Introduction

Top Abstract Introduction Materials and methods Results Discussion References

The conversion of L-arginine to L-citrulline is catalyzed by the enzyme nitric oxide synthase (NOS) and results in the release of nitric oxide (NO). This simple molecule is known to be involved in diverse intercellular communications, such as those leading to relaxation of smooth muscle, regulation of blood pressure, inhibition of blood clotting, and cytotoxic activity by activated macrophages (1). NO has also been reported to act as a neuronal messenger (2). NO not only plays a regulatory role in the central nervous system, but also in the peripheral nervous system: neurons and nerve fibers immunoreactive for NOS have been reported in several mammalian organs (6). Two major classes of NOS enzymes produce NO: the constitutive isoforms present in neurons (type I) and endothelial cells (type III), and the inducible isoform (type II) found in macrophages and other cells (1, 12).

In the respiratory tract, NO is believed to play a role in various pulmonary physiologic processes, such as vasodilation and bronchodilation (13). In addition to studies showing the presence of NO of endothelial origin (14, 16), several physiologic studies have indicated the involvement of NO in inhibitory nonadrenergic, noncholinergic (iNANC) neurotransmission in pulmonary smooth-muscle relaxation (17). On the other hand, immunoreactivity for NOS has been shown in pulmonary neurons and/or nerves of some mammalian species (20). Furthermore, immunostaining for NOS has also been found recently in the respiratory nerve plexuses of lower vertebrates (26). The presence and significance of NO in the mammalian lung is a matter of current interest (e.g., inhaled NO is being used as a new therapy for different respiratory syndromes, taking advantage of the vasodilatory action of NO both in newborns [27, 28] and adults [29, 30]).

At birth, the transition from fetal to neonatal life involves relaxation of both vascular and airway smooth muscle (31). As a consequence of the decline in the resistance of pulmonary vessels and airways, an increase in blood flow and oxygenation takes place. Several chemical factors seem to be involved in respiratory control at birth (32). Recent physiologic studies indicate that endothelial NO is present at birth and may have an important role in the relaxation of vascular smooth muscle in the perinatal period and in postnatal maturation (33, 34). Moreover, increases in both the production of NO by endothelium (in lambs, [35]) and in the expression of types I and III NOS in late fetal and neonatal lung (in rats [36, 37]) have been found through biochemical and immunocytochemical methods.

It is unknown whether NO of neural origin could be involved in a similar manner, in the relaxation of airways and/or blood vessels at the time of birth, acting as an intercellular signal and/or neurotransmitter. In fact, no histochemical studies have been reported for the presence of NOS in mammalian respiratory tract neurons during late fetal life. The aim of the present study was to investigate the presence of neuronal NOS (nNOS) in mouse respiratory nerve plexuses, paying special attention to the occurrence of nNOS during lung development and in perinatal life, which has received little attention. Using both enzyme histochemical and immunocytochemical techniques, we investigated the time of appearance of nNOS in the developing mouse lung, as well as the distribution of nNOS-containing structures in fetal, neonatal, and adult lung.

	Materials and Methods

Top Abstract Introduction Materials and methods Results Discussion References

Specimen Collection and Processing

A total of 52 Swiss mice were studied, consisting of eight sexually mature females (adult lungs, P-ad) and 44 mice at various stages of development, including gestational (E) days 13 to 19 and postnatal (P) days 0, 1, 2, and 6 (the age at which mouse lung is considered to be mature); four subjects of each age were studied.

The pregnant mice were anesthetized with 12.5% urethane (1 ml/100 g). The abdomen was opened and the fetuses were rapidly removed and chilled in ice. The newborn mice were anesthetized with ice. The chests of fetuses, newborn mice, and adults were opened and the tracheae and lungs were removed, except for fetuses on gestational days 13 and 14, which were immersed in toto in fixative.

Tissue for cryostat sections (lungs) and for whole mounts (tracheae) was fixed for 4 h at 4°C in 1% paraformaldehyde, and was then immersed overnight in 0.16 M phosphate-buffered saline containing 15% sucrose. Cryosections (6 to 8 µm thick) were collected on Vectabond-coated glass slides. For paraffin embedding, the material was immersed in Bouin's fixative for 20 to 24 h. Paraffin sections were 4 to 6 µm thick.

Conventional Staining

Paraffin sections were hydrated and stained with hematoxylin and eosin (H&E) or Masson's trichrome.

Histochemical Staining for Nicotinamide Adenine Dinucleotide Phosphate-Diaphorase

NOS-containing neurons were identified by the histochemical technique for nicotinamide adenine dinucleotide phosphate-diaphorase (NADPH-d), which is usually regarded as a marker for nitrergic neurons. To study the morphologic relationship of NOS-containing nervous structures, the NADPH-d technique was applied not only to cryostat sections but also to whole mounts of specimen material. However, because the small size of fetal lungs did not allow an in toto study, the NADPH-d technique was applied only to tracheae.

Lung cryostat sections or whole mounts of tracheae were washed in phosphate buffer (PB) 0.1 M and then treated with 0.3% Triton X-100 in PB for 20 min at room temperature. After washing in PB, the specimens were incubated with a reaction mixture consisting of 1 mg/ml NADPH and 0.2 mg/ml nitroblue tetrazolium in PB at 37°C in the dark. Incubation for 45 to 60 min provided optimal staining. After washing in PB, tracheae were slit open on the ventral side and placed with mucosa side down on slides. Both cryostat sections and tracheae was covered with glycerol mounting medium followed by the application of coverslips.

Immunocytochemistry

Two polyclonal antibodies to nNOS, diluted 1:500, were used. Both antibodies were characterized by Western blotting (38, 39). One of the antibodies was raised against whole nNOS, purified from an extract of rat brain. The second antibody was raised against a synthetic peptide from the deduced sequence (LPLLQANGNDPELFQIPPELC) of cloned neural NOS (40). The specificity of this antibody in mouse lung was confirmed by its preabsorption with the synthetic peptide. In addition, a polyclonal antibody raised against protein gene product 9.5 (PGP9.5) and diluted 1:5,000 (a kind gift of J. Polak of the Department of Histochemistry of the Royal Postgraduate Medical School, London, UK) was used as a general neuroendocrine marker.

Paraffin sections were dewaxed with xylol. Cryostat sections were pretreated with 0.3% Triton X-100 in Tris-HCl-buffered saline ([TBS] Tris buffer: 0.05 M, pH 7.36; NaCl 0.5 M) for 45 min and dehydrated through graded alcohols. Both paraffin and cryostat sections were treated with the avidin-biotin complex technique (41). Endogenous peroxidase was blocked by treatment with 3% H₂O₂ in methanol for 30 min. Sections were hydrated through alcohols and then placed in TBS. Nonspecific binding sites were blocked with 5% normal swine serum in TBS. After overnight incubation at 4°C with the primary antibody, sections were incubated for 30 min with biotinylated antirabbit immunoglobulins (E353; Dakopatts, Glostrup, Denmark), followed by a third incubation with avidin-biotin- peroxidase complexes (K355; Dakopatts) for 30 min. After each incubation, sections were rinsed in TBS. The peroxidase activity was demonstrated by means of 0.03% 3,3'-diaminobenzidine tetrahydrochloride (D-5637; Sigma, St. Louis, MO) in sodium acetate/acetic acid 0.1 M, pH 5.6, containing 2.5% ammonium nickel sulfate, 0.2% -D-glucose, 0.04% ammonium chloride, and 0.001% glucose oxidase (42). The sections were counterstained with hematoxylin, dehydrated, and mounted in DPX.

	Results

Top Abstract Introduction Materials and methods Results Discussion References

Structure of Developing Mouse Lung

On gestational days 13 to 16 the developing mouse lung is composed of epithelial tubules made up by cuboidal or columnar cells, which are surrounded by mesenchymal tissue (Figure 1a). On the following days, the epithelium flattens, and on gestational day 18, in addition to the immature bronchioles just described, saclike primitive alveoli with an irregular outline (Figure 1b) are present. The primitive alveoli are subdivided into several alveoli through the gradual appearance of secondary crests, until the lung acquires the morphology of the adult animal, at about postnatal day 6. In mouse respiratory tract, once the extrapulmonary bronchi penetrate the lungs, the cartilage rings disappear and give rise directly to the bronchiolar tree, without intermediate intrapulmonary bronchi (Figure 1c).

View larger version (86K): [in this window] [in a new window]   Figure 1.   Histology of developing mouse lung. (a) Gestational day 14 (E-14). Epithelial tubules (T) surrounded by mesenchymal conective tissue (M) are seen. Masson's trichrome. (b) E-19. A connection (star) is observed between the bronchiolar tree (b) and the primitive alveoli (asterisks). Masson's trichrome. (c) Postnatal day 0. Extrapulmonary bronchi (B) with cartilage rings (arrows) are seen in the hilar region. Note that the bronchiolar system (b) does not contain cartilage tissue. H&E stain. Bars: a, b = 100 µm; c = 250 µm.

Distribution of Nerve Plexuses in the Mouse Lung

In postnatal animals, nerve ganglia were easily identified by the presence of large neuron bodies with euchromatic nuclei and conspicuous nucleoli. In fetuses, the size of neuron bodies was smaller, obscuring their identification in conventionally stained ganglia. The distribution of mouse respiratory nervous structures was therefore studied by means of an antibody against the general neuroendocrine marker PGP9.5. With this, strong immunoreactivity was found in neural structures at all fetal stages studied (Figure 2). The nerve ganglia and nerves of both fetuses and adults were localized within or next to the walls of blood vessels, bronchi (Figures 2d and 2e), and bronchioles (Figure 2c), all three of which contained muscle cells. The size of the nerve ganglia decreased from the hilum to the bronchioles. The largest ganglia contained up to 35 neuron bodies per section, whereas the smallest ones contained only two or even one.

View larger version (155K): [in this window] [in a new window]   Figure 2.   Immunoreactivity for the general neural marker PGP9.5 in mouse fetal lungs. (a) Panoramic view of a sectioned gestational day 14 (E-14) mouse fetus. Neural structures (arrows) have been stained with an antibody against the general neural marker PGP9.5. The five pulmonary lobes (asterisks) characteristic of mouse lung are observed; n = neural tube; l = liver; k = kidney. (b) Detail of an extrapulmonary PGP9.5-positive nerve trunk (boxed area in a) that gives rise to a ganglionated branch (arrow) toward a pulmonary lobe; M = mesenchyme; T = pulmonary epithelial tubules; V = blood vessel. (c) E-14. Intrapulmonary PGP9.5-positive microganglia (arrows) and nerve fibers (arrowheads) close to epithelial tubules (T) (immature bronchioles). (d) E-18. Cross-sectioned nerves (N) and ganglia (arrows) immunostained with anti-PGP9.5 antibody are seen in the adventitial connective tissue of a bronchus (B); C = cartilage ring. (e) Detail of one of the nerve ganglia in d. E = bronchial epithelium. Bars: a = 1,000 µm; b, d = 100 µm; c = 25 µm; e = 20 µm.

Distribution of NADPH-d-Labeled Nerve Elements in Mouse Lung

NADPH-d staining was first detected in neuron bodies of mouse pulmonary nerve plexuses from gestational day 13 (Figures 3a and 3b) onward, being present during lung development (Figures 3a to 3f) and in adulthood (Figures 3g and 3h). The NADPH-d-labeled cells observed at gestational day 13 were interpreted as neuron bodies on the basis of their labeling and localization, although they were smaller and showed fewer cytoplasmic processes (Figures 3a and 3b) than in later fetal stages (Figures 3c to 3f).

View larger version (151K): [in this window] [in a new window]   Figure 3.   NADPH-d activity in neuron cell bodies of developing (a-f ) and adult (g, h) mouse lung. (a) Gestational day 13 (E-13). Putative NADPH-d-positive neuron body (arrow) near a blood vessel (V) in the pulmonary hilar region. (b) Detail of the neuron body. (c) E-15. Neuron cell bodies, apparently in contact through neuronal processes (arrowheads), displaying NADPH-d activity. (d) E-16. Two NADPH-d-positive neuron bodies (arrows) next to a bronchus (B). (e) Detail of one of the neurons in d. Neuronal processes (arrowheads) can be observed. ( f ) E-17. NADPH-d-positive neuron body next to a bronchiole (b); arrowheads = neuronal processes. (g) Adult animals. Hilar nerve ganglion containing two or three neurons positive for NADPH-d by histochemistry; arrowheads = neuronal processes. (h) Adult animals. NADPH-d-positive neuron body next to a bronchiole; arrowheads = neuronal processes; E = bronchiolar epithelium, also showing NADPH-d staining but not nNOS immunolabeling. Bars: a, d = 100 µm; b, e, f, h = 20 µm; c, g = 25 µm.

NADPH-d labeling was detected in both neuron cell bodies (Figure 3) and terminal nerve fibers (Figure 4) of mouse pulmonary nerve plexuses. Stained neuron bodies were found in the hilar region, next to blood vessels (Figures 3a and 3b) and bronchi (Figures 3d and 3e), as well as in the nerve ganglia located next to bronchioles (Figures 3c and 3f). Frequently, only one (Figures 3e and 3h) or a few (Figure 3g) positive neuron bodies were observed per nerve ganglion. Neuronal processes arising from the neuron bodies were neatly labeled with the NAPDH-d technique (Figures 3e to 3h). NADPH-d-positive neuron bodies, apparently in contact with one another through long neuronal processes, were also observed, especially in thicker cryostat sections (Figure 3c).

View larger version (159K): [in this window] [in a new window]   Figure 4.   Terminal NADPH-d- positive nerve fibers with varicosities (arrowheads) in bronchi (a, b), bronchioles (c-e), and blood vessels ( f, g). (a) Gestational day 19 (E-19) bronchus (B). (b) Adult bronchus (B); m = smooth muscle. (c) E-17 bronchiole (b). (d) E-18 mouse lung. Long cytoplasmic processes (double arrowheads) arising from neuron body (arrow) are seen to extend toward the upper bronchiolar wall (b), and probably toward a blood vessel (V). (e) Bronchiole (b) of an adult mouse lung. ( f  ) E-18 mouse lung. Nerve fibers (double arrowheads, bottom) extend toward and reach the wall of a pulmonary blood vessel (V). Two other nerve branches (double arrowheads, left and right), extending toward two bronchioles (b), are also present. (g) Adult animals. Tangential section of an intrapulmonary blood vessel (V). Several isolated nerve fibers can be observed within the wall. Bars: a-c, e = 20 µm; d, f = 40 µm; g = 25 µm.

Thin terminal NADPH-d-positive nerve fibers showing varicosities were observed among the muscle cells of bronchi (Figures 4a and 4b), bronchioles (Figures 4c to 4e), and vessels (Figures 4f and 4g). They were more frequent in vascular (Figure 4g) than in bronchial (Figures 4a and 4b) or bronchiolar (Figures 4c and 4e) walls.

Distribution of nNOS-Immunoreactive Neurons in Mouse Lung

Immunoreactivity for NOS was first detected on gestational day 15 (Figure 5a), and extended through development (Figures 5a to 5f) and adulthood (Figures 5g and 5h). By comparison with NADPH-d staining, labeled neural structures showed no apparent differences in terms of location or frequency of NOS in fetuses, newborns, or adults. NOS-immunoreactive neuron bodies were also observed in the hilar region (Figures 5b to 5d) and next to intrapulmonary bronchioles (Figures 5a and 5e to 5h). Also, with the specific anti-nNOS antibody, only a small proportion of neuron bodies were labeled: most of the histologic sections of nerve ganglia did not contain stained neurons, and in labeled ganglia, only one or a few positive neurons were observed (Figure 5d). The two anti-nNOS antibodies used in the study produced the same immunostaining. Preabsorption with the corresponding antigen abolished immunostaining (Figures 5h and 5i).

View larger version (166K): [in this window] [in a new window]   Figure 5.   nNOS immunoreactivity in developing (a-f  ) and adult (g-i) mouse lung. (a) Gestational day 15 (E-15) mouse lung. Neuron body immunoreactive to NOS (arrow) next to a bronchiole (b). (b) E-16. Immunostained neuron body (arrow) in the hilar region in the proximity of a blood vessel (V); B = bronchus; C = cartilage rings. (c) Detail of the neuron body shown in b. (d) Postnatal day 0 (P-0). Hilar nerve ganglion: only one (arrow) of the approximately 30 neuron bodies present in the section is clearly immunoreactive for nNOS. (e) P-0. Immunocytochemistry for nNOS in a cryostat section. Short, immunoreactive neuronal processes (arrowheads) can be observed; b = bronchiole; m = bronchiolar smooth muscle. ( f  ) P-6. Immunocytochemistry for nNOS in a paraffin section; m = smooth muscle. (g) Adult animal. Neuron body immunoreactive for nNOS. No immunostained cytoplasmic processes are observed. (h, i) Serial reversed-face paraffin sections of adult mouse lung. NOS immunoreactivity (h) is abolished with the preabsorbed antiserum; arrows = neuron body; V = blood vessel. Bars: a, d, e, h, i = 20 µm; b = 100 µm; c, f, g = 10 µm.

Both the NADPH-d technique and immunocytochemistry for nNOS produced staining of the same neuron cell bodies, as shown in the cryostatic serial sections in Figure 6. Nevertheless, some differences between NADPH-d staining and NOS immunocytochemistry were observed. First, as indicated, immunoreactivity to NOS was detected later (gestational day 15, Figure 5a) than was NADPH-d staining (gestational day 13, Figures 3a and 3b). Second, although NADPH-d-labeled processes arising from neuron cell bodies were conspicuous (Figures 3e, 3f to 3h, and 6a), NOS-immunoreactive neuronal processes were more infrequent and shorter (Figures 5e and 6b) or even absent (Figures 5c and 5g). Additionally, terminal nerve fibers were detected only with the NADPH-d technique (Figure 4).

View larger version (87K): [in this window] [in a new window]   Figure 6.   Serial cryostat sections. The same nerve ganglion (arrow) is stained with both NADPH-d histochemistry (a) and immunocytochemistry for nNOS (b). Note that neuronal processes (arrowheads) are neatly labeled with the NADPH-d technique but show almost no nNOS immunoreactivity; b = bronchiole; V = blood vessel. Bars: 20 µm.

NADPH-d Activity in the Mouse Tracheal Plexus

As indicated, the tracheal tubular structure allowed application of the histochemical technique in whole-mount preparations of both fetuses (Figures 7a to 7e) and adults (Figure 7f). NADPH-d-reactive neuron cell bodies and nerve fibers were found in tracheae from gestational day 15 (Figure 7a) onward. Mouse tracheae in earlier stages than gestational day 15 could not be examined because they were too small to be removed without damage. The NADPH-d-positive nerve plexuses were similar in fetal and adult tracheae. As in the lung, only a small proportion of neuron cell bodies showed NADPH-d activity (Figures 7e and 7f). Tracheal neurons also showed conspicuous processes that seemed to establish an anatomic link between them, forming a fine network (Figure 7c). Nerve processes showed conspicuous varicosities (Figures 7a and 7e).

View larger version (153K): [in this window] [in a new window]   Figure 7.   NADPH-d activity in whole-mount tracheae of developing (a-e) and adult ( f  ) mice. (a) Gestational day 15 (E-15). NADPH-d positive neuron body. Varicosities (arrowheads) of nerve processes and fibers can be observed. (b) E-17. Panoramic view. The dorsal face (DF) is in the center, edged by sectioned cartilage rings (C). (c) Detail of the boxed area in b. A fine network of neuron bodies (arrows) and nerve fibers (arrowheads) is present. (d) Detail of upper neuron body. (e) E-19. Tracheal nerve ganglion containing some neuron bodies displaying NADPH-d activity, two of them in close proximity (arrow); arrowheads = varicosities. ( f  ) Adult animals. Only two neuron bodies show NADPH-d activity. Bars: a, d = 20 µm; b = 250 µm; c = 100 µm; e, f = 25 µm.

	Discussion

Top Abstract Introduction Materials and methods Results Discussion References

The present histologic study identified NOS in the nerve plexuses of mouse lung and trachea from fetal life onward. The finding of both staining for NADPH-d and NOS immunoreactivity supports the presence of NOS in murine respiratory nerve plexuses. Labeling appeared during fetal development, at about gestational days 13-15, and remained throughout the postnatal and adult periods.

Studies of the functional significance of neurally derived NO in the mammalian respiratory tract have been done on postnatal specimens. We will therefore first discuss the functional significance of nNOS in the adult mouse respiratory tract, according to our findings.

iNANC Relaxation

The detection of NOS in neurons of murine and other mammalian lungs supports the hypothesis that NO is involved in iNANC-mediated relaxation of the respiratory tract. Nevertheless, the significance of NO involment in iNANC-mediated respiratory relaxation seems to differ among species and in different regions of the respiratory tract. It has been suggested that NO is the primary mediator of iNANC- mediated relaxation in some species (17, 19). Some authors have also stated that in humans, NO completely mediates iNANC-induced relaxation (13, 43), whereas others, using NOS inhibitors, have on the contrary shown that NO-mediated inhibition of iNANC-induced relaxation occurs to an extent of less than 50%, implying other mediators in the relaxation response (18). In agreement with the latter results is the finding that in human lung, only a small proportion of intrinsic neurons are immunoreactive for NOS (21). Similarly, only a few neurons in mouse respiratory plexuses are positive for NADPH-d or NOS, which also suggests a major involvement of other neurotransmitters in iNANC-mediated respiratory relaxation in this species.

NOS-immunoreactive neurons of both developing and adult mouse lung are localized in the hilar region and in the walls of bronchi and bronchioles. Previous studies of other mammals have also described NOS-positive neurons in the hilar region and in extra and intrapulmonary bronchi (21, 24), but not in bronchioles. These findings are in concordance with the existence in such mammals of nerve ganglia accompanying the large vessels and intrapulmonary bronchi, although only nerve fibers reach the bronchioles (44). The significance of NO in mammalian iNANC-mediated respiratory relaxation seems to be more important in trachea and bronchi than in peripheral airways (19). Such physiologic findings are in agreement both with the detection of NOS/NADPH-d-positive neurons and/or nerves in trachea and bronchi, and of only scarce NOS-positive nerve fibers in bronchioles (45), as well as with a reported decline in NOS-containing nerve fibers from upper to lower airways (22, 45). In the mouse lung, on the contrary, nerve ganglia, in addition to nerve fibers, accompany the bronchiolar tree. As shown in the present study, some NOS-positive neurons continue to be present in such small bronchiolar nerve ganglia. Both NADPH-d- labeled and NOS-labeled neurons and nerves are present in the bronchioles, which may indicate a more relevant role of NO in iNANC-mediated relaxation in mouse peripheral lung than in that of other species.

We observed thin, terminal NADPH-d-labeled nerve fibers with varicosities near smooth-muscle cellspresumably their target cellsof airways and blood vessels in the mouse. In other mammalian species, isolated nerve fibers have also been observed near muscle cells (24, 25, 45). In guinea pig lung (22) they have additionally been detected in the alveolar respiratory region. In the mouse, we found that the number of NADPH-d-positive nerve fibers in airways was smaller than in blood vessels at all stages studied. This observation suggests that the involvement of NO in iNANC-mediated relaxation of blood vessels may be greater than in the airways.

Pending its ultrastructural confirmation, our findings in both cryostatic thick lung sections and in tracheal whole mounts seem to indicate a structural relationship between NOS-containing neurons. In the myenteric plexus, electron microscopy has reportedly shown contacts between NOS-positive and both NOS-positive and NOS-negative neurons, and the involvement of NO not only in iNANC-mediated relaxation but also in the regulation of neuronal activity has been suggested (8).

Fetal Lung

To our knowledge, the earliest description of immunoreactivity for nNOS in mammalian pulmonary nervous structures was made in nerve fibers of the newborn pig (25). Although no other fetal detection of nNOS in mammalian species can be compared with our findings, the early detection of NOS/NADPH-d labeling in mouse fetal lung is consistent with the biochemical extraction of nNOS from rat fetal lung (36). The early presence of NOS on gestational days 13-15 mouse lung suggests a putative involvement of neurally derived NO in fetal lung, although the role of NO during this period is yet unknown. The presence from fetal life onward of NOS-containing nervous structures similar to those of adult lung suggests that the iNANC system exists in the developing mouse lung. In fact, it has recently been shown that before birth the porcine tracheobronchial tree appears to be functionally innervated by nitrergic input to the smooth muscle (46). However, other possibilities cannot be discarded (e.g., the involvement of NO in the establishment and maturation of synaptic connections, which has been indicated for the central nervous system [47, 48]).

Because the number and localization of NOS-containing nervous structures in the mouse is similar in the developing and adult respiratory tract, the iNANC for respiratory control mechanism does not seem to be strengthened in the perinatal period in this species. On the contrary, a decrease in NOS-containing nerve fibers with age has been reported in postnatal pig lung, suggesting that neurally derived NO may be important in pulmonary adaptation to extrauterine life (25). In fact, it has been shown in porcine bronchi that after birth there is a gradual transition away from NO toward a catecholaminergic pathway of relaxation (46). Similarly, an increase in NOS protein, with maximal levels in perinatal life, has been reported in the fetal rat (36). As indicated, the present study did not confirm similar changes in the expression of nNOS through variations in NOS-containing nervous structures in the late fetal period. However, it is difficult to reach definitive conclusions about a role for NO in early respiratory regulation because the number of neurons immunoreactive for nNOS in mouse respiratory nervous plexuses was small.

In summary, the results of the present study indicate that NO of neural origin is present in mouse lung from approximately gestational days 13 to 15 onward. Therefore, in addition to endothelial NOthe importance of which has already been claimedneurally derived NO may be considered to play a role in fetal and neonatal lung physiology. Further biochemical and physiologic studies are needed to identify the precise contribution of neurally derived NO in the developing mammalian lung.