INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Alpha 1-antitrypsin (1-AT) is a secretory protease inhibitor produced primarily in the liver (1). The 1-AT molecule is found abundantly in human plasma, with concentrations normally in the 20- to 53 µM range (2). The primary function of 1-AT is to irreversibly bind neutrophil elastase, a destructive serine protease released by neutrophils (3). In humans, a critical balance exists favoring a protective excess of 1-AT over neutrophil elastase in the lung. It is generally accepted that 1-AT produced and secreted by hepatocytes is the major contributor to the protective antiprotease shield in the lung alveolus (4, 5). Factors promoting a relative excess of protease creating an imbalance between 1-AT and neutrophil elastase can lead to progressive destruction of the lung parenchyma, which results in the loss of alveolar surface area and eventual respiratory failure (6).

This situation is best exemplified by patients suffering from the PiZ form of 1-AT deficiency, in which serum levels are < 15% of normal (7). Patients often develop emphysema by the third to fourth decade of life (8). Impairment of the protease-antiprotease balance may also have relevance to other pulmonary disease processes in which neutrophil-dominated inflammation occurs, such as cystic fibrosis and bronchitis (11). Recruitment of large numbers of inflammatory cells to the lung can lead to excessive protease release, a local excess of protease, and degradation of alveolar tissue.

Additional sources of 1-AT have been identified in humans, including production by peripheral blood monocytes and alveolar macrophages (12, 13). Production of 1-AT by phagocytic monocytes has been shown to be substantially lower in comparison to hepatocytes on a per-cell basis (14). For this reason, when considering the protease-antiprotease balance in the lung, the overall contribution of 1-AT protein production by monocytes and alveolar macrophages has received less attention. However, when considering that monocyte 1-AT production is augmented by neutrophil elastase and lipopolysaccharide (LPS), local regulation of the antiprotease shield may be an important front-line defense particularly in the microenvironment of lung inflammatory cells.

In addition to having a lower capacity to produce 1-AT, monocyte 1-AT regulation differs from that of hepatocytes in a number of ways. Previous work by Perlmutter and colleagues (15, 16) has demonstrated up-regulation of 1-AT protein synthesis by monocytes following stimulation with endotoxin and the acute-phase monokine interleukin-6 (IL-6). In similar experiments with a human hepatocyte-like cell line (HepG2), induction of 1-AT occurred after treatment with IL-6 but not IL-1. Furthermore, a unique serpin-enzyme complex (SEC) receptor present on hepatocytes and monocytes has been identified and shown to bind the 1-AT:neutrophil elastase complex or modified 1-AT inducing up-regulation of 1-AT in both cell types (17).

The regulation of 1-AT production is influenced by a variety of factors, a few of which have been described. Based on these data and that alveolar macrophages, the mononuclear cell of the lower respiratory tract, are present as a "front-line" defense in the lung, we hypothesized that other pro-inflammatory factors may directly influence control of 1-AT production by mononuclear cells. The cytokines interleukin-1 (IL-1) and tumor necrosis factor- (TNF) along with LPS are highly relevant pro-inflammatory factors known to trigger a variety of effects on monocytes and alveolar macrophages and therefore influence the lower airway environment (18, 19). This study demonstrates that 1-AT expression in monocytes is specifically regulated by physiologically relevant concentrations of IL-1 and TNF. Production of 1-AT by monocytes was also influenced by endotoxin as previously shown, with induction of secretion occurring at extremely low concentrations. Despite up-regulation of 1-AT, in serum-free conditions, functional antiprotease activity could not be detected. Furthermore, functional elastase-like activity was demonstrated with all conditions studied. Functional protease activity in peripheral blood mononuclear cell (PBMC) supernatants was inversely related to the concentration of 1-AT after treatment with LPS, IL-1, and TNF. These data implicate the importance of monocyte-derived 1-AT in the lower respiratory tract as a first-line defense against protease-induced tissue destruction.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell Purification

PBMC were purified from normal healthy volunteers. Heparinized (heparin sodium, 15 U/ml; Elkins-Sinn, Cherry Hill, NJ) blood was obtained (60 ml), and PBMC were purified using polysucrose/sodium diatrizoate (Histopaque; Sigma Diagnostics, St. Louis, MO) density gradient centrifugation. The PBMC (typically 20% monocytes, 80% lymphocytes) were counted, washed, and resuspended at a concentration of 5 × 10⁶/ml in RPMI 1640 (BioWhittaker, Walkersville, MD)/ 5% fetal bovine serum (Hyclone, Logan, UT) with 10 µg/ml of polymyxin B (Rorer Pharmaceuticals, New York, NY) to neutralize any contaminating endotoxin.

Cell Culture Conditions

LPS-, IL-1-, and TNF-induced 1-AT production by stimulated mononuclear cells. PBMC were either cultured alone or with increasing doses of LPS (LPS W, Escherichia coli 0127:B8; Difco Laboratories, Detroit, MI) in suspension overnight for 16 h at 37° C in 5% CO₂. Polymyxin B was not present in LPS-stimulated samples unless indicated. The cytokines recombinant human IL-1 (kind gift from the Biological Response Modifiers Program, National Cancer Institute, Frederick, MD) (0 to 10 ng/ml) and human TNF (kind gift from Knoll Pharmaceutical, Whippany, NJ) (0 to 100 ng/ml) were incubated with PBMC in suspension for 16 h as previously described. Polymyxin B was present in all studies involving the addition of cytokine reagents (to prevent nonspecific activation by potential endotoxin contamination).

Inhibition of LPS-, IL-1-, and TNF-induced 1-AT production by stimulated mononuclear cells. Combinations of soluble inhibitors were used to determine the specificity of IL-1-induced 1-AT production. Soluble type I interleukin-1 receptor (sIL-1 Ir) and soluble type II interleukin-1 receptor (sIL-1 IIr) (kind gift of J. E. Sims, Immunex, Seattle, WA) were added in combination to PBMC at a concentration of 1 µg/ml each. Additionally, soluble IL-1 IIr was added alone at a concentration of 10 µg/ml. Interleukin-1 receptor antagonist (IL-1ra) (kind gift from Dr. Daniel Tracey, Upjohn Laboratory, Kalamazoo, MI) was added in combination with sIL-1 IIr, both at a concentration of 1 µg/ml. All inhibitors were added to PBMC immediately before the addition of IL-1. For blocking experiments involving TNF, the neutralizing monoclonal antibody anti-TNF human (M199; Boehringer Mannheim, Indianapolis, IN) was added to PBMC at a concentration of 1 µg/ml immediately before the addition of TNF (100 ng/ ml). Additionally, the mouse monoclonal IgG₁ isotype control MOPC 21 (Sigma Diagnostics) was added (1 µg/ml) as an isotype control. In experiments involving LPS, polymyxin B was added to PBMC immediately before the addition of the highest dose studied (1 ng/ml). The PBMC incubated with the various blocking reagents were suspended in RPMI 1640  /  5% heat-inactivated fetal calf serum/polymyxin B (10 µg/ml) and incubated overnight for 16 h at 37° C with 5% CO₂. The following day, supernatants were removed and 1-AT release was measured by standard enzyme-linked immunosorbent assay (ELISA).

Inhibition of LPS-, IL-1-, and TNF-induced 1-AT transcription in stimulated mononuclear cells. Actinomycin-D (Sigma Chemical Co., St. Louis, MO) treatment of mononuclear cells was used to determine if up-regulation of 1-AT was the result of gene activation. Actinomycin-D (5 µg/ml) was added to mononuclear cells and incubated for 30 min followed by the addition of LPS, IL-1, and TNF as previously described. Cell-free supernatants were harvested 16 h later and measured for 1-AT concentration by ELISA. All samples were compared with similar samples not previously treated with actinomycin-D. Cell pellets were resuspended after supernatants were harvested and stained with trypan blue (Gibco-BRL, Gaithersburg, MD) to determine cell viability between treatment and control groups.

1-AT ELISA

1-AT release from PBMC was measured by a sandwich ELISA previously described but modified by our laboratory (20). The ELISA uses a goat anti-human 1-AT antiserum (Sigma Diagnostics) as the capture antibody and a rabbit anti-human 1-AT polyclonal antibody (Boehringer Mannheim) to complex the antigen. The complex was detected colorimetrically by an enzymatic reaction between a goat anti-rabbit IgG antibody conjugated to horseradish peroxidase (Bio-Rad Laboratories, Hercules, CA) and o-phenylenediamine (OPD) (Sigma Diagnostics). The ELISAs were read by comparison to human 1-AT derived from human plasma (Sigma Diagnostics) on a Dynatech MRX plate reader using Revelation Software (Immunosoft; Dynatech, McLean, VA). After reconstitution of 1-AT, the concentration of standard was determined by absorbance at 280 nm using the extinction coefficient of E1% = 5.3 (21).

Specificity of the 1-AT ELISA. The specificity of the 1-AT sandwich ELISA was determined by incubating a known concentration of 1-AT (10 ng/ml = 0.19 nM) with increasing amounts of human neutrophil elastase (CalBiochem, San Diego, CA) ranging from 0 to 20 nM. After complex formation for 15 min at room temperature, excess protease activity was blocked by the addition of 50 µg/ml of methoxysuccinyl ala-ala-pro-val-chloromethyl ketone (AAPV-cmk). The 1-AT:neutrophil elastase complexes were transferred to the detection plate, and ELISA analysis was completed as described. Additionally, the 1-AT ELISA was tested with human, rabbit, goat, and calf serum to demonstrate specificity for detection of the human protein.

Total RNA Extraction

Total RNA was purified by a modification of the method of Chomczynski and Sacchi (22). After overnight incubation, 1 × 10⁷ PMBC were centrifuged and the cell pellet was resuspended in 1 ml of Trizol Reagent (Gibco-BRL). Cells were lysed by repetitive pipetting, and RNA was extracted with chloroform and precipitated with isopropanol. The RNA pellet was washed two times in 75% ethanol, vacuum-dried, and resuspended in 20 µl of Rnase-free water. Total RNA concentration was measured by absorbance at 260  /280 nm. Purified RNA samples were maintained at 70° C until further use.

Northern Analysis

Ten micrograms of total RNA per sample was resuspended in an equal volume of RNA sample loading buffer (Sigma), denatured at 65° C for 15 min, chilled on ice, and then immediately loaded onto a 1% agarose/2.2 M formaldehyde gel. The ethidium bromide-stained samples were photographed, and the polaroid negative of the 28S ribosomal band of each sample was scanned and densitometrically recorded (NIH Image Software v1.6). Resolved RNA was blotted to nylon (Nytran; Schleicher & Schuell, Keene, NH) by capillary transfer with 20× SSPE and then immobilized by baking at 80° C for 1 h. Northern blots were probed with a ³²P-labeled human 1-AT cDNA probe (ATCC) followed by hybridization with a human -actin probe. Briefly, a 1.4 kilobase restriction fragment coding for the full-length human 1-AT M phenotype or a 1.1 kilobase fragment coding for human -actin was resolved on 1% agarose, isolated by electroelution, and labeled by the random priming method (RTS RadPrime DNA Labeling System; Gibco-BRL). Hybridization was performed overnight at 42° C in 50% formamide, 5× sodium chloride sodium phosphate, 0.5% EDTA, 5× Denhardt's solution, 100 µg/ml denatured salmon sperm DNA, and 2 × 10⁶ to 4 × 10⁶ cpm of labeled probe/ml of hybridization buffer. After hybridization, blots were washed twice at room temperature, 15 min each in 6× SSPE, 0.25% SDS; twice at 37° C, 15 min in 1× SSPE, 0.25% SDS; and once for 30 min at 55° C in 0.1× SSPE, 0.1% SDS. Autoradiography was performed overnight and quantitatively analyzed the following day using a Phospho-Imager (Molecular Dynamics, Sunnyvale, CA).

Western Analysis

All PBMC samples were incubated for 16 h in serum-free media alone or with polymyxin B, LPS (1 ng/ml), IL-1 (10 ng/ml) and polymyxin B, or TNF (100 ng/ml) and polymyxin B. Serum-free media was used to eliminate the exogenous addition of proteases or antiproteases present in fetal bovine serum. In similar experiments, the following protease inhibitors were added at final concentrations as follows: a mixture of aprotinin (0.15 U/ml), leupeptin (1 mM), and pepstatin (1 mM) or AAPV-cmk alone (0.2 mM). The inhibitors were added to PBMC in culture immediately before the addition of LPS, IL-1, or TNF and incubated for 16 h. Samples were run under denaturing conditions on 10% SDS-PAGE, blotted onto PVDF membranes (BioRad) by electrotransfer, and then probed with a rabbit anti-human 1-AT antibody (1:1,000) followed by an anti-rabbit peroxidase-conjugated antibody (1:5,000) (ECL-Kit; Amersham, Arlington Heights, IL). The blot was then stripped and reprobed in a similar fashion with a polyclonal antibody recognizing human neutrophil elastase (CalBiochem) or monoclonal antibody against human proteinase-3 (Research Diagnostics, Flanders, NJ) in an attempt to demonstrate complex formation between 1-AT and neutrophil elastase or proteinase-3. Additionally, a 10-fold molar excess of human neutrophil elastase was added to conditioned PBMC supernatants or 1-AT standard protein, incubated for 15 min at room temperature, and then analyzed under similar conditions. Molecular weights were determined by comparison to known molecular weight standards (Kaleidoscope-BioRad, Hercules, CA).

Measurement of Elastase-like Activity in Conditioned PBMC Supernatants

Supernatants from serum-free samples consisting of media alone, with polymyxin B, LPS (1 ng/ml), IL-1 (10 ng/ml) with polymyxin B, or TNF (100 ng/ml) with polymyxin B were incubated with a fixed concentration (100 µM) of the neutrophil elastase substrate methoxy-succinyl-ala-ala-pro-val-nitroanilide (MEOSAAPVNA). The supernatants and substrate were added together with 0.1 M Hepes (pH 7.5), 0.5 M NaCl, 0.1% Brij detergent as diluent, and the change in spectrophotometric absorbance at 410 nm was recorded over time. The change in absorbance over time was compared with standard dilutions of human neutrophil elastase ranging from 3 nM to 3 pM incubated with the same concentration of substrate.

Immunohistochemistry

Control and LPS-, IL-1-, and TNF-stimulated PBMC (5 × 10⁶/ml) were deposited on microscope slides by cytoprep centrifugation. The slides were fixed for 10 min in acetone at 4° C and then hydrated for 5 min with Tris-buffered saline (TBS) (0.05 M, pH 7.6). Rabbit serum (20%) was placed on the cells for 4 h at 37° C in a humidity chamber to block nonspecific binding of the antibody. A rabbit polyclonal anti-human 1-AT antibody (10 µg/ml) or isotype control IgG antibody (10 µg/ml) was placed on the cells and incubated in the humidity chamber at 37° C for 12 h. The slides were washed with TBS, incubated with a biotinylated anti-rabbit IgG (Vector Laboratories, Burlingame, CA) (dilution of 1:200) at 37° C for 10 min in a humidity chamber, and then washed again with TBS. Horseradish peroxidase- avidin D conjugate was added to the cells and incubated at 37° C for 10 min in a humidity chamber. After washing with TBS, the slides were placed in acetate buffer solution (0.02 M, pH 5.2) at 20° C for 5 min. The slides were then developed in 3-amino-9-ethylcarbazole, which had been activated with 1% hydrogen peroxide, for 1 min. The development was inhibited by the acetate buffer. The cells were then counterstained with hematoxylin for 1 min and prepared for viewing in glycerin jelly.

Statistical Analysis

All data were expressed as mean ± SEM. ANCOVA with Tukey's post-hoc testing was used to compare conditions (Systat, Evanston, IL). Statistical significance was defined as a p value  0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Detection of 1-AT by a Modified ELISA

Since contaminating proteases present in fetal calf serum or proteases secreted by monocytes in addition to 1-AT could potentially interfere with immunorecognition of 1-AT, the ability to detect free 1-AT or 1-AT complexed to neutrophil elastase by ELISA was determined before proceeding to analysis of conditioned PBMC supernatants. When increasing amounts of human neutrophil elastase (0 to 20 nM) were preincubated with a fixed concentration of 1-AT (0.19 nM), the ability to accurately detect 1-AT by ELISA was unaffected (data not shown), demonstrating that both free and bound 1-AT are measured accurately. Additional experiments were performed to determine the specificity of 1-AT detection by ELISA. Undiluted samples containing rabbit, goat, bovine, and human serum were tested with only human serum samples being immunoreactive, thus eliminating the possibility of immunoreactivity between the ELISA assay and culture media or detection antibodies (data not shown). Furthermore, 1-AT measurements taken from diluted human serum samples correlated well with normal serum levels ranging from 38 to 65 µM from normal healthy volunteers (data not shown).

Induction of 1-AT in Human PBMC by LPS

Up-regulation of 1-AT by LPS after 16 h of incubation represents data from seven separate experiments performed in healthy volunteers (Figure 1A). Freshly isolated unstimulated PBMC constitutively produced 1-AT that was detectable in supernatants after 16 h of culture (147 ± 19 ng/ml). Human PBMC cultured overnight in the absence of polymyxin B were exquisitely sensitive to bacterial LPS. A concentration of endotoxin as low as 40 pg/ml significantly increased 1-AT in PBMC supernatants when compared with unstimulated control supernatants (345 ± 62 ng/ml) (p  0.001). The induction of 1-AT by LPS was concentration dependent and maximal at 1 ng/ml. Up-regulation of 1-AT with 1 ng/ml of LPS resulted in a 3-fold enhancement above baseline (439 ± 66 ng/ ml). When polymyxin B was added to PBMC immediately before the addition of LPS at the highest dose (1 ng/ml), induction of 1-AT was inhibited (173.5 ± 35 ng/ml).

View larger version (14K): [in this window] [in a new window]   Figure 1.   (A) Induction of 1-AT in human PBMC by LPS. PBMC were incubated with increasing concentrations of LPS (0 to 1 ng/ ml) for 16 h. Cell-free supernatants were assayed for 1-AT release by ELISA. The results represent seven subjects and are reported as mean ± SEM. (B) Induction of 1-AT in human PBMC by IL-1. PBMC were incubated with increasing concentrations of IL-1 (0 to 10 ng/ml) for 16 h. Cell-free supernatants were assayed for 1-AT release by ELISA. The results represent five subjects and are reported as mean ± SEM. (C ) Induction of 1-AT in human PBMC by TNF. PBMC were incubated with increasing concentrations of TNF (0 to 100 ng/ml) for 16 h. Cell-free supernatants were assayed for 1-AT release by ELISA. The results represent six subjects and are reported as mean ± SEM.

Induction of 1-AT in Human PBMC by IL-1 and TNF

We next studied the effect of the pro-inflammatory cytokines IL-1 and TNF on 1-AT release from PBMC (Figure 1B and C). All experiments were performed in the presence of polymyxin B (10 µg/ml) to eliminate the potential for endotoxin contamination. IL-1 and TNF significantly enhanced the release of 1-AT into the cell culture media after overnight incubation (p  0.05). Secretion of 1-AT by PBMC increased after treatment with IL-1 (263 ± 37 ng/ml) and TNF (316 ± 59 ng/ml) in a dose-responsive fashion, resulting in approximately a 2-fold increase in measurable 1-AT protein in supernatants. Maximal stimulation occurred at an IL-1 dose of 10 ng/ml and a TNF dose of 100 ng/ml. The dosage range for each cytokine was chosen to approximate tissue levels during inflammation (23).

Specificity of 1-AT Up-regulation by LPS, IL-1, and TNF in Monocytes

To demonstrate that the effects observed with LPS, IL-1, and TNF were not due to contaminating factors, specific inhibitors for each agent were added to PBMC overnight along with either LPS, IL-1, or TNF. As previously shown, 1-AT production was inhibited to baseline when the highest dose of LPS (1 ng/ml) was coincubated with polymyxin B (Figure 2A). Coincubation of IL-1 (10 ng/ml) with the sIL-1 IIr (10 µg/ml) completely inhibited the enhanced release of 1-AT (Figure 2B). Furthermore, the complexation of IL-1 with sIL-1 IIr did not appear to alter the constitutive production of 1-AT. In a similar fashion, induction of 1-AT by TNF (100 ng/ml) was inhibited by coincubation with the blocking monoclonal antibody M199, whereas coincubation with an isotype control antibody (MOPC) did not significantly reduce detection of 1-AT in supernatants (Figure 2C).

View larger version (8K): [in this window] [in a new window]   Figure 2.   (A) Specificity of 1-AT up-regulation by LPS. PBMC were incubated for 16 h in media with polymyxin B (0 + PB), media with 1,000 pg/ml of LPS (LPS), or media with polymyxin B and 1,000 pg/ml of LPS (LPS + PB). Cell-free supernatants were assayed for 1-AT release by ELISA. The results are from one subject and are representative of six separate experiments. (B) Specificity of 1-AT up-regulation by IL-1. PBMC were incubated for 16 h in media with polymyxin B, media with polymyxin B and 10 ng/ml of IL-1, or media with polymyxin B, 10 ng/ml of IL-1, and soluble type II interleukin-1 receptor at a concentration of 10 µg/ml (IL-1 + sIL-1 IIr). Cell-free supernatants were assayed for 1-AT release by ELISA. The results are from one subject and are representative of two separate experiments. (C ) Specificity of 1-AT up-regulation by TNF. PBMC were incubated for 16 h in media with polymyxin B, media with polymyxin B and 100 ng/ml of TNF, media with polymyxin B, 100 ng/ml of TNF, and a neutralizing monoclonal antibody to human TNF at a concentration of 1 µg/ml (TNF + M199), or media with polymyxin B, 100 ng/ml of TNF, and an isotype control monoclonal antibody at a concentration of 1 µg/ml (TNF + MOPC). Cell-free supernatants were assayed for 1-AT release by ELISA. The results are from one subject and are representative of two separate experiments.

We next determined whether monocytes, lymphocytes, or both cell types were responsible for 1-AT production under the conditions previously described. A consistent pattern of immunostaining for 1-AT was demonstrated only in the monocyte population under all conditions studied (Figure 3). Immunodetection of 1-AT was absent in lymphocytes under all conditions, demonstrating that monocytes were responsible for constitutive and up-regulated expression of 1-AT.

View larger version (124K): [in this window] [in a new window]   Figure 3.   Production of 1-AT by monocytes. PBMC were incubated overnight with endotoxin (1 ng/ml) and then prepared for immunostaining. Panel A demonstrates stimulated PBMC stained with an IgG isotype control antibody. Panel B depicts stimulated PBMC that were stained with an antibody against 1-AT. The monocyte cytoplasm stains brown, indicating the presence of 1-AT.

Inhibition of 1-AT Up-regulation by Actinomycin-D

We hypothesized that increased detection of 1-AT in stimulated mononuclear cell supernatants was the result of a newly formed product. To rule out the possibility that up-regulation of 1-AT was due to enhanced release of a preformed protein, actinomycin-D treatment was used to inhibit transcription in untreated PBMC and in PBMC treated with LPS (1 ng/ml), IL-1 (10 ng/ml), and TNF (100 ng/ml). Actinomycin-D pretreatment resulted in inhibition of 1-AT in PBMC supernatants in all samples when compared with similar samples not preincubated with actinomycin-D (Figure 4). In fact, all samples treated with actinomycin-D had a similar baseline 1-AT concentration most likely representing constitutive production of 1-AT prior to inhibition of transcription. Trypan blue dye exclusion was used to determine cell viability between PBMC samples treated with or without actinomycin-D. Cell viability was > 95% in all samples, demonstrating that decreased 1-AT production was not a consequence of increased cell death with actinomycin-D treatment.

View larger version (14K): [in this window] [in a new window]   Figure 4.   Induction of 1-AT synthesis in monocytes by LPS, IL-1, and TNF. PBMC were preincubated for 30 min with 5 µg/ml of actinomycin-D. Then, PBMC were incubated in either standard growth media with polymyxin B (0 + PB) media without polymyxin B (0), LPS at a concentration of 1,000 pg/ml (LPS), IL-1 at a concentration of 10 ng/ml with polymyxin B (IL-1), or TNF at a concentration of 100 ng/ml with polymyxin B (TNF). Cell-free supernatants were harvested 16 h later along with similar samples that had not been preincubated with actinomycin-D. 1-AT concentration was measured by ELISA.

Up-regulation of the 1-AT Gene by LPS, IL-1, and TNF

To further characterize 1-AT up-regulation, Northern analysis was performed on unstimulated and stimulated PBMC as previously described (Figure 5). 1-AT mRNA was detected in control unstimulated PBMC in the absence or presence of polymyxin B, demonstrating constitutive expression of the 1-AT gene. When compared with unstimulated cells, overnight incubation of PBMC with LPS, IL-1, or TNF resulted in an increase in 1-AT mRNA above baseline. The induction of 1-AT mRNA by LPS, IL-1, and TNF occurred in a dose-responsive manner that correlated well with increased 1-AT protein secretion. The highest doses of LPS (1 ng/ml), IL-1 (10 ng/ml), and TNF (100 ng/ml) resulted in a 5.4-, 7.2-, and 15.5-fold increase in steady-state mRNA levels above baseline controls, respectively. Quantitation of mRNA was similar when data were normalized by the ethidium bromide stain or a human -actin cDNA probe (data not shown). These data taken with previous results suggest that up-regulation of 1-AT in monocytes by LPS, IL-1, and TNF is the net effect of increased protein synthesis. The disproportionate increase in 1-AT message levels relative to the increase in measured 1-AT protein in PBMC incubated with LPS, IL-1, and TNF imply that up-regulation of 1-AT is not exclusively pretranslational. Furthermore, comparison of LPS-, IL-1-, and TNF-induced up-regulation of 1-AT transcripts provides evidence that each factor may mediate a distinct form of regulation of the 1-AT gene.

View larger version (39K): [in this window] [in a new window]   Figure 5.   Induction of 1-AT mRNA in monocytes by LPS, IL-1, and TNF. 1-AT mRNA induction from human, stimulated PBMC was determined by Northern analysis after a 16-h incubation with media (0), media with polymyxin B (0 + PB), increasing doses of LPS (0 to 1,000 pg/ml), or LPS with polymyxin B (1,000 + PB) (left panel ) or with media without polymyxin B (0  PB), increasing doses of IL-1 (0 to 10 ng/ml) with polymyxin B, or increasing doses of TNF (0 to 100 ng/ml) with polymyxin B (right panel ). RNA samples were hybridized with a 32P-labeled 1-AT cDNA probe and washed thoroughly. Autoradiographs were developed overnight and quantitatively analyzed. Panel A depicts the polaroid negative image of the ethidium bromide-stained samples prior to capillary transfer. Panel B depicts the Northern blot of 1-AT mRNA from the endotoxin- and cytokine-treated samples. Panel C demonstrates the densitometric measurement of each lane and is expressed as fold increase over negative control (media control) after standardization of each sample with densitometry from the respective 28S ribosomal band.

Physical Characterization of Newly Synthesized 1-AT in PBMC

Serum-free supernatants from PBMC control conditioned media with or without polymyxin B, LPS, IL-1, and TNF were assayed by Western analysis under denaturing conditions to determine the molecular mass of the secreted 1-AT protein. Analysis in serum-free conditions was chosen to eliminate the exogenous addition of protease into the PBMC environment. The concentration of 1-AT detected in serum-free media was unaffected when compared with similar samples incubated in media containing serum as measured by ELISA. Interestingly, immunodetection of an approximately 80 kD polypeptide was present in control and LPS-, IL-1-, and TNF-stimulated samples (Figure 6C). The intensity of the polypeptide band in the respective samples corresponded to the amount of 1-AT present in the supernatants as measured by ELISA (Figure 6A). Unexpectedly, no 52 kD, secreted, unbound 1-AT was present in any samples, suggesting that the 1-AT was entirely complexed. When the 1-AT standard protein was incubated with a 10-fold molar excess of human neutrophil elastase and then analyzed under the same conditions, formation of an 80 kD polypeptide complex was detected that migrated at the same location as the protease-antiprotease complex harvested from PBMC supernatants. Similar results were obtained when cell lysates were analyzed under the same conditions. Addition of a 10-fold molar excess of human neutrophil elastase to conditioned supernatants before Western blotting did not influence migration of the complex or intensity of the band. Based on the molecular mass of the polypeptide complex, antibodies against human neutrophil elastase and human proteinase-3 were used in an attempt to co-detect protease bound to 1-AT. Nevertheless, concentrated supernatants failed to show either neutrophil elastase or proteinase-3 as an 80 kD complex or in the unbound native form. Experiments with known concentrations of neutrophil elastase and proteinase-3 free or bound to 1-AT suggested that our ability to demonstrate the presence of neutrophil elastase and/or proteinase-3 was limited by the sensitivity of the respective antibodies used and their ability to recognize the bound protease complex.

View larger version (25K): [in this window] [in a new window]   Figure 6.   Formation of 1-AT:protease complexes. PBMC were incubated in serum-free media with polymyxin B (0 + PB), serum-free media alone (0), serum-free media with 1,000 pg/ml of LPS (LPS), serum-free media with 10 ng/ml IL-1 and polymyxin B (IL-1), or serum-free media with 100 ng/ml TNF and polymyxin B (TNF). Cell-free supernatants were harvested 16 h later. Panel A demonstrates the nanomolar concentration of 1-AT present in supernatants as measured by ELISA. Panel B depicts the nanomolar concentration of functional elastase activity present in the same supernatants. The samples were incubated with a fixed concentration (0.5 mM) of the neutrophil elastase substrate MEOSAAPVNA in 0.1 M Hepes (pH 7.5), 0.5 M NaCl, 0.1% Brij detergent, and the change in spectrophotometric absorbance at 410 nm was recorded over time and compared with samples containing known amounts of neutrophil elastase. Panel C demonstrates the Western analysis of the same samples. Samples were run under denaturing conditions on 7.5% SDS-PAGE, blotted, and then probed with a rabbit anti-human 1-AT antibody (1:1,000) followed by an anti-rabbit peroxidase-conjugated antibody.

Functional Analysis of Monocyte-derived 1-AT

Having documented 1-AT release from monocytes, we next attempted to characterize the functional activity of the 1-AT. To test this, serum-free supernatants from PBMC were incubated with a constant amount of human neutrophil elastase followed by the addition of the neutrophil elastase substrate (MEOSAAPVNA). Our hypothesis was that increased 1-AT secretion into supernatants would result in an increased ability to inhibit neutrophil elastase activity. Unexpectedly, neither control nor stimulated PBMC supernatants showed inhibitory reactivity for neutrophil elastase. Based on this finding and the results from Western analysis, we next incubated the same conditioned PBMC supernatants used for Western analysis with neutrophil elastase substrate alone and monitored elastase-like activity over time (Figure 6B). All samples including untreated control supernatants demonstrated elastase-like activity. This elastase-like activity was inversely related to the amount of 1-AT present in supernatants as previously measured by ELISA and Western analysis. Despite up-regulation of 1-AT secretion by PBMC under the conditions studied, only bound 1-AT could be detected. These data imply that PBMC-derived 1-AT is released and rapidly complexed to another protein (presumably a protease) that inactivates it.

Inhibition of Protease-Antiprotease Complexation

Under the presumption that 1-AT was forming a complex with a serine protease, we attempted to inhibit antiprotease-protease complexation by incubating PBMC under previously described serum-free conditions with known serine protease inhibitors (Figure 7). When PBMC were incubated overnight with a mixture of aprotinin, leupeptin, and pepstatin or with AAPV-cmk alone, complexation of 1-AT in conditioned supernatants was not affected. Whereas, complexation of exogenous 1-AT and neutrophil elastase was clearly inhibited with an excess of the serine protease inhibitor, AAPV-cmk, under similar conditions. Trypan blue staining of cells under all conditions revealed > 95% viability. Similar results were obtained when PBMC cell lysates were examined from the study samples (data not shown).

View larger version (24K): [in this window] [in a new window]   Figure 7.   Inhibition of 1-AT:protease complexes. A total of 30 ng of 1-AT was incubated in serum-free media alone (lane 1) or with a 10-fold molar excess of neutrophil elastase (lanes 2 through 4) for 10 min at room temperature. Immediately before the addition of neutrophil elastase, a 100-fold (lane 3) or 1,000-fold (lane 4) molar excess of AAPV-cmk (0.02 and 0.2 mM) was added. PBMC were incubated in serum-free media with polymyxin B (lanes 5 and 6) or 1,000 pg/ml of LPS (lanes 7 and 8). Additionally, a mixture of aprotinin (0.15 U/ml), leupeptin (1 mM), and pepstatin (1 mM) (lanes 5 and 7) or AAPV-cmk (0.2 mM) (lanes 6 and 8) was added and incubated for 16 h. 1-AT:neutrophil elastase complexes and cell-free PBMC supernatants were analyzed by Western analysis for 1-AT as previously described.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

1-AT is generally found in relative abundance in human plasma (20 to 53 µM) following production and secretion by the hepatocyte (2). Serum levels of 1-AT can increase 3- to 4-fold during the hepatic acute-phase response, demonstrating enhanced production during acute inflammation (24). Although 1-AT diffuses throughout all tissue, protecting the lung parenchyma from protease destruction is its clinically relevant function. Lung disorders characterized by a relative excess of local proteases, such as 1-AT deficiency, cystic fibrosis, and the acquired disorders of tobacco-induced emphysema and bronchitis, can result in saturation of the protease shield and subsequent alveolar destruction (6, 11). Mature neutrophils are a primary provider of proteases, including neutrophil elastase and proteinase-3, to the lung matrix following activation or cell death (25). In addition to proteases provided by neutrophils, stimulated monocytes and macrophages actively secrete a number of proteases, including the metalloproteinase macrophage elastase as well as proteinase-3 (26, 27). Finally, lymphocytes present in our PBMC population synthesize and secrete granzymes which are also members of the serine protease family (28).

Recent work by Liou and Campbell has attempted to better define the nature of neutrophil elastase-mediated tissue destruction within the context of the lung microenvironment (29). Quantum release by neutrophils of azurophilic granules containing neutrophil elastase resulted in a rapid local increase of neutrophil elastase approximately two orders of magnitude greater than the local extracellular inhibitor (1-AT) concentration. The investigators concluded that the local antiprotease concentration in the lung is critical for confinement of pericellular microenvironment destruction to alveolar basement membranes.

In addition to hepatocyte-derived 1-AT, monocytes and alveolar macrophages have demonstrated the capacity to manufacture 1-AT in a regulated fashion (15, 16). The contribution to the lung antiprotease shield by mononuclear phagocytes has come under debate based on significantly lower amounts of 1-AT production when compared with hepatocytes. Nevertheless, investigators have suggested that mononuclear phagocyte-derived 1-AT may serve an important regulatory role in preventing protease destruction in the lung microenvironment when considering: (1) the direct proximity of the alveolar macrophage/monocyte to the basement membrane/interstitial matrix and (2) the large number of alveolar macrophages comprising approximately 10% of the alveolar cell population (14, 30).

Although 1-AT is the primary antiprotease that protects the lung from neutrophil elastase, additional antiprotease protection is provided by other sources. Work by Cichy and coworkers has demonstrated up-regulation of 1-antichymotrypsin production by primary lung epithelial cells following treatment with relevant pro-inflammatory factors, including IL-1 and IL-6 (31, 32). 1-Antichymotrypsin specifically inactivates neutrophil-derived cathepsin G and mast cell-derived chymase. Therefore, regulation of the protease-antiprotease balance in the lung is multifactorial, involving multiple cells that can up-regulate antiprotease protection during an acute-phase response.

Up-regulation of 1-AT in Monocytes

Up-regulation of monocyte/macrophage-derived 1-AT particularly during local inflammation in lung tissue may play a critical role in confining the quantum release and damage produced by neutrophil elastase or other relevant proteases such as proteinase-3. The local release of bacterial endotoxin and/ or the early production of potent inflammatory mediators such as IL-1 and TNF in lung tissue may also influence the extent of protease damage. In this study, we have demonstrated the exquisite sensitivity of monocyte 1-AT up-regulation to endotoxin ( 40 pg/ml). Additionally, this investigation is the first to demonstrate up-regulation of 1-AT protein in mononuclear cells by the cytokines IL-1 and TNF and then accurately quantitate the amount released. After treatment of PBMC with LPS, IL-1, or TNF, a similar 2- to 3-fold enhancement of 1-AT production over baseline, characteristic of an acute-phase response, was appreciated. In our experiments, only monocytes present in the PBMC cell population showed immunohistochemical detection of 1-AT. The effects of pro-inflammatory factors were shown to be specific in that coincubation of LPS-treated monocytes with polymyxin-B, IL-1-treated monocytes with sIL-1 IIr, and TNF-treated monocytes with an anti-TNF monoclonal antibody significantly decreased 1-AT in supernatants. Two lines of evidence strongly suggest that increased 1-AT was the result of gene activation and newly synthesized protein. Up-regulation of mRNA occurred in a dose-responsive manner and correlated with the level of 1-AT protein measured in the conditioned PBMC supernatants, and the increase in 1-AT over baseline production by LPS, IL-1, and TNF was completely inhibited by actinomycin-D pretreatment, an agent known to specifically inhibit gene transcription in mammalian cells.

Perhaps most intriguing was our inability to detect unbound 1-AT in PBMC supernatants under all conditions studied. The consistent detection of one 80 kD polypeptide that was immunoreactive with antibody against human 1-AT suggested that no functional 1-AT was present in the supernatants taken from PBMC. In accordance with our previous findings, both control and treated supernatants had residual protease activity confirming the absence of antiprotease activity. Furthermore, supernatants harvested from LPS-, IL-1-, and TNF-stimulated PBMC possessed less free protease activity when compared with untreated control samples. We believe this to likely be the result of increased complexation with 1-AT following increased expression of the antiprotease. Similar results were obtained from a second study volunteer.

Possible Role for Up-regulation of 1-AT in Monocytes

Our findings are in disagreement with previously reported results that have suggested an inability of monocytes to up-regulate 1-AT in the presence of IL-1 and TNF (16). These discrepancies may be explained by differences in the environments studied. We have focused our studies on PBMC cells in suspension during a relatively short 16-h period. Furthermore, all cytokine experiments were performed in the presence of polymyxin B to eliminate the potential for endotoxin contamination, which, as demonstrated, can significantly influence 1-AT detection in cell supernatants. Our findings are consistent with previous work by Perlmutter and colleagues demonstrating up-regulation of 1-AT by monocytes following endotoxin treatment or stimulation with other acute-phase cytokines such as interleukin-6 (12, 16).

When comparing mRNA from LPS-, IL-1-, and TNF-stimulated PBMC, disproportionate increases in 1-AT transcripts occurred, implicating the role for alternative pathways of 1-AT regulation. Previous investigations have identified three distinct species of 1-AT mRNAs present in monocytes under basal and modulated conditions (15). Our results from Northern analysis and actinomycin-D pretreatment show compelling evidence that up-regulation of 1-AT transcription does occur; however, we were not able to detect different species of 1-AT mRNA transcripts when comparing control and stimulated PBMC. In fact, one approximately 1.8 kilobase band was consistently detected in stimulated and unstimulated monocytes. This data in comparison to previous findings may best be explained by our inability to sensitively detect different mRNA species by Northern analysis. Of interest, an NF-B regulatory domain has been identified in the human 1-AT gene approximately 1,800 base pairs upstream of the transcription initiation site (33). Description of this transcriptional regulatory domain in the human 1-AT gene may provide an explanation to the results reported in the current study. NF-B encompasses a family of pleiotropic inducible transcription factors known to be activated by a number of extracellular signals and cytokines, including IL-1 and TNF (34).

In addition to providing antiprotease activity, monocytes contribute to the protease burden through release of other relevant proteases. Therefore, it is possible that 1-AT produced by the monocyte could primarily be involved in neutralization of proteases provided by monocytes or other cells present in the PBMC population. A similar phenomenon in neutrophils has recently been reported. In this study, the authors have suggested that intracellular regulation of the protease-antiprotease balance may also serve a role in cellular migration (35). To our knowledge, one other report has demonstrated the detection of 1-AT/protease complexes in the supernatants of unstimulated monocytes (12). In our analysis, human serum- derived 1-AT (Sigma Chemical Co.) (52 kD) was readily detected as a 52 kD protein with the same antibody. When native 1-AT was incubated with an excess of human neutrophil elastase and compared with serum-free PBMC supernatants, the protease-antiprotease complex was detected at the exact same mass as the PBMC-derived complex. Identical results were obtained when cell lysates taken from the same samples were analyzed. Furthermore, dissociation of the complex was not demonstrated when samples were coincubated with serine protease inhibitors. On the other hand, the inhibitors were able to prevent exogenous 1-AT and neutrophil elastase from forming a complex at similar concentrations. Therefore, we believe the additional mass of the complex to be a product of 1-AT binding to a protease provided by monocytes and approximately 29 kD in size.

Initial attempts to demonstrate functional (unbound) 1-AT activity in PBMC supernatants failed. Based on a lack of functional antiprotease activity and our inability to detect free 1-AT by Western analysis, we chose to measure elastase-like activity in serum-free PBMC supernatants. The serum-free samples derived from both stimulated and unstimulated PBMC were shown to have elastase-like activity that was inversely related to the amount of 1-AT measured in supernatants. However, we cannot exclude that protease concentrations may have varied in the different PBMC conditions and cannot exclude lymphocytes as a source of protease activity.

In summary, low-dose LPS and physiologically relevant doses of IL-1 and TNF up-regulate the expression of 1-AT secreted into the supernatants of PBMC. Up-regulation occurs in a dose-responsive manner resulting in a 2- to 3-fold increase in 1-AT secretion under the conditions studied. These findings further demonstrate the role of the monocyte as the provider of an accessory acute-phase response, particularly with respect to protease protection. Regulation of peripheral blood monocyte-derived 1-AT by IL-1 and TNF adds a new perspective to control of the local antiprotease microenvironment. As pro-inflammatory cytokines or LPS are introduced early into the local environment, recruitment of leukocytes via IL-8 production may be closely shadowed by a concomitant up-regulation of 1-AT. This in turn may serve as one of several early front-line defense mechanisms designed to protect the integrity of the alveolar tissue. Furthermore, regulation of 1-AT release from monocytes may primarily serve as a microenvironmental front-line defense against proteases provided by the monocyte.