INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Epithelial cells constitute the barrier that bacterial pathogens first encounter during invasion. In addition to serving as a physical barrier, these cells synthesize defense effector molecules to overcome pathogenic intruders. This defense barrier makes up a protective front of the innate human immune system. Besides effecting the destruction of potential pathogens, components of the innate defense system have an instructive role for the highly specific lymphocytes of the adaptive immune system, thereby amplifying the local clearance of pathogens and the potency of the defense mechanism. Thus, an effective, specific, and long-lasting immune response depends on the interplay between immune effector cells and molecules of the innate and adaptive systems (1, 2).

Antibacterial peptides and proteins are effector molecules in innate immunity, and are the main mediators of the killing of bacteria. They are constitutively synthesized at epithelial surfaces, and their expression is enhanced upon bacterial challenge or rupture of the epithelial barrier, as in wounds and inflammation (3). In addition to their role in surface defenses, antibacterial components are synthesized and stored by granulocytes that are recruited to sites of inflammation or immediately activated upon contact with bacteria that enter the circulation.

The lung epithelium covers a large surface that is constantly exposed to microorganisms. Imbalance in the innate immune system can result in repeated infections, as in cystic fibrosis (8). Inactivation of the salt-sensitive antibacterial peptide human -defensin-1 (HBD-1) has been suggested to be a decisive factor in the compromised immunity of the cystic fibrosis lung (9).

In order to compare the bactericidal capacity of lung mucosa in health and disease, we set out to determine the bactericidal activity in bronchoalveolar lavage fluid (BALF) from healthy individuals and Scandinavian patients with pulmonary sarcoidosis. Patients with this disease have an interstitial pulmonary inflammation characterized by a specific T-helper type 1 (Th 1) cell response to an unknown antigen (10). In this article we report that in sarcoidosis the antibacterial activity is enhanced and depends on multiple components as compared with that in healthy controls. We isolated several antibacterial peptides/proteins from BALF: LL-37,* -defensins (HNP 1-3), lysozyme, and antileukoprotease (ALP, also known as secretory leukocyte proteinase inhibitor [SLPI]), together with several unidentified components. At the messenger RNA (mRNA) level, LL-37, HBD-1, and -defensins were detected in healthy individuals and sarcoidosis patients. The enhanced antibacterial activity in sarcoidosis indicates an activation mechanism that may operate at the protein level through processing steps that deliver active components. Thus, it is justified to conclude that the airway epithelium is stringently protected by a complex antibacterial defense system. The enhanced activity of this system in patients with pulmonary sarcoidosis may be responsible for the low frequency of respiratory infections in these patients.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

BALF samples from 10 healthy individuals and 12 Scandinavian sarcoidosis patients were used (Table 1). The diagnosis of sarcoidosis was based on clinical presentations and radiographic findings, and was in five cases confirmed by positive biopsies. None of the patients was under treatment with corticosteroids, but Subject 11 had twice been treated with local injections (intraarticular) of corticosteroids (60 mg triamcinolone acetonide [Kenacort] T/injection), with one treatment given 4 wk and a second treatment 2 wk before BALF sampling. All healthy controls had a normal chest radiograph and had no respiratory infections during at least the 3 mo previous to the study.

Bronchoalveolar Lavage

Sterile saline solution at 37° C was instilled into a middle lobe bronchus in five aliquots of 50 ml. The fluid was gently aspirated after each instillation, and was collected in a siliconized bottle that was kept on ice. The BALF was strained through Dacron mesh and centrifuged at 400 × g for 10 min at 4° C, and the cell pellet was resuspended in RPMI 1640 (Sigma Chemical Co., St. Louis, MO). BALF cells were differentially counted and BALF supernatants were frozen for later analyses. Cytospins were prepared with aliquots of cell suspensions equivalent to 60,000 cells per slide. The material was centrifuged at 500 rpm for 3 min in a cytocentrifuge (Cytospin 2; Shandon, UK), and stained in May-Grünwald-Giemsa for differential cell counts or air dried and stored at 70° C for subsequent immunohistochemical staining.

Biopsies and Immunohistochemistry

Transbronchial specimens were obtained from 10 patients for histologic examinations. Samples were fixed in buffered formalin and embedded in paraffin; six of them were used for immunostaining to define the tissue distribution of the antibacterial peptide LL-37. This was done on 5-µm-thick paraffin sections of the transbronchial biopsies, using a rabbit polyclonal antibody specific for LL-37 that was obtained by a standard immunization scheme, using 100 µg synthetic peptide mixed with Freund's complete adjuvant as described (11). After deparaffinization and rehydration, the endogenous peroxidase activity was quenched by incubation of the sections with peroxide. The sections were then incubated with primary antibody at 4° C overnight. In seven cases (two healthy controls, five patients), cytospin preparations of BALF were also available. After quenching of the endogenous peroxidase activity, the slides were incubated with primary antibody for an hour at room temperature. For both the paraffin sections and the cytospin preparations, a secondary, biotinylated porcine antirabbit antibody (Dako AB, Glostrup, Denmark) was used for detection of bound anti-LL-37. The samples were incubated with the secondary antibody at room temperature for 40 min, treated with avidin-biotin-peroxidase complexes (Vector, Burlingame, CA) for 30 min, and incubated with 3,3'-diaminobenzidine as the chromogenic substrate. To ascertain the specificity of the immunostaining, anti-LL-37 was adsorbed with synthetic LL-37 prior to the immunohistochemical analysis (Figure 5B). This control was done in parallel for the different samples.

View larger version (94K): [in this window] [in a new window]   Figure 5.   Immunohistochemical staining with anti-LL-37 antibody. All preparations shown are from sarcoidosis patients. (A) Transbronchial biopsy (original magnification: ×125), showing positive signals in bronchial epithelium and bronchial glands. (B) A section parallel to A, in which the LL-37 peptide was preadsorbed with the antibody prior to immunodetection. (C ) Transbronchial biopsy with stained macrophages in lung parenchyma (original magnification: ×500). (D) Cytospin preparation of BALF cells showing signals in lung AM (original magnification: ×500).

Standardization of BALF Supernatants

To remove salt and low-molecular-weight components, the BALF supernatants were thawed and applied to Sep-Pak C₁₈ cartridges (Waters, Milford, MA) that were equilibrated with 0.1% trifluoroacetic acid (TFA). After washing the cartridges with 10% acetonitrile in 0.1% TFA, peptides/proteins were eluted with 80% acetonitrile in 0.1% TFA and lyophilized. This lyophilized peptide/protein fraction from each BALF supernatant served as starting material for all analyses at the protein level (i.e., high pressure liquid chromatographic fractionation, antibacterial assays, dot-blot analyses, and Western blot analyses).

Antibacterial Assay

Thin plates (1 mm) were made of 1% agarose in standard Luria-Bertani (LB) broth (containing 160 mM NaCl) and with approximately 6 × 10⁴ cells/ml of the test bacterium Bacillus megaterium Bm 11. Small wells of 3 mm were punched in the plates, and a sample (3 µl) was applied to each well. After overnight incubation at 30° C, the diameters of inhibition zones were recorded. In statistical analyses, p values were obtained by use of the nonparametric Mann-Whitney U-test.

Reversed-Phase High-Pressure Liquid Chromatographic Fractionation

A Waters system with photometric detection at 214 nm and 280 nm and a Vydac C₈ column (5 µm; 4.6 × 250 mm; Separations Group, Hesperia, CA), was utilized for isolation of peptides/proteins. Elution was done with a gradient of acetonitrile in 0.1% aqueous TFA at 1 ml/min.

Structural Analysis

A matrix-assisted laser-desorption/ionization instrument (Lasermat 2000; Finnigan MAT, San Jose, CA) was used for determination of peptide mass. A 10 mg/ml solution of -cyano-4-hydroxycinnamic acid (Sigma) in 70% acetonitrile containing 0.1% TFA was used as a matrix. Edman degradation of the isolated peptides was done with an Applied Biosystems 470 A instrument and PE-ABI Procise HT 494 protein sequencer (PE-Applied Biosystems, Foster City, CA) by means of reversed-phase high-pressure liquid chromatography (HPLC) and detection with phenylthiohydantoin.

Immunoassay

Detection of LL-37 immunoreactivity in the chromatographic fractions was done with a dot-blot assay, using the rabbit polyclonal antibody specific for LL-37. The second antibody was antirabbit IgG conjugated with alkaline phosphatase, obtained from Sigma. The filter was stained for enzymatic activity in 100 mM Tris-HCl, pH 9.5; 100 mM NaCl; and 5 mM MgCl₂ containing 4-nitroblue tetrazolium chloride (0.2 mg/ml) and 5-bromo-4-chloro-3-indolyl phosphate (0.1 mg/ ml), both from Boehringer Mannheim (Mannheim, Germany).

Western Blot Analysis

Discontinuous sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE), using 10 to 20% Tricine Ready Gels (Novex, San Diego, CA), was used for identification of LL-37 in the BALF supernatants. Samples from two healthy individuals and four sarcoidosis patients were separated on the gel. The material in the gel was further blotted onto polyvinyldifluoride (PVDF) membranes by electrophoretic transfer as previously described (12). Immunoreactivity was detected with the LL-37-specific antibody and antirabbit IgG conjugated with horseradish peroxidase (Bio-Rad Laboratories, Hercules, CA). The enhanced chemiluminescence (ECL) Western blotting detection system (Amersham, Little Chalfont, Buckinghamshire, UK) was used to record the results.

RNA Extraction and Reverse Transcription-Polymerase Chain Reaction

Total RNA was extracted from BALF, cells with RNAzol B (Tel Test, Inc., Friendswood, TX) according to the instructions of the manufacturer. The T-cell fraction was prepared as described (13) by flourescence-activated cell sorting (FACS). All RNA material was denaturated at 90° C for 5 min before first-strand complementary DNA (cDNA) synthesis, and was then chilled on ice. For the first-strand synthesis, random hexamer primers and 200 units of Moloney murine leukemia virus reverse transcriptase (GIBCO BRL, Gaithersburg, MD) were used in a reaction volume of 20 µl according to recommended conditions. The reaction was incubated at 40° C for 45 min and then heated at 95° C for 5 min. The following primer pairs (at 0.5 µM each) were used in separate polymerase chain reactions (PCRs): 5'-TGAAGGTCGGAGTCAACGGATTTGGT and 5'-CATGTGGGCCATGAGGTCCACCAC, purchased from Clontech (Palo Alto, CA) for glyceraldehyde-3-phosphate dehydrogenase (G3PDH), 5'-GAAGACCCAAAGGAATGGCC and 5'-TCAGAGCCCAGAAGCCTGAG for the CAMP gene transcript that encodes the antibacterial peptide LL-37, 5'-CTGAGCCACTCCAGGCAAGA and 5'-GCTCAGCAGCAGAATGCCCA for -defensins (HNP 1-3), and 5'-TTGTCTGAGATGGCCTCAGGTGGTAAC and 5'-ACACTTCAAAAGCAATTTTCCTTTAT for -defensin (HBD-1). cDNA template concentrations were adjusted to that of the housekeeping gene for G3PDH. PCR amplification was performed with the following thermal-cycle profile: 3 min denaturation at 94° C; 40 cycles of annealing for 1 min, extension at 72° C for 1 min, denaturation at 94° C for 1 min; and an extension step at 72° C for 7 min. The annealing temperatures were 60° C for G3PDH, 62° C for CAMP (LL-37), and 55° C for -defensins and -defensin. Analyses of the reaction mixtures were performed in 1.5% agarose gel, and the DNA was blotted onto a Hybond N nylon filter (Amersham) according to standard procedure (14). The filters were then prehybridized for 4 h in 6× standard saline citrate (SSC), 5× Denhardt's solution, 1% SDS, and denatured salmon sperm DNA 100 µg/ml at 64° C. Hybridizations were done overnight with defined probes under the same conditions as for the prehybridization. Cloned cDNAs were used as probes for the antibacterial peptide-encoding genes (see the subsequent discussion), and the G3PDH probe was purchased from Clontech. All probes were labeled with ³²P, using a Rediprime labeling kit (Amersham). After hybridizations, the filters were washed several times, finishing with 0.1× SSC at 64° C for 15 min. The results were analyzed in a PhosphorImager 445 SI (Molecular Dynamics, Sunnyvale, CA).

cDNA Cloning and Nucleotide Sequence Analysis

For cloning of the HBD-1, the same primers as in the reverse transcription-PCR (RT-PCR) were used in a PCR reaction with cDNA derived from one sarcoidosis patient. A DNA fragment of the expected size of 253 bp was obtained. This fragment was purified from a 1.5% Seaplaque agarose gel (FMC BioProducts, Rockland, ME), using -agarase (New England Biolabs, Inc., Beverly, MA) as recommended by the manufacturer, and was subcloned into pCR-script II (Stratagene, La Jolla, CA). The subcloned insert was sequenced through cycle sequencing, using dye-labeled terminators on a DNA sequenator (PE-Applied Biosystems).

Total RNA from buffy coat of a healthy blood donor and 3'-rapid amplification of cDNA ends-PCR (3'-RACE-PCR) approach (15) were used for cloning of -defensin cDNA. The primer used for RT in the 3'-RACE was 5'-TCGAATTCCTCGAGAAGC(T₁₈). The primers used for amplification were 5'-TCGAATTCCTCGAGAAGC and the -defensin-specific primer 5'-GCCATGAGGACCCTCGCCAT. A nested PCR was performed with the same -defensin-specific primers used after RNA extraction and RT-PCR. A band of the expected size (231 bp) was obtained, and was subcloned and sequenced in the same way as the HBD-1 insert. Cloning to the CAMP gene cDNA encoding LL-37 has been described (16).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Antibacterial Activity

BALF were collected from 10 healthy individuals and 12 patients with pulmonary sarcoidosis. The clinical data are summarized in Table 1. The BALF cells were pelleted through centrifugation, and each supernatant was enriched for peptide/ protein content on Sep-Pak C₁₈ cartridges. The lyophilized peptide/protein fractions were dissolved in 0.1% TFA, and samples of 60 µg were assayed for antibacterial activity against the test bacterium Bacillus megaterium Bm 11 in standard LB broth. Healthy controls had homogenous values ranging from 6.0 to 7.2 mm (Figure 1), with a mean value of 6.8 mm. In contrast, patient samples had heterogeneous values ranging from 4.8 to 9.3 mm, with a mean value of 7.2 mm. A very low value for one patient, the only one treated with corticosteroids prior to BALF collection, was noted (see arrow in Figure 1). Two patients with inactive disease also had low values. The two maximum values were found in patients with a high relative content of BALF neutrophils (Table 1). Excluding the patient with corticosteroid treatment and those with inactive disease, a significantly (p = 0.035) higher mean value was recorded in the patient group (7.65 mm) than in the control group (6.8 mm). With a low-salt medium (40 mM NaCl) in the antibacterial assay, substantially higher values were obtained in all cases, but with the same pattern of low values in healthy controls and in patients with inactive disease (not shown).

View larger version (12K): [in this window] [in a new window]   Figure 1.   Antibacterial activity in BALF. An inhibition-zone assay was used to measure the antibacterial activity in BALF from healthy controls and sarcoidosis patients. Each value is the mean of one to four experimental measurements. Filled circles indicate patients with inactive disease, and the arrow indicates a patient treated with corticosteroids prior to BALF collection. The values for these patients were not included in the p-value calculation. (Controls: n = 10, mean value = 6.8 mm; patients: n = 9, mean value = 7.65 mm.)

Isolation and Characterization of Antibacterial Components

Material from four patients (Subjects 12, 19, 22, and 25; Table 1) and two healthy individuals (Subjects 5 and 9; Table 1) were selected for separation on reverse-phase HPLC. After fractionation of 300 to 500 µg derived from the lyophilized BALF supernatants, the antibacterial activity was assayed in every fraction (shown in Figure 2 for Subjects 12, 22, and 5). In the material derived from sarcoidosis patients, several fractions contained antibacterial activity, but with individual differences in distribution and potency. In contrast, the material from healthy subjects contained only one fraction with antibacterial activity, and this was found to originate from lysozyme (Figure 2C, Peak 3). This lysozyme fraction, at 31 to 32 ml, was also detected in the sarcoidosis-patient material, but in variable amounts in different patients. The structural data for identification of lysozyme are shown in Table 2. For the other characterized fractions with antibacterial activity from the sarcoidosis-patient material (Figure 2A), Peak 1 was identified as the -defensins (HNP1, HNP2, and HNP3) by sequence analysis and mass value determination (Table 2), and Peak 2 as the elastase inhibitor antileukoprotease (ALP), also by sequence analysis and mass spectrometry (the mass value obtained was 11,623 Da) (17). Peak 4 was identified as LL-37 by dot-blot and Western blot analyses (see the subsequent discussion), complemented with mass value determination.

View larger version (16K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   Figure 2.   Reverse-phase HPLC of BALF material. (A, B) BALF supernatants derived from two sarcoidosis patients (Subjects 12 and 22 in Table 1) and (C ) from one healthy individual (Subject 5 in Table 1). A Vydac C8 column equilibrated with 15% acetonitrile containing 0.1% TFA was used for HPLC. The elution was done with a gradient of 15 to 60% acetonitrile with 0.1% TFA for 45 min, followed by 60 to 80% acetonitrile with 0.1% TFA for 10 min. The bars with double arrowheads indicate fractions containing antibacterial activity. Numbered peaks are identified as: (1) -defensins (HNP 1-3); (2) ALP; (3) lysozyme; and (4) LL-37. The numbers and the bars that indicate characterized peaks or active fractions, respectively, are depicted above the peaks or fractions to which they refer.

RT-PCR and Southern Blot Analyses

RNA was isolated from precipitated cells of BALF, from sarcoidosis patients and healthy controls. This RNA was analyzed with semiquantitative RT-PCR with four specific primer sets (Figure 3). The first primer set was specific for transcripts that encode the housekeeping gene G3PDH. This PCR reaction was used for adjusting the template concentrations in the other PCR reactions. The additional three primer sets were specific for transcripts corresponding to the antibacterial peptides LL-37, HNP 1-3, and HBD-1. The identity of the amplified fragments was confirmed by Southern blot analysis using characterized probes (Figure 3). Expression of all three genes was detected in samples from complete BALF cells of both sarcoidosis patients and healthy individuals. In purified T cells, only cDNA for -defensins was detected.

View larger version (72K): [in this window] [in a new window]   Figure 3.   Southern blot analysis of four different RT-PCR reactions. The amplified bands were of the expected sizes for each gene, and the identity of bands was further confirmed by hybridization with 32P-labeled probes. The gene identity is indicated on the left side, and the origin of the material loaded is indicated at the top of each lane. The result was recorded with a PhosphorImager; the signals for G3PDH and LL-37 were recorded at the same sensitivity, whereas those for the -defensins and -defensin-1 were recorded together at higher sensitivity. The RNA used for the T-cell lane was from 98% pure CD4+ T cells of sarcoidosis patients, sorted by FACS (13).

Detection of LL-37

The presence of mature LL-37 was investigated by Western blot analysis. Samples of 40 µg of lyophilized BALF material from two healthy individuals and four sarcoidosis patients were applied on a tricine polyacrylamide gel (Figure 4). The mature peptide (of approximately 4.5 kD) was detected at different levels in BALF from both the sarcoidosis patients and the healthy individuals. Additional bands, detected at about 16 kD and over 30 kD, most likely represent monomeric and dimeric forms of the precursor protein, respectively, whereas bands between 4.5 kD and 16 kD were intermediate processing forms or peptide oligomers (Figure 4). This interpretation is supported by the use of another antiserum, directed against the proprotein of LL-37 (18), which revealed the same bands (not shown). Furthermore, there was a single band at 4.5 kD detected in a Western blot of material in Peak 4 (Figure 2A) with the anti-LL-37 antibody (not shown). The molecular weight of the material in this peak was determined as being 4,493 D with mass spectrometry, which is again in agreement with its being LL-37.

View larger version (55K): [in this window] [in a new window]   Figure 4.   Western blot analysis. BALF supernatants from two healthy individuals (A and B represent Subjects 3 and 4 in Table 1, respectively) and four sarcoidosis patients (A, B, C, and D represent Subjects 10, 12, 19, and 22 in Table 1, respectively) were analyzed. Synthetic LL-37 (100 pg) was used as a positive control, and molecular weight markers are indicated on the right side. The ECL detection system was used for recording the result.

Cellular Localization

To determine the cellular localization of LL-37, cytospin centrifugation was performed on BALF from both healthy individuals and sarcoidosis patients, followed by immunodetection. Strong immunostaining was demonstrated in alveolar macrophages (AM), whereas no signal was observed in lymphocytes (Figure 5D). For tissue localization of LL-37 sections of paraffin-embedded transbronchial biopsy material from sarcoidosis patients were used for immunohistochemistry. Positive signals for LL-37 were demonstrated in epithelial cells, bronchial glands, and AM (Figures 5A and 5C). The specificity for anti- LL-37 was controlled by preabsorbing the antisera used in these studies with synthetic LL-37 (Figure 5B).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The innate and adaptive immune systems are intertwined in a complex network. The outcome of the immune response depends on the stimulatory pathways that are activated (2). Cytokines are well-known regulators of the immune system, but recently effector molecules of innate immunity, such as complement fragment C3d and -defensins, have also been found to affect the adaptive system (19, 20). Thus, in addition to direct pathogen elimination, the effector molecules of innate immunity have stimulatory functions on the adaptive system.

Antibacterial peptides and proteins are effector molecules in the innate immune system that are constitutively expressed by developing granulocytes and are an integral part of surface defenses (4, 7).

In order to maintain normal pulmonary function, it is of vital importance to curb potential pathogenic invasion of the lung. To study defense effector molecules of the lung in disease, we selected material derived from the lungs of sarcoidosis patients as compared with healthy individuals. Previously, the expression of lysozyme, ALP, two -defensins, and LL-37 has been documented in human lung (21). The presence of high levels of -defensins has been shown in BALF from patients with diffuse panbronchiolitis (25), and lower levels have been found in patients without signs of infection or inflammation (26). In addition, activation of complement factors within the lung occurs in certain pulmonary diseases such as sarcoidosis (27). Sarcoidosis is a systemic granulomatous disease of unknown etiology. The most common symptoms in Scandinavian patients originate in the lungs, where activated T lymphocytes characteristically accumulate in response to a postulated antigen (10), whereas influx of granulocytes is normally low.

We analyzed BALF from healthy individuals and sarcoidosis patients at the protein, mRNA, and cellular levels for antibacterial components in the lung. In the healthy lung we identified, lysozyme (Figure 2C) and LL-37 (Figure 4), whereas in the sarcoidosis-patient material we also found -defensins, ALP, and additional, unidentified components (Figures 2A and 2B). We further found that the total antibacterial activity was enhanced in BALF supernatants of the sarcoidosis patients' lungs. Thus, BALF seems well suited for identification of components with antibacterial activity. Interestingly, this enhancement may explain the common clinical observation of a low frequency of upper respiratory tract infections in Scandinavian patients with active sarcoidosis (A.E., personal observation). Previously, we have isolated LL-37 from granulocytes stimulated with phorbol ester (11). Furthermore, we have shown induction of the gene encoding LL-37 in keratinocytes during inflammation of the skin (4). In BALF cells we have with the present study detected at the mRNA level expression of LL-37 from both sarcoidosis patients' and healthy lungs (Figure 3). LL-37 peptide was identified in an HPLC fraction by Western blotting and by mass value determination, confirming that in vivo, the processed peptide is LL-37. In addition, we located LL-37 in AM and bronchial epithelial cells by immunohistochemistry. AM are known to produce cytokines and other inflammatory substances, and also synthesize the antibacterial peptide LL-37 (Figures 5C and 5D).

We have noted differences between sarcoidosis patients both in antibacterial protein profile and antibacterial activity distribution, indicating activation or recruitment of yet further components of the pulmonary antibacterial defense system. The highest antibacterial activity was detected in the BALF supernatant from Subject 12 (Figure 2A). This may reflect a high infiltration of granulocytes (Table 1), which are known to contain several antibacterial peptides and proteins. Interestingly, one patient (arrow in Figure 1) with a pronounced reduction in antibacterial activity had received corticosteroids before BALF collection. Glucocorticoids have multiple antiinflammatory activities that are partly mediated through their inhibitory effect on nuclear factor (NF)-B (28). Our finding suggests that glucocorticoids downregulate the synthesis of several human antibacterial components. This downregulation of antibacterial components may involve signal pathways that include NF-B.

At the mRNA level we detected LL-37, HNP 1-3, and HBD-1, but the genes for these did not seem to be induced at the transcriptional level in sarcoidosis (Figure 3). The increased antibacterial activity in sarcoidosis depends on several antibacterial components that may be activated by protein processing. In fact, it is known that antibacterial effector peptides are synthesized as preproforms and stored as inactive propeptides (29).

It is of interest that one of the antibacterial components in the BALF supernatant was identified as ALP. ALP is a potent reversible inhibitor of the protease elastase, and has previously been shown to have antibacterial, antifungal, and antiviral activities (30). The cellular localization of ALP coincides with that of LL-37 (i.e., tracheal glands and bronchiolar epithelial cells). The target enzyme for ALP inhibition, elastase, is probably the main processing enzyme for the proproteins of the cathelicidins, resulting in release of mature antibacterial peptides such as LL-37.

The total antibacterial activity of the epithelial defense barrier is highly influenced by the local salt concentration. HBD-1 is inactive at moderate to high salt concentration, and this has been suggested to be the main cause for the compromised innate defenses in the lungs of patients with cystic fibrosis (9). In contrast, LL-37 is more active at moderate to high salt concentrations (33), but the processing enzymes that generate it might be inhibited with salt.

In conclusion, our results indicate that the lung contains several antibacterial factors, some salt-sensitive, supporting the multicomponent character of this defense system. The antibacterial components are probably of clinical relevance for the status of different pulmonary diseases.