Evasion of myofibroblasts from immune surveillance: A mechanism for tissue fibrosis Shulamit B. Wallach-Dayan*†, Regina Golan-Gerstl*, and Raphael Breuer*‡ *Lung Cellular and Molecular Biology Laboratory, Institute of Pulmonary Medicine Hadassah, Hebrew University Medical Center, Jerusalem 91120, Israel; and ‡Department of Pathology, Boston University School of Medicine, Boston, MA 02118 Edited by Michael Sela, Weizmann Institute of Science, Rehovot, Israel, and approved October 19, 2007 (received for review June 27, 2007)

Tissue fibrosis evolving from impaired tissue remodeling after injury is characterized by myofibroblast accumulation. We propose that during the development of fibrosis myofibroblasts acquire an immune-privileged cell phenotype, allowing their uninterrupted accumulation. Using the murine model of bleomycin-induced lung fibrosis in mice, we show that myofibroblasts that accumulate in lungs with fibrosis, but not in normal lungs, kill Fasⴙ lymphocytes, resist Fas-induced apoptosis, and survive longer when grafted into allogeneic mice. In contrast, bleomycin-treated FasLigand (FasL)deficient (gld) chimeric mice did not accumulate myofibroblasts or collagen in their lungs, and their FasLⴚ myofibroblasts did not survive after alloengraftment. This finding indicates that myofibroblasts possess Fas/FasL-pathway-dependent characteristics that allow them to escape from immune surveillance and resulting organ fibrosis. Fas/FasL 兩 immune privilege

U

nder normal conditions, myofibroblasts play an essential role in the resolution of inf lammation and scar formation; however, recovery from injury requires their subsequent disappearance from the injured site (1). Myofibroblast eradication during normal lung repair depends on factors that maintain their survival (2) and on Fas Ligand (FasL)-induced apoptosis (3) primarily by immune T cells (4), maintaining tissue homeostasis (5). The pathogenesis of pulmonary fibrosis is typically characterized by abnormalities of alveolar structure accompanied by myofibroblast accumulation and collagen deposition in the extracellular matrix, with resulting lung scarring and inhibition of gas exchange (2). Intratracheal administration of bleomycin has been widely used as a model to study the mechanisms underlying lung injury, inflammation, and repair (6, 7). In contrast to normal regenerating tissue, resistance to Fasinduced apoptosis was detected in lung myofibroblasts of humans with idiopatic pulmonary fibrosis (IPF) (3), implying that fibrotic lung myofibroblasts may evade immune cell-induced apoptosis, thus allowing for their uninterrupted accumulation. Evasion of immune surveillance and the ability to undergo uninterrupted proliferation are characteristics of cancer cells, which possess features of immune-privileged tissues (8, 9). Their features also include the ability to resist Fas-induced apoptosis (10) after binding of FasL expressed by immune cells, and the expression of a functional FasL, which enables the counterattack of neighboring Fas-expressing immune cells, thus allowing for abnormal cell accumulation. The functional role of immune cells in lung injury (and repair) and fibrosis is multifaceted (1, 11–13). To shed more light on one aspect of this issue, we evaluated the specific interaction between immune T cells and lung myofibroblasts in vitro and in vivo. Our data suggest that myofibroblasts employ a mechanism involving immune-privileged cell phenotypes. This mechanism enables them to avoid clearance by cells maintaining tissue homeostasis, leading to a diversion from normal tissue remodeling and the development of lung fibrosis.

20460 –20465 兩 PNAS 兩 December 18, 2007 兩 vol. 104 兩 no. 51

Results Fibrotic Rather Than Normal Lung Myofibroblasts Overexpress Fas but Are Resistant to Fas- and Immune Cell-Induced Apoptosis. To assess

the sensitivity of lung myofibroblasts to Fas-induced apoptosis, we initially determined their Fas level of expression. When compared with normal myofibroblasts isolated from salinetreated lungs, myofibroblasts isolated from fibrotic lungs of bleomycin-treated C57BL/6 mice overexpress Fas molecules (Fig. 1A). We further tested Fas-induced myofibroblast apoptosis by assessing its response to Jo2 Fas-agonist mAb. Flowcytometry analysis revealed that, in comparison with normal lung myofibroblasts, myofibroblasts isolated from bleomycintreated lungs lost their sensitivity to Fas-induced apoptosis (3% apoptosis with IgG isotype-matched control vs. 3.4% with Jo2 mAb) (Fig. 1B Right). In contrast, the extent of Fas-induced apoptosis from normal saline-treated lung myofibroblasts increased (0% apoptosis with IgG control vs. 22% with Jo2 mAb) (Fig. 1B Left). The resistance to apoptosis exerted by fibrotic lung myofibroblasts was confirmed by parallel experiments where myofibroblasts were cocultured with 5 ⫻ 106 Jurkat T lymphocytes at a ratio of 1:5. Viable myofibroblasts were counted by using trypan blue staining and analyzed by confocal microscopy. Jurkat T cells induced the death of myofibroblasts from saline-, but not bleomycin-treated, mouse lungs (Fig. 1C). Myofibroblasts from Fibrotic Lungs Kill Lymphocytes by FasL/Fas Interaction. We have previously shown the expression of FasL on

myofibroblasts in fibrotic lungs of humans with IPF and mice with bleomycin-induced fibrosis (14). Constitutive expression of FasL, observed in immune-privileged tumor cells, has been shown to induce apoptosis of infiltrating lymphocytes (15, 16). We found that myofibroblasts isolated from fibrotic lungs of bleomycin-treated WT mice express a functional FasL (14) capable of inducing Fas/FasL-dependent apoptosis in unsorted naı¨ve thymocytes (Fig. 2A), activated primary T cells (Fig. 2B), and in the Jurkat T cell line (Fig. 2C). Myofibroblasts (5 ⫻ 105) were cocultured at a 1:2 ratio for 24 h, with 1 ⫻ 106 Fas⫹ or Fas⫺ target lymphocytes extracted from WT or Fas-defective (lpr) mice, respectively. No change in apoptosis levels was detected between Fas⫺ thymocytes in the control (17%) and coculture (14%) experiment (Fig. 2 A Right), whereas apoptosis in Fas⫹ thymocytes increased from 9% in controls to 34% when cocultured with myofibroblasts (Fig. 2 A Left). Similar results were obtained when fibrotic lung myofibroblasts were cocultured with Fas⫹ and Fas⫺ active T cells. The Author contributions: S.B.W.-D. designed research; S.B.W.-D. and R.G.-G. performed research; R.B. contributed new reagents/analytic tools; S.B.W.-D. and R.B. analyzed data; and S.B.W.-D. and R.B. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. †To

whom correspondence should be addressed. E-mail: [email protected].

© 2007 by The National Academy of Sciences of the USA

www.pnas.org兾cgi兾doi兾10.1073兾pnas.0705582104

basal apoptosis of WT-activated lymphocytes increased from 45% to 71% (Fig. 2B Left; Fas⫹). However, Fas⫺ T cells showed no change in cell apoptosis after coculture (Fig. 2B Right; Fas⫺). In addition, Jurkat T cell line apoptosis increased from 2% to 14% when cocultured with fibrotic lung myofibroblasts, with or without IgG (Fig. 2C Left and Center). However, preincubation of this culture with 10 ␮g of anti-FasL MFL3 Ab for 24 h led to comparable levels of apoptosis in Jurkat T cells alone (2%) and in cocultures with MFL3 (1.3%) (Fig. 2C Right and Left, respectively). Lung myofibroblasts from saline-treated mice had no cytotoxic effect on cocultured T cells (data not shown), consistent with their low level of FasL expression (14). Bleomycin-Treated Chimeric gld Mice, Compared with Control WT Mice, Exhibit Similar Inflammatory Responses but Decreased ␣SMAⴙ Cell Accumulation and Collagen Deposition. To address the effect of

myofibroblast FasL deficiency on bleomycin-induced lung fibrosis and exclude the effect of FasL-deficient hematopoietic cells, we created a chimeric gld mouse containing WT lymphoid organ cells (14). After bleomycin lung instillation, myofibroblast accumulation was markedly decreased in the lungs of chimeric gld mice, compared with their WT control counterparts, although CD3⫹ lymphocyte infiltrates were detected in both mice (Fig. 3A). Wallach-Dayan et al.

Fig. 2. Myofibroblasts from fibrotic lungs overexpress a cytotoxic FasL molecule. Shown is Annexin V staining of cocultured Fas⫹ vs. Fas⫺ lymphocytes isolated from WT and lpr mice, respectively. (A) Naı¨ve thymocytes. (B) Activated T cells separated by specific CD4 T cell columns (Biotex Laboratories) from the lymphatic tissues of bleomycin-treated WT and lpr mice, with the addition of anti-CD44 (supernatant from clone IM7.8.1, diluted 1:15) or 5 ␮g/ml anti-CD62L Abs. For FACS analysis, activated lymphocytes were further stained with phycoerythrin-conjugated anti-CD4 mAb. All target lymphocytes from WT, but not from lpr mice showed increased Annexin V staining compared with control levels. Results are representative of three experiments. (C) Jurkat T cells with the addition of antagonist MFL3 anti-FasL mAb or control IgG. The increased apoptosis of Jurkat T cells subjected to Jurkat cell/ myofibroblast coculture abated after the addition of MFL3 anti-FasL mAb. Results are representative of three experiments.

Consistent with these results, there was increased collagen deposition in the lungs of bleomycin-treated control WT mice (FasL⫹) but not in lungs of bleomycin-treated gld (FasL⫺) chimeric mice (Fig. 3B). However, lung inflammation, as assessed by bronchoalveolar lavage (BAL) fluid cellularity, was similar in the lungs of both WT and gld chimeric mice (Fig. 3C). Using confocal microscopy, we determined in lung sections in vivo whether bleomycin-treated lung ␣SMA⫹ cells induce FasLdependent apoptosis of adjacent Fas⫹ lymphocytes (Fig. 3D). To this end, lung sections from bleomycin-treated chimeric gld or control mice were triple stained with anti-CD3, anti-caspase-3, and anti-␣SMA Abs. CD3⫹ cells adjacent to ␣SMA⫹ cells from control mice stained positive for caspase-3 activity (Fig. 3D; FasL⫹), indicating their apoptotic state. In contrast, there was almost no caspase-3 activity in CD3⫹ cells or associated ␣SMA⫹ cell accumulation in lung sections of bleomycin-treated chimeric gld mice (Fig. 3D; FasL⫺). Fibrotic Lung Myofibroblasts Exhibit FasL-Dependent Prolonged Accumulation in the Air Pouches of Allogeneic Mice. FasL expression is

associated with escape from allogeneic immune surveillance (17, PNAS 兩 December 18, 2007 兩 vol. 104 兩 no. 51 兩 20461

IMMUNOLOGY

Fig. 1. Resistance to Fas-mediated apoptosis in fibrotic lung myofibroblasts. Lung cells were removed from bleomycin- or saline-treated mice 7 days after instillation. Lung myofibroblasts obtained from passages 2–10 were used. (A) CD95 (Fas) overexpression in ␣SMA-positive cells from bleomycin- or control/ saline-treated mice assessed by flow cytometry. (B) Myofibroblasts from lungs of bleomycin-treated mice, but not those from lungs of saline-treated mice, are resistant to Fas-induced apoptosis when incubated for 24 h with 10 ␮g/ml Fas agonist Jo2 mAb (Jo2) or control IgG (control) as assessed by flow cytometry of Annexin V and PI double staining. Results are representative of three experiments. (C) Myofibroblasts from saline- or bleomycin-treated mice were cocultured for 24 h with Jurkat T cells. Live myofibroblasts were quantified by confocal microscopy after staining with trypan blue at 0 and 24 h of coculture. The number of myofibroblasts per well are presented. Results are representative of two experiments.

Fig. 3. Decreased myofibroblast and collagen accumulation with decreased apoptosis of adjacent CD3⫹ lymphocytes in chimeric gld mice. (A) Lung tissues dissected 7 days after bleomycin- or saline-treated IT of chimeric gld (FasL⫺) and control chimeric WT (FasL⫹) mice. In spite of similar CD3⫹ cell distribution (green), a markedly decreased accumulation of ␣SMA⫹ cells (orange) was detected in lung sections of gld chimeric mice, as assessed by confocal microscopy with a ⫻40 lens (Zeiss). Data are representative of six experiments with similar results. (B) Analysis of collagen content in lungs, using the Sircoll assay 7 days after IT bleomycin-treated (filled squares) or saline-treated (open squares) of chimeric gld and WT mice detected an absence of collagen deposition in gld mice. (C) Similar inflammatory cell infiltration into the lungs of chimeric gld and WT mice. Shown are the total BAL cell count, percentage of lymphocytes, and percentage of neutrophils in bleomycin-treated (filled squares) or saline-treated (open squares) chimeric WT or chimeric gld mice. Comparisons were made between bleomycin and saline treatments. *, P ⬍ 0.05. (D) Apoptosis in CD3⫹ lymphocytes adjacent to FasL⫹ but not FasL⫺ myofibroblasts, demonstrated by confocal microscopy of cell triple staining using Cy2-anti-caspase-3 (green), Cy3-anti-␣SMA (red), and Cy5-anti-CD3 (blue) mAbs in lung sections from WT or gld mice. Differential interference contrast was according to Nomarski. Images from a ⫻63, 1.4 oil-immersion, zoom ⫻3 Zeiss lens are also presented. Representative staining is from five to six experiments.

18). Therefore, in an allogeneic inflammatory environment, we tested the in vivo characteristics of fibrotic lung myofibroblasts evading immune surveillance and their resulting accumulation. The air pouch is a well documented model for characterizing immune inflammatory responses (19, 20). We further verified, by using air pouches, whether myofibroblasts isolated from bleomycin-treated lungs are able to avoid immune surveillance after alloengraftment. Myofibroblasts from lungs of bleomycin- or saline-treated C57BL/6 mice were stained with CFSE and further injected into the dorsal air pouches of allogeneic BALB/c and syngeneic C57BL/6 mice. We initially confirmed an inf lammatory response to allogeneic transplantation of myofibroblasts from the lungs of either bleomycin- or saline-treated C57BL/6 mice. Differential analysis of inf lammatory cells in host air pouches revealed that the highest proportion of ⬇2 ⫻ 104 cells recovered from each air pouch were granulocytes (40 –50%), with smaller percentages of macrophages (30 – 40%) and lymphocytes (10 –20%) (Fig. 4A). No significant differences were observed in the 20462 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0705582104

Fig. 4. Fibrotic lung myofibroblasts show prolonged, FasL-dependent survival in the inflamed air pouches of allogeneic host mice. Grafts of CFSEstained lung myofibroblasts from bleomycin- or control/saline-treated WT or gld mice into air pouches of allogeneic BALB/c or control syngeneic C57BL/6 mice. (A) Cellular profile of fluid recovered from the pouches of BALB/c mice. Image is representative of six pouches. (B) RT-PCR showing induction of MCP-3 chemokine mRNA expression after engraftment of myofibroblasts isolated from normal/saline- or fibrotic/bleomycin-treated lungs. Results are representative of three similar experiments. (C) Myofibroblast survival in syngeneic C57BL/6 or allogeneic BALB/c host mice assessed by differences in CFSE/green fluorescence by using a fluorescent binocular microscope and a digital camera. Data are given as mean ⫾ SD fluorescence values (n ⫽ 5– 6), and images are representative of six separate experiments. (D) Individual values of IOD comparing the syngeneic and allo accumulation of bleomycin-treated (filled squares) or saline-treated (open squares) WT mice myofibroblasts. *, P ⬍ 0.01. (E) IOD values comparing the accumulation of myofibroblasts from bleomycintreated WT (FasL⫹) (filled squares) and gld (FasL⫺) (filled triangles) mice after engraftment into syngeneic C57BL/6 or allogeneic BALB/c mice. *, P ⬍ 0.01.

number or composition of pouch cells recovered after allogeneic transplantation of myofibroblasts from normal saline- and fibrotic bleomycin-treated mouse lungs (data not shown). A concomitant increase in myofibroblast expression of MCP-3 mRNA, a chemokine known to recruit inf lammatory cells (21), was found on days 1 and 7 after allogeneic transplantation from either fibrotic (bleomycin-treated) or normal (salinetreated) allogeneic mouse lungs (Fig. 4B). Myofibroblast survival was demonstrated on day 7 by detection of a green fluorescent signal using a fluorescent microscope (Fig. 4C). In syngeneic mice, positive green fluorescence was detected in myofibroblasts after transplantation from both normal (saline: FasL⫹) and fibrotic (bleomycin: FasL⫹) lungs (Fig. 4C Left and Center, respectively). However, in allogeneic BALB/c mice, only myofibroblasts transplanted from fibrotic lungs (bleomycin: FasL⫹), but not from normal lungs (saline: FasL⫹), showed a positive green fluorescence signal on day 7 (Fig. 4C Center and Left, respectively) and day 14 (data not shown). Fluorescence signals were further analyzed quantitatively by integrated optical density (IOD) arbitrary units using Image-Pro Plus software (Media Cybermetics). Normal and fibrotic lung CFSE⫹ myofibroblasts from saline- or bleomycinWallach-Dayan et al.

Discussion Apoptosis, generally induced by immune cells that are known to function as the so-called first line of defense after tissue injury (22), was shown to mediate decreased myofibroblast accumulation during the repair from injury (23). However, in many instances, tissue remodeling does not occur, and granulation tissue evolves into a hypertrophic scar containing myofibroblasts and an inappropriate production of extracellular matrix. This general process is a scheme for fibrosis formation, which applies to many organs after different types of injury (e.g., kidney, liver, heart, and lung). Our study was based on the initial finding that, in the experimental murine model of bleomycin-induced lung fibrosis, myofibroblasts from normal, but not from fibrotic lungs, are sensitive to the induction of apoptosis exerted by immune cells. We hypothesized that, in lung fibrosis, myofibroblasts adopt strategies to avoid apoptosis and consequent clearance by the immune system. Among the mechanisms that could be employed by fibrotic lung myofibroblast to avoid immune cell-induced apoptosis (24–27), we tested the involvement of the FasL pathway (10). We initially show that Jo2 anti-Fas mAb induces the membranal expression of the phosphatidylserine apoptosis marker, as detected by Annexin V binding, in normal, but not in fibrotic, lung myofibroblasts. This finding may explain the defects in their recognition and removal by immune cells (28–30), resulting in their accumulation in lungs. Resistance to Fas-induced apoptosis represents at least, in part, a mechanism by which fibrotic lung myofibroblasts appear invisible to immune cells. Importantly, similar sensitivity to Fas-induced apoptosis was previously reported in myofibroblasts from normal human lungs, compared with those from lungs of humans with IPF (5, 31, 32). These observations are also in accordance with a previous report showing loss of fibroblast resistance to apoptosis during normal repair after acute lung injury (33). The apparent contradiction between the findings that fibrotic lung myofibroblasts express relatively higher levels of Fas molecules, yet resist Fas-induced apoptosis, may be because of the up-regulation of the Fas apoptotic signal inhibitor, FLIP (34), or the release of soluble FasL, or the lack of Fas transmembrane or death domains, or a combination of these possibilities (35–37). The ability of the cell to counterattack immune cells by its expression of FasL (17) is an additional and complementary characteristic that supports avoidance from immune surveillance. We recently reported that fibrotic lung myofibroblasts overexpress FasL and function as effector cells capable of inducing Fas/FasL-dependent apoptosis in lung epithelial cells in vitro and in vivo (14). In this study, we show their ability to induce in vitro apoptosis in a variety of T cells, with Fas/FasL dependency. Concomitantly, chimeric FasL-deficient gld mice with FasLdeficient myofibroblasts exhibited no caspase-3 activity in adjacent CD3⫹ cells, compared with WT (FasL⫹) mice, suggesting Wallach-Dayan et al.

their inability to counterattack immune cells in vivo, as well as their dependence on FasL expression. Myofibroblasts’ ability to resist lymphocyte-induced apoptosis, as well as to induce apoptosis in lymphocytes, may be the cause of myofibroblast accumulation, with resulting collagen deposition, in the lungs of bleomycin-treated mice. These results support the notion that, just like cancer cells, fibrotic lung myofibroblasts accumulate because of their ability to evade immune surveillance. Moreover, using the air pouch model, which is an established tool used to assess immune inflammatory responses (19, 20) in allogeneic BALB/c mice, we showed that fibrotic lung myofibroblasts undergo a delayed rejection that is dependent on their FasL expression, implying their acquired ability to evade allogeneic immune surveillance. We propose that the differential survival of normal and fibrotic lung myofibroblasts in allogeneic mice has implications for myofibroblast accumulation in the inflammatory/fibrotic lung environment. Moreover, increased cell growth of FasLpositive myofibroblasts could indicate a growth advantage over other lung cells, not just a specific ability to counterattack T cells. Therefore, to address this question, we established the air pouch model in allogeneic mice. The lung environment is now excluded, thus relating fibrotic lung myofibroblast accumulation to evasion from immune surveillance rather than interaction with other cells of the lung. The loss of caspase-3 activity in ␣SMA⫹-adjacent CD3⫹ cells, detected in FasL-deficient bleomycin-treated chimeric mice, supports a specific interaction between the FasL on myofibroblasts and immune cells in vivo. However, because gld chimeric mice exclude FasL deficiency only on hematopoietic cells, it remains possible that, after the loss of their FasL expression, other cell types contribute to impaired accumulation of myofibroblasts in the lungs of these mice. Other FasL⫹ effector cells of the nonhematopoietic cell lineage, such as epithelial cells, could have affected myofibroblast accumulation in chimeric gld mice, via regulation of the inflammatory response (38) by FasL (39) after injury. However, the similar inflammation that we detected in bleomycin-treated chimeric gld and WT mice does not support this possibility. Additional mechanisms by which myofibroblast FasL expression is critical in their accumulation during lung fibrosis, besides their counterattack of immune cells, are possible. FasL in myofibroblasts may function as a proliferation signaling receptor, similar to T cells, where FasL was shown to positively regulate their proliferative capacity (40). Myofibroblast may perform an autocrine FasL/Fas-induced proliferation. This possibility can be supported by the detection of increased Fas/FasL expression (14), the previous report on myofibroblast Fas-induced proliferation (41), and the possible autocrine Fas/FasL interaction. In summary, our study identifies a previously unrecognized role for Fas/FasL signaling, which provides a mechanism for myofibroblast accumulation during lung fibrosis. Fibrotic lung myofibroblasts resist Fas-induced apoptosis and kill a variety of T cells. FasL contributes to prolonged myofibroblast survival after transplantation into allogeneic host mice, as well as to increased myofibroblast accumulation, with collagen deposition, in the lungs of bleomycin-treated mice. Moreover, cell-specific FasL deficiency reverses this phenotype, demonstrating its importance in myofibroblast acquisition of an immune privilege mechanism governing cell growth. Based on these findings, we propose that specific immune surveillance of lung myofibroblasts fails because of the emergence of their immune-privileged cell phenotype. The failure of this immune response appears to play a critical role in the evolution of lung fibrosis, and it may have similar consequences for fibrosis in other organs. PNAS 兩 December 18, 2007 兩 vol. 104 兩 no. 51 兩 20463

IMMUNOLOGY

treated mice showed similar in vivo accumulation capacities in the air pouches of syngeneic C57BL/6 mice (Fig. 4D; C57BL/6: saline and bleomycin). In allogeneic BALB/c mice, only myofibroblasts from fibrotic lungs (bleomycin) displayed a positive green fluorescence signal (Fig. 4D; BALB/c) (P ⬍ 0.01), which further decreased by day 30 (data not shown). We then determined whether FasL plays a role in the accumulation after alloengraftment (BALB/c) of fibrotic bleomycintreated lung myofibroblasts and found that FasL⫹ myofibroblasts survive in both BALB/c and in the control C57BL/6 mice. However, FasL⫺ myofibroblasts survive only in C57BL/6, but not in allogeneic BALB/c mice (Fig. 4 C Center and Right and E; FasL⫹ vs. FasL⫺ in BALB/c⫺, compared with C57BL/6, mice).

Materials and Methods Reagents. mAbs and other reagents included 1 mg/ml propidium iodide (PI) stock solution (Calbiochem) in PBS; CFSE (Molecular Probes); rabbit antihuman/mouse active-caspase-3 mAb (R&D Systems); mouse anti-mouse ␣SMA mAb (Sigma–Aldrich); Annexin V-FITC, phycoerythrin-conjugated Fas, and Jo2 anti-Fas mAb with IgG; rabbit anti-mouse CD4 and CD62L mAbs, rat antimouse CD31, and F4/80 mAbs (Pharmingen); and FITC-conjugated F(ab)2 goat anti-rat IgG (H&L), Cy2-conjugated goat anti-hamster IgG (H&L), Cy3conjugated donkey anti-rabbit IgG (H&L), and Cy5-conjugated goat antimouse IgG (H&L) (Jackson Immunoresearch Laboratories). Rat KM81 (IgG2b), anti-mouse pan CD44 mAbs (constant region-specific), and rat 2C11a (IgG1) anti-mouse CD3 mAb were generated from American Type Culture Collection hybridomas and further purified by protein S-Sepharose chromatography (42), Tri reagent (T9424; Sigma–Aldrich), a reverse transcription system (Promega, Madison, WI), and TaqDNA polymerase and ethidium bromide (Sigma–Aldrich). Animals. The 11- to 12-week-old male C57BL/6 and BALB/c mice were purchased from Harlan–Sprague–Dawley. The 11- to 12-week-old Tg(ACTBEGFP)10sb and 5- to 6-week-old lpr and gld C57BL/6-based mice were purchased from The Jackson Laboratory. Chimeric Mice and Lymphoid Organ Cell Transplantation. The generation of these mice has been reported (14). Briefly, after irradiation, mice received 50 –100 ⫻ 106 syngeneic splenocytes. Engraftment was confirmed by FACS analysis by using GFP C57BL/6 donor lymphoid organ cells (data not shown). Intratracheal Instillation (IT). The tracheas of anesthetized mice were cannulated, and 0.08 mg of bleomycin or sterile 0.9% saline was injected as we have previously described (7, 14). Bronchoalveolar Lavage (BAL) Analysis. The BAL analysis has been described (6, 7). A differential count was performed on 200 cells per animal and is expressed as a percentage of total cells recovered. Lung Collagen Content. Total soluble lung collagen was determined by using the Sircol collagen assay kit (Biocolor) as described previously (43). Briefly, the upper lobe of the right lung was homogenized in 5 ml of 0.5 mol/liter acetic acid containing 1 mg of pepsin (Sigma–Aldrich) per 10 mg of tissue residue. Samples were incubated at 21°C for 24 h and centrifuged. The optical density of 100 ␮l of supernatant was evaluated at 540 nm with a spectrophotometer. Collagen results are expressed in micrograms. Isolation of Lung Cells and Myofibroblasts. The isolation of lung cells and myofibroblasts was previously reported in detail (14). Briefly, right lungs were incubated at 37°C with collagenase and minced. Typically, 1 week after initiating myofibroblast culture, ⬎95% of the cells are morphologically myofibroblasts (43). Fas Expression Assayed by Flow Cytometry. Myofibroblast cell surface expression of Fas was assessed by indirect immunofluorescence and analyzed by flow cytometry as previously reported (14, 44). Activated T Cell Isolation. T cells were sorted from spleens of three WT mice and Fas⫺ T cells from lpr mice (Fas-deficient). Activated cells were separated from nonactivated cells by specific CD4 T cell columns (Biotex Laboratories) with the addition of anti-CD44 (supernatant from clone IM7.8.1, diluted 1:15) or 5 ␮g/ml anti-CD62L Abs. 1. Selman M, King TE, Pardo A (2001) Ann Intern Med 134:136 –151. 2. Phan SH (2002) Chest 122:286S–289S. 3. Moodley YP, Caterina P, Scaffidi AK, Misso NL, Papadimitriou JM, McAnulty RJ, Laurent GJ, Thompson PJ, Knight DA (2004) J Pathol 202:486 – 495. 4. Stalder T, Hahn S, Erb P (1994) J Immunol 152:1127–1133. 5. Nagata S (1996) Genes Cells 1:873– 879. 6. Adamson IY, Bowden DH (1974) Am J Pathol 77:185–197. 7. Izbicki G, Segel MJ, Christensen TG, Conner MW, Breuer R (2002) Int J Exp Pathol 83:111–119. 8. Kim R, Emi M, Tanabe K, Uchida Y, Toge T (2004) Cancer 100:2281–2291. 9. O’Connell J, Bennett MW, Nally K, Houston A, O’Sullivan GC, Shanahan F (2000) Ann NY Acad Sci 910:178 –192, 193–195. 10. Ashkenazi A, Dixit VM (1998) Science 281:1305–1308. 11. Helene M, Lake-Bullock V, Zhu J, Hao H, Cohen DA, Kaplan AM (1999) J Leukoc Biol 65:187–195.

20464 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0705582104

Coculture Experiments. Lung myofibroblasts and T cells were extracted ex vivo and cocultured for 24 – 48 h in RPMI medium 1640 (10% FCS) to determine the induction of cell apoptosis. Myofibroblasts were plated at a density of 2–3 ⫻ 105 per milliliter in six-well tissue culture plates. Lymphocytes were added when 1 ⫻ 106 myofibroblasts reached 70 – 85% confluence. T cells were harvested from the culture supernatant and analyzed for apoptosis by Annexin V staining and flow cytometry. Contaminating myofibroblasts were excluded by size. The death of myofibroblasts was determined by trypan blue cell staining.

In Vitro Detection of Apoptotic Cells by Annexin V Affinity Labeling. The 1 ␮g of FITC-conjugated Annexin V and 5 ␮g/ml PI (15) were added to 0.5 ⫻ 106 cells as previously described (14, 44). Immunofluorescence Staining of Tissue Sections. As previously reported (14), sections were incubated with mouse monoclonal anti-␣SMA, rabbit anti-CD3, and/or rat anti-caspase-3 and thereafter with anti-mouse-Cy5 conjugate, anti-rabbit Cy3, and anti-rat-Cy2 Abs, respectively (The Jackson Laboratory) for 30 min at 21°C. CFSE Staining of Myofibroblasts. After culture, 2 ⫻ 106 myofibroblasts per milliliter were resuspended in PBS and stained with CFSE as previously described (45). Briefly, an equal volume of 2.5 ␮M CFSE in PBS was added for 8 min at 21°C. Cells were immediately washed three times in RPMI medium 1640 or PBS. Murine Skin Air Pouch Model for Tracing Transplanted GFPⴙ Lung Myofibroblasts. Mice (6 – 8 weeks of age) were anesthetized with 40 ␮l of 85 mg/ml ketamine and 2 mg/ml xylazine, and dorsal air pouches were raised by injecting 3 ml of sterile air s.c. on days 0 and 3 as described previously (20). The 5 ⫻ 106 myofibroblasts resuspended in 1 ml of sterile PBS were engrafted into the air pouches of groups of five BALB/c or C57BL/6 mice. Mice were killed 7, 14, and 30 days after myofibroblast injection by a lethal dose (100 mg per mouse in 0.5 ml) of pentothal (CTS). After removal of dorsal skin, a fluorescent binocular microscope (Carl Zeiss) and a Nikon Coolpix digital camera were used to detect green fluorescence in the air pouches. Data quantitation in IOD arbitrary units was performed by using Image ProPlus software (Media Cybermetics). Pouch Lavage (PL) Analysis. A two-way syringe was injected into the air pouch, and PL was carried out with 3 ml of saline. The total number of cells was counted. A differential count was performed on 200 cells per animal and expressed as the percentage of total cells recovered. RT-PCR. Reverse transcription of 1 ␮g of total RNA and random primers in 20-␮l reactions was performed as previously reported (44) by using the following primers: ␤-actin sense, 5⬘-GTTGCCATCAATGACCCCTTC-3⬘; ␤-actin antisense, 5⬘-CATGT GGGCCATGAGGTCCAC-3⬘; MCP-3 sense, 5⬘-GTCAAGAAACAAAAGATCCCC-3⬘; and MCP-3 antisense, 5⬘-GGCTTTGGAGTTGGGGTTTTCATGTC-3⬘. Data Analysis and Statistics. Flow-cytometry experiments were repeated at least two to three times. Data of digital camera fluorescence were analyzed according to fluorescence arbitrary units, presented as mean ⫾ SD, and plotted on graphs. Statistical analysis was performed by using Student’s t test for experiments with single comparisons and by using ANOVA with the Bonferroni test for experiments involving multiple comparisons. ACKNOWLEDGMENTS. We thank Anita Kol and Reem Bader for performing experimental work and Shifra Fraifeld and Dr. Michael J. Segal for critical review and editorial assistance in preparing this manuscript. This work was supported by the David Shainberg Fund.

12. Hao H, Cohen DA, Jennings CD, Bryson JS, Kaplan AM (2000) J Leukoc Biol 68:515–521. 13. Huaux F, Liu T, McGarry B, Ullenbruch M, Xing Z, Phan SH (2003) J Immunol 171:5470 – 5481. 14. Golan-Gerstl R, Wallach-Dayan SB, Amir G, Breuer R (2007) Am J Respir Cell Mol Biol 36:270 –275. 15. Griffith TS, Yu X, Herndon JM, Green DR, Ferguson TA (1996) Immunity 5:7–11. 16. Griffith TS, Brunner T, Fletcher SM, Green DR, Ferguson TA (1995) Science 270:1189 – 1192. 17. Bohana-Kashtan O, Civin CI (2004) Stem Cells 22:908 –924. 18. Bellgrau D, Gold D, Selawry H, Moore J, Franzusoff A, Duke RC (1995) Nature 377:30 –32. 19. Clish CB, O’Brien JA, Gronert K, Stahl GL, Petasis NA, Serhan CN (1999) Proc Natl Acad Sci USA 96:8247– 8252. 20. Edwards JC, Sedgwick AD, Willoughby DA (1981) J Pathol 134:147–156. 21. Taub DD, Proost P, Murphy WJ, Anver M, Longo DL, van Damme J, Oppenheim JJ (1995) J Clin Invest 95:1370 –1376.

Wallach-Dayan et al.

34. Curtin JF, Cotter TG (2003) Cell Signal 15:983–992. 35. Wajant H (2002) Science 296:1635–1636. 36. Martin DA, Zheng L, Siegel RM, Huang B, Fisher GH, Wang J, Jackson CE, Puck JM, Dale J, Straus SE, et al. (1999) Proc Natl Acad Sci USA 96:4552– 4557. 37. Van Doorn R, Dijkman R, Vermeer MH, Starink TM, Willemze R, Tensen CP (2002) Cancer Res 62:5389 –5392. 38. Gochuico BR, Miranda KM, Hessel EM, De Bie JJ, Van Oosterhout AJ, Cruikshank WW, Fine A (1998) Am J Physiol 274:L444 –L449. 39. O’Connell J (2000) Arch Immunol Ther Exp (Warsaw) 48:73–79. 40. Suzuki I, Fink PJ (1998) J Exp Med 187:123–128. 41. Aggarwal BB, Singh S, LaPushin R, Totpal K (1995) FEBS Lett 364:5– 8. 42. Rochman M, Moll J, Herrlich P, Wallach SB, Nedvetzki S, Sionov RV, Golan I, Ish-Shalom D, Naor D (2000) Cell Adhes Commun 7:31– 47. 43. Bordin S, Page RC, Narayanan AS (1984) Science 223:171–173. 44. Wallach-Dayan SB, Izbicki G, Cohen PY, Gerstl-Golan R, Fine A, Breuer R (2006) Am J Physiol 290:L790 –L796. 45. Li X, Dancausse H, Grijalva I, Oliveira M, Levi AD (2003) J Neurosci Methods 125:83–91.

IMMUNOLOGY

22. Henson PM (2003) Am J Respir Cell Mol Biol 29:S70 –S76. 23. Desmouliere A, Redard M, Darby I, Gabbiani G (1995) Am J Pathol 146:56 – 66. 24. Cerwenka A, Bakker AB, McClanahan T, Wagner J, Wu J, Phillips JH, Lanier LL. (2000) Immunity 12:721–727. 25. Diefenbach A, Jamieson AM, Liu SD, Shastri N, Raulet DH (2000) Nat Immunol 1:19 –26. 26. Cerwenka A, Baron JL, Lanier LL (2001) Proc Natl Acad Sci USA 98:521–526. 27. Diefenbach A, Jensen ER, Jamieson AM, Raulet DH (2001) Nature 413:165–171. 28. Anderson HA, Maylock CA, Williams JA, Paweletz CP, Shu H, Shacter E (2003) Nat Immunol 4:87–91. 29. Hanayama R, Tanaka M, Miwa K, Shinohara A, Iwamatsu A, Nagata S (2002) Nature 417:82– 87. 30. Fadok VA, Bratton DL, Rose DM, Pearson A, Ezekewitz RA, Henson PM (2000) Nature 405:85–90. 31. Buhling F, Wille A, Rocken C, Wiesner O, Baier A, Meinecke I, Welte T, Pap T (2005) Respir Res 6:37. 32. Moodley YP, Misso NL, Scaffidi AK, Fogel-Petrovic M, McAnulty RJ, Laurent GJ, Thompson PJ, Knight DA (2003) Am J Respir Cell Mol Biol 29:490 – 498. 33. Polunovsky VA, Chen B, Henke C, Snover D, Wendt C, Ingbar DH, Bitterman PB (1993) J Clin Invest 92:388 –397.

Wallach-Dayan et al.

PNAS 兩 December 18, 2007 兩 vol. 104 兩 no. 51 兩 20465