rcmb%2e2002-0110oc

Neurotrophin and Neurotrophin Receptor Protein Expression in theHuman LungAlberto Ricci, Laura Felici, Salvatore Mariotta, Francesco Mannino, Giovanni Schmid, Claudio Terzano,Giuseppe Cardillo, Francesco Amenta, and Elena Bronzetti

Dipartimento di Scienze Cardiovascolari e Respiratorie, Universita` di Roma La Sapienza, Rome; Unita` Operativa di ChirurgiaToracica, Ospedale Carlo Forlanini, Rome; and Dipartimento di Farmacologia e Medicina Sperimentale, Universita` di Camerino,Camerino, Italy

Neurotrophins (NTs) promote survival and differentiation ofcentral and peripheral neurons, and display several activitiesalso in non-neuronal cells. Human lungs synthesize and releaseNTs, which are probably involved in the pathophysiology ofpulmonary disturbances. In this article the expression and ana-tomic localization of nerve growth factor, brain-derived neuro-trophic factor, and NT-3 and of corresponding high-affinityreceptors TrkA, TrkB (full-length and truncated [TR-] isoforms),TrkC, and of the low-affinity p75 receptor, were assessed in sur-gical samples from adult human lung by reverse transcriptasepolymerase chain reaction, Western blot, and immunohisto-chemistry. NTs and their cognate receptor mRNA and proteintranscripts were detected by reverse transcriptasepolymerasechain reaction and immunoblotting, respectively, nerve growthfactor (NGF) and brain-derived neurotrophic factor (BDNF)mRNA and corresponding protein transcripts being the mostexpressed. High levels of TrkB-[TR-] mRNA and of its proteintranscript were also demonstrated, whereas a low expressionof p75 mRNA and of corresponding protein transcript werefound. Microanatomic analysis of immunohistochemical studyrevealed that bronchial epithelial cells were immunoreactivefor different NTs, with a higher intensity of BDNF immune stain-ing compared with other NTs, but did not express NT receptorimmunoreactivity. Alveolar cells were immunoreactive for TrkAand TrkC receptor protein, but did not display immunoreactivityfor NTs or other receptors investigated. Gland cells expressedNT and high-affinity NT receptor immunoreactivity, but notp75 receptor immunoreactivity. NT and low-affinity receptorimmunoreactivity was observed within neurons and satellitecells of parasympathetic ganglia as well as in nerve fiberlikestructures supplying the bronchopulmonary tree. An obviousimmunoreactivity for NTs and NT receptor protein was alsoobserved in intrapulmonary branches of pulmonary artery. Pul-monary lymphocytes and macrophages express nerve growthfactor and high-affinity NT receptor immunoreactivity. The roleof NTs in non-neuronal tissue including lung has not been clari-fied yet. The widespread expression of NTs and their receptorsin different components of the lung suggests that these factorsmay contribute to regulate cell function in human lung.

(Received in original form July 10, 2002 and in revised form May 13, 2003)

Address correspondence to: Alberto Ricci, M.D., Dipartimento di ScienzeCardiovascolari e Respiratorie, Universita` di Roma La Sapienza, OspedaleSantAndrea, Via di Grottarossa, 10351039, 00189 Roma, Italy. E-mail:[email protected]

Abbreviations: bronchus-associated lymphoid tissue, BALT; brain-derivedneurotrophic factor, BDNF; nerve growth factor, NGF; neurotrophins, NTs;reverse transcriptasepolymerase chain reaction, RT-PCR.Am. J. Respir. Cell Mol. Biol. Vol. 30, pp. 1219, 2004Originally Published in Press as DOI: 10.1165/rcmb.2002-0110OC on June 5, 2003Internet address: www.atsjournals.org

Neurotrophins (NTs) are a family of polypeptide growthfactors, including nerve growth factor (NGF), brain-derivedneurotrophic factor (BDNF), NT-3, NT-4/5, and NT-6 (1).They bind two types of cell surface receptors characterizedby different binding affinities and molecular weight. High-affinity NT receptors have a molecular weight of 140145kD and display tyrosine-specific protein kinase (Trk) activ-ity. Low-affinity NT receptor (p75) is a glycoprotein recep-tor, with a molecular weight of 75 kD (1). NGF recognizesspecifically the TrkA receptor. BDNF and NT-4/5 acti-vate TrkB receptor, whereas NT-3 activates primarily theTrkC receptor, and to a lesser extent the TrkA and TrkBreceptors (2).

The physiologic role of NTs includes promoting differen-tiation and survival of developing neurons in the centraland peripheral nervous system (3). More recently, it hasbeen shown that NTs stimulate differentiation and prolifer-ation of cell types from all three germ layers (2, 4). Trkgenes are expressed by a variety of non-neuronal tissues(5), where NTs exhibit pleiotropic responses (2). The ex-pression of detectable amounts of high- and low-affinityNT receptor mRNAs, originally thought to be restricted tosensory, cranial- and dorsal root ganglia and to cells ofneural crest origin (6, 7), was reported in lung (5, 8, 9). Theexpression of BDNF mRNA in lung epithelium (10) as wellas of NTs and NT receptors in human pulmonary arterieswas also described (11).

It was hypothesized that NTs may play a role in lungfunction and in the pathophysiology of allergic inflamma-tion (12, 13), but the expression and localization of thesefactors in the lung was investigated only sparsely (5, 10,11). The present study was designed to assess the expressionand distribution of NTs and high- and low-affinity NT re-ceptors in human lung by molecular biology (reversetranscriptasepolymerase chain reaction [RT-PCR]), im-munochemical (Western blot analysis), and immunohisto-chemical techniques.

Materials and MethodsSubjects and Tissue PreparationSurgical samples of lung were obtained from smoking (10) andnonsmoking (5) patients (9 males and 6 females, age range 4565yr) undergoing pulmonary lobectomy for nonmalignant pulmonaryneoplasms. Specimens used were obtained from portions of paren-chyma surrounding the lesions and did not reveal pathologic find-ings. Specimens were dissected out and homogenized (for RT-PCR and Western blot analysis) or fixed in a buffered 10% formalin

Ricci, Felici, Mariotta, et al.: Neurotrophins in the Human Lung 13

solution for 24 h (for immunohistochemistry). Fixed specimenswere dehydrated in ethanol and embedded in paraffin. Serial 10-m-thick sections were obtained using a rotatory microtome,mounted on gelatine-coated coverslips, and processed for immuno-histochemistry (see below).

RT-PCR AnalysisTotal RNA was extracted from frozen adult human lung (n 15)at 37C by using Trizol reagent (Gibco BRL, Gaithersburg, MD)and purified over a cesium chloride cushion by ultra-centrifugation.Total RNA isolated from human lung specimens was treated withDNase I (RQI RNase-free DNase; Promega, Madison, WI) at 37Cfor 30 min, at 1 U/10g total RNA to eliminate contamination bygenomic DNA. One microgram of treated total RNA was subjectedto transcription into cDNA using a standard method using murineleukemia virus reverse transcriptase (Perkin-Elmer, Boston, MA).Total RNA nonincubated with transcriptase was used in subse-quent experiments as a negative control. Resulting cDNA productswere amplified by AmpliTaq polymerase (Perkin-Elmer) using oli-gonucleotides. Primers designed using published NT protein andNT receptor sequences are listed below. NGF 5CGCTCATCC-ATCCCATCCCATCTTC, 3CTTGACAAGGTGTGAGTCGT-GGT; BDNF 5AGGGTTCCGGCGCCACTCCTGACCCT, 3CT-TCAGTTGGCCTTTGTGATACCAGG; NT-3 5CGAAACGCG-TATCGCAGGAGCATAAG, 3GTTTTTGACTCGGCCTGGC-TTCTCTT; TrkA 5TCTTCACTGAGTTCCTGGAG, 3TTCTC-CACCGGGTCTCCAGA; TrkB full length 5TACACTTGTAC-TAAAATACA, 3GTGTCCCCGATGTCATTCGC; TrkB [TR-]truncated isoform 5TAAAACCGGTCGGGAACATC, 3ACCC-ATCCAGTGGGATCTTA; TrkC 5CATCCATGTGGAATACT-ACC, 3TGGGTCACAGTGATAGGAGG; p75 5AGCCCAC-CAGACCGTGTGTG, 3TTGCAGCTGTTCCACCTCTT. Theseprimers yield oligomer products of distinctive size: NGF, 267 bp;BDNF, 161 bp; NT-3, 203 bp; TrkA, 229 bp; TrkB, 245 bp; TrkB[TR-], 161 bp; TrkC, 228 bp; p75, 663 bp. After 95C hot start,cycling proceeded for 35 cycles with 2 min at 95C, 2 min at 62Cusing NGF, NT-3, and TrkC primers, at 65C using BDNF primer,at 60C with TrkA, TrkB, TrkB[TR-] and p75 primers, followedby 1 min at 72C. Products were then resolved by electrophoresis.

SequencingSequencing analysis was performed, for each primer set, for RT-PCR products from one adult human lung and from rat braincDNA. The RT-PCR products were run onto a 10% acrylamidegel and the excised bands were PCR-amplified using appropriateprimer sets. PCR products were then purified through MicroSpinS300 Columns (Pharmacia Biotechnology, Upsala, Sweden), se-quenced in both DNA strands by ABI PRISM Dye TerminatorCycle Sequencing Ready Reaction Kit (Perkin Elmer AppliedBiosystems, Fremont, CA) on an automated Applied ABI 377ADNA sequencer (Perkin Elmer). The data obtained were analyzedusing a computer assisted sequence software (Sequence NavigatorSoftware; Perkin Elmer). Sequences were comared with the Gen-Bank database (Genetics Computer Group, Madison, WI).

Assessment of Specificity of Antibodies UsedThe first step of the study consisted in assessment of the specificityof antibodies used that were obtained from commercial sources.For detecting of NTs and NT receptors the following antisera wereused: (i ) rabbit anti-NGF polyclonal antibody. It displays less than1% crossreactivity against recombinant human NT-3, NT-4, and

BDNF (sc-548; Santa Cruz Biotechnology, Santa Cruz, CA); (ii)rabbit polyclonal antibody anti-BDNF. It did not crossreact withNT-3 or NGF (sc-546; Santa Cruz); (iii) rabbit polyclonal antibodyanti NT-3. It did not crossreact with BDNF or NGF (sc-547; SantaCruz); (iv) rabbit polyclonal TrkA immunoglobulin. It recognizesan epitope corresponding to aminoacids 763 to 777, mapping adja-cent to the carboxy terminus of human trkA p140, noncrossreactivewith TrkB or TrkC (sc-118; Santa Cruz); (v ) rabbit polyclonalTrkB immunoglobulin it recognizes an epitope corresponding toaminoacids 794 to 808 of mouse trkB p145, noncrossreactive withTrkA or TrkC (sc-012; Santa Cruz); (vi) rabbit polyclonal TrkB[TK-] immunoglobulin. It recognizes the carboxy terminus of thetruncated [TK-] TrkB protein precursor, gp95 of mouse origin.Specific recognition of mouse, rat, and human TrkB [TK-] gp95,noncrossreactive with TrkB gp145, TrkA gp140 or TrkC gp140(sc-119; Santa Cruz); (vii) rabbit polyclonal TrkC immunoglobulin.It recognizes an epitope corresponding to aminoacids 798 to 812of porcine trkC p140 noncrossreactive with TrkA or TrkB (sc-117;Santa Cruz); (viii) goat polyclonal antibody to human p75 NTreceptor; It recognizes the amino acid sequence mapping the car-boxy terminus of the p75 NT receptor precursor of human origin,noncrossreactive with other growth factor receptors (sc-6188; SantaCruz).

The specificity of the Trk antibodies was assessed in membranesobtained from detergent lysates of SF9 cells, infected with recombi-nant baclovirus encoding human TrkA, TrkB, and TrkC receptors.Melanoma cells expressing human p75 were used as a referencecell population for p75 receptor analysis. Homogenates of rat brainwere used as reference tissue for both NT and NT receptor analysis(5). Immunochemical analysis displayed in material obtained fromreference cells or tissue the same migration profile typical of NTsand NT receptor protein investigated (see Results), suggesting thesuitability of antibodies for the subsequent analysis in pulmonarytissue.

Western Blot AnalysisLung homogenates were centrifuged at 1,500 g to remove nucleiand cell debris. The supernatant was re-suspended in an immuno-precipitation assay buffer containing phenylmethylsulfonylfluor-ide, aprotinin, and leupeptin. Aliquots of supernatant were usedfor protein assay against a standard of bovine serum albumin.Defined amounts (50 g) of proteins were loaded on to 10%stacking sodium dodecyl sulfatepolyacrylamide gel and electro-phoresed through a 10% sodium dodecyl sulfatepolyacrylamidegel. After electrophoresis, proteins were transferred to nitrocellu-lose paper. Antibodies were dissolved in 0.1 M phosphate-bufferedsaline containing bovine serum albumin (1%) and Tween 20(0.05%). Optimal antibody concentrations were established in aseries of preliminary experiments. The specificity of immune reac-tion was assessed using antibodies pre-adsorbed with correspond-ing peptides. Anti-NGF, anti-BDNF, and antiNT-3 antibodies(diluted 1:3,000) and anti-TrkA, anti-full, and truncated TrkB iso-forms anti-TrkC (diluted 1:500) and anti-p75NT (diluted 1:50)receptor antibodies were then applied. The product of immunereaction was revealed using secondary anti-rabbit (for all antiseraexcept those raised against p75 NT receptor) or anti-goat (for anti-p75 NT receptor antibodies) horseradish peroxidaseconjugatedimmunoglobulin (Ig)G. These IgGs were dissolved in phosphate-buffered saline containing nonfat milk (5%) and Tween 20. Theimmune reaction was then detected using a specific Western blottingdetection reagent (ECLTM RPN 2106; Amersham International,Buckinghamshire, UK) and developed using a chemiluminescencefilm (Hyperfilm; Amersham International). Positive bands were ana-lyzed by a scanning densitometer driven by an Image Quant software.

14 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 30 2004

ImmunohistochemistrySerial 5-m-thick sections were obtained from formalin-fixed tis-sues. Paraffin-embedded specimens were cut using a rotatorymicrotome. Sections were mounted on gelatin-coated slides andprocessed for immunohistochemistry as described elsewhere (11).Briefly, from each paraffin block, consecutive sections were ex-posed in sequence to anti-NGF, anti-BDNF, and antiNT-3 anti-bodies (diluted: 1:1,000) and to the same antibodies preadsorbedwith human NGF (10 g/ml), human BDNF (10 g/ml), humanNT-3 (10g/ml), or to anti-TrkA, anti-TrkB, anti-TrkB[TR-], anti-TrkC (diluted: 1:100), anti-p75 (diluted: 1:10) antibodies alone orpreadsorbed with the corresponding blocking peptides (10 g/ml).Optimal antisera dilutions and incubation times were assessed ina series of preliminary experiments. After incubation, slides wererinsed twice in phosphate buffer and exposed for 30 min at 25Cto anti-rabbit (for NT and Trk immunohistochemistry) or anti-goat(for p75NT receptor immunohistochemistry) secondary antibodiesdiluted 1:100. The product of immune reaction was revealed using0.05% 3,3-diaminobenzidine in 0.1% H2O2 as a chromogen. Sec-tions were then washed, dehydrated in ethanol, mounted in asynthetic mounting medium, and viewed at a light microscope.Endogenous peroxidase activity was blocked by H2O2, whereasnonspecific IgG binding to glass and tissue was prevented by addinga 3% fetal calf serum to the incubation medium. Further detailson the immunohistochemistry protocol are reported in a previousstudy by our group (11). The background of immune reaction wasevaluated by incubating some sections with a nonimmune serum,followed by processing with secondary antibodies.

Image AnalysisThe intensity of the immunoreaction developed within bronchialsurface and gland epithelium, bronchial and vascular smooth mus-cle, intrapulmonary ganglionic neurons, alveolar cells, intramuco-sal bronchus-associated lymphoid tissue (BALT), and alveolar orinterstitial macrophages was assessed microdensitometrically witha IAS 2,000 image analyzer (Delta Sistemi, Rome, Italy) connectedvia a TV camera to a light microscope. The system was calibratedtaking as zero the background obtained in sections exposed topreimmune serum. Ten cells of each type investigated were deline-ated by a diaphragm (diameter 100 m2) in seven slides per subject(five test slides and two control slides) exposed to different antisera.Analysis therefore included 70 different cells of each populationper subject and was made at a final magnification of 250. Theintensity of immune staining (test - control value) was assessed bya program of the image analyzer expressing the intensity of immunereaction in arbitrary units. Assessmant of the intensity of immunestaining on a linear scale based on the amount of deposition ofdiaminobenzidine product reaction was made according protocolsdeveloped in receptor histochemistry research (14).

StatisticsValues of Western blot densitometry and image analysis are meansof measurements of parameters examined per single subject. Quan-titative data of the size and intensity of immune bands of NT orNT receptors and of the intensity of immune staining for NTs andNT receptors in different cell populations was made by ANOVAfollowed by Duncans multiple range test as a post hoc test. A P 0.05 was taken as a cutoff of significance.

ChemicalsRabbit anti-NGF, anti-BDNF, antiNT-3, anti-TrkA, TrkB, TrkB-[TK-], TrkC, and goat anti-p75 NT receptor indicated above, -

NGF (sc-548P), BDNF (sc-546), NT-3 (sc-547P), and the followingblocking peptides used for raising the corresponding antibodies:TrkA (sc-118P), full length (sc-012P) and truncated isoforms ofTrkB (sc-119P), TrkC (sc-117P), and p75 NT (sc-6188P), werepurchased from Santa Cruz. Horseradish peroxidaseconjugatedsecondary antibodies for Western blotting, anti-goat (sc-2033), andanti-rabbit (sc-2004) were purchased from Santa Cruz. Anti-goatperoxidaseconjugated secondary IgGs and anti-rabbit IgG peroxi-daseconjugated secondary IgGs were purchased from SigmaChemical Co. (St. Louis, MO) and Boehringer Mannheim (Ger-many), respectively. Other chemicals were obtained from SigmaChemical Co. or Merck (Darmstadt, Germany).

ResultsRT-PCR Analysis

RT-PCR analysis was performed to identify NT and NTreceptor transcripts in specimens obtained from differentsubjects investigated. The specificity of each fragment waschecked by both electrophoresis and by sequencing. NGF,BDNF, and NT-3 primers amplified at 267 bp, 161 bp, and203 bp, respectively (Figure 1A). TrkA, TrkB, TrkB[TR-],TrkC, and p75 primers amplified at 229 bp, 245 bp, 161 bp,228 bp, and 663 bp, respectively (Figure 1) (15).

No apparent differences were found in RT-PCR analysisrelated to sex, age, and smoking or nonsmoking history fordifferent parameters assessed (data not shown).

Western Blot Analysis

The results of immunoblot analysis for NT immunoreactiv-ity and NT receptor immunoreactivity in pulmonary tissueand reference material are shown in Figures 2 and 3, respec-tively. NGF antibody was bound to a single band at 14kD in lung and in brain homogenates (Figure 2, lanes 1 and3). BDNF antibody reacted with a single band at 14 KDa(Figure 2, lanes 1 and 3). NT-3 antibody reacted with asingle band of 14 kD (Figure 2, lanes 1 and 3). The useof antibodies preadsorbed with corresponding NTs causedthe disappearance of bands of immunoreactivity in pulmo-nary and in brain tissue (Figure 2, lane 2).

TrkA antibody reacted with a single band of 140 kD bothin pulmonary tissue and in infected SF9 cells (Figure 3, lanes1 and 3), as well as TrkB antibody (Figure 3, lanes 1 and 3).TrkB-[TR-] antibody was bound to an immune band of 95kD (Figure 3, lanes 1 and 3), whereas TrkC antibody wasbound to an immune band of 140 kD (Figure 3, lanes 1and 3). p75-NT receptor antibody reacted with a band of 75kD both in pulmonary tissue and in human melanoma cells(Figure 3, lanes 1 and 3). The use of antibodies preadsorbedwith corresponding receptor blocking peptides caused thedisappearance of bands of immunoreactivity both in pulmo-nary tissue and in reference cells (Figure 3, lane 2).

NGF was the NT most expressed in human pulmonarytissue, followed in descending order by BDNF and NT-3(Figure 2) as revealed by microdensitometric analysis ofbands of immunoreactivity (Figure 2, right-hand panels)normalized to actin (Figures 2 and 3, lane 4). NT receptormicrodensitometry showed that the two isoforms of TrkBreceptor were the most expressed followed in descendingorder by TrkA, TrkC and p75 (Figure 3, right-hand panels).


Figure 1. NT and NT receptor transcripts expression in adult lung.Amplification of NGF, BDNF, and NT-3 (A ) and of TrkA, TrkB,TrkB[TR-], TrkC, and p75 (B ) transcripts in three pulmonarysamples (lung parenchyma) from male and female smoking (lanes1 and 2, respectively) and nonsmoking (lane 3) patients (Figure 2),analyzed for the presence of the different transcripts. The size ofthe RT-PCR products is indicated on the right (MW, lane 6).Accordingly, to ensure complete removal of potential DNA con-tamination; the total RNA isolated from human lung specimenswas treated with DNase I. Total RNA not incubated with thetranscriptase was used in the subsequent experiments as a negativecontrol (nRT, lane 5). Positive control was generated using ratbrain samples (Br, lane 4). No apparent differences were foundin RT-PCR analysis performed in all the subjects, related to sex,age, and smoking or nonsmoking history.

Immunohistochemistry

Neurotrophin immunoreactivity. In ciliated bronchial epi-thelium a slight NGF immunoreactivity, an intense BDNFimmunoreactivity and a slight NT-3 immunoreactivity wereobserved (Table 1 and Figure 4). In bronchial smooth mus-cle and in bronchial gland epithelium, the intensity of im-mune staining was similar for the 3 NTs, although bronchialgland immunoreactivity was stronger than in smooth muscle(Table 1 and Figure 4). An obvious NT immunoreactivitywas noticeable in neurons and to a lesser extent in satellitecells of intrapulmonary (parasympathetic) ganglia (Table 1and Figure 5). Intrapulmonary nerve fiberlike structureswere also immunoreactive for NTs, and their number washigher than that of immunostained ganglionic neurons(Table 1 and Figure 5). Microdensitometric analysis per-formed on nerve cell bodies revealed a moderate immuno-reactivity for NGF and BDNF and a slight immune reactionfor NT-3 (Table 1). No NT immunoreactivity was found inalveolar cells (data not shown), whereas BALT (Table 1 andFigure 6) and alveolar or interstitial macrophages displayedNGF but not BDNF or NT-3 immunoreactivity (Table 1).

Figure 2. Western blot analysis of NGF, BDNF, and NT-3 antibod-ies. Lane 1: Immunoblots with human -NGF and membranesobtained from rat brain exposed to different antibodies tested.Lane 2: Immunoblots with human -NGF and membranes ob-tained from rat brain exposed to different antibodies tested preab-sorbed with the corresponding NTs. The antibodies preabsorbedwith the corresponding NTs did not develop bands of immunoreac-tivity. Lane 3: Immunoblots with membranes obtained from humanlung specimens from adult male smoking patients undergoing pul-monary lobectomy for nonmalignant neoplasms. Lane 4: Immu-noblots with membranes obtained from human lung using antibod-ies preabsorbed with the corresponding NTs. The preabsorbedantibodies did not develop bands of immunoreactivity. Histogramssummarize values of microdensitometry for the respective blotsnormalized to actin. Densitometry values were obtained by multi-plying % optical density area occupied by immune bands (ex-pressed in m2). These values are the mean SEM of data ob-tained from single subjects examined within the different groups.Antibodies preadsorbed with corresponding peptides did not de-velop bands of immunoreactivity (lanes 2 and 4 ). *P 0.01 versusBDNF and NT-3.

In BALT, NGF immunoreactivity was located primarilyparafollicularly (Figure 6).

The tunica media of different-sized intrapulmonarybranches of the pulmonary artery displayed NT immunore-activity, slightly more pronounced for NGF and BDNFcompared with NT-3 (Table 1). The intensity of immunestaining was similar irrespective of the size of vessels investi-gated (data not shown).

Neurotrophin receptor immunoreactivity. Ciliated bron-chial epithelium did not display Trk or p75-NT receptorprotein immunoreactivity (Table 2). In bronchial smoothmuscle a moderate Trk but not p75-NT receptor proteinimmunoreactivity was noticeable, TrkB-[TR-] immunore-activity being the most expressed (Table 2). A faint Trkreceptor immunoreactivity was also observed in bronchialglands, which were negative for p75-NT receptor protein


Figure 3. Western blot analysis of TrkA, TrkB, TrkB [TR-], TrkC,and p75 antibodies. Lane 1: Immunoblots with membranes ob-tained from rat brain exposed to different antibodies tested. Lane2: Immunoblots with membranes obtained from infected SF9 cellsexposed to TrkA, TrkB, TrkB[TR-], and TrkC antibodies andfrom human melanoma cells exposed to p75-NT receptor antibody.Lane 3: Immunoblots with membranes obtained from rat brainexposed to different antibodies tested preabsorbed with the corre-sponding blocking peptides. The antibodies preabsorbed with thecorresponding peptides did not develop bands of immunoreactiv-ity. Lane 4: Immunoblots with membranes obtained from humanlung specimens from adult male nonsmoking patients undergoingpulmonary lobectomy for nonmalignant neoplasms. Lane 5: Immu-noblots with membranes obtained from human lung using antibod-ies preabsorbed with the corresponding blocking peptides. Thepreabsorbed antibodies did not develop bands of immunoreactiv-ity. Histograms summarize values of microdensitometry for therespective blots normalized to actin. Densitometry values wereobtained by multiplying % optical density area occupied byimmune bands (expressed in m2). These values are the mean SEM of data obtained from single subjects examined within thedifferent groups. *P 0.01 versus TrkA, TrkC, and p75.

immunoreactivity (Table 2 and Figure 4). Ganglionic neu-rons did not display Trk receptor immunoreactivity, butwere immunoreactive for p75-NT receptor protein (Table 2and Figure 5). Nerve fiberlike profiles displayed a moder-ate immunoreactivity for both high and low affinity NT

receptors (Table 2 and Figure 5). Pulmonary alveoli wereimmunoreactive for Trk A and Trk C, but not for otherNT receptors investigated (Table 2 and Figure 7). BALTwas immunoreactive for TrkA only (Table 2 and Figure 6),whereas macrophages displayed immunoreactivity for high-affinity NT receptors. Among these receptors, Trk B dis-played the most intense immunostaining, followed by TrkA,TrkC, and TrkB-[TR-] receptor protein immunoreactivity(Table 2).

High- and low-affinity NT receptor protein immunoreac-tivity was also observed within the wall of different-sizedpulmonary artery branches (Table 2).

DiscussionThe present study provides direct evidence that differentcomponents of human lung express NGF, BDNF, and NT-3as well as the corresponding high- and low-affinity NT re-ceptors. This work represents an extension of previous in-vestigations of our group that have demonstrated NTs andtheir receptors in alveolar macrophages (16), in the pulmo-nary vascular tree (11), and in lung cancer (17). Our findingsconfirm evidence that NTs are not confined to central ner-vous system or to neuronal elements, but are also presentin non-neuronal tissues such as human lung. The demonstra-tion of a different pattern of expression and localization ofNTs and their receptors in lung suggest that NTs may havea functional role.

The expression of NTs and NT receptors in target organsof the peripheral nervous system was reported. Thesegrowth factors promote the survival and function of localneurons (1). Nevertheless, the occurrence of NT synthesisand the expression of NT membrane receptors on non-neuronal cells of adult lung implies that locally producedNTs are able to exert direct functional and biological activi-ties on cells endowed with these receptors. Our observationssupport the hypothesis that biological actions of NTs withinlung can be exerted in a broader range of sites throughoutthe entire life span and in different pathologic conditions(17, 18).

Epithelial cells of adult human bronchial mucosa expressBDNF, and to a lesser extent NGF and NT-3 immunoreac-tivity. These data are in line with previous evidence of anabundant production of BDNF and NGF by adult visceralepithelia under basal and proinflammatory conditions (10,1921). In situ or cultured human bronchial epithelial cellsexpress NTs, and their production was significantly aug-mented if stimulated by interleukin-1 or tumor necrosisfactor- (20). In addition, BDNF mRNA was found in thelung epithelium (10, 19). Therefore, epithelial cells mayparticipate by increasing NT production in allergic inflam-mation (18, 22, 23). Lung epithelium is not immunoreactivefor NT receptors. This suggests that NTs may act as para-crine factors in regulating functional properties of neuronaland non-neuronal structures in adulthood as described invisceral epithelia (2, 5, 10). These observations are consis-tent with the findings of unaltered respiratory epitheliumin knockout BDNF / mice (10).

Bronchial smooth muscle expresses NTs and NT recep-tor immunoreactivity. Changes of bronchial and bronchiolarsmooth muscle are a prominent histopathologic feature of


TABLE 1

NT immunoreactivity microdensitometry in human lung

NGF BDNF NT-3

Ciliated epithelium 8.3 1.0 22.0 2.3* 7.8 0.4Bronchial smooth muscle 10.8 0.3 9.7 0.7 6.2 0.2**Bronchial glands 18.5 1.9 16.9 2.4 13 1.6**Ganglionic neurons 16.4 1.2 15.8 1.5 9.6 1.1**BALT 8.6 0.2 Nd NdMacrophages (alveolar, interstitial) 12.6 1.1 Nd NdPulmonary artery smooth muscle 7.5 0.6 8.2 0.5 5.8 0.4**

Definition of abbreviations: BALT, bronchus-associated lymphoid tissue; BDNF, brain-derived neurotrophic factor; Nd, not detectable; NGF, nerve growth factor;NT, neurotrophin.

Data are the means SE and are expressed in arbitrary units proportional to the intensity of staining. Values were obtained by subtracting data of immunostainingobtained with anti-NT antibodies preadsorbed with NTs from those obtained with anti-NT antibodies alone. For details on microdensitometry see Materials andMethods.

* P 0.01 versus NGF or NT-3 immunoreactivity.** P 0.05 versus NGF or BDNF immunoreactivity.

asthma and may contribute to airway hyperresponsiveness.NTs are potential mediators of airway inflammation duringasthma and lung injury (18, 23); it is therefore possible thatNTs may promote respiratory smooth muscle mitogenesis,

Figure 4. Micrographs of representative NT and NT receptor im-munostaining in sections of human lung specimens from adult malesmoking patients undergoing pulmonary lobectomy for nonmalig-nant neoplasms, showing bronchial epithelial cells, smooth muscle,and bronchial glands. Nonspecific (NS) immunostaining wasobtained when the section was exposed to the antibodies preab-sorbed with the corresponding blocking peptides. Note the localiza-tion of specific immunostaining using NGF and BDNF antibodieswithin ciliated cells (CC), basal cells (BC), and smooth musclecells (sm). Sierous bronchial gland cell bodies developed a specificNGF and TrkA immunoreactivity (arrowheads), as well as intersti-tial fusiform cell bodies that resemble fibroblasts (arrows). BM,basal membrane; M, mucous cells. Calibration bar: 25 m.

hyperplasia, and/or a proapoptotic effect and remodeling,similar to that described in hypertensive vascular smoothmuscle (24). The demonstration of an NT system in bron-chial smooth muscle adds new insights on a modulatoryrole of NTs on airway smooth muscle. On the other hand,NT production by smooth muscle may represent a trophicsignal promoting the development of innervation. In linewith this hypothesis is the demonstration of qualitativeand/or quantitative changes of peripheral neurons as a con-

Figure 5. Micrographs of NTs and NT receptors immunostainingin sections of human postganclionic neurons (n) and within nervefibers (f) from adult male or female smoking patients undergoingpulmonary lobectomy for nonmalignant pulmonary neoplasms.NS, nonspecific immunostaining. Some neurons express NT recep-tor and p75 immunoreactivity. Satellite cells display NT and NTreceptor immunoreactivity. Calibration bar: 25 m.


Figure 6. Micrographs of intramucosal bronchus-associated lymphoidtissue (BALT) from an adult smoking male patient exposed to NGFand TrkA antibodies. Note the intense immunoreaction developedwithin parafollicular area in which are located CD3 immunoreac-tive lymphocytes. No immunostaining was observed in the follicu-lar central area. Calibration bar: 25 m.

sequence of locally produced NTs as well as the develop-ment of hyperinnervation in transgenic mice overexpressingNGF (25). The occurrence of NT and NT receptor immuno-reactivity in bronchial glands suggests that NTs may alsocontrol glandular function in these glands acting as para-crine and/or autocrine factors.

Alveolar cells do not display NT immunoreactivity, butexpress TrkA and TrkC receptor immunoreactivity. Thesefindings, which are consistent with a previous study (19),suggest that these cells are sensitive to the influence of NTssynthesized and released elsewhere in lung.

NT and low-affinity NT receptor immunoreactivity wasalso demonstrated in pulmonary parasympathetic ganglia,with immune reaction located in nerve and satellite cells.The role of NTs in modulation of parasympathetic neuronphenotype and function has been already described in sev-eral organs (26, 27). Interactions between parasympatheticand sympathetic nerves represent an important regulatorymechanism control lung visceral target function. The pres-ence of parasympathetic intrapulmonary ganglion cellsdisplaying p75 receptor immunoreactivity is in line with

TABLE 2

NT receptor immunoreactivity microdensitometry in human lung

TrkA TrkB TrkB [TR-] TrkC p75

Ciliated epithelium Nd Nd Nd Nd NdBronchial smooth muscle 8.8 0.3 8.7 0.7 12.2 0.2* 8.5 0.7 NdBronchial gland cell body 4.5 1.9 4.9 2.4 5.6 1.6 4.9 0.1 NdGanglionic neurons Nd Nd Nd Nd 7.4 3.6Alveoli 8.4 0.7 Nd Nd 7.4 0.2 NdBALT 12.5 0.2 Nd Nd Nd NdMacrophages (alveolar, interstitial) 9.4 0.7 12.8 1.2 3.6 0.5* 8.5 0.5 NdPulmonary artery smooth muscle 15.4 1.3** 12.2 1.0 17.4 1.2** 11.8 0.9 10.6 0.8

For definition of abbreviations see Table 1.Data are the means SE and are expressed in arbitrary units proportional to the intensity of staining. Values were obtained by subtracting data of immunostaining

obtained with anti-NT receptor antibodies preadsorbed with receptor blocking peptides from those obtained with anti-NT receptor antibodies alone. For details onmicrodensitometry see Materials and Methods.

* P 0.01 versus TrkA, TrkB, TrkC, or p75 NT receptor immunoreactivity.** P 0.05 versus TrkB, TrkC, or p75 NT receptor immunoreactivity.

Figure 7. Micrographs ofTrkA and TrkC immuno-staining in the alveolus ofsmoking patients. Note theimmunoreactivity withinmembranous pneumocytesand type 2 cells (arrow-heads) that protrude intothe alveolar lumen (L). NS,nonspecific immunostain-ing. Calibration bar: 25m.

identification of a p75 receptor axonal transport mechanismfor NTs in efferent vagal neurons belonging to the dorsalmotor nucleus (28). In addition, neural tissue in fetal mouselung explants exhibits a striking increase in the amount ofp75 receptor when stimulated with neurotrophic factors,suggesting a role for this receptor during lung development(29). The larger number of NT-immunoreactive nerve fiberscompared with parasympathetic nerve cell bodies displayingNT immunoreactivity suggests that these fibers belong alsoto the sympathetic component of autonomic nervous systemin which the presence of NTs is documented (1).

The expression of NTs and NT receptors within immunecells, pulmonary macrophages, and in BALT confirms thepossible role of these growth factors in the modulation of


airway immune functions. The observations of the presentstudy extend and tend to support previous findings reportingNT and NT receptor immunoreactivity in alveolar macro-phages and pulmonary interstitium (16, 19). Both lympho-cytes and antigen-presenting cells like macrophages expressTrkA receptor and store and synthesize NTs. BALT ispresent in healthy adults and may be considered as a partof lung immunologic defense (16, 30). Our findings are inline with and support the view that NTs are in a key positionto exert an effect on immune-competent cell functions inlung health and disease (31).

The demonstration of NT and NT receptor immunoreac-tivity in human intrapulmonary arteries supports and ex-tends our previous study showing similar findings in themain branches of exptrapulmonary arteries (11). The ex-pression of NTs and NT receptors in intrapulmonary arter-ies also suggests the occurrence of a similar mechanism oftrophic control along the pulmonary arterial tree.

The above findings, which are consistent with those ofa recent investigation in mice (21), demonstrate that NTsare constitutively expressed by the resident cells of humanlung. NTs participate in the cross-talk between the lung resi-dent cells and immune cells that may be involved in thepathophysiology of a variety of lung disorders (2333). Fur-thermore, NTs, by regulating mesenchymal cell function, mayalso participate to abnormal repair processes resulting inprogressive fibrosis and end-stage lung disease (34).

References

1. Lewin, G. R., and Y. A. Barde. 1996. Physiology of neurotrophins. Annu.Rev. Neurosci. 19:289317.

2. Shibaiama, E., and H. Koizumi. 1996. Cellular localization of the trk neuro-trophin receptor family in human non-neuronal tissues. Am. J. Pathol.148:18071818.

3. Barbacid, M. 1994. The Trk family of neurotrophin receptors. J. Neuroimmu-nol. 25:13861403.

4. Thomson, T. M., W. J. Rettig, P. G. Chesa, S. H. Green, A. C. Mena, andL. J. Old. 1988. Expression of human nerve growth factor receptor oncell derived from all three germ layers. Exp. Cell Res. 174:533539.

5. Lomen, H. C., and E. M. Shooter. 1995. Widespread neurotrophin receptorexpression in the immune system and other non-neuronal rat tissue.J. Neurochem. 64:17801789.

6. Martin-Zanca, D., M. Barbacid, and L. F. Parada. 1990. Expression of thetrk proto-oncogene is restricted to the sensory cranial and spinal gangliaof neural crest origin in mouse development. Genes Dev. 4:683694.

7. Kaplan, D. R., B. L. Hempstead, D. Martin-Zanca, M. V. Chao, andL. F. Parada. 1991. The trk proto-oncogene product: a signal transducingreceptor for nerve growth factor. Science 252:554558.

8. Barbacid, M., F. Lamballe, D. Pulido, and R. Klein. 1991. The trk family oftyrosine protein kinase receptors. Biochim. Biophys. Acta 1072:235261.

9. Valenzuela, D. M., P. C. Maisonpierre, D. J. Glass, E. Rojas, L. Nunez, Y.Kong, D. R. Gies, T. N. Stitt, N. Y. Ip, and G. D. Yancopoulos. 1993.Alternative forms of rat trkC with different functional capabilities.Neuron 10:963974.

10. Lommatzsch, M., A. Braun, A. Mannsfeldt, V. A. Botchkarev, N. V. Botch-kareva, R. Paus, A. Fischer, G. R. Lewin, and H. Renz. 1999. Abundantproduction of brain-derived neurotrophic factor by adult visceral epitheliaimplications for paracrine and target-derived neurotrophic functions. Am.J. Pathol. 155:11831193.

11. Ricci, A., S. Greco, F. Amenta, E. Bronzetti, L. Felici, I. Rossodivita, M.Sabbatini, and S. Mariotta. 2000. Neurotrophins and neurotrophin recep-tors in human pulmonary arteries. J. Vasc. Res. 37:355363.

12. Bonini, S., A. Lambiase, S. Bovini, F. Angelucci, L. Magrini, L. Manni, andL. Aloe. 1996. Circulating nerve growth factor levels are increased inhumans with allergic diseases and asthma. Proc. Natl. Acad. Sci. USA 93:1095510960.

13. Virchow, J. C., P. Julius, M. Lommatzsch, W. Luttmann, H. Renz, andA. Braun. 1998. Neurotrophins are increased in bronchoalveolar lavagefluid after segmental allergen provocation. Am. J. Respir. Crit. Care Med.158:20022005.

14. Barili, P., E. Bronzetti, A. Ricci, D. Zaccheo, and F. Amenta. 2002. Microana-tomical localization of dopamine receptor protein immunoreactivity inthe rat cerebellar cortex. Brain Res. 854:130138.

15. Labouyrie, E., P. Dubus, A. Groppi, F. X. Mahon, J. Ferrer, M. Parrens,J. Reiffers, A. de Mascarel, and J. P. Merlio. 1999. Expression of neuro-trophins and their receptors in human bone marrow. Am. J. Pathol. 154:405415.

16. Ricci, A., S. Greco, S. Mariotta, L. Felici, F. Amenta, and E. Bronzetti.2000. Neurotrophin and neurotrophin receptor expression in alveolarmacrophages: an immunocytochemical study. Growth Factors 18:193202.

17. Ricci, A., S. Greco, S. Mariotta, L. Felici, E. Bronzetti, A. Cavazzana,G. Cardillo, F. Amenta, A. Bisetti, and G. Barbolini. 2001. Neurotrophinsand neurotrophin receptors in human lung cancer. Am. J. Respir. CellMol. Biol. 25:439446.

18. Hoglund, C. O., F. de Blay, J.-P. Oster, C. Duvernelle, O. Kassel, G. Pauli,and N. Frossard. 2002. Nerve growth factor levels and localization inhuman asthmatic bronchi. Eur. Respir. J. 20:11101116.

19. Yamamoto, M., G. Sobue, K. Yamamoto, S. Terao, and T. Mitsuma. 1996.Expression of mRNA for neurotrophic factors(NGF, BDNF, NT-3 andGDNF) and their receptors (p75NGFR, TrkA, TrkB and TrkC) in the adulthuman peripheral nervous system and non neuronal tissues. Neurochem.Res. 21:929938.

20. Fox, A. J., H. J. Patel, P. J. Barnes, and M. G. Belvisi. 2001. Release ofnerve growth factor by human pulmonary epithelial cells: role in airwayinflammatory diseases. Eur. J. Pharmacol. 424:159162.

21. Hikawa, S., H. Kobayashi, N. Hikawa, T. Kusakabe, H. Hiruma, T. Takenaka,T. Tomita, and T. Kawakami. 2002. Expression of neurotrophins andtheir receptors in peripheral lung cells of mice. Histochem. Cell Biol.118:5158.

22. Noga, O., G. Hanf, C. Schaper, A. OConnor, and G. Kunkel. 2001. Theinfluence of inhalative corticosteroids on circulating nerve growth factor,brain derived neurotrophic factor and neurotrophin-3 in allergic asthma-tics. Clin. Exp. Allergy 31:19061912.

23. Renz, H. 2001. The role of neurotrophins in bronchial asthma. Eur. J. Phar-macol. 429:231237.

24. Bono, F., I. Lamarche, and J. M. Herbert. 1997. NGF exhibits a pro-apoptoticactivity for human vascular smooth muscle cells that is inhibited byTGFbeta1. FEBS Lett. 416:243246.

25. Hoyle, G. W., R. M. Graham, J. B. Finkelstein, K. P. Nguyen, D. Gozal,and M. Friedman. 1998. Hyperrinnervation of the airways in transgenicmice overexpressing nerve growth factor. Am. J. Respir. Cell Mol. Biol.18:149157.

26. Hasan, W., and P. Smith. 2000. Nerve growth factor expression in parasympa-thetic neurons: regulation by sympathetic innervation. Eur. J. Neurosci.12:43914397.

27. Smith, P. G., J. D. Warn, J. J. Steinle, D. Krizsan-Agbas, and W. Hasan.2002. Modulation of parasympathetic neuron phenotipe and function bysympathetic innervation. Auton. Neurosci. 96:3342.

28. Helke, C. J., K. M. Adryan, J. Fedorowicz, H. Zhuo, J. S. Park, R. Curtis,H. E. Radley, and P. S. Distefano. 1998. Axonal transport of neurotroph-ins by visceral afferent and efferent neurons of the vagus nerve of therat. J. Comp. Neurol. 393:102117.

29. Tollet, J., A. W. Everett, and M. P. Sparrow. 2002. Development of neuraltissue and airway smooth muscle in fetal mouse lung explants. Am. J.Respir. Cell Mol. Biol. 26:420429.

30. Richmand, I., G. E. Pritchard, T. Ashcroft, A. Avery, P. A. Corris, andE. H. Walters. 1993. Bronchus associated lymphoid tissue (BALT) inhuman lung: its distribution in smokers and non smokers. Thorax 48:11301134.

31. Levi-Montalcini, R., S. D. Skaper, R. Dal Toso, L. Petrelli, and A. Leon. 1996.Nerve growth factor: from neurotrophin to neurokine. Trends Neurosci.19:514520.

32. Cosgaya, J. M., and A. Aranda. 1995. Nerve growth factor regulates trans-forming growth factor-beta 1 gene expression by both transcriptional andposttranscriptional mechanisms in PC12 cells. J. Neurochem. 65:24842490.

33. Olgart, C., and N. Frossard. 2001. Humal lung fibroblasts secrete nervegrowth factor: effect of inflammatory cytokines and glucocorticoids. Eur.Respir. J. 18:115121.

34. Micera, A., E. Vigneti, D. Pickholtz, R. Reich, O. Pappo, S. Bovini, F. X.Maquart, L. Aloe, and F. Levi-Schaffer. 2001. Nerve growth factor dis-plays stimulatory effects on human skin and lung fibroblasts, demonstra-ting a direct role for this factor in tissue repair. Proc. Natl. Acad. Sci.USA 98:61626167.

rcmb%2e2002-0110oc

Documents