscularization	PAH, MCT-PH, SuHx-PH, Bmpr2+/− rats	Ranchoux et al.34
SLUG	Snail family zinc finger 2	EndoMT	BMPR2 KO PAECs	Hopper et al.35

^NotebHLH: basic helix-loop-helix; EndoMT: endothelial-to-mesenchymal transition; FHR: fawn-hooded rat; HPH: hypoxia-induced PH; IPAH: idiopathic PAH; KO: knockout; MCT-PH: monocrotaline-induced PH; mRNA: messenger RNA; PA: pulmonary artery; PAECs: pulmonary arterial endothelial cells; PAH: pulmonary arterial hypertension; PAP: pulmonary artery pressure; PASMCs: pulmonary artery smooth muscle cells; PH: pulmonary hypertension; RVH: right ventricular hypertrophy; SSc-PAH: systemic sclerosis–associated PAH; SuHX-PH: hypoxia+SU5416–induced PAH (SU5416 is a vascular endothelial growth factor receptor [VEGFR-2] inhibitor).

Figure 1.

Role of transcription factors in the pathogenesis of PAH. Multiple pathological stimuli, such as hypoxia, shear stress, oxidative stress, mitogens, and inflammation (cytokines and chemokines), trigger downstream signaling cascades (MAPK, JNK, PI3K-Akt, JAK-STAT, ERK, PKA, PKG, PKC, etc.), which modulate the recruitment and activation of transcription factors that determine the stimulus-specific transcriptional responses in PAH. ERK: extracellular signal–regulated kinase; JAK-STAT: Janus kinase–signal transducer and activator of transcription; JNK: c-Jun N-terminal kinases; MAPK: mitogen-activated protein kinase; MEK: mitogen-activated protein kinase kinase; PAH: pulmonary arterial hypertension; PI3K-Akt: phosphatidylinositol 3-kinase-Akt; PKA: cAMP (cyclic adenosine monophosphate)-dependent protein kinase, or protein kinase A; PKC: protein kinase C; PKG: cGMP (cyclic guanosine monophosphate)-dependent protein kinase, or protein kinase G.

FoxO TFs

FoxO TFs belong to a family of transcriptional regulators characterized by a conserved DNA-binding domain termed the “forkhead box.”³⁶ FoxO proteins mainly act as potent transcriptional activators by binding to the conserved consensus core recognition motif TTGTTTAC.^37,38 FoxOs, when present in the nucleus and bound to promoters that contain the FoxO consensus motif, can act as transcriptional activators and repressors. In mammals, four FoxO isoforms have been identified: FoxO1, FoxO3, FoxO4, and FoxO6. Despite some redundant functions, genetic loss of individual FoxOs does result in specific phenotypes, suggesting divergent roles of FoxO isoforms.³⁹ FoxOs control various cellular responses, including proliferation, apoptosis, DNA repair, and metabolism, and they have been implicated in vascular structural maintenance.^40,41 FoxO activity is tightly controlled by posttranslational modifications (PTMs), such as phosphorylation, acetylation, methylation, ubiquitination, and binding of protein partners. These modifications alter transcriptional activity, DNA binding, subcellular localization, and protein stability. For example, phosphorylation of FoxO1 at Thr24 and Ser256 create binding motifs for the 14-3-3 scaffolding proteins, causing FoxO1 exclusion from the nucleus and promoting their cytoplasmic ubiquitination and degradation.^37,42

At the 2015 Grover Conference, one of us (SSP) presented evidence that among the FoxO family, FoxO1 is centrally involved in the hyperproliferative and apoptosis-resistant phenotype of pulmonary arterial SMCs (PASMCs), which represents a hallmark of PH. We observed that, in pulmonary vessels and PASMCs of human and experimental PH lungs, FoxO1 expression is downregulated and FoxO1 is inactivated via phosphorylation and nuclear exclusion. Importantly, stimulation of human PASMCs by growth factors and inflammatory cytokines typically involved in the pathogenesis of PH, such as PDGF (platelet-derived growth factor), IGF-1 (insulin-like growth factor), TNF-α (tumor necrosis factor alpha), and interleukin 6, led to downregulation of FoxO1 phosphorylation at Ser256 and Thr24, nuclear exclusion and inactivation, suggesting that FoxO1 is one of the central downstream TFs regulated by several PH stimuli. This postulated central role of FoxO1 in the pathogenesis of pulmonary arterial hypertension (PAH; i.e., group 1 PH) is further supported by in vitro and in vivo loss-of-function studies.³¹ These data indicate that depletion of FoxO1 specifically in SMCs is sufficient to induce PH and also synergizes with the hypertensive effects of hypoxia, resulting in more severe PH. Reconstitution of PASMC FoxO1 activity by an adenovirus carrying a phosphorylation-deficient FoxO1 mutant reversed the features of PAH both in vitro and in vivo. Transfection of PASMCs with FoxO1 fully normalized growth factor–driven abnormalities in proliferation, migration, and apoptosis in vitro. In monocrotaline (MCT)-induced PAH, inhalation of the adenovirus (after full establishment of disease) markedly reduced RV systolic pressure, pulmonary vascular resistance, and RV hypertrophy—the central clinical features of this disease. In parallel, adenovirus-treated lungs showed reversal of vascular remodeling and normalization of vascular cell hyperproliferation and cell cycle abnormalities to favor the quiescent PASMC phenotype.³¹

SSP’s group also demonstrated that the anti-neoplastic drug paclitaxel has a similar ability to reverse the PAH PASMC phenotype in vitro and the clinical features of established experimental PAH in vivo. Both intravenous and inhaled paclitaxel markedly improved hemodynamics, RV hypertrophy, RV function, and vascular remodeling in various experimental PAH models. The finding that several of these pathways converge on FoxO1 as a central downstream effector molecule in hyperproliferative PASMCs offers new options for therapeutic intervention. The preclinical evidence for the efficacy of this approach will have to be assessed further in clinical trials.

Importantly, understanding of other PTMs on FoxOs, such as acetylation and ubiquitination, in the PH disease setting and the impact of FoxOs on other cellular processes driving PH, such as metabolism and inflammation, is essential.¹³ In addition, the effects of FoxOs and FoxO-activating therapies directly on RV hypertrophy and failure should be studied.^13,43 Further characterizing the genes and gene networks regulated by FoxO1 in PAH-PASMCs, Savai et al.³¹ found that FoxO1-regulated genes are associated with cell cycle control (cyclin D1, p27), apoptosis (BCL6, GADD45), and bone morphogenetic protein (BMP) signaling (BMP receptor 2 [BMPR2], ID1, ID3). However, the microarray analysis also revealed changes in several genes that are not prototype target genes of FoxO TFs. An attractive possibility to explain these changes would be that FoxO has effects on chromatin properties more generally. Emerging evidence indicates that FoxO, similar to FoxA, has the capacity to bind to and remodel compacted chromatin.⁴⁴ Thus, in addition to regulating the expression of specific genes, FoxO PTMs, similar to histone PTMs, may regulate overall chromatin structure and participate in the epigenetic regulation of gene expression, which require deeper investigations.

Hypoxic inducible factors (HIFs)

HIFs are TFs and key regulators of the molecular response to hypoxia. The targets of HIFs include genes controlling vascularization, cellular proliferation, migration, and metabolism.^45,46 HIFs have therefore been strongly implicated in the pathogenesis of various forms of PH. Studies have shown that mice that are homozygous for either of the HIF isoforms (complete deletion of HIF1α or HIF2α is embryonic lethal) exhibit attenuated pulmonary pressures and vascular remodeling after experimental exposure to chronic hypoxia.^16,47 Further, recent studies show that conditional deletion of HIF1α in SMCs ameliorates the degree of remodeling in chronically hypoxic mice.^48,49 In addition, Kapitsinou et al.⁵⁰ have recently shown that endothelial HIF2α is required for the development of increased PAP in chronically hypoxic mice. These studies strongly support the importance of better understanding the role of HIFs in the pathogenesis of PH, which includes how they interact with other transcriptional factors, transcriptional coactivators, and repressors.

HIFs are heterodimeric TFs composed of an α-subunit (HIF1α, HIF2α, or HIF3α) and a β-subunit (HIF1β, also called the arylhydrocarbon receptor nuclear translocator [ARNT]). HIF is absolutely required for HIF-target gene activation under hypoxia, because knockdown of HIF1α and/or ARNT significantly reduces or completely blocks hypoxic induction of HIF-target genes. However, recent reports demonstrate that other TFs are also required for HIF-target gene expression. These factors have been found to physically interact with HIF1α or HIF2α in transcriptional complexes on HIF-target gene promoters, demonstrating the possibility that some of these factors may be components of what have been termed HIF enhanceosomes.⁵¹ The finding that HIFs interact with a diverse variety of cotranscriptional activators suggests a mechanism by which hypoxia activates HIF, as well as a number of physiologically and pathologically relevant signaling pathways, to induce HIF-target gene expression.^51-53 In addition, there are examples in which it has been shown that other TFs function to selectively enhance expression of HIF1α or HIF2α target genes, demonstrating that distinct HIF1α or HIF2α enhanceosome complexes exist and are composed of different sets of HIF-specific cotranscriptionally activating factors on the promoters of HIF-target genes.

Incorporation of multiple factors in enhanceosome complexes allows regulation of target gene transcription in a way that is highly sensitive to a variety of stimuli and can result in controlled levels of target gene expression under a variety of conditions. In addition, these complexes can allow for cooperation between factors acting within the same or overlapping signaling pathways and allow for cooperative or synergistic activation of target gene transcription in a context-specific manner (Fig. 1).⁵¹

It is noteworthy that HIF transcriptional effects are modulated by epigenetic factors described in more detail below. It is now clear that HIF activates its target gene expression by recruiting histone acetylases CBP and p300 via their C- and N-terminal activation domains.⁵⁴ Of importance in the similarity between many of the identified HIF cotranscriptional activators is their ability to bind and recruit the general coactivator proteins CBP and p300. It has been suggested, in fact, that HIF-dependent CBP and p300 recruitment contributes less to total CBP/p300 bound to HIF-target gene promoters than other coactivators, such as STAT3. These findings raise important questions for the future as to the specific functions of HIFs versus those of coactivators in transcriptional activation of HIF-target genes in PH and a myriad of other diseases including cancer.⁵¹

Transcription corepressor RBPJ/CSL

The Notch signaling pathway is a versatile regulator of cell fate decisions and plays an essential role for vascular development. Notch signaling is activated upon cell-to-cell contact as a result of interactions between Notch receptors and their ligands (Delta or Jagged).⁵⁵ Upon ligand binding, the Notch receptors undergo proteolytic processing, which involves ADAM metalloproteases and the γ-secretase complex, resulting in the intracellular domain of Notch (NICD) being released from the cell membrane and shuttling to the nucleus. This process activates the major downstream nuclear target for Notch, the CSL family of DNA-binding factors (CBF1/RBP-Jκ in mammals, Su(H) in Drosophila and Lag-1 in Caenorhabditis elegans). In the absence of active Notch, RBPJ is thought to be primarily a transcriptional repressor that exists in complexes with corepressors.⁵⁶ When bound to the active NICD, RBPJ recruits a coactivator complex, including a mastermind homolog (MAML1–3 in mammals), activates transcription of genes containing RBPJ binding sites, and drives a complex transcriptional program with phenotypic effects.^57,58

Notch signaling plays a critical role in controlling proliferation and differentiation of PASMCs. Upregulated Notch ligands and Notch3 receptors in PASMCs have been reported to promote the development of pulmonary vascular remodeling in patients with PAH and in animals with experimental PH.²⁰ At the 2015 Grover Conference, one of us (JY) explained how hypoxia, both acute and chronic, regulates Notch signaling and how hypoxia-mediated Notch signaling regulates cytosolic free Ca²⁺ concentration ([Ca²⁺]cyt).^59,60

Because Notch3 and Notch3 NICD expression is also increased in hypoxia-induced PH (HPH) in mice, JY’s group investigated the role of Notch signaling in the responses to acute and chronic hypoxia. Intriguingly, inhibition of Notch signaling with DAPT, a γ-secretase inhibitor (GSI), significantly attenuated the amplitude of acute alveolar hypoxia-induced increase in PAP in isolated perfused/ventilated mouse lungs. In addition, DAPT reversed the development of HPH in mice,^20,60 suggesting that Notch signaling is involved in the responses of the pulmonary vasculature to acute and chronic hypoxia.

Elevated [Ca²⁺]cyt is implicated in stimulating vascular SMC proliferation and inducing vasomotor tone and hence vasoconstriction. In PASMCs, [Ca²⁺]cyt can be increased by Ca²⁺ release from the intracellular stores and Ca²⁺ influx through plasmalemma Ca²⁺ channels.⁶¹ In PASMCs, increased [Ca²⁺]cyt may occur by activation of various Ca²⁺-permeable channels, including (1) activation of receptor-operated (or ligand-gated) Ca²⁺ channels (ROCs) that are opened, independent of membrane potential, by agonist-mediated activation of membrane receptors; (2) activation of store-operated Ca²⁺ channels (SOCs) that are opened, independent of membrane potential, by depletion of Ca²⁺ from intracellular stores; (3) activation of voltage-dependent Ca²⁺ channels that are opened by membrane depolarization; (4) activation of inositol triphosphate (IP3) receptor–mediated Ca²⁺ release from the IP3-sensitive sarcoplasmic reticulum (SR); and (5) activation of ryanodine receptor–mediated Ca²⁺ release from ryanodine-sensitive SR. Previous work demonstrated that the resting [Ca²⁺]cyt is increased and that expression and activity of ROCs and SOCs are both upregulated in PASMCs isolated from patients with idiopathic PAH (IPAH).^62-64 Furthermore, it was found that extracellular Ca²⁺ induces a large increase in [Ca²⁺]cyt in IPAH PASMCs and functionally upregulated extracellular Ca²⁺-sensing receptor (CaSR) is involved in the enhanced Ca²⁺ influx and proliferation in IPAH PASMCs.⁶⁵

Along similar lines, in HPH, increased expression of SOCs and enhanced store-operated Ca²⁺ entry (SOCE) was shown to contribute to increased [Ca²⁺]cyt.^64,66 SOCE is known to be important for cell proliferation and vascular remodeling in PH, and studies from JY’s lab have demonstrated increased expression of several proteins involved in SOCE, such as canonical transient receptor potential (TRPC) channels (TRPC3, TRPC6), stromal interaction molecule 2 (STIM2), and Orai2, in patients with PH.^62,64,67,68 Understanding the molecular mechanisms that regulate [Ca²⁺]cyt and PASMC proliferation is critical in understanding the pathogenesis of HPH.

Hypoxic pulmonary vasoconstriction (HPV) is an important physiological response that optimizes the ventilation/perfusion ratio.⁶⁹ In order to determine the involvement of Notch signaling in acute HPV, isolated perfused/ventilated mouse lungs and human PASMCs were treated with Notch ligand Jagged-1. Treatment with Jagged-1 resulted in a significant increase in PAP, demonstrating that Jagged-1 can mimic acute hypoxia to cause pulmonary vasoconstriction. In addition, interestingly, both short-term (15–30 min) and long-term (48 hours to days) activation of Notch by Jagged-1 enhances SOCE, which is closely associated with the level of NICD in PASMCs. Importantly, the short and rapid response to activation of Notch signaling in PASMCs can be explained by NICD functional interaction with TRPC6, leading to TRPC6 activation, enhanced SOCE, and an increase in [Ca²⁺]cyt in PASMCs. Moreover, employing several genetic and pharmacological approaches, Sylvester and colleagues⁶⁹ elegantly demonstrated that TRPC6 is required for SOCE and acute HPV in mice and that inhibition of Notch signaling markedly attenuates hypoxia-enhanced SOCE and acute HPV. However, the marked decrease of acute HPV mediated by Notch inhibition is not observed in TRPC6-deficient (TRPC6^−/−) mice, confirming the noncanonical form of Notch signaling, that is, NICD-TRPC6 interaction under acute hypoxic exposure.^59,60

Chronic hypoxia causes vascular remodeling, which is central to the pathogenesis of HPH.⁷⁰ As it was previously shown that Notch3 and SOCs (TRPC6, Orai1/2) are upregulated in PH induced by chronic hypoxia, it is interesting to decipher whether and how Notch signaling regulates SOCs or TRPC6 specifically in chronic hypoxic situations. Strikingly, JY’s group found that Notch signaling activation transcriptionally upregulates TRPC6, Orai1/2, and STIM2 via increased binding of RBP-Jκ with NICD, leading to enhanced SOCE under chronic hypoxia. Finally, pharmacological inhibition of Notch signaling attenuates TRPC6 expression, hypoxia-enhanced SOCE, and the development of HPH.^59,60 These studies suggest a novel dual role for Notch signaling on TRPC6 via direct functional interaction with TRPC6 and transcriptional upregulation of TRPC6, thereby maintaining increased levels of [Ca²⁺]cyt, that plays an important role in both acute HPV and HPH. Furthermore, recent work from this group suggests that Notch signaling regulates not only [Ca²⁺]cyt but also Ca²⁺ influx via regulation of CaSR function. Activation of Notch signaling by Jagged-1 enhances CaSR function and thus regulates Ca²⁺ influx and [Ca²⁺]cyt. Together, these data demonstrate the essential role of Notch-RBPJ signaling in initiation and perpetuation of HPH.

Transcriptional coactivator PKM2

PKM1 and PKM2 are pyruvate kinase isoforms expressed in different types of cells and tissues. Pyruvate kinase regulates the final rate-limiting step of glycolysis by catalyzing the transfer of a phosphate group from phosphoenolpyruvate to adenosine diphosphate to produce pyruvate and adenosine triphosphate (ATP),⁷¹ which suggests their major functions in glycolysis. In addition to its well-known role in glycolysis, PKM2 has also been reported to be involved in other cellular functions. Importantly, PKM2 has recently been found to translocate into the nucleus upon mitogenic and oncogenic stimulation.^72,73 In the nucleus, PKM2 functions as a transcriptional coactivator and a protein kinase that phosphorylates histones, highlighting the crucial role of PKM2 in the epigenetic regulation of gene transcription that is important for the G1-S phase transition and the “Warburg effect” (which states that most cancer cells produce energy by a high level of glycolysis, followed by lactic acid fermentation).^74,75 PKM2 functions as a coactivator that stimulates HIF-1 transactivation of target genes encoding GLUT1, LDHA, and PDK1 in cancer cells.⁷⁶ In addition to the crucial role of PKM2 in the G1-S phase, it phosphorylates important cell cycle regulators, such as the spindle checkpoint protein Bub3, to regulate chromatid segregation and the mitotic checkpoint during mitosis, and myosin light chain 2 (MLC2, encoded by MYL2) to initiate cytokinesis, leading to enhanced and governed tumor cell proliferation.⁷⁷ In line with its functions, PKM2 expression has been found to be increased in all cancer types examined.^78,79

At the 2015 Grover Conference, one of us (KS) briefly reviewed his group’s recent findings regarding the metabolic state of pulmonary arterial adventitial fibroblasts (PAAFs) isolated from hypertensive calves and humans with IPAH, the metabolic functions of PKM2, and the mechanisms that regulate PKM2 expression and its glycolytic enzymatic activity in these cells. Previous studies from the group suggested that some of the earliest pathophysiological changes in PH are observed in the pulmonary artery adventitia and include robust proliferation of fibroblasts and accumulation of macrophages.^80-83 It is increasingly recognized that increased proliferation and inflammatory activation can occur in the context of changes in cellular metabolism toward aerobic glycolysis (the Warburg effect), and these metabolic changes have been reported in PH in pulmonary arterial endothelial cells (PAECs), PASMCs, and, most recently, PAAFs and have formed the basis for a “metabolic theory of PH.”^84,85 However, the mechanisms that link metabolic reprogramming to inflammatory and proliferative pathways in PH remain largely unknown.

Elegant work employing RNA-Seq transcriptome profiling and nuclear magnetic resonance– and mass spectrometry–based metabolomics analysis showed significant metabolic reprogramming to aerobic glycolysis in PH-fibroblasts (PH-fibs). Glycolytic changes included sugar phosphates, ¹³C-glycolysis signature, ¹³C-lactate, d-glucose and glucose uptake, glyceraldehyde 3-phosphate, pyruvate, and lactate production. In addition, PH-fibs were characterized by alterations of redox homeostasis, as indicated by increased amounts of glutathione (GSH) and glutathione disulfide (GSSG), decreased GSH/GSSG ratios, evidence of utilization of the NADPH (nicotinamide adenine dinucleotide phosphate)-generating pentose phosphate pathway, and increased phosphogluconate and ribose phosphate (K. Stenmark, unpublished data). These data provide strong evidence that in vivo and in vitro aerobic glycolysis is a metabolic adaptation in PAAFs in the setting of chronic hypoxia as well as in IPAH. Metabolic data here suggest that, in a Warburg-like fashion, incomplete glucose oxidation affords PH-fibs building blocks to support the anaplerotic reactions (amino acid anabolism, fatty acid synthesis, urea cycle, gluconeogenesis) necessary to sustain proliferation, counteract oxidative stress, and promote resistance to apoptosis.

Importantly, these authors found an isoform switch from PKM1 to PKM2 in PH-fibs. Knockdown of PKM2 decreases PH-fib proliferation and lactate production, providing a link between transcriptional regulation and the metabolic status of the cell. The PKM1-to-PKM2 switch (i.e., PKM2 expression) is regulated at the level of transcribed PKM pre–messenger RNA (mRNA) by splicing factors. Polypyrimidine tract–binding protein 1 (PTBP1) is upregulated in PH-fibs by microRNA 124 (miR124) and subsequently binds to splicing signals that flank PKM exon 9, repressing the inclusion of exon 9 and thus promoting an enhanced expression of the PKM2 isoform. Knockdown of PTBP1, miR124 mimics, and histone deacetylase (HDAC) inhibitors reverses the PKM2-PKM1 switch in PH-fibs and reduces lactate production, suggesting that epigenetic regulation is an important mechanism in controlling PKM splicing in PH.

Transcriptional corepressor CtBP1

In the context of increased glycolysis, where increases in NADH/NAD (ratio of reduced to total nicotinamide adenine dinucleotide) are frequently observed, the functions of the CtBP family members, traditionally considered transcriptional corepressors, must also be considered. CtBPs comprise a dimeric family of transcriptional repressors encoded by two paralogous genes (CTBP1 and CTBP2) that have major roles in animal cell development and cancer and have been suggested to be “sensors” of the metabolic state of cells that can modulate many aspects of cell phenotype.⁸⁶ Overexpression of CtBP is observed in a number of cancers, including prostate, ovarian, colon, and breast.^74,87,88 Studies in cancer and the nervous system revealed that CtBPs promote cellular survival primarily through repression of Bcl-2 family members and other proapoptotic molecules (PERP, Bax, Bik, Puma, p21, and Noxa), as well as tumor suppressors (e.g., p16INK4a, p15INK4b).^86,88,89 Importantly, there is emerging evidence that CtBPs may also function to modulate the inflammatory response.⁹⁰ Further, a recent genome-wide analysis in breast cancer provided a more comprehensive description of CtBP repression targets.^88,91 CtBP targets were categorized into three main categories: genes that regulate cell renewal and pluripotency, genes that control genome stability, and genes that regulate epithelial differentiation and prevent epithelial-to-mesenchymal transition, all processes of potential importance in PH pathogenesis.^88,91 At the 2015 Grover Conference, one of us (KS) presented evidence for CtBP1 activation and control of PAAF proliferation, apoptosis resistance, and inflammatory gene expression in PAAFs derived from animals and humans with PH.

TF TWIST1

TWIST1 was recently demonstrated to play a crucial role in the pulmonary vascular remodeling responsible for human and experimental PAH through a process called endothelial-to-mesenchymal transition (EndoMT).^34,92 Interestingly, BMPR2 expression negatively correlated with Twist1 expression. The Twist proteins (Twist1 and Twist2) are highly conserved developmental proteins with key roles for transcriptional regulation in mesenchymal cell lineages. They belong to the superfamily of basic helix-loop-helix (bHLH) proteins and exhibit bifunctional roles as both activators and repressors of gene transcription. The Twist proteins are expressed at low levels in adult tissues but may become abundantly reexpressed in cells undergoing malignant transformation. This observation prompted extensive research on the roles of Twist proteins in cancer progression and metastasis. Twist1 suppresses BMP-induced osteoblast differentiation, and downregulation of endogenous Twist1 enhances BMP signaling. Maximal inhibition of BMP signaling was observed when Twist1 was bound to E47, which markedly enhanced the stability of Twist1. Co-immunoprecipitation assays revealed that Twist1 formed a complex with Smad4 and HDAC.⁹³

Very recent studies also indicate a novel role for Twist1 as a potential regulator of adipose tissue remodeling and inflammation. Several studies have suggested that Twist genes are important determinants of obesity, fat distribution, and remodeling capacity of different adipose depots. Moreover, Twist1 expression is strongly correlated with body mass index and insulin resistance in humans.⁹⁴ Interestingly, mice overexpressing Twist1 mutants in cardiomyocytes developed pathological cardiac remodeling. Twist1 may play a growth-inhibitory role in the postnatal heart, and phosphorylation may release their inhibitory effects, leading to hypertrophy.⁹⁵ Interestingly, the genes involved in oxidative phosphorylation and the Krebs (citrate) cycle were found with reduced levels in the hearts of mice overexpressing Twist1 mutants, suggesting that decreased metabolism may result in heart failure. Twist1 play also critical roles in diverse developmental systems, including myogenesis. Several experiments have demonstrated its role in inhibition of muscle cell differentiation. Overexpression of Twist1 can reverse muscle cell differentiation in the presence of growth factors.⁹⁶ Hence, Twist1 is potentially involved in all pulmonary and systemic abnormalities found in PAH. Regarding EndoMT as responsible for neointima and plexiform lesion formation and distal neomuscularization, Hopper et al.³⁵ also demonstrated that another TF might be involved. They showed in vitro that loss of BMPR2 in PAECs leads to heightened expression of the chromatin architectural factor high-mobility group AT-hook 1 (HMGA1), which increases the TF Slug. HMGA1 binds to DNA and may promote binding of additional pro-EndoMT TFs.³⁵

Targeting TFs: promising new strategies for PH therapy?

These studies strongly support the idea that modulation of TF and TF coregulator activity affects several pathological processes underlying PH and hence offers the possibility of future therapeutic interventions. Hence, the design of therapeutic agents targeted at pro–pulmonary hypertensive TFs regulating the initial expression should be the ultimate goal. Another approach is promoting the activity of anti–pulmonary hypertensive TFs. This will provide a protective mechanism to the cells in order to prevent further hypertensive derangement of gene expression. However, TFs are often considered chemically intractable because they lack surface involutions and hydrophobic pockets for high-affinity binding of small molecules and nuclear localization. Nonetheless, compounds have been developed to target TFs at different levels of function, including DNA binding, protein-protein interactions, and epigenetics. Direct inhibition of TF expression (via RNA interference or microRNAs [miRs]) and DNA binding (via oligodeoxynucleotide or pyrrole imidazole polyamides) are progressing into clinical trials with safer and more effective delivery systems. As mentioned above, miR124 mimics are useful tools to regulate PKM activity and thereby PH phenotype (K. Stenmark, unpublished data). For these approaches to enter the clinic arena, novel delivery systems must be developed.

Novel approaches in drug design that will selectively affect particular TFs and suppress their activity can be of great therapeutic interest. TFs have ligand-binding, dimerization, transeffector, DNA-binding, nuclear-localization, and regulatory domains, which may be targeted directly by small-molecule drugs. Other promising alternatives include artificial TF mimics, which consist of normal TF domains conjugated with synthetic compounds targeting DNA-binding domains or regulatory domains.

In addition, we need to carefully assess the effects of TFs-therapeutic targeting on other tissues. For example, modulating Notch signaling with GSIs in the early GSI trials for Alzheimer’s disease, which led to widespread gastrointestinal toxicity,^97,98 underscores that targeting of TFs has to be approached with a great deal of precision. The potential crosstalk of Notch signaling with several other signaling mechanisms, such as TRPC6, opens up interesting possibilities for combining Notch modulation with other therapeutic strategies. Indeed, the combined inhibition of Notch and TRPC6 signaling could be a novel and safe therapeutic approach.

Epigenetics: definition and mechanisms

Epigenetics is generally defined as heritable changes in gene activity and expression that occur without alteration in DNA sequence.⁹⁹ Epigenetic regulation of chromatin structure is fundamental to establish and maintain cell type–specific gene expression during development and disease states.

Epigenetic mechanisms tightly regulate chromatin accessibility. Double-stranded DNA does not exist naked in the nucleus of the eukaryotic cell but is packaged into a highly organized nucleoprotein complex known as chromatin. The basic unit of chromatin (a protein-DNA complex) is the nucleosome, which comprises 147 base pairs of DNA wrapped around an octameric histone core containing two copies each of the histones H2A, H2B, H3, and H4.¹⁰⁰ Overall, the higher-order chromatin structure facilitates the packaging, organization, and distribution of DNA but negatively affects several fundamental biological processes, such as transcription, replication, and DNA repair, by restricting the accessibility for multiprotein complexes.¹⁰⁰ Chromatin adopts different structural conformations, depending on the epigenetic modifications that occur in the DNA and histones.¹⁰⁰ Changes in the chromatin status of specific genes can lead to their repression or activation. Regulation of chromatin structure involves several interconnected mechanisms, such as (1) DNA methylation, (2) ATP-dependent chromatin remodeling, (3) histone PTMs, (4) ncRNAs, (5) replacement of canonical histones with histone variants, and (6) organization within the three-dimensional nuclear architecture (Fig. 2).¹⁰¹

Figure 2.

Epigenetic mechanisms in gene regulation. (1) DNA methyltransferases (DNMT) and the ten-eleven translocation (TET) methyl cytosine dioxygenase family modulate DNA methylation levels. (2) Chromatin remodeling complexes move, eject, or restructure nucleosomes. (3) Histone modifications are regulated by the counteracting activity of histone acetyltransferases (HATs) and histone deacetylases (HDACs) or lysine methyltransferases (KMTs) and lysine demethylases (KDMs). Acetylation is recognized by bromodomain-containing proteins and methylation by PHD, chromo, WD40, Tudor, double/tandem Tudor, MBT, ankyrin repeats, and PWWP domains. (4) Noncoding RNAs (long noncoding RNAs and microRNAs) affect gene expression by gene silencing or inhibition of protein translation. (5) Histone variants; these mechanisms or processes are the key regulators of genome-environment interactions. K-Ac: lysine acetylation; K-Me: lysine methylation; MBD: methyl CpG (cytosine-phosphate-guanine)-binding domain.

Why epigenetic alterations might be involved in PH

Despite our progressive understanding of the pathogenesis of PH and recent therapeutic advances, PH remains a fatal disease. PH is a complex disease with multifactorial etiology mediated by the interplay of genetic background, epigenetic changes, and exposure to injurious events such as viruses, drugs, toxins, hypoxia, and inflammation, which may explain the great variability in disease susceptibility. In PAH, the key genetic determinants, such as low-penetrance dominant BMPR2 mutations, account for ∼70% in familial and only ∼15% in idiopathic PAH cases.¹⁰² In a suitable genetic background, the interplay of epigenetics and injurious events may augment the severity of the disease, aggravate vascular remodeling and worsen clinical outcome.¹⁰³

We presume that persistence of hypertensive stimuli might disrupt cellular homeostasis and alter the existing epigenetic state (quiescent) of the pulmonary vascular cells and reestablish a new epigenetic signature that favors a disease phenotype exhibiting uncontrolled cell proliferation (Fig. 3). After acquisition of the disease phenotype, the pulmonary vascular cells contributing to the vascular remodeling process exhibit stable proproliferative, promigratory, antiapoptotic, proinflammatory, and profibrotic vascular cell phenotypes (Fig. 3).¹⁰⁴ Notably, we and others have observed that vascular cells (PAECs, PASMCs, PAAFs) isolated from hypertensive lung vessels and cultured ex vivo outside their vascular microenvironment maintain a hyperproliferative, apoptosis-resistant phenotype for a longer time than control cells¹⁹ and also maintain higher mesenchymal transitional potential,^34,35,105 demonstrating the presence of heritable and sustained phenotype.

Figure 3.

Phenotype alterations in pulmonary arterial hypertension (PAH). Pathological stimuli-induced recruitment of transcription factors and alterations in epigenetic mechanisms involves DNA methylation, histone modifications, and microRNAs. These epigenetic alterations in the gene expression programs facilitate the normal vascular cells to exhibit increased proliferation, migration, apoptosis, inflammatory gene expression, and increased mesenchymal transition and collagen extracellular matrix remodeling, as commonly observed in PAH pulmonary arteries. The persistence of pathological stimuli further establishes a phenotypic memory in PAH vascular cells that largely contributes to the development of complex vascular lesions during progressive vascular remodeling process in PAH.

Summary of epigenetics-related findings in PH/PAH

Emerging evidence suggests that the pathogenesis of PH is influenced by abnormalities in epigenetic mechanisms such as DNA methylation, HDAC activity, and miRs.¹⁰³

MicroRNA dysregulation in the pathogenesis of PH

MicroRNAs are small ncRNAs that negatively regulate gene expression by binding to the 3′ untranslated region of target mRNAs, thereby promoting mRNA degradation or inhibiting translation. Profiling of the miR expression in rodent models of PH and in human pulmonary arteries from PAH patients indicated dysregulation of several miRs, including miR204, the mir17/92 cluster, the miR143/145 cluster, and others.¹⁰³ To fine-balance these dysregulated miRs, different strategies, such as miR mimics or antagomirs, were employed and demonstrated to reverse the vascular remodeling in PH experimental models. Because of space limitations, we do not provide an extended analysis here, as the topic is covered in greater depth elsewhere.^103,106

Altered DNA methylation patterns in PH

DNA methylation is one of the best-studied epigenetic modifications occurring in mammals, almost exclusively on the 5′ position of the pyrimidine ring of cytosines in the context of CpG (5′-cytosine-phosphate-guanine-3′) dinucleotides. This DNA modification is mediated by DNA methyltransferases and generally results in tight packing of DNA and histones (heterochromatin) and in the long-term silencing of transcription. DNA methylation may result in gene silencing that can be propagated to daughter cells.

The influence of nongenetic factors in autoimmune and inflammatory diseases has been demonstrated in studies of genetically identical (monozygotic [MZ]) twins who show a variable degree of discordance with respect to different phenotypic traits, including susceptibility to these diseases. MZ twins show a concordance rate for different autoimmune diseases ranging from 12% for rheumatoid arthritis up to 70% for psoriasis. There is substantial epigenetic variation between MZ twin pairs. This variation increases with age and if twins live in different environments. These environment-host interactions might result in such severe epigenetic changes that the epigenetic homeostasis can no longer be maintained, leading to aberrant gene expression. Aberrant gene expression in specific cells of the immune system most likely contributes to the loss of immune tolerance, inflammation, and autoimmunity, which is characterized by the failure of an organism to recognize self-antigens. While a large number of candidate gene and genome-wide studies analyzing the DNA methylation profiles have been published for inflammatory and autoimmune diseases such as type 1 diabetes, multiple sclerosis, and systemic lupus erythematosus,¹⁰⁷ evidence for the implication of DNA methylation modifications in PAH is currently scarce.

Archer´s group¹⁰⁸ has identified a methylation-induced attenuation of Sod2 (mitochondrial superoxide dismutase 2) expression as an epigenetic mechanism in an animal model of spontaneous and heritable PAH (fawn-hooded rats). This may be relevant for clinical PAH because SOD2 is also downregulated in human PAH. The epigenetic downregulation of SOD2 impairs H₂O₂-mediated redox signaling, activates HIF-1, and creates a proliferative, apoptosis-resistant state. Both the mechanism of SOD2 downregulation and its consequences parallel those discovered in human cancer.¹⁰⁹

Moreover, DNA methylation for the measurement of a PAH marker has a number of attractive features. First, DNA molecules are very stable and, in contrast to mRNA and many proteins, can survive harsh conditions for long periods of time. Second, unlike proteins, nucleic acid can be amplified by polymerase chain reaction and related techniques, thus allowing measurements on small amounts of test sample.¹¹⁰ Moreover, blood levels of most currently used protein biomarkers are rarely increased in the early stage of the disease. Consequently, most existing blood protein biomarkers are of little value in either screening for or aiding the early diagnosis of PAH. On the other hand, aberrant methylation of the promoter regions of multiple genes may exist in both early and advanced PAH. As proof of principle, FP’s group^111,112 published an epigenetic study showing that DNA methylation alterations in the granulysin (GNLY) gene in explanted lungs and in peripheral blood mononuclear cells allow discrimination between pulmonary veno-occlusive disease (PVOD) patients (PVOD is a subgroup of PH with worse prognosis and risk of developing severe pulmonary edema with specific PAH therapy) and PAH patients.

At the 2015 Grover Conference, FP presented his ongoing work about the identification of a DNA methylation-based PAH signature. His group analyzed the promoters of protein-coding sequences, miRs, and long noncoding RNA in the genomic DNA from cultured PAECs isolated from IPAH and heritable PAH (BMPR2 mutation carriers) patients and controls. Enrichment analyses demonstrated that the differentially methylated promoters clustering in controls and PAH groups were enriched in pathways, networks, and functions related to metabolism, cardiovascular system development and function, cellular development, skeletal and muscular system development and function, cell morphology, cellular function and maintenance, cell death and survival, cell-mediated immune response, cellular development, cancer, and organismal injury and abnormalities (F. Perros, unpublished data). Further analysis of these data will enhance the discovery of novel regulatory pathways in PAH that require innovative therapy.

Alterations in histone levels in PAH

Linker histones (H1) have been found to play an important role in transcription, gene regulation, protein-protein interactions, and innate immunity and are associated with complexes to downregulate proinflammatory genes. Talati et al.¹¹³ employed histology-based matrix-assisted laser desorption ionization–mass spectrometry (MALDI-MS) and identified significant increases, compared to controls, in two fragments of histone H1 in the pulmonary arteries from the explanted lungs of the IPAH patients. However, a reduction in the nuclear histone H1 and an increase in the cytoplasmic H1 were observed in IPAH cells. The decreased nuclear H1 contributes to the less compact nucleosomal pattern in IPAH, and this, in turn, contributes to the increase in nucleosomal repeat length and ultimately to changes in transcription.¹¹³

Role of HDACs in PAH

Two different families of enzymes mediate the balance between the acetylated and deacetylated states of histone or nonhistone proteins: histone acetyltransferases and HDACs. The role of aberrant HDAC activity and histone acetylation has been strongly implicated in PAH pathogenesis on the basis of the promising therapeutic benefits observed upon the application of small-molecule HDAC inhibitors in different animal models of PH. Treatment with pan-HDAC inhibitors (suberoylanilide hydroxamic acid [vorinostat]) and class I selective inhibitors (MGCD0103, valproic acid) both attenuated and reversed the development of hypoxia- and MCT-induced PH in a rodent model.^114,115 Table 2 summarizes all research findings that reported the aberrant expression and activity of HDACs and the therapeutic effects of pan-HDAC inhibition. Moreover, Li et al.⁸³ demonstrated that class I HDAC inhibition reverses stable proinflammatory phenotype of PH. This study suggests that the persistent proinflammatory phenotype exhibited by activated PAAFs was associated with abnormal activity of HDACs, specifically, class I HDACs.⁸³ Importantly, catalytic inhibition of class I HDACs using the selective class I HDAC inhibitor apicidin significantly suppressed production of proinflammatory mediators by PH-fibs and caused a marked reduction in the ability of PH-fibs to induce monocyte migration and activation.⁸³ Using Bmpr2^+/− mice, Soon and coworkers¹²² recently demonstrated that BMPR2 deficiency promotes an exaggerated inflammatory response. In their study, BMPR2 deficiency in mouse and human PASMCs was associated with increased phospho-STAT3 and loss of extracellular superoxide dismutase (SOD3). Interestingly, the proinflammatory effect of constitutive BMPR2 haploinsufficiency could not be mimicked by short-term BMPR2 deficiency. The authors suggested that a potential mechanism is epigenetic modification of the promoter regions of genes as a consequence of chronic loss of BMPR2. Accordingly, they demonstrated that the loss of SOD3 expression in Bmpr2^+/− cells was reversed by incubation with trichostatin A.^122,123

Table 2.

Histone deacetylases in PH: published evidence

Species	Model/tissues/cells	Summary	Reference
Rat	PAB; RV	Suppression of HDACs (TSA) worsens RV dysfunction after PAB	Bogaard et al.116
Bovine	Hypoxia; adventitial fibroblasts	Alterations in class (I) HDAC1/2/3 expression and activity contribute to a proinflammatory phenotype of PH-fibroblasts and was attenuated by application of HDAC inhibitors (SAHA, apicidin)	Li et al.83
Human/bovine/rat		Increased expression of HDAC1/4/5 in IPAH lung homogenates, HI PH lungs and RV; HDAC inhibition (SAHA, VPA) attenuated the development of HI PH and hyperproliferative phenotype of PH-fibroblasts and R-cells from hypertensive bovine pulmonary arteries	Zhao et al.115
Human	IPAH; lungs
Bovine	Adventitial fibroblasts, R-cells
Rat	Hypoxia; lungs, RV
Rat	Hypoxia; RV, PASMCs	Selective class I HDAC inhibitors (MGCD0103 and MS-275) suppresses HI PH and RVH in a preclinical model of hypoxia-mediated PH	Cavasin et al.114
Ovine (sheep)	Serum; PASMCs	Inhibition of class I and II HDACs by apicidin and HDACi VIII suppressed serum-induced proliferation and migration of newborn PASMCs	Yang et al.117
Rat	PDGF; PASMCs	Sodium butyrate inhibits PDGF-induced proliferation and migration in PASMCs through Akt inhibition	Cantoni et al.118
Rat	PDGF; PASMCs	Class I HDAC inhibitor (MC1855), but not class II HDAC inhibitor (MC1575), counteracts PDGF-induced proliferation and migration of PAH PASMCs	Galletti et al.119
Bovine	Hypoxia; adventitial fibroblasts	Pan-HDAC inhibitors (SAHA, apicidin) rescued the HI downregulation of antiproliferative miR-124 expression in hypoxic adventitial fibroblasts	Wang et al.120
Rat	MCT, SuHx-PH, PAB	Pan-HDAC inhibitor TSA has no therapeutic or beneficial effects in SuHx-PH	De Raaf et al.121
Human	IPAH PAECs, MCT, SuHx-PH	Increased nuclear accumulation of class IIa HDACs HDAC4 and HDAC5 impaired MEF2 activity in PAH PAECs; pharmacological inhibition of class IIa HDACs using MC1568 rescued the impaired MEF2 transcriptional activity in PAH PAECs and also reduced PAH PAEC proliferation and migration; no evidence of myocardial fibrosis in MCT and SuHx groups upon HDAC inhibition	Kim et al.33

^NoteHDAC: histone deacetylase; HI: hypoxia-induced; IPAH: idiopathic PAH; MCT: monocrotaline; MEF2: myocyte enhancer factor; PAB: pulmonary artery banding; PAECs: pulmonary artery endothelial cells; PAH: pulmonary arterial hypertension; PASMCs: pulmonary artery smooth muscle cells; PDGF: platelet-derived growth factor; RV: right ventricle/ventricular; RVH: RV hypertrophy; SAHA: suberoylanilide hydroxamic acid; SuHX-PH: hypoxia+SU5416–induced PAH (SU5416 is a vascular endothelial growth factor receptor [VEGFR-2] inhibitor); TSA: trichostatin A; VPA: valproic acid.

Collectively, these findings also highlight the need for identification of deregulated isoforms in the PAH setting and to dissect the “histone code” that can mediate a unique cellular response, such as the proinflammatory/proproliferative PAAF phenotype, the hyperproliferative PASMC phenotype, and the angioproliferative PAEC phenotype. In order to elucidate the histone acetylation code underlying the PAH phenotype in human setting, SSP’s group employed three approaches: (1) investigate the expression, regulation, and functional role of histone acetylation modifiers, (2) explore the therapeutic potential of isoform-selective modulation of deregulated HDAC isoforms in PAH disease phenotype reversal ex vivo and in vivo, and (3) employ an unbiased next-generation sequencing–based approach to identify the disease-specific epigenetic signatures (histone modification patterns and transcriptome profiles) and the regulatory network that facilitates the vascular cells to acquire and sustain the disease phenotype. Interestingly, aberrant gene and protein expression profiles of class I HDAC isoforms across a variety of cardiopulmonary samples from human PAH was documented. The screening also revealed an organ- and cell-specific expression pattern of these isoforms between IPAH patients and donors. The stable disease phenotype exhibited by ex vivo isolated IPAH PAAFs was associated with aberrant expression and activity of class I HDACs isoforms. Importantly, integrative analysis of vascular cell–specific epigenomics data provides insights into the mechanistic link between recruitment of histone modifiers and the regulatory network of histone modifications that cooperatively regulates aberrant transcriptional responses in PAH (S. S. Pullamsetti and P. Chelladurai, unpublished data).

Epigenetic modulation: promising new strategies for PH therapy?

As miR inhibitors are currently tested for different indications in clinical trials, their clinical development as a new therapy for PH might therefore be imminent. It will be important to define the most suitable miR(s) to target and the optimum delivery strategy to address the pulmonary vascular compartment. A systematic evaluation of compartment- and cell-specific HDACs in experimental and human PAH and of the isoform-selective functions/role of HDACs in pulmonary vascular and RV remodeling is missing. In addition, detailed study of HDAC-mediated versus non-HDAC-mediated effects of pan-HDAC inhibitors is needed. Isoform-selective HDAC inhibitors should be evaluated for preclinical and clinical efficacy. In addition, we have to understand how the combination of different histone modifications (acetylation, methylation) appearing on different histones at different times (the “histone code”) can mediate a unique cellular response, such as the hyperproliferative PASMC phenotype.

For epigenetic modulation therapies, the lack of specificity of any interference is an important issue. For example, manipulation of global DNA demethylation via the use of nonselective agents can contribute to increased chromosomal instability and may even be transmitted between generations.¹²⁴ Therefore, interference has to be fine-tuned to focus on specific epigenetic regulators/modifiers linked with the genes most relevant for PH development.

Source of Support: Nil.

Conflict of Interest: None declared.