Mesenchymal Stem Cell Engraftment in Lung Is Enhanced in Response to Bleomycin Exposure and Ameliorates Its Fibrotic Effects

Home , Mesenchymal stem cell

Mesenchymal stem cell engraftment in lung is enhanced in response to bleomycin exposure and ameliorates its fibrotic effects

Luis A. Ortiz*†, Frederica Gambelli*, Christine McBride‡, Dina Gaupp‡, Melody Baddoo‡, Naftali Kaminski§, and Donald G. Phinney‡

*Division of Occupational Medicine, Department of Environmental and Occupational Health, and §Division of Pulmonary, Allergy, and Critical Care, University of Pittsburgh, Pittsburgh, PA 15261; and ‡Center for Gene Therapy, Tulane University Health Sciences Center, New Orleans, LA 70112

Communicated by Darwin J. Prockop, Tulane University, New Orleans, LA, May 15, 2003 (received for review April 2, 2003) Previously we described a reliable method based on immunodeple- demonstrate that early but not late administration of MSCs tion for isolating mesenchymal stem cells (MSCs) from murine bone ameliorated the fibrotic injuries observed in the lungs of BLM- marrow and showed that, after intracranial transplantation, the treated mice. cells migrated throughout forebrain and cerebellum and adopted neural cell fates. Here we systemically administered MSCs purified Materials and Methods by immunodepletion from male bleomycin (BLM)-resistant BALB͞c Isolation of Murine MSCs and Alveolar Type II Cells. MSCs were mice into female BLM-sensitive C57BL͞6 recipients and quantified isolated from mouse bone marrow as described (7) except that engraftment levels in lung by real-time PCR. Male DNA accounted whole bone marrow was plated at a density of 1.46 ϫ 106 cells for 2.21 ؋ 10؊5% of the total lung DNA in control-treated mice but per cm2 and cultured for 8–10 days before harvest. MSCs (up to in animals exposed to BLM before 40 ϫ 106 cells) were added to M-280 Dynabeads (five beads per (0.05 ؍ was increased 23-fold (P MSC transplantation. Fluorescence in situ hybridization revealed cell; Dynal, Oslo) conjugated to an anti-CD11b antibody (10 ␮g that engrafted male cells were localized to areas of BLM-induced per mg of beads; PharMingen) in a volume of 1 ml and incubated injury and exhibited an epithelium-like morphology. Moreover, on a rotator at 4°C for 45 min. Successive rounds of immu- purification of type II epithelial cells from the lungs of transplant nodepletion by using antibodies against CD34 and CD45 recipients resulted in a 3-fold enrichment of male, donor-derived (PharMingen) were conducted similarly. Immunodepleted cells cells as compared with whole lung tissue. MSC administration were suspended in Hanks’ balanced salt solution and used for immediately after exposure to BLM also significantly reduced the further experiments as described. Alveolar epithelial type II cells degree of BLM-induced inflammation and collagen deposition were isolated from the lung tissue as described (9). Approxi- within lung tissue. Collectively, these studies demonstrate that mately 2 ϫ 106 cells were used to prepare genomic DNA for murine MSCs home to lung in response to injury, adopt an epithe- real-time PCR. Alternatively, 5 ϫ 104 cells were cultured and lium-like phenotype, and reduce inflammation and collagen dep- fixed on positively charged slides for analysis by fluorescence in osition in lung tissue of mice challenged with BLM. situ hybridization (FISH).

ϫ 5 diopathic pulmonary fibrosis is a crippling disease character- Fluorescence-Activated Cell Sorting. Aliquots (2.5 10 ) of immu- ␮ ized by progressive dyspnea and is associated with a high nodepleted MSCs were suspended in 50 l of wash buffer (0.1% I ͞ mortality rate (1). Presently, no effective therapies to reverse or sodium azide 1.0% BSA in PBS) containing a rat anti-mouse ͞ retard the course of the disease are available (1). Several recent CD16 CD32 antibody (Fc Block, PharMingen) at a concentra- ␮ ϫ 6 studies have demonstrated that stem cells derived from adult tion of 1 g per 1 10 cells and incubated for 5 min at 4°C in ␮ ␮ tissues can home to and͞or participate in the development of the dark. Wash buffer (50 l) containing 5 g of the appropriate lung tissue (2–4), raising the possibility that stem cell-based fluorochrome-conjugated primary antibody (PharMingen) was therapies may be developed for effective intervention of lung added, and the cells were incubated for an additional 20 min. ␮ diseases. However, the nature of the signals involved in the Cells were washed twice with 200 l of wash buffer, and the recruitment of stem cells into the lung, the extent of stem cell extent of cell labeling was evaluated by using a Beckman Coulter engraftment, and the effect of injury or disease on these Model Epics XL. Isotype controls were run in parallel by using processes remains undetermined. the same concentration of each antibody tested. To begin to address these issues we compared the engraftment

of systemically administered mesenchymal stem cells (MSCs) in BLM-Induced Lung Injury and MSC Administration. Female 6- to MEDICAL SCIENCES ͞ lung tissue of normal mice to that of mice exposed to bleomycin 10-week-old C57BL 6 mice (Charles River Breeding Laborato- (BLM), which represents a well established model of lung injury ries) were anesthetized via an i.p. injection of tribromoethanol resulting in pulmonary fibrosis (5, 6). We also evaluated whether (Aldrich) and then exposed to BLM as described (10, 11). ϫ 5 ␮ MSC administration altered the clinical course of BLM-induced Immunodepleted male MSCs (5 10 in 200 l of PBS) were lung injury. Importantly, these studies used MSCs isolated from injected into the jugular vein immediately after or 7 days after murine bone marrow by immunodepletion, a method that re- challenge with BLM. Animals were killed 14 days after BLM ͞ moves contaminating hematopoietic cells, which possess an exposure by an injection of sodium pentobarbital (120 mg kg), appreciable engraftment potential in vivo, from plastic adherent perfused with cold 0.9% NaCl, and the left lungs were removed marrow cultures (7). Moreover, we used a real-time PCR assay and used for evaluation of collagen content or RNA isolation. that specifically targets sequences on the mouse Y chromosome The right lungs were fixed in situ for2hbytheintratracheal to quantify levels of male MSCs in the lungs of female transplant recipients (8). Our results reveal that MSCs engraft in the lungs Abbreviations: BLM, bleomycin; MSC, mesenchymal stem cell; FISH, ﬂuorescence in situ of normal mice at low levels, but engraftment is increased hybridization; HA, hyaluronan; MMP, matrix metalloproteinase; AU, arbitrary units. significantly in response to BLM-induced injury. In the latter †To whom correspondence should be addressed at: Department of Environmental and case engrafted donor cells adopted an epithelium-like morphol- Occupational Health, Graduate School of Public Health, University of Pittsburgh, A731 ogy and also copurify with type II epithelial cells. Our data also Crabtree Hall, 130 De Soto Street, Pittsburgh, PA 15261. E-mail: [email protected].

www.pnas.org͞cgi͞doi͞10.1073͞pnas.1432929100 PNAS ͉ July 8, 2003 ͉ vol. 100 ͉ no. 14 ͉ 8407–8411 Downloaded by guest on September 29, 2021 instillation of 10% neutral formalin (Sigma) and then postfixed for 24 h. Sagittal sections (4-␮m) of paraffin-embedded lung tissue were used for FISH. Collagen deposition was estimated by measuring the total hydroxyproline content of the lung as described (10, 11). Results were expressed as micrograms of hydroxyproline per lung. All experiments involving live animals were approved by the animal use and care committee of Tulane University Health Sciences Center.

Real-Time PCR. Real-time PCR was carried out as described (8) on a 7700 sequence detection system (Applied Biosystems) by using the PCR primers 5Ј-TTTTGCCTCCCATAGTAGTAT- TTCCT-3Ј and 5Ј-TGTACCGCTCTGCCAACCA-3Ј and the TaqMan probe 5Ј-FAM-AGGGATGCCCACCTCGCCAGA- TAMRA-3Ј. Standard curves were generated by serially diluting male mouse genomic DNA into female mouse genomic DNA prepared from liver.

FISH. Localization of male Y chromosome sequences in 4-␮m paraffin-embedded sections of mouse lung tissue was done by using the Universal ISH kit and a mouse Y chromosome paint probe (Innogenex, San Ramon, CA) as described (8). Sections were counterstained with 4Ј,6-diamidino-2-phenylindole or ethidium bromide (1 mg͞ml), photographed by using a Leica Fig. 1. Quantiﬁcation of MSC engraftment in mouse lung by real-time PCR. (Deerfield, IL) RX-DMV upright fluorescent microscope at- (A) Relationship between threshold cycle number and the percentage of male tached to a digital camera (Cooke Sensicam), and rendered by genomic DNA within the indicated samples. The standard curve (blue dots) using SLIDEBOOK software (Intelligent Imaging Innovations, was generated by using samples containing from 100% to 0% male genomic Denver). DNA. (B) Histogram of the data in A showing the percentage of male genomic DNA within lung tissue or type II epithelial cells isolated from female mice 14 days after exposure to BLM, MSC administration (MSCs), or MSC administra- Osteopontin and Metalloproteinase RNA Expression in Mouse Lung. tion after BLM exposure (BLMϩMSC). Plotted values represent the arithmetic Microarray analysis was used to quantify the effect of BLM mean, and error bars represent the standard deviation (*, P ϭ 0.05, and **, P ϭ exposure on osteopontin mRNA levels in lung as described (12). 0.001 by Student’s t test). Labeled cRNA was hybridized to the GeneChip Mu6500 array (Affymetrix, Santa Clara, CA). Data points for C57BL͞6mice represent total RNA pooled from six animals, whereas data points derived cells were localized in areas of BLM-induced lung injury for 129͞J mice represent total RNA isolated from individual mice and appeared to conform to the morphology of epithelial cells (n ϭ 3–5). Fold changes were determined by dividing the mean of (Fig. 2). To evaluate this engraftment further, we isolated the average differences in each experimental condition by that for epithelial type II cells from the lungs of female transplant the control-treated group. Metalloproteinase RNA expression was recipients and measured their content of male DNA by real-time evaluated by using a RiboQuant kit (PharMingen) following man- PCR. Epithelial type II cells isolated from the lungs of female ufacturer recommendations. mice exposed to BLM contained no detectable male DNA. In contrast, male DNA accounted for 1.37 ϫ 10Ϫ3% of the total Results DNA content of epithelial type II cells isolated from the lungs BLM-Induced Lung Injury Enhances MSC Engraftment. We exploited of BLM-treated female mice administered male MSCs (Fig. 1A). a well characterized murine model of BLM-induced pulmonary Therefore, purification of epithelial type II cells resulted in a fibrosis to evaluate how injury effects engraftment of stem cells 3-fold enrichment of male donor MSCs as compared with that in lung. Accordingly, we isolated by immunodepletion murine contained in whole lung tissue. This engraftment level also MSCs from the bone marrow of male BLM-resistant BALB͞c represents a 62-fold increase as compared with mice not exposed mice and transplanted them via the i.v. route into female to BLM (P ϭ 0.001) (Fig. 1B). BLM-sensitive C57BL͞6 mice that were challenged with BLM. Freshly isolated epithelial type II cells were identified by We then quantified engraftment levels of male, donor-derived Papanicolaou staining of lamellar bodies (Fig. 3A). Culture of cells in the lungs of female transplant recipients 14 days after the cells resulted in adhesion and loss of lamellar bodies. The BLM challenge using real-time PCR. No male DNA was de- cultured cells also expressed cytokeratins (Fig. 3B), reacted with tected in lung tissue isolated from control female mice, and male osmium tetroxide, but did not express vimentin and desmin (data DNA accounted for Ϸ2.21 ϫ 10Ϫ5% of total lung DNA in female not shown), indicating that the isolation procedure yielded a mice (n ϭ 7) that were administered MSCs. In contrast, male homogeneous population. FISH with a Y chromosome paint DNA accounted for 5.18 ϫ 10Ϫ4% of total lung DNA in female probe resulted in a positive hybridization signal for most of the transplant recipients (n ϭ 10) exposed to BLM before injection type II epithelial cells isolated from the lungs of male mice (Fig. of MSCs. This result represents a 23-fold increase in engraftment 3C). We also identified a few cells that demonstrated a positive levels of donor-derived cells as compared with mice not exposed hybridization signal in preparations of type II epithelial cells to BLM (P Ͻ 0.05) (Fig. 1). from female mice exposed to BLM and administered MSCs (Fig. To confirm these findings we localized engrafted male cells in 3D). These results are consistent with our real-time PCR data the lungs of BLM-treated female transplant recipients by FISH and confirm that male donor MSCs copurified with type II using a mouse Y chromosome paint probe. Relatively few cells epithelial cells from lung. were detected in lung tissue that demonstrated a positive hybridization signal against the Y chromosome, which was antic- BLM Enhances Osteopontin Expression in Mouse Lung. Immuno- ipated based on our quantitative analyses. The detected donor- depleted murine MSCs uniformly express CD44 but do not

8408 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.1432929100 Ortiz et al. Downloaded by guest on September 29, 2021 Fig. 4. Effect of BLM on osteopontin and MMP mRNA levels in lung. (A) Changes in osteopontin expression levels in lung tissue at various times after BLM exposure. The data points for C57BL͞6 mice represent total RNA pooled from lung tissue of six animals. The data points for 129͞J mice represent the mean of three to ﬁve separate expression values analyzing total RNA prepared from individual mice. (B) Total RNA was isolated from the lungs of untreated C57BL͞6 mice (lane 1) or mice 14 days after exposure to BLM alone (lanes 2 and 3) or BLM exposure followed by MSC administration (lanes 4 and 5) and subjected to ribonuclease protection analysis as described in Materials and Methods. Relative levels of each MMP transcript were determined by densi- tometry and normalized to the levels of GAPDH mRNA.

osteopontin mRNA in the lungs of BLM-sensitive strains (C57BL͞6 and 129͞J). Endotracheal challenge with BLM induce Fig. 2. (A)A4-␮m section of lung tissue from a male mouse hybridized with a time-dependent increase in osteopontin mRNA levels in lung, an FITC-conjugated Y chromosome paint probe and counterstained with resulting in a 3- and 5-fold increase as compared with untreated ethidium bromide. The arrows indicate representative nuclei that contain the mice at 7 and 14 days after BLM exposure, respectively (Fig. 4A). Y chromosome. (B) Localization of Y chromosome-positive nuclei (arrow) Therefore, two different ligands that bind to the CD44 receptor within lung tissue of a female mouse 14 days after BLM exposure and admin- are induced in lung tissue by BLM. istration of male MSCs. (C and D) A series (8–20) of images through the z axis were rendered from lung sections (as in B) to demonstrate that the probe ͞ hybridization signals (arrows) are associated speciﬁcally with cell nuclei. Engrafted MSCs Alter BLM-Induced Lung Injury in C57BL 6 Mice. The intratracheal exposure of mice to BLM but not to saline resulted in the development of subpleural areas of inflammation that express the v4 and v6 isoforms of this polymorphic glycoprotein encompassed Ϸ55–60% of the lung parenchyma, which was that functions as a cell-surface receptor for hyaluronan (HA) consolidated with loss of normal alveolar architecture, and also (13) and osteopontin (14). Interaction between CD44 and HA is involved the bronchi and vasculature (Fig. 5). Administration of fundamental for the normal development of the lung and the MSCs immediately after challenge with BLM reduced the extent repair of lung injury (15). Although BLM exposure is known to of inflammation within the lung as evidenced by large areas of induce expression of HA in lung (16, 17), its effect on osteopon- undamaged tissue with normal alveolar architecture. In contrast, MEDICAL SCIENCES tin expression is not well characterized. Accordingly, we evalu- no significant difference in the degree of inflammation was ated the effects of BLM exposure on expression levels of observed in animals administered MSCs 7 days after challenge

Fig. 3. Copuriﬁcation of engrafted MSCs with type II epithelial cells. (A) Freshly isolated alveolar epithelial type II cells show darkly stained cytoplasmic granules after pap staining. (B) Cultured epithelial type II cells stain positively with an anticytokeratin antibody. (C) FISH analysis of cytospin preparations of alveolar type II cells (50,000 per slide) isolated from the lungs of a male mouse. (D) FISH analysis of alveolar epithelial type II cells isolated from female mice 14 days after exposure to BLM and administration of male MSCs. (Scale bars, 100 ␮M; magniﬁcation, ϫ40.)

Ortiz et al. PNAS ͉ July 8, 2003 ͉ vol. 100 ͉ no. 14 ͉ 8409 Downloaded by guest on September 29, 2021 Fig. 5. Effect of MSC engraftment on BLM-induced lung injury in mice. Low-magniﬁcation photomicrographs of tissue sections prepared from the lungs of a C57BL͞6 mouse 14 days after saline exposure (A), 14 days after BLM exposure (B), and 14 days after BLM exposure and MSC administration (C). Note that the extent of BLM-induced inﬂammation, as evidenced by the wedge-shaped area of pneumonitis in B, is greatly reduced in C. TB, terminal bronchiole; PL, pleural surface. (Scale bar, 20 ␮m.)

with BLM as compared with those exposed to BLM alone An important aspect of this work is that the MSCs used in this (Fig. 5). study were enriched from plastic adherent cultures of murine BLM exposure in mice also results in an increase in collagen bone marrow by immunodepletion, which is in contrast to several deposition in the lung. Therefore, we compared the amount of previous reports that evaluated the engraftment and therapeutic hydroxyproline, a modified amino acid specifically found in potential of MCs by transplantation in vivo of plastic adherent collagen, in lung tissue of normal mice to those challenged with marrow cells (4, 18, 19). We and others (20–22) have shown that BLM with or without MSC administration. The hydroxyproline plastic adherent cultures elaborated from murine bone marrow content of lung tissue from control-treated mice was 48.1 Ϯ 7 ␮g contain a variety of hematopoietic cell types, which persist in the per lung. At 14 days after exposure to BLM this valued increased cultures even after serial passage and exhibit an appreciable significantly to 93 Ϯ 20 ␮g per lung (P Ͻ 0.05). In contrast, the engraftment potential in vivo. Therefore, the aforementioned hydroxyproline content of lung tissue from animals that were studies are confounded in that transplantation of plastic adher- administered MSCs immediately after BLM challenge was 67 Ϯ ent cells precludes a direct measure of the contribution made by 1 ␮g per lung, which was also significantly different from that of MSCs to the experimental outcome. Our experimental ap- control-treated mice but represented a statistically significant proach, in contrast, provides a more direct measure of the decrease (n ϭ 7; P Ͻ 0.05) compared with animals exposed to engraftment and therapeutic potential of MSCs in lung. BLM alone. Administration of MSCs 7 days after BLM exposure At present, the mechanism by which BLM-induced lung injury also resulted in a reduction in hydroxyproline content of lung 14 augments MSC engraftment in lung tissue is unclear. Endotra- days later (81 Ϯ 18 ␮g per lung), but this difference was not cheal challenge of mice with BLM leads to lung fibrosis and statistically significant as compared with animals exposed to occurs in three stages. The first stage, acute pulmonary toxicity, BLM alone (P Ͼ 0.05). results from BLM-induced DNA-strand scission and is charac- In addition to modulating the extent of collagen deposition, we terized by apoptosis and necrosis of alveolar epithelial cells found that MSC administration also altered the expression level (23–25). This is followed by an inflammatory phase, wherein of various matrix metalloproteinases (MMPs) in BLM-treated activated immune cells migrate into the lung and release a mice. Compared with control animals, BLM exposure increased variety of cytokines including tumor necrosis factor ␣, which the expression level of transcripts encoding MMP2 [27.6 Ϯ 6.9 promotes the development of lung fibrosis (26, 27). The final vs. 96.5 Ϯ 24.6 arbitrary units (AU), respectively] and MMP9 stage is characterized by enhanced collagen deposition within (5.48 Ϯ 1 vs. 7.51 Ϯ 1 AU, respectively) within lung tissue (Fig. lung, expansion of the lung interstitium due to the proliferation 4B). In contrast, administration of MSCs to animals immediately of fibroblasts and smooth muscle cells, and increased expression after BLM exposure reduced the extent of BLM-induced ex- of MMPs that participate in the remodeling of the injured tissue pression of MMP2 (67.3 Ϯ 16 AU) and completely attenuated (28). Each of these phases may affect engraftment of MSCs in BLM-induced expression of MMP9 (5 Ϯ 0.5 AU) in lung tissue. lung uniquely. For example, although we isolated MSCs from a BLM exposure also increased transcript levels of interstitial BLM-resistant strain of mice, it is possible that a significant collagenase (MMP13) in lung as compared with untreated mice percentage of administered cells were lost because of the acute (9.83 Ϯ 2 AU vs. 1.63 Ϯ 0.2 AU, respectively), which was reduced toxicity of BLM or immune rejection by the host, because our modestly in animals that were administered MSCs after BLM population of MSCs constitutes an allotransplant. Conceivably, exposure (6.43 Ϯ 0.6 AU). Collectively, these results indicate optimizing the dose of MSCs and timing of administration and that early but not late administration of MSCs reduces the transplanting autologous stem cells after BLM challenge may degree of inflammation and fibrosis in the lungs of mice enhance engraftment levels further. The homing and engraft- challenged with BLM. ment of MSCs in lung may have been augmented by engagement of their CD44 receptor by HA and osteopontin. BLM exposure Discussion is known to increase expression of HA in lung (16, 17), and we In this report we demonstrate that engraftment in lung tissue of have shown in this report that osteopontin mRNA levels are also systemically administered MSCs occurs at low levels in normal increased under these conditions. Alternatively, release of cy- mice but is increased significantly in response to BLM-induced tokines and mitogens by immune-infiltrating cells may also affect lung injury. We also show that administration of MSCs imme- the homing, engraftment, and proliferative status of MSCs in diately after BLM challenge protects lung tissue from BLM- lung. Last, breakdown of the microvasculature due to BLM induced injury, as evidenced by a significant reduction in in- toxicity may provide a nonspecific mechanism by which MSCs flammation, collagen deposition, and MMP activation within gain increased access to lung tissue. lung tissue. Collectively, these findings indicate that systemic Our data also indicate that MSC administration after BLM administration of MSCs may be beneficial in the treatment of exposure protects lung tissue from injury by significantly reduc- lung disease even if engraftment levels of the administered cells ing the extent of inflammation and fibrosis. Several explanations are comparatively low. may account for this effect of MSC administration. For example,

8410 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.1432929100 Ortiz et al. Downloaded by guest on September 29, 2021 MSCs may limit the injurious effects of BLM by replacing to fibrosis (32, 33). Recent studies have shown that scavenging alveolar epithelial type II cells, which are thought to function as of HA degradation products in lung by CD44-expressing cells is stem cells in lung and are known targets of apoptotic signals necessary for the resolution of BLM-induced lung injury (14). induced in lung by BLM (29). Differentiation of MSCs into Scavenging by MSCs of HA and osteopontin, which plays an epithelial type II cells may partially restore the stem cell pool, important role in the pathogenesis of BLM-induced fibrosis by leading to increased genesis of alveolar cells for the resolution of promoting the migration, adhesion, and proliferation of fibro- disrupted alveolar surfaces, thereby augmenting the repair pro- blasts in lung, may be an important mechanism of MSC action. cess. Generation of lung stem cells may also explain how low Clearly, the timing of MSC administration is an important levels of MSC engraftment in lung can produce a positive effect determinant of their biological response. Delaying MSC admin- on the health status of mice exposed to BLM. Notably, several istration by 7 days after BLM challenge did not inhibit engraft- recent reports have shown that fusion with somatic cells may ment but eliminated the ability of the cells to alter the course of enable stem cells to adopt unorthodox phenotypes (30, 31). At disease progression. Therefore, MSCs may produce factors that present we have not determined whether fusion of MSCs with impinge on molecules expressed early but not late during the somatic cells in lung is responsible for their epithelium-like course of BLM-induced injury. morphology or whether the cells change fate in vivo in response In summary, our data indicate that MSCs, given their pro- to transplantation to a novel residence. If a significant fraction pensity to engraft in lung tissue and ability to ameliorate the of engrafted MSCs do fuse with somatic cells, these events do not injurious effects of BLM, constitute an effective cellular vehicle seem to limit their ability to ameliorate the injurious effects of for the treatment of lung disease. BLM in lung. Alternatively, MSCs may protect against BLM-induced injury This work was supported in part by U.S. Public Health Service Grant by altering the microenvironment of lung at sites of engraftment. ES10859-01A1 (to L.A.O.) from the National Institutes of Environmental For example, MSCs may produce antagonists of tumor necrosis Health, and by the Louisiana Gene Therapy Research Consortium (New factor ␣ or other cytokines that disrupt signal pathways leading Orleans) and HCA-The Health Care Company (Nashville, TN).

1. American Thoracic Society͞European Respiratory Society (2002) Am. J. 18. Pereira, R. F., O’Hara, M. D., Laptev, A. V., Halford, K. W., Pollard, M. D., Respir. Crit. Care Med. 165, 277–304. Class, R., Simon, D., Livezey, K. & Prockop, D. J. (1998) Proc. Natl. Acad. Sci. 2. Jiang, Y., Balkrishna, N., Reinhardt, R. L., Schwartz, R. E., Keene, C. D., USA 95, 1142–1147. Ortiz-Gonzalez, X. R., Reyes, M., Lenvik, T., Lund, T., Blackstad, M., et al. 19. Jin, H. K., Carter, J. E., Huntley, G. W. & Schuchman, E. H. (2002) J. Clin. (2002) Nature 418, 41–49. Invest. 109, 1183–1191. 3. Krause, D. S., Theise, N. D., Collector, M. I., Henegariu, O., Hwang, S., 20. Kerk, D. K., Henry, E. A., Eaves, A. C. & Eaves, C. J. (1985) J. Cell. Physiol. Gardner, R., Neutzel, S. & Sharkis, S. J. (2001) Cell 105, 369–377. 125, 127–134. 4. Kotton, D. N., Yang-Ma, B., Cardoso, W. B., Sanderson, E. A., Summer, R. S., 21. Witte, P. L., Robinson, M., Henley, A., Low, M. G., Stiers, D. L., Perkins, Williams, M. C. & Fine, A. (2001) Development (Cambridge, U.K.) 128, 5181–5188. S., Fleischman, R. A. & Kincade, P. W. (1987) Eur. J. Immunol. 17, 5. Bowden, D. H. (1984) Lab. Invest. 50, 487–488. 1473–1484. 6. Harrison, J. H. & Lazo, J. S. (1987) J. Pharmacol. Exp. Ther. 243, 1185–1194. 22. Phinney, D. G., Kopen, G., Isaacson, R. L. & Prockop, D. J. (1999) J. Cell. 7. Kopen, G. C., Prockop, D. J. & Phinney, D. G. (1999) Proc. Natl. Acad. Sci. Biochem. 72, 570–585. 96, USA 10711–10716. 23. Harrison, J. H., Hoyt, D. G. & Lazo, J. S. (1989) Mol. Pharmacol. 36, 231–238. 8. McBride, C., Gaupp, D. & Phinney, D. G. (2003) Cytotherapy 5, 7–18. 24. Hagimoto, N., Kuqano, K., Nomoto, Y., Junitake, R. & Hara, N. (1997) Am. J. 9. Corti, M., Brody, A. R. & Harrison, J. H. (1996) Am. J. Respir. Cell Mol. Biol. Respir. Cell Mol. Biol. 16, 91–101. 14, 309–315. 25. Ortiz, L. A., Moroz, K., Liu, J.-Y., Hoyle, G. W., Hammond, T., Banks, W., 10. Ortiz, L. A., Champion, H., Lasky, J. A., Gozal, E., Hoyle, G., Friedman, M., Hamilton, R. F., Holian, A., Brody, A. R. & Friedman, M. (1998) Am. J. Physiol. Hyman, A. L. & Kadowitz, P. J. (2002) Am. J. Physiol. 282, L1209–L1221. 275, L1208–L1218. 11. Ortiz, L. A., Lasky, J. A., Lungarella, G., Cavarra, E., Martorana, P., Banks, W., 26. Phan, S. H. & Kunkel, S. L. (1992) Exp. Lung Res. 18, 29–43. Peschon, J., Brody, A. R. & Friedman, M. (1999) Am. J. Respir. Cell Mol. Biol. 20, 825–833. 27. Piguet, P. F., Collart, M. A., Grau, G. E., Kapanci, Y. & Vassalli, P. (1989) 12. Kaminski, N., Allard, J., Zuo, F., Griffiths, M. J. D., Morris, D., Huang, X., J. Exp. Med. 170, 655–663. Sheppard, D. & Heller, R. A. (2002) Proc. Natl. Acad. Sci. USA 97, 1778–1783. 28. Hoyt, D. G. & Lazo, J. S. (1998) J. Pharmacol. Exp. Ther. 246, 765–771. 13. Aruffo, A., Stamenkovic, I., Melnick, M., Underhill, C. B. & Seed, B. (1990) 29. Fehrenbach, H. (2001) Respir. Res. 2, 33–46. Cell 61, 1303–1313. 30. Ying, Q.-L., Nichols, J., Evans, E. P. & Smith, A. G. (2002) Nature 416, 14. Weber, G. F., Ashkhar, S., Glimcher, M. J. & Cantor, H. (1996) Science 271, 545–548. 509–512. 31. Terada, N., Hamazaki, T., Oka, M., Hoki, M., Mastalerz, D. M., Nakano, Y., 15. Teder, P., Vandivier, R. W., Jiang, D., Liang, J. Cohn, L., Pure, E., Henson, Meyer, E. M., Laurence, M., Petersen, B. E. & Scott, E. W. (2002) Nature 416, P. M. & Nobel, P. W. (2002) Science 296, 155–158. 542–545. 16. Teder, P. & Heldin, P. (1997) Am. J. Respir. Cell Mol. Biol. 17, 367–385. 32. Piquet, P. F. & Vesin, C. (1994) Eur. Respir. J. 7, 515–518. 17. Savani, R. C., Hou, G., Liu, P., Wang, C., Simmons, E., Grimm, P. C., Stern, 33. Ortiz, L. A., Lasky, J. A., Hamilton, R. F., Holian, A., Hoyle, G. W., Banks,

R., Greenberg, A. H., DeLisser, H. M. & Khalil, N. (2000) Am. J. Respir. Cell W., Peschon, J., Brody, A. R. & Friedman, M. (1998) Exp. Lung Res. 24, MEDICAL SCIENCES Mol. Biol. 23, 475–484. 721–743.

Ortiz et al. PNAS ͉ July 8, 2003 ͉ vol. 100 ͉ no. 14 ͉ 8411 Downloaded by guest on September 29, 2021