Immune surveillance of the airways is critical to maintain the integrity and health of the lung. We have identified a family of ligands expressed on the surface of stressed airway epithelial cells whose function is to bind the NKG2D-activating receptor found on several pulmonary lymphocytes, including natural killer cells, γδ+ T cells, and CD8+ T cells. We employed real-time PCR and flow cytometry in normal and transformed airway epithelial cell to demonstrate that major histocompatibility complex class I chain-related (MIC) B and the UL-16 binding protein (ULBP) ligands (ULBP1–4) are ubiquitously expressed at the mRNA level in all cell lines. MICA/B surface expression was present on 70% of transformed cell lines but was undetectable on primary cells. We demonstrate that MICA/B and ULBP 1, 2, 3, and 4 expression is rare or absent on the cell surface of unstimulated normal human bronchial epithelial cells although transcripts and intracellular proteins are present. Normal human bronchial epithelial cells exposed to 0.3 mM hydrogen peroxide exhibit an induction of all ligands examined on the cell surface. Surface expression is independent of changes in transcript level or total cellular protein and is mediated by the ERK family of mitogen-activated protein kinases. The induction of NKG2D ligands on stressed airway epithelial cells represents a potentially important mechanism of immune cell activation in regulation of pulmonary health and disease.
- MHC class I chain-related molecules A/B
- UL-16 binding proteins
- natural killer
- T cell
the airways are continually exposed to pathogenic organisms, allergens, and environmental contaminants. Therefore, the respiratory tract must efficiently and effectively neutralize the potentially deleterious effects of these exposures to protect the delicate tissue of the air spaces and preserve their gas exchange capability. The respiratory tract is equipped with an array of increasingly complex structural, physical, chemical, and cellular mechanisms that protect the composition and function of the lung. The cellular components of pulmonary host defense include cells of the myeloid lineage, as well as lymphoid cells that mediate the innate and acquired immune responses.
Epithelial cells undergoing physical or chemical stress and compromised function must be efficiently removed to control inflammation and promote cellular repair. Multiple mechanisms for the detection and elimination of stressed cells have been described including immune cell activation (16). One system that may provide a mechanistic link between epithelial cell stress and immune cell activation in the lung involves NKG2D receptor activation. The NKG2D receptor is expressed on circulating and tissue lymphocytes. This receptor directly recognizes transformed or infected cells through structurally related ligands that are expressed on the surface of stressed cells (2, 39, 48). The NKG2D receptor functions as a homodimer that binds distinct but structurally related ligands using structural pattern recognition (30, 37). The receptor is constitutively expressed on the cell surface of human natural killer (NK) cells (10), γδ T cells, and CD8+ αβ T cells (2). The proposed role of the NKG2D receptor in innate and adaptive immune responses to cellular and tissue stress is based on the ability of the receptor to stimulate cytotoxic effects of NK cells and T cells against virally infected cells and tumor cells in vitro and in vivo (6). Specifically, NKG2D receptor activation can induce target cell lysis and can trigger the production of cytokines (8, 18) and chemokines (8, 27, 44), as well as perforin and granzymes involved in cellular lysis (3). NKG2D engagement also serves as an enhancing or costimulatory signal for CD8+ T cell activation (18) or when stimulated in conjunction with T cell receptor cross-linking (22).
Two families of NKG2D ligands have been identified in humans: the MHC class I chain-related (MIC) molecules, MICA and MICB (2), and the UL-16 binding proteins 1, 2, 3, and 4 (ULBP1, 2, 3, 4) (8, 23). MICA and MICB expression is restricted or absent on normal tissues but is induced in response to various stresses and in some pathological conditions including epithelial-derived tumors (19, 24). MICA and MICB mRNA expression increases in response to heat shock in HeLa cells, and hydrogen peroxide (H2O2) increases MICA and MICB transcript levels in a human colon carcinoma cell line (49), suggesting that these ligands are regulated by cellular stress (17). Additionally, MICA/B protein has been detected in intestinal epithelial cells infected with Mycobacterium tuberculosis or Escherichia coli (46), and MICA has been demonstrated on the surface of cytomegalovirus-infected fibroblasts (18, 47). ULBP ligands are expressed at the mRNA level in many tissues and cell lines including lung, heart, liver, testis, brain, and colon, but cell surface expression by normal cells has not been detected (8, 38).
The majority of our knowledge of how stressed or damaged epithelium communicates with the immune system comes from experiments utilizing ozone-exposed mouse models and generally supports a role for T cells in the protection of lung tissue from oxidant induced damage (4, 7, 11, 26). Although these studies describe roles for T cells in the regulation of airway inflammation and epithelial cell sloughing, they do not provide mechanisms by which the immune system is activated, nor do they specify mediators involved in the signaling of cellular stress to the pulmonary immune system.
Despite increasing knowledge of the importance of the immune system in a variety of pulmonary diseases, the expression of NKG2D ligands has not been examined in airway epithelium. There is currently no evidence that stress can upregulate the surface expression of NKG2D ligands in the airways. Our study investigates whether airway epithelial cells express these molecules on the cell surface following exposure to oxidative stress. We examined mRNA transcript levels, cell surface ligand expression, and total cellular protein levels of NKG2D ligands in a panel of transformed and primary airway epithelial cell lines. Furthermore, we assessed the regulation of surface expression of these ligands in response to oxidative stress.
We report that NKG2D ligand transcripts are constitutively expressed in airway epithelial cell lines, whereas cell surface ligand expression is restricted to transformed cell lines. Additionally, cell surface expression of all NKG2D ligands studies is induced following H2O2-induced oxidative stress via the ERK family of mitogen-activated protein kinases (MAPK). Furthermore, data obtained using real-time PCR, Western analyses, and permeabilization studies indicate that these ligands are posttranscriptionally regulated.
MATERIALS AND METHODS
Normal human bronchial epithelial (NHBE) and normal small airway epithelium cells were obtained from a commercial supplier (Cambrex Bioproducts, Walkersville, MD) and cultured in airway epithelial cell basal medium [bronchial epithelial growth medium (BEGM)] with the addition of supplements and growth factors according to the manufacturer's recommendations (Cambrex). Airway epithelial cell tumor lines H292, H358, Calu-3, Nu6–1, H345, H596, H125, A549, NE-18, H441, ChaGo-K-1, SKMES-1, SW9000, Calu-1, H322, SKLU-1, and the SV-40 transformed NHBE cell line, 9HTEo-, were purchased from American Type Culture Collection (ATCC, Manassas, VA) and maintained according to ATCC protocols. Differentiated NHBE cells were obtained by culturing cells under air-liquid interface conditions (21, 43). NHBE cells were seeded at a density of 2 × 104 cells/cm2 on six-well culture inserts (0.4-μm pore size; Costar, Cambridge, MA) and cultured in BEGM for 1 wk in submerged culture conditions. NHBE cells differentiated culturing in BEGM at an air-liquid interface for 2 wk. Differentiated cells were identified on the basis of the appearance of secretory granules by light microscopy. All cells were maintained at 37°C in a 5% CO2/95% air humidified incubator. Cells were harvested using a nonenzymatic cell dissociation solution [10 mM EDTA in phosphate-buffered saline (PBS) without Ca+2/Mg+2].
Initial screening of NKG2D ligand expression (mRNA and MICA/B cell surface expression) was assessed in untreated primary airway epithelial cells, as well as transformed and tumor cell lines grown to ∼80% confluence. Regulation of NKG2D ligand expression was examined in NHBE, NCI-H292, and 9HTEo- cells in response to H2O2-induced oxidative stress. Cells were grown in six-well plates, and experiments were performed using cells at ∼80% confluence. In brief, cells were replenished with fresh media, allowed to equilibrate for 2 h (37°C in 5% CO2), and exposed to H2O2 or PBS. In preliminary studies, cells were treated to 0, 0.03, 0.1, 0.3, or 1.0 mM H2O2 for 12, 24, 48, and 72 h. For studies examining the ERK pathway inhibitor PD-98059, cells were pretreated with 3, 10, or 30 μM of inhibitor (or 0.25% DMSO) in PBS for 4 h before H2O2 exposure. All experiments with NHBE cells were conducted at passages 2–4 in our laboratory.
Total RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA) according to manufacturer's protocol. Total RNA from each sample was reverse transcribed into cDNA using the following reaction mixture: 2.0 μg total RNA in 10 μl RNase-free water, 1 μl oligo dT-15 (Promega, Madison, WI), and 1 μl 10 mM deoxynucleotide triphosphate (Invitrogen). The reaction mixture was incubated at 65°C for 5 min and snap-cooled (4°C). First-strand buffer (5×, 4 μl), 2 μl 0.1 M dithiothreitol, 1 μl SuperScript II (Invitrogen), and 1 μl RNase inhibitor (Promega) were then added to the reaction and incubated at 42°C for 1 h. The reaction was terminated by heating the mixture (70°C, 5 min), and the cDNA was stored at 4°C. Real-time PCR analysis was performed using an ABI 7900HT sequence detection system (Applied Biosystems, Foster City, CA) and a Quantitect SYBR Green PCR kit (Qiagen, Valencia, CA) according to each manufacturer's specifications. Samples (100 ng of cDNA per well) were performed in triplicate in 20-μl reactions with 3 μM primers. The real time PCR cycling conditions were 95°C for 15 min, followed by 45 cycles of 94°C for 30 s, 60°C for 15 s, and 72°C for 15 s, followed by a dissociation curve. Threshold cycle (Ct) values were determined using Sequence Detection Software version 2.1 (Applied Biosystems). Transcripts were considered present at a Ct ≤35. Relative expression was calculated as: 2−ΔCt, where ΔCT = CTNKG2D Ligand − CTGAPDH. Fold induction was calculated as 2/2, where ΔCT = CTLigand − CTGAPDH. The primers were as follows: ULBP1 forward 5′-TGC AGG CCA GGA TGT CTT GT-3′, ULBP1 reverse 5′-CAT CCC TGT TCT TCT CCC ACT TC-3′; ULBP2 forward 5′-CCC TGG GGA AGA AAC TAA ATG TC-3′, ULBP2 reverse 5′-ACT GAA CTG CCA AGA TCC ACT GCT-3′; ULBP3 forward 5′-AGA TGC CTG GGG AAA ACA ACT G-3′, ULBP3 reverse 5′-GTA TCC ATC GGC TTC ACA CTC ACA-3′; ULBP4 forward 5′-TAT GTC GAC CTC CAC AGT ATG CGA AGA ATA TCC CTG-3′, ULBP4 reverse 5′-ATA GGC GGC CGC AGA CTA AGA CGT CCT CAA-3′; MICA forward 5′-ACA ATG CCC CAG TCC TCC AGA-3′, MICA reverse 5′-ATT TTA GAT ATC GCC GTA GTT CCT-3′; MICB forward 5′-TGA GCC CCA CAG TCT TCG TTA C-3′ MICB reverse 5′-TGC CCT GCG TTT CTG CCT GTC ATA-3′; GAPDH forward 5′-GCC ATC AAT GAC CCC TTC ATT-3′, GAPDH reverse 5′-TTG ACG GTG CCA TGG AAT TT-3′. Appropriate amplicons sizes were verified on a 2% agarose gel. DNase-treated RNA was used as a control in PCR reactions to confirm that the only cDNA was being amplified.
Cells used for flow cytometry were harvested in 2 ml of nonenzymatic cell dissociation solution. Cells were washed in 0.5% BSA, 0.05% sodium azide in PBS [fluorescence-activated cell sorting (FACS) buffer]. For cell surface staining of NKG2D ligands, cells were resuspended in 100 μl of FACS buffer and incubated with 10 μg/ml of primary antibodies for 60 min at 4°C. The primary mouse monoclonal IgG antibodies used were as follows: anti-MICA/B (clone 6D4) (kindly provided by Dr. Thomas Spies, University of Washington, Seattle, WA), anti-ULBP1 (clone M295, AMGEN), anti-ULBP2 (clone M311, AMGEN), anti-ULBP3 (clone M550, AMGEN), anti-ULBP4 (clone M475, AMGEN), anti-human leukocyte antigen (HLA)-A, B, C (BD Pharmingen), and anti-β-actin (Santa Cruz Biotechnologies, Santa Cruz, CA). Isotype control mouse IgG (clone MOPC-173) was obtained from BD Pharmingen.
Cells were washed and resuspended in 100 μl of FACS buffer. Cells were then incubated with 20 μg/ml phycoerythrin-conjugated goat anti-mouse IgG secondary antibody (BD Biosciences, San Jose, CA) for 30 min at 4°C. Cells were washed and fixed in 400 μl of 2.0% paraformaldehyde (pH 7.2). For total cell staining, cells were fixed in 2.0% paraformaldehyde at room temperature for 20 min, washed with FACS buffer, and permeabilized in 0.05% Triton X-100 (Sigma, St. Louis, MO) for 5 min at room temperature. NKG2D ligand staining was then performed as described for cell surface ligands. Flow cytometry was performed using a Beckman-Coulter flow cytometer (Epics XL-MCL, Fullerton, CA), and the data were analyzed using WinList 5.0 software (Verity, Topsham, ME).
Western blot analysis.
Protein for Western blot analysis was isolated using mammalian protein extraction reagent (Pierce, Rockford, IL) according to manufacturer's instructions. Supernatants were aliquoted for storage at −80°C, and protein was quantitated using a bicinchoninic acid assay (Pierce). Western blot protein samples (40 μg) were loaded into a NuPAGE 4–12% Bis-Tris gel (Invitrogen) and electrophoresed using the protocol for XCell Surelock Mini-cell in NuPAGE MES-SDS Running Buffer (Invitrogen). Nondenaturing conditions were used for the assessment of phosphorylated ERK1/2 protein. After electrophoresis, gels were transferred onto polyvinylidene difluoride membranes. Membranes were incubated overnight with α-MICA (clone E-16, Santa Cruz Biotechnologies), α-ULBP2 (R&D Systems), or anti-phospho-p42/44 MAPK antibody (Upstate, Charlottesville, VA). Membranes were then incubated with species-specific α-IgG-horseradish peroxidase-conjugated secondary antibody, and detection was performed using ECL Reagents (Amersham, Piscataway, NJ). Membranes were exposed to Biomax XAR Film (Kodak, Rochester, NY).
Parametric data were analyzed for statistical significance by one-way ANOVA followed by Student's t-tests, with differences between means considered significant when P < 0.05.
Expression of NKG2D ligands in human airway epithelial cells.
The expression of NKG2D ligands in healthy adult tissues is rare or absent. NKG2D ligand transcript expression and the presence of the representative NKG2D ligands, MICA/B, on the cell surface were examined in 17 transformed airway epithelial cell lines of variable histology, as well as both primary bronchial airway epithelial cells and primary small airway epithelial cells (Table 1). Cell surface expression of MICA/B was detected at various levels on the majority of transformed cell lines but was absent on the primary airway epithelial cells. Real-time RT-PCR analyses demonstrate that MICB and ULBP1–4 transcripts are ubiquitously expressed regardless of the cell derivation. In contrast, MICA transcripts are consistently expressed in normal bronchial epithelial cells but are only detectable in three of 17 transformed epithelial cell lines. MICA and MICB ligand expression was not distinguished because the antibody (clone 6D4) does not discriminate between the two proteins.
To examine the variability and relative abundance of NKG2D ligand transcripts in NHBE cells, mRNA levels were quantitatively assessed in NHBE cells derived from four separate donors (Table 2). These data demonstrate that all examined NKG2D ligand transcripts are expressed in primary airway epithelial cells. ULBP2 is consistently expressed at the highest levels in these cells, whereas ULBP4 is consistently present at a level two orders of magnitude below the other ligands.
Oxidative stress induces cell surface NKG2D ligand expression.
To begin to understand the regulation of NKG2D ligand cell surface expression in response to stress, the induction of representative NKG2D ligands, MICA/B and ULBP2, was examined in response to H2O2 treatment in airway epithelial cells. We examined primary cells (NHBE) grown in submerged culture and at the air-liquid interface, a tumor cell line (NCI-H292), and a virally immortalized cell line (9HTEo-). The various cell types exhibited different constitutive and inducible expression of MICA/B and ULBP2 as assessed by flow cytometry (Fig. 1). Unstimulated NHBE cells grown in an immersed culture displayed very low levels of MICA/B and undetectable levels of ULBP2. Similarly, unstimulated NHBE cells grown at the air-liquid interface also displayed low levels of MICA/B and undetectable levels of ULBP2. Unstimulated NCI-H292 tumor cells expressed significant levels of MICA/B and ULBP2, whereas 9HTEo- cells expressed significant levels of MICA/B and low levels of ULBP2. Exposure to 0.3 mM H2O2 for 48 h induced the expression of MICA/B and ULBP2 in immersed and differentiated NHBE cells, increased the expression of MICA/B (but not ULBP2) in NCI-H292 cells, and increased both MICA/B and ULBP2 surface expression in 9HTEo- cells. Time-course and dose-response experiments indicated maximal induction of NKG2D ligands under these conditions (data not shown). These data demonstrate that the NHBE cells most accurately reflect the expression patterns in vivo (i.e., little or no baseline expression in healthy tissues, increased expression in transformed cells).
We further characterized the expression of additional NKG2D ligands (ULBP1, 3, and 4) and MHC class I molecules (HLA-A, B, C) in NHBE cells (Fig. 2). Unstimulated NHBE cells exhibited little or no surface expression of ULBP1, 3, and 4. H2O2 exposure significantly induced all three ULBP ligands (Fig. 2, A–C). In contrast, significant levels of MHC class I molecules were observed on the surface of unstimulated NHBE cells, and exposure to 0.3 mM H2O2 for 48 h attenuated this expression (Fig. 2D). Staining with an antibody to the intracellular protein β-actin revealed no staining after exposure to 0.3 mM H2O2 for 48 h, indicating that the treatment did not increase the permeability of the cells (not shown).
Regulation of NKG2D ligand expression in response to oxidative stress.
Intracellular staining of NKG2D ligands was performed on permeabilized, unstimulated, and H2O2-exposed NHBE cells to examine whether increased total cellular protein accompanied the increased surface expression. These studies demonstrate that all NKG2D ligands examined are present inside unstimulated NHBE cells. Exposure to 0.3 mM H2O2 for 48 h did not significantly increase the total cellular staining of the ligands (Fig. 3). Western blot analyses of total cellular protein (utilizing antibodies different from the flow studies) in unstimulated and H2O2-exposed NHBE cells demonstrated no noticeable increase in the levels of MICA/B or ULBP2 (Fig. 4). These data support the flow cytometry data obtained using permeabilized cells. The MICA/B Western antibody (clone E-16), which does not discriminate between the two proteins, recognized a single band of ∼43 kDa, which is consistent with published data (17). The ULBP2 Western antibody recognized a band at ∼37 kDa corresponding to the expected size of the mature glycosylated protein. We verified that the ULBP2 band by treating the cell lysates with glycanase to remove the sugar moieties. This resulted in a shift of the 37-kDa band to ∼28 kDa (not shown) consistent with the size of the unmodified peptide (8).
We performed quantitative real-time RT-PCR on all the NKG2D ligands to assess whether steady-state mRNA levels were increased coincident with the increases in cell surface ligand expression in NHBE cells. Steady-state transcript levels of MICA, ULBP2, ULBP3, and ULBP4 do not change following exposure to 0.3 mM H2O2 (Fig. 5). In contrast, MICB exhibited a twofold induction of steady-state mRNA levels after 24 h, and ULBP1 exhibited an increase in message levels after 6 and 24 h. Together, these data indicate that a general mechanism controlling cell surface expression of NKG2D ligands on airway epithelial cells involves stimulation of ligand translocation from the cytoplasmic compartment to the cell surface.
Additionally, we examined the role of the ERK signaling pathways in H2O2-induced NKG2D ligand expression. Exposure of NHBE cells to 0.3 mM H2O2 induced rapid (10 min) phosphorylation of ERK1/2 proteins that was sustained for several hours before returning to baseline (Fig. 6A). Exposure of NHBE cells to DMSO alone or the specific inhibitor of the ERK pathway, PD-98059, had no effect on ULBP ligand expression. However, H2O2-induced NKG2D ligand expression was dose dependently inhibited by pretreatment with the inhibitor (Fig. 6B).
The immune system is activated by diverse signals from cells exposed to pathogens, environmental stimuli, or mechanical damage, and these signals can be constitutive and/or inducible. The underlying function of stress immunosurveillance is to augment tissue repair and maintenance by eliminating damaged cells and expedite the establishment of healthy cells (16, 29). Constitutive signals primarily consist of intracellular products released following cellular necrosis (32) and inducible signals are molecules synthesized (5, 16) or modified (45) in response to stress. The induction of NKG2D ligands on airway epithelial cells following oxidative stress represents a potential mechanism whereby this class of molecules communicates tissue stress and damage to the local pulmonary immune system.
Lymphocyte populations known to express NKG2D receptors (i.e., NK cells, CD8+ cells, γδ+ cells) are intriguing in the context of airway injury for several reasons, including their relative abundance in the airways (9) and the functional implications of their activation. NKG2D receptor activation can induce target cell lysis and trigger the production of inflammatory cytokines (IFN-γ, TNF-α, IL-2, IL-4, granulocyte-macrophage colony-stimulating factor) (8, 18) and chemokines (macrophage inflammatory protein 1-β and I-309) (8, 27, 44), as well as perforin and granzymes involved in cellular lysis (3).
Presently, the T cell subpopulations and corresponding mechanisms that modulate the responses to pulmonary injury are unknown. Dziedzic and White (11) demonstrated a regulatory role for T cells in response to ozone-induced pulmonary injury. Repeated exposure of athymic mice to ozone resulted in the accumulation of greater amounts of damaged epithelium and an increase in inflammatory lesion volumes compared with control mice. Similar results were observed when mice were treated with the immunosuppressive drug cyclosporine A (4). Specific T cell populations and mediators controlling the inflammation and epithelial damage were not investigated. More recently, studies using mice specifically deficient in γδ+ T cells have revealed distinct roles for these cells in the regulation of acute pulmonary inflammation and the maintenance of epithelial integrity (26). Specifically, in mice deficient in γδ+ T cells, ozone exposure resulted in increased epithelial necrosis, decreased neutrophil accumulation, and decreased macrophage accumulation in the bronchoalveolar lavage compared with wild-type mice (26).
The induction of NKG2D ligands on the cell surface of airway epithelial cells is consistent with the idea that stressed airway epithelium possess the capacity to directly activate the pulmonary immune system and contribute to tissue repair and inflammation. Along these lines, we observed a decrease in MHC class I expression (HLA-A, -B, -C) coincident with NKG2D ligand induction following cell stress (Fig. 2). This is consistent with the idea that the epithelium activates the local immune system because the downregulation of MHC class I molecules, as seen in certain tumors or during viral infections, leads to sensitivity to lysis by NK cells (25). The results reported here indicate that NKG2D ligand upregulation on airway epithelial cells may represent an important mechanism in airway inflammation and repair present in many pulmonary diseases. The possible role of NKG2D receptor ligand interactions in these systems remains to be investigated.
We utilized H2O2 exposure in these studies because it is a well-characterized agonist used to induce oxidative stress, injury, and cell death in NHBE cells in vitro (12, 34, 36). These reports demonstrate that similar concentrations of H2O2 (0.2–1.0 mM) induce oxidative stress and a continuum of pathologies including injury and both apoptotic and necrotic cell death. As both apoptotic and necrotic epithelial cell death are known to occur in the many lung pathologies, the induction of NKG2D ligands following H2O2 exposure represents a relevant model system in which to investigate the regulation of these ligands in the airways.
It is important to know which NKG2D ligand genes are expressed and their relative roles in activating the immune system. Because there are at least six NKG2D ligands potentially expressed in the lungs, it is also important to determine whether they are redundant in function or exhibit specificity. In addition, determining the biological effects elicited by this pathway will significantly advance our knowledge of pulmonary immunobiology and contribute to our understanding of the pathophysiology of pulmonary diseases. Consistent with the theory that NKG2D receptor-ligand activation occurs in response to cellular or tissue stress, upregulation of NKG2D ligands has only been demonstrated in response to nonspecific stimuli such as infection, stress, DNA damage, and Toll-like receptor (TLR) activation. Initially, it was observed that MICA and MICB mRNA expression increase in response to heat shock in HeLa cells, suggesting that these ligands are regulated by cellular stress (17). Subsequently, it was demonstrated that H2O2 increases MICA/B transcript levels in a human colon carcinoma cell line (49). Expression of MICA/B is upregulated on the surface of many tumor cell lines including epithelial-derived tumors (19, 24). Additionally, MICA/B has been detected in intestinal epithelial cells infected with M. tuberculosis or E. coli (46), and MICA has been demonstrated on the surface of cytomegalovirus-infected fibroblasts (18, 47). ULBP ligands are expressed at the mRNA level in many tissues and cell lines including lung, heart, liver, testis, brain, and colon, but cell surface expression by normal cells has not been detected (8, 38). However, ULBP1 and ULBP2 cell surface expression is induced in cytomegalovirus-infected fibroblasts (40). The mouse Rae1 ligands are similarly expressed on tumor cell lines (1, 10), and Rae1 transcripts are increased in carcinogen-treated mouse skin and in the subsequent carcinomas (15). Rae1 ligands, but not H60, have been demonstrated on the surface of mouse fibroblasts infected with mouse cytomegalovirus (28), and Rae1 transcripts and cell surface protein are markedly induced in peritoneal macrophages in response to TLR ligands,. including LPS, zymosan, and double-stranded RNA (20). Most recently, Gasser et al. (14) demonstrated that NKG2D ligands are upregulated in response to activation of the DNA damage response program.
In our studies, NKG2D ligand transcripts, with the exception of ULBP1, are not significantly altered (Fig. 5). Thus our data are not consistent with the idea that increased transcripts are responsible for the induction of cell surface expression in airway epithelial cells following oxidative stress in vitro. On the contrary, our data support the notion that NKG2D ligands are constitutively expressed within airway epithelial cells, and cell surface expression is likely regulated by a mechanism controlling translocation from the cytoplasm to the cell surface. This is based on the flow cytometry data using permeabilized cells that indicate no significant upregulation of protein except ULBP4 (Fig. 3). Western analyses of MICA/B and ULBP2 further support this notion. Parallel flow cytometry experiments (both cell surface and permeabilized cell staining following H2O2 exposure) were performed using the 9HTEo- cell line and demonstrated similar results for all ligands (data not shown). This suggests a mechanism of NKG2D ligand regulation common to airway epithelial cells.
We further examined the regulation of NKG2D ligands in airway epithelial cells by investigating signal transduction pathways activated following oxidative stress. Oxidative stress induced by H2O2 activates the MAPK signaling cascade (including ERK1/2, JNK, and p38) in airway epithelial cells (36) and is important for epithelial cell functions including survival (41) and chemokine elaboration (36). MAPK activation in airway epithelial cells has been implicated in the pathophysiology of several airway diseases including asthma (35), chronic obstructive pulmonary disease (COPD) (31), and fibrosis (13). Along these lines, data demonstrating that H2O2-induced NKG2D ligand expression is partially dependent on the ERK1/2 signaling cascade (Fig. 6) are consistent with the functional importance of this pathway in airway pathologies. The transient induction of phosphorylated ERK1/2 in response to H2O2 relative to the delayed induction of cell surface NKG2D ligand expression suggests that secondary or intermediate events are involved ligand expression. Given the established role of ERK1/2 proteins in fundamental processes of cell survival and the nuclear response to stress (42), we speculate that the de novo generation mediators or alterations in the integrated cell survival signals in response to H2O2 exposure mediated the latent induction of NKG2D ligand expression on airway epithelial cells.
Although there is a pattern of regulation similar amongst all ligands (i.e., all are detectable at the transcript and protein level inside airway epithelial cells, and all are upregulated at the cell surface following oxidant exposure) there are clear distinctions in gene expression and regulation within this class of ligands. For example, ULBP2 mRNA is consistently expressed at the highest levels between donors and ULBP4 mRNA is consistently expressed at very low levels (Table 2). However, transcript level may not correspond to the total protein levels in or on cells as ULBP2 and ULBP4 are expressed at similar levels on the cell surface. Along these lines, only ULBP1 demonstrated a significant increase (10-fold) in transcript levels (Figs. 3 and 5) in response to H2O2 exposure, yet the amount of total cellular protein was not significantly altered. Similarly, ULBP4 was the only NKG2D ligand that appears to increase total cellular protein levels in response to H2O2 exposure (Fig. 3). The reasons for these discrepancies are presently not known. However, these discrepancies demonstrate that although the ligands are very similar in primary sequence and function, cells may utilize different mechanisms of controlling NKG2D ligand expression on the surface of stressed cells. These issues also emphasize the need to determine how each ligand is regulated in the cells of interest and suggest that the ligands are uniquely expressed for distinct physiological purposes. Few studies have addressed the issue of NKG2D ligand regulation in detail, and the data are primarily derived from transformed cell lines (17, 24, 49) or cell lines with constitutive surface expression (18, 47). The use of primary cells that exhibit little or no surface expression at baseline (such as occurs in healthy tissue) may be a more representative model of NKG2D ligand regulation. The discrepancy in the levels of cell surface ligands between transformed and primary airway epithelial cell lines underscores the importance of studying and understanding the expression, regulation, and function of these ligands in appropriate model systems. There are advantages and disadvantages associated with the use of primary cell cultures. The disadvantages include increased time to establish cultures in sufficient quantities, increased cost compared with tumor cell lines, and increased variability between donors. The advantages to primary cells are that they more accurately reflect the status of NKG2D ligand expression in vivo, they circumvent the genomic problems associated with genetically altered tumor cells, and the derivation of data from multiple donors enhances the relevance and significance of the results.
Given the number of pulmonary diseases associated with immune dysregulation, it is important to gain a better understanding of how epithelial cell stress/injury modulates the function of the pulmonary immune system in the development and exacerbation of disease. Most pulmonary diseases are associated with injury because of infection, stress [either endogenous (inflammation) or exogenous (air toxics)], or both. An increased understanding of how airway epithelial cells signal the immune system to maintain healthy tissue is critical for future investigations to determine the role of lymphocyte subpopulations in pulmonary diseases where injury and repair are in disequilibrium. For example, a role for T cells in the pathogenesis of COPD is suggested from correlations between increased CD8+ T cells in the airways and parenchyma and airflow limitation in patients with COPD. Specifically, CD8+ T cells are increased in the airways of smokers who developed COPD (33). This represents a condition in which chronic stress may inappropriately stimulate the immune system and have an adverse impact on disease progression or exacerbation. Similarly, patients with asthma, a disease of the airways caused by exaggerated immune responsiveness, could be significantly affected by exogenous airway stress.
These data demonstrate that NKG2D ligands are induced on the surface of airway epithelial cells following oxidant stress. This is an important finding because it represents a novel mechanism for the immune system to recognize injured or stressed pulmonary tissue. This mechanism may represent a critical pathway involved in the tissue repair and remodeling processes required for the preservation of pulmonary tissue and the gas exchange function of the lung. These results underscore the importance of determining the levels of control for cell surface expression of NKG2D ligands to gain a better understanding of pulmonary immunosurveillance in health and disease.
This study was supported by the National Institute of Environmental Health Sciences Center for Environmental Genetics (P30-ES-06096-02).
We acknowledge the technical contributions of Eric Shifferd and the valuable discussions with David Hoffman, Dr. Jay Tichelaar, and Dr. George Leikauf.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society