Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Despite significant inroads in the understanding of cellular and molecular mechanisms regulating lung development, little is known of mechanisms regulating pulmonary epithelial maintenance and remodeling during adulthood. In most regenerative organ systems, recovery from epithelial injury involves the participation of two different types of regenerative cell populations: stem cells and transit-amplifying (TA) cells. Characteristics that are invariably associated with the stem-cell pool include a relatively undifferentiated phenotype, infrequent proliferation in the steady state, capacity for self-renewal, and ability to produce daughter cells capable of functioning as TA cells (1). In contrast, progenitor or TA cells have the capacity for only a finite number of cell divisions, have more limited differentiation capacity, and generally fulfill certain differentiated functions (1, 4, 5).

Several epithelial cell types have been identified in conducting airways that have properties of TA cells. These include Clara/secretory cells, pulmonary neuroendocrine cells (PNECs) and basal cells (6). However, stem cells capable of replenishing depleted TA cells of the pulmonary epithelium have not been definitively identified (10). In general, differentiated functions of TA cells increase their susceptibility to environmental agents, thus requiring a stem cell pool to effect their replacement after injury. This property of TA cells has been exploited to reveal the identity and location of stem cells within different regenerative organs (11, 12). Such a scenario has been clearly demonstrated within the cornea, where injury to the central corneal epithelium is associated with accelerated proliferation of limbal cells, resulting in replenishment of depleted TA cells and subsequent regeneration of the injured epithelium (5, 11, 13).

Clara cells are the major TA cell population of the rodent conducting airway epithelium (6, 14). Moreover, Clara cells also have the capacity to metabolize lipophilic compounds through cytochrome P450-mediated oxidation, a function that renders them susceptible to injury by lipophilic compounds (15, 16). We have previously demonstrated that epithelial renewal after naphthalene-induced TA (Clara)-cell depletion involves two proliferative cell types, Clara cell secretory protein (CCSP)-expressing (CE) Clara cells and PNECs, that colocalize predominantly within the airway bifurcation zone of bronchioles (8, 17, 18). These data support the involvement of naphthalene-resistant regenerative cells in the process of epithelial renewal after TA cell depletion. Either naphthalene-resistant CE cells or PNECs may fulfill this role. CE cells of the neuroepithelial body microenvironment were previously shown to include a subpopulation, termed variant CE (vCE) cells, that lacked detectable cytochrome P450 2F2 isozyme (CYP2F2) protein, the cytochrome P450 isoenzyme responsible for the generation of toxic metabolites of naphthalene (8, 19). These data support the potential existence of a naphthalene-resistant subpopulation of CE/vCE cells that are specifically maintained within the neuroepithelial body (NEB) microenvironment.

PNECs, in contrast to Clara cells, lack differentiated functions that render them susceptible to lipophilic pollutants (8). PNECs are commonly organized into innervated clusters, called NEBs, that have been proposed to serve various functions, including regulation of embryonic lung growth and maturation through elaboration of a variety of potent neuropeptides (20). Several studies have suggested that PNECs are quiescent cells with limited self-renewal capacity (23). However, it was recently demonstrated that PNECs do have a self-renewal capacity and can be activated to undergo multiple rounds of proliferation after TA (Clara) cell depletion (8, 9, 18, 27). Even though the differentiation potential of PNECs has not been elucidated, increases in the proliferative fraction of PNECs after naphthalene-induced Clara cell ablation and the finding of cells with promiscuous expression of both PNEC and Clara cell markers within the NEB microenvironment, identify PNECs as candidate stem cells (8, 18).

The present study was designed to test the hypothesis that PNECs represent a stem cell pool that is sufficient for epithelial renewal after ablation of TA cells. The rate of cell proliferation was measured as a criterion for identification of stem cells within the NEB microenvironment. In addition, transgenic mice allowing conditional ablation of CE cells were used to determine the contribution that NEB-associated vCE cells and PNECs make to airway regeneration. We demonstrate that the NEB microenvironment harbors slow-cycling populations of cells that represent a candidate stem-cell pool and that PNECs are unable to effect epithelial renewal after elimination of CE cells.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals

Generation and characterization of CCSP-herpes simplex virus thymidine kinase (HSVtk) (transgene in which HSVtk gene is placed at the downstream of mouse CCSP promoter) (CCtk) transgenic mice have been previously described (9). CCtk transgenic mice used in this study were maintained as a specific pathogen-free in-house colony. They were allowed food and water ad libitum and maintained on a 12-h daylight/dark cycle. Representative animals from the colony were screened quarterly for the absence of pathogens using a comprehensive 16-agent serologic panel (Microbiological Associates, Rockville, MD). CCtk male mice used in this study were 2 to 4 mo old.

Naphthalene and Ganciclovir Treatments

Naphthalene treatment was carried out as previously described (8), except that animals received 275 mg/kg body weight dose. Animals were injected intraperitoneally with 5 mg/ml bromodeoxyuridine (BrdU) (Sigma, St. Louis, MO) in phosphate-buffered saline (PBS) solution to yield a dose of 50 mg/kg body weight 70 h after naphthalene treatment, and were killed 2 h later.

CCtk transgenic mice were exposed to 10 mg of ganciclovir (GCV) over a 24-h period using miniosmotic pumps (ALZET Osmotic Pumps, #2001D; ALZA Corp., Palo Alto, CA). Cytovene-IV (GCV sodium; Hoffmann-La Roche, Inc., Nutley, NJ) was dissolved in sterile normal saline solution to give a final concentration of 50 mg/ml. One-day osmotic pumps (8 µl/h) were charged with 200 µl of 50 mg/ml GCV solution and primed by incubating in normal saline for 3 h at 37°C under sterile conditions. CCtk mice were anesthetized with 37.5 mg/kg Avertin intraperitoneally, the dorsal flank was shaved and disinfected, and the primed miniosmotic pumps were inserted through a 1.5-cm incision into small subcutaneous pockets. Incision wounds were closed using metal wound clips. Animals were allowed to recover for the indicated durations before analysis.

Labeling of Proliferative Cells

For continuous [³H]thymine deoxyribose (TdR) labeling and determination of label retention, male FVB/n mice were treated with 275 mg/kg naphthalene in corn oil by intraperitoneal (i.p.) injection. Six hours after naphthalene treatment, a 7-d miniosmotic pump charged with 230 µl [³H]TdR (1 mCi/ml) was implanted as described earlier. Pumps were removed 9 d later. Animals were killed 9 (n = 6) or 45 (n = 6) d after naphthalene treatment. Tissue was fixed and 5-µm sections were stained for CCSP or calcitonin gene-related peptide (CGRP) as described later. Slides were washed overnight in 1× PBS, dehydrated, coated in NTB2 emulsion (Kodak, Rochester, NY), and autoradiographed for 25 to 35 d as previously described (18). Tissue was counterstained with Mayer's hematoxylin after autoradiography.

To continuously label proliferating cells after GCV treatments in CCtk mice, GCV pumps were removed at Day 2 of recovery and replaced with 14-d miniosmotic pumps delivering 0.5 µl/h of sterile 20 mg/ml BrdU in normal saline solution. Animals were killed on Day 10 of recovery (continuous labeling for 8 d). Control CCtk transgenic mice (n = 4) were continuously administered with BrdU in the same manner for 8 d, without the initial GCV treatment.

Tissue Collection

Control and GCV-treated mice were killed by i.p. injection of 100 mg/kg pentobarbital followed by exsanguination. Lungs were exposed, left lobes were recovered for isolation of total RNA, and right lobes were inflation-fixed using methods described previously (8, 17, 28, 29).

RNA Analysis

Total RNA was isolated as previously described (28, 29). Changes in the relative levels of CCSP and CYP2F2 messenger RNAs (mRNAs) were determined by S1 nuclease protection assay using L32 for normalization as previously described (28, 29). CCSP and CYP2F2 were used as Clara cell-specific markers (30, 31). A phosphoimager and ImageQuant analysis software (Molecular Dynamics, Sunnyvale, CA) were used to quantify band intensities. The mean ± standard error of the mean (SEM) is reported for each treatment group. Differences between means were considered statistically significant when P < 0.05, as determined by Student's t test.

Immunohistochemistry

Adjacent serial 5-µm sections of lung from each control and GCV-treated lung tissue were stained for CCSP, CGRP (8, 18, 26), and acetylated tubulin (ACT) (32) to detect Clara cells, PNECs, and ciliated cells, respectively. Standard immunohistochemical techniques were used (33). Rabbit antirat CCSP was obtained from G. Singh (University of Pittsburgh, Pittsburgh, PA) and used at a dilution of 1:12,000. Rabbit antirat CGRP and mouse monoclonal anti-ACT were purchased from Sigma and used at dilutions of 1:10,000 and 1:8,000, respectively. Controls were immunostained in the absence of primary antibody (data not shown).

Tissue from each continuously BrdU-labeled control and GCV-treated CCtk animal was subjected to adjacent section analysis for CCSP, BrdU, and CGRP. CCSP and CGRP immunostaining was done as described earlier. Sheep anti-BrdU was purchased from Fitzgerald Industries International, Inc. (Concord, MA) and used at a dilution of 1:3,000 to 1:6,000. For BrdU immunostaining, sections were treated sequentially with 0.05% Proteinase K solution in PBS for 2 min at room temperature, 1 N HCl for 45 min, and 3% H₂O₂ (aq.) for 15 min, and blocked with 5% bovine serum albumin (BSA) in PBS for 30 min at room temperature. Primary antibody was diluted in 3% BSA in PBS and incubated with the tissue overnight at 4°C in a humidified chamber. Antigen-antibody complexes were detected using an ABC Vectastain Kit (Vector Laboratories, Burlingame, CA) and a DAB substrate kit (Vector Laboratories), according to the manufacturer's instructions. Sections were counterstained in either hematoxylin or Nuclear Fast Red (Vector Laboratories).

Morphometric Analysis of PNEC Hyperplasia

Morphometry for PNEC hyperplasia was performed as previously described (18). Lung tissue sections, 5 µm, from each of the control and GCV-treated animals (Days 6 and 10 of recovery) were immunostained for CGRP and counterstained with hematoxylin as described earlier. The basement membrane (BM) of conducting airway epithelium on each section was traced using the measurement function of Image-Pro Plus (Media Cybernetics, Silver Spring, MD) to determine its total length under a magnification of ×100. The BM subtending CGRP-immunoreactive (IR) regions was measured similarly, except it was done at ×400 magnification. The percentage of BM subtended by CGRP-IR was defined as (CGRP-IR BM/total BM) × 100. CGRP-IR cells containing nuclear profiles were counted on each section to determine the number of PNECs per millimeter of BM (#PNEC/ mm BM). The total number of PNECs in the NEBs was divided by the total number of NEBs to determine the average number of PNECs per NEB. NEBs were defined as clusters of PNECs containing more than two CGRP-IR cells. The mean ± SEM for four untreated control and four treated mice (for each recovery group; for Day 10, n = 3) is expressed in bar graphs. Differences between means were considered statistically significant when P < 0.05, as determined by Student's t test.

Dual Immunofluorescent Labeling and Analysis

Standard immunofluorescence methods were used to simultaneously detect either CGRP and CCSP or HSVtk and CCSP on adjacent serial sections from untreated CCtk lung tissue. Polyclonal rabbit anti-HSVtk (serum, diluted 1:10) was purchased from W. C. Summers (Yale University, New Haven, CT) and goat antirat CCSP was supplied by G. Singh (University of Pittsburgh). CGRP and CCSP were detected with rabbit antirat CGRP (1:4,000 dilution) and goat antirat CCSP (1:3,000 dilution), respectively. HSVtk and CCSP were detected with rabbit anti-HSVtk (1:1,000 dilution) and goat antirat CCSP (1:3,000 dilution), respectively. All primary antibodies were diluted in 3% BSA/PBS. After an overnight incubation at 4°C, all sections were incubated simultaneously with Texas Red-mouse antigoat immunoglobulin (1:100) and Cy2-donkey antirabbit F(ab')₂ (1:100) (Jackson ImmunoResearch Laboratories, West Grove, PA) for 30 min at room temperature in the dark. The slides were mounted with 1 µg/ ml 4,6-diamidino-2-phenylindole (DAPI) (Sigma) in glycerol vinyl alcohol aqueous mounting solution (Zymed Laboratories, South San Francisco, CA).

Three adjacent 5-µm serial sections of 72-h recovery naphthalene-treated CCtk lung tissues were dually labeled for CGRP/ BrdU, CCSP/BrdU, and HSVtk/CCSP as described earlier. CGRP/ BrdU and CCSP/BrdU sections were pretreated with 0.05% Proteinase K and 1 N HCl. CGRP/BrdU and CCSP/BrdU sections were stained with sheep anti-BrdU (1:1,000 dilution) in combination with rabbit antirat CGRP (1:4,000 dilution) or rabbit anti-HSVtk (1:500 dilution). HSVtk/CCSP sections were stained with rabbit anti-HSVtk (1:1,000 dilution) in combination with goat antirat CCSP (1:3,000 dilution). Simultaneous detection of CCSP and BrdU, or CGRP and BrdU, in continuous BrdU-labeling control and GCV-treated CCtk mice was carried out in a similar fashion.

Stained sections were analyzed with an AX70 microscope equipped with a DAPI/Texas Red dual optical excitation filter cube and a fluorescein isothiocyanate optical excitation filter cube (Olympus, Lake Success, NY). Images were captured as described earlier. Cell counts and classification were performed on a cell-by-cell basis along the entire major axial pathway, beginning at the lobar bronchus and including terminal bronchioles where applicable. Only those cells that contacted the BM and exhibited a positive nuclear profile were counted.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The NEB Microenvironment Harbors Label-Retaining Cells and Serves as a Focal Site for Regeneration of CCSP-Expressing Cells after Naphthalene-Induced TA (Clara)-Cell Depletion

We have previously demonstrated that the NEB serves as a focal site for epithelial regeneration after naphthalene-induced ablation of airway progenitor (Clara) cells (8). Several cell types exist within this microenvironment that have potential to serve as a stem/progenitor-cell pool, all of which can be defined by their ability to express either CCSP (a Clara-cell marker), CGRP (a PNEC marker), or combinations of these antigens (8). The kinetics of cellular proliferation was determined among cells of the NEB microenvironment as an additional criterion for the identification of stem cells at this location. A low rate of cell cycling is commonly accepted as a criterion for cells with stem cell-like character (1, 11, 34, 35). We determined the rate of cell cycling through incorporation of [³H]TdR into nuclear DNA after naphthalene-induced airway injury and measurement of the rate of dilution of this incorporated label as a function of time. Using this approach, slow-cycling stem cells have been identified in other regenerative organs on the basis of their higher retention of nuclear label (11). Mice were exposed to 275 mg/kg naphthalene for depletion of Clara cells and activation of naphthalene-resistant regenerative cells within the NEB microenvironment. [³H]TdR was continuously administered to naphthalene-exposed mice for the first 7 d of the recovery and was followed by chase periods of varying duration for dilution of incorporated label. Immediately after the labeling period, foci of regenerating CCSP-IR Clara cells were found to be spatially localized to CGRP-IR NEBs, and all epithelial cells within and adjacent to NEBs exhibited extensive incorporation of [³H]TdR (Figures 1A and 1B). The distribution of nascent CCSP-IR cells was coincident with the pattern of [³H]TdR labeling at this time point. Recovery for 45 d after naphthalene exposure was associated with expansion of these regenerating regions of airway epithelium. [³H]TdR-labeled cells residing outside the NEB microenvironment included a highly labeled population of ciliated cells and CCSP-IR cells that exhibited varying degrees of label dilution (Figure 1E). Label-retaining epithelial cells within NEBs included most CGRP-IR cells and a subpopulation of CCSP-IR cells (Figures 1C and 1D). The majority of CCSP-IR cells located within the NEB showed extensive depletion of [³H]TdR (Figure 1D). No further dilution of label or variation in the distribution of label- retaining cells was observed after a 100-d chase period (data not shown). These data identify two label-retaining/ slow-cycling cell populations within the NEB microenvironment of conducting airways: CGRP-IR cells and a subpopulation of CCSP-IR cells. This observation further supports the findings from our previous study (8) that cell populations with stem cell-like character are maintained within the NEB microenvironment.

View larger version (101K): [in this window] [in a new window]   Figure 1.   Slow-cycling cells of the conducting airway epithelium. Mice were exposed to 275 mg/kg naphthalene to deplete the Clara cell population, the major progenitor cell population of the airway epithelium. Proliferating cells were continuously labeled for the first 7 d of recovery with [3H]TdR delivered through a subcutaneous miniosmotic pump. Lung tissue was recovered either immediately after the labeling period (A and B) or following 45 d of recovery from naphthalene exposure (C-E). Adjacent serial sections of lung tissue were immunostained for CGRP (A and C) or CCSP (B, D, and E) using diaminobenzene detection (brown coloration) followed by autoradiographic detection of [3H]TdR labeling. Black arrows in A and B identify boundaries of regenerating epithelium defined by [3H]TdR-labeled nuclei (dark silver grains). White arrowheads in C and D identify [3H]TdR-retaining CGRP-IR cells and CCSP-IR cells, respectively. Black arrowheads in D mark [3H]TdR-deficient CCSP-IR cells. Arrows in E identify CCSP-IR cells showing varying degrees of [3H]TdR dilution (indicated by varying numbers of overlying silver grains). Asterisks indicate locations of [3H]TdR-labeled ciliated cells. Scale bars: 10 µm (A and B); 20 µm (C-E).

Conditional and Selective Ablation of CE Cells in CCtk Mice after Acute Administration of GCV

To define the contributions of CCSP-IR (CE) and CGRP-IR (PNEC) cell populations within the NEB microenvironment to airway regeneration, we used a line of transgenic mice (CCtk) in which HSVtk coding sequences were placed under the regulatory control of the Clara cell-specific mouse CCSP promoter (9). Transgenic mice expressing the HSVtk enzyme under the control of cell type-specific promoter elements have been used to sensitize subsets of cells to GCV toxicity and thus allow dissection of cell lineage relationships in the context of organ regeneration (36). Conditional ablation of CCSP/HSVtk-expressing cells within the pulmonary epithelium of CCtk transgenic mice is effected through administration of GCV (9). To ensure that NEB-associated CE cells express the HSVtk transgene and are therefore susceptible to GCV-mediated ablation, the distribution of HSVtk-IR protein was determined among NEB-associated cells both in the steady-state lung and 72 h after naphthalene exposure. Adjacent lung-tissue sections from untreated CCtk mice were double stained for either CGRP and CCSP or HSVtk and CCSP. In the normal conducting airway epithelium of CCtk mice, CCSP-IR cells were often associated with NEBs (Figure 2A). All NEB-associated Clara cells expressed HSVtk transgene, although the staining intensity varied among these cells (Figure 2B). Foci of regenerating CCSP/HSVtk coexpressing cells were also observed in the repairing lung of CCtk mice 72 h after naphthalene exposure (Figures 2D and 2E, and data not shown), indicating that CCSP-IR cells contributing to epithelial renewal maintain expression of the HSVtk transgene. As previously demonstrated for naphthalene-exposed wild-type mice (8), two types of proliferating cells were identified within regenerating regions: CGRP-IR PNECs and CCSP-IR Clara cells (Figure 2D and 2E). Thus, the molecular characteristics of CE cells within the NEB microenvironment are consistent with a GCV-susceptible phenotype.

View larger version (37K): [in this window] [in a new window]   Figure 2.   Expression of HSVtk-IR protein among NEB-associated CCSP-expressing (CE) cells of the steady-state and injured lungs of CCtk transgenic mice. Adjacent serial sections of untreated CCtk mouse lung were labeled by dual immunofluorescence either for CGRP (Texas Red; red fluorescence) and CCSP (Cy2; green fluorescence) (A) or for CCSP (green fluorescence) and HSVtk (red fluorescence) (B) and DAPI (blue fluorescence) nuclear counterstain. Flattened and morphologically distinct CE cells are shown to be associated with a NEB (A). All CE cells, including those in direct contact with CGRP-IR cells, are shown to coexpress HSVtk, evident by the orange-yellow color of the overlaid image (B). C-F show immunofluorescent staining of mouse airways 72 h after naphthalene exposure. Adjacent serial sections were stained for either CGRP (red fluorescence)/BrdU (green fluorescence) (D), CCSP (red fluorescence)/BrdU (green fluorescence) (E), or HSVtk (red fluorescence)/CCSP (green fluorescence) (F) with DAPI (blue fluorescence) nuclear counterstain. C shows a phase-contrast image of an airway region (black box) shown in D-F. BrdU-labeled (green nuclei) epithelial cells were spatially colocalized to NEBs after naphthalene administration and included both CGRP-IR (D) and CCSP-IR (E) subsets. Regenerating CCSP-IR cells also express HSVtk-IR protein (F). Scale bars: 10 µm (A, B, D-F ); 20 µm (C).

To determine the contribution of CE cells and PNECs to airway renewal, repair was evaluated after GCV-mediated ablation of CCSP/HSVtk-expressing cells in CCtk transgenic mice. We have previously demonstrated that chronic exposure of CCtk transgenic mice to GCV results in selective ablation of CE cells and PNEC proliferation. However, the differentiation potential of PNECs could not be evaluated in these experiments due to the continual ablation of CE cells by chronic administration of GCV and the possibility that nascent CE cells represented a critical TA cell population to effect airway renewal (9). To overcome this problem, CCtk transgenic mice were acutely exposed to GCV followed by a recovery period, allowing assessment of the differentiation potential of residual regenerative cells. Adult male CCtk transgenic mice were exposed to 10 mg GCV over 24 h, followed by recovery for 0, 1, 3, 6, or 10 d. Total lung RNA was subjected to S1 nuclease protection analysis for the quantification of cell-specific marker gene expression. The abundance of CCSP and CYP2F2 (Clara cell-specific markers) mRNAs in lungs of GCV-exposed CCtk mice were reduced to 53.6 and 31.0% of the untreated control mRNA levels, respectively, immediately after GCV exposure (Day 0 of recovery) (Figure 3). Expression of CCSP and CYP2F2 mRNAs continuously decreased throughout the recovery time points and was 5.7 and 8.5% of the control levels, respectively, at the 10-d recovery time point (Figure 3). This level of Clara-cell depletion was similar to that achieved after exposure to > 200 mg/kg naphthalene (17).

View larger version (58K): [in this window] [in a new window]   Figure 3.   Decreases in Clara cell-specific mRNA expression among GCV-exposed CCtk mice. CCtk male mice were exposed to 10 mg GCV over a 24-h period and allowed to recover for 0, 1, 3, 6, or 10 d. Total lung RNA samples isolated from untreated control CCtk mice and GCV-treated CCtk mice were analyzed by S1 nuclease protection assay to measure changes in the abundance of Clara cell-specific markers (CCSP, CYP2F2). An autoradiogram showing analysis of representative samples from each group appears in A, with time of recovery in days indicated above the autoradiogram and protected probe fragments corresponding to CCSP, CYP2F2, and the L32 internal control mRNA indicated on the left. mRNA abundance was quantified using a PhosphorImager and normalized to that for L32 mRNA (B). The y-axis shows relative mRNA abundance in arbitrary units; x-axis shows the days of recovery following GCV exposure. Bar graphs are presented as means ± SEM (n = 4; for Day 10, n = 3) *P < 0.05 and **P < 0.01 relative to levels in untreated control mice.

The spatial context and cellular specificity of GCV-mediated airway epithelial cell depletion was further investigated by immunohistochemical analysis. Adjacent lung tissue sections from animals in each group were immunostained to determine the distribution and abundance of CCSP- and ACT (ciliated cell-specific)-IR proteins. As shown in Figure 4, CCSP-IR Clara cells in the conducting airway epithelium were reduced immediately after GCV exposure and continued to decrease throughout the recovery period (Figures 4A-4D). CCSP-IR cells were rarely detected within the airway epithelium of CCtk mice 10 d after GCV exposure (Figure 4D), which was consistent with the level of Clara cell ablation (approximately 94%) suggested by the marked decrease in relative abundance of Clara cell-specific mRNAs (Figure 3). The magnitude and pattern of Clara cell ablation observed after acute administration of GCV was similar to that observed after chronic GCV treatment (9). The decline in CCSP-IR cells was accompanied by a decrease in airway epithelial cell density (Figure 4D). However, in contrast to the dramatic decline in abundance of CCSP-IR cells, the abundance of ACT-IR ciliated cells of the proximal bronchiolar epithelium was not affected by GCV treatment (Figures 4E-4H). These results demonstrate the specificity of Clara cell ablation in bronchiolar airways of CCtk transgenic mice after acute administration of GCV. These data suggest that CE cells continued to be ablated during the recovery period after acute exposure to GCV and that airway repair was compromised after depletion of CCSP/HSVtk-expressing cells. Failure of airways to restore epithelial integrity resulted in the progressive decline of lung function among GCV-exposed CCtk mice, leading to death as early as 12 d after exposure (Reynolds and colleagues, unpublished observations). As such, it was not possible to assess epithelial renewal after prolonged recovery. Moreover, efficient depletion of HSVtk-expressing Clara cells by acute exposure to GCV in CCtk mice does not require cell proliferation, a finding that is consistent with the proliferation-independent mechanism of GCV-mediated cell ablation that has been proposed by Wallace and coworkers (40). The kinetics and duration of Clara/CE-cell ablation observed in the present study may reflect such a proliferation-independent mechanism of GCV-mediated ablation.

View larger version (53K): [in this window] [in a new window]   Figure 4.   Selective ablation of CE cells among GCV-exposed CCtk mice. The abundance and distribution of Clara/CE and ciliated cells were determined within airways of control and GCV-exposed CCtk mice by immunohistochemical detection of CCSP and ACT, respectively. Adjacent serial sections were immunostained for CCSP (A-D) and ACT (E-H ) and counterstained with hematoxylin. Untreated control (A and E); 3-d recovery (B and F ); 6-d recovery (C and G); and 10-d recovery (D and H ). Photomicrographs are representative images taken at proximal bronchiolar regions within conducting airways. Scale bars: 20 µm.

PNEC Hyperplasia without Regenerating Foci of CE Cells during Recovery from GCV Exposure

Early events associated with repair of naphthalene-induced injury to airway epithelium are characterized by PNEC proliferation and hyperplasia (8, 18, 41), coupled with the appearance of distinct foci of regenerating CE cells that localize to NEBs (8, 17) (Figure 2). In addition, identification of CCSP and CGRP dual IR cells in NEB microenvironments suggested that PNECs may serve as a pluripotent progenitor cell population capable of regenerating Clara cells (8). To test this hypothesis we determined the extent of PNEC hyperplasia and the pattern of Clara cell regeneration associated with recovery of CCtk mice acutely exposed to GCV. Adjacent serial sections of lungs from GCV-treated CCtk mice were immunostained for CGRP and CCSP to identify PNECs/NEBs and to determine whether these regions included regenerating Clara cells. As ablation of Clara cells progressed, a trend toward increasing numbers of total PNECs and NEBs and the average size of NEBs was apparent and reached significance at the 10-d time point (Figures 5 and 6). In addition, instantaneous [³H]TdR labeling of GCV-exposed animals at each recovery time point indicated an increased labeling index of PNECs starting at Day 3 of recovery (data not shown). Hyperplastic NEBs, containing 10 to 20 PNEC nuclei within the plane of the section, were frequently observed among GCV-exposed CCtk mice at both 6- and 10-d recovery time points (Figures 5B and 5C). In contrast, NEBs of control untreated lungs generally included no more than five to eight PNEC nuclei within the plane of the section (Figure 5A and data not shown). Morphometry demonstrated that by 10 d of recovery, GCV-exposed CCtk mice exhibited a 3.2-fold increase in CGRP-IR cell (PNEC) nuclei per millimeter of BM (Figure 6A) and a 3.4-fold increase in the percentage of BM subtending CGRP-IR regions (Figure 6B), compared with untreated controls. In addition, there was an approximate 1.5-fold increase in the average number of PNECs per NEB (data not shown). These results indicate that PNEC hyperplasia is closely associated with TA (Clara)-cell depletion and protracted airway regeneration. This association between severe Clara-cell injury and development of PNEC hyperplasia was consistent with previous results from the naphthalene-induced Clara-cell injury model, in which significant PNEC hyperplasia was demonstrated after 5 d of recovery (8, 18, 41).

View larger version (44K): [in this window] [in a new window]   Figure 5.   Development of PNEC hyperplasia in response to Clara/ CE cell ablation and absence of NEB-associated Clara cell regeneration. The abundance and distribution of PNECs and Clara cells were determined within airways of control and GCV-exposed CCtk mice by immunohistochemical detection of CGRP and CCSP, respectively. Adjacent serial sections were immunostained for CGRP (A-C) and CCSP (D-F ). Control (A and D); 6-d recovery (B and E); and 10-d recovery (C and F ). The total number of PNECs and proportion of BM subtending CGRP-IR regions increased with time of recovery (B and C). In contrast to naphthalene-induced injury, no distinct foci of regenerating Clara cells were observed around hyperplastic NEBs at Day 10 (F ). Scale bars: 20 µm.

View larger version (18K): [in this window] [in a new window]   Figure 6.   Morphometry of PNEC hyperplasia during recovery from GCV-induced Clara/CE cell ablation. Increased numbers of CGRP-IR PNECs are presented as either CGRP-IR cells per millimeter of BM (#PNECs/mm BM; A) or as percentage of total airway BM (% BM subtended by CGRP-IR regions; B). Results are presented as means ± SEM for each treatment group. *P < 0.05 with respect to the untreated control.

Despite similarities in the development of PNEC hyperplasia between naphthalene- and GCV-induced Clara-cell depletion models, no regenerating foci of CCSP-IR cells were identified within the NEB microenvironment of GCV-exposed CCtk mice (Figures 5 and 7). At the 10-d recovery time point, residual CCSP-IR cells were randomly scattered throughout the airway and rarely found to be spatially associated with NEBs (Figures 4D, 5C and 5F, 7D and 7F). This observation further supports our previous conclusion that PNECs function as a self-renewing cell population. However, these data indicate that PNECs are unable to differentiate into CE cells despite the fact that they were induced to proliferate after CE cell depletion.

View larger version (33K): [in this window] [in a new window]   Figure 7.   Airway epithelial cell proliferation either in steady-state CCtk mice or after GCV-induced CE cell ablation in CCtk mice. GCV-exposed CCtk mice or corresponding untreated controls were continuously labeled with BrdU from Days 2 to 10 of recovery (8 d). Adjacent serial sections from control (A-C) and treated (D- F ) lung tissues were immunostained for CGRP (A and D), BrdU (B and E), and CCSP (C and F ). Whereas BrdU-labeled nuclei in the control conducting airway epithelium appear randomly distributed throughout the airway (B), those in the treated tissues appear in clusters and localized to the NEBs (E). CCSP-IR cells are not associated with foci of BrdU-labeled epithelial cells 10 d after GCV exposure (F ). Scale bars: 10 µm.

To determine whether residual CCSP-IR cells observed within airways of GCV-exposed CCtk mice were generated through proliferation and differentiation of another progenitor cell pool, CCtk mice (controls and GCV-exposed) were continuously labeled with BrdU to label cells passing through S-phase of the cell cycle. BrdU labeling was performed either for an 8-d period in control mice or during the 2- to 10-d recovery period in mice previously exposed to GCV. BrdU labeling within the conducting airway epithelium of control mice showed a significant colocalization with CCSP-IR cells but infrequent labeling of CGRP-IR cells (Figures 8A and 8C and Table 1). These BrdU/CCSP dual-positive cells appeared to be randomly distributed throughout airways with no specific association with NEBs (Figures 7A-7C). CCSP-IR cells represented 82.2% of BrdU-labeled cells, with 8.3% of all Clara cells showing BrdU labeling after the 8-d exposure period (Table 1). In contrast, mice recovering from GCV exposure showed a pattern of BrdU labeling that was highly restricted and closely associated with CGRP-IR cells organized into NEBs (Figures 7D, 7E, and 8D). All NEBs examined contained at least one cell with a BrdU-labeled nucleus (Figure 8D and data not shown). CGRP-IR cells accounted for 45.8% of the total BrdU-labeled nuclei of the conducting airway epithelium, and 39.1% of total CGRP-IR cells were labeled with BrdU. In contrast, only 4.1% of CGRP-IR cells of the control tissue were labeled with BrdU (Table 1). Residual CCSP-IR cells detected at Day 10 of recovery were infrequently labeled with BrdU (Figure 8B). Moreover, these residual CCSP-IR cells showed no spatial association with NEBs harboring abundant BrdU-labeled CGRP-IR cells (Figures 7D-7F). As such, residual CCSP-IR cells present at Day 10 of recovery are not a result of active regeneration or differentiation of potential progenitor pools. Although a small percentage of residual CCSP-IR cells showed BrdU labeling, this was observed only among two of the four GCV-exposed mice at the 10-d recovery time point (Table 1). We attribute this finding to incomplete Clara-cell ablation that may result from inter-individual variability in GCV susceptibility. These data demonstrate that CE cell-depleted airway epithelium of GCV-exposed CCtk mice failed to effect regeneration of CE cells and restore functional airway epithelium within the time frame of recovery.

View larger version (36K): [in this window] [in a new window]   Figure 8.   Identification of BrdU-labeled cells in control and GCV-treated tissues after continuous BrdU labeling. BrdU (green fluorescence) labeling within epithelial cells lining conducting airways of control tissues was observed predominantly within CCSP-IR cells (red fluorescence; A) and not among CGRP-IR cells (red fluorescence; C). Mice treated with GCV and recovered for 10 d exhibited PNEC hyperplasia with CGRP-IR cells (red fluorescence) showing extensive BrdU (green fluorescence) labeling (D). In contrast, none of the few remaining CCSP-IR cells (red fluorescence) in the GCV-treated animals showed BrdU (green fluorescence) labeling at the 10-d recovery time point (B). Scale bars: 10 µm.

An interesting finding of the present study was the observation that CGRP-IR cells account for only approximately 46% of the proliferative fraction of airway epithelial cells after elimination of CE cells from the regenerative pool. The remaining BrdU-labeled epithelial cells were located predominantly within proximal airways, with a distribution that was distinct from NEB-associated proliferation observed within the bronchiolar epithelium. The morphology and distribution of BrdU-labeled cells within proximal airways of GCV-exposed CCtk transgenic mice are consistent with those of basal cells. However, additional studies are needed to validate this theory and fully investigate regenerative properties at this airway location.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Despite the identification of progenitor cell populations that contribute to maintenance and renewal of the pulmonary epithelium, the existence and identity of airway stem cells remains elusive (10). We have previously identified the NEB microenvironment at airway branchpoints as a site of epithelial regeneration after naphthalene-induced TA (Clara)-cell ablation (8). Clara cell depletion results in the activation of two proliferative cell populations located within the NEB microenvironment, CGRP-IR PNECs and CCSP-IR Clara cells (8). This mode of TA cell depletion is similar to that used to reveal the identity and location of stem cells within other regenerative organs (11, 12). In the present study, using retention of nuclear [³H]TdR as an indication of infrequent cell cycling, we demonstrate that the NEB harbors slow-cycling populations of epithelial cells that are candidate stem cells. On the basis of these observations and earlier studies identifying proliferative CGRP-IR and CCSP-IR cells within the NEB microenvironment, experiments were designed to investigate the alternate hypotheses that either PNECs or a population of naphthalene-resistant, NEB-associated Clara (vCE) cells serve as pluripotent regenerative cells after ablation of naphthalene-sensitive progenitor/TA cells.

Transgenic mice expressing HSVtk under the regulatory control of the CCSP promoter (CCtk) (9) were used to investigate the contribution that CE cells and PNECs make to epithelial renewal. Our previous study using this model in combination with chronic administration of GCV to achieve Clara cell ablation revealed a self-renewing progenitor function for PNECs, yet was unable to determine their capacity for pluripotent differentiation through a CE intermediate (9). In the present study we used acute exposure of CCtk transgenic mice to GCV to reveal the regenerative capacity of progenitor cells in the absence of CE cells. Acute exposure of CCtk mice to GCV resulted in selective ablation of CCSP/HSVtk-expressing cells, an associated increase in the proliferative fraction of PNECs, and PNEC hyperplasia. However, continual decrease in the number of CE cells during recovery and the absence of labeling in the residual CE cells detected at the latest recovery time point after continuous BrdU administration indicate that airways depleted of CE cells were unable to effectively regenerate a normal epithelium. Moreover, after GCV-mediated CE cell depletion, regenerating foci of CE cells were not observed around NEBs despite the increased proliferation of PNECs and subsequent development of hyperplasia. This observation is in contrast to the focal pattern of NEB-associated epithelial regeneration observed after naphthalene-induced Clara-cell injury (8, 17). Findings of this study demonstrate that PNECs are not themselves capable of effectively renewing CE cell- depleted airway epithelium and that naphthalene-resistant CE (vCE) cells of the NEB microenvironment serve a critical function in airway renewal after TA cell depletion.

In the label-retention experiment of the present study, most of the CE cells located within NEBs showed extensive depletion of label (Figure 1D). However, CE cells located outside of the NEB microenvironment and within the regenerated epithelium had diluted nuclear label to a lesser extent (Figure 1E). These findings suggest that CE TA cells of the adult repairing lung have a higher proliferative rate in close proximity to NEBs than do those located more distant from the NEB. Similar findings have been demonstrated in the developing lung, in which epithelial cell proliferation decreased as a function of increasing distance from the NEB (42, 43), suggesting a mitogenic influence of NEB-derived factors that decreases in potency with increasing distance from the NEB. Detection of highly labeled ciliated cells within the regenerative portion of the airway epithelium suggests that proliferation of NEB-associated progenitor cells early in the response to napthalene-induced injury is followed by epithelial cell migration and differentiation. Evans and colleagues (6, 14) demonstrated that ciliated cells are generated through proliferation and subsequent differentiation of Clara cells, but that ciliated cells do not themselves have proliferative capacity. The finding of label retention among ciliated cells located at the distal boundary of the regenerating epithelium highlights the importance of considering proliferative capacity in addition to label retention as a criterion for defining cells with stem cell-like character.

CE/Clara cells appear to play an indispensable role in maintenance of airway epithelium as indicated by the absence of repair after GCV-mediated ablation of CE cells in CCtk mice. The contrast in regenerative capacity of the conducting airway epithelium after ablation of all CE cells in the present study versus naphthalene-sensitive Clara-cells described previously (8) highlights the importance of naphthalene-resistant CE (vCE) cells of the NEB microenvironment to effect epithelial renewal. Findings from the present study are therefore consistent with the notion that the NEB microenvironment is multifunctional, serving to maintain slow-cycling epithelial cells in the steady-state epithelium and to stimulate the proliferation of TA cells either after airway injury or during airway development (Figure 9). Mason and colleagues postulated that specific microenvironments may be present within the pulmonary epithelium that maintain distinct subsets of regenerative cells with stem cell-like properties (10). Although of nebulous character, stem-cell niches are thought to comprise a unique microenvironment capable of maintaining the undifferentiated phenotype and pluripotent differentiation potential of stem cells (1, 4, 44, 45). Characteristics of stem-cell niches vary among those tissues in which they have been identified, with the critical characteristic being attributed to location within the tissue, cellular interactions, and cell/matrix interactions (11, 46). The NEB microenvironment may represent an analogous structure within the conducting airway epithelium for maintenance of an airway stem-cell pool. It may influence the phenotype of CE cells, blocking Clara to ciliated cell differentiation and preserving a population of regenerative cells (i.e., vCE cells) that can contribute to epithelial renewal after exposure to Clara cell toxicants. However, due to the precedent for cellular and paracrine interactions as key determinants of stem cell maintenance within other regenerative systems, our finding that CE cells are required for NEB-associated epithelial regeneration does not unequivocally identify CE cells as the stem cell. CE cells may either represent a stem cell pool or function as an accessory cell that is required for the appropriate differentiation of proliferating CGRP-IR cells.

View larger version (64K): [in this window] [in a new window]   Figure 9.   Model describing proposed cellular interactions within and adjacent to the NEB microenvironment allowing epithelial renewal in the steady state and after progenitor/TA cell depletion. In the steady state and after depletion of terminally differentiated cells (e.g., ciliated cells), airway epithelium is maintained through proliferation of Clara (CE) cells that are capable of self-renewal (curved red arrows) and in most cases, differentiation into ciliated cells (green arrow) (6, 14). CE cells of the steady-state epithelium also fulfill differentiated functions, including synthesis and secretion of lumenal components, and metabolism of both endogenous and xenobiotic compounds. A variant subpopulation of CE (vCE) cells is specifically maintained within the NEB microenvironment (8). vCE cells are both deficient in xenobiotic metabolizing enzymes and resistant to injury by the Clara cell toxicant naphthalene. Moreover, vCE cells and NEB-associated CE cells share unique cell-cell interactions with PNECs, a normally quiescent epithelial cell population of airways, and include a subset of slow-cycling cells (potentially overlapping with the vCE subset). Both PNECs and vCE cells have proliferative potential (curved yellow arrows) and are activated after depletion of naphthalene-sensitive CE cells (Ref. 8 and Figure 2). However, PNECs by themselves are not capable of effectively renewing CE cell-depleted airway epithelium. vCE cells are required for airway repair, as is evident from contrasting patterns of NEB-associated epithelial renewal observed between naphthalene exposures, in which vCE cells are specifically preserved allowing for airway regeneration (Ref. 8 and Figure 2), and GCV-mediated ablation of both naphthalene-sensitive and vCE cells in CCtk mice, in which case airway renewal is compromised (see Figures 4 and 5). The NEB microenvironment serves to maintain the undifferentiated properties of vCE cells in the steady-state lung, thus preserving a critical pool of regenerative cells for airway repair after depletion of pollutant-sensitive CE cells. Our previous study demonstrating the existence of NEB-associated cells capable of coexpressing both Clara and neuroendocrine markers (8) also suggests a potential for CE or vCE cells to PNEC differentiation, a property of CE cells that was not addressed in the present study.

Properties of the NEB microenvironment that are uniquely suited for the maintenance of vCE cells are currently unclear. It has been suggested that PNEC-derived paracrine factors might play a role in regulation of epithelial cell differentiation and proliferation during fetal lung development and possibly in the normal or injured adult lung (50). The appearance of CE cells during human fetal lung development shows a close association with NEBs (50, 51) and is recapitulated during repair from naphthalene-induced acute Clara-cell depletion (8). These data argue that either cell-cell interactions or paracrine factors unique to the NEB microenvironment may play a role in sequestration of cells with stem cell-like properties during lung development and provide a permissive environment for their maintenance through adulthood. Further studies are needed to define essential components of the NEB microenvironment that are permissive for maintenance of this putative stem-cell pool and to both identify and characterize cells maintained therein. Development of methods for either the culture or transplantation of vCE cells and other cells of the NEB microenvironment will aid in their classification.