Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Peribronchial inflammation is a characteristic feature of allergic asthma (1, 2) and reducing inflammation is an important strategy for treating symptoms of allergic airway disease. T cells that produce T helper (Th) 2-type immunoregulatory cytokines play an important role in orchestrating immune and inflammatory processes that lead to immunoglobulin (Ig) E production, allergic inflammation, and accompanying tissue injury. T cells producing the Th2 cytokines interleukin (IL)-4 and IL-13 are increased in the respiratory mucosa of patients with allergic asthma (2, 3), and are further increased after allergen challenge (4). Successful treatment of allergic airway disease with glucocorticoids reduces the expression of Th2-type cytokines in the lung (5).

Chemokines are 8- to 12-kD chemoattractant cytokines that bind to specific chemokine receptors on leukocytes and regulate leukocyte migration. The chemokine superfamily can be divided into four groups on the basis of the number and spacing of the amino-terminal cysteines, and can also be generally grouped by their target cells (6). The CC chemokines can variably recruit eosinophils, monocyte/macrophages, and T cells. CC chemokines such as eotaxin; eotaxin-2; regulated on activation, normal T cells expressed and secreted (RANTES); monocyte chemotactic protein (MCP)-3, and MCP-4 are upregulated in the airways of individuals with asthma (7), likely contributing to the inflammatory cell infiltrate observed in asthmatic airways. The CXC chemokines can be divided into two groups, ELR motif (e.g., IL-8, GRO family members, and ENA-78) and non-ELR motif (e.g., IP-10, MIG, I-TAC) chemokines, which recruit neutrophils and T cells, respectively. Several chemokines known to be upregulated in asthma, including eotaxin, eotaxin-2, RANTES, and the MCP family, can signal to memory T cells (6), but this does not explain the preferential increase in Th2-type T cells compared with Th1-type T cells in the asthmatic airways.

Th1 and Th2 cytokine-producing CD4 T cells express different chemokine receptors and this enables their selective recruitment. For example, Th2 cytokine-producing CD4 T cells preferentially express CCR4, the receptor for TARC and MDC (10), whereas activated/memory interferon (IFN)--producing Th1-type CD4 T cells express CXCR3 that can bind the IFN--inducible chemokines IP-10, MIG, and I-TAC (12). Signaling of Th1 cells through CXCR3 and Th2 cells through CCR4 selectively induces the migration of those cells, as shown in chemotaxis assays (12). The selective expression of receptors for different chemokines by Th2 and Th1 CD4 T cells suggests a potential role for chemokines in the pathophysiology of allergic inflammation. Consistent with this notion, both TARC and MDC were upregulated in a murine model of atopic dermatitis (13) and MDC was upregulated in a murine model of allergic airway inflammation (14). Given the pivotal location of bronchial epithelial cells at the interface between the external environment of the host and the host's internal milieu, we postulated that bronchial epithelial cells may be a source of chemoattractants for Th2-type T cells in the respiratory tract.

The studies herein demonstrate the regulated production of TARC, but not MDC, by normal human bronchial epithelial (NHBE) cells in vitro and by bronchial epithelial cells in human lung xenografts in vivo in response to stimulation with proinflammatory cytokines. Increased TARC levels were also observed in human bronchial alveolar lavage (BAL) fluids (BALFs) of patients with allergic asthma after allergen challenge. In addition, TARC is shown to be a nuclear factor (NF)-B target gene. These findings broaden current concepts of the capacity of bronchial epithelial cells to initiate and regulate mucosal immune and inflammatory responses in human airways.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

Recombinant human (rh) tumor necrosis factor (TNF)-, IL-1, IL-4, and IL-13 were obtained from PeproTech (Rocky Hill, NJ), IFN- from Biosource (Camarillo, CA), and TARC and MDC from R&D Systems (Minneapolis, MN). Monoclonal mouse antihuman TARC and MDC antibodies and polyclonal goat antihuman TARC were from R&D Systems. Polyclonal rabbit antihuman MDC antibody was from PeproTech. Dexamethasone (DEX) and MG-132 were from Sigma Chemical Co. (St. Louis, MO).

Cell Lines

NHBE cells were from Clonetics (Walkersville, MD). The human epithelial lung cancer cell line A549 and the simian virus 40 (SV-40)-transformed human normal bronchial epithelial cell line BEAS-2B cells were from the American Type Culture Collection (Manassas, VA). A549 cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal calf serum (FCS) and 2 mM L-glutamine. BEAS-2B cells were grown in DMEM/F12 with 10% FCS and 2 mM L-glutamine. NHBE cells were maintained in serum-free bronchial epithelial cell basal medium (Clonetics) supplemented with 50 µg/ ml bovine pituitary extract, 50 ng/ml human epidermal growth factor, 0.5 µg/ml hydrocortisone, 0.5 µg/ml epinephrine, 10 µg/ml transferrin, 5 µg/ml insulin, 0.1 ng/ml retinoic acid, 6.5 ng/ml triiodothyronine, 50 µg/ml gentamicin, and 50 ng/ml amphotericin-B (all from Clonetics). Cells were grown to confluence in six-well tissue culture plates before stimulation with cytokines. After stimulation, supernatants from airway epithelium were centrifuged to remove cellular debris and stored at 80°C until use.

Infection with Recombinant Adenovirus

Recombinant adenovirus containing an IB-AA superrepressor (Ad5IkB-A32/36) or the Escherichia coli -galactosidase gene (Ad5LacZ) was constructed as previously described (15, 16). Ad5IB-A32/36 expresses a hemagglutinin epitope-tagged form of IB with serine to alanine substitutions at amino acid residues 32 and 36. Mutant IB does not undergo signal-induced phosphorylation and degradation, and acts as a potent superrepressor of NF-B activation (15, 16, and data not shown).

A549 cells were grown to confluence, and cells were infected with Ad5IB-A32/36 or Ad5LacZ in serum-free media (Opti-MEM; Gibco BRL Life Technologies, Grand Island, NY) for 1 h at 37°C at a multiplicity of infection of 50. The adenovirus was then washed off, fresh serum-containing media was added to the cells, and cells were incubated for 16 h before stimulation with cytokines (17).

Human Fetal Lung Xenografts

Human fetal lung xenografts in SCID mice were generated using a protocol we and others have previously described for generating SCID mice carrying human fetal intestinal xenografts (18- 20). Human fetal lung (gestational age 17 to 18 wk) was transplanted subcutaneously onto the backs of C57BL/6 SCID mice. Subsequently, xenografts were allowed to develop for 10 to 12 wk during which they mature into lung tissue that contains bronchial structures lined with pseudostratified ciliated epithelium, alveolar sacs, and submucosal glands, as also described by others (21, 22). Xenograft-bearing mice were injected intraperitoneally with cytokines in phosphate-buffered saline (PBS); and 6 h later, xenografts were harvested and segments were either frozen in liquid nitrogen for subsequent RNA isolation or embedded in OCT and snap-frozen for subsequent immunohistochemistry.

BAL

Mild asthmatics with allergy to cat, dog, or house dust mite, as confirmed by skin tests and inhalation allergen challenge, were used in the study. Mean age of subjects was 22.8 ± 1.1 yr, and mean baseline forced expiratory volume at 1 s (FEV₁) (percent predicted) scores were 90 ± 2. Endobronchial allergen challenge methods were as previously published (23). Briefly, 1 ml of a 10% vol/vol solution of the PD₂₀ FEV₁ allergen dose (concentration of allergen that caused a 20% decrease in FEV₁) was instilled into the posterior segment of the right lower lobe, and 1 ml of diluent (as negative control) was instilled into the anterior segment of the right lower lobe. After 24 h, a post-allergen BAL was obtained from both sites. BALF was centrifuged to remove cells, and the supernatant was concentrated 10-fold using the Centricon system (3000 MW cut-off; Amicon, Bedford, MA) before assaying for TARC or MDC by enzyme-linked immunosorbent assay (ELISA). The statistical significance of differences between diluent and allergen challenge TARC levels was assessed by a paired nonparametric signed rank test. All studies were approved by the UCSD Human Subjects Committee.

Reverse Transcriptase/Polymerase Chain Reaction Analysis

Total cellular RNA was extracted using an acid guanidinium-phenol-chloroform method (Trizol; GIBCO BRL) and treated with ribonuclease (RNase)-free deoxyribonuclease (Stratagene, La Jolla, CA). RNA integrity was confirmed by electrophoresis on 1% agarose gels and ethidium bromide staining. Total cellular RNA, 1 µg, was reverse transcribed at 37°C for 70 min in 20 µl containing 2.5 U Superscript-II reverse transcriptase (RT) (GIBCO BRL); 10 mM dithiothreitol, 1 mM each of deoxyadenosine triphosphate (dATP), deoxythymidine triphosphate (dTTP), deoxycytidine triphosphate (dCTP), and deoxyguanidine triphosphate (dGTP); and 5 µg/ml oligo-dT primer (Pharmacia, Piscataway, NJ). Reactions were stopped by heat inactivation for 10 min at 85°C. Sequences were amplified from complementary DNA (cDNA) by polymerase chain reaction (PCR) using specific primers for TARC, MDC, and -actin. Primers for TARC and MDC were designed from available sequences from GenBank and yielded PCR products of 221 and 425 base pairs, respectively. Primer sequences for TARC and MDC were as follows: TARC sense primer 5-CAC GCA GCT CGA GGG ACC AAT GTG-3, antisense primer 5-TCA AGA CCT CTC AAG GCT TTG CAG G-3; MDC sense primer 5-GAG ACA TAC AGG ACA GAG CAT GGC T-3, antisense primer 5-ATG GAG ATC AGG GAA TGC AGA GAG-3. The primers for -actin were as previously described (24): sense primer 5-TGA CGG GGT CAC CCA CAC TGT GCC CAT CTA-3, antisense primer 5-CTA GAA GCA TTG CGG TGG ACG ATG GAG GG-3. TARC and MDC primers did not amplify products from mouse spleen cDNA. PCR amplifications were performed in 50 µl containing 2 µl of cDNA, 25 pmol of each primer, 1.5 mM MgCl₂, 4.0 U Taq polymerase (GIBCO BRL), and 200 µM each of dATP, dTTP, dCTP, and dGTP in RNase-free distilled water. The amplification profile was 45 s of denaturation at 95°C, followed by 2.5 min of annealing and extension at 60°C for TARC, 64°C for MDC, and 72°C for -actin. For negative controls, RNA was omitted from the RT reactions and cDNA was omitted from the PCR reactions. RNA from lipopolysaccharide (2 µg/ml)-stimulated and phytohemagglutinin-P (10 µg/ml)-stimulated human peripheral blood mononuclear cells served as a positive control for TARC, and RNA from human monocyte-derived macrophages was used as a positive control for MDC. After amplification, aliquots of the PCR products were separated on a 1.2% agarose gel containing ethidium bromide and photographed.

Real-Time PCR

Real-time PCR was performed using an ABI Prism 7700 Sequence Detection System (PE Applied Biosystems, Foster City, CA). Each reaction contained 25 µl of 2× SYBR Green Master Mix (containing 200 nM dATP, dGTP, and dCTP; 400 nM deoxyuridine triphosphate; 2 mM MgCl₂; 0.25 units of uracil N-glycosylase; 0.625 units of Amplitaq Gold DNA polymerase); 25 pmol of the sense and antisense TARC primers described earlier; and 2 µl of cDNA in a final volume of 50 µl. The reactions were incubated at 50°C for 2 min followed by 95°C for 10 min. The amplification profile was 95°C for 15 s and 62°C for 1 min for 40 cycles. Amplification of the expected single product was confirmed by running on a 1% agarose gel stained with ethidium bromide. Data analysis was performed using PE Biosystems 7700 sequence detection system software. Real-time PCR data were plotted as the R_n fluorescence signal versus the cycle number. R_n was calculated using the equation R_n = (R_n⁺)  (R_n) where R_n⁺ is the fluorescence signal of the product and R_n is the fluorescence signal of the baseline emission. C_t is the cycle number at which the R_n crosses threshold. Fold changes in TARC cDNA were determined as fold change = 2^Ct.

TARC and MDC ELISA

TARC and MDC in supernatants and BALF were measured by sandwich ELISA. Briefly, 96-well polystyrene plates (Immulon-4; Dynex Technologies Inc, Chantilly, VA) were coated with mouse antihuman TARC or anti-MDC monoclonal antibody (1 µg/ml) in carbonate buffer as capture antibody for 4 h at 37°C, and blocked overnight at 4°C with 5% nonfat milk in PBS with 0.1% Tween-20. Recombinant TARC or MDC as a standard or culture supernatants diluted in PBS containing 1% bovine serum albumin (BSA) and 0.1% Tween-20 were added to the plates and incubated for 3 h at room temperature. Biotinylated goat antihuman TARC or rabbit anti-MDC (100 ng/ml) was used as detection antibody, followed by horseradish peroxidase (HRP)-conjugated streptavidin (Amersham Life Science, Inc., Arlington Heights, IL) or HRP-conjugated donkey antirabbit (Amersham) to detect TARC or MDC, respectively. Bound HRP was visualized with TMB substrate for 30 min, and the reaction was stopped by addition of 1.2 M H₂SO₄. Absorbance was measured at 450 nm. The ELISAs for TARC and MDC were sensitive to 30 pg/ml recombinant cytokine.

Immunohistochemistry

Xenograft tissue was embedded in OCT compound and snap-frozen in isopentane/dry ice. Cryostat sections (5 µm) were air-dried, fixed with acetone, and then blocked with 1% BSA/PBS. Endogenous biotin was blocked using an avidin/biotin blocking kit (Zymed, South San Francisco, CA). Sections were incubated overnight at 4°C with a biotinylated goat antihuman TARC (2.5 µg/ml) or a biotinylated goat IgG (2.5 µg/ml; Jackson ImmunoResearch Laboratories, West Grove, PA) as a capture antibody. Binding of biotinylated goat IgG was detected with Cy3-labeled streptavidin (Jackson) at a 1:250 dilution. Sections were counterstained with the nuclear dye Hoechst 33258 (Calbiochem, San Diego, CA).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Regulated TARC Expression by A549 Human Lung Carcinoma Cells

As shown in Figure 1A, A549 cells expressed TARC messenger RNA (mRNA) in response to TNF- or IL-1 stimulation. Moreover, as recently reported (25), TARC mRNA expression by A549 cells was markedly increased in cells stimulated with a combination of TNF- and IL-4, although IL-4 alone did not upregulate TARC mRNA expression. As assessed using real-time PCR (Figure 1B), TNF- increased TARC expression 30-fold. In contrast, IL-4 increased TARC expression 2-fold, and IFN- increased expression by 3-fold. TNF- plus IL-4 synergistically increased TARC expression 3,250-fold, whereas TNF- plus IFN- induced an intermediate 280-fold increase in TARC expression compared with unstimulated cells. TARC protein secretion by A549 cells was also markedly increased by stimulation with TNF- or IL-1, and potentiation was observed between IL-4 and these proinflammatory cytokines (Figure 1A).

View larger version (33K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   Figure 1.   Cytokine-induced TARC protein and mRNA expression by A549 cells. (A) A549 cells were stimulated with the indicated cytokines (20 ng/ml) for 12 h before isolation of total RNA, and RT-PCR for TARC (30 cycles) or -actin (20 cycles) was performed as described in MATERIALS AND METHODS. TARC protein expression at 24 h (as measured by ELISA) is shown at the top of the figure for each stimulation condition indicated at the bottom. Values are means ± standard error of the mean (SEM) for at least three independent experiments. (B) TARC mRNA expression in control or agonist-stimulated A549 cells assessed by real-time PCR. Values are means of triplicate samples.

TARC Is a NF-B Target Gene

Because TNF- alone can upregulate TARC mRNA and protein production, and TNF- is known to activate signal transduction pathways that culminate in activation of the transcription factor NF-B, we used two approaches to determine whether TARC functions as a NF-B target gene. The proteasome inhibitor MG-132 has previously been shown to block NF-B activation in A549 cells (26). Therefore, A549 epithelial cells were first pretreated with MG-132 for 1 h before stimulation with TNF-, alone or in combination with IL-4. As shown in Figure 2A, MG-132 significantly decreased TARC protein production in response to those stimuli, suggesting that NF-B is important for upregulated TARC production. Because pharmacologic approaches are not always completely specific, we used a second approach to block NF-B activation. For these studies, cells were infected with recombinant adenovirus expressing a mutant IB protein that has serine to alanine substitutions at positions 32 and 36 (Ad5IB-A32/36) and acts as a superrepressor of NF-B activation (15, 17, 27). As shown in Figure 2B, TARC mRNA expression in response to stimulation with TNF- or TNF- plus IL-4 was completely inhibited in cells infected with the Ad5IB-A32/36 superrepressor, but not in cells infected with control adenovirus expressing -galactosidase. Consistent with the dependence of TARC expression on NF-B activation, DEX (a glucocorticoid that has been shown to inhibit NF-B and activator protein-1 activation in bronchial epithelial cells [28]) inhibited TARC production in response to TNF- or TNF--plus-IL-4 stimulation by 70 to 80% (0.15 ± 0.05 versus 0.46 ± 0.10 ng/ml for TNF-, and 8.3 ± 2.1 versus 32.7 ± 10.2 ng/ml for TNF- plus IL-4).

View larger version (20K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   Figure 2.   Inhibition of NF-B activation inhibits TARC secretion. (A) A549 cells were pretreated for 60 min with 10 µM MG-132 before stimulation with TNF- (20 ng/ml) or with TNF- (20 ng/ml) plus IL-4 (20 ng/ml) for an additional 30 min. Cells were washed and cultured for an additional 24 h, after which supernatants were assayed for TARC. Values are means ± SEM. (B) A549 cells were left uninfected or were infected with an adenovirus encoding an IB superrepressor (Ad5IB-A32/36) or -galactosidase (Ad5LacZ) as a virus control. Cells were then stimulated with cytokines (20 ng/ml) for 6 h before isolation of total RNA and RT-PCR analysis for TARC (34 cycles) and -actin (20 cycles).

Regulated TARC mRNA Expression in Normal Bronchial Epithelial Cells

Th2 but not Th1 cytokine-producing CD4 T cells express receptors for TARC. The marked potentiation of TARC expression in A549 carcinoma cells by the Th2 cytokines IL-4 and IL-13 suggest that Th2 cytokines may participate in a positive feedback loop to markedly amplify epithelial cell TARC production. Because asthma is primarily a disease of the conducting airways, we tested whether a positive feedback loop similar to that shown in A549 cells, which are thought to be alveolar in origin, could be demonstrated in bronchial epithelial cells. We first assessed TARC mRNA expression using BEAS-2B, a SV-40-transformed human bronchial epithelial cell line. In contrast to A549 cells, TARC mRNA expression in BEAS-2B cells was upregulated in response to stimulation with TNF- in combination with IFN-, but not in response to TNF- in combination with the Th2 cytokines IL-4 or IL-13 (Figure 3A). These data suggest that regulation of TARC expression in bronchial epithelial cells may differ from that of the A549 cell line.

View larger version (34K): [in this window] [in a new window]   View larger version (17K): [in this window] [in a new window]   Figure 3.   Expression of TARC mRNA in BEAS-2B and NHBE cells. (A) BEAS-2B and NHBE cells were stimulated with 20 ng/ ml of the indicated cytokines for 12 h before isolation of total RNA and RT-PCR analysis for TARC (35 cycles) and -actin (25 cycles). (B) TARC mRNA expression in control or agonist-stimulated NHBE cells assessed by real-time PCR. Values are means of triplicate samples.

On the basis of apparent differences in TARC regulation in the different lung epithelial cell lineages, and the potential importance of TARC in the pathogenesis of bronchial inflammation, we next assessed regulated TARC mRNA expression in NHBE cells. As shown in Figure 3A, TARC mRNA expression was upregulated in NHBE cells in response to TNF- stimulation. As with BEAS-2B cells, TARC mRNA expression in NHBE cells was not potentiated by IL-4 or IL-13, but was upregulated by costimulation with IFN- in combination with TNF-. TARC regulation in NHBE cells was further investigated using real-time PCR (Figure 3B). TNF- alone upregulated TARC expression 5-fold, IFN- did so 3-fold, and the two together upregulated TARC expression 18-fold compared with unstimulated NHBE cells. IL-4 alone or with TNF- did not increase TARC expression compared with unstimulated or TNF-- stimulated cells, respectively. Together, these data suggest that Th2 cytokines have little, if any, role in amplifying bronchial epithelial cell TARC responses, whereas the Th1 cytokine IFN- in combination with TNF- can upregulate epithelial cell TARC expression.

Expression of the other known CCR4 ligand, MDC, was also examined in NHBE, BEAS-2B, and A549 cells. MDC mRNA expression was not seen in any of these cell lines after stimulation with TNF-, IL-1, IFN-, IL-4, or IL-13 alone or in combination (data not shown).

Regulated TARC Expression in Human Fetal Lung Xenografts

To determine whether the data obtained for TARC expression and regulation using human bronchial epithelial cells lines had an in vivo correlate, we used a human fetal lung xenograft model. Human fetal lung xenografts implanted subcutaneously into SCID mice were allowed to develop for > 10 wk, at which point the tissue contained well-defined alveolae and bronchial structures with surrounding cartilage (representative histology shown in Figure 4). To determine constitutive and inducible TARC production in human lung, mice bearing lung xenografts were injected intraperitoneally with rhIL-1 alone or in combination with rhIL-4 or rhIFN-, after which TARC mRNA expression and TARC protein production were assessed in the xenografts. As shown in Figure 5, little TARC mRNA was detected in control human lung xenografts from mice injected with PBS. In contrast, TARC mRNA expression increased in the lung xenografts after IL-1 injection. IL-1-stimulated TARC mRNA expression was further increased by costimulation of xenografts with IFN-, but not with IL-4.

View larger version (131K): [in this window] [in a new window]   Figure 4.   Human fetal lung xenografts. Hematoxylin and eosin- stained sections of human fetal lung tissue (gestational age 18 wk), before transplantation (A and C ), or 10 wk after (B and D-H ) subcutaneous transplantation in a SCID mouse. (A) Lung tissue before transplantation. Alveolar spaces are not yet apparent, although airways lined by columnar epithelium are present. (Original magnification: ×200). (B) Lung xenograft 10 wk after transplantation, showing alveolar development (original magnification: ×200). (C ) Lung tissue before transplantation, showing pseudostratified epithelium surrounded by cartilage. (Original magnification: ×200). (D) Lung xenograft 10 wk after transplantation. Bronchial structure surrounded by alveolae. (Original magnification: ×200). (E ) Lung xenograft 10 wk after transplantation, showing area of cartilage (arrow) located adjacent to a bronchial structure. (Original magnification: ×200) (F ) Lung xenograft 10 wk after transplantation, showing alveolar spaces lined by pneumocytes (arrows). (Original magnification: ×1000). (G) Lung xenograft 10 wk after transplantation, showing a columnar epithelium lining a small bronchial airway. (Original magnification: ×1000). (H ) Lung xenograft 10 wk after transplantation, showing pseudostratified ciliated epithelium lining a large bronchial airway, with a layer of smooth muscle below the epithelium. (Original magnification: ×1000).

View larger version (39K): [in this window] [in a new window]   Figure 5.   Expression of TARC mRNA in human fetal lung xenografts. SCID mice bearing human fetal xenografts were injected intraperitoneally with PBS, rhIL-1 (1 µg), rhIFN- (10 µg), or rhIL-4 (3 µg). After 6 h, xenografts were removed and tissue was snap-frozen for subsequent RNA isolation. Total xenograft RNA was used for RT-PCR analysis of TARC (35 cycles) or -actin (20 cycles). Shown is a representative PCR result. Of four experiments performed, variable levels of constitutive TARC expression were noted, but cytokine stimulation consistently increased TARC mRNA expression.

To localize TARC protein expression in the xenografts, sections were immunostained for TARC. As shown in Figure 6, TARC immunostaining was localized to the ciliated pseudostratified epithelium of the bronchial structures within the xenografts. TARC staining was not observed under any stimulation conditions in the alveolar epithelium. Low levels of constitutive TARC staining were observed in the bronchial epithelium, which was not noticeably increased by IL-1 stimulation alone or with IL-4. Consistent with the mRNA data, immunostaining was increased in the bronchial epithelium of human xenografts from IL-1- and IFN--injected mice.

View larger version (135K): [in this window] [in a new window]   Figure 6.   Immunohistochemical detection of TARC in human fetal lung xenografts. Sections of unstimulated (A and B) or IL-1 plus IFN--stimulated (C-F ) human lung xenograft were immunostained using biotinylated goat antihuman TARC antibody (A-D) or a control biotinylated goat IgG (E-F ) and Cy3-labeled streptavidin. Small letters indicate examples of bronchial (b) and alveolar (a) structures. Specific TARC staining shown in red was restricted to ciliated epithelium of the bronchial structures, and no staining was observed in the alveolar regions. Sections were counterstained with Hoechst 33258 to show nuclei in blue. Original magnification: A, C, and E, ×100; B, D, and F, ×1,000.

TARC Is Present in BALF

Consistent with our studies of regulated TARC expression, TARC immunostaining was recently noted in bronchial epithelium of asthmatic subjects (25). To determine whether TARC is also secreted into airway fluid, we assessed TARC in BALF from patients with asthma after endobronchial allergen challenge, or challenge with diluent as control (Figure 7). TARC was detectable in all samples measured, and TARC was significantly (P = 0.016) increased in BALF after local allergen challenge compared with diluent challenge. TARC levels in BALF after diluent challenge were similar to nonchallenge values in a group of nonatopic, nonasthmatic control subjects, and in a group of nonchallenged asthmatic subjects (data not shown). Low levels of another CCR4 ligand, MDC, were observed in control and asthmatic subjects (ranging from < 0.045 to 0.15 ng/ml), but did not increase after allergen challenge.

View larger version (14K): [in this window] [in a new window]   Figure 7.   TARC in BALF. TARC was assayed by ELISA in 10-fold concentrated BALF from allergic asthmatic subjects 24 h after challenge with allergen or diluent as negative control (n = 9). Horizontal lines are geometric means. The difference in TARC after diluent challenge compared with allergen challenge was statistically significant (P = 0.016).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Airway epithelial cells produce chemokines that have the capacity to orchestrate inflammatory cell migration in the lung. We demonstrate herein the regulated expression and production of TARC, a known Th2 CD4 T-cell chemoattractant, by normal bronchial epithelial cells in vitro, by pseudostratified bronchial epithelium in vivo in a human lung xenograft model, and in BALF of allergen-challenged patients with allergic asthma. Epithelial cells from patients with allergic asthma have been reported to express chemokines (e.g., eotaxin, eotaxin 2, MCP-3, and MCP-4) that are known to recruit cells important in host innate immunity (8, 9). These studies, coupled with a prior report demonstrating the regulated production by normal bronchial epithelium of IFN-inducible chemokines that can chemoattract T cells that produce Th1-type cytokines (29), establish the capacity of normal bronchial epithelium to play a role in signaling cells important for mediating host acquired immunity in the conducting airways.

We note there is a significant difference in the regulation of TARC between bronchial epithelial cells in vitro compared with A549 alveolar epithelial cells, which have been widely used as a model to study epithelial cytokine production in the respiratory tract. Although proinflammatory mediators (i.e., TNF- or IL-1) consistently upregulated TARC in normal bronchial epithelium and in A549 alveolar cells, the marked potentiation of that response by the Th2 cytokines IL-4 and IL-13 was a feature restricted to the A549 cell line. This difference may reflect fundamental differences in bronchial (BEAS-2B, NHBE) versus alveolar (A549) epithelial cell TARC regulation. To more closely model the regulation of TARC expression in vivo, we used a human fetal lung xenograft model. Although this model has been described in the literature (21, 22), it has not been widely used as an in vivo model of epithelial gene expression, although one recent study used a fetal tracheal xenograft model to examine the production of inflammatory mediators in normal versus cystic fibrosis epithelium (30). One advantage of our xenograft model is the presence of many different phenotypes of lung epithelium in the xenograft tissue. Using this model, we were able to demonstrate that pseudostratified ciliated epithelial cells of bronchial structures contained TARC protein after IL-1 and IFN- stimulation. Costimulation with IL-1 and IL-4 did not induce alveolar expression of TARC as was observed with A549 cells, suggesting that the epithelial source of TARC in vivo is limited to the bronchial epithelium, regardless of the cytokine milieu. These findings support the hypothesis that chemokine gene expression by normal lung epithelium in vivo is specific to both region and cell type.

The upregulated expression of TARC in response to TNF- and IL-1 in airway epithelial cells led us to test the possibility that TARC functions as a NF-B target gene. Although structural information on the promoter region of human TARC is not currently available, our studies clearly demonstrate that inhibiting the activation of NF-B completely abrogates TNF--stimulated TARC production. These studies further demonstrate the central role NF-B plays in regulating TARC production in epithelial cells, inasmuch as the marked potentiating effect of IL-4 on TARC expression in A549 cells was also completely abrogated by a superrepressor of NF-B activation. Airway eosinophilia in response to allergic sensitization is dependent on NF-B, as shown using mice deficient in the p50 subunit of NF-B, and such studies have demonstrated that several chemokines, including eotaxin, macrophage inflammatory protein (MIP)-1, and MIP-1, are not induced in p50 knockout mice after allergen challenge (31). Together with our studies, those findings suggest that NF-B is a central regulator in the development of allergic airway inflammation and that this is mediated, at least in part, through the regulation of chemokine gene expression.

The mechanisms responsible for the accumulation of Th2 T cells in the asthmatic lung are not known. Because TARC is a chemoattractant for Th2 T cells, one would expect that TARC levels might be elevated in disease states characterized by elevated Th2 cytokine levels. In this regard, elevated production of TARC has been documented in skin lesions in a mouse model of atopic dermatitis associated with infiltration of IL-4-producing T cells (13), and increased TARC expression was observed in Reed-Sternberg cells from patients with Hodgkin's lymphoma, with an accompanying infiltrate of CCR4⁺ cells with a Th2 phenotype (32). Our studies and a recent study by others (25) suggest parallel findings in allergic lung disease, which is associated with an increased infiltration of Th2-type T cells (4). Thus, as shown herein, TARC was increased in BALF after allergen challenge of allergic asthmatics, and TARC immunostaining of bronchial epithelial cells has been noted in patients with allergic asthma (25). Finally, our finding that TARC expression was restricted to pseudostratified epithelial cells in the human lung xenografts, combined with increased immunostaining for TARC in bronchial epithelium of allergic asthmatics (25), supports the concept of bronchial epithelium as a source of TARC in BALF.

To the extent TARC produced by epithelial cells can play a role in chemoattracting Th2 cells at mucosal sites, the studies herein suggest the notion that different mediators may play a role in chemoattracting Th2 CD4 T cells in different mucosal sites. For example, both TARC and MDC are known ligands for CCR4 expressed on Th2 cytokine-producing cells. However, we demonstrate that airway epithelial cells express and produce TARC, but not MDC, in a regulated manner. In striking contrast to bronchial epithelial cells, human intestinal epithelial cell lines and human intestinal epithelial cells in vivo produce MDC, but not TARC, in response to stimulation with TNF- or IL-1 (33). Mouse-skin keratinocytes appear to be more similar to bronchial epithelial cells in that they produce TARC, but not MDC, in response to TNF-, IL-1, or IFN- (13). Although TARC expression in epithelial cells is primarily stimulated by proinflammatory cytokines, expression of TARC in macrophages and dendritic cells has been shown to be upregulated in response to IL-4 but not TNF- or IL-1 stimulation (34). This suggests that a different cytokine milieu may result in different cellular sources of TARC production.

Human lung xenografts in SCID mice contain several epithelial phenotypes, smooth muscle, and cartilage that may each participate in regulation of mucosal inflammation in the lung. We used the xenograft model to study early in vivo lung epithelial cell responses to stimulation with one or more human cytokines. The ability to directly stimulate human lung epithelium in xenograft-bearing SCID mice with human Th1 or Th2 cytokines provides a powerful tool to quantitatively and qualitatively assess regulated epithelial cell gene expression in an in vivo setting. The ability to study epithelial physiology in a controlled in vivo setting may be of value not only in the study of inflammatory gene expression as it was used in the studies herein, but also in the understanding of cell-cycle regulation, including epithelial proliferation and apoptosis in lung growth and repair.