INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE RECENT DEMONSTRATION that gene-targeted disruption of the neurokinin-1 (NK-1) receptor (the main receptor for substance P) protects murine lungs from immune complex injury (5) has bolstered the notion of a tachykinin-based mechanism for neural amplification of immune inflammation in the lungs. The existence of a similar, but smaller-scale, mechanism had been suspected since the discovery that airway sensory fibers contain substance P and other tachykinins encoded by the preprotachykinin (PPT)-A gene and release them in response to nociceptive signals (48).

Sensory fibers, however, may not be the only source of tachykinins in the lungs. Substance P immunoreactivity has been observed in airway ganglia from several species, including humans (7-10, 44, 49). These observations have two intriguing implications. First, airway neurons may function as a tachykinin reservoir available even when the sensory innervation of the airways is disrupted (e.g., after lung transplantation). In addition, because airway ganglia are themselves innervated by peptidergic sensory fibers (42) and undergo membrane depolarization when exposed to substance P (43, 50), ganglion neurons are likely to contain NK-1 receptors and may even coexpress NK-1 receptor and PPT-A genes. Indeed, it is conceivable that the intrinsic neuronal network of the airways contributes to the amplification and propagation of immune-mediated inflammation in the lungs through a system of synaptic and parasynaptic peptidergic interactions, independent of the sensory nerves.

The present study was designed to characterize the expression of PPT-A and NK-1 receptor mRNAs and protein by intrinsic airway neurons in the rat. First, we applied immunohistochemical and mRNA detection techniques to elucidate whether PPT-A and NK-1 receptor gene products are expressed independently or are coexpressed by airway neuronal bodies. Second, we investigated whether the expression of these genes is altered by the removal of vagal and sensory inputs in tracheal xenografts, syngeneic lung grafts, and cultured parasympathetic ganglia. Finally, we took advantage of the requisite role of the NK-1 receptor for the injury produced by antigen-antibody complexes in the lung to determine whether nonsensory sources of substance P can support the development of this type of injury in syngeneic lung grafts.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All experiments were performed following protocols approved by the Washington University Animal Studies Committee. Rodents were housed in a climate-controlled room at 22°C with full access to a standard rat chow and water.

Preparation of trachea and tracheal xenografts. Tracheae were removed from 10- to 12-wk-old Sprague-Dawley rats (Charles River, Wilmington, MA) under pentobarbital sodium anesthesia (100 mg/kg ip). After separation from the esophagus, each trachea was divided into two halves, upper and lower, by a transverse coronal section at the level of the thoracic inlet. Each half was randomly assigned for immediate preservation as a control or for implantation as a xenograft into immunodeficient mice. The dissection of the trachea from the neighboring tissue was not designed to preserve the tracheal nerves or the longitudinal trunk parasympathetic ganglia attached to these nerves (2, 52). At least in the ferret, the majority of the neurons in these ganglia are cholinergic and lack substance P expression (7). The tracheal xenografts were prepared with a modification of a technique previously described (13). Immediately after removal, the selected tracheal segments were rinsed and soaked for 4 h at 4°C in antibiotic-antimycotic solution (Sigma, St. Louis, MO) reconstituted in PBS to concentrations of 100 U/ml penicillin, 100 µg/ml streptomycin, and 0.25 µg/ml amphotericin B. The segments were then implanted subcutaneously into male C57BL/6J nu/nu mice (Jackson Laboratory, Bar Harbor, ME) under pentobarbital sodium anesthesia (50 mg/kg ip). The xenografts were removed at 1 (n = 8), 3 (n = 5), 5 (n = 5), 7 (n = 7), 14 (n = 16), or 28 (n = 22) days after implantation and placed in fixative solution (4% paraformaldehyde or 2% paraformaldehyde, 0.1% glutaraldehyde, and 15% picric acid, both in 0.1 M sodium phosphate-buffered saline) for 24 h. Although some mucus accumulated in the tracheal lumen, the xenografts did not appear distended even after 28 days. Therefore, no provisions were made for continuous drainage.

Syngeneic lung transplantation. Lung transplantation was performed in 12-wk-old inbred Fischer rats (n = 16; Charles River). The trachea of the donor rat was cannulated, and the lungs were ventilated with a rodent ventilator (Harvard Apparatus, S. Natick, MA) under pentobarbital sodium anesthesia (50 mg/kg ip). The thoracic and abdominal organs were exposed through a midline incision. Heparin (1,000 U/kg) was injected into the inferior vena cava, which was then sectioned and allowed to bleed until exsanguination. The pulmonary vessels were flushed with cold saline through a pulmonary artery cannula, and the lungs were removed into a cold saline bath. The recipient rat was anesthetized with halothane (0.5-3%) piped into an induction box or, after cannulation of the trachea, into the inspiratory limb of the rodent ventilator. After a lateral thoracotomy and left pneumonectomy, the pulmonary artery and vein were attached to the corresponding vessels of the donor rat's lung by use of polyethylene cuffs (41, 47). The left main stem bronchus was anastomosed with an 8-0 Prolene running suture. The thoracotomy was closed by layers, with care taken to remove all air remaining in the pleural cavity through a silicone rubber catheter. On recovery from anesthesia, the pleural catheter was removed, and after a short period of observation, the rat was transferred to the rodent holding facility without supplemental oxygen.

Immune complex injury. The objective of the experiments was to determine whether the destruction of peptidergic sensory nerve fibers during syngeneic lung transplantation prevents the development of an immune complex injury. In mice, it is known that this kind of injury is attenuated by specific pharmacological NK-1 receptor antagonists (24) and prevented by disruption of the NK-1 receptor gene (5). Lacking similar information in other species, we investigated first whether the NK-1 receptor plays a similar conditional role in the rat. To this effect, we compared the responses of five rats pretreated with the NK-1 receptor antagonist LY-306740 (38) (a gift of Lilly Research Laboratories, Indianapolis, IN; 30 mg/kg ip in 2 ml of normal saline) and six rats pretreated with saline vehicle alone to the simultaneous injection of chicken ovalbumin intravenously (20 mg/kg in 2 ml of normal saline; Sigma) and a rabbit polyclonal antibody against ovalbumin intratracheally (0.9 mg/kg in 1 ml of normal saline; Chemicon, Temecula, CA). The rats were injected 2.5 h later with FITC-labeled albumin (5 mg/kg iv; Sigma), and after an additional 1.5 h, the lungs were exposed through a median sternotomy under pentobarbital sodium anesthesia. A clip was placed on the left main stem bronchus, and the left lung was removed for analysis. The right lung was then lavaged three times, each with 2 ml of normal saline. Disruption of the permeability of the alveolar-capillary membrane was quantified as the ratio of lavage fluid to serum FITC albumin concentrations measured spectrofluorometrically (excitation = 495 nm, emission = 520 nm; model SPF 500, American Instruments, Silver Spring, MD).

Primary culture of airway neurons. Neurons were cultured from tracheae of 10- to 12-wk-old Sprague-Dawley rats as previously described (14, 17). Each culture combined the tracheae from four rats. The posterior quadrant of each trachea was separated through bilateral longitudinal incisions at the level of the trachealis muscle's attachments to the tracheal cartilage. The resultant tissue strips were denuded of epithelium and digested for 30-45 min by placement in two consecutive 37°C baths of sterile low-calcium Krebs-Henseleit solution with 2.5 mg/ml BSA (Sigma), 3 mg/ml (1st bath) or 1.5 mg/ml (2nd bath) collagenase I (Sigma), and 0.15 mg/ml (1st bath) or 0.50 mg/ml (2nd bath) elastase IV (Sigma). The digestion product was washed with a solution containing 10% FCS, 100 U/ml penicillin, 100 µg/ml streptomycin, and 0.25 µg/ml amphotericin B (Sigma) in DMEM-F-12 (equal volumes of DMEM and Ham's F-12 medium) and centrifuged twice (1,000 rpm for 10 min at 4°C). After cell clumps were broken with a heat-smoothed glass pipette, the cells were seeded onto Falcon Primaria plates (Becton Dickinson, Lincoln Park, NJ) and incubated overnight at 37°C in 5% CO₂. On the following day, the plates were rinsed with serum-free medium to remove nonadhered cells. The cell suspension was washed twice in fresh serum-free medium and centrifuged. The cell pellet was then resuspended in DMEM-F-12 containing antibiotics (see above for concentrations), 100 ng/ml -nerve growth factor (Harlan Bioproducts for Science, Indianapolis, IN), 1 µg/ml bovine insulin (Sigma), 20 µg/ml human transferrin (Jackson ImmunoResearch Laboratories, West Grove, PA), 20 µg/ml glutamine (Sigma), and 30 nM sodium selenite (Sigma) and seeded on Matrigel-coated round coverslips (Becton Dickinson, Bedford, MA). (Some of the cultures were performed in the absence of -nerve growth factor.) Cytosine arabinoside (1 µM; Sigma) was added 24 h later. The medium was replenished every 48 h for the duration of the culture (7-14 days). Neuronal cultures were preserved by 30 min of fixation in solutions containing 4% paraformaldehyde and 0.25% glutaraldehyde (immunohistochemistry) or 2% paraformaldehyde, 0.1% glutaraldehyde, and 15% picric acid (in situ hybridization).

Immunohistochemical analysis. Paraffin-embedded tissue sections (5 µm) were deparaffinized with xylene and hydrated in incremental concentrations of ethanol. Tissue sections and neuronal cultures were permeabilized with 0.3% Triton X-100 for 15 min and treated with 0.1% hyaluronidase in 0.1 M sodium acetate buffer for 30 min at 37°C. Tissue sections and cultures were then exposed to donkey serum (Sigma) diluted 1:50 in Tris-buffered saline (TBS) for 30 min and incubated overnight at 4°C with the specific primary antisera (Table 1). The samples were then washed thoroughly and placed in a 1:50 solution of the specific fluorescent-labeled (FITC, tetramethylrhodamine isothiocyanate, or aminomethylcoumarin acetate) or biotinylated anti-IgG antiserum for 3-4 h at room temperature. All anti-IgG antisera were raised in donkey and purchased from Jackson ImmunoResearch Laboratories. After a final wash, each sample was mounted on a glass slide and protected with a sealed coverslip. The specificity of substance P and NK-1 receptor antibodies was confirmed by showing no staining in samples chosen at random for each processing batch when 10 µM immunizing peptide was added to the primary antiserum solution (1). The specificity of the macrophage metalloelastase antisera was corroborated by the absence of staining when the primary antiserum was replaced with preimmune serum (46). Biotinylated antibody labeling was demonstrated by incubation with an avidin-horseradish peroxidase complex (Vectastain ABC kit, Vector Laboratories, Burlingame, CA) followed by staining with 0.05% diaminobenzidine tetrahydrochloride and gold intensification. Macrophage metalloelastase-immunoreactive cells were counted at ×200 magnification after lung tissue sections were divided into eight equal sectors, and two of them were selected randomly. Cell density was calculated by dividing the total number of immunoreactive cells in each sector by the sector's area measured with a 10 × 10 40-µm-square reticle.

In situ hybridization. Samples (tissue slides or culture coverslips) were prepared for hybridization as previously described (6). Digoxigenin-labeled cRNA probes were prepared for the detection of PPT-A and NK-1 receptor gene transcripts with a commercial digoxigenin-U labeling kit (Roche Molecular Biochemicals, Indianapolis, IN). PPT-A probes were derived from a cDNA cloned into the BamH I site of pGEM1 [pG1-b-PPT (36)]. The antisense probe was transcribed with T7 RNA polymerase (Promega Biotec, Madison, WI) after the plasmid was linearized with Hind III (Promega Biotec). This resulted in a cRNA complementary to a 508-bp fragment of the -PPT mRNA, including the 75-bp 5'-untranslated region, the entire protein-coding region, and a 43-bp fragment of the 3'-untranslated region (36). A sense -PPT probe was transcribed with SP6 RNA polymerase (Promega Biotec) to provide a control for the specificity of the antisense probe. NK-1 receptor probes were derived from cDNAs inserted into the Hinc II site of pBluescript (21). Antisense probes complementary of the 5' (nt 1 to 642)- and 3'-coding regions (nt 637 to 1224) of the NK-1 receptor mRNA were transcribed with T3 RNA polymerase (Promega Biotec) after the plasmid was linearized with Xho I (Promega Biotec) and Xba I (GIBCO BRL, Life Technologies, Gaithersburg, MD), respectively. Sense NK-1 receptor probes were transcribed with T7 RNA polymerase after the plasmid was linearized with Xba I and Xho I. Hybridization was carried out overnight by exposing the samples to a solution containing 2.5 µg of probe per milliliter at 42°C. Each sample was then washed with saline-sodium citrate buffer for 15 min, incubated with 20 µg/ml RNase A for 30 min at 37°C to remove nonhybridized probe, and washed again in saline-sodium citrate buffer. Digoxigenin-labeled probe was detected by enzyme-linked immunoassay followed by an overnight color reaction (BCIP, Genius 3 kit, Roche Molecular Biochemicals) according to the recommendations of the manufacturer. The samples were counterstained with eosin and sealed with coverslips.

RT-PCR amplification of PPT transcripts. Total RNA was extracted from whole lung, tracheal tissue, or neuronal cultures, as appropriate, with RNAzol B (Tel-Test, Friendswood, TX). Single-stranded DNA was synthesized from 2 µg of RNA with use of dT priming and Moloney murine leukemia virus RT (GIBCO BRL). PCR reactions contained 6 µl of 10× reaction buffer, 30 pmol of each primer in 1 µl of solution, 2 µl of 5 mM deoxynucleotide triphosphate mixture, 1 U of Taq polymerase, and 66 µl of sterile water. Amplification was conducted in a programmable thermal controller (model PTC-100, MJ Research, Watertown, MA) and involved denaturation at 96 and 94°C for 3 min and 30 s, respectively, annealing at 57°C for 1 min, and extension at 72°C for 4 min for a total of 25 cycles. Primers for RT-PCR of PPT-A transcripts were as follows: sense 5'-ATGCCCGAGCCCTTTGAGC-3' (exon 2) and antisense 5'-ATTGCGCTTTCATAAGCC-3' (exon 7). These primers allowed discrimination of the three alternatively spliced transcripts of the PPT-A gene by the size of the PCR products: 181-bp -PPT, 235-bp -PPT, and 190-bp -PPT. As a control for the stability of the neuronal populations present in the tracheal xenografts, we performed RT-PCR of the microtubule-associated protein 2 (MAP-2), a constitutive protein expressed exclusively in the axon, dendrites, and cell bodies of neurons (23). The primers were as follows: sense 5'-ACGAAGGAAAGGCACCACACT-3' and antisense 5'-GTATCTGAATAGGTGCCCTGT-3'. The 148-bp product was sequenced to confirm its identity with the corresponding segment of the rat MAP-2 gene. All PCR products were electrophoresed on 1.5% agarose gels and stained with ethidium bromide. Southern blotting was performed with ³²P-labeled antisense probes: pG1--PPT and a probe generated by cloning the MAP-2 PCR product into a pCR-TOPO vector (Invitrogen, San Diego, CA).

Casein zymograms and Western blotting. Immediately after collection, lung lavage fluid was separated from suspended cells by centrifugation at 4°C and frozen at 70°C. Casein zymograms were performed as described previously (12). For Western blot analysis, 30 µl of lavage fluid were loaded on a 10% SDS-polyacrylamide gel and electrophoresed. The gel was then transferred to a polyvinylidene difluoride membrane, which was placed in a blocking solution of 3% powdered milk in TBS-Tween for 1 h. The membrane was incubated in a solution of primary antibody (Table 1) in equal parts of TBS-Tween and 3% powdered milk for 1 h at room temperature and washed thoroughly. Antibody binding was detected by incubation for 45 min with a horseradish peroxidase-conjugated secondary antibody (donkey anti-rabbit, 1:2,000) at room temperature and development by chemiluminescence (Amersham Pharmacia Biotech, Piscataway, NJ).

RIA. Substance P RIA was performed as previously described (37). Bolton-Hunter-iodinated substance P (2,000 Ci/mmol) was purchased from NEN Life Sciences (Boston, MA). The anti-substance P antiserum R5 (30) was used at a dilution of 1:400,000. Samples were assayed in duplicate at multiple dilutions and compared with synthetic substance P standards. The linear range of the standard curve was 25-300 pg/ml.

Data analysis. Southern blots of the RT-PCR products were processed in a Storm phosphor screen and analyzed for pixel intensity with ImageQuant (Molecular Dynamics, Sunnyvale, CA). Pixel intensities or, when appropriate, their ratios were compared by Student's t-test or Wilcoxon signed rank test (depending on whether the data were or were not normally distributed) for single comparisons or by ANOVA with replication for multiple comparisons (effects of time on mRNA levels in tracheal xenografts).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Baseline expression of PPT-A and NK-1 receptor mRNA and protein by airway ganglia. Both -PPT and NK-1 receptor mRNAs were detected by in situ hybridization in tracheal superficial muscular plexus neurons, peribronchial neurons, and occasionally in longitudinal trunk ganglia (Fig. 1). PPT-A and NK-1 receptor proteins were also demonstrated by immunohistochemistry, coexisting in neuronal bodies identified by neurofilament M or protein gene product 9.5 immunoreactivity in the same locations (Fig. 2). -PPT mRNA staining was similar between rats and, within each rat, among all the stained neurons. In contrast, the intensity of immunostaining for PPT-A protein (and presumably the individual cellular levels of the peptides) varied substantially. In some rats, virtually all the neurofilament M-stained cells in the area of the muscular superficial plexus were immunoreactive for PPT-A protein (Fig. 2). In other rats, only a portion of the cells contained detectable PPT-A protein. We found no discrepancies in the immunofluorescent patterns generated in adjacent slides by three antisera of different specificity against PPT-A-encoded peptides (Table 1). We confirmed the specificity of double-fluorescent staining for -PPT and NK-1 receptor proteins by the disappearance of the appropriate fluorescent signal in the samples pretreated with immunizing antigens. We used these controls and the differences in the intracellular distribution of the fluorescent signals to distinguish specific antibody labeling from bleed-through and autofluorescence artifacts.

View larger version (152K): [in this window] [in a new window]   Fig. 1.   A: preprotachykinin (PPT)-A gene expression in rat tracheal parasympathetic ganglia demonstrated by in situ hybridization with a -PPT antisense digoxigenin-labeled cRNA probe. PPT-A-expressing neurons were located in close relationship to trachealis muscle. B: PPT-A gene expression in isolated intrapulmonary peribronchial ganglia (left) and in a cluster of ganglion cells near tracheal bifurcation (right). C: neurokinin-1 (NK-1) receptor mRNA expression in posterior tracheal membrane ganglia demonstrated by in situ hybridization with an antisense digoxigenin-labeled cRNA probe. A and C: areas enclosed by squares are shown enlarged on right. Scale bars, 25 µm.

View larger version (86K): [in this window] [in a new window]   Fig. 2.   Coexpression of substance P and NK-1 receptor in tracheal superficial muscular ganglia demonstrated by immunohistochemical staining of rat's posterior tracheal membrane with antisera against neurofilament M (A, aminomethylcoumarin acetate-conjugated secondary antibody), -PPT [B and inset, tetramethylrhodamine isothiocyanate (TRITC)-conjugated secondary antibody], and NK-1 receptor (C, FITC-conjugated secondary antibody). With few exceptions (arrowhead), neurofilament M-immunoreactive cell bodies contained -PPT and NK-1 receptor proteins. On high magnification (B, inset), projections from tachykinin-immunoreactive neurons were thin and disappeared after a short distance into tissue. TM, trachealis muscle; arrow, longitudinal nerve fibers.

Expression of PPT-A and NK-1 receptor after denervation. PPT-A and NK-1 receptor expression persisted after the airway ganglia were separated from their motor and sensory inputs, independent of whether this separation was accomplished by grafting tracheal segments into nu/nu mice, by syngeneic lung transplantation (Fig. 3), or by isolation of the ganglion neurons in primary culture (Fig. 4). Substance P-immunoreactive fibers, which could be found easily at high-power magnification in the subepithelial region of control tracheae and bronchi, were absent from the tracheal xenografts and the bronchi contained in the syngeneic lung grafts. Substance P immunoreactivity was apparent, however, in the projections from cultured neurons, often colocalizing with synaptophysin immunoreactivity (Fig. 4). Neuronal NK-1 receptors were located predominantly in perikarya in tissues and culture. Similar to Dey et al. (7), we found that only a minority of the ganglia in the superficial muscular plexus of the trachea were immunoreactive for choline acetyltransferase. These neurons appeared to retain their cholinergic phenotype in the tracheal xenografts.

View larger version (132K): [in this window] [in a new window]   Fig. 3.   Preservation of PPT-A and NK-1 receptor mRNA expression by denervated parasympathetic ganglia. A and B: despite disruption of tissue structure after 4 wk of subcutaneous implantation, in situ hybridization of rat tracheal xenograft with antisense digoxigenin-labeled cRNA probes shows cell bodies containing -PPT (A) and NK-1 receptor mRNAs (B) in proximity of tracheal adventitia. Insets: enlargements of area enclosed by squares. C: coexpression of substance P and NK-1 receptor by airway neurons in tracheal xenograft 4 wk after implantation demonstrated by triple immunohistochemical staining with antisera against neurofilament M (left), -PPT (middle), and NK-1 receptor (right). TM, trachealis muscle; arrowheads, ganglion neurons. D: substance P immunoreactivity (right, R140 antiserum, TRITC-conjugated secondary antibody) in ganglia identified by neurofilament M staining (left, FITC-conjugated secondary antibody) in vicinity of bronchus (Aw) and pulmonary vessel (V) in syngeneic left lung graft (coronal cut at level of hilum).

View larger version (85K): [in this window] [in a new window]   Fig. 4.   A: expression of PPT-A gene in cluster formed by cultured rat airway ganglia demonstrated by in situ hybridization with -PPT antisense digoxigenin-labeled cRNA probe. -PPT expression levels varied considerably among different neurons in culture, suggesting phenotypical differences. B: immunostaining of cultured ganglia denoting substance P presence in neuronal projections. C: substance P and synaptophysin colocalized in areas where cultured nerve fibers came in contact, suggesting that substance P is released at synaptic sites. D: coexpression of substance P (left, R140 antiserum, TRITC-conjugated secondary antibody) and NK-1 receptor (right, FITC-conjugated secondary antibody) in rat tracheal neurons in primary culture. Substance P was present in cell bodies and nerve fibers, but NK-1 receptor immunoreactivity was more prominent on cell bodies.

Cell populations expressing PPT-A and NK-1 receptor genes. PPT-A gene-encoded mRNA and peptides were detected only in neurofilament M or protein gene product 9.5-immunoreactive cells. NK-1 receptor immunoreactivity was present frequently in macrophages located in the wall of the tracheal xenografts or in the alveolar spaces of the lung (Fig. 5).

View larger version (66K): [in this window] [in a new window]   Fig. 5.   NK-1 receptor immunoreactivity in macrophages infiltrating wall of a tracheal xenograft. A, B, and C: simultaneous immunostaining of tracheal xenograft for neurofilament M (aminomethylcoumarin acetate-conjugated secondary antibody), -PPT (TRITC-conjugated secondary antibody), and NK-1 receptors (FITC-conjugated secondary antibody; enlarged in inset), respectively. D: cells immunoreactive for NK-1 receptors identified as macrophages with a mouse-derived anti-rat macrophage antibody in an adjacent tissue section stained with diaminobenzidine tetrahydrochloride.

Analysis of PPT mRNAs by RT-PCR. Tracheal xenografts and lung syngeneic grafts contained greater amounts of - and -PPT mRNA than matching tracheal segments and contralateral lungs, respectively. In the tracheal xenografts, the increase in PPT mRNAs reached its peak at ~3 days and became attenuated thereafter (Fig. 6). MAP-2 mRNA levels decreased uniformly after implantation, suggesting neuronal loss. In the lung grafts, the increase in PPT mRNA was still detectable 2 wk after transplantation (Fig. 7). PPT mRNA levels were similar after intratracheal instillation of anti-ovalbumin antibodies (10 ± 2% of brain control) or normal saline alone (21 ± 4%; P = 0.40 by ANOVA) into rats previously injected intravenously with ovalbumin. This finding suggests that, at least during the 4-h duration of the experiments, immune complex inflammation does not upregulate local substance P production or attract foreign substance P-producing cells into the lungs. There was no difference in PPT mRNA levels between the left and right lungs of the six control rats subjected to a left thoracotomy. MAP-2 mRNA was not affected by lung transplantation or thoracotomy alone. RT-PCR products from -PPT were ~2.5 times more abundant than those from -PPT in trachea and lung. This ratio was not altered in the lung grafts.

View larger version (22K): [in this window] [in a new window]   Fig. 6.   A: upregulation of PPT-A gene transcription in rat tracheal xenografts after implantation into nu/nu mice detected by Southern analysis of RT-PCR product after hybridization with 32P-labeled cRNA probe. Only bands corresponding to - (235 bp) and -PPT (190 bp) mRNA amplification products were identified in tracheal tissue. B: microtubule-associated protein 2 (MAP-2) mRNA, which was used as an internal RT-PCR control, decreased in many tracheal xenografts, suggesting neuronal loss after implantation. C: combined - and -PPT mRNA levels normalized to MAP-2 mRNA were higher in tracheal xenografts than in matched tracheal segments preserved at time of implantation (n = 8 at 1 day, 5 at 3 days, 5 at 5 days, 7 at 7 days, and 7 at 14 days; error bars, SE; P = 0.025, xenografts vs. corresponding control segments, by 2-factor ANOVA with repeated measures). Upregulation of PPT in xenografts reached its maximum 3 days after implantation.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   A: PPT mRNA detected by Southern blotting of RT-PCR products was more abundant in syngeneic lung graft than in contralateral native lung (control), lungs injured by injection of ovalbumin intravenously and anti-ovalbumin antibodies intratracheally (Ag + Ab), or lungs from rat injected with ovalbumin intravenously and normal saline intratracheally (Ag + NS). B: combined - and -PPT mRNA levels (normalized to brain - and -PPT mRNA) were higher in 4 of 5 syngeneic lung grafts than in native contralateral lungs after immune complex injury (P = 0.08, by Wilcoxon signed rank test).

Effects of NK-1 receptor antagonist on immune complex-mediated injury in rat lungs. Intravenous injection of ovalbumin followed by intratracheal instillation of anti-ovalbumin antibodies caused a diffuse lung injury characterized by edema, patchy hemorrhage, and infiltration with polymorphonuclear leukocytes and macrophages. Casein zymograms of the lavage fluid revealed a band of casein lysis immediately above the 20-kDa marker, which was identified by Western blotting analysis as corresponding to the activated forms of macrophage metalloelastase and matrilysin. Pretreatment with LY-306740 inhibited the activation of both proteases but had only a limited effect on the increase in alveolar-capillary permeability observed after the immune complex injury (Fig. 8).

View larger version (41K): [in this window] [in a new window]   Fig. 8.   A: pretreatment with NK-1 antagonist LY-306740 (LY) mitigated increase in alveolar-capillary permeability (measured as percent lavage fluid-to-serum ratio of FITC-labeled albumin) in rats injected with ovalbumin intravenously and anti-ovalbumin antibodies intratracheally (error bars, SE; P = 0.03, by unpaired t-test). B: LY-306740 reduced amount of macrophage metalloelastase (top blot) and matrilysin (bottom blot) detected in lavage fluid by Western blot analysis. Nos. on left, molecular mass.

Immune complex-mediated injury in syngeneic lung grafts. There were no differences in respiratory rate, activity, or food consumption between syngeneic lung recipients and control rats before the administration of ovalbumin and anti-ovalbumin antibodies. After immune complex formation, however, the transplanted rats developed greater respiratory difficulty than the control rats. Five of 12 lung graft recipients untreated with LY-306740 died after developing chest wall retractions and a gasping pattern of breathing. None died in the control group during the same period (² = 5.9; P = 0.02). At postmortem examination, none of the lung grafts had signs of vascular or bronchial obstruction.

View larger version (72K): [in this window] [in a new window]   Fig. 9.   A: casein zymograms showing a band of increased protease activity immediately above 20-kDa (molecular-mass) marker after immune complex injury. Band was present after injections of ovalbumin and anti-ovalbumin antibodies in lungs of control (Ag + Ab) and transplanted rats (Tx: Ag + Ab; G, grafted lung) and absent after injections of ovalbumin intravenously and normal saline intratracheally (Ag + NS). B: macrophage metalloelastase (top blot) and activated matrilysin (bottom blot) demonstrated by Western blot at similar concentrations in lavage fluid obtained from native lungs and syngeneic lung grafts after immune complex injury.

View larger version (142K): [in this window] [in a new window]   Fig. 10.   Occasional macrophages immunoreactive for macrophage metalloelastase were found in lungs of rats injected with ovalbumin intravenously and normal saline intratracheally (A and B). Density of macrophages increased in response to formation of ovalbumin immune complexes (C and D). Response was more severe in syngeneic lung grafts, which often contained areas of immunostaining in their matrix (E and F). Pretreatment with LY-306740 attenuated metalloelastase immunoreactivity in macrophages and matrix (G and H). B, D, F, and H, enlargements of A, C, E, and G, respectively. All samples are from left lungs.

Substance P concentrations in lung lavage fluid. Equivalent amounts of substance P (measured by RIA) were recovered in the lavage fluid from lung grafts (median substance P recovered: 350 pg; 10th and 90th percentiles: 231 and 471 pg, respectively) and control lungs (median: 360 pg; 10th and 90th percentiles: 251 and 1,350 pg) after immune complex formation. Substance P was usually undetectable in lavage fluid from rats for which the intratracheal antibody was replaced with normal saline.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Overwhelming evidence has accumulated in recent years placing the tachykinins and their receptors at the center of a mechanism by which the nervous system can influence the course of inflammation in the lungs (3, 5). The prevailing understanding of this mechanism is that substance P and other products of the PPT-A gene [which are often referred to as "sensory neuropeptides" (48)] are released exclusively by lung sensory fibers in response to inflammatory mediators, mechanical and chemical irritation, and antidromic stimulation of the vagus nerve. Here, we demonstrate that airway intrinsic neurons not only contain PPT mRNA and mature substance P but also express the NK-1 receptor. In addition, we present evidence that, lacking other apparent tachykinin sources, peptidergic airway neurons can support the development of an injury that requires activation of the NK-1 receptor. These observations challenge the concept that sensory fibers are the only source of PPT-A-derived peptides in the airways. Moreover, they suggest that tachykinins act as messengers in a neuronal-based system for the amplification of inflammation in the lungs.

PPT-A and NK-1 receptor expression by parasympathetic ganglia. The presence of substance P in airway ganglia has been refuted (32, 35) and affirmed (8, 9, 44) in multiple studies involving several species. The controversy was clarified substantially after a detailed quantitative study by Dey et al. (7) demonstrated that a large proportion of the neurons in the superficial muscular plexus of the ferret's trachea are immunoreactive for substance P. The present results confirm the findings of these investigators by showing the presence of PPT-A mRNA and protein products in the airway ganglia of the same plexus in the rat.

Upregulation of PPT-A gene expression after denervation. Tracheal and pulmonary PPT mRNA content increased after these tissues were implanted into nu/nu mice and syngeneic rats, respectively. Although alveolar macrophages and intestinal eosinophils can produce substance P (29, 39), we did not detect -PPT mRNA or substance P in the cellular infiltrates found in tracheal xenografts and injured lungs. Furthermore, there was no rise in PPT mRNA after the intense infiltration of control (nongrafted) lungs by inflammatory cells after immune complex injury. Thus we ascribe the increase in PPT transcripts to a neuron-specific response to the grafting process.

Substance P-dependent inflammation in syngeneic lung grafts. There is evidence that activation of the NK-1 receptors is necessary for the inflammatory response produced by the deposition of complexes between a soluble antigen and preformed antibodies in the airways (Arthus reaction). Gene-targeted disruption of the receptor protects mice from this response (5). Pretreatment with NK-1 receptor antagonists reduces the presence of inflammatory cells at the peak of inflammation in murine lungs by 18-64%, depending on cell type (24). In our study, LY-306740 markedly reduced the induction of lung metalloproteases after immune complex injury but had only a limited (30-40%) protective effect against the increase in alveolar-capillary permeability produced by the same injury. These findings confirm our hypothesis that NK-1 receptor stimulation by tachykinins contributes to the Arthus reaction in the rat's lungs. They also suggest the existence of alternative pathways for the immune complex-induced disruption of alveolar-capillary permeability in this species, perhaps involving other tachykinin receptors.

Summary. In this study, we demonstrate the existence of an intrinsic population of airway neurons that contain PPT-A-derived tachykinins and NK-1 receptors. In addition, we provide evidence suggesting that these neurons participate in the amplification of immune complex-mediated inflammation in the lungs. It is tempting to speculate that, by a similar mechanism, denervated neurons contribute to local airway inflammation in the course of disease processes such as bronchiolitis obliterans after lung transplantation.