Perinatal Hypoxia-Inducible Factor Stabilization Preserves Lung Alveolar and Vascular Growth in Experimental Bronchopulmonary Dysplasia

Abstract

Rationale: Antenatal inflammation with placental dysfunction is strongly associated with high bronchopulmonary dysplasia (BPD) risk in preterm infants. Whether antenatal or postnatal HIF (hypoxia-inducible factor) augmentation can preserve lung structure and function and prevent pulmonary hypertension after intrauterine inflammation is controversial.

Objectives: To determine whether antenatal or postnatal prolyl-hydroxylase inhibitor (PHi) therapy increases lung HIF expression, preserves lung growth and function, and prevents pulmonary hypertension in a rat model of chorioamnionitis-induced BPD caused by antenatal inflammation.

Methods: Endotoxin (ETX) was administered to pregnant rats by intraamniotic injection at Embryonic Day 20, and pups were delivered by cesarean section at Embryonic Day 22. Selective PHi drugs, dimethyloxalylglycine or GSK360A, were administered into the amniotic space at Embryonic Day 20 or after birth by intraperitoneal injection for 2 weeks. Placentas and lung tissue were collected at birth for morphometric and Western blot measurements of HIF-1a, HIF-2a, VEGF (vascular endothelial growth factor), and eNOS (endothelial nitric oxide synthase) protein contents. At Day 14, lung function was assessed, and tissues were harvested to determine alveolarization by radial alveolar counts, pulmonary vessel density, and right ventricle hypertrophy (RVH).

Measurements and Main Results: Antenatal PHi therapy preserves lung alveolar and vascular growth and lung function and prevents RVH after intrauterine ETX exposure. Antenatal administration of PHi markedly upregulates lung HIF-1a, HIF-2a, VEGF, and eNOS expression after ETX exposure.

Conclusions: HIF augmentation improves lung structure and function, prevents RVH, and improves placental structure following antenatal ETX exposure. We speculate that antenatal or postnatal PHi therapy may provide novel strategies to prevent BPD due to antenatal inflammation.

At a Glance Commentary

Scientific Knowledge on the Subject

Antenatal inflammation, as observed in the clinical setting of chorioamnionitis, is strongly associated with a high risk for bronchopulmonary dysplasia (BPD) in preterm infants, but causal mechanisms are poorly understood. Early disruption of vascular growth contributes to the pathobiology of BPD, and HIF (hypoxia-inducible factor), a key regulator of angiogenesis, may be decreased in BPD. Prolyl-hydroxylase inhibitor (PHi) drugs augment HIF signaling, but whether antenatal or postnatal PHi therapy can preserve lung structure and function and prevent pulmonary hypertension in experimental BPD due to intrauterine inflammation is uncertain.

What This Study Adds to the Field

Intrauterine inflammation impairs HIF signaling in the developing lung, and augmentation of HIF-1a and HIF-2a with PHi therapy in utero or after birth preserves lung vascular and alveolar growth and prevents PH during infancy. Antenatal PHi therapy inhibits the adverse effects of intrauterine inflammation on placental structure, including spiral-artery remodeling, which may provide a mechanism through which antenatal intervention can enhance cardiopulmonary outcomes after exposure to intrauterine stress. Targeted enhancement of placental vascular function may provide a novel strategy to improve cardiorespiratory outcomes in at-risk preterm infants exposed to antenatal stress.

Bronchopulmonary dysplasia (BPD) is the chronic lung disease of prematurity that is caused by early lung injury during antenatal and early postnatal life (1–5). Characterized by an arrest of lung growth (3), BPD is associated with high mortality, a prolonged need for mechanical ventilation and neonatal ICU hospitalization, and other comorbidities, including pulmonary hypertension (PH) (4–10). After discharge from the hospital, infants with BPD continue to have significant cardiorespiratory disease due to impaired lung and cardiac structure and function, leading to persistent cardiovascular and pulmonary abnormalities and substantial healthcare needs and costs throughout childhood and into adulthood (6–10).

Historically, BPD has been considered to primarily result from postnatal insults to the preterm lung due to ventilator-induced injury, hyperoxia, infection, and inflammation (1). Although advances in clinical practice have mitigated the negative impact of postnatal insults, the overall incidence of BPD has remained unchanged over the past several decades (2, 4, 6). Although the high risk for BPD reflects increasing survival of extremely low-birth-weight preterm infants, there has also been a growing recognition that antenatal determinants are strong contributors to the etiology and severity of BPD beyond postnatal injury alone (11–13). Epidemiologic data show strong associations of antenatal factors, such as chorioamnionitis (CA), and placental dysfunction with BPD susceptibility (14–26). Importantly, antenatal risk factors as identified on the first day of life are strongly associated with the subsequent development of BPD and late respiratory disease during infancy (13, 19, 26, 27) and suggest that the early identification of at-risk preterm infants may provide a greater opportunity for applying interventions that target BPD prevention.

Exposure to postnatal hyperoxia has been most often used as a preclinical model to study the pathogenesis of BPD; however, the increased recognition of the importance of antenatal factors on respiratory outcomes has stimulated the development of animal models to explore fetal mechanisms that contribute to BPD (13). Past studies have shown that antenatal injection of endotoxin (ETX) to induce inflammation and mimic CA causes sustained abnormalities of impaired lung alveolar and vascular growth throughout infancy, which mimics features of human BPD, even in the absence of postnatal hyperoxia or mechanical ventilation (27–29). Diverse inflammation-mediated mechanisms can contribute to the abnormal lung development that causes BPD, including early disruptions to critical growth-factor signaling, such as those related to VEGF (vascular endothelial growth factor) (30).

Pharmacologic or genetic disruption of angiogenesis is sufficient to impair alveolar and vascular growth in the developing lung, causing sustained abnormalities of lung structure and function that have features of human BPD lungs (31–37). Past studies have shown that exogenous VEGF therapy can enhance endothelial survival and function and improve lung-structure damage due to intrauterine stress (37) or postnatal hyperoxia (38–40). HIFs (hypoxia-inducible factors), known as potent upstream regulators of diverse signaling pathways, including VEGF-induced angiogenesis, may play an important role in supporting normal lung development (41–45). In the presence of oxygen, HIF protein expression is negatively regulated by a family of prolyl-hydroxylases that lead to HIF ubiquitination and proteasomal degradation (41–45). In the low-Po₂ environment of fetal life, strong HIF expression is detected in the developing human lung as early as at 8 weeks of gestation, remains elevated throughout fetal life, and then rapidly declines after preterm or term birth (46–54). HIF-1a and HIF-2a isoforms have distinct patterns of spatial expression in the lung, which likely reflect functional differences between these subtypes (41–44, 55). Genetic homozygous deletion of HIF-1a and HIF-2a isoforms causes fatal cardiovascular malformations, impaired lung development, and severe respiratory distress at birth, further emphasizing the importance of HIF signaling in the immature lung (41–44, 55). Experimentally, HIF signaling can be augmented by pharmacologic treatment with prolyl-hydroxylase inhibitor (PHi) drugs, which stabilize HIF (41, 42, 45–49). PHi therapy has promising potential for therapeutic roles in diverse clinical settings, such as anemia with chronic renal failure, and past studies have shown that dimethyloxalylglycine (DMOG), an established PHi, can improve lung structure in rodent models of BPD due to hyperoxia and mechanical ventilation (46, 54, 55).

Despite these findings, much controversy exists regarding the potential role for augmentation of HIF signaling in different diseases. Whereas postnatal models of BPD suggest a critical role for HIFs in disease pathobiology and potential therapy (47, 48, 56, 57), little is known about the role of HIFs in the setting of antenatal models of BPD, such as observed with CA, or about whether early augmentation of HIF signaling in the perinatal period can cause sustained improvement in lung growth and prevent PH throughout infancy. Importantly, HIF signaling is vital for placental vascular development (58–60); however, controversies persist regarding potential adverse effects of sustained HIF upregulation in the placenta, which may potentially worsen fetal distress before birth because of placental dysfunction (58–63). Adding to further controversy, experimental models of adult pulmonary fibrosis and PH suggest that enhanced HIF signaling, especially through increased HIF-2a expression and impaired prolyl-hydroxylase activity, is sufficient to induce severe PH and right ventricle (RV) hypertrophy (RVH) (44, 45, 64–69). Whether sustained upregulation of HIF during late fetal life would improve or worsen the risk for abnormal lung development and placental structure, especially in the setting of antenatal stress, is uncertain.

Thus, to further explore the potential roles of HIF signaling in disease pathogenesis and the impact of PHi therapy in experimental BPD due to antenatal mechanisms of disrupted lung development, we proposed a series of studies in a model of BPD caused by intrauterine inflammation, as observed in CA. We hypothesized that intrauterine inflammation impairs HIF signaling in the developing lung and that augmentation of HIF with PHi therapy in utero or during the early postnatal period after birth, would preserve lung vascular and alveolar growth and prevent PH during infancy. We further tested the hypothesis that antenatal PHi therapy would mitigate the adverse effects of intrauterine inflammation on placental structure, which may provide a mechanism through which antenatal intervention enhances pulmonary outcomes of the exposed offspring.

Methods

For details regarding the methods used in this study, see the online supplement.

Animals

All protocols and procedures were approved by the Animal Care and Use Committee at the University of Colorado Anschutz Medical Center. Time-dated pregnant Sprague-Dawley rats were purchased from Charles River Laboratories and were maintained in room air for 1 week before giving birth.

Study Design

At Day 20 of gestation (term = Day 22 of gestation), pregnant rats received intraamniotic injections of ETX to induce antenatal inflammation (28, 29). As illustrated in Figure 1, pregnant rats were prepared for receiving intraamniotic injections with laparotomy at 20 days of gestation (term = 22 d) (28, 29). Pregnant rats were randomly assigned to receive normal saline (NS; 50 μl/amniotic sac) for the control group or ETX (10 μg of Escherichia coli 055:B55 diluted with 50 μl of NS/sac) for the CA model.

Figure 1.

Experimental model of chorioamnionitis caused by intraamniotic injections of ETX, with timeline to illustrate the study design used for determining the antenatal and postnatal effects of prolyl-hydroxylase inhibitor treatment on lung structure and function in infant rats. C-section = cesarean section; DMOG = dimethyloxalylglycine; e = Embryonic Day; ETX = endotoxin.

Studies on the Effects of Antenatal PHi Treatment

For studies examining the effects of antenatal PHi therapies, pregnant rats that received NS (control animals) or ETX were randomized to receive intraamniotic injections either of NS or of DMOG (Sigma-Aldrich) or GSK630A (GlaxoSmithKline). Two PHi agents were included in this study to provide stronger evidence regarding the role of HIF signaling in this model. Either DMOG (10 mg/s) or GSK360A (1 mg/s) was injected into the amniotic sac after NS or ETX had been administered. Two days after intraamniotic injections, cesarean section was performed on pregnant rats under general anesthesia with isoflurane inhalation (28, 29). All rat pups in the injected amniotic sacs were delivered within 5 minutes after the onset of anesthesia. Some pups from each of the study groups were randomly selected and killed on the first day of life (Day 0) to harvest tissue for Western blot analysis. The remaining pups were weighed and maintained with their littermates over the 2-week postnatal period and were killed at Day 14 for study endpoints.

Studies on the Effects of Postnatal PHi Treatment

To examine the effects of postnatal HIF stabilization, neonatal pups from litters of pregnant rats that received either NS (control animals) or ETX were randomized for treatment either with NS or with DMOG (5 mg/kg) or GSK360A (5 mg/kg). Beginning on Day 1, doses were administered by intraperitoneal injections every other day for 2 weeks (Figure 1).

Study Endpoints

Several newborns from each of the study groups were killed shortly after cesarean-section delivery on Day 0 to obtain lung tissue. Placentas were collected and fixed in paraformaldehyde immediately after cesarean-section delivery. At 14 days of age, tissues were harvested for histologic and morphometric analysis and assessment of RVH after performing lung-function tests.

Placental studies

To assess the effects of ETX and PHi treatment with GSK360A on placental structure, placentas were harvested at birth and immediately placed for tissue fixation in 4% paraformaldehyde until study. Tissue analysis included hematoxylin and eosin staining for routine histologic assessment, immunofluorescent staining for cytokeratin, and DAPI to evaluate spiral-artery transformation and trophoblast distribution within the placenta, as previously described (60).

Lung function

Lung function was determined in 14-day-old anesthetized rats by using the flexiVent system (flexiVent; SCIREQ), which measures total respiratory system resistance and compliance according to standard methods from the manufacturer, as previously described (29).

Tissue for histologic analysis

Lungs were inflated with 4% paraformaldehyde and maintained at 20-cm H₂O pressure for 60 minutes. A 2-mm-thick transverse section was taken from the midplane of the right lower lobe and left lobe of the fixed lungs of each animal, respectively, to process and embed in paraffin wax.

Immunohistochemistry

Slides with 5-μm paraffin sections were stained with hematoxylin and eosin for assessment of alveolar structure by morphometric analysis and vWF (von Willebrand Factor), an endothelial cell–specific marker, for assessment of vascular density.

Morphometric analysis

Radial alveolar counts (RAC) and pulmonary vessel density (PVD) were determined by standard morphometric techniques (70, 71). Alveolarization was assessed by standard RAC methods, as previously described (28, 29, 70, 71). PVD was determined by counting vWF-stained vessels with an external diameter of ≤50 μm/high-power field.

Assessment of RVH by Fulton’s index

The RV and left ventricle plus septum were dissected and weighed. The weight ratios of the RV to the left ventricle plus septum and the RV to the body were determined to evaluate RVH.

Western blot analysis

Frozen lung tissue was homogenized and separated in nuclear and cytosolic extracts. Nuclear-extract membranes were incubated overnight in either anti–HIF-1a monoclonal antibody (mAb; 1:2,000 dilution) or anti–HIF-2a mAb (1:2,000 dilution). Whole-lung lysates were incubated overnight with anti-eNOS (anti–endothelial nitric oxide synthase) mAb (1:2,000 dilution) or anti-VEGF mAb (1:2,000 dilution).

Statistical Analysis

Statistical analysis was performed with the GraphPad Prism 8.0 software package (GraphPad Software). Statistical comparisons were made between groups using the t test or ANOVA with Kruskal-Wallis/Dunn post hoc analysis for significance. P < 0.05 was considered significant. Data are presented as mean ± SEM.

Results

Effects of Antenatal DMOG Treatment in a CA-induced Model of BPD

Intraamniotic ETX injection causes sustained changes in lung alveolar and vascular structure in the offspring in the absence of postnatal hyperoxia or ventilation (Figure 2 and Figure E1 in the online supplement). As assessed at Postnatal Day 14, antenatal ETX administered to rats without hyperoxia decreased distal airspace growth as reflected by a 42% decrease in RAC in comparison with administration of saline to control animals (P < 0.01) (Figure 2). ETX-exposed fetal rats that were treated with concurrent intraamniotic DMOG injections demonstrated improved distal lung structure, as quantified by a 45% increase in RAC (P < 0.01) in comparison with the group exposed to ETX alone. Animals that received DMOG injections without ETX had RAC that were not different from control animals receiving saline alone. ETX exposure reduced PVD in comparison with saline, and antenatal DMOG treatment increased PVD by 56% as compared with ETX alone (P < 0.05) (Figures 2 and E1).

Figure 2.

Effects of antenatal dimethyloxalylglycine (DMOG) treatment on distal lung structure in infant rats after antenatal intraamniotic injection of endotoxin (ETX) in experimental chorioamnionitis. (A–C) In comparison with saline treatment (A) and DMOG treatment (B) alone, prenatal exposure to ETX caused lung simplification (C). (D) Antenatal DMOG treatment improved lung histology, as quantified by changes in radial alveolar counts and vessel density (presented as histograms). As shown, antenatal DMOG treatment reduced right ventricle hypertrophy (bottom row, left panel), increased lung compliance (middle panel), and decreased total lung resistance (right panel) in rats exposed to ETX before birth. Micrographs are representative and were obtained at the same magnification (scale bars, 250 μm). Histogram data are shown as mean ± SEM. *P < 0.01 versus ETX and ^#P < 0.05 versus ETX. Animal numbers per study group: control, n = 10; ETX, n = 15; DMOG, n = 10; and ETX + DMOG, n = 15. LV = left ventricle; RV = right ventricle; S = septum.

To determine the long-term physiologic effects of antenatal DMOG after fetal exposure to ETX, we assessed RVH and lung mechanics at Postnatal Day 14 (Figure 2). When compared with saline control animals, rats exposed to ETX had an increase in RVH of 70% (P < 0.01), and antenatal DMOG treatment significantly reduced RVH (P < 0.01). Lung mechanics as measured at Day 14 demonstrated improvements in total respiratory system compliance and resistance after antenatal DMOG treatment. ETX exposure alone decreased lung compliance by 24% (P < 0.01) and increased lung resistance by 29% (P < 0.01). When compared with ETX exposure alone, ETX with DMOG treatment increased lung compliance by 59% (P < 0.01) and decreased lung resistance by 43% (P < 0.01).

Effects of Postnatal DMOG Treatment

To determine the effects of postnatal DMOG treatment after exposure to intrauterine ETX, we compared the effects of DMOG or saline treatment (control animals) during the first 2 weeks of life on lung airspace (Figure 3) and vascular structure (Figures 3 and E2). As with antenatal DMOG treatment, postnatal DMOG improved alveolarization as assessed by RAC and increased PVD to values observed in control animals. Postnatal DMOG treatment reduced RVH and improved lung mechanics (Figure 3).

Figure 3.

Effects of postnatal dimethyloxalylglycine (DMOG) treatment on distal lung structure in infant rats after antenatal endotoxin (ETX) exposure in experimental chorioamnionitis. (A–C) In comparison with saline (A) and DMOG (B) control treatments, prenatal exposure to ETX caused lung simplification (C). (D) Postnatal DMOG treatment improved lung histology, as quantified by changes in radial alveolar counts and vessel density, as shown by histogram. As shown, antenatal DMOG treatment reduced right ventricle hypertrophy (bottom row, left panel), increased total respiratory system compliance (middle panel), and decreased total respiratory system resistance (right panel) in rats exposed to ETX before birth. Micrographs are representative and were obtained at the same magnification (scale bars, 250 μm). Histogram data are shown as mean ± SEM. *P < 0.01 versus ETX and ^#P < 0.05 versus ETX. Animal numbers per study group: control, n = 10; ETX, n = 15; DMOG, n = 10; and ETX + DMOG, n = 15. For definition of abbreviations, see Figure 2.

Effects of Antenatal GSK360A Treatment

To further explore the role of HIF augmentation, we performed similar experiments with a second PHi, GSK360A, in infant rats exposed to intraamniotic ETX. As shown in Figure 4, antenatal GSK360A treatment improved lung airspace structure and increased RAC when compared with exposure to ETX alone (Figure 4). Similarly, antenatal GSK360A treatment improved PVD by 52% when compared with ETX exposure alone (P < 0.01) and preserved vascular density at control amounts in ETX-exposed pups (P < 0.01; Figures 4 and E3). Antenatal GSK360A treatment of ETX-exposed pups caused a 46% reduction in RVH, increased total lung compliance by 40%, and decreased total lung resistance by 35%, as compared with no GSK360A treatment of ETX-exposed pups (Figure 4).

Figure 4.

Effects of antenatal GSK360A (GSK) treatment on distal lung structure in infant rats after antenatal intraamniotic injection of endotoxin (ETX) in experimental chorioamnionitis. In comparison with (A) saline and (B) GSK control treatments, (C) prenatal exposure to ETX caused lung simplification. (D) Antenatal GSK treatment improves lung histology, as quantified by changes in radial alveolar counts and vessel density, as shown by histogram. In comparison with ETX exposure combined with saline (control treatments), antenatal GSK prevented right ventricle hypertrophy (bottom row, left panel), improved total respiratory system compliance (middle panel), and improved total respiratory system resistance (right panel) in infant rats. Micrographs are representative and were obtained at the same magnification (scale bars, 250 μm). Histogram data are shown as mean ± SEM. *P < 0.01 versus ETX. Animal numbers per study group: control, n = 10; ETX, n = 15; GSK, n = 10; and ETX + GSK, n = 15. For definition of abbreviations, see Figure 2.

Effects of Postnatal GSK360A Treatment

To evaluate the role of postnatal PHi on lung growth after antenatal ETX exposure, GSK360A was administered to ETX-exposed pups during the postnatal period alone. Pups exposed to intraamniotic ETX and treated after birth with GSK360A demonstrated an increase in RAC of 48% (P < 0.01) and a 36% increase in PVD (P < 0.01; Figure 5) as compared with control animals. Postnatal GSK360A decreased RVH by 41%, increased lung compliance by 33% (P < 0.01), and reduced lung resistance by 40%, as compared with ETX exposure alone (Figures 5 and E4).

Figure 5.

Effects of postnatal GSK360A (GSK) treatment on distal lung structure in infant rats after antenatal endotoxin (ETX) exposure in experimental chorioamnionitis. (A–C) In comparison with saline (A) and GSK (B) control treatments, prenatal exposure to ETX caused lung simplification (C). (D) Postnatal GSK treatment improved lung histology, as quantified by changes in radial alveolar counts and vessel density, as shown by histogram. In comparison with ETX exposure and saline treatment (control treatment), antenatal GSK prevented right ventricle hypertrophy (bottom row, left panel), improved total respiratory system compliance (middle panel), and improved total respiratory system resistance (right panel) in infant rats. Micrographs are representative and were obtained at the same magnification (scale bars, 250 μm). Histogram data are shown as mean ± SEM. *P < 0.01 versus ETX. Animal numbers per study group: control, n = 10; ETX, n = 15; GSK, n = 10; and ETX + GSK, n = 15. For definition of abbreviations, see Figure 2.

Effects of Antenatal PHi Treatment on Lung HIF-Isoform, VEGF, and eNOS Protein Expression

To determine the effects of antenatal ETX exposure and PHi treatment on HIF–VEGF signaling, we performed Western blot studies of whole-lung homogenates from rat pups on the first day of life after exposure to saline, ETX, or ETX with DMOG treatment (Figure E5). As shown, animals receiving DMOG treatment showed increased protein expression of downstream proangiogenic factors, VEGF and eNOS, when compared with control animals. As HIF is transcriptionally active in the nucleus, we further performed assays using nuclear extractions from lung tissue to assess the effects of antenatal PHi on active HIF-isoform protein concentrations. Antenatal treatment with either DMOG or GSK360A increased expression of HIF-1a and HIF-2a when compared with saline or ETX alone (Figure 6).

Figure 6.

Effects of antenatal prolyl-hydroxylase inhibitor treatment on lung HIF-1a (hypoxia-inducible factor-1a) and HIF-2a nuclear protein expression after birth after intrauterine endotoxin (ETX) treatment. (A–D) As shown, intraamniotic injections of DMOG (A and B) and GSK360A (GSK; C and D) markedly increased lung HIF-1a (A and C) and HIF-2a (B and D) protein in both saline- and ETX-exposed newborn rat pups. Histograms present data as mean ± SEM. *P < 0.05 versus ETX and ^#P < 0.05 for comparisons indicated by brackets in C and D. n = 3 for each study group. DMOG = dimethyloxalylglycine.

Effects of Antenatal PHi Treatment on Placental Vascular Structure

To explore the effects of HIF augmentation on placental structure, we examined placental tissue from pregnancies after saline injections (control animals), ETX exposure, and treatment with GSK360A. Antenatal GSK360A treatment normalized spiral-artery morphology and histologic pattern of the surrounding trophoblast invasion (Figure 7). As shown, antenatal ETX exposure caused an involution of spiral-artery transformation (Figure 7A) with a concomitant decrease in surrounding cytokeratin-positive invasive trophoblast cells (Figure 7B). Intraamniotic GSK360A treatment restored placental spiral-artery transformation in comparison with saline treatment with ETX exposure. Quantitative analysis showed a marked reduction of the percentage of transformed blood vessels per 40× field but did not show a reduction in the percentage of total blood vessels after ETX exposure, which was increased with GSK360A treatment (Figure 7C).

Figure 7.

Effects of antenatal GS360A (GSK) treatment on placental vascular structure. (A and B) In comparison with saline control treatment, intraamniotic endotoxin (ETX) exposure altered spiral-artery remodeling (A) and decreased surrounding trophoblast invasion (B). GSK treatment restored placental spiral-artery transformation in comparison with saline-treated ETX exposure. (C) Quantitative analysis shows a marked reduction of the percentage of transformed blood vessels per 40× field but does not show a reduction in the percentage of total blood vessels after ETX exposure, which was increased with GSK treatment. (A) Representative images of hematoxylin and eosin histology of rat spiral-artery morphology. (B) Representative images of immunofluorescent staining of pan cytokeratin, a marker of invasive cytotrophoblasts. Red stain = cytokeratin; blue stain = DAPI nuclear stain. (C) Quantitative analysis of the percentage of transformed blood vessels counted per 40× visual field as well as total blood vessels examined per condition. For all placental analysis, n = 5 animals/condition. *P < 0.05 versus ETX. Scale bars: A, 12 μm; B, 50 μm. BV = blood vessel; ns = nonsignificant; Sal = saline.

Discussion

Antenatal stress due to CA is associated with a high risk for developing BPD and PH after preterm birth; however, mechanisms that increase susceptibility for poor outcomes are incompletely understood. We found that intraamniotic ETX in the absence of postnatal hyperoxia or mechanical ventilation impairs lung growth and causes sustained abnormalities of lung alveolar and vascular structure and function with RVH in infant rats. To determine whether early augmentation of HIF signaling could improve late outcomes in infant rats, we examined the effects of two different PHi drugs, DMOG and GSK360A, on lung structure and function and found that both agents, when administered before or after birth, preserved lung growth and function and reduced RVH at 2 weeks of age. Interestingly, antenatal PHi therapy alone was effective in preventing the adverse effects of intraamniotic ETX on infant lung function and structure. We further show that antenatal PHi treatment increases the lung protein content of both HIF isomers, HIF-1a and HIF-2a, as well as proangiogenic factors, including VEGF and eNOS, which are well-known downstream targets of HIF. Finally, we report that late intrauterine exposure to ETX impairs spiral-artery growth in the placenta, which is restored with PHi therapy. Overall, these findings suggest that antenatal inflammation during late pregnancy impairs HIF signaling and that HIF augmentation before or after birth preserves lung structure and function and prevents RVH in experimental CA.

Although past studies have shown that PHi-induced enhancement of HIF signaling has protective effects on lung structure in postnatal models of BPD caused by hyperoxia and mechanical ventilation (45, 46, 53, 56), the effects of PHi therapies on lung development in the setting of antenatal stress have not been studied. This is especially important in light of strong clinical data supporting the important role of antenatal factors, such as exposure to CA, as being major determinants of BPD, PH, and late respiratory disease in preterm infants (14–25). In addition, the presence of placental abnormalities with maternal vascular underperfusion has been strongly associated with an increased risk for BPD and BPD with PH, especially in the setting of IUGR (intrauterine growth restriction), a biomarker of fetal distress (25, 72–74). Data from this study support the hypothesis that intrauterine inflammation is sufficient to alter placental vascular structure as reflected by abnormal spiral-artery remodeling after intraamniotic injection of ETX. This suggests that in addition to the direct adverse effects of inflammation on the fetal lung and risk for BPD after birth, antenatal inflammation may also increase susceptibility for lung disease because of altered placental vascular remodeling.

These placental findings are especially important to address in the context of HIF signaling, as controversy exists in the literature regarding the potential adverse effects of HIF on placental structure and function, especially in the setting of antenatal stress. Whereas HIF plays an important role in normal placental development (58–60), sustained HIF activity can potentially impair placental structure, especially during late gestation and in the setting of intrauterine inflammation (58–60). Inflammation-induced fetal growth restriction in rats is associated with increased placental HIF-1α expression (61), and persistent elevation of HIF accelerates spiral-artery remodeling in the placenta in experimental preeclampsia (73, 74). Our data show that intraamniotic PHi treatment attenuates the adverse effects of inflammation on placental spiral-artery remodeling, suggesting that HIF stabilization may actually improve placental structure and function with antenatal inflammation, which can then protect the fetus and contribute to the benefit of antenatal PHi therapy beyond its putative direct effects on the fetal lung itself.

Past experimental studies have shown that early disruption of lung VEGF signaling in utero, due to either hemodynamic stress from high intravascular pressure or pharmacologic VEGF inhibition with a selective VEGF aptamer, a VEGF receptor inhibitor, or soluble Flt-1 (fms-related receptor tyrosine kinase 1), inhibits lung vascular and alveolar growth and causes PH (29, 32, 35, 37). Inhibition of angiogenesis by either hemodynamic stress or VEGF inhibitors impairs alveolarization and increases susceptibility for PH, which has been strongly associated with downregulation of eNOS. Past work has shown an important role for eNOS in lung vascular angiogenesis during development (75–77) and that intrauterine stress decreases placental and lung eNOS expression in models of intrauterine growth restriction and in related settings (78–80). As shown in this study, antenatal PHi treatment increases both lung VEGF and eNOS expression as measured at birth, suggesting that enhanced VEGF–eNOS activity may have contributed to the preservation of lung structure and function and the prevention of RVH in this model of antenatal CA. This concept is supported by past studies of BPD models caused by hyperoxia, in which recombinant human VEGF protein or adenoviral vector–delivered VEGF gene therapies improved lung growth and PH in rats (37–40). In addition, further subgroup analysis showed no difference between male and female pups regarding the severity of injury after antenatal ETX exposure or the protective effects of PHi therapy, suggesting that disease severity and PHi efficacy are not related to sex.

We have previously reported that antenatal ETX impairs angiogenesis and lung airspace growth that persists throughout infancy (12), which reflects the dysmorphic lung microvasculature observed in human BPD. Early biomarkers, such as placental signs of underperfusion, altered circulating angiogenic factors in cord blood, and early echocardiogram changes suggestive of pulmonary vascular disease, are strongly associated with the subsequent risk for developing human BPD and BPD with PH (25, 81–84). These preclinical and clinical findings further support to the “vascular hypothesis” of BPD, in which early disruption of lung angiogenesis may play a key role in disease pathogenesis, including sustained impairment of alveolar growth and lung function (30–36). This current report provides further evidence that early disruption of HIF signaling associated with antenatal stress contributes to persistent respiratory and pulmonary vascular abnormalities throughout infancy and that early PHi treatment can play a protective role in BPD prevention as induced by antenatal risk factors.

Decreased lung HIF1a, VEGF, and eNOS has been reported in postnatal BPD models caused by hyperoxia or mechanical ventilation (14, 24–26). In a primate model of ventilator- and hyperoxia-induced BPD, PHi therapy improved lung structure and gas exchange (45, 46), enhanced VEGF expression, and increased vascular-tube formation in vitro (45). In studies of hyperoxia models, administration of deferoxamine specifically upregulated HIF-1a, whether administered intraperitoneally (56) or by aerosolization (57). Furthermore, HIF-1a gene therapy preserved lung VEGF expression, vessel density, and alveolarization after postnatal hyperoxia (46). Our findings add further evidence supporting the critically important role of HIF signaling in the perinatal lung and demonstrate that enhancement of HIF signaling is protective not only in hyperoxia-induced BPD but also in an antenatal model of BPD without postnatal injury.

Interestingly, we show that PHi treatment upregulates both HIF1a and HIF2a in the developing lung, but these data do not distinguish separate or distinct mechanistic roles of these HIF subtypes in the antenatal and early postnatal periods. Although HIF signaling presents an attractive target for future development of BPD prevention therapy, the role of HIF subtypes in PH remains controversial in both adult and neonatal lung diseases. This is especially important, as past work suggests that elevated HIF-1a expression contributes to the pathogenesis of PH in adult animal models (41–43, 65). Conversely, Kim and colleagues (85) have suggested that HIF-1a actually maintains low pulmonary vascular tone in studies of a genetic mouse model of selective HIF-1a deletion in pulmonary-artery smooth muscle cells.

Recent studies have further shown that upregulation of HIF-2a in lung vascular endothelial cells causes severe pulmonary vascular remodeling and PH in adult rodents (66, 67) and that pharmacologic inhibition of HIF-2a appears to be protective (66). In striking contrast, we report that PHi-induced upregulation of lung HIF1a and HIF2a expression prevented RVH in the developing lung, which suggests that mechanisms related to PH pathobiology in the developing lung are distinctly different from the mechanisms related to pathogenesis of PH in the adult lung. Further studies are needed to specifically define the different roles of HIF isoforms in in the lung circulation, especially regarding the specific impact of developmental timing, disease progression, and severity of disease in established PH models across the lifespan (82, 86–88). Finally, because of persistent uncertainties regarding the potential for adverse effects of HIF stabilization in pregnant adults, more preclinical work and caution is urged before translating these findings to the clinical setting.

Thus, we conclude that intrauterine inflammation impairs HIF signaling in the developing lung and that augmentation of HIF signaling with PHi therapy, either in utero or during the early postnatal period shortly after birth, preserves lung vascular and alveolar growth and prevents PH during infancy. Antenatal PHi therapy mitigates the adverse effects of intrauterine inflammation on placental structure, including spiral-artery remodeling, which suggests an additional mechanism through which antenatal intervention can enhance cardiopulmonary outcomes after exposure to intrauterine stress. Overall, we speculate that targeted enhancement of placental vascular growth and function may provide a novel strategy to improve early and late cardiorespiratory outcomes in at-risk preterm infants.