Hypoxia Inducible Factor-1α in Pulmonary Artery Smooth Muscle Cells Lowers Vascular Tone by Decreasing Myosin Light Chain Phosphorylation

Abstract

Rationale

Hypoxia inducible factor-1α HIF-1α an oxygen (O₂)-sensitive transcription factor, mediates transcriptional responses to low O₂ tension states. While acute hypoxia causes pulmonary vasoconstriction and chronic hypoxia can cause vascular remodeling and pulmonary hypertension, conflicting data exists surrounding the role of HIF-1α in modulating pulmonary vascular tone.

Objective

To investigate the role of smooth muscle cell (SMC) specific HIF-1α in regulating pulmonary vascular tone.

Methods and Results

Mice with a SMC specific deletion of HIF-1α (SM22α-HIF-1α^−/−) were created to test the hypothesis that pulmonary artery smooth muscle cell (PASMC) HIF-1α modulates pulmonary vascular tone and the response to hypoxia. SM22α-HIF-1α^−/− mice exhibited significantly higher right ventricular systolic pressure (RVSP) compared to wild-type (WT) littermates under normoxia and with exposure to either acute or chronic hypoxia in the absence of histologic evidence of accentuated vascular remodeling. Moreover, myosin light chain (MLC) phosphorylation, a determinant of SMC tone, was higher in PASMC isolated from SM22α-HIF-1α^−/− mice compared to WT PASMC, during both normoxia and after acute hypoxia. Further, over-expression of HIF-1α decreased MLC phosphorylation in HIF-1α-null SMC.

Conclusion

In both normoxia and hypoxia, PASMC HIF-1α maintains low pulmonary vascular tone by decreasing MLC phosphorylation. Compromised PASMC HIF-1α expression may contribute to the heightened vasoconstriction that characterizes pulmonary hypertension.

Introduction

Organismal survival requires cells to sense and respond to changes in oxygen tension. Hypoxia inducible factor-1α (HIF-1α, an oxygen (O₂)-sensitive transcription factor, facilitates cellular responses to hypoxia by regulating genes involved in angiogenesis, oxygen transport, and energy metabolism.¹ The pulmonary vasculature responds to hypoxia with vasoconstriction and remodeling, however, the role of HIF-1α in the pulmonary circulation remains unclear.

Sustained hypoxic exposure results in pulmonary hypertension, a disease characterized by pulmonary artery (PA)vasoconstriction and neomuscularization.² In a murine model, haploinsufficiency of HIF-1α attenuates hypoxia-induced increases in pulmonary artery pressure, right ventricular hypertrophy, and pulmonary vascular remodeling, suggesting that HIF-1α might play a role in increasing pulmonary vascular tone.³ However, recent data in pulmonary artery smooth muscle cells (PASMC) indicates that HIF-1α regulates the expression of an ion channel subunit that mitigates vasoconstriction.⁴ Insight into the role that HIF-1α plays in modulating pulmonary vascular tone and in the pathogenesis of pulmonary hypertension is limited, in part, by neonatal lethality in mice lacking HIF-1α.^5,6 Thus, to test the hypothesis that PASMC HIF-1α modulates both basal PA tone and the response to hypoxia, we created mice with a SMC-specific deletion of HIF-1α.

Methods

Transgenic mice with selective deletion of HIF-1α in SMC were created by crossbreeding SM22α -promoter-driven Cre mice expressing the Cre reporter, ROSA26 (R26R), (kindly provided by Marlene Rabinovitch M.D., Stanford University)⁷ with HIF-1α^flox/flox mice (JAX 007561, B6.129-Hif1α^tm3Rsjo/J; Jackson Laboratories). HIF-1α homozygous floxed mice contain the HIF-1α exon 2 flanked by LoxP sites. The Institutional Animal Care and Use Committee at Stanford University approved all the procedures and protocols governing the care and use of laboratory animals. RVSP was measured after exposure to either acute or chronic hypoxia. Histologic analysis was performed on lung tissue. Western blotting was performed to determine myosin light chain (MLC) phosphorylation in PASMC. Full details can be found in the Online Supplement.

Results

SM22α-HIF-1α^−/− mice

The smooth muscle specific protein, SM22, is ubiquitously expressed in SMC. SM22 promoter activity was found in the arterial SMC, including the aorta and PA.⁸ LacZ reporter staining demonstrated the absence of β-gal activity in tissues derived from wild-type (WT) SM22α-HIF-1α^+/+ mice (Figures 1A and 1C). In contrast, prominent X-gal staining was found in SMC of the PA and aorta from SM22α-HIF-1α^−/− mice (Figures 1B and 1D, respectively). Both HIF-1α mRNA and protein were undetectable in aortic (Ao) SMC and PASMC isolated from SM22α-HIF-1α^−/− mice, confirming deletion of HIF-1α in vascular SMC (Figures 1E and 1F, respectively). In addition, PASMC HIF-2α protein expression was not different between WT and HIF-1α^−/− mice (Online Figure I). While the myocardium of SM22α-HIF-1α^−/− mice demonstrated patchy X-gal staining, HIF-1α protein expression in the heart did not differ between WT and HIF-1α^−/− mice under either normoxic or hypoxic (21 days) conditions (data not shown). In addition, left ventricular function, as measured by echocardiography did not differ between the two groups (Online Table I).

Figure 1. Smooth muscle specific deletion of HIF-1a in SM22a-HIF-1a^−/− mice and HIF-1a^−/− SMC.

Pulmonary artery and aorta tissues from SM22a-HIF-1a^−/− (B and D, respectively) and WT (SM22a-HIF-1a^+/+) mice (A and C, respectively) were stained with X-Gal. Scale bar = 100mm. (E) Real-time quantitative RT-PCR and (F) Western immunoblot to detect HIF-1α mRNA and protein in AoSMC and PASMC isolated from SM22α-HIF-1α^−/− and WT mice (left and right panels, respectively). Bars represent means ± SEM (n=3).

SMC specific loss of HIF-1α increases pulmonary vascular tone

At baseline, RVSP was higher in SM22α-HIF-1α^−/− mice compared to WT (Figure 2A), despite only normoxic exposure. After chronic hypoxia, RVSP increased in both groups, but was significantly higher in SM22α-HIF-1α^−/− mice compared to controls. Heart rate, cardiac output, left ventricular function, hematocrit, and body weight did not differ in the two genotypes, either at baseline or after chronic hypoxia (Online Table I). Moreover, the increased RVSP in SM22α-HIF-1α^−/− mice was observed in the absence of differences in the number of muscularized arterioles between the two groups (Figure 2B), suggesting that the relatively higher RVSP in SM22α-HIF-1α^−/− mice is not attributable to differential vascular remodeling.

Figure 2. Smooth muscle specific loss of HIF-1α increases pulmonary vascular tone.

(A) RVSP in SM22α-HIF-1α^−/− and WT mice in normoxia and following chronic hypoxia (10% O₂, 3 wks) (n=14-18 per group). (B) Percent of muscularized peripheral PA to total PA (n=5-9 per group). (C) RVSP in SM22α-HIF-1α^−/− and WT mice under baseline, in response to hypoxia for 15 minutes, and after recovery (n=12-16 per group). Bars represent means ± SEM. * p < 0.05 and ** p < 0.01, SM22α-HIF-1α^−/− vs. WT; §§ p < 0.01 and §§§ p < 0.001, hypoxia vs. normoxia for each genotype.

To address the potential that hypoxic pulmonary vasoconstriction differs between SM22α-HIF-1α^−/− and WT mice, RVSP were measured during 15 minutes of acute hypoxia (10% O₂) and then during exposure to 40% O₂ (Figure 2C). Acute hypoxia increased RVSP in both groups, but RVSP remained higher in SM22α-HIF-1α^−/− mice, compared to WT mice. With exposure to 40% oxygen, RVSP decreased in both groups, but remained higher in the SM22α-HIF-1α^−/− mice.

Loss of HIF-1α in PASMC increases myosin light chain phosphorylation

MLC phosphorylation augments the contractile state of vascular SMC by facilitating myosin and actin filament interaction.⁹ To investigate the molecular mechanism leading to increased pulmonary vascular tone in SM22α-HIF-1α^−/− mice we measured MLC phosphorylation (pMLC) in PASMC isolated from the two groups of mice. pMLC was more than 2-fold higher in PASMC isolated from SM22α-HIF-1α^−/− mice, compared to WT mice under baseline normoxia (Figure 3A). While acute hypoxia increased pMLC in both groups, pMLC remained significantly higher in HIF-1α^−/− PASMC, compared to WT PASMC.

Figure 3. Loss of HIF-1α in PASMC increases myosin light chain phosphorylation.

(A) pMLC was measured in PASMC isolated from SM22α-HIF-1α^−/− and WT mice by Western immunoblot under normoxia and after exposure to acute hypoxia. Bars represent means ± SEM (n=3). * p < 0.05 and ** p < 0.01, HIF-1α^−/− vs. WT; § p < 0.05 and §§ p < 0.01, hypoxia vs. normoxia. (B) Expression of pMLC in hPASMC transfected with siNC (non-targeted control siRNA) or siHIF-1a. Nuclear extracts demonstrate expression of active HIF-1a in siNC-transfected cells. TATA bp serves as a nuclear loading control. * p < 0.05 and ** p < 0.01, siHIF-1a vs. siNC; §§§ p < 0.001, hypoxia vs. normoxia. (C) Expression of pMLC in mPASMC null for HIF-1a expression, HIF-1a^−/−, transfected with empty vector, pcDNA3, or constitutively active HIF-1a, HIF-1a (CA). * p< 0.05 and *** p < 0.001, HIF-1a (CA) vs. pcDNA3.

To ensure that HIF-1α modulates MLC phosphorylation in human as well as murine PASMC, human PASMC (hPASMC) were transfected with HIF-1α-targeted siRNA, siHIF-1α. Depletion of endogenous HIF-1α increased pMLC expression (Figure 3B). Under hypoxic conditions, pMLC expression increased in both groups. However, pMLC expression increased significantly more in HIF-1α depleted hPASMC compared to cells transfected with non-targeting control siRNA.

To investigate whether over-expression of HIF-1α would rescue the enhanced pMLC observed in the mouse PASMC (mPASMC)null for HIF-1α, HIF-1α^−/− PASMC were transfected with empty vector, pcDNA3, or a constitutively active form of HIF-1α, HIF-1α (CA) (Figure 3C). Transfection with the constitutively active construct effectively restored HIF-1α expression in the null cells. Over-expression of HIF-1α decreased pMLC under both normoxic and hypoxic conditions.

To address the phosphorylation of MLC in vivo, lung tissues from SM22α-HIF-1α^−/− and WT mice exposed to normoxia or chronic hypoxia were examined by Western immunoblot (Figure 4). Overall, pMLC was increased in SM22α-HIF-1α^−/− mice compared to WT. After chronic hypoxia, pMLC expression increased in both groups, but was significantly higher in SM22α-HIF-1α^−/− mice compared to WT. Taken together, these results demonstrate that in PASMC, HIF-1α decreases pMLC, providing evidence that HIF-1α modulates pulmonary vascular tone through effects on MLC phosphorylation.

Figure 4. Loss of HIF-1α in PASMC increases myosin light chain phosphorylation in vivo.

Lung tissues from SM22α-HIF-1α^−/− and WT mice exposed to normoxia or chronic hypoxia were examined for pMLC by Western immunoblot. Bars represent means ± SEM (n=3). ** p < 0.01, SM22α-HIF-1α^−/− vs. WT; §§ p < 0.01 and §§§ p < 0.001, hypoxia vs. normoxia.

Discussion

This is the first report to demonstrate that selective deletion of HIF-1α in SMC increases pulmonary vascular tone under both normoxic and hypoxic conditions and increases PASMC pMLC. These novel results suggest that SMC HIF-1α plays a previously undescribed role in maintaining the normally low pressure of the pulmonary vasculature at baseline, and in mitigating hypoxic pulmonary hypertension. Taken together, the observations that in SM22α-HIF-1α^−/− mice pMLC is increased and RVSP is elevated in the absence of accentuated muscularization, indicate a primary role for SMC HIF-1α in modulating vascular tone specifically, and not vascular remodeling.

HIF-1α mediates, in part, SMC cell proliferation in response to hypoxia.¹⁰ In a previous study, as compared to controls, mice haploinsufficient for HIF-1α had decreased RVSP, fewer muscularized vessels, and reduced PA wall thickness after chronic hypoxia, suggesting a blunted vasoproliferative response. In contrast, in this report, complete loss of HIF-1α in SMC raised RVSP in the absence of an increase in the number of muscularized arteries. Thus, a global decrease of HIF1α may mitigate against the development of pulmonary hypertension, while an absence of HIF-1α in SMC alone potentiates the response to hypoxia.HIF-1α may play distinct cell-specific roles in the pulmonary vasculature. Since HIF-1α was deleted only in SMC in this report, HIF-1α activation in other cell types, such as endothelial cells, may increase arterial muscularization under hypoxia via HIF-1α mediated induction of growth factors.¹¹ Thus cell-specific alterations of HIF-1α expression in the pulmonary vasculature, including decreased expression in SMC, may play an etiologic role in pulmonary hypertension. Consequently, cell-specific modulation of HIF-1α might represent a viable therapeutic strategy to address pulmonary hypertension, a disease without either cure or definitive treatment.

Loss of HIF-1α in PASMC increased pulmonary vascular tone and pMLC under both normoxic and hypoxic conditions, though the physiologic significance of the mild RVSP elevation under normoxic conditions is uncertain. Over-expression of HIF-1α decreased MLC phosphorylation, providing further evidence for the importance of HIF-1α activity in regulating MLC phosphorylation and vascular tone. While it appears that HIF-1α plays a role in maintaining the normally low pulmonary vascular tone by decreasing pMLC in SMC, the mechanistic link between HIF-1α and MLC phosphorylation remains unknown. HIF-1α may directly or indirectly induce factors that reduce phosphorylation of MLC (e.g. MLC phosphatase) or repress the activity of proteins that increase pMLC (e.g. Rho kinase and MLC kinase).

We conclude that SMC specific HIF-1α plays a previously undescribed role in maintaining low tone in the normoxic pulmonary circulation and in regulating the response to hypoxia by decreasing MLC phosphorylation. We speculate that compromised PASMC HIF-1α expression or activity contributes to the vasoconstriction that characterizes pulmonary hypertension and thus represents a novel therapeutic target.

Supplementary Material

Online Figure I. Loss of HIF-1α in mPASMC does not affect HIF-2α expression. Expression of HIF-1α and HIF-2α in PASMC isolated from SM22α-HIF-1α^−/− and WT mice by Western immunoblot under normoxia and after exposure to acute hypoxia.

Online Table I. Hemodynamic Assessments of SM22α-HIF-1α^−/− and WT Mice

Values of parameters are expressed as means ± SEM. (); number of mice per group.

Novelty and Significance.

What Is Known?

• Hypoxia inducible factor-1α (HIF-1α is an oxygen (O₂)-sensitive transcription factor that facilitates the cellular adaptation to hypoxia.

• Sustained hypoxic exposure results in pulmonary hypertension, a disease characterized by pulmonary artery vasoconstriction and neomuscularization.

• In murine models, haploinsufficiency of HIF-1α protectsagainsthypoxia-induced pulmonary hypertension, but the specific function of HIF-1α in pulmonary artery smooth muscle cells (PASMC) is not known.

What New Information Does This Article Contribute?

• Tissue specific deletion of HIF-1α in PASMC increases pulmonary vascular tone, both at baseline and in response to hypoxia.

• HIF-1α decreasesthephosphorylationof myosin light chain (MLC), a key determinant of SMC tone.

Pulmonary hypertension causes substantial morbidity and has a high mortality rate. While haploinsufficiency of HIF-1α protects against hypoxia-induced pulmonary hypertension, cell specific functions of HIF-1α are unknown. This study demonstrates that SMC specific loss of HIF-1α increases pulmonary vascular tone, without increased pulmonary artery muscularization. Moreover, we report that SMC specific loss of HIF-1α increases pulmonary vascular tone by augmenting MLC phosphorylation. Thus, SMC specific HIF-1α plays a previously undescribed role in maintaining the low tone of the normoxic pulmonary circulation, and in mitigating hypoxia-induced increases in tone. Compromised PASMC HIF-1α activity may contribute to the vasoconstriction that characterizes pulmonary hypertension.

Nonstandard Abbreviations

AoSMC

Aortic smooth muscle cell(s)

HIF-1α

Hypoxia inducible factor-1α

hPASMC

Human pulmonary artery smooth muscle cell(s)

MLC

Myosin light chain

mPASMC

Mouse pulmonary artery smooth muscle cell(s)

PA

Pulmonary artery

PASMC

Pulmonary artery smooth muscle cell(s)

pMLC

Phosphorylated myosin light chain

RVSP

Right ventricular systolic pressure(s)

SMC

Smooth muscle cell(s)

WT

Wild-type