INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: fibrosing alveolitis; NF-B; gene transcription; pulmonary inflammation

Fibrosing alveolitis (FA) is a chronic inflammatory process involving the pulmonary interstitium and alveoli, resulting in excessive collagen deposition and impaired organ function. The current understanding of disease pathogenesis envisages that FA is initiated by alveolar epithelial and/or capillary endothelial cell injury by inhaled particles (e.g., asbestos), physical injury (e.g., radiation), or antigens that are deposited in the pulmonary vasculature (e.g., drugs like bleomycin, amiodarone, or nitrofurantoin) (1-3). Frequently, an initiating antigen is not identified, and the term "cryptogenic fibrosing alveolitis" is applied to the disease. These diverse antigens trigger a common inflammatory process through cytokine, chemokine, and adhesion molecule production by alveolar macrophages (AMs) and other immune effector cells within the lung. Chemokines and adhesion molecules recruit monocytes, neutrophils, and eosinophils to the lower airway, while cytokines amplify lung injury through the activation of these cells and fibroblasts that deposit collagen (4, 5). The disease process is frequently progressive, resulting in a five year mortality of 50% and chronic respiratory morbidity in many survivors (6). The failure of existing therapies to significantly improve survival highlights the need for new treatment strategies.

There are two lines of evidence to support the hypothesis that NF-B activation is a critical event in the pathogenesis of FA. First, NF-B activation is required for the expression of cytokines (e.g., tumor necrosis-factor [TNF]-, interleukin [IL]-6), chemokines (e.g., IL-8), and critical enzyme systems (e.g., nitric oxide synthetase) that earlier studies have determined to be important in the pathogenesis of FA (7, 8). Second, there is now data from animal models directly linking nuclear factor-B (NF-B) activation to the pathogenesis of FA, which contrast the low basal levels of NF-B activity in normal AMs (9) with the increased activity of the transcription factor in AMs and respiratory epithelial cells of rats with FA (10, 11). It has been proposed that selective inhibition of molecular pathways that regulate NF-B activation may be a future therapeutic strategy in FA and other chronic inflammatory lung diseases, including chronic obstructive pulmonary disease (COPD) and acute respiratory distress syndrome (ARDS) (12, 13).

The cytoplasmic binding of NF-B to IB proteins prevents gene transcription under resting conditions (14, 15). A kinase cascade that regulates IB degradation through phosphorylation of two N-terminal serine residues activates NF-B-dependent gene transcription (16). IB phosphorylation allows recognition of the protein by a specialized E3 ubiquitin ligase complex (E3^IB) that ubiquinates lysine residues, targeting the molecule for degradation by the 26S proteosome (17, 18). Once free of IB, NF-B is free to translocate into the nucleus and activate gene transcription. Serine phosphorylation is the only regulated step of IB degradation (19), identifying this process as a potential target for antiinflammatory therapies.

Numerous studies involving cell lines and transgenic mice suggest that TNF- and IL-1 degrade IB though a common kinase cascade (20-29). Binding of TNF- to the type 1 receptor (TNFR1) recruits the receptor-associated antigens, TNF receptor-associated death domain protein (TRADD), TNFR-associated factor-2 (TRAF-2), and receptor-interacting protein (RIP) to the cell surface. Similarly, binding of IL-1 to its type 1 receptor (IL-1R1) and receptor accessory protein (AcP) facilitates an interaction between IL-1 receptor-associated kinase (IRAK) and TRAF-6. Both sets of receptor-associated proteins form an active signaling complex that binds to NF-B-inducing kinase (NIK). The requirement for NIK in TNF- and IL-1 mediated NF-B activation was established by Malinin and colleagues, when over-expression of kinase defective NIK in a 293-human embryonic kidney (EBNA) cell line was shown to inhibit nuclear activity of the transcription factor (20). NIK itself does not directly phosphorylate IB, but rather activates an intermediate "IB kinase" (IKK) complex that performs this function (21, 22). IKK is composed of two structural proteins, IKK (or NEMO) and IKK complex-associated protein (IKAP) and two catalytic subunits (IKK and IKK), the functions of which have been largely determined using cell lines and knock-out mice (23-25, 27). Over-expression of catalytic subunits in cell lines indicates that IKK, but not IKK, is required for TNF- and IL-1 mediated IB degradation. The death of IKK knock out (IKK^/) mice in utero of hepatocyte apoptosis, due to the absence of TNF--induced NF-B activation, provides further evidence that IKK is critical for cytokine mediated NF-B activation (25). The function of IKK is less clear, but based on studies involving IKK^/ mice, it is likely to regulate NF-B activity during cellular differentiation (26).

Although studies involving IKK^/ mice have provided important information concerning the function of the IKK complex, the role of NIK is less certain. Recent studies involving transgenic mice that express a defective NIK with a mutation at the TRAF2 interacting site ("aly/aly mice") suggest that, at least within murine cells, NIK may be only selectively required for lymphotoxin and not TNF- or IL-1-mediated NF-B activation (30-32). Furthermore, we have previously demonstrated that other stimuli, also likely to be involved in the activation of AMs that occurs in FA, result in proinflammatory gene transcription independent of NF-B (33, 34). Taken together, this evidence suggests that the existing model of NIK function, derived from studies involving transformed cell lines, may not be applicable to fully differentiated primary macrophages. Before selective inhibition of proinflammatory gene transcription can be considered as a therapeutic option in pulmonary disease, the mechanism of NF-B activation that operates in primary human AMs will need to be determined. In particular, defining the point at which TNF- and IL-1- responsive kinase cascades converge, as inhibition of the pathways before this point is unlikely to influence NF-B dependent gene expression in vivo.

Recently, we demonstrated successful in vitro delivery of Adv transgenes to AMs obtained from patients with fibrotic lung disease, but not from normal volunteers (35). The differential transgene expression is related to upregulation on diseased AMs of the CAR, v3, and v5 integrins that are required for Adv infection (36, 37). The purpose of this study was to exploit this property of diseased AMs to deliver transgenes encoding components of the kinase cascade, believed to regulate NF-B activation, to investigate the disease-specific mechanisms of proinflammatory gene transcription in FA. Substitution of 429/430 lysine for alanine within the ATP binding site of the NIK molecule (NIKkd) and substitution of alanine 44 for arginine within the kinase domain of the IKK subunit (IKKkd) produces proteins previously determined to be kinase-defective when expressed in both primary cells and transformed cell lines (20, 38, 39). Using Adv constructs encoding these kinase defective proteins (AdvIKKkd and AdvNIKkd), we determined that constitutive and cytokine mediated NF-B activation and IL-6 production requires IKK, but not NIK. This study provides evidence that functionally important IKK kinases, other than NIK, contribute to constitutive and cytokine induced NF-B activation in primary human macrophages. These findings suggest that therapies targeting components of the kinase cascade above IKK are unlikely to effectively inhibit proinflammatory gene transcription.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cells

BALs were obtained from patients undergoing diagnostic bronchoscopy for suspected FA (40). For inclusion, patients were required to fulfill the criteria for the diagnosis of cryptogenic FA (41). BALs were collected in a siliconized media bottle, centrifuged (1,500 rpm for 10 minutes), washed, and resuspended in serum free Roswell Park Memorial Institute medium (RPMI) (5 × 10⁶ cells/ml). An enriched AM population was obtained using magnetic polystirene beads coated with mAb to CD2 (Dynal CD2 CELLection kit; Dynal, Bromborough, UK). Beads were added at five times the number of alveolar T cells calculated by haemocytometer and incubated for 30 minutes on a rotator before transfer onto a magnetic particle concentrator (Dynal) for 3 minutes to facilitate attachment of the rosetted T cells to the test tube wall. The fluid containing the negatively selected AMs was then aspirated. BALs processed using this technique were routinely determined to contain more than 97% AMs by fluorescence activated cell sorter (FACS) analysis.

Adv Constructs

Adv constructs were generated according to the protocol described by He and colleagues (42). Constructs encoded wild-type (wt) and kinase defective (kd) NIK, (AdvNIKwt and AdvNIKkd; Prof Liellach, Weizmann Institute, Givatoyim, Israel), a "Green Fluorescent Protein" (GFP) reporter (AdvGFP; Dr Mahon, KIR, London, UK), an IKKkd, and porcine IB (AdvIKKkd and AdvIB; Dr de Martin, Vienna, Austria). A FLAG tag was incorporated into the IKKkd, NIKwt, and NIKkd genes. Substitution of alanine 44 for arginine within the IKK kinase domain, and lysine 429/430 for alanine within the NIK ATP binding site, produced kinase-defective proteins (20, 38). Porcine IB has more than 95% homology with the human molecule and associates with human NF-B (43). Constructs were propagated in 293 human embryonic kidney cells and purified by ultra-centrifugation through cesium chloride gradients. The virus titer was determined by plaque assay in 293 cells.

Infection Techniques

AMs were infected for 60 minutes at a multiplicity of infection (m.o.i.) of 150 plaque forming units (pfus)/cell in serum free RPMI. The medium-containing virus was then removed and replaced with complete media (RPMI with 5% fetal calf serum (FCS), 25 mM Hepes, 2 mM L-glutamine and 100 units/ml penicillin/streptomycin).

Western Immunoblotting

IKK and IB were analyzed by Western Immunoblotting, while immuno-precipitation was performed to analyze NIK expression. After 48 hours, AMs were removed from the culture plate using "Cell Dissociation Solution" (Sigma, Poole, UK). Cytosolic and nuclear extracts were prepared as described by Whiteside and colleagues (44). Protein was quantified by Bradford assay and 100 µg loaded for SDS/ PAGE separation on a 10% (wt/vol) polyacrylamide gel, before electrotransfer onto polyvinyl difluoride (PVDF) membranes (Millipore, Bedford, MA). -IB, IKK, and IKK mAbs (Santa Cruz Biotechnology, Santa Cruz, CA) were used as primaries, while the secondary was a horseradish peroxidase-conjugated (HRP) donkey -rabbit (Amersham International, Oxford, UK). An -FLAGM2-agarose affinity gel (Sigma, Poole, UK) was used for immunoprecipitation of NIK proteins, with an -NIK (Santa Cruz Biotechnology, Santa Cruz, CA) and HRP-conjugated -goat (Dako, Cambridge, UK) as the respective primaries and secondaries.

Electrophoretic Mobility-Shift Assay

Twenty-four hours after infection with the stated Adv construct, AMs were treated with either TNF- 10 ng/ml or IL-1 10 ng/ml before being dissociated from the wells and the nuclear proteins extracted as described by Dent and Latchman (45). The cells were lysed in hypotonic buffer (0.5% Nonidet P-40, 10 mM Hepes [pH 7.9], 10 mM KCL, 1 mM DTT, 2 mM PMSF, 30 µg/ml leupeptin, 10 µg/ml aprotinin, 10 µg/ml pepstatin) before the nuclei were harvested by centrifugation (1,200 rpm for five minutes) and resuspended in a hypertonic extraction buffer (5 mM Hepes [pH 7.9], 1.5 mM MgCl₂, 0.2 mM EDTA, 0.5 M NaCl, 25% glycerol, 1 mM DTT, 2 mM PMSF, 30 µg/ml leupeptin, 10 µg/ml aprotinin, 10 µg/ml pepstatin). Samples were agitated for 60 minutes at 4° C before being centrifuged (12,000 rpm for 10 minutes), and the supernatant containing the nuclear proteins aspirated. Protein concentration was quantified by Bradford assay and 20 µg run on a 5% TBE gel with a [-³²P]-ATP labeled NF-B consensus oligonucleotide (Promega, Madison, WI). The gels were dried onto filter paper and retarded DNA complexes visualized using Hyperfilm (Amersham Pharmacia Biotech, Cambridge, UK). Competition and supershift assays were performed to confirm that the highlighted bands were NF-B. Competition assays involved incubating the nuclear extracts with [-³²P]-ATP labeled NF-B and either 100 × unlabeled NF-B or AP-1 (non-specific) oligonucleotides. For the supershift assays, nuclear proteins were incubated for two hours at 4° C with antibodies specific to the known NF-B components (p50, RelB, c-Rel and p65) before the [-³²P]-ATP labeled NF-B oligonucleotide was added.

Cytokine Analysis

Supernatants were harvested at 24 hours and centrifuged at 1,500 rpm. TNF-, IL-6 and IL-8 production at 24 hours was analyzed by sandwich enzyme-linked immuosorbent assay (ELISA) as previously described (32).

Statistical Methods

Cytokine production by Adv infected cells was expressed as a percentage of that produced by uninfected cells. The mean and standard deviation (SD) of percentage cytokine production by Adv-infected cells was then calculated relative to uninfected cells from the same specimen.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Selective Over-Expression of IKK, NIK, and IB in AMs

The efficiency of Adv-mediated delivery of transgenes encoding IKK and IB to primary AMs was assessed by western immunoblotting. Low levels of NIK expression and the absence of a high affinity anti-NIK monoclonal requires the use of immunoprecipitation for analysis. Figure 1 demonstrates the cytosolic expression of the protein of interest 48 hours after infection with the stated Adv construct. Over-expression of (a) NIK, (b) IKK, and (c) IB was detected in AMs infected with the AdvNIKwt and kd, AdvIKKkd, and AdvIB constructs respectively. Endogenous NIK was undetected in uninfected and AdvGFP infected cells. Endogenous IKK expression was unaffected by IKK over-expression.

View larger version (61K): [in this window] [in a new window]   Figure 1.   Cytosolic over-expression of Adv transgenes cytosolic extracts prepared from 4 × 106 AMs were analyzed for NIK, IKK, IKK, and IB expression by western immunoblotting and immuno-precipitation. Cells were either uninfected (negative control) or infected with AdvGFP (positive control), AdvNIKkd/wt, AdvIKKkd, or AdvIB at a multiplicity of infection of 150 plaque forming units:1 (A) Over-expression of NIK was detected in cells infected with AdvNIKwt and AdvNIKkd, but not uninfected or AdvGFP infected cells. Similarly over-expression of (B) IKK and (C) IB was detected only in AdvIKK and AdvIB infected cells respectively. Endogenous IKK expression was not influenced by infection with AdvIKK.

Over-Expression of IKKkd and IB, but Not NIKkd, Inhibits Constitutive NF-B Activation in AMs

In contrast to normal AMs, there is significant constitutive NF-B activity in diseased AMs (9, 35, 46). To determine if cytoplasmic over-expression of IKKkd, IB, and NIKkd would influence this constitutive NF-B activity, electrophoretic mobility-shift assay (EMSAs) were performed on the nuclear extracts of uninfected, AdvGFP, AdvNIKkd, AdvIKKkd, and AdvIB infected AMs. Competition assays (Figure 2) indicated that the two bands are NF-B specific with a subsequent supershift assay confirming that the upper band (arrow) is the p50/p65 heterodimer and the lower band represents unknown NF-B components (see online data supplement). NF-B activity was inhibited by IKKkd, and IB, but unaffected by NIKkd and GFP over-expression.

View larger version (100K): [in this window] [in a new window]   Figure 2.   IKKkd and IB, but not NIKkd over-expression inhibits constitutive NF-B activation in AMs. EMSAs were performed on the nuclear extracts of 4 × 106 AMs using a P32 labeled NF-B consensus oligonucleotide 48 hours after infection with the stated Adv construct to assess the effect of NIKkd, IKKkd, and IB over-expression on constitutive NF-B activity. Constitutive NF-B activation was reduced by over-expression of IKKkd and IB, but not NIKkd or infection with AdvGFP.

Infection of AMs with AdvIKKkd and AdvIB, but Not AdvNIKkd, Inhibits Constitutive Cytokine Production

The data suggest that constitutive NF-B dependent gene expression by AMs from patients with FA would be inhibited by IKKkd and IB, but not by NIKkd or GFP over-expression. To test this hypothesis, the constitutive production of TNF-, IL-6, and IL-8 by AMs, obtained from five consecutive BAL specimens that were either uninfected or infected with the AdvGFP, AdvNIKkd, AdvIKKkd, and AdvIB constructs, was analyzed (Figure 3). TNF- production (n = 5, range = 401-1390 pg/ml) by AdvIKKkd and AdvIB infected AMs was reduced to 34.1 ± 6.6% and 32.4 ± 11.1% of control, respectively, whereas TNF- production by AdvNIKkd infected cells was only minimally reduced to 84.5 ± 8.6% (Figure 3A). IL-6 production (n = 5, range = 432-3821 pg/ml) by AdvIKKkd and AdvIB infected AMs was reduced to 24.1 ± 8.2% and 26.1 ± 14.2% of control, respectively, whereas IL-6 production by AdvNIKkd infected cells was unaffected (98.0 ± 8.7%)(Figure 3B). IL-8 production (n = 5, range = 47,268- 6,0034 pg/ml) by AdvIKKkd and AdvIB infected AMs was reduced to 32.4 ± 9.4% and 22.4 ± 5.7% of control, respectively, whereas AdvNIKkd infection (86.4 ± 3.8%) had minimal effect on chemokine production (Figure 3C). There was no difference in either TNF-, IL-6, or IL-8 production by AdvGFP infected AMs relative to uninfected cells.

View larger version (29K): [in this window] [in a new window]   Figure 3.   IKKkd and IB, but not NIKkd over-expression inhibits constitutive cytokine production by AMs. Constitutive (A) TNF-, (B) IL-6, and (C) IL-8 production by AMs that were left uninfected or infected with the stated Adv construct at a titer of 150 plaque forming units:1 was analyzed by ELISA at 24 hours. (Bottom panel ) The mean percentage cytokine production by cells infected with the AdvGFP, AdvNIKkd, AdvIKKkd, and AdvIB constructs was calculated relative to uninfected cells from the same specimen (error bars = ± 1 SD).

Over-Expression of IKKkd and IB, but Not NIKkd, Inhibits TNF- Mediated NF-B Activation in AMs

The data presented so far indicate that NF-B activation in AMs is IKK, but not NIK, dependent. The in vivo stimuli involved in this process are not, however, fully understood, and may involve cytokines (e.g., TNF-) that have been shown in transformed cell lines to activate NF-B via NIK dependent pathways (20, 47, 48). To further define the role of NIK, IKK, and IB in NF-B activation, NF-B activity was analyzed in TNF--treated AMs infected with AdvGFP, AdvNIKkd, AdvIKKkd, or AdvIB. Paired wells containing 4 × 10⁶ AMs from a single BAL were infected with the stated Adv construct. Forty-eight hours later, one well from each pair was treated with TNF- 10 ng/ml for 45 minutes before all cells were lysed and nuclear extracts prepared (Figure 4). NF-B activation in AdvGFP infected AMs following treatment with TNF- was largely inhibited by over-expression of IKKkd and IB, but not NIKkd.

View larger version (95K): [in this window] [in a new window]   Figure 4.   IKKkd and IB, but not NIKkd over-expression inhibits cytokine mediated NF-B activation in AMs. Paired wells containing 4 × 106 AMs were infected with AdvGFP, AdvNIKkd, AdvIKKkd, and AdvIB. EMSAs were performed using a P32 labeled NF-B oligonucleotide. The increased NF-B activity observed in AdvGFP infected cells following TNF- activation, was largely inhibited by IKKkd and IB, but not NIKkd over-expression.

Infection of AMs with AdvIKKkd and AdvIB, but Not AdvNIKkd, Inhibits Cytokine-induced IL-6 Production

The data indicate that cytokine-mediated NF-B activation in AMs requires a functional IKK subunit, but not NIK. To further define the role of these two kinases in the activation of NF-B, IL-6 production by uninfected and AdvGFP, AdvNIKkd, AdvIKKkd, or AdvIB infected AMs, treated with either TNF- or IL-1 from three to five consecutive FA patients, was analyzed. There was no difference in IL-6 production by AdvGFP and AdvNIKkd infected AMs relative to uninfected cells. In contrast, IL-6 production by AdvIKKkd and AdvIB infected AMs was substantially reduced (Table 1 and Figure 5). These results confirm that over-expression of IKKkd and IB, but not NIKkd, inhibits NF-B-dependent proinflammatory gene transcription in primary AMs.

View larger version (29K): [in this window] [in a new window]   Figure 5.   Infection of AMs with AdvIKKkd and AdvIB, but not AdvNIKkd inhibits cytokine induced IL-6 production. AMs that were left uninfected or infected with the stated Adv construct (multiplicity of infection of 150 plaque forming units:1) were activated with TNF- 10 ng/ml or IL-1 10 ng/ml. The mean percentage IL-6 production by cells infected with AdvGFP, AdvNIKkd, AdvIKKkd, and AdvIB was calculated relative to uninfected cells from the same specimen (error bars = ± 1 SD). There was no significant difference in IL-6 production by uninfected, AdvGFP, and AdvNIKkd infected AMs. In contrast, IL-6 production by AdvIKKkd and AdvIB infected AMs was substantially reduced relative to uninfected and AdvGFP infected cells.

Because the molecular pathways involved in cytokine gene transcription depend not only on the mechanism of cell activation, but also lineage (49-51), it was possible that the data obtained from AMs relating to the function of NIK, IKK, and IB are not applicable to other primary human cells. To investigate this possibility, identical studies were performed using human umbilical vein endothelial cells (HUVECs) that, like diseased AMs, are permissive to Adv infection. Again, IL-6 production by AdvIKKkd and AdvIB, but not AdvGFP or AdvNIKkd infected cells, was reduced relative to uninfected cells (see online data supplement).

NF-B Activation in AMs Secondary to NIKwt Over-Expression Is Inhibited by Coinfection with AdvNIKkd, AdvIKKkd, and AdvIB

The failure to demonstrate an absolute requirement for NIK in constitutive and cytokine mediated NF-B activation suggests the presence of an alternative mitogen-activated protein-3 kinase (MAP3K), able to degrade IB. Alternatively, despite previous studies demonstrating that the substitution of lysine 429/430 for alanine within the NIK ATP binding site produced a kinase defective protein in both primary cells and transformed cell lines (20, 38), it was possible that this protein was functional in AMs and HUVECs. AdvNIKkd infected AMs were also treated with zymosan, lipopolysacharide (LPS), and -CD45 mAb in an attempt to identify a stimulus for NF-B activation inhibited by NIKkd over-expression, and therefore establish the kinase-defective nature of NIKkd in these cells. TNF- production by AdvNIKkd infected AMs was not reduced relative to uninfected or AdvGFP infected AMs (results not shown), failing to confirm that the NIKkd was kinase defective. Transfection of NIKwt into transformed cell lines has been shown to activate NF-B-dependent gene transcription (52, 53). In preliminary studies, Adv-mediated delivery of NIKwt to AMs was also shown to be a potent stimulus for NF-B dependent gene expression, with a progressively increasing titer of AdvNIKwt resulting in a dose-dependent augmentation of IL-6 production (see online data supplement). It was possible that this stimulus for NF-B activation would be inhibited by coexpression of NIKkd.

To validate data obtained from studies involving coinfection of AMs with two different Adv constructs, it was first necessary to demonstrate that coexpression of two virally encoded proteins was possible. Western immunoblotting and immunoprecipitation was performed on the cytosolic extracts of AMs 48 hours after coinfection with AdvNIKwt and either AdvGFP, AdvNIKkd, AdvIKKkd, or AdvIB. Over-expression of two virally encoded proteins was demonstrated in AMs coinfected with AdvNIKwt, AdvNIKkd, AdvIKKkd, and AdvIB (Figure 6).

View larger version (63K): [in this window] [in a new window]   Figure 6.   Coexpression of Adv gene products following infection of AMs with AdvNIKwt and either AdvNIKkd, AdvIKKkd, or AdvIB. Wells containing 4 × 106 AMs were infected with the stated Adv constructs. After 48 hours the cells were dissociated from the plates, lysed, and cytosolic extracts prepared. Cytosolic coexpression of Adv transgenes was demonstrated with western immunoblotting and immunoprecipitation.

EMSAs were performed on the nuclear extracts of AMs that were uninfected, infected with GFP or AdvNIKwt alone, or coinfected with the stated Adv constructs (Figure 7). As expected, there was augmented NF-B activity in cells infected with AdvNIKwt alone relative to uninfected and AdvGFP infected cells. There was no difference in the NF-B activity of AMs infected with AdvNIKwt alone compared with those coinfected with AdvNIKwt and AdvGFP, indicating that the higher virus titer per se (total m.o.i. 200 pfus) had little effect on NF-B activity. NF-B activity in AMs coinfected with AdvNIKwt and either AdvNIKkd, AdvIKKkd, or AdvIB was similarly reduced indicating that when the stimulus was NIKwt, NIKkd was as effective as IKKkd and IB in inhibiting NF-B activity.

View larger version (76K): [in this window] [in a new window]   Figure 7.   Coinfection of AMs with AdvNIKkd, AdvIKKkd, and AdvIB inhibits the augmented NF-B activation produced by NIKwt over- expression. EMSAs performed on the nuclear extracts of 4 × 106 AMs that were uninfected or infected with the stated Adv constructs indicate that the NF-B activation secondary to NIKwt over-expression is attenuated by coexpression of the NIKkd, IKKkd, and IB virally encoded proteins.

IL-6 Production by AMs Activated by NIKwt Over-Expression Is Inhibited by Coinfection with AdvNIKkd, AdvIKKkd, and AdvIB

To provide further evidence that the NIK encoded by the AdvNIKkd construct was kinase defective, IL-6 production by AMs coinfected with AdvNIKwt and either AdvGFP, AdvNIKkd, AdvIKKkd, or AdvIB was analyzed. Augmented IL-6 production was demonstrated in the AdvNIKwt infected AMs relative to the uninfected (negative control) and AdvGFP infected cells (positive control) (see online data supplement). The mean and SD of percentage IL-6 production by cells coinfected with the stated Adv construct was calculated relative to cells infected with AdvNIKwt alone. IL-6 production by AMs coinfected with AdvNIKwt and AdvGFP did not significantly differ from cellular preparations infected with AdvNIKwt alone, indicating that increased viral titer per se had little effect on IL-6 gene expression (Figure 8). In contrast, IL-6 production by AMs coinfected with AdvNIKwt and either AdvNIKkd, AdvIKKkd, or AdvIB was reduced to 23.7 ± 13.9%, 24.1 ± 4.9%, and 22.1 ± 7.2%, respectively (range: 4,832-6,943pg/ml). Again, identical studies performed using HUVECs produced similar results, with AdvNIKkd, AdvIKKkd, or AdvIB coinfection resulting in inhibition of IL-6 gene expression (see online data supplement).

View larger version (23K): [in this window] [in a new window]   Figure 8.   IL-6 production by AMs that over-express NIKwt is inhibited by coinfection with AdvNIKkd, AdvIKKkd, and AdvIB. IL-6 production by AMs infected with AdvNIKwt alone or coinfected with AdvNIKwt and either AdvGFP, AdvNIKkd, AdvIKKkd, or AdvIB was measured at 24 hours. The mean percentage IL-6 production by cells coinfected with the stated Adv constructs was calculated relative to cells infected with AdvNIKwt alone (error bars = ± 1 SD). IL-6 production by cells infected with AdvNIKwt alone and cells coinfected with AdvNIKwt and AdvGFP did not significantly differ. In contrast, IL-6 production by cells coinfected with AdvNIKwt and either AdvNIKkd, AdvIKKkd, or AdvIB was substantially reduced.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study, we have investigated the molecular mechanisms of NF-B activation in primary AMs using a novel Adv-mediated gene delivery system. Upregulation of the CAR, v, and v5 integrins on AMs obtained from patients with fibrosing lung diseases allows efficient Adv infection and analysis of molecular signaling pathways (35). Using Adv constructs encoding a NIK and IKK subunit previously determined to be kinase defective (20, 38, 39), we demonstrated that constitutive and cytokine induced NF-B activation by AMs requires a catalytically active IKK subunit, but not NIK dependent signaling.

Adv mediated over-expression of IB in primary human macrophages inhibits nuclear translocation of NF-B and IL-6 production, providing direct evidence that NF-B activation is required for cytokine gene transcription (33, 35, 43, 54). The observation that Adv mediated over-expression of IKKkd also inhibits NF-B activation and cytokine production establishes a similar requirement for IKK in the process of NF-B activation within primary AMs. The current model of NF-B activation is based largely on studies involving cell lines and transgenic mice that die in utero and envisages that IKK phosphorylation is required for cytokine mediated IB degradation (24, 25). Some studies have suggested, however, that the role of IKK may depend on cell lineage and the state of cellular differentiation (22, 55). For example, Fischer and coworkers reported that LPS induced activation of IKK in THP-1 monocytes occurred primarily as a result of IKK phosphorylation, but TNF- and IL-1 treatment resulted in rapid IKK phosphorylation with only minimal alteration in IKK activity (55). As the mechanism NF-B activation in FA is likely to involve TNF- and IL-1 (4, 56), it was possible that a functional IKK was not required for IB degradation in AMs. The data presented in this study does, however, support this existing model of IKK function, indicating that the different mechanisms of IKK activation observed in macrophage cell lines are not of functional significance in primary AMs. We have, however, noted that LPS induced NF-B activation in fibroblasts and peripheral blood monocytes is only minimally affected by IKKkd over-expression (unpublished observations), suggesting that there could be variation in IKK function between different primary macrophages. This study is important because it confirms that IKK is required for NF-B activation in primary AMs and suggests that Adv- mediated transgene delivery can be used to investigate possible differential IKK function.

The failure of NIKkd to inhibit TNF and IL-1 mediated NF-B activation was unexpected as, when delivered to HeLa and A293 cell lines by transfection and Adv constructs, this protein has been shown to effectively inhibit NF-B activation (20, 38). The observation that constitutive and cytokine mediated NF-B dependent gene expression in AMs does not require NIK, highlights the limitations of studies involving cell lines in predicting disease specific molecular signaling pathways within primary human cells. NIK was originally identified as the critical IKK kinase required for cytokine mediated NF-B activation by Malinin and coworkers, who demonstrated that in 293-EBNA cells over-expression of the same NIK used in this study ablated TNF--induced NF-B activation (20). A number of subsequent studies involving cell lines and utilizing NIKkd constructs also demonstrated a similar requirement for NIK in NF-B-dependent gene transcription (47, 48, 57). Recent studies involving the aly/aly mouse that expresses a NIK with a single amino acid substitution within the C-terminal interaction domain, preventing association with TRAF2, have, however, called into question the applicability of this data to the molecular regulation of NF-B within primary cells. The aly/aly phenotype is characterized by disorganized thymic and splenic structure, absent lymph nodes, and humoral immunodeficiency that can be completely reversed by NIKwt expression (30). Despite expressing a NIK that does not recognize TRAF2 and 6, the aly/aly mouse and embryonic fibroblasts still activate NF-B in response to TNF- and LPS, suggesting that there is an alternative MAP3K capable of activating IKK. It is possible that the mitogen activated protein kinase/ERK kinase kinase (MEKK1) is the alternative MAP3K involved in cytokine mediated NF-B activation (Figure 9). MEKK1 is a component of the stress activated c-Jun N-terminal kinase (JNK) pathway that has been shown in one study to directly activate IKK (58). The observation that chimeric TRAF2 proteins that are unable to interact with NIK retain the ability to activate IKK provides further evidence that there are NIK-independent mechanisms of NF-B activation (59). Our study establishes that in primary AMs, NIK-independent mechanisms of IKK activation are functionally important and indicates that targeted inhibition of NIK is unlikely to reduce NF-B dependent gene expression.

View larger version (36K): [in this window] [in a new window]   Figure 9.   Putative mechanism of NF-B activation by TNF- and IL-1 involving the stress activated c-Jun N-terminal kinase (JNK) pathway. TNFR, tumor necrosis factor receptor; TRADD, TNFR-associated death-domain protein; TRAF2/6, TNFR-associated factor 2 and 6; RIP, receptor interacting protein; IL-1R, interleukin-1 receptor; ACP, accessory membrane spanning protein; IRAK, IL-1R-activated kinase; IKK, IB kinase complex; IKK/,  and  subunits of IKK; NEMO (or IKK), IKK docking protein; IKAP, IKK complex-associated protein; MEKK1, mitogen activated protein kinase/ERK kinase kinase; PAK, protein kinase; NIK, NF-B inducing kinase; p50/p65, NF-B heterodimer; P, phosphorylated serine residues; Ub, ubiquinated lysine residues.

There is mounting evidence that NF-B activation in AMs is important in the pathogenesis of many pulmonary diseases (7, 8). Elevated levels of NF-B have been detected in AMs obtained from patients with ARDS (13), but not from cells obtained from healthy volunteers (9). It has also been reported that in vitro exposure of AMs to asbestos fibers increases NF-B activity at sites within the promoter region of the IL-6 and IL-8 genes (60). Corticosteroids inhibit NF-B activation and highlight the central role of this transcription factor in the pathogenesis of pulmonary inflammation. Corticosteroids inhibit NF-B activation through upregulation of IB gene expression and an interaction between the ligand bound glucocorticoid receptor and NF-B (61, 62). It has been proposed that targeted inhibition of NF-B activation may provide the immunosuppressive benefits of corticosteroids, while avoiding many of the unwanted side effects. The data presented in this study indicates that inhibition of the kinase cascade above the level of IKK is unlikely to result in effective inhibition of NF-B-dependent gene expression. Our study indicates that strategies preventing IKK phosphorylation or degradation of IB would be more likely to attenuate the constitutive and cytokine mediated activation of NF-B in AMs.

Until recently, thoracic molecular research has focused on the role of the eosinophil and lymphocyte in the pathogenesis of airway-centered inflammation. The recognition by the World Health Organization that COPD will become the third most common cause of death in industrialized nations by 2020 (63) has, however, led to increased interest in the molecular mechanisms of this disease, which involves activation of AMs in the distal airspaces. Our study highlights the potential of Adv-mediated transgene delivery as a technique for investigating dysregulated AM function in pulmonary disease. Using Adv-mediated gene delivery, we have made some potentially, clinically relevant observations regarding NF-B regulation in AMs. We have also achieved more than 95% -galactosidase reporter gene expression in freshly prepared AMs from patients with COPD and ARDS (unpublished observations). It is likely, therefore, that the inflammatory process in COPD and ARDS also upregulates the CAR and v integrins and will permit the application of this technique to the investigation of the molecular mechanisms of inflammation in these conditions.

In this study, we have established in AMs that NIK is not required for constitutive or cytokine-mediated activation of IKK, and that IKK is the critical IKK subunit necessary for NF-B activation. The requirement for a functional IKK in cytokine-mediated NF-B activation within primary AMs was predicted by earlier studies involving cell lines and knockout mice. We found no evidence that the different patterns of IKK phosphorylation demonstrated in macrophage cell lines are of functional significance in primary AMs. In contrast, the failure of NIKkd overexpression to inhibit NF-B activation did not support the existing model of NIK function and indicates that there are important differences in the regulation of NF-B between primary cells and transformed cell lines. The data suggest that AMs possess an alternative IKK kinase capable of activating NF-B. In contrast to studies involving cell lines, we determined that in primary AMs and HUVECs, IKK, not NIK, is the point where the pathways responsible for NF-B activation converge. This study highlights the potential value of Adv mediated gene delivery as a tool to investigate the disease specific mechanisms of NF-B regulation in FA.