Vitamin A (i.e., retinol) and its derivates (i.e., retinoids) are indispensable for various aspects of early embryogenesis. Over the years, several knockout models have indicated a role of the retinoid signaling pathway and its downstream targets in the pathogenesis of CDH (38). For instance, mice lacking both subtypes of retinoic acid receptors α and β (Rarα and Rarβ) have been demonstrated to generate offspring with posterolateral diaphragmatic defects identical to those observed in human patients (39–44), and similar to the vitamin A-deficient CDH mouse model as previously reported by Anderson (45, 46). Surprisingly, single Rar null mutant mice did not display any of the predicted malformations that were seen in rats with vitamin A deficiency (31). Nevertheless, when the activity of several receptors was inhibited, various deformities were noted including right-sided CDH in Rarα/β2 and left-sided CDH in Rarα/β2^+/– animals. Moreover, these mice exhibited severe pulmonary hypoplasia (31). Despite the convincing data, these genetically modified mouse models manifest only a comparatively low incidence of diaphragmatic defects and a high rate of additional comorbidities (e.g., cranial, cardiac, vertebral and limb), which do not accurately depict the human situation (39, 42). Still, mutations in the stimulated by retinoic acid gene 6 (STRA6), a membrane receptor that controls the cellular uptake of vitamin A and cellular retinoic acid binding protein 1 (CRABP1), which is located on chromosome 15, have been found to lead to a spectrum of developmental anomalies including CDH and hyperplastic lungs (47, 48).

Another important gene that is linked with the retinoid signaling pathway is chick ovalbumin upstream promoter transcription factor II (COUP-TFII), a transcription factor that is affiliated with the nuclear steroid/thyroid hormone receptor superfamily, whose DNA-binding site has been shown to reduce the induction of retinoic acid receptors (49–51). COUP-TFII was recently renamed as nuclear receptor subfamily 2 group F (NR2F2), which is expressed in the diaphragm and lungs during early gestation (34). Mapped to chromosome 15q26 in humans, the NR2F2 gene is situated on a recognized CDH hotspot region, thus making it a likely contributor to the etiology of diaphragmatic defects. On the basis of this observation, You et al. (52) have created a tissue-specific Nr2f2^–/– mouse model that features left-sided Bochdalek-type CDH and pulmonary hypoplasia similar to the human situation. Through targeted ablation of Nr2f2 in the foregut mesenchyme and pleuroperitoneal folds (PPFs), posterolateral diaphragmatic defects presumably arise because of the failure of the posthepatic mesenchymal plate to merge with the lateral body wall, thus enabling stomach and liver to protrude into the chest (52).

The creation of genetic mouse models for various other applications has revealed several genes, which one would not necessarily immediately associate with CDH. Initially introduced as a model for the investigation of early urogenital organogenesis (53), Wilm’s tumor 1 (Wt1) null mutant mice die during mid-gestation, displaying posterolateral diaphragmatic defects and lung hypoplasia alongside urogenital abnormalities (54–56). Heterozygous mutations of the WT1 gene, which encodes a transcription factor that contains four zinc finger motifs, is known to produce distinct syndromes with clinical overlap that include CDH (e.g., Denys-Drash syndrome or Meacham syndrome) (57, 58). Wt1^–/–, vitamin A-deficient and nitrofen mouse models of CDH each implied a mutual pathomechanism for the formation of diaphragmatic defects with several analogies to the condition in humans (59). More recently, Carmona et al. (60) and Cleal et al. (61) have reported that conditional deletion of Wt1 in the mesenchyme of the septum transversum can cause CDH in mice. Today, it is proven that Wt1 and Couptf2 both interact with the retinoid signaling pathway during embryonic development (3). Surprisingly, Wt1 and Couptf2 are not found in the muscle precursors but in the non-muscular mesenchymal compartment of the PPFs (3). Paris et al. (62) have developed a novel genetically modified mouse model of CDH, demonstrating that Wt1-induced β-catenin loss-of-function produces posterior diaphragmatic defects, bilateral pulmonary hypoplasia and liver herniation, comparable to the phenotypes associated with CDH in human patients. Additionally, a decreased mesothelial proliferation and increased rate of cell death was identified in the posterior diaphragm mesenchyme, and all mouse pups died postnatally with malformed lung lobes and collapsed distal air spaces (62). Loss of Wt1 has also been associated with lung branching defects before diaphragm closure in another genetic model of CDH (63).

GLI-Kruppel family member 2 (Gli2) and Gli3 and are both members of a highly conserved morphogenetic family, belonging to the sonic hedgehog (Shh) signaling pathway (64). This pathway is thought to be crucial during normal diaphragmatic development (36). A murine model of the VACTERL-like syndrome (i.e., vertebral, anorectal, cardiac, tracheoesophageal, renal and limb anomalies) created by Kim et al. (65) involved Gli2^–/–;Gli3^–/– and Gli2^–/–;Gli3^+/– mice that developed left-sided posterior CDH and pulmonary hypoplasia besides the observed VACTERL components. This was the first experimental model that reproduced the human VACTERL association, indicating that disruptions in Shh signaling might contribute to the pathogenesis of VACTERL syndrome. Likewise, as Gli2, Gli3 and Wt1 all encode important zinc finger proteins, further transcription factors of this type have been hypothesized through the generation of newer genetic animal models of CDH. For example, kinesin family member 7 (Kif7) and pre-B-cell leukemia transcription factor 1 (Pbx1) were recently recognized as indispensable components of the Shh signaling pathway, functioning as regulators during early embryogenesis (66, 67). Kif7 encodes a motor protein that functions downstream of the transmembrane receptor smoothened, and interacts with both Gli2 and Gli3 (68). Furthermore, Kif7 was found to coordinate cell proliferation, central tendon patterning and differentiation of the primordial diaphragm in a genetically modified mouse model of CDH (69). Homozygous Kif7^dd/dd mutant mice and Pbx1^–/– knockout mice both display left-sided posterior diaphragmatic defects and hypoplastic lungs (36, 69–71). In turn, haploinsufficieny of PBX1 has been associated with various congenital anomalies including CDH (72). Moreover, two predicted variants in the KIF7 gene were recently detected in patients with CDH (73). Additionally, mice lacking chromatin target of protein arginine methyltransferase 1 (Chtop) have numerous developmental abnormalities including posterolateral defects in the diaphragm, pulmonary hypoplasia and liver herniation (74–76). High-resolution 3D imaging further characterized these diaphragmatic defects in Chtop^–/– mice embryos (77).

Zinc finger protein 2 (ZFPM2), formerly known as friend of GATA-binding protein 2 (FOG2), encodes another zinc finger-containing protein that regulates the transcriptional activity of GATA4, hereby controlling a number of developmental mechanisms in the forming diaphragm and lung (78–81). In mice, Fog2 was initially found to be expressed in the embryonic septum transversum of the diaphragm (81). In humans, ZFPM2 is located on chromosome 8p23 and has been demonstrated to interact with COUP-TFII (82, 83). However, only a single mutation in the ZFPM2 gene has been identified in isolated patients with non-syndromic CDH to date (31). In a cohort of 275 patients with CDH, Longoni et al. (84) have recently reported the incidence of ZFPM2 mutations to be nearly 5%. In addition, their genetic analysis of a multigenerational family revealed a heritable intragenic ZFPM2 deletion with an approximated penetrance for clinical relevant diaphragmatic defects of around 37.5% (84). On the other side, mice exposed to the chemical mutagen N-ethyl-N-nitrosourea (ENU) generated Fog2^–/– offspring with bilateral hypoplastic lungs and a defective posterolateral diaphragm characteristic of CDH (80), while 70% of mice heterozygous for a Gata4 deletion mutation (i.e., Gata4^+/Δex2) displayed retrosternal diaphragmatic defects, dilated distal airways and thickened pulmonary mesenchyme (85). Using genetically modified mice, Merrell et al. (86) have shown that Gata4 mosaic mutations in PPF-derived muscle connective tissue fibroblasts led to the development of localized amuscular regions of the diaphragm, which were biomechanically weaker and subsequently caused CDH. GATA4 and ZFPM2 genes have been both found to be absent in humans with CDH (78), emphasizing their roles as possible candidate genes for CDH. Moreover, Zfpm2 is known to interact with Nr2f2, indicating that these two transcription factors together with Gata4 may contribute to diaphragm formation (83). Recurrent microdeletions of 8p23.1, including GATA4 and the sex determining region Y-box 7 (SOX7) gene are accompanied with a significant risk of CDH and cardiovascular anomalies (87). Even though mice lacking the Gata4 gene display both diaphragmatic and cardiac defects, no human patient with cardiac anomalies and GATA4 mutations have been identified with CDH so far (87). However, Wat et al. (87) have recently demonstrated that haploinsufficiency of Sox7 or Gata4 is enough to cause retrosternal diaphragmatic defects in mice and that haploinsufficiency of SOX7 and GATA4 may in turn be involved in the pathogenesis of CDH in patients with 8p23.1 deletions.

Several basic helix-loop-helix transcription factors have been shown to support the development of the primordial diaphragm in mice. Capsulin (Cap) is one of those, which is strongly expressed in the fetal diaphragm and in mesenchymal cells of the lung (88). As one might expect, Cap^+/– mice have severe defects in lung morphogenesis and lack alveoli (89). On the other hand, mice homozygous deficient for both cap^–/– and the related myogenic bHLH transcription factor R (MyoR^–/–) lack facial musculature and exhibit posterior diaphragmatic defects. This double mutant mouse model, which was generated at first to study the formation of facial muscles, displayed not only CDH and defective lung branching morphogenesis but also severe facial muscle abnormalities (90). Although these genetically modified mice died soon after birth because of pulmonary and cardiac malformations, the type of diaphragmatic defect seen in this model indicates that both Cap and MyoR are necessary for the integrity of the developing diaphragm. Previously, these genes were referred to as transcription factor 21 (Tcf21) and musculin (Msc), respectively (90, 91).

Eyes absent (Eya) genes and the transcription factor sine oculis homebox 1 (Six1) form an important signaling network, which plays a central role during embryonic development (92, 93). Eya1 and Six1 together constitute an evolutionary conserved transcriptional complex that coordinates multiple integrated processes needed for normal growth of the primordial diaphragm and lung (94). Further research work has confirmed that the Eya1-Six1 pathway has a key role in lung maturation by regulating its branching morphogenesis (95). Mice deficient in Eya1^–/– and Eya2^–/+ have no diaphragm (94), whereas single mutant Eya1 mice die shortly after birth due to respiratory failure, having severely hypoplastic lungs with reduced epithelial branching and increased mesenchymal cellularity (95). In Six1^–/– mice, the diaphragm is also absent, and the Six1 deletion causes pulmonary hypoplasia with greatly reduced epithelial branching, narrow bronchi, dense mesenchyme and obvious failure of normal lung maturation (96, 97). These findings indicate that disruption of the Eya1-Six1 signaling pathway may lead to neonatal lethality as a consequence of an absent diaphragm and hypoplastic lungs.

H2.0-like homeobox (Hlx) is a protein coding gene that is relatively conserved across various species (98). This homeobox transcription factor has been found to be highly expressed during early organogenesis in the septum transversum of the diaphragm and lung mesenchyme (99, 100). Hlx^–/– mice suffered early demise and showed diaphragmatic defects (101). Additionally, Farrell et al. (102) reported two human fetuses with multiple congenital anomalies including CDH that were homozygous for a missense variant in the HLX gene. The Hox gene family encodes for multiple transcription factors that have crucial regulatory functions during embryonic development (103). Targeted mutation of the homeobox B4 (Hoxb4) gene in mice resulted in offspring with poorly formed diaphragms and diaphragmatic defects, strikingly similar to the phenotype seen in humans with anterior CDH (104).

The Slit guidance ligand (Slit) family of proteins comprise a group of molecules with crucial functions in cell migration and adhesion through interaction with roundabout (Robo) receptors. Slit genes are expressed in the mesothelium of the diaphragm during embryogenesis (105). Homozygous Slit3^–/– mice experience faulty detachment of the central tendon region of the diaphragm from the underlying liver due to connective tissue defects, thus causing central-type (i.e., septum transversum) CDH (105, 106). Therefore, this genetically modified model is facing the disadvantage of having the diaphragmatic defect on or near the ventral midline portion of the central tendon as opposed to the posterolateral diaphragm, thus representing less than 5% of CDH cases seen in human patients. Further malformations in this mouse model include ureteric and renal agenesis in combination with intrathoracic herniation of liver and gallbladder (106), which again occurs infrequently in humans with CDH. Until now, no SLIT3 mutations have been identified in CDH patients. Robo genes encode large transmembrane receptors that are involved together with their ligands in numerus developmental mechanisms (107–109). For example, the Slit-Robo signaling pathway has been reported to have various fundamental functions including axon guidance, neural crest cell migration, epithelial cell adhesion, embryonic heart formation as well as diaphragm and kidney development (105–107, 110–114). Inactivation of Robo1 and Robo2 genes in mice has been shown to cause diaphragmatic defects and subsequent herniation of the stomach into the thorax, which leads to poor lung inflation and perinatal death, similar to human CDH cases (107). Homozygous mice with targeted deletion in the Dutt1/Robo1 gene often die at birth due to respiratory failure, demonstrating delayed lung maturation and diaphragmatic defects in some instances (115). More recent studies identified the heparan sulfate proteoglycan as an essential part of the Slit-Robo signaling complex, which stabilizes the Slit-Robo interaction (116). Furthermore, Zhang et al. (117) have noted that absence of the heparan sulfate biosynthetic enzyme N-deacetylase-N-sulfotransferase-1 (Ndst1) in the mouse endothelium interferes with vascular development in the primordial diaphragm, resulting in hypoxia as well as diaphragmatic hypoplasia and central-type CDH. The observed phenotypes in these animals mirror the congenital anomalies seen in Slit3 knockout mice. In addition, implementation of a heterozygous mutation in the Robo4 gene, which encodes the receptor of Slit3, exacerbated the defect in vascular and diaphragmatic formation (117). Thus, these findings suggest that loss of Ndst1 may lead to abnormal vasculogenesis in the diaphragm and CDH and that heparan sulfate in turn promotes the angiogenic Slit3-Robo4 signaling cascade during normal vascular patterning. Apart from this, mice homozygous for null mutations in the platelet-derived growth factor receptor α (Pdgfra) gene exhibit not only posterolateral diaphragmatic defects, they also develop a spectrum of other comorbidities including cardiovascular anomalies, renal and urogenital malformations, facial clefts, lung hypoplasia and failure of alveogenesis (118, 119).

Inbred C57BL/6J mice chemically mutagenized with ENU displayed a previously unknown mutation in the fuzzy planar cell polarity protein (Fuz) that was associated with CDH, liver protrusion into the chest cavity and pulmonary hypoplasia with a single left lung lobe (120). The mesenchymal-epithelial transition factor (Met) gene encodes for a receptor tyrosine kinase that is necessary for the migration of muscle precursor cells into the forming diaphragm (121), whereas fibroblast growth factor 10 (Fgf10) is crucial for early organogenesis of the lung (122). Oral administration of the herbicide nitrofen in Met^–/– mice with amuscular diaphragms and Fgf10^–/– mice with hypoplastic lungs resulted in CDH in both murine models, indicating that diaphragmatic defects may develop independently of myogenesis and pulmonary development (123). A novel missense variant affecting the actin-binding domains of plastin 3 (PLS3) was recently identified in eight unrelated families, causing X-linked CDH and body wall defects. A genetically modified mouse model of this Pls3^W499C variant resulted in perinatal death and reproduced the main features of the human phenotype, including diaphragmatic and body wall abnormalities (124). An abnormal plastin-actin interaction is the most likely explanation for the observed congenital malformation in both humans and mice.