Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Cystic fibrosis (CF) is primarily an ion transport disease caused by mutations in the gene encoding the cystic fibrosis transmembrane conductance regulator (CFTR) (1). It is becoming more evident, however, that altered regulation of inflammatory responses is an important factor in the progression of CF lung disease. Transforming growth factor (TGF)-1 has several roles in influencing inflammatory responses, and increased expression of TGF-1 has been reported to be associated with more severe pulmonary disease in CF (2). Despite evidence of TGF-1 involvement in CF disease, an examination of TGF-1-mediated signaling in CF epithelial cells has yet to be performed.

Various Smad proteins carry out intracellular signaling in response to TGF- family members. Smad proteins are classified into three categories: (1) receptor-regulated Smads (R-Smads); (2) common mediator Smads (Co-Smads); and (3) inhibitory Smads (I-Smads) (reviewed in Refs. 3 and 4). The R-Smads are directly phosphorylated by the receptor and are translocated to the nucleus, where they bind various response elements to regulate gene expression. Smad2 and Smad3 are R-Smads that are phosphorylated and activated by TGF- and activin receptors, whereas Smads 1, 5, and 8 are R-Smads that are activated in response to the various bone morphogenic proteins (BMP). Smad4 is the only known mammalian Co-Smad and acts by forming complexes with either Smad2 or Smad3 or other proteins such as the forked head activin signal transducer-1 (FAST-1) to facilitate signaling. FAST-1 is an important element in the binding complex specific for the activin response element (ARE). The I-Smads (Smad6 and Smad7) inhibit TGF- signaling by interfering with R-Smad/Co-Smad association.

Although there is some evidence that TGF-1 is involved in the progression of CF airway disease (2), the specific manifestations have not been identified. Our initial goal in studying TGF-1 signaling was an attempt to identify a mechanism that would explain why nitric oxide synthase (NOS)2 expression is reduced in CF epithelium (5). TGF-1 is known to be an effective negative regulator of NOS2 expression in a variety of cell types, including vascular smooth muscle, peritoneal macrophages, and retinal epithelial cells (8). The TGF-1 null (TGF-1^/) mice spontaneously produce increased levels of NO and exhibit increased NOS2 expression compared with wild-type sibling mice (11). Therefore, increased TGF-1 signaling could be a possible mechanism for reduced NOS2 expression in CF. Another possible effect of increased TGF-1 signaling in CF is the progressive accumulation of fibrotic tissue. TGF-1 is a key factor in the expression of extracellular matrix (ECM) proteins that are important factors in fibrosis as well as overall airway remodeling (12- 16). The integrins are ECM proteins particularly susceptible to TGF-1 regulation (13), and one report has demonstrated that v, 5, and 6-integrins are highly expressed in areas of undifferentiated epithelium in CF lung tissue (17). The expression of integrins has been postulated to be associated with an increased susceptibility to both viral and bacterial infections by providing binding sites for these pathogens (18).

Although our original hypothesis was that increased TGF-1 signaling in CF could be responsible for NOS2 regulation, reduced TGF-1 signaling in CF epithelium could also have disease-related consequences. A characteristic of CF airway disease is an overzealous inflammatory response. Given the antiinflammatory nature of TGF-1, it is possible that reduced TGF-1 signaling could lead to improper inflammatory regulation. It has been reported that TGF-1 can have differential effects on airway cytokine production, leading to reduced interleukin (IL)-8 expression and increased RANTES (regulated upon activation, normal T cells expressed and secreted) expression (19). The opposite ratios have been shown in CF airways with increased IL-8 production and reduced RANTES expression, possibly indicating reduced TGF-1-mediated antiinflammatory signaling (20).

Given the potential relevance of TGF-1 to CF airway disease and the absence of direct study, the goal of this manuscript is to examine TGF-1-mediated signaling in CF epithelial cells. Of specific interest are the TGF-1 responsive R-Smads (Smad2 and Smad3), the Co-Smad (Smad4), and the potential role of TGF-1 in inflammatory regulation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Transfection and Luciferase Assay

9/HTEo- Cells overexpressing the CFTR R domain (pCEPRF) and mock-transfected 9/HTEo- cells (pCEP2) were a generous gift from the lab of Dr. Pamela B. Davis (Case Western Reserve University, Cleveland, OH). Cells were cared for as previously described (23). Cells were seeded at 6.5 × 10⁵ cells/well in 6-well culture dishes 24 h before transfection. For each transfection, 1 µg of 3TP-Lux (provided by Dr. Joan Massague, Sloan Kettering Institute, New York, NY) or 1 µg each of A3-Luc and pCS2, which expresses mouse FAST-1 (provided by Dr. Malcolm Whitman, Harvard Medical School, Boston, MA), were placed into 100 µl of serum-free culture medium. Lipofectamine (Gibco BRL, Gaithersburg, MD) (5 µl) was also placed into 100 µl of serum-free culture medium. The two solutions were mixed at room temperature and allowed to incubate up to 40 min. After rinsing, cells were placed in 0.8 ml serum-free culture medium. The DNA-liposome mixture (0.2 ml) was added to each well and mixed gently. Cells were incubated with the complex for 5 h at 37°C in 95% O₂/5% CO₂. Following incubation, 1 ml of growth medium (20% serum) was added to each well for a final concentration of 10% serum and incubated over night. TGF-1 (2 ng/ml) was then incubated for ~ 12 h at 37°C in 95% O₂/5% CO₂. Cells were then lysed and assayed for luciferase activity according to manufacturer instructions (Promega, Madison, WI). Cells transfected with either the IL-8-Luc construct (provided by Dr. Allan Brasier, University of Texas Medical Branch, Galveston, TX), the NOS2-Luc construct (provided by Dr. Joel Moss, NIH, Bethesda, MD), or the NF-B- Luc construct (Clontech Laboratories, Palo Alto, CA) were stimulated for 4 h (12 h for NOS2-Luc) with a mixture of IL-1 (0.5 ng/ml) and tumor necrosis factor (TNF)- (1.0 ng/ml) with the addition of interferon (IFN)- (100 U/ml) for the NOS2-Luc construct. Wild-type and dominant-negative Smad3 expressing vectors, pEXL-Flag-Smad3 and pEXL-Flag-Smad3a, were used as described in the text (provided by Dr. Xuedong Liu, University of Colorado, Boulder, CO). Luciferase activity was determined on a Tropix (Bedford, MA) automatic injection luminometer. Results were controlled for transfection efficiency between cell lines by normalizing results to luciferase activity in parallel experiments of cells transfected with pTK-Luc obtained from Clontech to obtain relative light units (RLU).

Western Immunoblotting

Antibodies against Smad3(rabbit), Smad4(rabbit), and erk(rabbit) were obtained from Santa Cruz biotechnologies (Santa Cruz, CA). Antibody against Smad2(rabbit) was obtained from Zymed (San Francisco, CA). Protein samples were prepared by homogenizing either excised nasal epithelium or 60 mm dishes of cultured cells in lysis buffer (50 mM Tris, pH 7.5, 1% Triton X-100, 100 mM NaCl, 50 mM NaF, 200 µM Na₃VO₄, and 10 µg/ml pepstatin and leupeptin [all chemicals from Sigma Chemical Co., St. Louis, MO]) on ice. Protein concentration of samples was measured using the Bio-Rad Dc protein assay system (Bio-Rad, Hercules, CA). Proteins were separated using sodium dodecyl sulfate/ polyacrylamide gel electrophoresis (SDS-PAGE) and samples were transferred to Immobilon-P membrane (Millipore, Bedford, MA). Blots were blocked in phosphate-buffered saline (PBS) (NaCl [138 mM], Na₂HPO₄ [15 mM], KCl [1.5 mM], and KH₂PO₄ [2.5 mM]) containing 5% nonfat dehydrated milk and 0.1% Tween-20 (Sigma Chemical Co.). Blots were incubated for 1 to 2 h at room temperature in PBS with primary antibody (1:800 to 1:1,000 dilution) and then washed three times in PBS with 0.1% Tween-20. Blots were then incubated with secondary antibody conjugated to horseradish peroxidase (1:4,000 dilution) (Sigma Chemical Co.) for 1 h at room temperature and washed again as described. Signal was visualized by incubating with Super Signal chemiluminescent substrate (Pierce, Rockford, IL) and exposing the membrane to Kodak scientific imaging film (Kodak, Rochester, NY). Quantification of protein expression was accomplished by densitometry using Kodak Digital Science 1D software. Arbitrary density units refer to mean pixel density.

Mice

Mice lacking CFTR expression (CFTR^tm1Unc) were obtained from Jackson Laboratories, Bar Harbor, MA. CFTR wild-type and wild-type heterozygous mice were siblings of cftr^/ mice. All mice used were between six and eight weeks of age. CF mice were fed a liquid diet as described by Eckman and colleagues (24). Mice were cared for in accordance with Case Western Reserve University IACUC guidelines by the CF Animal Core Facility.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

TGF-1-Mediated Signaling in a Model of CF Epithelial Cells

To assess the ability of TGF-1 to initiate signaling in CF airway epithelial cells, we used a model cell line in which the regulatory (R) domain of CFTR is overexpressed in 9/HTEo- cells (pCEPRF) (23). The overexpression of the R domain has been shown to effectively inhibit CFTR function and to cause characteristic CF phenotypes in these cells, including an overly aggressive inflammatory response, increased bacterial binding, and altered cell surface glycosylation (23, 25). This model system has the advantage of having a matched control cell line, which expresses an empty vector and maintains a wild-type phenotype (pCEP2).

We examined TGF-1-mediated signaling in these cells using two TGF-1-responsive, luciferase-expressing constructs, 3TP-Lux and A3-Luc. Although not limited to specific complexes, 3TP-Lux is optimally stimulated by a Smad3/Smad4 complex, whereas the ARE sites in A3-Luc prefer a Smad2/Smad4/FAST-1 signaling complex (28, 29). We found differential levels of TGF-1-mediated activation of these constructs in the pCEP2 and pCEPRF 9HTEo- cells. Luciferase activity is reported as a percent of TGF-1-stimulated RLU in pCEP2 9/HTEo- cells. Basal luciferase expression in pCEP2 cells transfected with 3TP-Lux was 48.4 ± 4.0% of control (n = 13). However, basal and TGF-1-stimulated luciferase expression in 3TP-Lux transfected pCEPRF 9/HTEo- cells was only 2.2 ± 0.5% (n = 13) and 5.8 ± 1.6% (n = 13), respectively, compared with control cells. 3TP-Lux luciferase expression is reduced almost 20-fold in pCEPRF cells compared with pCEP2 cells in both treated and untreated conditions (P < 0.0001 for each). Although fold stimulation was similar in each cell line (~ 2-fold), basal levels and TGF-1-stimulated levels of total luciferase activity were significantly reduced in the CF-phenotype pCEPRF cells compared with control pCEP2 9/HTEo- cells (Figure 1A). The data demonstrating fold-stimulation is similar, but total activity is reduced, suggesting that levels of signaling intermediates may be altered as opposed to an inability to normally activate these intermediates. Surprisingly, however, there is very little difference in stimulated and total luciferase expression in these two cell lines when using the A3-Luc vector (Figure 1B). Basal luciferase expression in A3-Luc/FAST-1 transfected pCEP2 9/HTEo- cells was found to be 43.4 ± 8.5% (n = 15) of TGF-1 (2 ng/ml)-stimulated cells. In the CF-like pCEPRF 9/HTEo- cells, basal expression was only 14.3 ± 1.8% (n = 15) of stimulated pCEP2 cells, but increased to 104.9 ± 26.5% (n = 15) of control with TGF-1. Basal A3-Luc-mediated luciferase expression was significantly (~ 3-fold) reduced in the pCEPRF cells (P = 0.002), but TGF-1-stimulated luciferase expression reached essentially identical levels between the two cell types. Because Smad3 is key to the activation of the 3TP-Lux construct, we expressed Smad3 in pCEPRF cells using the pEXL-FLAG-Smad3 construct and pEXL-FLAG-GFP as a transfection control to determine if Smad3 would restore signaling in these cells (Figure 1C). Basal luciferase activity in Smad3 expressing pCEPRF cells increased almost 10-fold over GFP-expressing cells. These results suggest the possibility that Smad3-specific TGF-1-mediated signaling pathways are altered in CF while other Smad-dependent pathways are left intact. Therefore, we decided to examine protein expression levels of the TGF-1 signaling proteins, Smads 2, 3, and 4. 

View larger version (18K): [in this window] [in a new window]   View larger version (23K): [in this window] [in a new window]   View larger version (21K): [in this window] [in a new window]   Figure 1.   TGF-1-mediated cell signaling in pCEP2 and pCEPRF 9/HTEo- cells. (A) Cells were transfected with 3TP-Lux and stimulated with 2 ng/ml TGF-1 (+TGF) as described. The number of experiments (n) is shown in parentheses above each bar. Error bars represent SEM. Asterisk represents significance as measured by t test between pCEP2 and pCEPRF cells under each experimental condition (*P < 0.0001). (B). Cells were cotransfected with A3-Luc and pCS2 (FAST-1) and stimulated with 2 ng/ml TGF-1 (+TGF) as described. (C) Cells were cotransfected with 3TP-Lux and either pEXL-Flag-Smad3 (a Smad3-expressing vector) or pEXL-Flag-GFP as a control and stimulated with 2 ng/ml TGF-1 (+TGF) as described. The number of experiments (n) is shown in parentheses above each bar. Error bars represent SEM. Asterisk represents significance as measured by t test between pCEP2 and pCEPRF cells under each experimental condition (*P = 0.002). All data are normalized for transfection efficiency as determined by parallel transfections with pTK-Luc and presented as relative light units (RLU). Solid bars: TGF; striped bars: +TGF.

Smad2, Smad3, and Smad4 Protein Expression Levels in pCEP2 and pCEPRF 9/HTEo- Cells

The observation of a differential ability for the CF-like pCEPRF 9/HTEo- cells to support the stimulation of two separate TGF-1-responsive promoters prompted us to examine the protein expression levels of Smad2, Smad3, and Smad4 (Figure 2). We found that Smad2 levels were slightly, but significantly (P < 0.001) reduced in pCEPRF cells compared with pCEP2 cells, although there was variability in its expression. Smad4 levels, which always appear as a doublet in our samples (possibly due to cross reactivity of the antibody with another Smad), were identical between the two cell lines. Smad3 levels, however, were dramatically reduced in the CF-phenotype pCEPRF cells compared with the control pCEP2 9/HTEo- cells (P < 0.0001). As a control, erk levels were measured to verify protein concentration in each preparation, and no differences were found in tested samples.

View larger version (34K): [in this window] [in a new window]   Figure 2.   Smad2, Smad3, Smad4, and erk expression in pCEP2 and pCEPRF 9/HTEo- cells. Representative blots of Smad2 (A), Smad3 (B), Smad4 (C), and erk (D) protein expression in pCEP2 9/HTEo- control cells (lanes 1-3) and CF-phenotype pCEPRF 9/HTEo- cells (lanes 4-6). Densitometry analysis (E) of Smad2, Smad3, Smad4, and erk expression. Erk levels were measured as a control for protein loading. Number of samples (n) is shown in parentheses above each bar. Significance determined by t test (*P < 0.001; **P < 0.0001). Solid bars: pCEP2; striped bars: pCEPRF. Error bars represent SEM.

Impact of Reduced Smad3 Expression on Epithelial Inflammatory Responses

TGF-1 is a known antiinflammatory cytokine capable of modulating the production of proinflammatory agents such as IL-8. A defining characteristic of CF airway disease is overly aggressive inflammation, and the abnormal regulation of the inflammatory response in CF may be due in part to altered TGF-1-mediated signaling. To test this hypothesis, we examined the ability of TGF-1 to modulate IL-8 promoter activity in response to a mixture of IL-1 (0.5 ng/ml) and TNF- (1.0 ng/ml) (IL-1/TNF-) in the pCEP2 and CF-phenotype pCEPRF 9/HTEo- cells using a construct in which the human IL-8 promoter drives luciferase expression (IL-8-Luc). In the normal phenotype pCEP2 cells, the addition of TGF-1 to IL-1/TNF--treated cells reduced IL-8-Luc expression to 68.1 ± 5.9% (n = 8; P = 0.0003) of the stimulation level in cells treated with IL-1/TNF- alone (Figure 3A). However, the CF-phenotype pCEPRF cells exhibited no response to TGF-1 in modulating IL-8 promoter activity in response to IL-1/ TNF- (Figure 3B).

View larger version (23K): [in this window] [in a new window]   Figure 3.   Smad3 is necessary for TGF-1-mediated regulation of IL-8 and NOS2 promoter activity. TGF-1 (2 ng/ml) reduces IL-1 (0.5 ng/ml)- and TNF- (1.0 ng/ml)-stimulated luciferase expression driven by the human IL-8 promoter (IL-8-Luc) in 9/HTEo- pCEP2 cells (A), but not in CF-phenotype pCEPRF cells (B). These results are mimicked in A549 cells by expressing either wild-type Smad3 (pEXL-Flag-Smad3) or a dominant-negative Smad3 (pEXL-Flag-Smad3a). TGF-1 (2 ng/ml) reduces IL-1 (0.5 ng/ ml)- and TNF- (1.0 ng/ml)-stimulated IL-8 promoter activity in Smad3 expressing A549 cells (C), but not in dominant-negative Smad3a expressing A549 cells (D). Results are presented as % of stimulated luciferase activity in each cell line with n = 8 for each sample. TGF-1 (2 ng/ml) also reduces cytomix (CM) (IFN- [100 U/ml], IL-1 [0.5 ng/ml], and TNF- [1.0 ng/ml]) stimulated NOS2 promoter activity in Smad3 expressing A549 cells (E), but not in dominant-negative Smad3a expressing A549 cells (F). Results are presented as % of stimulated luciferase activity in each cell line with n = 3 for each sample. Significance determined by t test (*P < 0.001; **P < 0.01; ***P < 0.05; NS = not significant). (A, B, C, and D) Solid bars: not treated; open bars: IL-1/TNF; striped bars: IL-1/TNF + TGF. (E and F) Solid bars: not treated; open bars: CM; striped bars: CM + TGF. Error bars represent SEM.

There is a clear difference in TGF-1-mediated regulation of IL-8 promoter activity between the pCEP2 and pCEPRF 9/HTEo- cells. To determine if this CF-phenotype could be mimicked by manipulating Smad3, we used another airway epithelial cell line known to respond to TGF-1 (A549) and cotransfected the IL-8-Luc reporter construct with either a wild-type Smad3-expressing vector (pEXL-Flag-Smad3) or a dominant-negative Smad3- expressing vector (pEXL-Flag-Smad3a). TGF-1 reduced IL-8 promoter activity in A549 cells expressing the wild-type Smad3 stimulated with IL-1/TNF- to 68.4 ± 6.3% compared with IL-8 promoter activity in cells stimulated with IL-1/ TNF- alone (Figure 3C). TGF-1 had no effect on IL-8 promoter activity in A549 cells expressing a dominant-negative Smad3 (Figure 3D). These results are identical to those observed in the pCEP2 and pCEPRF 9/HTEo- cell models and demonstrate directly the role of Smad3 in regulating IL-8 promoter activity. We also examined the role of Smad3 in the regulation of the NOS2 promoter. As was observed with the IL-8 promoter, the dominant-negative form of Smad3 (Smad3a) prevented the negative regulation of the NOS2 promoter by TGF-1 in response to stimulation by cytomix (CM: IFN- [100 U/ml], IL-1 [0.5 ng/ml], and TNF- [1.0 ng/ml]).

Both the IL-8 and NOS2 promoters are regulated in part by NF-B activity. To determine if TGF-1-mediated regulation of these promoters was due to an influence over NF-B activation, we used an NF-B-Luc construct as a reporter system. We tested the ability of TGF-1 to influence NF-B activity in the presence of wild-type or dominant-negative Smad3 to examine the role of Smad3 in antiinflammatory regulation. We found that in the presence of wild-type Smad3, TGF-1 inhibited IL-1/TNF--stimulated NF-B activation 50.0 ± 1.4% (Figure 4). However, coexpression of the dominant-negative form of Smad3 (Smad3a) reduced this TGF-1-mediated inhibition to only 26.3 ± 3.3% (P < 0.001; n = 7 for each). These data suggest that TGF-1 may be able to partially regulate the IL-8 and NOS2 promoters through its ability to influence NF-B activity via a Smad3-dependent mechanism, although other mechanisms are likely involved.

View larger version (19K): [in this window] [in a new window]   Figure 4.   TGF-1-mediated regulation of NF-B activity in A549 cells. TGF-1 (2 ng/ml) reduces IL-1 (0.5 ng/ml)- and TNF- (1.0 ng/ml) (I/T)-stimulated luciferase expression driven by NF-B activity (NF-B-Luc) in A549 cells. Results are presented as % of stimulated luciferase activity in each cell line with n = 7 for each sample. The role of Smad3 in this process was examined by expressing either wild-type Smad3 (pEXL-Flag-Smad3) or a dominant-negative Smad3 (pEXL-Flag-Smad3a) in A549 cells. TGF-1 (2 ng/ml) reduces IL-1- and TNF--stimulated NF-B activity 50.0 ± 1.4% in Smad3 expressing A549 cells, but only 26.3 ± 3.3% in dominant-negative Smad3a expressing A549 cells (solid bars: not treated; open bars: IL-1 and TNF- treated; striped bars: IL-1/TNF- + TGF treated). Results are presented as % of stimulated luciferase activity in each cell line with n = 7 for each sample. Significance determined by t test (*P < 0.001). Error bars represent SEM.

Smad2, Smad3, and Smad4 Protein Expression Levels in Nasal Epithelium Isolated from cftr^+/ and cftr^/ Mice

Although the pCEP2 and pCEPRF 9/HTEo- cells have proven to be useful models in the study of various CF-phenotypes, they are clonal, stably transfected cultured cells which may be prone to cell signaling artifacts. To verify our CF-related findings of reduced Smad3, we examined excised nasal epithelium for protein expression levels of Smad2, Smad3, and Smad4. Mouse nasal epithelial (MNE) samples yielded results identical to those found in the pCEP2 and pCEPRF cells (Figure 5). Smad2 protein levels were slightly but significantly reduced in the cftr^/ MNE (P < 0.001). A doublet occurs in the cftr^+/ samples, which coincides with cross reactivity of the antibody with Smad3. In agreement with our observations in the cultured cell model, Smad3 levels were found to be dramatically reduced in the cftr^/ MNE samples compared with cftr^+/ MNE samples (P < 0.0001). No differences were found in either Smad4 or erk expression levels. Smad4 appeared as a doublet in the MNE samples, just as seen in the pCEP2 and pCEPRF samples. These results confirm in two separate systems that there is a CF-related reduction in Smad3 protein expression.

View larger version (34K): [in this window] [in a new window]   Figure 5.   Smad2, Smad3, Smad4, and erk expression in mouse nasal epithelium excised from cftr+/ and cftr/ mice. Representative blots of Smad2 (A), Smad3 (B), Smad4 (C), and erk (D) protein expression in excised nasal epithelium from cftr+/ mice (lanes 1-3) and cftr/ mice (lanes 4-6). Each lane represents tissue isolated from an individual mouse. Densitometry analysis (E) of Smad2, Smad3, Smad4, and erk expression. Erk levels were measured as a control for protein loading. Number of samples (n) is shown in parentheses above each bar. Solid bars: cftr+/; striped bars: cftr/. Significance determined by t test (*P < 0.001; **P < 0.0001). Error bars represent SEM.

Nonepithelial Smad3 Expression in cftr^+/ and cftr^/ Mice

Because CFTR is expressed primarily in epithelial tissue, we wanted to determine if alterations in Smad3 expression were limited to epithelium (Figure 6). We examined Smad3 and erk levels in extract from whole liver tissue from cftr^+/ and cftr^/ mice. No significant difference in either Smad3 or erk levels was found in liver tissue. These results indicate that there is a tissue-specific component to CF-related changes to Smad3 expression.

View larger version (46K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   Figure 6.   Smad3 and erk expression in whole liver tissue excised from cftr+/ and cftr/ mice. (A) Blot of Smad3 and erk (simultaneously probed on the same gel) protein expression in whole liver tissue from cftr+/ mice (lanes 1-4) and cftr/ mice (lanes 5-8). Each lane represents tissue isolated from an individual mouse. (B) Densitometry analysis of Smad3 and erk expression (n = 4 for each). Erk levels were measured as a control for protein loading. No significant differences were observed. Solid bars: cftr+/; striped bars: cftr/. Error bars represent SEM.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We have identified a significant alteration in the TGF-1 signaling cascade in CF epithelium. Two R-Smads responsive to TGF-1, Smad2 and Smad3, exhibit reduced protein expression in both a cultured cell model of CF and in the nasal epithelium of cftr^/ mice. Smad2 exhibits only slightly reduced expression and, although statistically significant, it is unlikely biologically significant as demonstrated by essentially normal signaling in the CF phenotype pCEPRF cells. Smad3, in particular, exhibits significantly reduced protein expression in CF epithelial cells, and this reduction is apparently sufficient to influence the proper transmission of certain TGF-1-stimulated signaling events, including antiinflammatory regulation.

The potential consequences of altered Smad3 signaling in epithelial cells need to be explored. Overzealous inflammation in response to bacterial pathogens, or even in the basal state according to some reports, is a defined characteristic of CF (1, 30). This inflammatory response is reportedly modulated by antiinflammatory cytokines such as IL-10 (31), and the lack of efficient antiinflammatory signaling by TGF-1 in airway epithelial cells due to reduced levels of Smad3 may be a contributing factor leading to increased inflammatory cytokine production. It has already been shown that TGF-1 effectively reduces IL-8 expression and increases RANTES expression in airway epithelial cells (19). A loss of TGF-1 signaling in CF epithelial cells would be consistent with reports that demonstrate CF-related increases in IL-8 expression and decreases in RANTES expression (20). We have shown in A549 cells that TGF-1-mediated negative regulation of the IL-8 promoter is Smad3-dependent, results identical to those we observe in the pCEP2 and pCEPRF 9/HTEo- cells. We also examined the ability of TGF-1 to regulate NOS2 promoter activity and the role of Smad3 in that regulation. We found that a dominant-negative Smad3 prevented TGF-1-mediated regulation of NOS2 in A549 cells, data consistent with previous reports examining NOS2 regulation in macrophages (32). The original impetus to examine TGF-1 signaling in CF was to attempt to identify a mechanism responsible for reduced NOS2 expression in CF epithelium. Based on both our findings and the findings of others on NOS2 regulation in macrophages, it is unlikely that TGF-1 and Smad3 play an active role in the altered NOS2 expression in CF.

Although not likely a contributor to altered NOS2 expression characteristic of CF, reduced Smad3 expression may be involved in inflammatory regulation. Common to both IL-8 and NOS2 expression is their dependency on NF-B activity. Consistent with other reports (33), our data demonstrate that TGF-1 acts as a negative regulator of NF-B activation and that Smad3 plays at least a partial role in this regulation. Smad3 deficiency represents a possible mechanism responsible for aberrant inflammatory responses characteristic of CF due to partially reduced negative regulation of NF-B activity and the subsequent regulation of the expression of inflammatory mediators such as IL-8.

It is currently unclear why CF epithelium would exhibit this alteration of Smad3 levels. One possibility is that a feedback regulatory mechanism is responsible for the control of Smad3 expression. It has been previously shown that chronic treatment of airway epithelial cells with TGF-1 causes reduced expression of Smad protein levels and message levels (34). It is possible that a loss of CFTR function alters cell surface interactions sufficiently that the cell interprets this as a stimulation of the TGF-1 pathway. It is also possible that altered Smad3 expression is a response by the CF epithelium to restore NOS2 expression by eliminating a negative regulatory process. These potential mechanisms responsible for reduced Smad3 expression in CF epithelium are currently being explored.

Our findings identify a unique, tissue-specific alteration in an important signaling pathway in CF epithelial cells that may have implications involving the modulation of inflammatory responses. A better understanding of the role of altered TGF-1 signaling in CF will hopefully lead to more effective antiinflammatory therapies. CF-specific alterations in Smad3 expression also provide a naturally occurring system to study the in vivo role of Smad3 in regulating various cellular events.