Aberrant PGC-1α Signaling in a Lamb Model of Persistent Pulmonary Hypertension of the Newborn

Abstract

Background:

Persistent Pulmonary Hypertension of the Newborn (PPHN) is characterized by elevated pulmonary vascular resistance (PVR), resulting in hypoxemia. Impaired angiogenesis contributes to high PVR. Pulmonary artery endothelial cells (PAECs) in PPHN exhibit decreased mitochondrial respiration and angiogenesis. We hypothesize that Peroxisome Proliferator-Activated Receptor Gamma Co-Activator-1α (PGC-1α) downregulation leads to reduced mitochondrial function and angiogenesis in PPHN.

Methods:

Studies were performed in PAECs isolated from fetal lambs with PPHN induced by ductus arteriosus constriction, with gestation-matched controls and in normal human umbilical vein endothelial cells (HUVECs). PGC-1α was knocked downed in control lamb PAECs and HUVECs and overexpressed in PPHN PAECs to investigate the effects on mitochondrial function and angiogenesis.

Results:

PPHN PAECs had decreased PGC-1α expression compared to controls. PGC-1α knockdown in HUVECs led to reduced Nuclear Respiratory Factor-1 (NRF-1), Transcription Factor-A of Mitochondria (TFAM), and mitochondrial electron transport chain (ETC) complexes expression. PGC-1α knockdown in control PAECs led to decreased in vitro capillary tube formation, cell migration, and proliferation. PGC-1α upregulation in PPHN PAECs led to increased ETC complexes expression and improved tube formation, cell migration, and proliferation.

Conclusion:

PGC-1α downregulation contributes to reduced mitochondrial oxidative phosphorylation through control of the ETC complexes, thereby affecting angiogenesis in PPHN.

Introduction

Persistent pulmonary hypertension of the newborn (PPHN) is a clinical syndrome characterized by sustained elevation of pulmonary vascular resistance (PVR) after birth, resulting in extra-pulmonary right-to-left shunting and severe hypoxemia.^1,2 This condition affects 1.8 out of 1,000 newborn infants,³ and respiratory failure from PPHN is a major contributor to morbidity and mortality.⁴ While PPHN primarily affects term and near-term infants,^1,3,4 some extremely preterm infants can present with acute PPHN associated with certain risk factors, such as oligohydramnios, sepsis, and preterm premature rupture of membranes.^2,5-8 Inhaled nitric oxide (iNO) can be effective in alleviating PPHN symptoms.⁹ Yet, 40% of infants with PPHN fail to improve with such therapy.¹⁰ For these non-responders, Extracorporeal Membrane Oxygenation (ECMO) is a rescue treatment option;¹⁰ however, it is more invasive and associated with increased complications.^10,11 Although widely investigated, antioxidant agents have not been translated to therapies for PPHN, which suggests that targeting specific signaling pathways to restore normal adaptation is required for successful treatment.¹² While the role of energy depletion in the pathogenesis of PPHN is unknown, it is fundamental to understanding the metabolic dysregulation occurring at the molecular level.

PPHN manifests as a failure of normal postnatal transition in the pulmonary circulation, which can be a consequence of impaired angiogenesis, remodeled vascular structure, and reduced vasoreactivity.¹³ Angiogenesis dysfunction is a result of metabolic dysregulation in pulmonary artery endothelial cells (PAECs), smooth muscle cells, and fibroblasts.^14,15 Our lab uses a well-accepted lamb model of PPHN in which ligation of the fetal ductus arteriosus (DA) creates elevated postnatal PVR.^16-19 PAECs from PPHN lambs show reduced angiogenesis compared to controls.^20,21 Angiogenesis requires ECs to proliferate, migrate, and form tubes, all of which are energy-dependent processes.²² In addition to stabilizing cellular redox balance, mitochondria provide adenosine triphosphate (ATP) and metabolic intermediates of oxidative phosphorylation needed for proper EC function.^22,23 Our previous studies have shown that PAECs from PPHN lambs demonstrate impaired mitochondrial biogenesis and reduced mitochondrial number,^20,21 which may serve as a pathologic mechanism leading to impaired angiogenesis. Another cell type frequently studied in cardiovascular diseases is human umbilical vein endothelial cells (HUVECs)²⁴; specifically, HUVECs taken from human babies suffering from PPHN and bronchopulmonary dysplasia (BPD) show similar bioenergetic dysfunction that correlates with underlying disease.²⁴

A recent observation from our lab focused on the role that 5’-AMP-Activated Protein Kinase (AMPK) plays in the pathology of PPHN.²⁰ AMPK, a critical energy sensor in the cell, is activated by decreased ATP and increased adenosine monophosphate (AMP) levels. AMPK function appears to be required for angiogenesis in fetal lamb PAECs.^15,20,25 Previous studies in our lab revealed that AMPK activity was downregulated in PAECs isolated from fetal lambs with PPHN.¹⁵ These changes in AMPK function were associated with reduced levels of Peroxisome Proliferator-Activated Receptor Gamma Co-Activator-1α (PGC-1α), a transcription co-activator, in PPHN.^20,26 PGC-1α is activated by AMPK-mediated phosphorylation,²⁷ and following this stimulus, PGC-1α upregulates mitochondrial transcription factors to maintain metabolic homeostasis^26,28 and reactive oxygen species (ROS) detoxification machinery to defend against oxidative stress.²⁹ Our lab has previously found that not only is PGC-1α protein expression decreased in PPHN PAECs relative to controls but the mitochondrial transcription factors, Nuclear Respiratory Factor-1 (NRF-1) and Transcription Factor-A of Mitochondria (TFAM), and electron transport chain (ETC) complexes I-V are also reduced.³⁰ Downregulation of PGC-1α expression and activity can cause a decrease in mitochondrial biogenesis,^29,31 and as a result, critical pathways responsible for efficient oxidative phosphorylation and maintenance of redox balance fail to be activated, potentially leading to impaired angiogenesis function.^28,32,33

Currently, the contribution of PGC-1α signaling to PAEC dysfunction and impaired angiogenesis in PPHN is unknown. The potential downregulation of PGC-1α as a mechanistic link to reduced angiogenesis in PPHN is a plausible hypothesis. This protein is typically upregulated in the face of increased metabolic demands, which leads to augmentation of mitochondrial biogenesis to restore energy balance.^15,28 Failure of its induction in the presence of decreased ATP levels in PPHN PAECs may lead to increased apoptosis, which has been previously reported in PPHN.^15,20,34 The increased apoptosis and decreased proliferation of PAECs has been found to be associated with reduced migration and capillary tube formation.^15,20 Since failure of angiogenesis is a central feature of PPHN, defining this mechanism is critical to improve our understanding of PPHN. We hypothesize that downregulated PGC-1α expression, secondary to altered AMPK function, in PPHN leads to reduced mitochondrial function and impaired angiogenesis in PPHN. We further posit that enhancing PGC-1α expression through genetic approaches will restore angiogenesis in PPHN. Our objectives for this study are to investigate whether: (1) Downregulation of PGC-1α protein levels using silencing RNA (siRNA) impairs mitochondrial function and angiogenesis in normal endothelial cells (ECs), and (2) Overexpression of PGC-1α protein levels in PPHN lamb PAECs via MycPGC-1 plasmid restores mitochondrial oxidative efficiency and angiogenesis function.

Methods

Detailed methods can be found in Data Supplement 5. Experiments were performed in control and PPHN PAECs isolated from lambs with or without fetal pulmonary hypertension induced by prenatal DA constriction and in cultured HUVECs. All fetal lamb studies were performed in compliance with NIH guidelines and were approved by the Medical College of Wisconsin IACUC.

Fetal lamb model:

Constriction of fetal DA was performed at 128 ± 2 days of gestation (term = 144 days), as previously described.^35,36 Sham-operated, gestation-matched twin fetal lambs were used as controls. After 8 days of ductal constriction, fetal lungs were harvested for PAEC isolation. Previous studies have shown that these lambs consistently develop severe PPHN with this intervention not only due to high PVR but also pulmonary vascular remodeling.^20,30,35,37

Endothelial Cell Isolation and Culture:

We used PAECs obtained from fetal lambs with PPHN and from gestation-matched control lambs without PPHN. Additionally, we used normal HUVECs for selected experiments to knock down PGC-1α using siRNA directed to the human PGC-1α sequence. Both PAECs and HUVECs (passage of 6 or less) were grown in a humidified incubator at 37°C in 21% O₂ + 5% CO₂.

Transient siRNA-mediated knockdown and plasmid- or adenoviral-mediated overexpression of PGC-1α:

Transient knockdown of PGC-1α in control lamb PAECswas achieved via transfection with siRNA specific for human PGC-1α (3147831, Qiagen). The control treatment for the knockdown studies was non-silencing RNA (nsRNA, 1027292, Qiagen). Of note, siRNA transfections were also performed in HUVECs, in addition to control lamb PAECs, because PGC-1α knockdown in lamb PAECs led to a smaller percent decrease in PGC-1α (Supplemental Data 1 and 4). The reduced transfection efficiency (~30-40% knockdown) may be due to differences in human and sheep mRNA sequences for PGC-1α. HUVECs were previously used as representative ECs in cardiovascular disease^38,39, including in neonates with PPHN^24,40 and BPD,²⁴ because HUVECs perform angiogenesis in response to angiogenic stimuli^38,39 and display mitochondrial dysfunction in humans suffering from PPHN and BPD.²⁴ Transfecting normal HUVECs with human PGC-1α siRNA created at least a 50% knockdown of protein levels, which allowed for more specific assessment of the downstream effects on mitochondrial transcription factors and ETC complex proteins. We found that transient overexpression of PGC-1α in PPHN lamb PAECs can be achieved via transfection with either: 1. MycPGC-1 plasmid (10974, Addgene), or 2. PGC-1α adenoviral construct with Green Fluorescent Protein (GFP) tag (032008, ViraQuest, Inc). The control treatment for the overexpression studies in PPHN PAECs was: 1. Empty plasmid (Addgene) when transfecting with MycPGC-1 plasmid, and 2. No treatment when transfecting with PGC-1α adenovirus. Additionally, to assess how excess PGC-1α protein levels may affect healthy cells, adenovirus was also transfected in control lamb PAECs. Transfection of each cell group was performed when cells reached ≥70% confluence, and cells were collected 48 hours after transfection for additional studies.

Immunoblotting:

PAECs or HUVECs treated for their respective analyses were lysed in modified radio immune precipitation assay (RIPA) buffer. Using the modified Bradford method, protein concentrations were determined in the supernatants via the Pierce^™ BCA Protein Assay Kit (23225). Proteins were separated by SDS PAGE and transferred to a nitrocellulose membrane. Proteins were blotted with specific antibodies for PGC-1α (Abcam, Millipore), NRF-1 (LSBio), TFAM (Abcam), mitochondrial ETC complex proteins I-V (Abcam), GFP (Cell Signaling), and β-actin (Sigma-Aldrich). Representative blots with catalog number, MW, and dilution for these proteins are included in Supplemental Data 3. Membranes were blotted with horseradish peroxidase (HRP)-conjugated anti-mouse or -rabbit IgG antibody and exposed to CL-XPosure films (Pierce) after treatment with SuperSignal West Pico (Pierce). Signals were analyzed with ImageJ and normalized to β-actin.

In Vitro Angiogenesis Assays:

We assessed angiogenesis function of lamb PAECs with in vitro assays for tube formation, cell migration, and cell proliferation.

Tube Formation Studies:

PAEC tube formation was assessed in vitro using Matrigel, as previously reported.³⁴ After 48 hours of each respective transfection in a 6-well plate, the PAECs were added on top of the Matrigel for 8-14 hours and stained with eosin or calcein AM. We then took one representative picture per well using an Olympus IX50 inverted microscope 4x. We compared the total tube length and number of branches, nodes, and/or meshes for control and PPHN PAECs, with or without treatments. We only included tubular structures connecting two cell clusters in our measurements and considered cell clusters with at least three tubular structures emanating out to be a branching point. Quantification was performed using ImageJ Analyzer and confirmed with manual assessment.

Cell Proliferation Studies using BrdU Cell Proliferation Assay Kit (6813, Cell Signaling Technology):

Following 48 hours of each respective transfection in 4-well chamber slides, the PAEC media was replaced with diluted 10X-BrdU labeling solution with media for a 1-hour incubation at 37°C. Next, supernatant was aspirated, and cells were washed. Then, Fix/Denature Solution was added, cells were washed, and antibody diluent was aliquoted to each well. After incubation for 1 hour, wells were washed, and then diluted Anti-BrdU antibody was added; next, there was a wash followed by the addition of a secondary antibody HRP Conjugate to each well. Then, Substrate Solution was aliquoted to each well, including the blank wells, and solutions were incubated at RT on the orbital shaker for 30 minutes. Lastly, Stop Solution was added to each well. Results were read immediately using a spectrophotometer set at 450 nm absorbance for primary wavelength. Proliferation was assessed as being proportional to the increase in absorbance.

Cell Migration Studies using a Monolayer Scratch Recovery Test:¹⁵

After PAECs were treated for 48 hours for their respective transfection in a 6-well plate, we made a scratch line with a 1-ml pipette tip and washed away the scraped cells. After cells were observed to migrate into the margins (≥7 hours), cells were fixed with paraformaldehyde (PFA). Then, eosin or calcein AM stain was added. Cells were imaged with a fluorescent microscope, measuring the narrowest distance of the gap between frontlines of recovery.

Quantification of Mitochondrial DNA (mtDNA) Copy Number:

MtDNA and nuclear DNA copy numbers were quantified using ViiA 7 System multicolor real-time PCR detection system (Life Technologies). Following 48 hours of each respective transfection, genomic DNA was extracted from cultured PAECs grown in 6-well plates using Quick-DNA Miniprep Plus Kit (D4068, Zymo Research). 30 ng of purified genomic DNA was used for PCR as described previously.⁴¹ NADH Dehydrogenase 1, 2, 5, and 6 (ND1, ND2, ND5, and ND6) genes were selected from the D-loop of mtDNA and PCR sheep primers were designed using USC Genome Browser with sequences as follows: ND1 – forward 5′-ACGAGCCACAGAAGCATCAA -3′ and reverse 5′-GCCTGAGGATAGGGGAATGC-3′; ND2 – forward 5′-TACTTCAACCCATCGCCGAC-3’ and reverse 5′-ACGGCTAGGCTTGATATGGC-3′; ND5 – forward 5’-ACCCTTTCTCACCGGATTCT-3’ and reverse 5’-TTTGGGGGATTGTTATTGGA-3’; ND6 – forward 5’-GGGTTGGGTGTGGTATTGTG-3’ and reverse 5’-AACCCAGAATCCCCCGTAT-3’; β-Actin – forward 5’-CAGGCTGAAGTCAGGGTCTC-3’ and reverse 5’-GACTGAAGCCAGGACTGAGG-3’. Each primer’s specficity was confirmed by performing a blast search to ensure primer sequences were unique for the template sequence. The PCR cycle was started at 95°C for 10 minutes, followed by 40 cycles of 95°C for 15 seconds; the last cycles were 60°C for 30 seconds, followed by 72°C for 35 seconds. Melting temperatures were monitored and single-peak melting temperatures were observed for all primer pairs. MtDNA copy number was normalized to β-Actin copy number. The number of the threshold cycle (Ct) for each target DNA was corrected against the corresponding Ct of β-Actin to obtain the ΔCt, and then 2^−ΔΔCt was calculated against the corresponding control for estimation of DNA abundance.

Measurement of ATP Concentration:

To measure ATP levels in live cells, Luminescent ATP Detection Assay Kit (ab113849, Abcam) was utilized. After 48 hours of each respective transfection in a 96-well plate, media was replaced with detergent, including the wells containing ATP standards. After sealing and shaking the plate to lyse cells and stabilize ATP, Substrate Solution was added to each well. The plate was sealed and shaken for 5 minutes, and then allowed to dark adapt for 10 minutes. Luminescence was measured using SpectraMax i3x. ATP concentration was determined using the standard curve created from the luminescence values of the ATP standards.

Assessment of Mitochondrial Superoxide (O₂⁻) Levels:

MitoSOX Red Mitochondrial Superoxide Indictor (Invitrogen, M36007) quantifies mitochondrial O₂⁻ levels using selective fluorogenic dyes. After 48 hours of each respective transfection in a 6-well plate, cells were incubated with MitoSOX for 30 minutes to detect mitochondrial O₂⁻ levels. Fluorescence signals were quantified using fluorescence detection microscopy and analyzed with MetaView software. DAPI (4’,6-diamidino-2-phneylinodole) was used as a contrasting fluorescent signal to highlight the nucleus blue, and results were expressed as ratio of MitoSOX/DAPI intensity signals.

Statistical Analysis:

Data are shown as means ± SD. Graphpad Prism version 10.1 (Graphpad software, San Diego, CA) was used for statistical analysis of data, and normality was assessed using Shapiro-Wilk’s test. For comparison of two samples, unpaired T test was used for normally-distributed data, and Mann-Whitney U-test was used for data that did not pass the normality test. When two groups with different sample sizes were compared, i.e., imaging data with ≥2images per sample, Welch’s correction was used with the unpaired T test.

Results

Comparing control to PPHN lamb PAECs, PGC-1α protein level was significantly decreased in PPHN PAECs (Figure 1: p=0.0012). To determine whether angiogenesis dysfunction characteristic of the PPHN phenotype could be reproduced in vitro in ECs, PGC-1α protein level was knocked down in control lamb PAECs using siRNA (Supplemental Data 1); although PGC-1α knockdown achieved was statistically significant, the degree of knockdown was only 30-40%. We did not observe a statistically significant change in the downstream mitochondrial ETC complexes (Supplemental Data 4) which may be due to this smaller knockdown of PGC-1α in control lamb PAECs. Since HUVECs are commonly used ECs to define mitochondrial biology and angiogenesis alterations^24,38,39, we chose this cell type as an alternative for PGC-1α siRNA knockdown to study the downstream effects on mitochondrial function and angiogenesis. PGC-1α protein level was knocked down by 50-60% in HUVECs via siRNA (Figure 2a: p=0.0085). HUVECs treated with siRNA had significant reductions in the protein levels of NRF-1 (Figure 2b: p=0.0032), TFAM (Figure 2c: p=0.0032), and mitochondrial ETC complexes I-V compared to controls treated with nsRNA (Figures 2d-h: ETC I - p=0.0002, ETC II - p=0.0496, ETC III - p=0.0336, ETC IV - p=0.0006, and ETC V - p=0.0351).

Figure 1:

PGC-1α Protein Level in Lamb PAECs - PGC-1α protein is decreased in PPHN PAECs compared to controls; normalized to β-Actin. Data expressed as mean ± SD for n=4, and ** indicates p <0.01 by T test.

Figure 2:

PGC-1α Knockdown (KD) in HUVECs – (a) PGC-1α protein level is decreased in HUVECs with siRNA KD compared to HUVECs treated with nsRNA. (b) and (c) NRF-1 and TFAM protein levels are decreased in HUVECs with siRNA KD compared to nsRNA. (d)-(h) The mitochondrial ETC protein levels for complexes I-V are decreased in HUVECs with siRNA KD. All results normalized to β-Actin. Data expressed as mean ± SD for n=3, and * indicates p <0.05, ** p <0.01, and *** p <0.001 by T test.

Conversely, we assessed whether PGC-1α upregulation restored mitochondrial ETC complex expression in PPHN lamb PAECs. We found increased expression of PGC-1α after transfecting hypertensive PAECs with MycPGC-1 plasmid (Figure 3a: p=0.0122). PPHN PAECs transfected with MycPGC-1 plasmid showed significant increases in protein levels of mitochondrial ETC complexes I-III and V compared to PPHN PAECs treated with the empty plasmid, but there was no significant change in complex IV (Figures 3b-f: ETC I - p<0.0001, ETC II - p=0.0433; Figure 13 and 14: ETC III - p=0.0384, ETC IV - p=0.3924; Figure 15: ETC V - p=0.0462).

Figure 3:

PGC-1α Overexpression (OE) in PPHN Lamb PAECs – (a) PGC-1α protein is increased in hypertensive PAECs with MycPGC-1 OE compared to PPHN PAECs treated with an empty plasmid. (b)-(f) Mitochondrial ETC protein levels for complexes I-III and V are increased in hypertensive PAECs with MycPGC-1 OE. All results normalized to β-Actin. Data expressed as mean ± SD for n=3, and * indicates p <0.05 and *** p <0.001 by T test.

After observing the correlation between PGC-1α expression and mitochondrial function, we assessed whether altering PGC-1α levels affected angiogenesis in fetal lamb PAECs. While PGC-1α siRNA’s knockdown in control lamb PAECs was insufficient to assess the downstream effects on mitochondrial transcription factors, we assessed the effects on angiogenesis(Supplemental Data 1: p=0.0171). In the tube formation studies (Figure 4), control PAECs transfected with PGC-1α siRNA had a significantly reduced ability to form tubes compared to nsRNA-treated controls in terms of decreased total tube length (p<0.0001) and branch number (p=0.0004). In contrast, PPHN PAECs treated with MycPGC-1 plasmid had a significantly increased ability to form tubes compared to PPHN PAECs treated with an empty plasmid based on elevated total tube length (p<0.0001) and branch number (p=0.0003). For cell migration (Figure 5a), control PAECs with PGC-1α knockdown had a significantly greater gap, indicating decreased cell migration, compared to control PAECs treated with nsRNA (p<0.01). In contrast, PPHN PAECs with PGC-1α overexpression revealed a significantly reduced gap, suggesting increased cell migration, compared to PPHN PAECs treated with an empty plasmid (p<0.01). For cell proliferation (Figure 5b), control PAECs with PGC-1α knockdown had a significantly lower proliferation rate compared to control PAECs treated with nsRNA (p<0.0001); PPHN PAECs with PGC-1α overexpression displayed a significantly increased proliferation rate compared to PPHN PAECs treated with an empty plasmid (p<0.0001).

Figure 4:

Tube Formation: Control lamb PAECs with PGC-1α KD using siRNA have decreased total tube length and branch number compared to normotensive controls treated with nsRNA. PPHN lamb PAECs with PGC-1α OE using a myc-plasmid show increased total tube length and branch number compared to PPHN PAECs treated with an empty plasmid. Analysis performed manually on 40x images, but images shown are 10x. Scale bar is 100 μm. Data expressed as mean ± SD for n=3 (with at least two images analyzed per sample), and *** indicates p <0.001 by T test, with Welch’s correction if unequal sample sizes given replicate images reviewed.

Figure 5:

(a) Monolayer Scratch Recovery Test - Control lamb PAECs with PGC-1α KD show a significantly greater gap compared to normotensive controls treated with nsRNA. PPHN lamb PAECs with PGC-1α OE show a significantly decreased gap distance compared to hypertensive controls treated with an empty plasmid. Images shown and analyzed at 4x, and scale bar is 100 μm. Data expressed as mean ± SEM for n =3 (at least two image analyzed per sample): nsRNA - 572.5±10.8 μm, siRNA - 768.7±19.6 μm, empty plasmid - 742.1±19.2 μm, mycPGC-1 - 637.5±13.6 μm. ** indicates p <0.01 by T test, with Welch’s correction if unequal sample sizes given replicate images reviewed. (b) BrdU Incorporation Assay – Control lamb PAECs with PGC-1α KD display significantly less cell proliferation compared to nsRNA treatment. PPHN lamb PAECs with PGC-1α OE show significantly more cell proliferation compared to hypertensive PAECs treated with an empty plasmid. Data expressed as mean ± SEM for n=3, and *** indicates p <0.001 by T test.

To assess the effect of PGC-1α overexpression in control lamb PAECs, an adenoviral construct was used to enhance PGC-1α’s expression, and transfection was confirmed by immunoblotting for the GFP tag (Supplemental Data 2a: GFP – p<0.0001). For tube formation, there was a significant increase in the number of nodes (p=0.0049), number of meshes (p=0.0052), and branching length (p=0.0124) for control PAECs treated with PGC-1α adenovirus compared to untreated controls (Supplemental Data 2b). For cell migration, there was a significant decrease in the gap between frontlines of recovery in control PAECs with PGC-1α overexpression relative to untreated controls (Supplemental Data 2c: p= 0.0108).

To define PGC-1α’s role in regulating mitochondrial biogenesis, relative mtDNA expression was evaluated. In control lamb PAECs with PGC-1α knockdown via siRNA, there was no significant change in relative mtDNA expression (Figure 6a- ND1, p=0.7602; ND2, p=0.7007; ND5, p=0.3778; and ND6, p=0.2316), and in PPHN lamb PAECs with PGC-1α overexpression via adenovirus, there was also no significant change in relative mtDNA expression (Figure 6b- ND1, p=0.7601; ND2, p=0.3609; ND5, p=0.2661; and ND6, p=0.2680). Given PGC-1α’s more focused regulation of ETC complex expression, we measured the total levels of cellular ATP to assess PGC-1α’s effect on oxidative phosphorylation efficiency. PGC-1α knockdown in control PAECs did not affect ATP production (Figure 6c: p=0.7798) likely due to inadequate decrease in PGC-1α protein levels; contrastingly, PGC-1α overexpression in PPHN PAECs did generate higher ATP levels compared to untreated PPHN PAECs (Figure 6d: p=0.0167). Lastly, we evaluated mitochondrial O₂⁻ levels in control and PPHN lamb PAECs following PGC-1α manipulation. In control PAECs treated with PGC-1α siRNA, there was a significant increase in mitochondrial O₂⁻ levels compared to control PAECs treated with nsRNA (Figure 6e: p=0.014). In PPHN PAECs treated with PGC-1α adenovirus, there was a significant decrease in mitochondrial O₂⁻ levels relative to untreated PPHN PAECs (Figure 6f: p=0.0367). The increase in O₂⁻ levels seen even with limited (~30-40%) knockdown of PGC-1α protein in control lamb PAECs was likely due to decreased transcription or function of antioxidant genes. In contrast, we did not see a change in levels of ATP (Figure 6c) and mitochondrial ETC complex expression (Figure 4b-f) with 30% knockdown of PGC-1α levels. Our data suggests that redox balance is more sensitive to small perturbation in PGC-1α protein levels compared to ETC complex protein expression. Despite the mitochondrial ETC complexes being unaffected by limited PGC-1α knockdown in control lamb PAECs, angiogenesis function worsened (Figures 4-5) likely due to heightened oxidative stress, as we previously reported.³⁴

Figure 6:

(a)-(b) Mitochondrial DNA Copy Number Studies – There is no significant change in mtDNA expression in control lamb PAECs treated with PGC-1α siRNA compared to nsRNA. There is no significant change in mtDNA expression in PPHN lamb PAECs treated with PGC-1α adenovirus compared to untreated PPHN PAECs. (c)-(d) Luminescent ATP Detection Assay – There is no significant change in cellular ATP levels in control lamb PAECs treated with PGC-1α siRNA compared to nsRNA. There is a significant increase in cellular ATP levels in PPHN lamb PAECs treated with PGC-1α adenovirus compared to untreated PPHN PAECs. (e)-(f) Mitochondrial Superoxide (O₂⁻) Level Assessment via MitoSOX – There is significantly higher detection of O₂⁻ levels (red) in control lamb PAECs treated with PGC-1α siRNA compared to nsRNA. There is significantly lower detection of O₂⁻ levels in PPHN lamb PAECs treated with PGC-1α adenovirus compared to those without treatment. Images also stained with DAPI to highlight the nucleus as blue. Images analyzed and shown at 40x, and scale bar is 20 μm. For Figure 6 (a)-(f), data expressed as mean ± SD for n=3, and * indicates p <0.05 by T test.

Discussion

Our studies reveal that the expression of PGC-1α, a key mitochondrial transcription factor, is decreased in PAECs in our lamb model of PPHN. Using gain- and loss-of-function studies, we discovered that PGC-1α regulates the expression of the mitochondrial ETC complexes and transcription factors, NRF-1 and TFAM, in ECs. While PGC-1α regulates the expression of the ETC complexes I-V, an aspect of mitochondrial biogenesis, it does not appear to control mtDNA copy number as it does in other cell types.⁴² Our results reveal that in lamb PAECs, PGC-1α regulates aspects of mitochondrial biogenesis that enhance the efficiency of oxidative phosphorylation by not only producing higher ATP levels via increasing the expression of the ETC complexes but also by maintaining redox balance by limiting O₂⁻ levels. These enhancing effects on mitochondrial oxidative phosphorylation and redox balance may serve as the mechanistic link explaining PGC-1α’s regulation of angiogenesis in PAECs.³⁴ Our data suggest that upregulating PGC-1α protein is a potential therapeutic target to enhance the efficiency of mitochondrial oxidative phosphorylation and minimization of oxidative stress to improve angiogenesis in PPHN.

As an important regulatory protein in metabolism, PGC-1α has been studied extensively in other disease processes. One such disease is diabetes mellitus, with research particularly focusing on how PGC-1α overexpression can restore metabolic homeostasis via its control of mitochondrial biogenesis and upregulation of oxidative phosphorylation genes. A study by Patti M et al compared gene expression in human skeletal muscle tissue from non-diabetic and diabetic samples, and not surprisingly, found that PGC-1α and its downstream effector, NRF-1, are both decreased in diabetic subjects with altered glucose regulation and metabolism.⁴³ Another study found that when murine mature adipocytes are treated with a commonly-used diabetic medication, a sodium glucose cotransporter 2 (SGLT2) inhibitor, energy expenditure increased via induction of the AMPK-Sirtuin-1-PGC-1α pathway due to elevated mitochondrial number and immunofluorescent intensity, indicative of enhanced mitochondrial biogenesis.⁴⁴ Our results also support the role of PGC-1α in maintaining aspects of mitochondrial biogenesis pertaining to oxidative phosphorylation efficiency.

Focusing on blood vessel growth, Saint-Geniez M et al reported that PGC-1α regulates angiogenesis within multiple retinal blood vessel cell types by inducing vascular endothelial growth factor (VEGF) expression.⁴⁵ While our observation of PGC-1α’s control of mitochondrial biogenesis is not a newly-identified role for PGC-1α, its influence over mitochondrial biogenesis as it relates to oxidative phosphorylation efficiency and redox stress minimization in PAECs is novel, especially as a mechanistic link to angiogenesis. Interestingly, Sawada N et al studied mice with PGC-1α overexpression specifically in ECs treated with streptozotocin (STZ) to induce a diabetic phenotype, and found reduced EC migration in those animals with PGC-1α overexpression.⁴⁶ Their hypothesis was that angiogenesis required increased glycolytic flux at the expense of fuel consumption by mitochondria in the setting of diabetes, a stressed metabolic state.⁴⁶ While this study’s findings contrasted our data, our models differ in that Sawada’s study assessed PGC-1α’s effect on angiogenesis after stress was induced with STZ, causing insulin resistance;⁴⁶ however, the PPHN phenotype in our model develops after creation of a high-pressure system in the PA, following ductal constriction, which is associated with elevated ROS in mitochondria. It is possible that PGC-1α may have distinct functions depending on the type of stress encountered. If PGC-1α fails to be induced in the setting of increased ROS and reduced ATP in PPHN, cells are unable to perform oxidative phosphorylation, the most efficient way to produce energy. Consequently, cells switch to glycolysis to produce ATP, generating significantly less energy. Without efficient energy production, they are unable to adapt to metabolic and redox stress with increased angiogenesis, resulting in PPHN.²⁰

Our studies are limited in that all experiments are performed in vitro. Further studies are needed to investigate whether these same effects of PGC-1α manipulation occur in large animal models of PPHN. The specific knowledge gaps remaining include: 1. How PGC-1α alters lung histology in terms of vascular remodeling and alveoli oversimplification; and 2. Global effects of PGC-1α overexpression in blood vessels of other tissues, i.e., brain or eyes. Additionally, given our lower sample sizes, we were unable to assess the sex differences in PGC-1α expression. Moreover, while our animal model is one of the possible etiologies resulting in the PPHN phenotype in neonates, replicating the PPHN subtype that presents with severe hypoxemia at birth, it does not encompass all etiologies of PPHN.

Since we achieved insufficient knockdown of PGC-1α protein levels in control lamb PAECs, we could not provide a direct comparison of PPHN lamb PAECs with PGC-1α overexpression to control lamb PAECs with PGC-1α knockdown. However, our HUVEC studies support our hypothesis that a reduction of PGC-1α generates findings consistent with the PPHN phenotype, including elevated ROS, reduced mitochondrial oxidative phosphorylation efficiency, and decreased angiogenesis.

In conclusion, our experiments may have translational significance as this mechanism offers a new potential therapeutic target in PPHN. If PGC-1α is mechanistically linked to PPHN, strategies to restore its function may improve postnatal transition in PPHN. One such approach is to use allosteric AMPK activators, including Metformin, A769662, and α-lipoic acid, which increase AMPK function and result in enhanced PGC-1α protein levels.²⁰ Our studies provide the rationale and foundational evidence for testing these therapies in PPHN.

Impact.

• Reveals a novel mechanism for angiogenesis dysfunction in persistent pulmonary hypertension of the newborn (PPHN).

• Identifies a key mitochondrial transcription factor, Peroxisome Proliferator-Activated Receptor Gamma Co-Activator-1α (PGC-1α), as contributing to the altered adaptation and impaired angiogenesis function that characterizes PPHN through its regulation of mitochondrial function and oxidative phosphorylation.

• May provide translational significance as this mechanism offers a new therapeutic target in PPHN, and efforts to restore PGC-1α expression may improve postnatal transition in PPHN.

Data Availability Statement:

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.