Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Idiopathic pulmonary fibrosis (IPF)/usual interstitial pneumonia (UIP) is a progressive and usually lethal lung disorder characterized by fibroblast proliferation and abnormal accumulation of extracellular matrix (ECM) molecules, especially fibrillar collagens (1). An important feature in IPF/UIP is the presence of fibroblast foci, widely dispersed throughout the lung parenchyma (1). Fibroblast foci seem to represent microscopic zones of acute lung injury where fibroblasts migrate, proliferate, and contribute to the accumulation of ECM molecules in the damaged alveolus. Subsequently, abnormal remodeling of lung architecture results from interstitial and intraluminal deposition of connective tissue (2).

However, studies dealing with the proliferative characteristics and collagen synthesis capacity in fibroblasts obtained from normal human lungs and IPF lungs are scant and results have been inconclusive (3). Thus, for example, Jordana and colleagues (3) showed that human lung fibroblasts derived from fibrotic tissues proliferate significantly faster than do those obtained from normal lungs. On the other hand, Raghu and associates (4) found that fibroblast proliferation was highest in cells derived from areas of early fibrosis compared with normal areas, whereas the lowest proliferation was observed in cells obtained from dense fibrosis.

Additionally, fibroblasts are not a homogeneous population, and in the last decade the concept of phenotypic diversity has emerged. Thus, phenotypically distinct populations of lung fibroblasts that differ in surface markers, receptor expression, cytoskeletal arrangement, and cytokine profile have been described (6, 7). In this context we have recently demonstrated significant differences in the growth rate of normal human lung fibroblast subpopulations separated by size, with an inverse relationship between cell size and growth rate (8).

Myofibroblasts are a unique subpopulation of fibroblasts that express features of smooth-muscle differentiation. -Smooth muscle actin (-SMA)-positive fibroblasts increase in IPF lungs and constitute the major component of the fibroblasts' foci (1, 2). In addition, these cells acquire an aggressive phenotype and seem to be the main subset responsible for collagen accumulation (9).

On the other hand, the abnormal ECM remodeling observed in IPF lungs is at least partially due to an imbalance between some components of the matrix metalloproteinase (MMP) family, i.e., collagenase-1 (MMP-1), gelatinases A and B (MMP-2 and -9, respectively), and the tissue inhibitors of metalloproteinase (TIMPs) (10). Fibroblasts are good candidates to play an important role in this aberrant remodeling process because they are responsible for the exaggerated secretion of most of the ECM components; however, their participation through MMP and/or TIMP expression is largely unknown.

In this study we examined the following in lung fibroblasts derived from IPF patients and control human lungs: growth rate; apoptotic index; percent of myofibroblasts; and 1-(I) collagen, MMP-1, MMP-2, MMP-9, TIMP-1, TIMP-2, TIMP-3, TIMP-4, and transforming growth factor (TGF)- gene expression.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

Materials for cell isolation and culture were from GIBCO BRL (Rockville, MD). Trypsin, propidium iodine (PI), and fluorescein isothiocianate (FITC)-conjugated monoclonal antibody (mAb) to -SMA (clone B4A) were obtained from Sigma Chemical Co. (St. Louis, MO). Avidin-FITC, FITC-conjugated antimouse immunoglobulin G, deoxyribonuclease (DNAse)-free ribonuclease (RNAse) (cytometry grade), and proliferation reagent WST-1 were from Boehringer Mannheim (Indianapolis, IN). All other chemicals were of reagent grade.

Fibroblast Isolation and Culture

Primary human lung fibroblasts were obtained in our laboratory as previously described (13). Briefly, fibroblasts were derived from lung tissue obtained from IPF patients (n = 5) by open lung biopsy. IPF was diagnosed as described elsewhere, including the characteristic morphology of UIP (1). Patients (three male and two female; four nonsmokers and one ex-smoker) aged 57 ± 5 yr (mean ± standard deviation [SD]) were inpatients at Instituto Nacional de Enfermedades Respiratorias, Mexico DF, Mexico. Time elapsed to first visit after the beginning of symptoms, usually progressive dyspnea, was 22 ± 6 mo. All patients showed a restrictive functional pattern (forced vital capacity: 57 ± 10% of predicted; total lung capacity: 62 ± 9% of predicted) and hypoxemia at rest (53 ± 7 mmHg; normal values for Mexico City: 59 ± 8 mmHg), worsening with exercise. Lung samples were taken by open lung biopsy usually 1 wk after hospital admission. None of the patients had been treated with corticosteroids or immunosuppressive drugs at the time of biopsy. Control fibroblasts (n = 5) were derived from lungs of patients undergoing lobectomy/pneumonectomy for removal of a primary lung tumor which showed no histologic evidence of disease. Lung fibroblasts were isolated by trypsin dispersion, and cells were grown in Ham's F-12 medium (GIBCO Laboratories, Grand Island, NY) supplemented with 10% fetal bovine serum (FBS) (GIBCO BRL). Fibroblasts (passages 5-8) were cultured at 37°C in 5% CO₂/95% air in T-25 cm² Falcon flasks, using Ham's F-12 medium supplemented with 10% FBS, 100 U/ml penicillin, and 100 g/ml streptomycin.

Growth Rate Assay

Fibroblasts were plated in 96-well culture plates at a cell density of 15 × 10³ cells/well and incubated at 37°C in 5% CO₂/95% air in Ham's F-12 medium supplemented with 10% FBS. After 24 h, the medium was replaced using 1% or 10% FBS and the growth rate was followed for 13 d.

Cell number was determined using the cell proliferation reagent WST-1 (Boehringer Mannheim) as previously described (8). WST-1 is a tetrazolium salt cleaved by the mitochondria of viable cells to yield a soluble formazan chromophore. The medium in corresponding wells was replaced with fresh medium, and the dye solution was added to each well according to manufacturer's instructions. Absorbance was measured on an enzyme-linked immunosorbent assay (ELISA) plate reader at a wavelength of 450 nm using a reference wavelength of 620 nm. Absorbance changes were taken as percentage of growth rate increase related to basal values (Day 0).

Flow Cytometry Analysis for -SMA and Apoptosis

Flow cytometry analysis was performed on a Partec CA-III flow cytometer equipped with a 25-mW argon ion laser for excitation at 488 nm. PI fluorescence was acquired through a 610-nm long-pass filter, and fluorescent FITC through an EM 520 band-pass filter. After standardization with fluorescence microspheres (Coulter, Hialeah, FL), amplifier gains were not changed throughout an experiment.

Fibroblasts for cytometry analyses were grown at mid-log phase in culture media supplemented with 10% FBS. Cells were recovered by trypsyn/ethylenediaminetetraacetic acid (EDTA) digestion, centrifuged, and fixed in ice-cold 70% ethanol while shaking for 5 min, and stored at 20°C until assayed.

For apoptosis analysis, a bivariate assay of in situ end labeling (ISEL) versus log forward angle light scatter (FALS, an index of cell size [8]) was performed. Cells were labeled with biotinylated deoxyuridine triphosphate (dUTP) and detected with avidin-FITC as described by Gorczyca and coworkers (14). Briefly, cells were washed with phosphate-buffered saline (PBS) for 10 min at 4°C, and incubated with saline sodium citrate buffer containing 1% bovine serum albumin (BSA) for 45 min at 37°C. After washing with PBS, cells were incubated with 100 µl ISEL buffer containing Bio-dUTP for 2 h at 17°C. Finally, fibroblasts were washed again with PBS and incubated with 100 µl avidin-FITC/ PBS for 45 min at 37°C.

To evaluate -SMA expression, fibroblasts were incubated for 1 h at 37°C with FITC-conjugated monoclonal antihuman -SMA antibody diluted 1:400 in 1% BSA in PBS, pH 7.3, with DNAse-free RNAse (25 µg/ml). DNA distribution experiments were conducted with PI as described earlier (15).

Microscopy and Image Analysis of -SMA

Fibroblasts were grown at mid-log phase in culture media supplemented with 10% fetal calf serum (FCS) and were labeled with antihuman -SMA (clone BA4) antibody in a way similar to that performed for flow cytometry analysis. Photomicrography was accomplished with an Olympus EMT-2 epifluorescence phase-contrast microscope equipped with band-pass filters for detection of FITC and fitted with color and gray-scale charge-coupled device cameras. Cell diameters of -SMA-positive and -negative cells were determined by automated measurement of ferret diameters of 50 cells/group with the image-analysis program MOCHA (Jandel Scientific, San Rafael, CA).

Fluorescence Microscopy for Apoptosis

Detection of apoptotic cells with PI was conducted as described earlier after digestion of ethanol-fixed cells with DNAse-free RNAse in PBS containing 5 µg/ml PI (8, 15). In these assays detached cells were retained by centrifugation of the 24-well culture vessels during fixation with 70% ethanol for at least 30 min at 4°C; red fluorescence intensity (> 590 nm) of chromatin-bound PI is proportional to DNA content, allowing visualization of chromatin condensation. For quantitation, fragmented nuclei were scored as a percentage of total nuclei in a minimum of four microscopic fields per culture vessel; at least four culture vessels per fibroblast strain were scored. Data from three normal and three fibrotic strains were compiled and are reported as means ± SD.

Reverse Transcription/Polymerase Chain Reaction

Total RNA from lung fibroblasts was isolated according the acid guanidium isothiocianate/phenol chloroform extraction method. Complementary DNA (cDNA) from IPF (n = 3) and control (n = 3) fibroblasts was synthesized by reverse transcription of 5 µl total RNA, using a cDNA synthesis kit (GIBCO BRL).

Polymerase chain reaction (PCR) amplification (Perkin-Elmer 9600, Perkin-Elmer, Branchburg, NJ) was performed with a cDNA working mixture in a 25-µl reaction volume containing 20 mM Tris-HCl (pH 8.3), 50 mM KCl, 2 mM MgCl₂, 200 µM deoxynucleotide triphosphates, 1 µM specific 5' and 3' primers and 1 U Taq DNA polymerase (Perkin-Elmer).

To quantify the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH), a competitor was constructed by cutting with NcoI an internal fragment of 155 base pairs (bp) of a GAPDH cDNA of originally 1,233 bp cloned in a PBR 322 plasmid. The modified GAPDH cDNA was subsequently religated. The competitor (c) GAPDH sequence was obtained by PCR amplification of the modified plasmid using the primers for GAPDH. The competitor size was 240 bp.

Four-fold serial dilutions of the standard competitor (5, 10, 15, and 20 pg) were coamplified with a constant amount of cellular cDNA (1 µl). Cycling conditions were: 95°C for 10 min for one cycle; 95°C for 30 s, 58°C for 30 s, and 72°C for 120 s for 40 cycles; and the final incubation at 72°C for 7 min. Aliquots (5 µl) of the PCR product were resolved in 1.5% agarose gel containing ethidium bromide. Band intensities were quantitated by scanning densitometry using a Kodak digital science electrophoresis documentation and analysis system 120 (Kodak, Rochester, NY). The logarithm of GAPDH/cGAPDH ratio was plotted as a function of the logarithm of known cGAPDH amount. The point of equivalence represents the concentration of GAPDH in the cDNA sample. Dilutions were performed to reach 50 to 750 fentograms of GAPDH for each amplification as specified in Table 1.

cDNA was amplified with specific primers (16-21, Table 1) by PCR. Amplification conditions for all reactions were done at: 94°C for 5 min for one cycle; 94°C for 30 s, 55 to 60°C for 30 s, and 72° C for 30 s for a number of cycles (Table 1); and the final incubation at 72°C for 5 min. Aliquots (5 µl) of the PCR products were resolved in 1.5% agarose gel containing ethidium bromide. Band intensities were quantitated by scanning densitometry.

Collagen Synthesis Assay

Collagen synthesis was evaluated according to Peterkofsky and Diegelmann (22). Fibroblasts (2.5 × 10⁵ per well) were cultured in six-well tissue culture multiwell plates in Ham's F-12 medium supplemented with 10% FBS. After 24 h, the medium was replaced by fresh serum-free medium containing 15 µCi/ml [³H]proline (New England Nuclear, Boston, MA), 50 µg/ml ascorbic acid, and 50 µg/ml -aminopropionitrile. After 6 h labeling, supernatants were collected and dialyzed against distilled water containing protease inhibitors (25 mM N-ethylmaleimide, 0.1 mM phenylmethylsulfonyl fluoride, and 10 mM EDTA) and 0.02% sodium azide. Samples were lyophilized and resuspended in 50 mM Tris-HCl and 10 mM CaCl₂ (pH 7.6). Aliquots were incubated with or without 36 µg/ml bacterial collagenase (Sigma type VII) for 3 h at 37°C and precipitated with cold 10% trichloroacetic acid and 0.5% tanic acid. Collagen synthesis was evaluated in the supernatants and noncollagenous proteins in the precipitates after HCl hydrolysis. Samples were counted in a scintillation counter (Beckman LS6000 SE). Collagen synthesis was calculated as: % collagen synthesis = disintegrations per min (dpm) collagen × 100/(dpm noncollagen protein × 5.4) + dpm collagen.

TGF-1 Quantification

Measurement of TGF-1 protein in fibroblast conditioned media was performed by using a commercial sensitive and specific ELISA, following the manufacturer's instructions (R&D Systems, Minneapolis, MN). TGF-1 was expressed as picograms per microgram of protein.

Statistical Analysis

Results are expressed as means ± SD. Comparisons were made using a Student's t test for unpaired observations. Values of P < 0.05 were considered statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Growth Rate Analysis

To determine the growth rate of control and IPF-derived fibroblasts, cell number was determined up to 13 d, using the cell proliferation reagent WST-1. As shown in Figure 1A, fibroblasts derived from IPF patients exhibited a significantly lower growth rate when compared with control cells from the fourth day until the end of the experiment (IPF: 13.3 ± 38.5% versus 294.6 ± 57% in controls at 13 d; P < 0.0001). Similar results were obtained when cells were incubated in media containing 10% FCS (Figure 1B).

View larger version (20K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   Figure 1.   Growth rate analysis from primary culture of human lung fibroblasts: normal human lung fibroblasts (control strains, circles); fibroblasts derived from IPF strains (squares). Cells were grown in Ham's nutrient medium supplemented with 1% FCS for 13 d. (A) Fibroblasts incubated in media containing 1% FCS. (B) Fibroblasts incubated in media containing 10% FCS. At the indicated times, cells number was estimated by WST-1 assay (see MATERIALS AND METHODS). Each point represents the mean ± SD of five different primary strains of fibroblasts, each one assayed in quadruplicate; *P < 0.05 at 4 d; **P < 0.01 at 8 and 11 d; ***P < 0.0001 at 13 d (analysis of variance-Bonferroni analysis).

Flow Cytometry Analysis of Apoptosis and -SMA Expression

ISEL of fragmented DNA was used to evaluate the percent of fibroblasts displaying spontaneous apoptosis. Cells cultured in medium supplemented with 10% FCS at mid- log phase were harvested and subjected to cytofluorometric analysis of ISEL in relation to cell size (Figure 2). As shown in Table 2, fibroblasts derived from IPF lungs displayed a significant increase in the percent of apoptotic cells when compared with controls (IPF versus controls: 31.4 ± 6.9% versus 15.0 ± 7.4%; P < 0.008). In addition, the presence of apoptotic cells was analyzed by PI staining (Figures 3A and 3B). As shown in Figure 3C, a significant increase of apoptosis was observed in IPF fibroblasts (430 ± 49% versus 100 ± 16% in controls; P < 0.01).

View larger version (27K): [in this window] [in a new window]   Figure 2.   Bivariate flow cytometric analysis of ISEL for DNA fragments, versus log FALS (Size) of primary human lung fibroblasts. Cells were grown at mid-log phase in media supplemented with 10% FCS, harvested, and fixed in ice-cold 70% ethanol while shaking. Assay for ISEL is described in MATERIALS AND METHODS. (A) Human fibroblast cell line derived from IPF patient; (B) human control cell line.

View larger version (155K): [in this window] [in a new window]   View larger version (32K): [in this window] [in a new window]   Figure 3.   Detection of apoptotic fibroblasts with PI under fluorescence microscopy. Subconfluent cultures of fibroblasts were fixed with 70% ethanol in a manner that retained detached cells (see MATERIALS AND METHODS) and were subsequently exposed to 5 µg/ml PI in the presence of DNAse-free RNAse. (A) Normal nuclei in a control fibroblast strain. (B) Normal and fragmented nuclei (arrowheads) containing condensed chromatin, indicative of apoptosis in an IPF fibroblast strain. (C) Quantitation of apoptotic fibroblasts by the PI assay. Fragmented nuclei were scored as a percentage of total nuclei as described in MATERIALS AND METHODS. Data are means ± SD from three normal and three fibrotic strains compiled. *P < 0.01.

The proportion of -SMA immunoreactive cells was analyzed by immunocytochemistry followed by microscopy image analysis and cytometry analysis in relation to DNA content, as shown in Figure 4. As seen in Table 2, several strains of lung fibroblasts derived from IPF patients showed a significant increase in the percent of -SMA- positive cells when compared with control lung fibroblasts (IPF versus controls: 62.8 ± 25.2 versus 14.8 ± 11.6%; P < 0.01).

View larger version (39K): [in this window] [in a new window]   View larger version (33K): [in this window] [in a new window]   Figure 4.   Relationship of -SMA immunoreactivity to DNA content in human primary lung fibroblasts. Cells were cultured to mid-log phase with 10% FCS, washed, harvested, and fixed in ice-cold 70% ethanol while stirring for flow cytometry detection. Fibroblasts were incubated with FITC-conjugated mAbs to -SMA (anti--SMA; see MATERIALS AND METHODS) and subjected to cytometry analysis versus DNA content by means of PI stain. (A') A human fibroblast cell line derived from an IPF patient. (B') A normal human lung fibroblast. Left: immunofluorescent analysis of the same cell lines shown on the right. (A) Human fibroblast cell line derived from an IPF patient. (B) Control fibroblasts. (C) Negative control omitting the antihuman -SMA antibody.

Additionally, the cell diameters of -SMA-positive and -negative cells were determined by automated measurement of ferret diameters of 50 cells/group with the image-analysis program MOCHA. Results showed that in both IPF and control fibroblast strains, cells immunoreactive to -SMA exhibited a significantly higher diameter than did negative cells (IPF: 38.1 ± 8.2 µm versus 26.9 ± 7.4 µm, P < 0.05; controls: 35.8 ± 7.0 µm versus 25.7 ± 5.8 µm, P < 0.05). This result was consistent with immunocytochemical observation where -SMA immunoreactive cells appeared larger than negative cells (Figures 4A and 4B).

IPF and Control Fibroblast Gene Expression

Semiquantitative assessment of messenger RNA (mRNA) of selected molecules expressed by IPF and control fibroblasts was done by serial dilutions of cDNA samples adjusted to equal quantified housekeeping gene GAPDH. Densities of bands of PCR products achieved with these primers were analyzed by agarose gel electrophoresis (Figure 5). GAPDH band intensities were quantitated by scanning densitometry (controls: 5,314 ± 96; IPF fibroblasts: 5,377 ± 85), demonstrating that the DNA used from different cell lines was similar.

View larger version (42K): [in this window] [in a new window]   View larger version (49K): [in this window] [in a new window]   Figure 5.   Expression of mRNA by three control and three IPF fibroblasts using a semiquantitative RT-PCR methodology (see MATERIALS AND METHODS). Total fibroblast RNA was extracted and subjected to RT-PCR amplification using specific primers (Table 1). PCR products were electrophoresed and visualized by ethidium bromide staining. (A) 1-(I) collagen, TGF-1, collagenase-1, gelatinases A and B, and the housekeeping gene GAPDH. (B) TIMP-1, -2, -3, and -4, and GAPDH.

IPF-derived fibroblasts consistently revealed a higher expression of 1-(I) collagen gene as compared with control fibroblasts (Figure 5A). The expression increased almost 2-fold when quantitated by scanning densitometry (9,418 ± 1,217 versus 5,173 ± 1,559; P < 0.02). When these cell lines were analyzed for collagen synthesis a similar result was observed (53.1 ± 5.2% versus 28.6 ± 3.7%; P < 0.01). Concerning TGF-1 expression (Figure 5A), mRNA transcripts were usually more abundant in IPF than in control fibroblasts (12,881 ± 2,033 versus 6,902 ± 2,782; P < 0.05); however, at the protein level, results were heterogeneous and no differences were observed (96 ± 23 versus 120 ± 34 pg/mg protein).

Collagenase-1 (MMP-1) was expressed by one out of three control cell strains and equally by one from three IPF-derived fibroblast cell strains with no difference in the net intensity of the band in the collagenase-expressing cell lines. No significant differences were found among fibroblasts expressing gelatinase A (MMP-2) (3,922 ± 664 in normal versus 6,037 ± 1,641 in IPF fibroblasts). By contrast, gelatinase B (MMP-9) was expressed in two out of three IPF-derived strains (30,414 ± 1,416) and faintly (3,615) by one out of three control cell lines (Figure 5A).

Results of reverse transcriptase (RT-PCR) for all four TIMPs are illustrated in Figure 5B. TIMP-1 revealed a 2-fold increase in IPF compared with control fibroblasts (75,195 ± 10,119 versus 38,479 ± 13,846; P < 0.02). TIMP-2 was also highly expressed in IPF fibroblasts showing a 2-fold increase in net intensity in IPF as compared with controls (20,935 ± 6,451 versus 9,449 ± 4,322; P = 0.05). Likewise, TIMP-3 and TIMP-4 were mainly expressed in IPF-derived fibroblasts and faintly detectable in one control cell strain.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

IPF is a progressive and usually fatal lung disorder characterized by the occurrence of widely scattered small clusters of fibroblasts, presumably provoked by multiple microscopic lung injuries, and abnormal tissue remodeling (1).

A growing body of evidence has demonstrated that there is phenotypic and behavioral heterogeneity among tissue fibroblasts. Myofibroblasts, which express features of smooth-muscle differentiation, constitute one of these subpopulations and are usually increased in human and experimental lung fibrosis (2, 9).

In the present study, fibroblasts derived from IPF lungs displayed a marked increase in the percentage of myofibroblasts, paralleling the in vivo observations. As examined by fluorescence-activated cell sorter and immunocytochemistry, most of the fibroblasts derived from IPF lungs expressed -SMA. The cell origin of lung myofibroblasts remains unknown, but interstitial fibroblasts and smooth muscle-like cells of the alveolar duct septal tips are considered as a potential source (23). Increased myofibroblasts in the multiple fibroblast foci of IPF lungs may have a number of deleterious consequences, primarily contributing to active parenchymal contraction, decreased compliance, and distorted architecture (2). Actually, myofibroblasts play a critical role in healing because they express high levels of -SMA, form tight adhesions to the substrate, and arrange the stress fibers along the long axis, facilitating tissue contraction and reducing the amount of denuded surface area of wounded tissue (24). In addition, myofibroblasts produce apoptotic factor(s) for alveolar epithelial cells in vitro (25), and have been localized close to denuded alveolar epithelium in vivo (26). Moreover, it has been suggested that myofibroblasts may be responsible for most of the increased lung collagen gene expression in pulmonary fibrosis (9). Nevertheless, this last observation had not been documented in vitro. Raghu and colleagues (5) found that fibroblasts from normal and fibrotic lungs synthesize similar amounts of collagen even when stimulated with TGF-. In our study, however, fibroblasts derived from IPF lungs, usually presenting high proportions of -SMA-positive cells, expressed higher levels of type I collagen mRNA and synthesized higher amounts of collagen compared with normal lung fibroblasts.

On the other hand, IPF fibroblasts showed a marked decrease of growth rate when compared with lung fibroblasts derived from normal subjects. The cell kinetic assay showed a marked difference from the fourth day on. This finding differs from those previously reported. Thus, Jordana and associates (3) showed that primary fibroblasts derived from IPF patients proliferated faster than did normal lung fibroblasts; whereas Raghu and coworkers (4), who obtained fibroblasts from different areas of the same fibrotic lung, found a higher rate of proliferation in cells derived from inflammatory areas and the lowest in fibroblasts from dense fibrotic scars. Given the new histologic classification of IPF (1), it is not certain whether IPF fibroblasts in those studies were obtained from UIP cases.

Our results substantiate unpublished observations found repeatedly over many years in our laboratory. Usually, fibroblast strains derived from IPF lungs are very difficult to grow in vitro and exhibit a limited number of duplications. The reduced growth rate of the IPF fibroblasts could have several possible explanations. Thus, for example, it has been shown that rat lung myofibroblast differentiation in vitro results in reduced telomerase activity (27). The decreased expression of this enzyme may account for a reduced proliferative capacity. Likewise, we have previously demonstrated that myofibroblasts, when separated by gravity sedimentation, are usually large cells that grow markedly slower than intermediate or small size fibroblasts (8). Coincidentally, in this study we also observed that the diameters of the -SMA-positive cells, independently if derived from normal or IPF lungs, were significantly larger than those of -SMA-negative cells.

In addition, slower growth rate may be also explained by an increase in programmed cell death. Our findings support this notion because IPF fibroblasts exhibited a higher basal rate of spontaneous apoptosis than did normal cells. In studies of primary fibroblasts isolated from normal human lung, Akamine and colleagues (6) identified two subpopulations which were isolated on the basis of differential expression of the receptor for C1q, one of the first subcomponents of complement. The fibroblast subsets separated by C1q receptor expression displayed two distinct morphologies: spindle-shaped cells with elongated processes (high binding), or larger and more flattened cells (6). The subgroups also differed in rates of proliferation and especially collagen synthesis, both under basal conditions and after inhibition by interferon- or stimulation by TGF-. More interesting in the context of our findings, the low C1q-binding subset also displayed a poor rate of growth in culture despite containing a higher percentage of cycling cells identified by flow cytometry (6).

This paradox might be explained by a higher rate of spontaneous apoptosis within this subpopulation, which exhibits the same morphologic characteristics (large and flat) as the "large" subgroup identified in our earlier work (8).

Regardless, in this report we showed that IPF fibroblast isolates contain an increased subpopulation of myofibroblasts and exhibit decreased proliferative capacity and an increased rate of spontaneous apoptosis. A number of cytokines might participate in the induction of this apoptosis, including interleukin-1, which has been reported to induce apoptosis and inhibit cell proliferation (28). In addition, the increased basal apoptosis might be related to the production by myofibroblasts of angiotensin peptides (29), which induce apoptosis in a variety of cell types, including alveolar epithelial cells of the lung (29, 30). With regard to the molecules analyzed in this study, although primarily related to ECM remodeling, some of them may also contribute to increased apoptosis. Particularly, overexpression of TIMP-3 induces programmed cell death in a variety of cell types (31). Inhibition of tumor necrosis factor--converting enzyme may account at least partially for its ability to induce apoptosis (31).

The determination of which of these factors is involved in the induction of fibroblast apoptosis, and which phenotypes within the lung fibroblast isolate undergo apoptosis, will be an interesting topic for future experiments.

The final stage of IPF is an extensive structural disorganization of the lung parenchyma with the subsequent progressive loss of the alveolar-capillary units. The molecular mechanisms behind this erratic tissue remodeling that results in an exaggerated ECM accumulation have not been elucidated but likely involve a loss in the highly regulated balance between MMPs and TIMPs (10). In this context there were no differences in collagenase-1 mRNA expression, which corroborates previous findings with MMP-1 protein levels in IPF and control fibroblasts (32). With regard to gelatinases, MMP-2 has been shown to be constitutively expressed in lung fibroblasts and we did not find differences in IPF and normal lung-derived cells. However, normal lung fibroblasts usually do not express MMP-9 in vitro and thus our recent findingcorroborated in this studythat fibroblasts obtained from IPF lungs strongly expressed the MMP-9 transcript was quite interesting (12). Further, the expression of gelatinase B was closely related with the percent of myofibroblasts (12). Gelatinase B expression may have a role in basement membrane disruption and fibroblast migration into the alveolar spaces.

The gene expression of all four TIMPs was usually increased in fibroblasts derived from IPF lungs. Although different TIMPs bind tightly to most MMPs, differences in inhibitory properties, expression patterns, and other propertiessuch as association with latent MMPs, cell growth- promoting activity, cell survival-promoting activity, and apoptosishave been reported (31, 33). In this context, upregulation of TIMP gene expression could contribute to a prevailing lung fibrotic microenvironment by inhibiting MMP activity and matrix degradation; supporting this point of view, we recently found in vivo that in IPF lungs there is higher interstitial expression of TIMPs compared with collagenases (12). On the other hand, because TIMPs may play a role in regulating apoptosis or cell survival, the increase of TIMP expression may also contribute to the pathogenic processes observed in IPF lungs by either increasing proliferation of some cell populations or inducing the death of others. In this context, it is interesting to emphasize that TIMP-2 is almost exclusively expressed by fibroblasts/myofibroblasts in the characteristic fibroblast foci considered to be the sites of ongoing alveolar epithelial injury and activation associated with evolving fibrosis (11, 12). Fibroblast TIMP-2 expression could be associated with collagenase inhibitory effect, but also with activation of latent gelatinase A and stimulation of fibroblast proliferation. TIMP-3 is singular because it binds strongly to insoluble extracellular components, mainly heparan sulfate. It has been observed to have antiadhesive properties in fibroblasts, and TIMP-3 overexpression can induce apoptosis (31). TIMP-3 is the only TIMP to inhibit members of the ADAM (a disintegrin and metalloprotease domain) and to restrain shedding of L-selectin (31). Interestingly, it has been reported that the levels of TIMP-3 mRNA were elevated in systemic sclerosis fibroblast cell strains (34). Studies on TIMP-4 are scanty and their functional activities, besides MMP inhibition, are essentially unknown. Recent evidence demonstrates a temporal relationship between upregulation of TIMP-4 and the onset of collagen deposition in injured arterial walls, suggesting an important role in the proteolytic balance of the vasculature (35).

Certainly, other cell types besides fibroblasts/myofibroblasts may play a role in the protease-antiprotease imbalance in IPF. Thus, resident cells (i.e., alveolar epithelial cells and endothelial cells) as well as those derived from the inflammatory response are able to produce different MMPs and TIMPs in vivo. Particularly, MMP-1, MMP-2, MMP-9, and TIMP-2 have been detected in epithelial cells close to fibroblast foci. MMP-1 has also been observed in alveolar macrophages, MMP-9 in neutrophils, TIMP-1 and TIMP-3 in interstitial macrophages, and TIMP-4 in plasma-cells (11). Therefore, during the development of IPF, ECM remodeling is regulated by a complex interplay among epithelial and endothelial cells, fibroblasts, and inflammatory cells, and the ECM in the local mielieu.

In summary, although the number of fibroblast strains explored in this study is small, our in vitro findings correlate with some previous in vivo observations and suggest that fibroblasts/myofibroblasts from IPF lungs exhibit a profibrotic secretory phenotype with lower growth rate and increased spontaneous apoptosis.