Peribronchial smooth muscle constriction causes airway stretch, an important mechanical force in developing lung. Little is known about factors influencing these spontaneously active muscle elements. We measured contractile activity of neurokinin (NK) receptors on fetal intrapulmonary smooth muscle by tracheal perfusion assay (n = 11). Injecting either capsaicin or the NK2 receptor agonist [NLE10]NKA resulted in significant (P < 0.05) bronchoconstriction. A specific NK2 receptor antagonist inhibited constriction caused by endogenous tachykinins released by capsaicin. We then examined NK2 receptor (n = 44) and NKA (n = 23) ontogeny in human lung. NKA immunostaining was identified in peribronchial nerves in samples with gestational age >12 wk. NK2 receptor protein was identified in peribronchial and perivascular smooth muscle. These results indicate that endogenous tachykinins released by the developing lung act via NK2receptors to cause smooth muscle constriction. We speculate that tachykinins could modulate lung development.
- mechanical transduction
the ontogeny of the tachykinins, an important family of peptides, has not been studied extensively. Specifically, the ontogeny of the neurokinin-2 (NK2) receptor has not been elucidated. We studied the ontogeny of the NK2 receptor. Our interest in this receptor stems from the fact that the most potent native ligand of the NK2 receptor is neurokinin A (NKA), an extremely effective bronchoconstrictor. Activity of this receptor-ligand system would likely be to affect airway tone, which could in turn alter the mechanical forces affecting the developing lung.
Mechanical forces are known to influence pulmonary organogenesis. The severe pulmonary abnormalities associated with clinical syndromes, such as congenital diaphragmatic hernia (6, 21, 49, 52, 54, 59) and alterations in the normal amount of lung fluid (including oligohydramnios and laryngeal atresia; see Refs. 8,20, 45), are largely caused by disruption of the normal physical forces affecting the developing lung. Animal models using several different experimental approaches, including abrogating or minimizing the pressure differences during fetal breathing movements (3, 17, 26, 30), intrathoracic crowding during gestation as a result of intrathoracic balloon placement (18) or diaphragmatic hernia (5, 47), and changes in transpulmonary pressure as a result of altering the amount of lung fluid (3, 4, 22, 25, 43), have demonstrated that normal pulmonary organogenesis depends on complex interactions among distinct types of physical forces.
The mechanisms transducing the effects of mechanical forces on the developing respiratory system have not yet been completely elucidated. It is possible that components of the lung regulate some of these factors. Peribronchial smooth muscle is ideally located to modulate the physical forces acting on the developing lung.
Peribronchial smooth muscle begins to differentiate from the primitive mesenchyme at gestational week 6 in humans (37). Investigations using confocal microscopy have demonstrated that the peribronchial muscle is covered by a network of nerves, and neural extensions to the smooth muscle can be visualized by gestational week 7.5 (55). Immunohistochemical analyses of fetal rat lung have demonstrated an early muscle protein, α-actin, in the clefts of branching airways (29, 41,42). This location of immature peribronchial smooth muscle makes it a plausible candidate to modulate airway stretch during morphogenesis. Regular spontaneous peribronchial smooth muscle contractions have been observed in organ cultures from first-trimester human fetal lung (40). Although the amplitude and frequency of these contractions varied in response to smooth muscle agonists and relaxants, including isoproterenol and carbechol, possible responses to tachykinins were not evaluated (40). The potent contractile activity of the tachykinins in adult airway smooth muscle suggested to us that developing airway smooth muscle might constrict in response to endogenously released tachykinins.
Tachykinins are endogenous bronchoconstrictors with greater potency [that is, producing 50% of their maximal effects (ED50) at lower doses] than methacholine (1). This family of neuropeptides has diverse functions, including bronchoconstriction and vasodilation (9, 46). Two of the most common tachykinins, substance P and NKA, are released by unmyelinated sensory C fibers. These nerves also contain other neuropeptides, such as calcitonin gene-related peptide (CGRP). The neuropeptides released by C fiber nerves act via specific receptors: substance P is the most potent ligand of the NK1 tachykinin receptor, NKA is the most potent ligand of the NK2 receptor, and CGRP has two types of specific receptors (9, 10). Because NKA is a more potent agonist of the NK2 receptor in human bronchial tissue than either substance P or CGRP (9), tachykinin-related effects on airway stretch are likely mediated by the NKA-NK2 receptor system. NKA functions are well characterized in the adult lung and include bronchoconstriction and vasodilation (9, 46). In addition, NKA is a growth factor and chemoattractant for cultured human lung fibroblasts (19). It is therefore plausible that NKA could modulate airway development and/or repair. Although the maturing bronchial smooth muscle of the human neonate can constrict after NKA stimulation (14), little is known regarding possible actions of the tachykinins in the more primitive developing lung.
These observations led us to speculate that NK2receptor-mediated actions of NKA might contribute to the mechanical forces in developing lung. We hypothesized that the NK2receptor-NKA receptor-ligand system would be functional in the developing lung. To test this hypothesis, we examined the capacity of the developing lung to constrict to exogenous tachykinin agonists and to endogenous tachykinins released by capsaicin. Next, we investigated the expression of the NK2 receptor gene during gestational periods correlating with the pseudoglandular and canalicular stages of pulmonary organogenesis. Finally, we examined the cellular localization and ontogeny of NKA and NK2 receptor protein in developing human lung. Our investigations indicate that NKA and NK2receptors are both expressed and functional during these critical periods of pulmonary development.
Sample acquisition and specimen processing.
Tissue samples of normal adult and first- and second-trimester human lung were obtained from discarded surgical specimens. The protocols for this study were approved in advance by the Brigham and Women's Hospital Human Research Committee. Each tissue sample was partitioned, as size allowed, for RNA and immunohistochemical analyses. Samples for immunohistochemistry (n = 37) were fixed in 4% paraformaldehyde and then processed into paraffin-embedded tissue blocks. Samples for RNA analysis (n = 22) were frozen in liquid N2 and kept at −70°C until the RNA was extracted. Samples evaluated by the tracheal perfusion assay (n = 11) were immersed in ice-cold buffer (137 mM NaCl, 1.8 mM CaCl2, 1.05 mM MgCl2, 1 g/l dextrose, 0.6 mM NaHCO3, 0.13 mM NaH2PO4, and 0.896 Na2HPO4, pH 7.4) until placed in the perfusion chamber. Samples evaluated by both tracheal perfusion and RNA analysis (n = 9) were frozen in liquid N2at the completion of the tracheal perfusion assay.
Tracheal perfusion assay.
Eleven lung samples were evaluated by tracheal perfusion assay. Because of size constraints of the perfusion equipment and limits of sensitivity of the pressure transducer, only samples of at least 20 wk gestation were evaluated in this assay. Tracheal perfusion was performed as previously described (32, 33, 39). Briefly, the lungs were dissected en bloc and placed in ice-cold low-potassium perfusion buffer (137 mM NaCl, 1.8 mM CaCl2, 1.05 mM MgCl2, 1 g/l dextrose, 0.6 mM NaHCO3, 0.13 mM NaH2PO4, and 0.896 mM Na2HPO4, pH 7.4). A tracheal catheter was placed, and the lungs were suspended in a 37°C, 100% humidity, Plexiglas perfusion chamber. Perfusion buffer (low-potassium perfusion buffer with 2.68 mM KCl added) was warmed to 45°C and pumped at 5 ml/min through a bubble trap before being cooled to 37°C and administered to the lungs via the tracheal cannula. Perfusate exited the fully expanded lungs via small holes placed in the pleura. The openings in the pleura served to remove contributions of visceral pleural pressure to the airway opening pressure (Pao). The “back pressure” resulting from continuous-flow tracheal perfusion represents Pao and was recorded from a side tap at the tracheal cannula with a pressure transducer (P23Db; Statham Instruments, Oxnard, CA). Prior investigations demonstrated that, in isolated lung preparations with continuous flow, changes in Pao directly reflect changes in the contractile state of the lung (32, 33, 39). The lungs were allowed to equilibrate in the perfusion chamber for 15 min. After this, contractile agonists except [NLE10]NKA (NLE-10) were diluted in 100-μl volumes and injected in the tracheal cannula (Table1). The diluent for all agonists except NLE-10 was perfusion buffer. The NLE-10 solution was prepared by dissolving 1 mg of NLE-10 in 300 μl of DMSO (Sigma, St. Louis, MO) and then diluting this in 700 μl of perfusion buffer. NLE-10 solution (74.8 μl; 10 μmol) was injected in the tracheal cannula. Tracheal injection of either diluent did not affect Pao. All neuropeptides were obtained from Peninsula Laboratories (Belmont, CA). Because of the limited number of samples available for the tracheal perfusion studies, doses for the peptide agonists and SR-48968 were chosen from published studies examining human and animal model airway responses. Doses used in the tracheal perfusion assays were those that caused bronchoconstriction in adult guinea pig samples studied with tracheal perfusion assay (31, 32) and those demonstrated to be effective bronchoconstricting doses in other models (13, 48,63). At the completion of the assays, the intact capacity of the pulmonary smooth muscle to respond to agonists was demonstrated by administering 6.1 mg (0.025 mol) of the nonspecific smooth muscle agonist BaCl2.
RNA extraction and analysis.
Total RNA was prepared from frozen tissue using the Stratagene RNA isolation kit (Stratagene, La Jolla, CA). RNA integrity was evaluated by examination of 18S and 28S RNA bands after electrophoresis using ethidium bromide-stained 1.5% agarose formaldehyde gels.
RT preparation and PCR analysis.
cDNA was prepared using RT from 22 samples with gestational ages between 10 and 23 wk. Each RT reaction consisted of 1 μg of total RNA, Moloney murine leukemia virus RT (2 units; Bethesda Research Laboratories, Gaithersburg, MD), RNasin (5 units; Sigma), and random hexamers [1 unit, 5′-pd(N)6; Pharmacia LKB Biotechnology, Piscataway, NJ] in a total volume of 10 μl of 50 mM Tris · HCl (pH 8.3), 75 mM KCl, 10 mM dithiothreitol, and 3 mM MgCl2 for 60 min at 42°C. Each PCR contained 3 μl of the RT solution. Negative controls consisted of substituting an equal volume of diethyl pyrocarbonate-treated water for the volume of RNA in the corresponding reaction.
Synthetic oligodeoxynucleotides were purchased from Ransom Hill Bioscience (Ramona, CA). All primer pairs yielded products that spanned at least one intron to permit distinction between cDNA and any contaminating genomic DNA. Amplification of primers for β-actin served as the internal positive control, as has been previously reported (15, 16). Primers used for actin were 5′-GGG-CAC-GAA-GGC-TCA-TCA-TTC (antisense primer) and 5′-GGC-CCC-TCC-ATC-GTC-CAC-CGC (sense primer). Primers used for the NK2 receptor were 5′-GCA-TCA-CAG-CCT-TCT-CCA-T (antisense) and 5′-GGT-CTT-CAC-AAA-CTT-CTT-C (sense).
PCRs were carried out using Taq polymerase (Boehringer Mannheim, Indianapolis, IN), as specified by the manufacturer. Briefly, 3 μl of the RT mixture were added to 5 μl of 20× reaction buffer (20× buffer = 1.0 M Tris · HCl, pH 9.0; 400 mM ammonium sulfate; and 30 mM MgCl2; Promega, Madison, WI), 1 μl (1 unit) of Taq polymerase, 1.25 μl of 2 mM dATP, dGTP, dCTP, and dTTP (final concentration 20 μM; Promega), 1 μl of each primer (final concentration 20 μM), 70 μM DMSO (Sigma), and 29.5 μl of sterile deionized water in 500-μl Eppendorf tubes; each reaction was overlaid with 30 μl of light mineral oil (Sigma). The reaction mixture was first incubated at 94°C for 5 min and then subjected to 35 cycles, each consisting of denaturing (1.0 min, 93°C), annealing (1.0 min, 55°C), and extension (1.0 min, 72°C) using a programmable thermal cycler (M. J. Research, Watertown, MA). After the final cycle, the reaction mixture was incubated at 55°C for 2 min, then at 72°C for 5 min, and finally held at 4°C. The reaction products were analyzed on 1.5% agarose gels containing 0.5 μg/ml ethidium bromide in 1× buffer containing 10 mM Tris base, 10 mM boric acid, and 40 mM EDTA (pH 7.6) and blotted on nitrocellulose membranes as previously described (34).
Analysis of PCR products.
The PCR products obtained from amplifying the fetal lung cDNA samples using the NK2 receptor primers were subjected to restriction enzyme digest analysis. AvaII andHaeIII digests were performed according to the manufacturer's instructions (New England BioLabs, Beverly, MA). Restriction enzyme digest products were analyzed on 1.5% agarose gels to determine their size and then were transferred to nylon membranes and hybridized to an end-labeled cDNA probe specific for adult NK2 receptor, as described below. This probe was obtained from amplification of normal adult lung cDNA using the NK2 receptor primers, which resulted in the expected 689-bp product. This 689-bp product, corresponding to the region of interest in the NK2 receptor, was cloned using a TA cloning kit (Invitrogen, Carlsbad, CA), and its sequence was confirmed by direct cycle sequencing. The 689-bp product was then labeled with γ-ATP using a Boehringer Mannheim kit (Boehringer Mannheim). The probe (100 μl) was incubated with the Southern blot at 65°C for 2 h, and then the blot was washed first in 1× SSC (150 mM NaCl and 15 mM sodium citrate) with 0.1% SDS (Fisher Chemicals, Springfield, NJ) for 15 min at 65°C followed by 1× SSC at room temperature for 10 min. After the stringency washes, the blot was placed on film (Kodak, Rochester, NY) overnight at −80°C for autogradiography.
The relative expression of mRNA for the NK2 receptor vs. mRNA for actin was determined by densitometry analysis. The autoradiograms were scanned into a Hewlett-Packard Vectra personal computer (Hewlett-Packard, Palo Alto, CA) using a Microtek Scan Maker E6 model MRS3–1200E6 (Microtek International, Taiwan, China). Densitometry was performed using Media Cybernetics Gel-Pro Analyzer version 3.0 (Media Cybernetics, Silver Spring, MD).
Immunohistochemical analyses were performed in 37 samples with gestational ages between 10 and 23 wk using a modified avidin-biotin complex (ABC) technique (23). Rabbit polyclonal antibodies included affinity-purified anti-NKA used at 1:200 dilution (Accurate Chemical and Scientific, Westbury, NY), antisera to the neuronal and neuroendocrine cell marker protein gene product 9.5 (PGP9.5) used at 1:1,000 (Ultraclone, Isle of Wight, UK), antisera to the Clara cell marker CC10 used at 1:2,500 (generous gift from Dr. Gurmukh Singh, University of Pittsburgh, Pittsburgh, PA), and antisera to the epithelial cell marker cytokeratin used at 1:300 (DAKO, Carpinteria, CA). Murine monoclonal IgG1 antibodies included anti-NK2receptor (supplied by Krause) used at 1:50, the muscle marker anti-desmin used at 1:100 (DAKO), the neuroendocrine cell marker anti-chromogranin A used at 1:500 (Boehringer Mannheim), and the muscle marker anti-α-actin (clone HHF35) used at 1:30 (Enzo Diagnostics, Farmingdale, NY) dilution. Negative controls included substituting the primary antibody with the irrelevant murine IgG1 MOPC-21 (Sigma). Preabsorbance of the diluted primary antibody against 10–50 μg of specific peptide overnight at 4°C was used as an additional negative control for the NKA and NK2 receptor antibodies.
Briefly, paraffin-embedded lung tissue sections were dewaxed in xylenes and rehydrated in graded alcohols. Nonspecific immunoglobulin binding was blocked with 10% normal goat serum (GIBCO BRL, Gaithersburg, MD) for the rabbit antisera and with 10% normal horse serum (GIBCO BRL) for the murine antibodies and MOPC. For all samples except NKA, the primary antisera, diluted as above in PBS with 2% BSA, was applied to tissue sections and incubated at 4°C overnight in a humidified chamber. The primary antisera for NKA was diluted in Tris-buffered saline (TBS) with 2% BSA. The slides, except for the NKA samples, were then incubated with biotinylated IgG secondary antibody (goat anti-rabbit or horse anti-mouse depending on the primary antibody host; Vector Laboratories, Burlingame, CA) diluted 1:200 in 5% powdered milk in PBS with 5 μl/ml normal human serum, at 4°C for 2 h. The NKA slides were incubated in goat anti-rabbit antisera (Jackson Immunoresearch Laboratories, West Grove, PA) diluted 1:100 in TBS-BSA at room temperature for 1 h. Endogenous peroxidase activity for all slides was quenched using methanol containing 1% hydrogen peroxide. For all samples except NKA, ABC standard (Vector Laboratories) was made according to the manufacturer's instructions, applied to sections, and incubated at room temperature for 1 h. The NKA slides were incubated for 1 h at room temperature in peroxidase-conjugated rabbit anti-peroxidase (Jackson Immunoresearch) diluted 1:400 in TBS. For NKA, NK2 receptor, and anti-actin slides, additional amplification with tyramide (TSA Biotin System; NEN, Boston, MA) was performed. Biotinylated tyramide was diluted 1:50 in amplification buffer and incubated for 6 min and 30 s at room temperature. This was followed by streptavidin-horseradish peroxidase diluted 1:100 in PBS and incubation for 30 min at room temperature. Immunopositivity was visualized using the chromagen diaminobenzidine (0.025%) in PBS and 0.1% hydrogen peroxide. Immunostaining using the negative controls proceeded as described for the primary antibodies. All sections were counterstained with 2% methyl green (Sigma).
The immunostaining results were evaluated by an experienced reader (K. J. Haley) using a semiquantitative analysis. Cell types were identified by colocalizing immunostaining for the markers described above in serial slides having an average thickness of 3–5 μm. The results of this analysis were confirmed by a Brigham and Women's Hospital staff pathologist (M. E. Sunday).
The results of the tracheal perfusion assays are reported as means ± SE. Data were tested for normalcy, and one-way ANOVA was used to evaluate the differences in the change in the Pao. Differences were regarded as statistically significant at P < 0.05.
Tracheal perfusion assay.
We found that the NK2 receptor agonist NLE-10 was a highly effective contractile stimulus in our samples of midtrimester human lung. Indeed, 100 nmol of NLE-10 were more effective than either the NK1 receptor agonist [Sar9,Met(O2)11]- substance P (Sar-9) or 10 μmol of methacholine. Comparison of the contractile activity of the neurokinin receptors with the cholinergic receptors in human fetal lung was accomplished in this system by evaluating the changes in Pao after smooth muscle agonists were administered via tracheal perfusion. The use of changes in Pao to indicate the contractile state of the lung has been validated in prior investigations (32, 33, 39). The agonists used in the comparison studies (Fig. 1) included perfusion buffer, methacholine, the NK1 receptor agonist Sar-9 (13, 63), the NK2 receptor agonist NLE-10 (48), the diluent for NLE-10, and BaCl2. We found that NLE-10 caused the greatest amount of bronchoconstriction in the samples with Pao after buffer (n = 10; 0.05 ± 0.158 cmH2O) and Pao after NLE-10 (n = 8; 4.5 ± 1.79 cmH2O; P < 0.05). All samples demonstrated significantly increased Pao after BaCl2 (data not shown).
Our tracheal perfusion assay system also demonstrated that capsaicin administration significantly increased Pao compared with buffer. Intratracheal challenge (n = 11) with capsaicin resulted in significantly increased Pao of 2.72 ± 2.09 cmH2O compared with buffer (0.05 ± 0.158 cmH2O; P < 0.05). To determine whether the response to capsaicin was due to release of endogenous tachykinins activating the NK2 receptor, we studied samples treated with the NK2 receptor antagonist SR-48969 (2,27). SR-48969 is a competitive antagonist of the NK2receptor that binds the NK2 receptor with an affinity at least 1,000-fold greater than either the NK1 or NK3 receptor; SR-48969 is effective in human bronchial smooth muscle at the dose used in the tracheal perfusion assay [30 nmol (2)]. Addition of 30 nmol of SR-48969 (n = 5) to the perfusion buffer completely ablated the contractile activity associated with capsaicin (Fig.2).
Immunostaining in 23 lung samples with gestational ages ranging from 10 to 23 wk identified peribronchial immunopositive cells in all but one sample. The sole negative sample was of 12 wk gestational age. However, three additional samples with gestational age 11–12 wk demonstrated immunopositive cells in the airway peribronchial nerves. Representative immunostaining for NKA in lung with gestational age 19–22 wk is shown in Fig. 3,A and C. The immunostaining colocalized with the neural marker PGP9.5, consistent with peribronchial nerve immunopositivity (Fig. 3, B and D). The immunostaining was predominantly identified in the airway-associated nerves of primitive airways lined with cuboidal epithelium in samples with <18 wk gestation (n = 11); in older samples, NKA was identified in the peribronchial/peribronchiolar nerves of airways lined with either columnar or cuboidal epithelium (n = 12). The profusion of immunostaining increased with gestational age.
In samples at least 19 wk gestational age, perivascular nerves (n = 4) were also noted to be immunopositive (Fig.3 E). As with the nerve fibers associated with airways, the perivascular staining for NKA colocalized with the neural marker PGP9.5 (Fig. 3 F). Immunostaining was absent after the primary antibody was preabsorbed with NKA (data not shown). Thus NKA protein was identified in lung samples during the pseudoglandular and canalicular stages of lung development, which correlate with the time of branching morphogenesis (gestational age 10 through ∼14–16 wk). During this period, NKA protein localized to peribronchial nerves. In samples of at least 22 wk gestational age, NKA was identified in both peribronchial and perivascular nerves, consistent with the localization reported in human adult lung (9).
RT-PCR analysis was used to identify the mRNA for the NK2receptor. A single PCR product of the size expected for the NK2 receptor was detected in 19 of 22 samples. Restriction digest analysis with HaeIII and AvaII was used to evaluate the identity of these PCR products. The uncut PCR product size obtained from fetal lung was identical to that obtained from adult lung, 689 bp, and contained restriction sites for bothHaeIII and AvaII. The fragments resulting from restriction enzyme digest were of the sizes expected for the NK2 receptor (231 and 458 bp for AvaII and 524 and 165 bp for HaeIII). The uncut PCR products and the restriction digest fragments bound the radiolabeled NK2receptor probe (Fig. 4). Expression of mRNA for actin was detected in all samples (Fig.5, A and B). In specimens that had not been used in the tracheal perfusion assay, NK2 receptor mRNA was identified in 19 of 22 samples from gestational ages 10 to 22 wk (Fig. 5 A). All of the samples without detectable NK2 receptor mRNA were between 16 and 18 wk gestation. The pattern of NK2 receptor mRNA was the same with and without normalization for actin mRNA expression. In separate experiments, nine of the samples used in the tracheal perfusion assay were also evaluated for the presence of mRNA for NK2receptor, and all nine demonstrated abundant expression of mRNA for the NK2 receptor (Fig. 5 B). Because the samples used in the tracheal perfusion assay were evaluated separately from the other samples, the densitometry is not directly comparable between these experiments. The samples used in the tracheal perfusion assays were at least 18 wk gestational age and demonstrated abundant NK2 receptor mRNA.
Immunostaining for NK2 receptor protein was evaluated in 27 samples of human lung tissue with gestational age ranging between 10 and 23 wk and included 7 samples that had also been evaluated for NK2 receptor mRNA expression. All samples demonstrated immunopositive cells surrounding both large and small airways (Fig.6 A). Examination of serial sections with the smooth muscle marker desmin showed that the airway-associated NK2 receptor immunostaining colocalized with the smooth muscle marker desmin in all samples (Fig.6 B). In contrast to the mRNA expression pattern, the profusion of immunostaining consistently increased with increased gestational age. The immunopositive airways in samples of <16 wk gestational age (n = 7) were predominantly bronchioles with columnar epithelium (large bronchioles). In samples of at least 16 wk gestational age (n = 20), both large bronchioles and smaller airways demonstrated immunopositive cells, although the immunostaining continued to be more prominent in larger airways. This pattern of peribronchial and peribronchiolar localization is similar to that reported in adult human lung (7, 38). In addition, in these samples of immature lung, scattered loose mesenchymal cells were also positive for the NK2 receptor (Fig. 6 C). The greatest profusion of mesenchymal staining was in the samples of <16 wk gestational age and in all samples was identified in <1% of the mesenchymal cells.
In samples of at least 17 wk gestation, NK2 receptor immunostaining was demonstrated in epithelial cells (Fig.7 A). These cells were rare (<5% of epithelial cells in all samples) single epithelial cells located in cartilaginous bronchi and were not observed in the small membranous airways. Colocalization with keratin in serial sections confirmed the epithelial location of these cells (Fig. 7 B). The NK2 receptor-positive cells were not positive for the muscle marker desmin (data not shown). Furthermore, markers for two airway epithelial cell subtypes, Clara cells (CC10) and neuroendocrine cells (chromogranin A and PGP9.5), did not demonstrate colocalization with NK2 receptor immunostaining (data not shown).
Perivascular smooth muscle was positive for NK2 receptor in samples of at least 16 wk gestational age (n = 20). The staining was more prominent in thicker-walled vessels, consistent with arterioles, compared with thinner-walled vessels, consistent with pulmonary veins (Fig. 6 B), and colocalized with perivascular staining for desmin and α-actin in serial sections (data not shown). Immunostaining for the NK2 receptor was completely abolished after preincubation with NK2 receptor peptide (Fig. 6, E and F).
The NK2 receptor immunostaining therefore demonstrates abundant protein distribution during both the pseudoglandular and canalicular stages of lung development. The location of the receptor includes the peribronchial and perivascular smooth muscle, consistent with the adult distribution of the NK2 receptor in both guinea pigs (56, 64) and humans (7, 38). However, in the developing lung, additional sites of immunopositivity are noted in the loose mesenchyme and epithelial cells.
In this study, we report that the NKA-NK2receptor-ligand system is present during pulmonary organogenesis and transduces pulmonary constriction in tracheally perfused lungs more effectively than the cholinergic system in developing lung. We found that the mRNA and protein for the NK2 receptor and the protein for its most potent native ligand, NKA, are expressed during the pseudoglandular and canalicular stages of lung development. NK2 receptor protein expression was demonstrated throughout gestational ages 10–23 wk. The protein for NKA was abundant in all samples of >12 wk gestation, and samples of at least 22 wk gestational age demonstrated immunopositive cells in both the peribronchial and perivascular nerves. This is similar to the distribution reported by other investigators in human adult lung (9, 46, 53) where NKA has been identified in C fiber nerves (9, 46). The tracheal perfusion assay demonstrated greater contractile responses to the NK2 receptor agonist than the NK1 receptor agonist. In addition, a greater contractile response was observed after 100 nmol of the NK2 receptor agonist than after 10 μmol of methacholine. The potent contractile effects of the NK2agonist led us to examine the NKA-NK2 receptor system.
NK2 receptor protein, the preferred receptor for NKA, was identified in all samples of at least 10 wk gestational age and was located predominantly in the peribronchial and perivascular smooth muscle. The localization to airway smooth muscle is similar to that reported for adult human lung (7, 38). In our samples of developing lung, two unanticipated cell types were also positive for NK2 receptor protein. In samples of at least 18 wk gestation, NK2 receptor protein localized to rare (<5% in all samples) airway epithelial cells in cartilaginous airways. In addition, isolated loose mesenchymal cells were positive for NK2 receptor, with this staining being most prominent in the samples of <16 wk gestational age. The mesenchymal and epithelial NK2 receptor immunostaining suggests that mediators acting via the NK2 receptor, such as NKA, might contribute to the growth and/or maturation of these cell types. NKA has been demonstrated to be a growth factor and chemoattractant for cultured human bronchial epithelial cells (28) and human pulmonary fibroblasts (19). Our findings suggest that NKA might have similar roles in vivo.
Tracheal perfusion assays were used to define the functional capacity of the tachykinin receptors present in our samples. Because of the limited number of samples available for study, we chose to use substantial doses of agonists that had been previously demonstrated to be effective in causing bronchoconstriction in the tracheal perfusion model (31-33). In our studies, tracheal perfusion assay of capsaicin-challenged lungs demonstrated that the midgestational (i.e., canalicular stage) developing lung can release endogenous neuropeptides that activate neurokinin receptors on smooth muscle to cause bronchoconstriction. Furthermore, pretreatment with a specific NK2 receptor antagonist prevented the response to capsaicin, demonstrating that the effects of endogenous tachykinins in the immature lung are mediated predominantly through the NK2 receptor. The endogenous ligand that most closely matches the mediator identified by the tracheal perfusion studies is NKA. Thus the tracheal perfusion data imply that the presence of the NKA-NK2 receptor-ligand system is functional during lung development. The presence of both the NK2 receptor and NKA immunostaining as early as 11 wk gestation suggests that this system may be active even before 20 wk.
The greater potency (a lower ED50) of NKA compared with methacholine for causing bronchoconstriction in intact lungs is similar to that previously observed in isolated human neonatal bronchi (14). In isolated bronchial rings from neonates (age 1–60 days postnatal), NKA demonstrated the lowest ED50compared with carbachol, histamine, or KCl (14). In contrast, a greater maximal degree of constriction was observed from complete carbachol dose-response curves compared with partial dose-response curves for NKA (14). Our data indicate that canalicular stage peribronchial smooth muscle in the intact lung is more responsive to tachykinin stimulation. The greater potency (lower ED50) of NLE-10 compared with cholinergic stimulation observed in the present study is similar to that observed in adults (1).
There are at least two plausible mechanisms by which the NKA-NK2 receptor system could contribute to lung development. The first is by modulating the mechanical forces affecting the airway. Stimulation of the NK2 receptor via endogenous tachykinins results in peribronchial smooth muscle constriction, which alters airway stretch. Therefore, it is feasible that the NKA-NK2 receptor system could modify the mechanical forces impacting on the immature lung by altering airway stretch. The second mechanism would be by direct proliferative and/or maturational effects of NKA on the cells of the developing airway wall.
Several investigators have examined the effects of periodic stretch on pulmonary cells (11, 35, 36, 50, 51, 57, 61, 62). These effects include stimulating proliferation (11, 35), causing the release of growth factors capable of acting in a paracrine fashion on surrounding cells (36, 51, 61), enhancing responses to growth factors (58) and mediators such as parathyroid hormone-related peptide (57), altering the secretion of components of the extracellular matrix such as proteoglycans (61), and increasing the secretion of surfactant apoproteins (50). The effects of stretch depend on the maturational stage of the exposed cells; fetal rabbit type II epithelial cells demonstrated increased [3H]choline uptake after stretch, but these effects were decreased by exposure to fibroblast-conditioned medium, which promotes differentiation (51). Thus stretch has been demonstrated to be a mechanical force that modulates the lung through several mechanisms. Therefore, factors that alter stretch have the potential to affect lung development.
In addition to effects secondary to changes in airway stretch, the NKA-NK2 receptor system could have direct effects on the developing lung. NKA has been demonstrated to be a growth factor and chemoattractant for human pulmonary fibroblasts (19), human bronchial epithelial cells (28), and guinea pig tracheal epithelial cells (28, 60). NKA also induces rat aortic smooth muscle cell proliferation in culture (24,44). Our present observation of the location of the NK2 receptor in the epithelium and smooth muscle of human fetal lung suggests that such in vitro effects may also be present in vivo.
Our findings extend prior investigations regarding potential roles of the tachykinin receptors during lung development. Adult NK1receptor knockout mice have unremarkable pulmonary histology at baseline but have diminished inflammatory responses during immune complex-mediated alveolitis (12). These findings suggest that the NK1 receptor itself is not absolutely required for development of normal pulmonary morphology. These findings are consistent with our findings of a lower degree of contractile effect of an NK1 receptor agonist in tracheally perfused lung. The greater potency of the NK2 receptor suggests that it may be relevant to pulmonary development. The importance of the NK2 receptor for murine lung growth and/or maturation is not known. Many critical physiological processes employ redundant pathways that allow compensation in the absence of any one component. Although it is possible that the NKA-NK2 receptor system may have important roles in lung development, it is also possible that the absence of the NKA-NK2 receptor system could be compensated for by other factors. It is clear that endogenously produced tachykinins and the neurokinin receptors are present and functional in immature lung. These mediators are the most potent known components of a system producing mechanical forces that are crucial to normal lung development.
In conclusion, this investigation describes the ontogeny of the NK2 receptor and its most potent native ligand, NKA, in human lung. We have demonstrated that both NKA and NK2receptors are present during critical phases of pulmonary organogenesis and that the NK2 receptor transduces peribronchial smooth muscle constriction after at least 20 wk gestation. This study extends the results of other investigators by finding that the NK2receptor is the primary tachykinin receptor involved in peribronchial smooth muscle constriction in developing human lung. Because the tachykinins are present and their receptors are functional during pulmonary organogenesis, they may modulate lung development and/or maturation.
Address for reprint requests and other correspondence: C. M. Lilly, Respiratory and Critical Care Medicine, Brigham and Women's Hospital, 75 Francis St., Boston, MA 02115.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2001 the American Physiological Society