Introduction

Congenital diaphragmatic hernia (CDH) is a severe birth defect characterized by disrupted lung organogenesis that results in an underdeveloped lung (pulmonary hypoplasia) with structural alterations of the wall of pulmonary vessels, also known as vascular remodeling.1 The hypermuscularized pulmonary arterial bed, a hallmark of vascular remodeling, leads to a strong vasoconstrictive response, resulting in increased pulmonary vascular resistance.2 Physiologically, pulmonary arterial vasodilation is necessary at birth for the transition from fetal to neonatal circulation. However, the pulmonary vasoconstriction that occurs in hypoplastic lungs translates clinically into postnatal pulmonary hypertension (PHTN), one of the main determinants of morbidity and mortality in CDH patients.3 In fact, the mortality rate at discharge in children with CDH is known to be directly related to the severity of PHTN, as shown by a prospective multicenter study.3 PHTN in neonates with CDH can lead to circulatory shunting, hypoxia, hypercapnia, and cardiac dysfunction.4 PHTN secondary to CDH is often refractory to treatments, which are effective in PHTN of other etiologies, and survivors often carry a long-term burden into childhood and adulthood.5

The pathogenesis of pulmonary vascular remodeling and the link between remodeling and PHTN are still incompletely understood. To investigate PHTN pathogenesis, several experimental models have been developed, including hypoxia models, genetic models, and models based on the administration of drugs, such as monocrotaline.6 The nitrofen rat model of pulmonary hypoplasia is of particular interest to investigate vascular remodeling secondary to CDH, as it reproduces pulmonary vascular changes similar to those observed in human neonates with CDH.7 The nitrofen model relies on the administration of a herbicide (2,4-dichlorophenyl-p-nitrophenyl ether) to a pregnant rat at 9 days of gestation (E9), which causes pulmonary hypoplasia in 100% of the offspring and a diaphragmatic defect in 50–60% of the litter.8 In the nitrofen model, fetuses with CDH have decreased distal vessel density, increased muscularization of small arteries, resulting in increased pulmonary vascular wall thickness.7,9 In fact, both medial and adventitial layers of the vascular walls are thickened in the experimental and human CDH.10

Over the years, the nitrofen model has been extensively employed, but to the best of our knowledge, no study has synthesized the accumulated data on the pathogenesis of PHTN in this experimental model of CDH. The aim of the present study was to comprehensively analyze all factors and pathways involved in vascular remodeling secondary to CDH in the nitrofen rat model and to understand their relevance to the human condition. This study could help researchers better understand PHTN pathophysiology, which is essential to design novel preventative and treatment strategies for these babies.

Materials and methods

A systematic review of the literature was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement.11 Eligible studies were identified by searching scientific databases (PubMed, Medline, Cochrane Collaboration, Embase, and Web of Science) involving studies published in English from the first available date in each database to August 2018. The search strategy combined the keywords: “congenital diaphragmatic hernia” and “nitrofen”. Reference lists were searched to identify relevant cross-references. Case reports, reviews, and opinion articles were excluded from the review. All gray literature publications (i.e., reports, theses, conference proceedings, bibliographies, commercial documentations, and official documents not published commercially) were excluded. Publications using other models of CDH or the murine nitrofen model were excluded. The full text of potentially eligible studies was retrieved and assessed for eligibility. Inclusion criteria were experimental studies reporting at least one dysregulated factor with a potential role in vascular remodeling.

Dysregulation of the factor was defined as a significant decrease or increase of mRNA and/or protein and/or enzyme activity quantification in the lungs of nitrofen-exposed rats with CDH compared with control rats (statistical significance defined as p < 0.05). Factors were classified according to whether they were involved in the dysfunction of endothelial cells (ECs) or pulmonary artery smooth muscle cells (PASMCs), or in the crosstalk between these two cell populations. To investigate the potential translational effects of the dysregulated factors, we then searched for reported similarities in human CDH.

Results

Study selection and characteristics

Within the 358 articles screened, 75 reported dysregulated factors with a potential role in pulmonary vascular remodeling and were included in the review (Fig. 1). The articles were published between 1995 and 2018, with an interesting increment in the last decade (Fig. 2).

Fig. 1
figure 1

PRISMA flowchart of search results

Full size image
Fig. 2
figure 2

Date of publication of the articles included in the review

Full size image

Factors involved in EC dysfunction

During fetal development, the pulmonary vasculature develops in close synergy with the airways in a process regulated by interactions between the developing epithelium and endothelium (Fig. 3 and Table 1). In pulmonary hypoplasia secondary to CDH, the impaired homeostasis of the respiratory epithelium results in decreased signaling to ECs.9,12

Fig. 3
figure 3

Factors involved in endothelial cell dysfunction in the nitrofen model of congenital diaphragmatic hernia

Full size image
Table 1 Factors involved in endothelial cell (EC) dysfunction
Full size table

Vascular endothelial growth factor (VEGF) is a signal protein released by the epithelial cells to regulate differentiation of mesenchymal cells into ECs, EC proliferation, and vascular development.13 ECs express two receptors for VEGF: VEGFR-2 that is essential for angiogenesis in early stages and has a proliferative role in ECs, and VEGFR-1 that is a negative regulator of EC division in later embryonic stages, and allows for EC maturation, tube formation, and integrity of the vessel wall. VEGF can be upregulated by lung expansion and hypoxia, via hypoxia-inducible factor-1 (HIF-1).13,14 VEGF and both receptors have been frequently reported as downregulated in the lungs of nitrofen-exposed CDH fetuses throughout development, from the pseudoglandular stage to the saccular stages.15,16,17,18,19,20,21,22,23,24 VEGF production by fetal lung epithelial cells can be upregulated by fibroblast growth factor (FGF)-10/FGFR-2 signaling, which is downstream of canonical Wnt signaling.25 In nitrofen-exposed lungs of fetuses with CDH, mesenchymal FGF-10 and canonical Wnt signaling are decreased, resulting in a decrease of VEGF production by epithelial cells.26,27 Ventilation is another regulator of VEGF signaling. An experimental study where nitrofen pups were ventilated after birth showed that ventilation increased VEGFR-1 and decreased VEGFR-2 expression in lungs of pups with CDH.28

In human CDH pregnancies, VEGF levels have been reported to be decreased at amniocentesis, as well as in lungs of CDH fetuses.29,30 Although pulmonary expression of VEGF is similar between newborns with or without CDH at birth, increased levels of pulmonary VEGF have been reported in newborns with CDH and confirmed PHTN compared with controls.31,32,33 Similarly, although plasma levels of VEGF at birth are similar between newborns with or without CDH, increased plasma levels of VEGF in CDH newborns on the third day of life have been reported to be predictive of severe PHTN and mortality.34,35 Endothelial colony-forming cells derived from cord blood of CDH newborns have blunted responses to VEGF and high levels of nitric oxide production, as well as reduced potential for proliferation and migration, which suggests a reduced number of both VEGF receptors in endothelial cells of CDH newborns.36

Krüppel-like factor 2 (KLF-2), a central regulator of endothelial function, is known to mediate the VEGF-induced EC maturation.37 In lungs of nitrofen-exposed fetuses with CDH, KLF-2 is downregulated.38 As a result, the transmembrane tyrosine kinase receptor c-Kit, and its ligand stem cell factor (SCF) have an increased expression from E15 to E21 in ECs of nitrofen-exposed rat lungs compared with controls.39 c-Kit+ and SCF+ cells in developing lungs are markers of endothelial cell progenitors, and the increase in their expression suggests a delay in differentiation of the ECs, which remain immature and dysfunctional.40 KLF-2 is also known for regulating the crosstalk with PASMCs by upregulating endothelial nitric oxide synthase (eNOS) and downregulating endothelin-1 (ET-1) and angiotensin-converting enzyme (ACE).41

Forkhead box F1 (FoxF1) is a transcription factor essential for pulmonary angiogenesis, required for VEGF signaling in ECs, and is reported as downregulated in lungs of nitrofen-exposed CDH pups.42,43 A decrease in FoxF1 results in decreased expression of EC genes essential for vascular development, such as VEGF receptors and Ephrin B2.44 FoxF1 acts downstream of epithelial sonic hedgehog signaling, which is downregulated during late gestation in nitrofen-exposed lungs of rats with CDH, as is kinesin family member 7 (Kif-7), an essential component of sonic hedgehog signaling to ECs.45,46,47

Bone morphogenetic protein (BMP) signaling is crucial for lung angiogenesis, and loss of the receptor BMPR-II in ECs has been shown to lead to PHTN.48 During lung development, BMPR-II is mainly expressed in ECs, and its activation by BMP-2 and BMP-4 results in EC proliferation and migration through phosphorylation of Smad 1 and 5.49 In lungs of nitrofen-exposed rats with CDH, BMP4/BMPR-II signaling is downregulated from E17 to E21, as well as several downstream targets of BMP signaling in ECs, such as phosphorylated Smad 1/5/8 and Apelin (APLN).27,50,51,52,53,54 APLN is essential for EC homeostasis and attenuates the response of SMCs to growth factors, through its receptors that are present on ECs and PASMCs. APLN and APLN receptors are both downregulated in nitrofen CDH lungs at E21 (Figs. 3 and 5).54

The loss of BMPR-II signaling results in the production of reactive oxygen species (ROS) in ECs through the activation of a RhoA/ROCK1 pathway, which causes EC injury and enhances ROS production by inhibiting eNOS. Moreover, ROS production can result from the decrease of peroxisome proliferator-activated receptor-γ (PPAR-γ), which protects against oxidative stress and inflammation by inhibiting monocyte chemoattractant protein-1 (MCP-1) and NADPH oxidase (Nox)-4.55,56 Nox has been reported to be a major source of hydrogen superoxide in the vasculature, contributing to EC dysfunction and PASMC proliferation. Hydrogen superoxide reacts with nitric oxide (NO) to form the ROS peroxynitrite. The activation of oxidative stress through these pathways is present in lungs of nitrofen-exposed rats with CDH, as shown by the increase in RhoA and its activator Wnt11 in the endothelial layer and the decrease in PPAR-γ associated with an increase in Nox-4 and MCP-1 in the vascular wall.57,58,59,60,61 Nox-4 can be activated by PDGF-A which is increased in the lungs of nitrofen-exposed CDH fetuses.62 As a result, increased levels of hydrogen superoxide have been reported in the lungs of nitrofen-exposed CDH fetuses, as well as in the increased number of oxidative-damaged proteins.62,63 Moreover, expression and activities of antioxidant enzymes, such as superoxide dismutase (SOD) and catalase, are decreased in lungs of nitrofen-exposed CDH fetuses, suggesting that the negative feedback regulatory loop is affected.60,64 These changes are observed also in PASMCs (Fig. 5), where a cyclic stretch due to excessive vasoconstriction further increases Nox4 and the production of ROS in fetal pulmonary arteries.65

Endoglin (ENG) is a transmembrane accessory receptor for transforming growth factor (TGF)-β signaling in ECs, and it is crucial for the activation of EC proliferation and migration through activin receptor-like kinase-1 (ALK1)/Smad1/5.66 This accessory receptor also interacts with VEGFR2 to promote VEGF-A-induced angiogenesis and VEGF signaling.67 ENG has been reported as downregulated in the lungs of CDH pups.68

The C-reactive protein (CRP) is a well-known marker of inflammation and a risk factor for endothelial dysfunction and, when present in the vessel wall, induces the expression of the vascular adhesion molecule-1 (VCAM-1) and intercellular adhesion molecule-1 (ICAM-1) by ECs and serves as a chemoattractant for monocytes.9 CRP and both adhesion molecules were reported as significantly increased in pulmonary arteries of newborn pups with CDH.9,69 Another trigger for expression of these adhesion molecules by ECs is tumor necrosis factor-α (TNFα), which was increased in the lungs of nitrofen-exposed pups with CDH.70 In human CDH, inflammation seems to play a role in endothelial dysfunction, with increased levels of adhesion markers in pulmonary artery ECs and in the plasma of CDH newborns and stillborns, and increased TNF-α expression in pulmonary epithelial cells.71,72

Disrupted crosstalk between ECs and PASMCs

As ECs modulate PASMC proliferation, apoptosis, and contraction, EC dysfunction results in a disrupted secretion of paracrine factors and PASMC dysfunction (Fig. 4 and Table 2).

Fig. 4
figure 4

Factors involved in the disrupted crosstalk between endothelial cells and pulmonary artery smooth muscle cells in the nitrofen model of congenital diaphragmatic hernia

Full size image
Table 2 Factors involved in disrupted crosstalk between endothelial cells (ECs) and pulmonary artery smooth muscle cells (PASMCs)
Full size table

Endothelin-1 (ET-1) is produced by pulmonary artery ECs and is known to be proliferative and vasoactive in PASMCs.73 ET-1 acts through two different receptors: ET-a that is present on PASMCs and mediates vasoconstriction by stimulating cytosolic calcium release, and ET-b that is present on PASMCs and ECs, and mediates vasodilation by stimulating the release of prostacyclin and NO from ECs. Both receptors mediate PASMC proliferation. After stimulation by angiotensin II or ROS, ET-1 is converted from its precursor into its active form by endothelin-converting enzymes (ECE). Lungs of nitrofen-exposed rats with CDH had increased levels of ET-1, ECE, and both receptors as early as E15.9,51,73,74,75,76,77,78 The increase in receptors was confirmed in a study showing that arterioles from nitrofen-exposed CDH lungs constricted more than those in controls in response to ET-1.79 Interestingly, administering an inhibitor of ECE prenatally to dams exposed to nitrofen was reported to increase the survival rate of pups with CDH.80

Human babies with CDH have been reported to have increased plasma levels of ET-1 and pulmonary levels of both ET-1 receptors.77,81,82,83,84 Moreover, lungs from babies who died with CDH have been reported to have ET-a receptors in very small capillaries (<25 μm), whereas these receptors are limited to small and large capillaries in control lungs.77 Finally, babies with CDH had increased levels of ECE during fetal life, whereas in the lungs of control fetuses, ECE expression is low during fetal life and increases just before birth.77

Angiotensin-converting enzyme (ACE) is produced by the pulmonary vascular endothelium and converts angiotensin I into angiotensin II, a potent vasoconstrictor. Along with angiotensin II receptors, AT-R1 and AT-R2, ACE is expressed in the developing lung as early as the pseudoglandular stage.85 In lungs of nitrofen-exposed rats with CDH, ACE had an increased activity, whereas the expression of angiotensin receptors was decreased, possibly due to a negative regulatory feedback loop.21,86,87 Angiotensin II enhances PASMC migration and proliferation through AT-R1.88,89 Angiotensin II also has autocrine effects on ECs, mediating endothelial dysfunction by increasing the production of ROS, which in turn induce endothelial production of ET-1, ICAM-1, and VCAM-1.90

Nitric oxide (NO), produced by nitric oxide synthase (NOS) in pulmonary ECs, is a potent vasodilator, whose production is stimulated by oxygenation, ventilation, and shear stress at birth. ECs produce the constitutive endothelial isoform, eNOS, and an inducible isoform, iNOS. eNOS expression is increased in normal lung development in late gestation and maximal levels are reported in the near term. Caveolin-1 is a scaffolding protein of caveolae, which are invaginations of the cell surface plasma membrane in ECs and PASMCs, allowing and regulating crosstalk between these cell populations. Caveolin-1 stabilizes eNOS, and regulates its activity, resulting in decreased NO synthesis. Loss of caveolin-1 has been shown to result in endothelial dysfunction.91 In the nitrofen CDH rat model, studies on NOS have yielded conflicting results. NO production and eNOS expression have been mainly reported as decreased during the saccular stage in nitrofen-exposed fetuses with CDH.9,19,21,51,92,93,94 eNOS expression has also been reported as increased at E21 in CDH lungs, with a concomitant decrease of caveolin-1.77,91,95 A possible explanation for the increased level of eNOS at E21 in these three studies could be found in the different fetal harvesting techniques. In these studies, pups were delivered by C-section and then euthanized,77,91,95 whereas in the other studies, the dam was euthanized and the pups were harvested after the dam’s death, preventing any breathing for the pups.9,19,21,51,92,93,94 Indeed, ventilation has been shown to increase eNOS expression and decrease iNOS in pups with CDH.28,96,97

High levels of endothelial NO have anti-proliferative effects on PASMCs and cause vasodilation through the activation of soluble guanylate cyclase (sGC) that generates cyclic guanosine monophosphate (cGMP).98 The level of cGMP is controlled by a family of catabolic enzymes, cyclic nucleotide phosphodiesterases (PDEs). Several PDEs are found expressed in fetal rat lungs, but only PDE2 and PDE5 levels were increased in lungs of nitrofen-exposed rats with CDH, resulting in decreased levels of cGMP.19,95,99 Pulmonary arteries from lungs of nitrofen-exposed CDH pups have blunted responses to NO and GMP activators, but these are restored back to normal by the administration of a PDE5 inhibitor.100 This suggests the absence of a physiological downregulation of PDE5, which is seen postnatally in lungs of control rats.100 In lungs of nitrofen-exposed CDH pups, an increased expression of the protein kinase Prkg2, downstream of PDE5, confirms the upregulation of PDE5 expression, resulting in increased elimination of cGMP by PDE5, which participates in the lack of vasodilation seen in these lungs.95

In human CDH, eNOS expression has been reported as decreased at birth and earlier in CDH fetuses,101,102 whereas increased levels of eNOS have been reported in neonates with CDH after a few days of life or when treated with extracorporeal membrane oxygenation (ECMO).103,104,105

Serotonin (or 5-hydroxytryptamine, 5-HT) can be produced by pulmonary vascular ECs, neuroendocrine cells, and neuroepithelial bodies in response to hypoxia. Released 5-HT can bind to serotonin receptors and transporters on PASMCs, such as 5-HTT and 5-HT2a, inducing PASMC proliferation and contraction (Fig. 5). In lungs of nitrofen-exposed rats with CDH, the expression of 5-HTT and 5-HT2a is increased compared with that of controls during the canalicular stage of lung development.106

Fig. 5
figure 5

Factors involved in pulmonary artery smooth muscle cell dysfunction in the nitrofen model of congenital diaphragmatic hernia

Full size image

Two studies have investigated the role of the prostaglandin pathway in experimental and human hypoplastic lungs secondary to CDH. A decrease in prostaglandin synthase and thromboxane synthase was reported in experimental CDH, and a decrease in the prostaglandin receptor has been found in human CDH newborns.77,95

Factors involved in PASMC dysfunction

Several studies have reported an increase in proliferation of PASMCs in the small pulmonary vessels of CDH pups at E21 through several mechanisms (Fig. 5 and Table 3).95,107,108 One of these mechanisms involves the activation of signal transducer and activator of transcription 3 (STAT-3) signaling.109 STAT-3 is increased in the vasculature of lungs of nitrofen-exposed fetuses with CDH from E17.5 to E21, and results in the upregulation of Pim-1, KLF-5, and Survivin, transcription factors involved in PASMC proliferation and resistance to apoptosis.110,111,112,113 Pim-1 is a target of the micro-RNA miR-33, which is downregulated in fetal lungs of nitrofen-induced CDH rats.114 STAT-3 is activated by phosphorylation in response to vasoconstrictive agents (such as endothelin-1 and angiotensin-II), growth factors, and cytokines.115 In PASMCs of nitrofen-exposed lungs of rats with CDH, the increased expression of the receptor for advanced glycation end products (RAGE) strongly activates STAT-3 but also induces BMPR-II and PPARγ downregulation.50,51,52,58,111,116,117 Loss of BMPR-II in PASMCs has been shown to result in phosphorylation of the anti-apoptotic p38 MAPK, which is increased in lungs of nitrofen-exposed pups with CDH, resulting in a reduced PASMC apoptosis rate.50,118

Table 3 Factors involved in pulmonary artery smooth muscle cell (PASMC) dysfunction
Full size table

STAT3 also upregulates NFATc2 (nuclear factor of activated T-cell, cytoplasmic, calcineurine-dependent-2), a transcription factor that is upregulated in lungs of nitrofen-exposed pups with CDH and that plays an important role in modulating vascular tone response.110,111,119 NFATc2 actively suppresses the expression of voltage-gated potassium channels, which are important regulators of calcium channels.119 In nitrofen-exposed lungs of rats with CDH, the inhibition of potassium channels in the vascular smooth muscle layer depolarizes PASMCs to a threshold that opens voltage-gated calcium channels, increasing cytosolic-free calcium concentration and resulting in PASMC vasoconstriction and proliferation.119,120,121 In parallel, an increase in canonical transient receptor channel 6 (TRPC6) and calcium-sensing receptor (CaSR), which interact together to increase the cytosolic concentration in calcium, has been shown.122 Finally, impaired vascular tone in the nitrofen model has been confirmed, with an overconstriction of pulmonary arterioles in response to endothelin-1, a blunted vasodilation in response to oxygen, and a blunted vasoconstrictive response to hypoxia.79,123,124

The sphingosine-1-phosphate receptor 1 (S1P1) activates STAT-3 and increases the expression of Ras-related C3 botulinum toxin substrate 1 (Rac1), which is an important mediator of pulmonary vascular remodeling by promoting PASMC proliferation. In normal conditions, PASMC proliferation is negatively regulated by a decrease in expression of S1P1 and an increase in S1P2 and S1P3.125 In lungs of nitrofen-exposed pups with CDH, there is an increase in S1P1 and a decrease in S1P2 and S1P3 compared with those of control lungs.126 This is confirmed by the upregulation of downstream targets of S1P1, Rac1, and STAT3, participating in the excessive proliferation of PASMCs.126

In the developing lung, it has recently been shown that TGF-β can stimulate PASMCs through ALK-1/Smad 1,5 signaling.104 In nitrofen CDH lungs, expression of ALK-1 is upregulated at E21 in the pulmonary vasculature.127 The expression of TGF-β1 has been reported as increased in PASMCs of newborn pups with CDH compared with those of controls.128 Elastin microfibril interface-located protein 1 (Emilin-1), expressed by PASMCs, is essential for elastogenesis and inhibits TGF-β signaling by binding to its precursor.129 In nitrofen CDH lungs, Emilin-1 is downregulated.130 Taken together, these results could suggest an increased autocrine TGF-β signaling in PASMCs, resulting in PASMC proliferation through ALK-1 and increased deposition of collagen.131

In human CDH pregnancies, levels of TGF-β are decreased at amniocentesis, whereas TGF-β3 and its regulator microRNA 200b have been reported as increased in pulmonary arteries and lungs of CDH newborn lungs.132,133 An increased TGF-β signaling to PASMCs results in an altered extracellular matrix gene expression, such as collagen, elastase, and osteopontin, which can lead to thickening of the medial and adventitial layers of pulmonary arteries. Osteopontin is highly expressed by PASMCs from patients with idiopathic pulmonary hypertension, and mediates signals of proliferation to other PASMCs.134 Pulmonary arteries from rat pups with CDH present an extracellular matrix with an increased proteolytic activity (serine elastase and matrix metalloproteinase activity) and enriched in growth factors, such as epidermal growth factor and osteopontin, enhancing PASMC proliferation and muscularization of pulmonary arteries.75,107,135 In parallel, another growth factor mitogenic for PASMCs and fibroblasts, PDGF-B is increased in the medial layer of pulmonary arteries of lungs of nitrofen-exposed pups with CDH.136

Embryonic essential myosin light-chain (MLC)-1a and regulatory MLC (MLC-2), which are normally expressed in PASMCs and parabronchial smooth muscle cells, are absent in lungs of nitrofen-exposed pups during the pseudoglandular stage, while their expression is recovered during late stages.137 The absence of these proteins suggests an early impairment in PASMCs during the pseudoglandular stage, before the establishment of vascular remodeling (late canicular and saccular stages).

Discussion

This systematic review confirms that in experimental CDH, the impaired homeostasis of the respiratory epithelium, which is a hallmark of pulmonary hypoplasia, leads to pulmonary vascular remodeling (Fig. 6). The main signaling pathways disrupted in the epithelium, such as BMP, Shh, Wnt, and VEGF, are at the origin of the pulmonary endothelial dysfunction, thus confirming the close relationship between epithelial and vascular development during fetal lung morphogenesis. The resulting endothelial dysfunction is characterized by the immaturity of the EC, the decreased expression of transcription factors essential for EC function in angiogenesis (e.g., KLF-2, Apelin, and FoxF1), and the imbalanced signaling to PASMCs. This impaired crosstalk between ECs and PASMCs leads to an increase in PASMC proliferation rate, resistance to apoptosis, and constriction, leading to increased medial wall thickness. Moreover, the abnormal secretion of factors, such as collagen, by PASMCs into the extracellular matrix allows PASMCs to create their own favorable extracellular environment to maintain further proliferation along with autocrine stimulation and increased adventitial wall thickness. Eventually, all these modifications to normal lung growth lead to the postnatal development of PHTN.

Fig. 6
figure 6

Pathogenesis of pulmonary hypertension secondary to congenital diaphragmatic hernia

Full size image

Interestingly, several studies have shown that the pathways leading to vascular remodeling were reported to be predominantly disrupted in nitrofen-exposed pups that developed CDH.7,17,22,138 In these studies, in fact, control pups and pups exposed to nitrofen that did not have CDH had normal levels of factors, such as VEGF and eNOS. Nonetheless, we acknowledge that most studies included in this systematic review focused only on nitrofen-exposed pups, which developed CDH, thus making it challenging to draw conclusions on this observation.

Over the years, many studies have reported a number of factors and signaling pathways that are affected in CDH and result in vascular remodeling. The present systematic review of the literature shows for the first time in a comprehensive manner how these factors and pathways are affected. For this reason, the nitrofen-induced model of CDH has great value not just for the study of pulmonary hypoplasia, but also for the understanding of PHTN pathogenesis. In the literature, there are mainly two other models of CDH, which are both based on the surgical creation of the diaphragmatic defect in the rabbit and in the lamb. Although these two models have been used for preclinical studies, there are only a few articles that have reported disrupted factors and pathways involved in vascular remodeling in these experimental animals. In fact, in the rabbit model, only two studies have been published and have reported a decrease in eNOS in CDH lungs compared with controls, but no disruption in VEGF or ET pathways.139,140 In the sheep model of CDH, there are a few more studies, indicating similar findings to those observed in the nitrofen model: an altered interaction between dysfunctional EC and PASMCs via increased expression of ET-1 and decreased eNOS.141,142,143 However, contrarily to what has been reported in the nitrofen model, ECs of CDH lamb lungs had a higher expression of VEGF and its receptors compared with those of controls.141 These differences with the nitrofen model can be explained by the fact that the diaphragmatic defect is created late in gestation. As these studies on surgical models were not sufficient to investigate the pathogenesis of vascular remodeling in CDH, we focused on the nitrofen model of CDH in this review.

The literature reported in the nitrofen model of CDH seems to be relevant for potential translation into clinical practice, as the observations made mirror the pathway disruption reported in human fetuses with CDH. As these changes occur in utero, fetuses with CDH could benefit from an antenatal treatment that induces epithelial maturation prenatally, to prevent the establishment or to reverse the development of vascular remodeling. Although this seems to be a promising strategy, one needs to remember that, at present, translating novel fetal therapies into clinical application remains difficult. In fact, the fetus is a particularly challenging patient that responds to interventions in a distinctive way. Challenges include addressing several dysfunctional pathways at the same time with one single treatment, risk of side effects due to unintended targeting of other organs, treatment delivery route, and differences in lung development between experimental models and human fetuses. For the latter, lung development stages in rat and human fetuses occur at different time points during gestation. The present review has shown that the changes observed in vascular remodeling occur at the same stage of development in both rat and human fetuses, and that is the saccular stage. This stage has a different length as it occurs at the end of pregnancy between E20 and term in rats, and from 24 weeks of gestation to birth in human fetuses. Despite these differences, the saccular stage seems to be the ideal time frame for a potential translation of antenatal therapies in human fetuses.

We acknowledge the limitations of this systematic review, whose quality is dependent on the quality of the papers published. For instance, within the same model, we have observed discordant findings on some factors or pathways at the same time point. The possible explanation for this could be due to study design, variations in assays employed, and influence of the size and/or side of the diaphragmatic defect that may affect the severity of vascular remodeling. Furthermore, the studies included in this review were focused only on factors that were already known to be dysfunctional in other models of PHTN associated to other diseases. Unbiased investigations were limited to the whole-lung RNA sequencing, and thus were not informative on pathways that might be disrupted at the cellular level at different developmental time points. To minimize the differences and improve the quality of the results, the authors of future studies designed to address the changes in vascular remodeling in experimental CDH should systematically verify and report the presence and laterality of the diaphragmatic defect.

In conclusion, this systematic review has shown that, over the years, there has been an increasing interest in vascular remodeling secondary to CDH, with exponentially more research articles dedicated to this important aspect of the disease. However, a full understanding of the pathogenesis of vascular remodeling in CDH fetuses has yet to be achieved. For this reason, there is sparse literature on novel potential strategies to prevent or treat vascular remodeling in utero. Further studies are required to better explore the pathophysiology and possible translational strategies for PHTN, which still remains a critical determinant of morbidity and mortality in CDH infants.