OUP user menu

HIF-1 inhibition decreases systemic vascular remodelling diseases by promoting apoptosis through a hexokinase 2-dependent mechanism

Caroline M. Lambert, Mélanie Roy, Geneviève A. Robitaille, Darren E. Richard, Sébastien Bonnet
DOI: http://dx.doi.org/10.1093/cvr/cvq152 196-204 First published online: 24 May 2010

Abstract

Aims Vascular remodelling diseases are characterized by the presence of proliferative and apoptosis-resistant vascular smooth muscle cells (VSMC). There is evidence that pro-proliferative and anti-apoptotic states are characterized by metabolic remodelling (a glycolytic phenotype with hyperpolarized mitochondria) involving Akt pathway activation by circulating growth factors. Hypoxia-inducible factor-1 (HIF-1) is involved in different vascular diseases. Since this transcription factor is implicated in metabolic responses, we hypothesized that HIF-1 activity could be involved in vascular remodelling in response to arterial injury.

Methods and results Our findings indicate that growth factors, such as platelet-derived growth factor (PDGF), activate the Akt pathway (measured by immunoblot) in human carotid artery VSMC. Activation of this pathway increased HIF-1 activation (measured by immunoblot), leading to increased glycolysis in VSMC. Expression and mitochondrial activity of hexokinase 2 (HXK2), a primary initiator of glycolysis, are increased during HIF-1 activation. The mitochondrial activity of HXK2 in VSMC led to the hyperpolarization of mitochondrial membrane potential (measured by tetramethylrhodamine methyl-ester perchlorate) and the suppression of apoptosis (measured by TUNEL assay and 3 activity), effects that are blocked by HIF-1 inhibition. Additionally, HIF-1 inhibition also decreased VSMC proliferation (proliferating cell nuclear antigen and Ki-67 assays). In vivo, we demonstrate that localized HIF-1 inhibition, using a dominant-negative HIF-1α adenoviral construct, prevented carotid artery post-injury remodelling in rats.

Conclusion We propose that HIF-1 is centrally involved in carotid artery remodelling in response to arterial injury and that localized inhibition of HIF-1 may be a novel therapeutic strategy to prevent carotid stenosis.

  • Vascular diseases
  • Glycolysis
  • Warburg effect
  • Hypoxia-inducible factor-1
  • Apoptosis
  • Hexokinase

1. Introduction

Vascular remodelling diseases (VRD) such as atherosclerosis, hypertension, and carotid stenosis are characterized by increased proliferation and decreased apoptosis of vascular smooth muscle cells (VSMC). Despite greater energy needs (due to the proliferation) and the presence of oxygen, cancer cells use glycolysis as a main energy source, allowing the shutdown of mitochondrial function including pro-apoptotic pathways.1 This phenomenon was first described in the early 1930s by Warburg.2,3 We previously described that VRD share many features with cancer cells, including the expression of the proto-oncogene-like protein survivin and the activation of the transcription factor NFAT (nuclear factor of activated T-cells).4,5 Furthermore, the metabolic modulator dichloroacetate (DCA), an inhibitor of the pyruvate dehydrogenase kinase (PDK), increased the glucose oxidation/glycolysis ratio and restored mitochondrial-dependent apoptosis, reversing both cancer and vascular remodelling.68 Taken together, these findings suggest significant similarities between cancer cell and VSMC metabolism and proliferation.

One unique feature of cancer is its glycolytic metabolic phenotype.9,10 Glycolysis is associated with pro-proliferative and anti-apoptotic states, and glycolytic enzymes, such as hexokinase 2 (HXK2), are now recognized as also being anti-apoptotic1115. Furthermore, many anti-apoptotic pathways, such as the activation of the phosphatidylinositol 3-kinase (PI3K)/Akt pathway, c-myc, or the inactivation of p53, also appear to induce a glycolytic phenotype.10,14,1620 Therefore, cellular metabolism can be viewed as an important integrator of different molecular pathways involved in the regulation of apoptosis and proliferation.

Hypoxia-inducible factor-1 (HIF-1) is a key transcription factor for responses to low oxygen. HIF-1 regulates different cellular responses such as proliferation and survival, erythropoiesis, and angiogenesis.21 A well-known inducer of glycolysis, HIF-1, is active in both vascular diseases and cancer.2228 HIF-1 increases the expression of several pro-glycolytic enzymes, such as HXK223,29 and PDK,30 both of which are also known inhibitors of apoptosis.1,14 However, the links between HIF-1, cell metabolism, and VSMC survival and remodelling are not clearly understood. Here, we demonstrate that in proliferative [platelet-derived growth factor (PDGF)-treated] human carotid artery smooth muscle cells (hCASMC), HIF-1 activation occurs secondary to PI3K/Akt pathway activation. By increasing HXK2 expression, HIF-1 promoted glycolysis and mitochondrial membrane potential hyperpolarization (↑ΔΨm), thus decreasing apoptosis and promoting hCASMC proliferation. In vivo, localized HIF-1 inhibition using a dominant-negative (DN) adenoviral construct prevented carotid artery post-injury remodelling in rats.

2. Methods

All the experiments were performed with the approval of the Université Laval Ethic and Biosafety Committee. The investigation conforms to 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). It conforms to the principles outlined in the Declaration of Helsinki.31

2.1 Cell culture

hCASMC isolated from a healthy donor (organ transplant patient) were grown in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% foetal bovine serum (Invitrogen) and 1% of an antibiotic–antimycotic mixture (GIBCO 15240, Invitrogen) and maintained in a humid atmosphere, at 37°C with 5% CO2. Cells were used between passages 3 and 9 as previously described.32

2.2 Drugs and peptides

hCASMC were treated for 48 h with PDGF-BB (30 ng/mL, Millipore) and PI3K inhibitor LY294002 [10 µM, EMD Chemicals (Gibbstown, NJ)], and a cell-permeable peptide (carrying an N-terminal antennapedia homeoprotein sequence) termed hexokinase 2 voltage-dependent anion channel (VDAC)-binding domain peptide (Hxk2VBD, Calbiochem, San Diego, CA, USA; RQIKIWFQNRRMKWKK-MIASHLLAYFFTELN) was used to inhibit hexokinase localization to the mitochondrion (100 µM for 48 h).

2.3 Transfection

hCASMC were transfected by CaPO4 precipitation with 20 nM siRNA oligonucleotides. Twenty-four hours post-transfection, the cell medium was changed, and 48 h post-transfection, cells were treated as indicated. All siRNAs were obtained from Applied Biosystems and the specific sequences used are as follows: human HIF-1α (accession no. NM_001530; sense, 5′-AGGACAAGUCACAACAGGAUU-3′. As a control oligonucleotide, ‘Silencer Negative Control #2 siRNA’ was used.

2.4 Immunoblotting

Immunoblotting was performed with HIF-1α (1:1500, antiserum raised in rabbits immunized against the last 20 amino acids of the C-terminal human protein),33 Akt (1:1000, Cell Signaling), phospho-Akt (1:1000, Cell Signaling), GSK3β (1:1000, Cell Signaling), phospho-GSK3β (1:1000, Cell Signaling), HXK2 (1:1000, Cell Signaling), caspase 3 (1:1000, Cell Signaling), and actin (1:300, Santa Cruz). Twenty-five micrograms of protein were loaded. Expression was normalized to actin to correct for loading differences.

2.5 Lactate measurement

Following indicated cell treatment in section 2.2, the medium was removed from cells and lactate levels in the extracellular medium were measured using the Lactate Colorimetric Assay Kit (Abcam, Cambridge, MA, USA). This kit detects lactate levels in samples from concentrations of 0.02–10 mM. Lactate concentration was normalized to sample cell number.

2.6 Confocal microscopy

Mitochondrial membrane potential was determined using 10 nM tetramethylrhodamine methyl-ester perchlorate (TMRM, Invitrogen). Nuclei were stained using 50 nM Hoechst 33342 (Invitrogen). Cell imaging was performed using an FV1000 confocal microscope equipped with a live cell apparatus (Olympus, Center Valley, PA, USA).

2.7 Immunofluorescence

hCASMC were fixed with 1% paraformaldehyde and permeabilized with 0.2% Triton X-100. Apoptag apoptosis detection kit (TUNEL Serologicals, Norcross, GA, USA), Annexin V (Clontech, Canada), the proliferating cell nuclear antigen (PCNA), and Ki-67 (1:200, Millipore) were used to measure apoptosis after starvation and proliferation by calculating the per cent of TUNEL, Annexin V, PCNA, and Ki-67-positive cells. Antibodies to HXK2 (1:100) and Mitotracker Red (250 nM, Invitrogen) were also used. Primary antibody detection was performed using Alexa Fluor (488 for GFP staining; 647 for TUNEL, Ki-67, and HIF-1α; and TRITC 543 for PCNA and smooth muscle actin) secondary antibodies (Invitrogen). Nuclei were stained with DAPI (Invitrogen). Rat carotid arteries were fixed with 4% paraformaldehyde. Immunofluorescence was performed on 5 µm carotid slices. Antibodies to HIF-1α (1:200), GFP [1:400, Living Colors A.v. Monoclonal Antibody (JL-8), Clontech, CA, USA], PCNA (1:400), Ki-67 (1:200), and α-smooth muscle actin (1:400, Sigma) were used.

2.8 Carotid artery balloon injury model

Male Sprague–Dawley rats (350 g) were used. Under anaesthesia, a neck incision was made. A 20 mm section of the right common carotid artery was isolated and temporarily occluded to prevent retrograde blood loss. After proximal arteriotomy, a 2 F Fogarty embolectomy catheter was introduced to perform an antegrade balloon inflation along a 15 mm segment of the common carotid artery. The catheter, containing either saline, a GFP adenoviral construct, or a GFP-HIF-1α dominant-negative (DN HIF-1α) adenoviral construct (1.8 × 108 p.f.u) was maintained in place for 30 min to assure efficient adenoviral infection. The lumen was then flushed with heparinized saline, the arteriotomy was closed with 8-0 Prolene sutures, and the perfusion was restored. Haematoxylin and eosin staining was performed on carotid slides and vascular wall, media wall, and neointima thickness were measured 14 days post-injury (five slides per rat/five rats were studied). Experiments were performed with the approval of the Laval University Animal Ethics Committee.

2.9 Statistics

Values are expressed as the mean ± SEM. ANOVA was performed with the Newman–Keuls test for post hoc analysis. A value of P < 0.05 was considered significant (GraphPad Prism 5.0b).

3. Results

3.1 HIF-1 is activated by PDGF in normoxic hCASMC and is regulated by the PI3K/Akt/GSK3β pathway

As seen in Figure 1A, HIF-1α protein expression is significantly up-regulated in hCASMC by PDGF (five-fold) in normal oxygen conditions. Targeting HIF-1α with a specific siRNA confirmed the detection of HIF-1α in western blots. The transfection of PDGF-treated hCASMC with an HIF-1α siRNA decreased HIF-1α protein expression by ∼70% compared with cells transfected with a control siRNA (n = 3, P < 0.05) (Figure 1A). Interestingly, HIF-1 activation was associated with an activation of the PI3K/Akt pathway as shown by a significant increase in phosphorylation of Akt (20-fold) in PDGF-treated hCASMC vs. control cells (Figure 1B). To determine the dependency of PDGF-induced HIF-1α expression in hCASMC on PI3K/Akt pathway activation, cells were treated with the potent PI3K/Akt inhibitor, LY294002. As seen in Figure 1B, LY294002 significantly decreased (2.6-fold) HIF-1α expression in hCASMC treated with PDGF (n = 3, P < 0.05) (Figure 1B). It is important to note that levels of Akt phosphorylation were not modified when hCASMC were transfected with HIF-1α siRNA (n = 3, P < 0.05) (Figure 1C), demonstrating that Akt activation is upstream of HIF-1 activation. Furthermore, the treatment of hCASMC with PDGF led to the inhibitory phosphorylation (Ser21/9) of GSK3β (two-fold). Again, levels of GSK3β phosphorylation were not modified when hCASMC were transfected with HIF-1α siRNA (n = 3, P < 0.05) (Figure 1D). These results suggest that HIF-1α expression is dependent on PI3K/Akt/GSK3 pathway activation by PDGF in hCASMC.

Figure 1

HIF-1 is activated by PDGF in normoxic hCASMC and is regulated by PI3K/Akt/GSK3β pathway. (A) hCASMC stimulated with PDGF-BB (30 ng/mL) for 48 h has elevated expression of HIF-1α (five-fold) by western blot analysis. Transfection with HIF-1α silencer RNA decreased about 70% of its expression (n = 3). (B) Inhibition of HIF-1α expression by the PI3K/Akt inhibitor LY294002 (10 µM) in PDGF-stimulated cells (2.6-fold), as shown by western blot analysis (n = 3). (C) Phosphorylation and activation of Akt in PDGF-stimulated cells (20-fold). HIF-1α inhibition by siRNA on PDGF-stimulated cells did not alter P-Akt (n = 3). (D) Phosphorylation and inhibition of GSK3β in PDGF-treated cells (two-fold). HIF-1α inhibition by siRNA on PDGF-stimulated cells did not alter P-GSK3β, as shown by western blot analysis (n = 3). *P < 0.05.

3.2 HIF-1 activation increases glycolysis in hCASMC

As seen in Figure 2A, HXK2 protein expression was increased (two-fold) by the treatment of hCASMC with PDGF. Targeting HIF-1α with a specific siRNA significantly blocked (1.5-fold) the increase in HXK2 protein levels in PDGF-treated hCASMC when compared with the control siRNA (n = 3, P < 0.05) (Figure 2A). Lactate concentration in hCASMC media was also significantly increased by PDGF (2.3-fold). This increase in extracellular lactate concentration was also blocked by transfecting hCASMC with an HIF-1α siRNA (1.7-fold) (n = 5, P < 0.05) (Figure 2B). These results show that in PDGF-treated hCASMC, HIF-1 regulates HXK2 expression and increases glycolysis by enhancing the transformation of pyruvate into lactate.

Figure 2

HIF-1 activation increases hCASMC glycolysis. (A) HXK2 expression is increased in PDGF-stimulated cells (two-fold). HIF-1α inhibition by siRNA decreased HXK 2 expression in PDGF-stimulated cells (1.5-fold), as shown by western blot analysis (n = 3). (B) Spectrophotometry analysis of lactate concentration in media showed that concentration is increased in PDGF-treated cells (1.0–2.3 mM). HIF-1α inhibition in PDGF-stimulated cells decreased lactate concentration (2.4–1.4 mM) (n = 5). *P < 0.05.

3.3 Translocation of HXK2 to mitochondria hyperpolarizes mitochondrial membrane, decreases apoptosis, and enhances hCASMC proliferation

As seen in Figure 3A, HXK 2 (green) colocalized (yellow) with mitochondria (red) in PDGF-treated hCASMC. This colocalization was lost when PDGF-treated hCASMC were transfected with an HIF-1α siRNA when compared with a control siRNA (n ∼ 50) (Figure 3A). We then used confocal microscopy to analyse TMRM levels in live cells in order to demonstrate that PDGF treatment led to hyperpolarized mitochondrial membrane potential (↑ΔΨm) (n ∼ 500, P < 0.05) in hCASMC. Again, increased mitochondrial membrane hyperpolarization was blocked when PDGF-treated hCASMC were transfected with an HIF-1α siRNA when compared with a control siRNA or treated with a cell-permeable competitive peptide (100 µM), inhibiting the interaction between hexokinase 2 and VDAC (Figure 3B; see Supplementary material online, Figure 1A). These findings confirm that mitochondrial hyperpolarization is indeed mediated by the mitochondrial translocation of HXK2. Additionally, TUNEL staining demonstrated that the percentage of apoptotic hCASMC in response to 24 h of serum deprivation significantly decreased during PDGF treatment when compared with untreated hCASMC (60–20%). Decreased apoptosis was inhibited when PDGF-treated hCASMC were transfected with an HIF-1α siRNA (Figure 3C) (≈16–47%) (n ∼ 150, P < 0.05) or the cell-permeable competitive peptide (100 µM), inhibiting the interaction between HXK2 and VDAC (see Supplementary material online, Figure 1) (≈10–32%) (n ∼ 30, P < 0.05). These findings were further confirmed by cytochrome c release from the mitochondria and Annexin V staining (see Supplementary material online, Figure 2). These results suggest that the mitochondrial apoptosis resistance effects of HXK 2 are HIF-1-dependent.

Figure 3

Translocation of HXK2 to mitochondria hyperpolarizes the mitochondrial membrane, decreases apoptosis, and enhances proliferation of hCASMCs. (A) PDGF-stimulated hCASMC showed significant colocalization between HXK2 (green) and mitochondria (red), giving a yellow pattern in the merged photographs. HIF-1α inhibition in PDGF-treated cells showed diffuse cytoplasmic staining of HXK2 (no colocalization of HXK 2 to mitochondria). (B) PDGF-stimulated hCASMC have hyperpolarized ΔΨm vs. control cells. HIF-1α inhibition induced mitochondrial depolarization (n ∼ 500). PDGF-stimulated hCASMC treated for 48 h with a cell-permeable competitive peptide (100 µM) inhibiting the interaction between HXK2 and VDAC significantly depolarized the mitochondrial ΔΨm to a level similar to the one seen in hCASMC treated with siHIF-1 (n ∼ 100 cells). (C) In PDGF-treated cells, HIF-1α inhibition promoted apoptosis measured by TUNEL staining (n ∼ 150). (D) HIF-1α inhibition in PDGF-stimulated cells decreased proliferation measured by Ki-67 staining (n ∼ 150). *P < 0.05.

Finally, as measured by PCNA (n ∼ 150, P < 0.05) (see Supplementary material online, Figure 2) and Ki-67 (n ∼ 150, P < 0.05) (Figure 3D) staining, cell proliferation was also enhanced in PDGF-treated hCASMC, an effect which is also blocked when PDGF-treated hCASMC were transfected with an HIF-1α siRNA (Figure 3D; see Supplementary material online, Figure 2) or cell-permeable HXK2 and VDAC competitive peptide (100 µM) (see Supplementary material online, Figure 1C) (n ∼ 150, P < 0.05).

3.4 HIF-1α inhibition decreases vascular wall thickness in the model of carotid stenosis in rats

As seen in Figure 4, media and neointima wall thicknesses in saline-treated injured rat carotids (saline carotids) were significantly increased compared with sham-operated rats (sham carotids) (n = 5 per group, P < 0.05). Injured carotids infected with a DN HIF-1α adenoviral construct (DN HIF-1α carotids) showed decreased media (∼22%; P < 0.05) and neointima thicknesses (∼30%; P < 0.05) when compared with injured carotids infected with a GFP adenoviral construct (GFP carotids) (n = 5 per group, P < 0.05) (Figure 4). Infection efficiency was confirmed by GFP staining (see Supplementary material online, Figure 3D). No significant change of thickness was observed between saline and GFP carotids. Immunofluorescent analysis of carotid slides with an anti-α-smooth muscle actin antibody (in green) indicates that SMCs are present in the entire vascular wall and neointima of injured carotids and contribute to their thickening. Triple staining with HIF-1α (in red), α-smooth muscle actin (green), and DAPI showed that HIF-1 activation is increased in VSMC of saline and GFP carotids (showed by the colocalization of DAPI/HIF-1α/α-smooth muscle actin) compared with DN HIF-1α carotids, where HIF-1α is decreased and mainly present in the cytosol (n ∼ 300 from five rats, P < 0.05). Ki-67 (in red), α-smooth muscle actin (green), and DAPI (nucleus blue) showed that saline and GFP carotids presented more proliferative VSMC than DN HIF-1α carotids (measured by the colocalization of Ki-67 with α-smooth muscle actin and the nucleus) (n ∼ 300 from five rats, P < 0.05). Sham carotids had practically no proliferative cells. GFP carotids also showed no effect on Ki-67 staining. All these findings were confirmed by PCNA staining (see Supplementary material online, Figure 3A). Triple staining with TUNEL (in red), α-smooth muscle actin (green), and DAPI showed that DN HIF-1α carotids presented significantly more apoptotic VSMC than saline and GFP carotids (n ∼ 300 from five rats, P < 0.05). Again, GFP carotids showed no significant increase in the percentage of apoptotic cells when compared with saline carotids. Increased apoptosis by HIF-1α inhibition was confirmed by caspase 3-activity assay (see Supplementary material online, Figure 3B). Finally, the decrease in CASMC proliferation and the increase in CASMC apoptosis by HIF-1α inhibition decrease the amount of smooth muscle actin-positive cells (CASMC) within the neointima (see Supplementary material online, Figure 3C). These results indicate that localized HIF-1α inhibition decreases carotid artery post-injury remodelling in rats.

Figure 4

HIF-1α inhibition decreases vascular wall thickness in a model of carotid stenosis in rats. Localized HIF-1α inhibition using DN adenovirus decreased injured carotid artery media and neointima wall thickness (five slides/rat, five measurements/slide). Triple staining with HIF-1α (in red), α-smooth muscle actin (green), and DAPI showed that HIF-1α activation is increased in VSMC of saline and GFP carotids (showed by the colocalization of DAPI/HIF-1α/α-smooth muscle actin) compared with DN HIF-1α-treated carotids (n ∼ 300 from five rats, P < 0.05). Triple staining with Ki-67 (in red), α-smooth muscle actin (green), and DAPI (nucleus blue) showed that saline and GFP carotids presented more proliferative VSMC than DN HIF-1α carotids (measured by the colocalization of Ki-67 with α-smooth muscle actin and the nucleus) (n ∼ 300 from five rats, P < 0.05). Sham carotids had practically no proliferative cells. GFP carotids also showed no effect on Ki-67 staining. Triple staining with TUNEL (in red), α-smooth muscle actin (green), and DAPI showed that DN HIF-1α carotids presented significantly more apoptotic VSMC than saline and GFP carotids (n ∼ 300 from five rats, P < 0.05). Again, GFP carotids showed no significant increase in the percentage of apoptotic cells when compared with saline carotids. These results indicate that localized HIF-1α inhibition decreases carotid artery post-injury remodelling in rats. *P < 0.05.

4. Discussion

We show for the first time that in human carotid artery smooth muscle, growth factors like PDGF activate the transcription factor HIF-1 through a PI3K/Akt/GSK3β-dependent mechanism. As in cancer cells, HIF-1 activation increased glycolysis, by promoting HXK2 expression, and hyperpolarized mitochondrial membrane potential by promoting HXK2 translocation to mitochondria suppressing hCASMC ability to undergo apoptosis and promoting hCASMC proliferation. These phenomena are reversed by HIF-1 inhibition (Figure 5). In vivo, we demonstrate that localized HIF-1 inhibition using a DN construct prevented carotid artery post-injury remodelling in rats by decreasing hCASMC proliferation and resistance to apoptosis.

Figure 5

Injury or endothelial dysfunction increase circulating growth factors and agonists like PDGF and ET-1. PDGF activates the PI3K pathway, promoting the phosphorylation and activation of Akt. P-Akt decreases GSK3β activation with its phosphorylation. Activation of the Akt/GSK3β axis leads to HIF-1α stabilization, which promotes VSMC proliferation. Inhibition of GSK3β activity and increased HXK2 expression lead to its binding to VDAC. This induces mitochondrial hyperpolarization and decreases apoptosis, both of which lead to the development of VRD.

Our findings are the first evidence of similarities between VRD and cancer. Indeed, similar to the described Warburg effect in cancer, proliferative VSMC are characterized by enhanced glycolysis and high lactate level, despite normal oxygen levels.2,9 In agreement with our findings in VSMC, Pastorino et al.14 previously showed in cancer cells that this phenotype resulted from the activation of the Akt pathway, promoting HXK2 mitochondrial translocation inhibiting both glucose oxidation and apoptosis. Because of its implications in the regulation of mitochondrial functions (promotes mitochondrial hyperpolarization),22 cell metabolism (regulates the expression of several metabolic enzymes including HXK2),23 cell apoptosis, and proliferation,22 the transcription factor HIF-1 can on its own explain the ‘cancer phenotype’ in VSMC. Interestingly, previous studies have shown that the Akt pathway,34,35 along with the glycolytic end-product lactate, promotes HIF-1 activation in cancer cells,36,37 reinforcing the existence of similarities between proliferative VSMC and cancer cells. Therefore, our findings provide a better explanation of how metabolic modulators like DCA1,38 are so efficient in vascular disease like pulmonary hypertension.22,39 Indeed, by blocking the PDKs, DCA will force the pyruvate translocation within the Krebs cycle, decreasing lactate generation and thus HIF-1 activation.

In the present findings, we provide evidence that HIF-1 activation is dependent on the activation of a PI3K/Akt/GSK3 axis and that inhibition of this axis resulted in the inhibition of HIF-1 activation. This finding is very appealing as it provides a better understanding of the mechanism of action of the widely used PI3K/Akt inhibitors in both cancer40 and cardiovascular diseases.41 For example, we have recently shown that dehydroepiandrosterone (DHEA) reverses vascular remodelling through the inhibition of the Akt pathway resulting in the inhibition of HXK2-dependent resistance to apoptosis.32 In the current study, we describe HIF-1 activation as a new mechanism linking Akt activation to enhanced HXK2 expression and thus apoptosis resistance in VSMC. Our findings could be of great therapeutic interest due to the high translational potential of DHEA (being a cheap and non-toxic drug)32 and provide evidence that Akt inhibition by DHEA can explain the inhibition of HIF-1 by this hormone in pulmonary circulation.42

Along with apoptosis resistance, VRD are associated with increased VSMC proliferation as shown in the current study. We extensively explained that the mechanism accounting for VSMC proliferation in VRD involved the down-regulation of K+ channels (Kv1.5) resulting in VSMC depolarization, L-Type Ca2+ channel activation increasing [Ca2+]I, promoting the activation of the pro-proliferative factor NFAT.22,43,44 In the present study, we demonstrated that HIF-1 inhibition reverses VSMC proliferation. This finding is, in fact, in agreement with previous findings45 including ours, showing that HIF-1 activation accounts for K+ channels down-regulation (including Kv1.5).22,44 Therefore, by blocking HIF-1, we promote K+ channels expression, decreasing [Ca2+]I and proliferation.22 Because our group has already published the implication of HIF-1 in VSMC proliferation22 and has asserted that Kv1.5 is down-regulated in PDGF-treated hCASMC,32 the effect of HIF-1 inhibition on hCASMC proliferation was limited to the presented PCNA and Ki-67 assays. In accordance with our study, Karshovska et al.45 had previously shown that HIF-1-specific inhibition by siRNA induced significant inhibition of neointimal growth. In addition to the decreased VSMC proliferation and resistance to apoptosis that we demonstrate, they showed that the reduction in neointimal area by HIF-1 inhibition was attributed to a diminished accumulation of neointimal SMC (also observed in our model) secondary to a down-regulation of the HIF-1-regulated SDF-1 pathway. Although the effects of HIF-1 inhibition on SDF-1 were not tested in our study, the study provides an additional mechanism demonstrating that HIF-1 inhibition could be of a great therapeutic interest in vascular injury.

Owing to the numerous mediators involved in VRD, it is likely that in addition to the Akt pathway, other mechanisms might explain in part our results. Indeed, growth factors and cytokines effects (which trigger vascular remodelling)28,46,47 are mediated by the activation of receptor tyrosine kinase, which can activate several different downstream pathways including Akt, STAT3, or Src to name a few.41,48,49 All of which can also activate HIF-150 and promote apoptosis resistance35 and cell proliferation.48,49 Therefore, it is believed that whatever signalling pathway triggers HIF-1 activation in VSMC will promote cell proliferation and resistance to apoptosis. This indeed reinforces the importance of HIF-1 in vascular remodelling and clearly presents HIF-1 as a very attractive therapeutic target for vascular diseases as suggested by our in vivo data and previous findings.22,45

Indeed, after demonstrating the implication of HIF-1 in hCASMC proliferation and resistance to apoptosis, we demonstrated for the first time the implication of HIF-1 activation in vascular remodelling in the rat carotid injury model. Moreover, we provide evidence that HIF-1 can be therapeutically targeted. Therefore, we propose that localized HIF-1 inhibition might be a novel therapeutic strategy to prevent VRD. In conclusion, our findings open new avenues of investigation in VRD and could be of great therapeutic interest for the treatment of vascular diseases.

Supplementary material

Supplementary material is available at Cardiovascular Research online.

Conflict of interest: none declared.

Funding

This work was supported by grants from the Canadian Institutes of Health Research (CIHR; entitle Experimental therapies for vascular remodeling diseases to S.B. and MOP-49609 to D.E.R.) and the Heart and Stroke Foundation of Canada (S.B.). S.B. holds a Canada Research Chair. D.E.R. is the recipient of a CIHR New Investigator Award. C.M.L. is a recipient of a Graduate Scholarship from La Société Québécoise d'Hypertension Artérielle (SQHA).

References

View Abstract