Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Following lung injury, fibroblasts can respond to the subsequent inflammatory milieu in different ways, resulting eventually in either restorative repair or disordered remodeling. It is increasingly evident that not all fibroblasts are equal. Heterogeneous subpopulations of fibroblasts have separate, distinct roles in tissue homeostasis and response to injury (1). We have previously described that primary rat lung fibroblasts sorted on the basis of surface expression of the lipid raft glycoprotein Thy-1 (Thy-1 and Thy-1+) demonstrate differential proliferative responses to the profibrotic platelet-derived growth factor-AA (PDGF-AA), mediated by differential expression of the  isoform of PDGF receptor (PDGFR) (5). Proliferative responses to PDGF-BB and expression of PDGFR were not found to differ. Interleukin (IL)-1 has been shown to be an important regulator of the PDGFR in lung myofibroblasts in a number of models (6, 7). We therefore hypothesized that Thy-1 and Thy-1+ fibroblasts would differ significantly in their response to IL-1.

IL-1 is a paradigmatic cytokine and a central regulator of inflammation. Its effects on parenchymal cells mediate tissue remodeling at sites of inflammatory injury. Cellular responses to IL-1 are regulated by complex, interactive pathways involving three ligands (IL-1 and -1, and IL-1 receptor antagonist [IL-1RA]), and two receptor subtypes, IL-1RtI and IL-1RtII. The latter of these acts as a "decoy" (nonsignaling) receptor, and both can be shed from the cell surface. A receptor-accessory protein (IL-1RAcP) is necessary to initiate signaling, which proceeds through a complex network of interacting intracellular signaling intermediates to regulate dozens of downstream genes (8, 9). This is an ancient and canonical signal response pathway, with homologs in plants and insects (10). Differential responses to IL-1 therefore indicate markedly differing cellular functions. In fibroblasts, both  and  isoforms of IL-1 are known to stimulate proliferation, largely through an autocrine PDGF-AA-mediated pathway (11, 12). Fibroblast responses to IL-1 thus serve as a link between inflammatory and remodeling (or fibroproliferative) events. IL-1 also affects fibroblasts' ability to feed back and modulate inflammatory responses, such as through induction of IL-6 (13).

Here we report significant differences in fibroblast subpopulation responses to IL-1. In Thy-1, but not Thy-1+, rat lung fibroblasts, IL-1 enhances, and IL-1 RA inhibits, expression of PDGFR. Binding of IL-1 and expression of receptor subunits are equivalent in the two subpopulations, as are downstream activation of binding of both nuclear factor (NF)-B and CAATT-enhancer binding protein (C/EBP) transcription factors, and activation of p38 mitogen-activated protein kinase. However, in addition to PDGFR expression, both the IL-1-induced expression of IL-6 and IL-1-induced proliferation are discordant in fibroblast Thy-1 subpopulations. These differential IL-1-mediated responses, despite equivalent surface binding and receptor expression, indicate that divergent signaling pathways are activated in response to IL-1, and that these differ between subpopulations of fibroblasts. Our findings reinforce the importance of Thy-1 as a marker of biologically relevant subsets of fibroblasts, and underscore the existence of separate, distinct roles for fibroblast subpopulations in the interface between inflammation and tissue remodeling.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cells and Reagents

The sorting and characterization of primary rat lung fibroblast Thy-1 subpopulations have been previously described (5, 14). U937 cells were the kind gift of Alan Cooper, M.D. (Birmingham, AL). Recombinant murine IL-1 and human IL-1 RA were from R&D Systems, Inc. (Minneapolis, MN). Antibodies to IL-1 receptor type I, PDGFR, and PDGF-AA were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Antibody to IL-1 receptor type II and the IL-1 Fluorokine-binding assay were obtained from R&D Systems. The rat IL-1 enzyme-linked immunosorbent assay (ELISA) kit was obtained from BioSource International (Camarillo, CA). Recombinant murine IL-1RtII was the kind gift of Dr. Mark Wewers (Ohio State University). Antibody to IL-1 receptor accessory protein (IL-1RacP) was obtained from Pharmingen/Transduction Laboratories (San Diego, CA). Antibodies to total and phosphorylated p38 were obtained from Cell Signaling Technology (Beverly, MA). A 900-bp fragment of murine IL-6 cDNA was used for Northern blotting (15). Oligonucleotides corresponding to the consensus binding sites for NF-B (GATCCAT GGGGAATTCCCCATG) (16) and C/EBP (TGCAGATTGCGC AATCTGCA), as well as antibodies to p50 and p65 subunits of NF-B, all obtained from Santa Cruz Biotechnology, were used in electrophoretic mobility shift assays (EMSAs).

Proliferation Assay

Fibroblasts were plated at 15,000 cells/250 µl in 24-well plates and allowed to attach overnight. Monolayers were washed twice with serum-free medium (SFM) and rendered quiescent in medium containing 0.4% fetal bovine serum (FBS) for 48 h. Recombinant murine IL-1 was added as indicated in SFM for 8 h, after which [³H]-thymidine (Amersham, Arlington Heights, IL) was added to a final concentration of 5 µCi/ml for an additional 16 h (24 h total). Wells were gently aspirated, washed three times with SFM, and placed on ice. Monolayers were treated with ice-cold 5% trichloroacetic acid (TCA) for 15 min., washed, and solubilized in pre-warmed (37°C) buffer (0.2 N NaOH, 0.1% sodium dodecyl sulfate (SDS) for 30 min at 37°C before scintillation counting. Each condition was assayed in triplicate. Wells in which medium alone, either SFM or containing 10% FBS, was added were used to determine [³H]-thymidine uptake in quiescence and log-phase growth, respectively. The counts per minute obtained for cells cultured in SFM were averaged and subtracted from all experimental values, and these are expressed as arbitrary units, with average cpm in wells exposed to 10% FBS alone set at 100. We have previously compared this assay to one using direct cell counts and found excellent correlation (data not shown).

IL-1 Binding by Flow Cytometry

IL-1 binding to Thy-1 and Thy-1+ fibroblasts was determined by flow cytometry following binding of biotinylated IL-1 (Fluorokine Kit; R&D Systems). Briefly, fibroblasts were suspended at subculture and washed twice in phosphate-buffered saline (PBS) at 4°C. 10⁵ cells were incubated with biotinylated IL-1 for 60 min, after which avidin-fluorescein isothiocyanate (FITC) was added for an additional 30 min. The cells were then washed before analysis by flow cytometry. As a control for specificity, cells were blocked with immunoglobulin G (IgG) to block Fc-mediated interactions, incubated with biotinylated IL-1 which had been preincubated with anti-human IL-1 blocking antibody, then stained with avidin-FITC. No significant binding was demonstrated over that seen when avidin-FITC alone was added to cells (data not shown).

Western Immunoblotting

Thy-1 and Thy-1+ fibroblasts were grown in 100-mm dishes to near confluence and rendered quiescent. Cells were then exposed to the mediators indicated for an additional 24 h in SFM. Protein from whole cell lysates prepared in the presence of protease inhibitors was loaded onto 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gels at 10 µg/lane, based on Bradford assay (Pierce Chemical Co., Rockford, IL), and separated electrophoretically along with markers of known molecular weight. Protein was transferred onto polyvinylidine fluoride filters which were blocked with 3% milk/PBS and probed with rabbit anti-mouse PDGFR at 2 µg/ml, followed by a second block and goat anti-rabbit horseradish peroxidase (BioRad, Hercules, CA) at 1:2,000. The filters were then developed with an enhanced chemiluminescent (ECL) detection system and autoradiographs generated.

Cells cultured in the presence or absence of dexamethasone for 24 h in serum-free medium, as above, were used to determine expression of IL-1 receptors and IL-1RacP. Supernatants were removed and briefly centrifuged to remove nonattached cells and debris, then concentrated using centrifugal concentrators (Millipore, Bedford, MA) before SDS-PAGE under reducing conditions. Cell lysates were prepared as above in the presence of protease inhibitors. Supernatants and lysates were immunoblotted with polyclonal antibodies to IL-1Rt I (Santa Cruz Biotechnology, Santa Cruz, CA) or Rt II (R&D Systems). For IL-1RtI, a biotinylated goat-anti-rabbit secondary antibody was used at 10 ng/ml followed by Extravidin Peroxidase (Sigma, St. Louis, MO) at 1:40,000. Lysates only were used to determine IL-1RacP.

For determination of total and phospho-specific p38 mitogen-activated protein (MAP) kinase, quiescent monolayers were stimulated with 5 ng/ml rmIL-1 for 30 min, and cell lysates prepared in the presence of 1.0 mM Na Orthovanadate (Sigma), before SDS-PAGE (40 µg cell lysate protein/well) and immunoblotting with phosphorylation-specific antibodies to p38, then stripped and reprobed with antibody to total p38.

PDGF-AA Determination

Subconfluent monolayers of sorted fibroblasts were rendered quiescent in medium containing 0.1% FBS, then stimulated with either rmIL-1, 25 or 100 pg/ml in SFM, or SFM control, for an additional 24 h. Conditioned media (CM, 2 ml) were collected in the presence of protease inhibitors, centrifuged to remove debris, and concentrated 5-fold using 3,000 MWCO filters (Centricon-3; Millipore). Cells were collected by scraping into cold PBS with 2% FBS and protease inhibitors, and lysed by sonication, followed by centrifugation. Capture antibody (goat anti-PDGF-A antiserum; Santa Cruz Biotechnology) was added to wells of a 96-well microtiter plate and incubated overnight at 4°C. After washing the wells three times with PBS + 0.1% Tween-20 (PBST) and blocking with 1% BSA in PBST, cell lysates, concentrated supernatants, or standards (rhPDGF-AA, 1-100 ng/ml; Genzyme, Inc., Cambridge, MA) were added in duplicate or triplicate to individual wells for 90 min at 37°C. Wells were washed three times with PBST and incubated with detection antibody (rabbit anti-PDGF-A antiserum; Santa Cruz Biotechnology), 1.6 µg/ml, 1 h at 37°C. After washing, wells were incubated with biotin-conjugated goat anti-rabbit IgG (Southern Biotechnology Associates, Birmingham, AL), 1 µg/ml, 30 min at 37°C, then washed again and incubated with streptavidin-horseradish peroxidase (Pierce), 1 µg/ml, 20 min, prior to development with slow TMB (Pierce). Development was stopped with 20% H₂SO₄ and absorbance read at 450 nm in a microplate reader. Resulting absorbances were converted to concentrations by comparison to standards using the Microplate Manager software (BioRad).

IL-1 Determination

Subconfluent monolayers of sorted fibroblasts were treated with IL-1 or tumor necrosis factor (TNF)- and processed as for PDGF-AA determination, above. A commercial ELISA kit was used to determine the concentration of IL-1 in cell lysates and conditioned media (BioSource International). In each plate, rmIL-1 was added in known concentrations to determine possible cross-reactivity with IL-1. None was detected.

IL-6 mRNA Expression

Sorted Thy-1 and Thy-1+ fibroblasts were rendered quiescent in 0.4% FBS and then stimulated with rm IL-1 (R&D Systems) at 0.2-5 ng/ml for 8 h. Total RNA was prepared as previously described (17) and 10 µg in denaturation buffer was added to each well of a 1.2% agarose gel and electrophoretically separated. RNA was then transferred to a nylon membrane (Hybond-N; Amersham-Pharmacia Biotech, Piscataway, NJ) by capillary action. Prehybridization and hybridization were performed in NorthernMAX buffer (Ambion, Austin, TX). Membranes were prehybridized 3 to 5 h at 62°C and hybridized overnight at 62°C, with a cDNA probe specific for murine IL-6 labeled with [³²P]-dCTP using random primers. Hybridized membranes were washed with 2× SSC, 0.5% SDS, then 0.2× saline sodium citrate, 0.5% SDS at 37°C and 62°C. The resulting hybridized signal is detected by autoradiography and quantified using image phosphor analysis (Phosphor Imager and ImageQuant; Molecular Dynamics, Sunnyvale, CA).

EMSA

Thy-1 and Thy-1+ fibroblasts were exposed to rmIL-1 (1 and 5 ng/ml) or SFM for 45 min and nuclear extracts were prepared as described (18, 19). Washed cells (10⁶-10⁷) were pelleted in PBS (4°C) and lysed by repeated pipetting at 4°C (in 10 mM Hepes, pH 7.9, 1.5 mM MgCl₂, 10 mM KCL, 0.5 mM dithiothreitol [DTT] and 0.2 mM phenylmethylsulfonyl fluoride [PMSF]). Nuclear proteins were extracted (420 mM NaCl, 20 mM Hepes, pH 7.9, 1.5 mM MgCl₂, 0.5 mM DTT, and 0.2 mM PMSF) and total protein was measured by commercially available methods (Bradford method; Pierce). Oligonucleotides (see CELLS AND REAGENTS, above) were end-labeled with [-³²P]-ATP using the T4 polynucleotide kinase (Roche Molecular Biochemicals, Indianapolis, IN) reaction resulting in a specific activity of at least 10⁶ cpm/ng DNA. The labeled oligonucleotide (200,000 cpm) was allowed to combine with 1 µg of nuclear protein for 10 min at 25°C (10 mM Tris-Cl, pH 7.5, 10% glycerol, 50 mM KCl, 2 mM MgCl₂, 200 µg/ml bovine serum albumin, 0.05% Nonidet P-40, 1 mM DTT, and 0.2 mM PMSF) in the presence of the nonspecific competitor poly dI-dC (100 µg/ml). For specific competition experiments, excess (100-fold) of "cold" (unlabeled) oligonucleotide was added to the labeled probe before binding reaction. For "supershifting," antibodies to the p50 or p65 subunit of NF-B (Santa Cruz Biotechnology) were added at 2 µg/reaction for an additional 30 min at 25°C. In all cases, the entire reactions (20 µl) were electrophoresed under nondenaturing conditions (6% acrylamide/0.5XTBE, 10V/cm for 90 min at 4°C) followed by fixation in 10% methanol/10% glacial acetic acid, vacuum drying, and autoradiography. Autoradiograms were made in light-tight cassettes with intensifier screens at 70°C (Kodak XAR film, Eastman Kodak Co., Rochester, NY).

Statistical Analysis

To test for differences in signal from mediator-exposed samples and controls in Western and Northern blotting and ELISA, or differences in proliferation, one-way analysis of variance or paired Student's t test was performed on the interval data generated by scanning of autoradiographs or storage phosphor scanning of hybridized membranes, or on data calculated from measured absorbance in ELISA or measured cpm in proliferation assays. Where significant differences were noted, a Dunnett's method multiple comparisons procedure was employed to test for differences at particular mediator concentrations. Significance was accepted at a P value of < 0.05 for all analyses (20).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

PDGFR Expression Is Modified by IL-1 and IL-1RA in Thy-1, but not Thy-1+, Fibroblasts

To determine the basis for differential expression of PDGFR which we had previously described in Thy-1 and Thy-1+ fibroblast subpopulations, we explored the regulation of PDGFR by IL-1 and IL-1RA. Quiescent Thy-1 and Thy-1+ fibroblasts in 100-mm dishes were exposed to recombinant murine IL-1 and human IL-1RA (R&D Systems) for 24 h in SFM. Protein from whole cell lysates was immunoblotted with antibodies to PDGFR. Figure 1 demonstrates a representative autoradiograph (inset) and quantitation of densitometric scanning. The data reported in Figure 1 demonstrate that IL-1 stimulates increased PDGFR expression by Thy-1, but not Thy-1+ fibroblasts. The IL-1RA data (designated "RA" in graph) show that the baseline and FBS-mediated differences in R expression are partly, but not entirely, due to autocrine IL-1 signaling, likely mediated by IL-1.

View larger version (33K): [in this window] [in a new window]   Figure 1.   IL-1 effects on PDGF  receptor. Thy-1 and Thy-1+ fibroblast lysates were immunoblotted for the  isoform of the PDGF receptor as described in MATERIALS AND METHODS. A representative ECL autoradiograph is shown (inset). Fibroblasts were incubated with rmIL-1 (5 ng/ml) or rhIL-1RA (500 ng/ml) for 24 h in SFM or 10% FBS. Numbers at left of autoradiographs indicate relative migration of molecular size standards (kD). The graph indicates densitometric analysis of immunoblot autoradiographs (n = 3). *P = 1 × 105 versus Thy-1(+)/SFM; +P = 0.0004 versus Thy-1()/SFM; ^P = 0.02 versus Thy-1()/SFM.

There Is No Significant Difference in Binding of IL-1 to Thy-1 and Thy-1+ Fibroblasts

As a first step to determine a mechanism for differential responses to IL-1, its binding to Thy-1 and Thy-1+ fibroblasts was measured by flow cytometry. Figure 2 shows that biotinylated IL-1 is able to bind to a qualitatively similar degree to both subpopulations.

View larger version (29K): [in this window] [in a new window]   Figure 2.   Flow cytometric analysis of IL-1 binding. Thy-1 and Thy-1+ fibroblasts were incubated with biotinylated IL-1 and avidin-FITC as described in MATERIALS AND METHODS. Representative flow cytometry histograms are shown. "Con" indicates fibroblasts incubated with avidin-FITC alone.

IL-1 Receptor Components Are Expressed at Equivalent Levels in Both Subpopulations

Because there are IL-1-binding events which do not result in signaling (e.g., binding to IL-1RtII), we explored the expression of IL-1 receptor subtypes in fibroblast subsets. In whole cell lysates, equivalent levels of IL-1RtI are detected in both subpopulations using a biotin-amplified immunoblotting technique (Figure 3, upper panel). Soluble IL-1RtI was not detected in conditioned medium of either subpopulation (data not shown). IL-1RtII, a "decoy" receptor for IL-1, was not detected in cell lysates or conditioned media of either subpopulation. Positive control protein (recombinant murine sIL-1RtII) was readily detected at 10 ng/lane (not shown).

View larger version (47K): [in this window] [in a new window]   Figure 3.   Immunodetection of IL-1 receptor type I (upper panel) and IL-1 receptor-associated protein (lower panel). Whole-cell lysates (20 µg) from quiescent fibroblasts exposed to either SFM (upper panel, lanes 2 and 4) or 107 M dexamethasone (upper panel, lanes 3 and 5) were subjected to SDS-PAGE under reducing conditions and immunoblotted with anti-IL-1RtI as described in MATERIALS AND METHODS. Lane 1 is dexamethasone-stimulated U937 whole cell lysate. A representative ECL autoradiograph is shown. The expected molecular size of IL-1RtI is 80 kD. Whole-cell lysates (10 µg) from quiescent fibroblasts exposed to either SFM (lanes 1, 2, 5, and 6, lower panel) or 107 M dexamethasone (lanes 3, 4, 7, and 8, lower panel) were subjected to SDS-PAGE under reducing conditions and immunoblotted with anti-IL-1 RacP as described in MATERIALS AND METHODS. A representative ECL autoradiograph is shown. The expected molecular size of IL-1RacP is 66 kD.

Expression of the IL-1 Accessory Protein Is Equivalent in Fibroblast Thy-1 Subpopulations

When IL-1 binds to the type I (signaling) receptor, the IL-1 receptor accessory protein (IL-1RacP) must be present, in order for association of the IL-1 receptor-associated kinase (IRAK) and signaling to occur (21). Immunoblotting of lysates of both Thy-1 fibroblast subpopulations demonstrates equivalent levels of IL-1RacP (Figure 3, lower panel).

IL-1 Stimulates Nuclear DNA Binding of NF-B and C/EBP Transcription Factors in Both Thy-1 and Thy-1+ Fibroblasts

To determine whether binding of IL-1 to its receptor initiates distinct intracellular signaling events in Thy-1 subsets, we explored events immediately upstream of gene expression, namely induction of transcription factor binding. Nuclear extracts from IL-1-stimulated fibroblasts were used in EMSA with oligonucleotides representing binding sites for NF-B. Figure 4 demonstrates that stimulation with IL-1 results in induction of nuclear proteins that bind specifically to NF-B DNA-binding elements (appearance of band B and increased intensity of band D) in both subpopulations. Unlabeled probes at 100-fold molar excess significantly compete for binding to labeled probe (including bands C, E, and F, which are present constitutively). Supershifting indicates that band B represents p65 homodimer and that band D likely represents p50/p65 heterodimer. An additional p65-containing band, band A, is seen in IL-1-stimulated Thy-1+ cells. Band A also corresponds to the retarded migration of p50 in the presence of antibody, as seen in Thy-1 lanes. EMSA with C/EBP (not shown) demonstrates increased binding of nuclear protein to consensus oligonucleotide in both subpopulations after IL-1 stimulation.

View larger version (111K): [in this window] [in a new window]   Figure 4.   Electrophoretic mobility shift assay for NF-B. Nuclear extracts (NE) were prepared from Thy-1 and Thy-1+ rat lung fibroblasts exposed to either SFM alone or SFM with IL-1 (5 ng/ml) for 45 min, and incubated in the presence of 32P-labeled NF-B consensus oligonucleotides, in the presence or absence of 100-fold excess unlabeled oligonucleotide (102×) or antibody (Ab) to NF-B subunits p50 or p65, as decribed in MATERIALS AND METHODS. Protein-bound and free oligonucleotides were separated by electrophoresis followed by autoradiography. Retarded bands are indicated by letters A-F at left (see RESULTS).

IL-1 Stimulates Phosphorylation of p38 MAP Kinase in Thy-1 and Thy-1+ Fibroblasts

Because IL-1-induced activation of p38 MAP kinase has been shown to stabilize PDGFR mRNA in myofibroblasts (22), we measured baseline and IL-1-induced phosphorylation of p38 in sorted Thy-1 fibroblast subsets, by immunoblotting for total and phospho-p38. Figure 5 demonstrates that p38 is phosphorylated in both subsets in response to IL-1.

View larger version (49K): [in this window] [in a new window]   Figure 5.   Immunoblot of total and activated p38 MAP kinase. Thy-1 and Thy-1+ fibroblast lysates (SFM control and IL-1-stimulated) were immunoblotted using antibodies to total p38 (upper panel) and phosphorylated p38 (lower panel). 3T3: unstimulated NIH 3T3 fibroblasts (control for total p38). ECL autoradiographs shown are representative of three separate experiments. Numbers at right indicate relative migration of molecular size markers (kD).

IL-1 Stimulates Differential Proliferation and Expression of IL-6 mRNA in Thy-1 and Thy-1+ Fibroblasts

Because of the near equivalence of binding of IL-1, expression of receptor components, and induction of transcription factor binding and p38 MAP kinase phosphorylation, we determined whether other downstream events controlled through IL-1 signaling differ in Thy-1 subsets, namely proliferation and expression of IL-6.

To assess the overall effect of IL-1 stimulation on proliferation of fibroblast subpopulations, we determined uptake of [³H]thymidine following stimulation of fibroblasts with IL-1. Figure 6 demonstrates that IL-1 induces proliferation in Thy-1, but not in Thy-1+, fibroblasts. Both subsets have proliferative responses to serum (not shown).

View larger version (14K): [in this window] [in a new window]   Figure 6.   Mitogenic responses of fibroblast subpopulations to IL-1. Thy-1 and Thy-1+ fibroblasts were cultured in the presence of rmIL-1 as described in MATERIALS AND METHODS, in the presence of [3H]thymidine. Mean values for each condition were adjusted by subtracting the mean value of control monolayers cultured in SFM without added mediators. The maximum rate of proliferation obtained in cells cultured in 10% FBS without added mediators was set at 100%. Mean values (± SD) for each condition were calculated from four separate wells for each mediator concentration.

The results of Northern blotting of total RNA from IL-1-stimulated Thy-1 and Thy-1+ fibroblasts indicates that IL-1 stimulates IL-6 expression in both subpopulations, but the level of induction is significantly higher in Thy-1 cells (33.4 ± 0.7-fold versus 15.7 ± 1.5-fold induction in Thy-1+ fibroblasts, not shown). Although the level of induction varied slightly between experiments, the difference between subpopulations was maintained (Thy-1 IL-6 expression = 2.30 ± 0.59 higher than Thy-1+, P = 0.008).

There Are No Significant Differences in Expression of IL-1 or PDGF-AA in Thy-1 and Thy-1+ Fibroblasts

To more clearly define autocrine proliferative signaling through the IL-1/PDGFR pathway, we performed ELISA for IL-1 or PDGF-AA in Thy-1-sorted lung fibroblasts. Cells were stimulated in duplicate and three separate stimulations were performed. Neither Thy-1 nor Thy-1+ fibroblasts had measurable levels of cell-associated IL-1, but both increased to 600-700 pg/100-mm dish on stimulation with 125 pg/ml IL-1 (data not shown). Although proliferation is also stimulated in Thy-1 fibroblasts at the same concentrations, at 24 h the accumulated IL-1 in the dishes is likely reflective of the cells originally plated. Levels of PDGF-AA secreted protein in both Thy-1 subpopulations were 8-9 ng/ml in CM and did not change significantly following 24 h stimulation with IL-1 or TNF- (data not shown).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Our findings indicate that the expression of both PDGFR and IL-6 and proliferation are differentially regulated by IL-1 in fibroblast Thy-1 subpopulations. Both subsets are equally capable of binding IL-1, and both have type I IL-1 receptors and IL-1 receptor-associated protein. Increased binding of nuclear proteins to consensus sequences for both NF-B and C/EBP cis-acting promoter regions indicates that intracellular signaling in response to IL-1 occurs in both subpopulations. However, the differing expression of downstream targets indicates that other parallel intracellular signaling cascades are differentially activated by IL-1 in Thy-1 fibroblasts.

IL-1 is a potent, multifunctional proinflammatory cytokine, with complex receptor and signaling pathways that result in control of signaling at multiple levels. A group of fibroblasts which responds to IL-1 by upregulating other autocrine growth pathways such as those regulated by PDGF-AA could mediate rapid fibroproliferation following a multitude of inflammatory events in which IL-1 is involved. The findings reported here imply that there are distinct roles for fibroblast subpopulations in lung fibrosis. However, the same fibroblast-cytokine pathways are central to other remodeling processes, such as joint remodeling in rheumatoid arthritis, and skin scarring in keloid formation (23, 24). The finding that Thy-1 fibroblasts have a greater mitogenic response to IL-1, and therefore may be predisposed to "pro-fibrotic" responses, is consistent with previous work from this laboratory demonstrating in Thy-1 fibroblasts greater expression of and response to connective tissue growth factor (CTGF), a TGF--regulated peptide growth factor central to many fibrotic processes (25, 26). Thy-1 fibroblasts may become activated by IL-1 during an acute inflammatory event, leading to induction of autocrine mitogenic pathways and therefore to dysregulated proliferation. Elucidating the mechanisms by which such differential responses are controlled is essential to understanding the biologic interface between inflammatory signaling and fibrotic remodeling events.

There is a clear difference in PDGFR expression at baseline (Figure 1). Our prior studies indicate that this difference is specific to the  isoform of PDGFR, and occurs at the level of mRNA expression (5). The moderate increase in R expression in Thy-1 cells by IL-1 (20% increase over SFM, P = 0.0004) likely represents the fact that baseline expression is already markedly different from that of Thy-1+ fibroblasts. Inhibition of IL-1 binding by IL-1RA results in a significant reduction of R expression (~ 20%, P = 0.02) in the Thy-1 subset, confirming the presence of autocrine signaling. The incompleteness of the reduction reflects either the lower affinity of IL-1RA versus IL-1 or  for IL-1 receptors (9), or the presence of other non-IL-1 stimuli regulating PDGFR.

There are several possible mechanisms for the differential regulation of PDGFR by IL-1. One of the central features of IL-1 signaling is the exquisite sensitivity of the IL-1RtI even with very low receptor numbers (9). Our finding that both Thy-1 and Thy-1+ fibroblasts have detectable levels of IL-1RtI indicates that the initial binding and signaling events likely occur in both subpopulations. This finding, however, does not preclude the possibility that in an inflammatory milieu IL-1RtI may be differentially regulated in the two subsets. Others have demonstrated, for example, that TGF- decreases expression of IL-1RtI preferentially in Thy-1+ fibroblasts, making them less responsive to IL-1-mediated induction of IL-6 (27). We considered the possibility that Thy-1+ (nonresponsive) fibroblasts may have higher expression of IL-1RtII, the nonsignaling or "decoy" receptor, which could act as a "sink" for IL-1. However, we were unable to detect type II receptors by immunoblotting of lysates or conditioned media or by flow cytometry, consistent with prior reports that the type II receptor is not expressed in fibroblasts (28). Another possibility is that there is a higher level of IL-1 receptor antagonism in Thy-1+ subpopulations. However, equivalent levels of IL-1RA secretion by murine Thy-1 subpopulations have been previously demonstrated (29). IL-1 receptor-associated protein (IL-1RacP) is essential for activation of the IL-1 receptor-associated kinase (IRAK) and stress-activated protein kinase pathways (8). Our findings indicate no differences in expression of the IL-RacP in Thy-1 and + subpopulations. Thus all of the necessary proximal signaling components of the IL-1 pathway are present in both types of cells.

Surprising, therefore, given the differing responses seen in the two subpopulations, is the finding that the induction of distal signaling components, such as transcription factors, seem to be equivalent as well. Most of the inflammatory genes induced by IL-1 are regulated by NF-B (10). That IL-1 induces NF-B binding to its consensus DNA sequence in both subpopulations is, on face value, contradictory to differential intracellular signaling in Thy-1 subsets. It has previously been clearly demonstrated in unsorted rat lung fibroblasts that IL-1 increases PDGFR gene expression and response of fibroblasts to PDGF-AA in part through NF-B activation (6). IL-1 induces increased NF-B transcription factor binding, via p50 and p65 subunits, and inhibition of induced NF-B binding with specific inhibitors suppresses PDGFR upregulation. However, other stimuli for NF-B activation and promoter binding, such as TNF-, fail to upregulate PDGFR, indicating that IL-1-induced NF-B binding is necessary, but not sufficient, to increase R expression (30). A parallel IL-1- mediated signaling pathway is required. It has recently been demonstrated that IL-1-induced p38 MAP kinase stabilizes PDGFR mRNA (22). Our findings indicate activation of p38 in both Thy-1 and Thy-1+ fibroblasts (Figure 5). Others have demonstrated in other cell types that it is possible to inhibit IL-1-induced NF-B-dependent transcription without inhibiting DNA binding (31, 32). Ongoing studies in our laboratory address this level of regulation.

Uncontrolled fibroproliferative responses to injury and inflammation represent a major clinical challenge in the lung and other organs. One of the characteristics shared by many fibrotic disorders is the continuation of fibroblast activation and proliferation after the initial injurious or inflammatory phase has subsided. The persistence of this "young connective tissue" has very recently been shown to be a major predictor of clinical outcome in idiopathic pulmonary fibrosis (33). Initiation of autocrine pathways by early inflammatory mediators, such as IL-1, may provide in part an explanation for this persistence, and the failure of fibrotic lesions to resolve. Although both fibroblast subsets make PDGF-AA and IL-1, differential responsiveness to these factors, mediated by differential PDGFR expression and IL-1-induced signaling, respectively, likely result in increased autocrine proliferative signaling in Thy-1 fibroblasts. Thus Thy-1 fibroblasts may be predisposed to mediate dysregulated fibroproliferation following inflammatory stimuli. Further study of mechanisms underlying differential signaling in fibroblast subpopulations, such as Thy-1 subsets, may yield novel biological insights into the fibroproliferation which is characteristic of fibrosis.