Engraftment of bone marrow (BM)-derived stem cells (SC) to the lung has been reported in several in vivo studies (Kotton et al., 2001; Krause et al., 2001; Grove et al., 2002; Abe et al., 2003, 2004; Ortiz et al., 2003; Yamada et al., 2004; Rojas et al., 2005; Wang et al., 2005). High doses of radiation, known to produce alveolitis and injury to lung resident stem cells, were required to achieve significant grafting of BM-derived cells as epithelial-like (Krause et al., 2001; Grove et al., 2002; Abe et al., 2003). Other evidence for lung engraftment of BM-derived cells was found in biopsies of human lung after lung or bone marrow transplantation under conditions of immunosuppression (Kleeberger et al., 2003; Suratt et al., 2003; Spencer et al., 2005). Results of these studies have been challenged by other observations demonstrating only leukocyte-type engraftment to the lung (Wagers et al., 2002). The role of marrow cells in populating the alveolar epithelium remains controversial, as some recent publications (Chang et al., 2005; Kotton et al., 2005) suggest that possible technical difficulties and complex alveolar structures result in artifacts of overlapping fluorescent signals from endogenous and donor-derived cells. Achievements and problems in this area have been recently reviewed (Prockop et al., 2003; Pitt and Ortiz, 2004; Van Haaften and Thebaud, 2006; Weiss et al., 2006). In previous studies, attention was focused mostly on the alveolar epithelium (Krause et al., 2001; Kotton et al., 2003; Yamada et al., 2004). Epithelium of the distal conducting airways is also an attractive target for stem cell-based therapy of such diseases as cystic fibrosis, chronic or acute bronchitis, or asthma. Engraftment of BM-derived cells in the epithelium of the bronchial tree has been less investigated. Recent studies evaluating engraftment of adult marrow-derived stem cells after specifically targeted injury to airways by endotoxin, NO2 (Beckett et al., 2005), or naphthalene (Loi et al., 2006) did not find significant levels of engraftment in irradiated mice. We have also reported (Serikov et al., 2005) that, in sublethally irradiated animals transplanted with whole bone marrow, the amount of chimerism in the bronchial epithelium was substantial; however, most of these cells were positive for leukocyte markers. Therefore, to avoid participation of leukocyte lineages in recovery of bronchial epithelium, we further used a population of cultured plastic-adherent stromal cells from bone marrow. We hypothesized that engraftment of BM-derived mesenchymal cells into the airway epithelium might be promoted by the specificity and severity of epithelial injury by naphthalene.
The biochemical mechanisms of naphthalene lung toxicity are well characterized (Mahvi et al., 1977; Stripp et al., 1995; Buckpitt et al., 2002). Pulmonary toxicants (furans, hydrocarbons) are activated by cytochrome P-450 monooxygenaze system in Clara cells (Buckpitt et al., 2002). Clara cells are the main source of epithelial cell renewal after injury and development (Stripp et al., 1995). At 1 to 2 days after naphthalene, Clara cells die and exfoliate; this results in the appearance of numerous dead cells and debris in the airways; markers of Clara cells disappear, but gradually return back by day 14 (Van Winkle et al., 1995).
Therefore, the goal of this study was to investigate the temporal sequence of engraftment of transplanted BM-derived stem cells to the bronchial epithelium in naphthalene injury. We used a green fluorescent protein (GFP)-tagged plastic-adherent mesenchymal-enriched stromal stem cell line (termed below MSC) from BM. Our first aim was to develop a reliable method of tracking GFP+ cells in the bronchial epithelium with several controls: by immunofluorescence (IF) staining with Z-stack analysis, immunohistochemistry (IHC), and fluorescence in situ hybridization (FISH). The second aim was to investigate engraftment of BM-derived cells in the epithelium of airways at a minimal rate of lung radiation damage after naphthalene injury. Our third aim was to determine the temporal pattern of engraftment of MSC, and characterize the phenotype of BM-derived cells grafted to the epithelium of conducting airways.
MATERIALS AND METHODS
Donor bone marrow was harvested from 2–4 week old C57BL/6 male GFP+ Tg mice (C57BL/6-Tg[ACT6EGFP]1Osb/J), Jackson Laboratories (Bar Harbor, ME). Recipients were strain-matched 3- to 5-week-old female C57BL/6 mice. Animal studies were performed according to protocols approved by institutional animal use committee.
We isolated MSC by a similar protocol as laid out for plastic-adherent mesenchymal-enriched bone marrow cell cultures (Friedenstein, 1976; Pittenger et al., 1999; Badoo et al., 2003; Meirelles and Nardi, 2003; Peister et al., 2004). Therefore, our definition of MSC is similar to the above term.
Bone marrow cells from GFP Tg mice were obtained by flushing out of tibias and femurs using 1 ml of growth medium (GM): Dulbecco's modified Eagle's medium with 25 mM HEPES and 10% of fetal bovine serum (FBS, Gibco BRL). Cells were washed in GM by spinning out at 400 × g for 10 min, counted with Trypan Blue to evaluate their viability and resuspended. The resuspended cells were cultured immediately at density of 2 × 105/ml in four six-well tissue culture plates in a humidified 5% CO2 incubator at 37°C. After 72 hr, the nonadherent cells were removed. Cells were propagated the same way up to 40 passages. After 5 and 25 passages, cells were checked for the presence of surface antigens CD45, CD11b, and CD34 by FACS, and these were not observed. For animal transplantation, we used cells at 15–25 passages.
Differentiation assays were performed similarly to described by Phinney et al. (1999). Adipogenic differentiation was induced by seeding the MSC for 3 weeks in GM supplemented with 10−8 ml/L dexamethasone (Sigma, St. Louis, MO) and 5 μg/ml insulin (Sigma). The cells containing droplets of fat were identified by staining with 3.75% Oil Red (Sigma). Osteogenic differentiation was induced by culture in GM containing 10 mM β-glycerol phosphate, 50 μg/ml ascorbic acid, and 10−8 ml/L dexamethasone for 3 weeks. The calcium-containing precipitates were visualized after staining with 2% Alizarin Red S (Sigma) adjusted to pH 4.2 with ammonium hydroxide. Myogenic differentiation was induced by culture for 2 weeks in GM containing 1.5 μg/ml amphotericin B (Sigma) and then for 1 week in the medium without amphotericin. Cells were then fixed for 10 min in 4% paraformaldehyde and 70% ethanol overnight at 4°C followed by IF evaluation of expression of the early (desmin) and late (myosin heavy chain) muscle cell differentiation markers.
Recipients received sublethal dose of whole-body irradiation (5.05 Gy) the day before transplantation. Under general anesthesia (pentobarbital, 15 mg/kg), incision was made on the neck and animals were infused with one million MSC in 0.2 ml of phosphate buffered saline (PBS) into the jugular vein. Animals were allowed to recover and some were followed up to 9 months.
In the first experimental group, 1 month after transplantation of MSC, chimeric animals were subjected to 250 mg/kg naphthalene (Sigma) IP in corn oil (n = 18). Control chimeric animals received corn oil without naphthalene (n = 6). Animals were killed and lungs removed for examination at 2, 6, 12, and 30 days after naphthalene administration. Six nonchimeric animals with naphthalene injury were used as secondary controls.
Intratracheal MSC Instillation
Six animals received intratracheal instillation of PBS (n = 3, control) or 0.75 × 106 MSC, (n = 3). These animals were killed immediately after instillation and lungs were fixed for IF.
Immunohistological Analysis of the Lung
Animals were euthanized in 100% CO2. Chest was opened, and incision was made in the left atrium, and the pulmonary artery was flushed with ice-cold PBS with 10% sucrose (5 ml) under pressure 20 cm H2O. Catheter was inserted into the trachea and lungs were fixed through the airways with pure ethanol under pressure 25 cm H2O. Lungs were then removed from the chest and lung lobes were kept in pure ethanol at 4°C for further analyses. One lung was used for cryosectioning, and the other for paraffin sectioning. Before cryo-sectioning, lungs were washed in a large volume of PBS for 24 hr, placed in 30% sucrose (in PBS) for 12 hr, frozen in liquid nitrogen, briefly embedded in OCT compound, immediately frozen and cut in Tissue-Tek Cryostat (Miles Labs, IN) at −25°C. Tissue slices 10 μm thick were mounted on poly-lysine coated glass and kept at −20°C for immunostaining. To avoid development of nonspecific tissue fluorescence, tissues were processed within several days from the date of fixation.
IF protocol was adopted from (Zalenskaya and Zalensky, 2004). Slides were washed with PBS, post-fixed with ethanol, washed twice, exposed to 0.2% Triton X-100 in PBS for 5 min, and then washed with PBS twice again. Tissues were further blocked with 3% goat serum, 2% horse serum, 3% bovine serum albumin (BSA), 0.1% Tween 20 in 4 × standard saline citrate (SSC) for 20 min. Appropriate primary antibody (in 1% goat serum, 1% horse serum, 1% BSA, 0.1% Tween 20 in 2 × SSC) was added for 1 hr at 37°C, tissues were washed 3 times with PBS and blocked with 3% goat serum, 2% horse serum, 3% BSA/0.1% Tween 20 in 4 × SSC for 20 min. Red fluorescent dye (Rhodamine, Texas Red, Alexa Fluor or Cy5) -conjugated secondary antibody (in 1% goat serum, 1% horse serum, 1% BSA in 2 × SSC/0.1% Tween 20) was added for 1 hr at 37°C. Control tissues were stained with Isotypic anti-mouse or anti-rabbit primary antibody. Tissues were washed 3 times with PBS, mounted on glass slides in Fluoro-Guard antifade reagent (Bio-Rad, Hercules, CA), and coverslips were applied. Laser confocal fluorescence microscopy for dual color staining (Axiovert 100, LSM 510, Zeiss, Germany) was performed at the UC Davis Department of Medicine Confocal Imaging Core Facility. Fluorescence microscopy for tri-color staining was performed using Olympus IX70 fluorescent microscope with CCD camera and “InVivo” digital imaging software by Media Cybernetics, Inc.
Paraffin Sections and Immunohistology
Paraffin sections of the lung, IHC, and hematoxylin–eosin staining were performed at Pathology Department, UCSF (San Francisco, CA). IHC staining for GFP was performed using standard horseradish peroxidase–diaminobenzidine technique.
Staining for Y-chromosome
Staining for Y-chromosome by FISH was performed in paraffin sections using Y-chromosome stain from Cambio Ltd. (Cambridge, UK), according to the manufacturer's instructions.
The following antibodies were used: anti-GFP rabbit IgG (1:50), anti-GFP mouse monoclonal antibody (1:250; Molecular Probes, Eugene, OR), and anti-GFP rabbit IgG G1544 (1:50; from Sigma); staining for epithelial cells: mouse monoclonal anti-pancytokeratin (1:50; Calbiochem, San Diego, CA, Clone B311.1) and rabbit anti-mouse pancytokeratin IgG (1:10; from Zymed Labs, South San Francisco, CA); anti-CCSP antibody (1:10,000) was a kind gift from Dr. Barry R. Stripp, University of Pittsburg, Pennsylvania. Antibody to epithelial marker Peroxiredoxin V (Kinnula et al., 2002) was from our own source (Kropotov and Tomilin, 1997), anti α-ENaC (1:100) was from Calbiochem. Staining for leukocytes was done with rat anti-mouse CD45 antibody, 1:20 (BD Biosciences, clone 30-F11). Rabbit antibody to PCNA (proliferating cell nuclear antigen, 1:200) was from ABCAM. Isotype anti-mouse and anti-rabbit antibodies (prediluted) were both from Zymed Labs. For characterization of MSC by IHC, antibodies against the following antigens were used: Vimentin (Ab-2, Clone V9, NeoMarkers), recognizing 57-to 60-kDa vimentin; mouse monoclonal antibody against actin (smooth muscle Ab-1 (Clone 1A4, NeoMarkers); rabbit polyclonal antibody against β-catenin (Sigma); rat monoclonal antibody against CD45 (Serotec, UK); and rabbit polyclonal antibody against factor VIII (vonWillebrand factor). The desmin and myosin heavy chain (MHC) expression in MSC were determined by using the D3 and MF20 monoclonal antibodies (Developmental Studies Hybridoma Bank, University of Iowa). Secondary Rhodamine, Cy-5, Texas-red or Alexa-Fluor® 633 or Alexa Fluor® 586 labeled anti-mouse, anti-rat and anti-rabbit antibodies were from Molecular Probes (Invitrogen). When only green fluorescent antibodies were used, or only GFP signal was analyzed, tissues were co-stained with propidium iodine (PI). For FACS analyses (FACS Calibur, Becton-Dickinson, Oxnard, CA) antibodies against CD11b, CD45, CD31, CD34, CD117 (c-kit) were from Zymed Labs, labeled with PE or Cy5.
At least six different sections from each lung were used for analyses. Two hundred visual fields were analyzed in each section. Numbers of GFP+ cells or cell clusters (more than three adjacent cells) were determined per total number of cells in a section of the whole lung and expressed per 105 cells. Total number of cells was determined from surface area of section and averaged number of cell nuclei in 10 visual fields (1.2 × 105 μm2). Statistical analysis was performed using Mann–Whitney–Wilcoxon test.
Characterization of MSC
According to fluorescence-activated cell sorting analyses (FACS), our GFP+ MSC were negative for CD45, CD11b (Supplementary Figure S1), CD31, CD34, CD117 (c-kit), and a fraction of cells was positive for CD90 (4%; data not shown). By IF, MSC were positive for actin, vimentin, β-catenin, factor VIII. Western blot analyses did not show expression of pancytokeratin in these cells, as compared with cell lines A549 and Calu-3 (data not shown). Results of differentiation assays are shown in Supplementary Figure S1. MSC demonstrated ability for differentiation into multiple lineages after 25 passages.
One month posttransplantation of MSC in sublethally irradiated animals, we observed partial engraftment of MSC to the bone marrow, with chimerism ranging from 1 to 12% (mean, 4 ± 2%). We also observed some degree of chimerism in liver and spleen, but not in other organs. Chimerism in the lungs before naphthalene injury was not observed. FACS analyses of peripheral blood demonstrated that less than 0.001% of white blood cells were GFP-positive.
Morphological features of naphthalene injury in the lung were similar to previously described. We observed massive exfoliation of epithelial cells along conducting airways, which was manisfested at 2–12 days after naphthalene (Supplementary Figure S2).
Authentification of GFP+ Cells in the Lung
Specific attention was given to distinguishing between authentic GFP fluorescence and nonspecific green fluorescence of tissue. For this purpose, we performed several internal controls. As we determined in our preliminary experiments, procedures of paraffin sectioning could quench GFP fluorescence. Therefore, paraffin and cryosections were immunostained for GFP. A typical appearance of a rare cluster of GFP-positive cells is presented in Figure 1 (IF). Other examples of such structures, as well as appropriate controls for primary antibodies, positive and negative controls are shown in Supplement Figure 3. Signal for GFP was strong and allowed to clearly distinguish GFP from nonspecific green fluorescence. GFP+ cells were not present in tissues stained with isotype primary antibody. IHC images, which allow to avoid artifacts due to nonspecific fluorescence of tissues, are given in Figure 2, altogether with positive and negative controls. Lungs of GFP Tg mouse, shown as positive controls in Figures 2A and 3D,E, demonstrate expression of GFP in airway epithelial cells, and pattern of this expression is similar to the pattern observed in patches of GFP+ epithelium of naphthalene-injured chimeric mice. In the lungs of control chimeric noninjured animals, we observed occasional GFP+ cells, but these cells were present in the lung parenchyma, not in the epithelium of airways.
To distinguish staining for GFP from nonspecific fluorescence of dead cells, cell clusters (rafts) and debris, lungs of nonchimeric animals were used as controls at 2–6 days after naphthalene. In the lungs of nonchimeric animals injured with naphthalene, we did not observe GFP+ cells, after parafin fixation and staining for GFP at the same time periods after injury (Fig. 3F). Although intra-airway cell debris was autofluorescent, laser and detection setting of microscope allowed to clearly separate this phenomenon from GFP signal by IF (Supplement Fig. 3A). Controls for specificity of antibody staining and discrimination of GFP fluorescence from nonspecific fluorescence of epithelial cells due to fixation procedures were further done (Supplementary Figure S3). As yet an additional control for the presence of GFP+ cells in the lung, we instilled cultured GFP+ MSC into the airways of anesthetized mice. Mice were killed immediately, and we observed multiple GFP+ cells in the lungs of animals instilled with MSC (Supplementary Figure S3G), but not in PBS-instilled controls (Supplementary Figure S3H).
Naphthalene Lung Injury
We did not observe GFP+ cells in the airway epithelium of noninjured chimeric animals transplanted with MSC. In animals transplanted with MSC, we identified patches of epithelial lining, completely composed of only donor-derived GFP+ cells by IHC (Fig. 2). Further Z-stack confocal analyses with IF costaining for pancytokeratin was done (Fig. 4, Supplementary Figure S4). Z-stack analyses by confocal microscopy showed location of signals from pancytokeratin and GFP in the same plane (Fig. 4A,B). Staining for pancytokeratin, GFP and nuclear staining by DAPI demonstrated colocalization of these signals in nucleated cells (Fig. 4C–L). FISH for Y-chromosome and costaining for pancytokeratin also confirmed these findings (Fig. 5). GFP+ cells in the bronchial epithelium and interstitium were positive for PCNA, nuclear marker of proliferating cells (Fig. 6).
Most of the GFP+ cells in the epithelial layer were positive for other markers of airway epithelial cells: α-subunit of ENaC and peroxiredoxin V (data not shown). There was no clear positive staining of GFP+ cells in the epithelium for CCSP, marker of Clara cells (data not shown), vimentin, CD34, and SP-B. Thus, GFP+ cells located in the bronchial epithelium have the following phenotype according to IF analyses: CD45−, CD34−, vimentin−, CCSP−, SP-B−. They were positive for pancytokeratin, ENaC, peroxiredoxinV, PCNA. Some GFP cells present in the interstitium of airways, were CD45+, vimentin+. As shown in Table 1, PCK-positive GFP+ cell clusters were observed in the epithelium within first 6 days after injury. In 30 days after naphthalene, the number of GFP-positive cells in the epithelium of conducting airways was very limited and observed only in 1 of 4 animals. Differences between control noninjured animals and animals after naphthalene injury at 2–6 days were statistically significant, the same as the difference between groups at 2–6 days and 30 days.
Table 1. Mean number of cell clusters of GFP+ cells in lungs of recipients following naphthalene injury per 105 lung cellsa
No. of animals
No. of cell clusters per 105 lung cells
No. of cells per cluster
Mean ± SD values are given. At least six different sections from each lung were used for analyses. Two hundred visual fields were analyzed in each section. Numbers of GFP+ cells or cell clusters (more than three adjacent cells) were determined as total number of cells per total cells in section of the whole lung and expressed per 105 cells. Total number of cells was determined from surface area of section and averaged number of cell nuclei in ten visual fields (1.2 × 105 μm2). Statistical analysis was performed using Mann–Whitney–Wilcoxon test with significant difference accepted at α < 0.05. GFP, green fluorescent protein; MCS, plastic-adherent mesenchymal-enriched stromal stem cell line.
Significant difference between control non-injured group and MSC-transplanted animals at 2–6 days after naphthalene.
Significant difference between MSC-transplanted animals at 2–6 days after naphthalene and 30 days after naphthalene administration.
After naphthalene injury in MSC-transplanted chimeric animals, we identified patches of GFP+ cells in large and small conducting airways by IF, IHC, and FISH. GFP+ cells were present in lung parenchyma and epithelium of conducting airways at 2–6 days after naphthalene. GFP+ cells in the epithelium of airways were positive for pancytokeratin, ENaC α-subunit, but not CCSP. GFP+ cells formed clear isolated patches of the bronchial epithelium, consistent with clonal formation, as they were also positive for PCNA, marker of proliferating cells. At day 30, only very rare GFP+ cell patches were present in the epithelium.
These data suggest that after acute naphthalene-induced injury, bone marrow-derived cells participate in lung epithelial reparation in a time-dependent manner. After initial development of patches in the epithelium, consistent with clonal formation, these cells nearly disappear from the epithelial lining at the time of complete recovery from injury.
We used accepted procedures to isolate MSC. The ability to adhere to plastic is a primary feature, underlying the general procedure for MSC isolation and usage. This approach, developed by Friedenstein (1976) is now accepted and widely used (Prockop, 1997; Phinney et al., 1999; Pittenger et al., 1999; Spees et al., 2003; Meirelles and Nardi, 2003; Peister et al., 2004). This procedure of MSC isolation from bone marrow does not include initial elimination of CD45- and/or CD11b-positive cells. In the first few passages, mouse bone marrow plastic-adherent cells represent a mixed population of stromal and hematopoietic cells. Negative selection of this population with antibody against CD11b, CD34 and CD45 leads to concentration of MSC, but does not change their stem cell characteristics (Phinney et al., 1999; Baddoo et al., 2003). More recent investigations (Peister et al., 2004) have revealed that MSC from later passages (after seven) prepared from five different mouse strains varied by the expression of stem cells markers as CD34 and Sca-1, but did not express lineage-specific markers like CD45, CD11b, and others. This finding is similar to other publications, showing that MSC from different sources do not express markers of hematopoietic cells (De Ugarte et al., 2003; LeBlank et al., 2003). Cells derived by us did not display CD11b or Cd45 after passage 5. Obviously, the more accurate term for MSC that we used would be “plastic-adherent mesenchymal-enriched bone marrow cell cultures,” as we specified in the Materials and Methods section.
Some controversies have been considered regarding authentification of donor-derived cells (discussed in Weiss et al., 2006). We used GFP+ and Y-chromosome+ cells as markers of BM-derived cells. We presumed that green GFP fluorescence might not be a reliable cell tag and, therefore, performed several controls to ensure presence of cells with GFP. Processing of tissue rapidly and often irreversibly quenches GFP fluorescence. Nonspecific fluorescence of tissues (bronchial epithelial lining), hardly distinguishable from GFP fluorescence, rapidly appears unless tissues are cryosectioned after a few hours of fixation in 100% ethanol. To verify our results with different approaches, we further performed IHC, which was free of nonspecific fluorescence-related errors, and FISH for Y-chromosome. Both confirmed our results, although FISH might be subject to the same sources of errors as IF for GFP. Naphthalene injury is specifically prone to false-positive identification of donor-derived cells, as cell debris is present in airways.
We used additional controls, in which naphthalene injury was induced in nonchimeric mice. In these cases, nonspecific fluorescent cells were observed at days 2–6, but their fluorescence intensity was different and lower than that of GFP-positive cells in cryosections, and many-fold lower than fluorescence of sections stained for GFP with FITC or Alexa Fluor-labeled antibodies, both in cryo- and in paraffin sections. Further, colocalization of GFP signal and signal from epithelial-specific marker pancytokeratin were confirmed by Z-stack analysis.
We observed only a temporary presence of GFP+ cells in the injured lung. Question remains, whether some pool of circulating GFP+ cells marginated to the pulmonary vessels and migrated to epithelium after naphthalene injury. Our experience from other similar studies shows that initially, 1–3 days after transplantation of MSC, some GFP-positive cells can be observed inside lung circulation, preferentially adherent to walls of large veins. However, after 7–10 days, these cells are not observed morphologically. In our opinion, GFP+ MSC migrated to the lungs only after injury by means of circulation from the place of their homing. Severe immunological conflict is unlikely between sublethally irradiated recipients and donors, both C57BL/6. However, it is likely that small radiation damage with minimal effects on the lung's own stem cells resulted in insufficient proliferative prevalence for transplanted stem cells. Requirement for radiation-induced lung injury for engraftment in otherwise noninjured lung is consistent with previous reports. High levels of engraftment reported previously in the lung were achieved at clearly damaging levels of radiation (Krause et al., 2001; Abe et al., 2003). These levels of radiation damage the pool of stem cells in the lung, which are important for lung recovery (Reynolds et al., 2004) and, thus, promote repopulation by SCs from other sources. High degrees of lung chimerism were reported in humans after BM transplantation (Kleeberger et al., 2003; Suratt et al., 2003). Immunosuppression and possible damage of lung SCs are inevitable in humans undergoing BM transplantation. In studies of human lungs after lung transplantation, significant epithelial pulmonary chimerism up to 2.5–8.0% was observed (Kleeberger et al., 2003). In another study, chimerism in bronchial epithelium, type II pneumocytes, and seromucous glands was reported after BM transplantation, and epithelial structures with signs of chronic injury demonstrated chimerism up to 24% (Suratt et al., 2003). Our current data confirm recent observations by Loi et al. (2006) and Beckett et al. (2005), as well as older observations by Ortiz et al. (2003) and Yamada et al. (2004) that overall level of lung engraftment of BM-derived cells is very small and unlikely to have physiological significance. This finding contrasts with other studies that reported high levels of chimerism in experimental animals after lethal irradiation, as stated above. Previously, we reported participation of bone marrow cells in restoration of some properties of naphthalene-injured airway epithelium (Serikov et al., 2005). The current study has confirmed our initial observations with several different methods and the use of MSC. Both of our studies are in good agreement with recent findings by Loi et al. (2006), although the experimental protocols were different. Loi et al. used Cftr knockout mice as recipients of either total bone marrow cells in irradiated animals (800 rads) or cultured stromal bone marrow in nonirradiated animals, and administered naphthalene (275 mg/kg) to induce airway epithelial injury. In nonirradiated animals, stromal bone marrow cells were administered 3 days after naphthalene. Under this protocol, in situ examination of recipient mouse lungs demonstrated rare (0.025%) chimeric airway epithelial cells, some of which (0.01%) expressed Cftr protein, as well as cytokeratin, CCSP, and pro-SP-C. Naphthalene-induced airway remodeling nonsignificantly increased the number of chimeric airway epithelial cells expressing Cftr. Naphthalene injury in irradiated whole bone marrow-transplanted Cftr knockout mice did not result in appearance of CD45−, CCSP-positive cells up to 1 month after naphthalene (only pro-SP-C cells, which were also located in alveolar septa). Data on cytokeratin-positive donor-derived cells under this protocol were not given. Our protocol was different in several ways. First, our animals were irradiated with a much lower sublethal dose (5 Gy vs. 8 Gy in study by Loi et al., 2006), which makes a major difference in the levels of lung engraftment (discussed in Weiss et al., 2006). Second, we assessed the number of cell clusters (at least three cells adjacent to each other) per lung, rather than total quantity of donor-derived cells per recipient lung. Finally, we investigated the presence of donor-derived cells only in the epithelium of airways, not in the alveoli. Unlike Loi et al., we did not observe the presence of donor-derived GFP+ cell clusters in airway epithelium of noninjured lungs, although these cells were certainly present in alveoli and peribronchial interstitium in our study. As Loi et al. (2006), we did not observe CCSP-positive donor-derived cells clusters in irradiated animals, though our data show cytokeratin-positive donor derived cell clusters. Similar to these investigators, we found a very limited number of donor-derived cells in the airway epithelium 1 month after naphthalene injury, even though a significant amount of BM-derived cells was present in the lung parenchyma during the acute phase of injury (2–6 days). Although Loi et al. (2006) presented evidence for clustering of donor-derived cells and the appearance of adjacent donor-derived cells, their illustrations did not show large patches of donor-derived epithelium. In previous studies (Grove et al., 2002; Abe et al., 2003, 2004; Rojas et al., 2005), signs of clonal formation by BM-derived cells were not reported. However, we observed morphological signs of clonal formation in the bronchial epithelium: BM-derived cells formed clear patches, and some were positive for the nuclear marker of proliferating cells PCNA. In our opinion, it is also unlikely that cell fusion may result in appearance of GFP+ and Y-chromosome+ cells organized in clusters. Question of cell fusion as a mechanism of chimerism is fully discussed by Spees et al. in (2003). These findings allow us to hypothesize that clonal formation from donor-derived cells can occur in injured epithelium. It is possible that some specific means of delivery of donor-derived stem cells are required to increase effectiveness of repopulation of epithelium. This may include intra-airway delivery of cell suspensions or preformed epithelial layers.
We conclude that toxic injury to the Clara cells in the bronchial epithelium triggers appearance of bone marrow-derived cells at a few sites of severe bronchial epithelial injury. During the acute phase of injury, there is evidence that some of these cells display proteins that are specific for epithelial cells. Bone marrow-derived stem cells may play a transient, regulatory role in the repair of injured bronchial epithelium with minimal long-term engraftment.
M.A.M was funded by the NIH, N.G. received a CIRM Stem Cell Research Grant, and V.M.M. and B.P. were funded by the Russian Fund of Basic Research. Some of the initial results were published in part earlier (Serikov et al., Cytotherapy 2005;7:483–493). The naphthalene model of lung injury was suggested to us by Dr. C. Plopper (UC Davis).