Intrinsic lung neurons, whose cell bodies reside within the lungs, are derived from neural crest cells (NCC), a highly migratory, transitory cell population that gives rise to a wide diversity of cell types throughout the embryo including all neurons and glia of the peripheral nervous system [22]. Their development begins with the migration of NCC from the dorsal aspect of the neural tube into and along the embryonic foregut. Fate mapping studies in chick embryos demonstrated that NCC arising in the vagal (hindbrain) region of the neural tube, adjacent to somites 1–7, migrate in a rostrocaudal direction along the entire length of the gut to form the enteric nervous system (ENS), the intrinsic innervation of the gastrointestinal tract [23], and reviewed in [24–26]. As these vagal NCC migrate along the foregut, a subpopulation of cells moves tangentially from the foregut into the adjacent lung buds, a process revealed by NCC tracing in the chick [11] and mouse [19], and by antibody labelling of NCC in the human embryo [12]. This subpopulation of NCC therefore appears to undergo a two-step migration from one organ to another: first from the vagal region of the neural tube into and along the gut, and then tangentially from the (fore)gut into the lungs (Fig. 4.2). Concomitant with this early vagal NCC migration, the trachea and lung buds are in the process of separating from the foregut, so the lung buds and gut endoderm are physically connected, facilitating the direct migration of NCC into the lung buds. At these early stages of lung bud development, NCC are present in relatively high numbers compared to the size of the buds (Fig. 4.2f–i). However, this appears to be a transitory situation, as the lung buds subsequently increase in size dramatically, decreasing this ratio.

Within the chick lung , vagal NCC have been shown to differentiate to form neurons that are present adjacent to the future airways and are closely associated with ASM [11]. These intrinsic lung neurons appear to aggregate and form the airway parasympathetic ganglia. Lung innervation studies in human [12, 29], pig [30] and mouse [15, 20] demonstrated that intrinsic lung innervation is closely associated with ASM in the distal lung. Surrounding the future airways, a neuronal plexus lies on top of the ASM and extends into the distal lung over time along the growing airways [15]. Intrinsic airway ganglia are often found at conducting airway branching points and act as sites of convergence for several different classes of pulmonary innervation [31]. They receive input from extrinsic and other intrinsic neurons, and may act as sites of signal integration before signaling to downstream targets of innervation in the ASM [32].

Although the mechanisms and signaling cues that direct the tangential migration of NCC from the gut into the lungs have not yet been elucidated, perhaps the most plausible explanation for this behavior is a cell autonomous (chemotactic) response to a lung -specific signal. Such mechanisms, involving the guidance proteins Netrins (netrin-1 and -3) and netrin receptors (DCC, deleted in colorectal cancer, and the adenosine A2b receptor) have been reported to be involved in the perpendicular “secondary” migration of NCC from the outer (myenteric) to the inner (submucosal) layers of the gut and also, interestingly, out of the foregut into the developing pancreas [33], where NCC also form ganglia [34]. However, to date no specific molecular signature, or unique expression of cell surface receptors, has been identified in the NCC that colonize the lung. In fact, in chick embryo studies addressing this question, lung NCC were found to express many genes in common with enteric NCC, including Sox10, EdnrB, and Ret [11]. A large body of evidence has demonstrated the importance of the RET signaling pathway in ENS development , and activation of Ret receptor and Gfra1 co-receptor on migrating enteric NCC by the GDNF ligand is essential for NCC migration, survival, proliferation, and differentiation (reviewed in [35]). Due to this key role for RET in NCC development, and that Ret ^−/− mice have a depressed ventilatory response to inhaled CO₂ [36] the potential role of RET signaling in the development of lung innervation has been investigated by a number of groups. Our group showed that Ret-expressing lung NCC are attracted to GDNF in cultured lungs [19]. Further, studies by Langsdorf et al. [20] demonstrated that in Ret ^−/− mice the number of intrinsic lung neurons is reduced by 80 % in the trachea and primary bronchi, but these effects were not evident in Gdnf ^−/− and NRTN ^−/− mice suggesting redundancy in signaling through the Gdnf family ligands, Gdnf, NRTN, artemin and persephin [37]. However, in conflicting studies, when Freem et al. examined the lungs of Ret ^−/− and Gfra1 ^−/− mice, no major differences in the extent of lung innervation were observed [19]. Nevertheless, Langsdorf et al. reported that in Ret mutant embryos intrinsic lung ganglia were reduced in size, further pointing to a possible role for the RET signaling pathway in survival and/or proliferation of lung NCC [20]. Thus although some studies have implicated signal(s) such as Ret in NCC colonization of the lung, other studies raise alternate possibilities. Freem et al. demonstrated that NCC following the enteric migration route along the gut can equally re-migrate to the lungs after back-transplantation from the gut to their initial site of migration within the vagal neural tube [27]. This suggests that vagal NCC are not prespecified to colonize either the gut or the lung and therefore they may migrate into the lung passively or by responding to (as yet unknown) signaling cues expressed by the lung and/or the gut. How NCC discriminate between signals such as Ret/Gdnf in the gut and lung remains to be determined but differing levels of Ret expression by these cell types may be important, as has previously been shown for vagal and sacral NCC that differ in their ability to colonize the gastrointestinal tract [38].

Another possible mechanism for facilitating migration of NCC from the gut into the lung is via the vascular system, which has been suggested to act as a substrate for the directed migration of NCC in the gut [39]. In support of this idea, the vascular system begins to form within the developing lung buds concomitant with the migration of vagal NCC into the lung [28]. However, our recent studies in chick embryos do not support this idea, as it was found that in the lung the innervation and the vascular system initially develop on opposite sides of the lung buds, clearly establishing two distinct early spatiotemporal patterns [28] (Fig. 4.3a–c). Interestingly though, the nerve and vascular networks are found in very close proximity at later stages of development , suggesting that neurovascular units within the lungs are functionally important (Fig. 4.3d) [28].

Throughout their early development , lung NCC are closely associated with the developing vagus nerves (see Fig. 4.2). Tollet et al. using a p75^NTR antibody to label neural crest-derived cells in mouse embryos, revealed NCC in the vagus nerve and in processes extending from the vagus to the lung [15]. This suggests that NCC may also migrate along these paired extrinsic nerves to gain access to the lung. Indeed, following preestablished nerve fibers has already been recognized as a migratory behavior for another population of NCC, the sacral NCC, which contributes to the hindgut ENS. Studies on chick [40, 41] and mouse [42] have shown that the sacral NCC, which are known to share molecular signatures with vagal NCC [38], follow already established nerve projections from the extrinsic nerve of Remak in the chick, or pelvic plexus in the mouse, as a route for gaining entry to the hindgut. Moreover, this nerve-associated migratory behavior seems to be more common than previously recognized. In particular, a recent study from the Brunet laboratory showed that migrating cranial NCC can only form parasympathetic ganglia after being guided to the site of ganglion formation by nerves laying down fibers in “preganglionic” territories [43]. Interestingly, it can also be noted that an organ like the liver, which is only innervated by extrinsic nerve fibers after the initial wave of NCC migration along the gut has passed, does not contain any intrinsic ganglia [44]. Together, these studies suggest that preestablished neural projections offer an additional pathway for migrating NCC to colonize the lungs and that an absence of projecting nerve fibers could even prevent NCC entering an organ, such as the liver, during early development. However, detailed analysis of the interrelationships between intrinsic and extrinsic innervation is hampered by the fact that there are no known developmental models with a selective deficiency in intrinsic lung neurons, and by a lack of molecular markers that distinguish intrinsic from extrinsic neurons.

As discussed above, extrinsic pulmonary innervation begins to develop during embryonic stages concomitant with intrinsic lung innervation. The extrinsic innervation of the lung, which will allow direct bronchopulmonary communication with the central nervous system (CNS), arises from the vagal and sympathetic nerve trunks of the autonomic nervous system. A proportion of extrinsic lung innervation, the vagus nerve fibers from the vagal nodose ganglia, is derived from the nodose epibranchial placode [45]. Other sensory innervation, including that from the dorsal root ganglia, vagal jugular ganglia, and cervical thoracic ganglia, is neural crest-derived [46]. The vagal fibers that innervate the mouse lungs have been shown to contain a mix of placode-derived and neural crest-derived fibers [47].

Parasympathetic innervation from the brainstem, nodose ganglia, and jugular ganglia enters the lung via the vagus nerve, while sympathetic pulmonary innervation comes from the dorsal root ganglia and cervical thoracic sympathetic ganglia [48]. In contrast to intrinsic lung neurons, whose numbers have been reported to be reduced in Ret mutant mice [20], ASM innervation in the distal lung remains unaffected in these mutants, suggesting that extrinsic lung innervation may require a different neurogenic signal [21, 49]. Studies of E13.5 mouse embryos have shown that extrinsic axons extend into the distal lung and are present in the vicinity of developing ASM [15]. This association between ASM and growing axons is maintained during later embryonic stages and elaborates into a complex neural network closely associated with the smooth muscle [15]. This temporal and spatial interrelationship between axon outgrowth and ASM expression led to the investigation of potential ASM-derived neurotrophic factor(s) coordinating these two aspects of lung development [49, 50]. Xingbin’s group showed that, as Shh signaling induces the specification of ASM, it also represses an inhibitory microRNA (miR-206), which leads to the activation of Brain-derived-neurotrophic-factor (BDNF) expression in the smooth muscle. BDNF is then used as an ASM-derived neurotrophic signal by the extrinsic neurons to extend their axons into the distal lung [49]. In accordance with this result, a study using TrkB (the receptor for BDNF) null mice, showed that the lungs of these mutants had thinner bronchial epithelium , a larger airway luminal diameter, and larger alveolar spaces than their wild-type siblings. This suggests that BDNF/TrkB is important for cell proliferation and/or migration in the lung and ultimately for adequate development of epithelium and ASM [50, 51].

Again, as for the NCC that migrate from the gut into the lung , the guidance cues that direct axons into and within the lungs are elusive, although parallels can be made with signals guiding extrinsic axons innervating the gut. In this organ vagal sensory axons have been reported to respond to chemoattractive netrin signaling cues mediated through the receptor Deleted in Colon Cancer (DCC) as shown in in vitro experiments. Vagal axons in the gut terminate their response to netrin in the presence of laminin, expressed in the extracellular matrix of the gut, suggesting a possible mechanism for vagus nerve terminal positioning [52].

Central congenital hypoventilation syndrome (CCHS) is a disorder that results in an absence of adequate autonomic control of respiration with decreased sensitivity to hypoxia and hypercapnia. In CCHS respiration slows dramatically during sleep and does not increase in response to hypoxia, causing suffocation. CCHS is primarily considered a disorder of central respiratory pattern generation and can cause death before the age of 2 years (reviewed in [80]). Mutations in PHOX2B are strongly associated with CCHS in humans [80–83] and have a similar effect on respiration in mouse models [84].

Phox2b is a paired homeobox transcription factor expressed by neurons in the central and peripheral nervous system [85]. It is necessary for the development of much of the autonomic nervous system, including neural crest-derived peripheral autonomic innervation and neural crest-derived intrinsic enteric neurons [86, 87]. Phox2b is also important in the development of neurons in the brainstem that regulate ventilation in response to carbon dioxide levels [88]. There also exists a weaker association between CCHS and RET mutations, in which RET mutations increase the risk of intestinal aganglionosis when accompanying PHOX2B mutations [89]. Phox2b has been reported to regulate the downstream expression of Ret and Gdnf [86]. As Ret ^−/− mice have aganglionosis due to failure of NCC to colonize the entire gut and also a depressed ventilatory response to inhaled CO₂ [36], there may be an, as yet unidentified, link between NCC development and breathing control.