The composition of the conducting airway epithelium varies significantly along the proximal to distal axis, with that of the tracheal epithelium exhibiting the greatest complexity. A number of progenitor cells have been proposed to contribute to the maintenance of this cellular diversity both in the steady state and in response to injury. However, individual roles for each progenitor cell type are poorly defined in vivo. The present study was undertaken to investigate the hypothesis that basal cells represent a multipotent progenitor cell type for renewal of the injured tracheal epithelium. To understand their contribution to epithelial repair, mice were exposed to naphthalene to induce airway injury and depletion of the secretory cell progenitor pool. Injury resulted in a rapid induction of cytokeratin 14 (K14) expression among the majority of GSI-B4-reactive cells and associated hyperplasia of basal cells. Restoration of depleted secretory cells occurred after 6 days of recovery and was associated with regression of the basal cell hyperplasia, suggesting a progenitor-progeny relationship. Multipotent differentiation of basal cells was confirmed using a bitransgenic ligand-regulated Cre-loxP reporter approach in which expression of a ubiquitously expressed LacZ reporter was activated within K14-expressing progenitor cells during airway repair. With the use of this approach, it was determined that K14-expressing cells include subsets capable of either multipotent or unipotent differentiation in vivo. We conclude that basal cells have the capacity for restoration of a fully differentiated epithelium.
- Cre recombinase
the conducting airway is a highly dynamic structure, both in its cellular composition and function, with deviations from the steady state contributing significantly to deterioration in lung function typical of patients with chronic lung disease (12). Changes in the cellular composition of the surface epithelium can be a direct consequence of ongoing injury and repair, or may result from alterations in the differentiation potential of epithelial cells driven by the aberrant cytokine and growth factor milieu (4, 9, 25, 27, 28). One approach used to understand mechanisms of epithelial remodeling has been to define progenitor and/or stem cell populations capable of contributing to the repair process and how their differentiation potential is modified by paracrine factors.
Progenitor cells with the potential to contribute to airway repair include secretory, neuroendocrine, and basal cell populations (2, 8, 15–17, 19, 20). The individual contributions made by these progenitor cells toward epithelial maintenance and repair following injury are not fully appreciated. There is a significant body of data suggesting that the selectivity of injury toward specific airway epithelial cell types is an important determinant of progenitor cell activation. Evans and colleagues (7, 8) demonstrated that secretory cells of both the proximal and distal airway epithelium of rats function as the principal progenitor cell for replenishment of the conducting airway epithelium following oxidant-induced injury to terminally differentiated ciliated cells. Even though cells with morphological characteristics of basal cells were shown to represent a progenitor population in these studies, their contribution to the proliferative fraction was lower in the setting of airway injury than in the steady state, raising questions as to their role in restoration of columnar epithelial cells (8). In contrast, naphthalene-induced ablation of secretory cells was accompanied by proliferation of three principal cell types, neuroendocrine cells, a pollutant-resistant subpopulation of Clara cell secretory protein (CCSP)-expressing cells associated with either the neuroepithelial body (NEB) or the bronchoalveolar duct junction, and a population of bronchial Griffonia simplicifolia isolectin (GSI)-B4-reactive, cytokeratin 14 (K14)-negative cells that rapidly assumed a K14-expressing phenotype following injury (11, 13, 19, 23). Even though NEB-associated pulmonary neuroendocrine cells (PNEC) were shown to represent a progenitor population that was activated following secretory cell depletion, their differentiation potential was limited to self-amplification, a property that results in PNEC hyperplasia but not epithelial regeneration (19, 20, 23). In contrast, cells of the bronchial epithelium that assumed a K14-expressing phenotype following injury were shown to exhibit multipotent differentiation potential (13a).
Approaches to investigate the differentiation potential of airway epithelial progenitor cells ex vivo have principally involved seeding of isolated cells into denuded tracheas before their implantation into nude mice (5, 6, 14, 15, 18). With this approach Engelhardt and colleagues (6) were able to demonstrate that individual human bronchial epithelial cells exhibited high clonogenic capacity with multipotent differentiation potential. However, the identity of putative stem or progenitor cells tagged using this approach could not be directly determined (6). The finding that isolated epithelial cell populations enriched for either basal or secretory cells are able to reconstitute a fully differentiated airway epithelium within tracheal xenografts suggests that either of these populations may be multipotent. A caveat with these studies is that small numbers of cells with high clonogenic capacity and multipotent differentiation, such as those identified by Engelhardt and colleagues (5, 6), could account for most, if not all, of the regenerative capacity of isolated cell preparations, making it impossible to rule out contributions made by contaminating cell types to establishment of the engrafted epithelium.
The purpose of the present study was to define roles for basal cells in renewal of the tracheal epithelium through investigation of the hypothesis that basal cells represent a multipotent progenitor for epithelial renewal. We determined that a significant subpopulation of basal cells within the steady-state tracheal epithelium express K14, a property that distinguished them from bronchial basal cells (13a). Naphthalene treatment was used for ablation of secretory cells and recruitment of nonsecretory cell progenitors to the regenerative pool. Secretory cell ablation led to a dramatic increase in the abundance of GSI-B4-reactive/K14-immunoreactive (IR) basal cells with evidence of differentiation into secretory cell populations with extended recovery. To unambiguously define the differentiation and clonogenic potential of this population, LacZ-expressing lineage tags were activated through Cre-mediated recombination within K14-expressing cells, and the cellular composition of tagged cells was determined after prolonged recovery. Some tagged clusters of cells included basal, ciliated, and secretory cell types, demonstrating the multipotent differentiation potential of K14-expressing basal cells. Other tagged clusters included only basal and, in some cases, only ciliated and secretory cell subsets. The implications of these findings are discussed.
MATERIALS AND METHODS
Animals and naphthalene treatment. All procedures involving animal use in this study were reviewed and approved by the University of Pittsburgh animal use and care committee. Female FVB/n mice used in this study were 2–4 mo old and were maintained as a specific pathogen-free, in-house colony. Naphthalene treatment was carried out as previously described (24). Each animal received 275 mg of naphthalene/kg body wt (ip). Mice were recovered in filtered air for 1.5, 3, 6, 9, 21, and 43 days. At least four mice were used for each recovery time point.
Tissue collection and immunohistochemical analysis. Control (untreated) and naphthalene-treated animals were killed by injection of 100 mg of pentobarbital sodium/kg (ip) followed by exsanguination. Trachea and tongue of each mouse were immersion fixed in 10% neutral buffered formalin overnight at 4°C and stored in PBS at 4°C. Tissues were then embedded in paraffin, sectioned in 5-μm thickness, and mounted on superfrost microscope slides (VWR, West Chester, PA).
Immunohistochemical detection of CCSP and K14 and histochemical detection of GSI-B4 binding sites have been previously described (13a). GSI-B4 is a plant-derived protein that shows specificity toward glycoconjugates present on the surface of basal cells of various mouse and rat epithelia, including tracheobronchial epithelium (10, 21). Images of representative fields were acquired using an Olympus Provis AX70 microscope (Olympus, Lake Success, NY) equipped with a Spot RT color digital camera (Diagnostic Instruments, Sterling Heights, MI) linked to a PC running Image-Pro Plus (Media Cybernetics, Silver Spring, MD).
K14/recombination substrate bitransgenic mice and treatments. Transgenes used in this study and the strategy for the introduction of lineage tags within K14-expressing cells are shown in Fig. 1. A transgene composed of the human K14 promoter linked to coding sequences for a fusion protein between bacteriophage P1 Cre recombinase and a mutated estrogen receptor ligand binding region (Cre-ERt) has been previously described (26) and was used to generate transgenic mouse lines in the FVB/n background. Mice harboring a Cre recombination substrate (RS) allele were generated previously by Soriano (22) and contain a loxP-neo-loxP-LacZ cassette inserted downstream from the endogenous ROSA26 promoter. K14/RS mice were treated with naphthalene as described above on day 0 and with 4 mg of Tam in corn oil ip on recovery days 2, 3, and 4. Treated mice were killed on days 4, 6, 21, or 43. Tissues of treated RS mice or untreated K14/RS mice served as negative controls.
β-Galactosidase histochemistry. Trachea, esophagus, and tongue were harvested from the naphthalene- and Tam-treated K14/RS and RS mice. Tissues were fixed in 4% paraformaldehyde for 20 min at room temperature and stored in PBS at 4°C. Tissues were incubated in X-gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) staining solution [5 mM K3Fe(CN)3, 5 mM K4Fe(CN)6, 2 mM MgCl2, 0.02% Nonidet P-40, 0.01% sodium deoxycholate, 1× PBS, and 1 mg/ml X-gal] for 4–5 h at 37°C in the dark. After histochemical analysis, tissues were washed in PBS and postfixed in 4% paraformaldehyde at 4°C overnight, embedded in paraffin, and sectioned at 5 μm. Representative sections containing X-gal-stained cells were counterstained with Nuclear Fast Red (Vector Laboratories, Burlingame, CA) for morphological analysis or immunohistochemically for analysis of cellular phenotype. Sections containing blue tracheal epithelial cells were deparaffinized, hydrated, coverslipped in PBS, and photographed. Sections were then immunostained for CCSP, acetylated tubulin, GSI-B4 binding sites, or K14 using a horseradish peroxidase (HRP)-3,3′-diaminobenzidine (DAB) system as described previously (13a). Phenotypes of the X-gal-stained epithelial cells were determined by photographing regions of the section containing X-galstained cells and comparing them to images taken before immunohistochemical staining.
Repair of the tracheal epithelium after naphthalene-induced Clara cell injury. Injury and repair of the tracheal epithelium after naphthalene-mediated secretory cell depletion was assessed by immunohistochemical detection of the Clara cell marker CCSP and the basal cell markers GSI-B4 or K14. Untreated mice were used as controls. Clara cells were numerous within tracheas of control mice and were distributed throughout the epithelium (Fig. 2A). GSI-B4-reactive basal cells formed a continuous layer in the lower part of the pseudostratified epithelium and had no access to the airway lumen (Fig. 2B). K14-immunoreactive (IR) cells were a subpopulation of GSI-B4-positive cells (Fig. 2, B and C), were usually found in small clusters, and were often located above cartilaginous rings (Fig. 2C). Naphthalene treatment of mice resulted in a dramatic decrease in CCSP-IR cells. By day 1.5 of recovery, nearly all Clara cells had exfoliated, a process accompanied by a marked decrease in the overall epithelial cell density (Fig. 2D). In contrast, there was no noticeable change in the number or distribution of basal cells at this time point (Fig. 2, E and F). By day 3 of recovery, hyperplastic basal cells were evident, as was an increase in the overall epithelial cell density (Fig. 2H). At this time point, CCSP-IR cells were still absent from the epithelium (Fig. 2G). Basal cell hyperplasia observed between recovery days 1.5 and 3 was accompanied by expression of K14 in virtually all GSI-B4-reactive basal cells (Fig. 2I). Restoration of a CCSP-IR population and reestablishment of a pseudostratified epithelium was evident on days 6 and 9 of recovery and was associated with regression of hyperplastic K14-IR and GSI-B4-reactive cells (Fig. 2, J–L). At this time point, a number of columnar epithelial cells that clearly lacked morphological features of basal cells were found to be K14-IR (Fig. 2L). The epithelium returned to its normal composition by recovery day 21 (Fig. 2, M–O). No further changes in epithelial morphology were observed between recovery days 21 and 42.
In vivo lineage tagging of basal cells of the tracheal epithelium. The differentiation potential of tracheal basal cells was determined using a transgenic approach for in vivo lineage tagging of K14-expressing cells. Adult K14/RS bitransgenic mice were treated with 275 mg of naphthalene/kg body wt to induce Clara cell injury, and with 4 mg of Tam on recovery days 2–4 to activate CreERt recombinase and recombination of the RS allele. Mice were killed on day 4, 6, 21, or 43 of recovery. Various tissues, including trachea, tongue, and esophagus from each animal, were fixed in 4% paraformaldehyde and stained as whole mounts for β-gal-expressing cells. Tongue and esophagus exhibited a high frequency of recombination within the basal cell layer at all time points (data not shown) and served as positive controls for Tam administration and X-gal staining. No recombination was observed in the tracheal epithelium of K14/RS mice in the absence of Tam (data not shown). Specificity of CreERt expression was determined by combining X-gal staining with immunohistochemistry and demonstrated that all β-gal-positive cells examined at day 4 of recovery exhibited a basal cell-like morphology and were immunoreactive for K14 (Fig. 3). These data demonstrate that recombinational activation of β-gal expression occurred selectively within K14-expressing basal cells and was dependent on Tam administration.
Basal cells are a multipotent progenitor of the tracheal epithelium. Tracheas of naphthalene- and Tam-treated K14/RS mice from days 6, 21, and 43 of recovery were examined after whole mount X-gal staining. On day 6, groups of β-gal-positive cells were visible throughout the tracheal epithelium and varied in size from 2 to 15 cells (Fig. 4A). β-gal-Positive cells were usually tightly associated with one another, suggestive of a clonal expansion of a tagged cell (Fig. 4D). As the epithelium returned to its normal composition on day 21, a general increase in the size of the β-gal-positive groups was noted (Fig. 4B). By day 43, there was a further enlargement of these groups, some of which contained hundreds of tightly clustered β-gal-positive cells, whereas others appeared to be mixtures of β-gal-positive and -negative cells (Fig. 4, C and F). These data indicate that K14-expressing basal cells, which had been labeled during Tam administration, had undergone rounds of proliferation and differentiation during epithelial repair, resulting in expansion of the β-gal-positive groups within the epithelium. The existence of large groups of LacZ-positive cells that were uninterrupted by β-gal-negative cells suggests that the initially tagged precursor cell(s) exhibit significant self-renewal capacity (Fig. 4, C and F).
The phenotype of cells present in each β-gal-positive group was determined on day 43 using a combination of histochemical and immunohistochemical techniques. Morphological analysis demonstrated that the majority of β-gal-positive cell clusters harbored multiple cell types, including basal cells and columnar epithelial cells (Fig. 5A). However, some β-gal-positive groups containing only basal cells were observed (Fig. 5A, bottom right). GSI-B4 lectin histochemistry revealed that the β-gal-positive clusters included both GSI-B4-reactive basal and GSI-B4-negative nonbasal/columnar cell phenotypes (Fig. 5B). Similar analysis of basal (K14), ciliated (acetylated tubulin), or secretory (CCSP) cell markers confirmed that K14-expressing cells tagged during repair from naphthalene-induced injury give rise to all cell types common to the tracheal epithelium (Fig. 5C). Together, these results demonstrate that K14-expressing basal cells are induced to proliferate and exhibit multipotent differentiation potential after secretory cell depletion.
The present study used in vivo progenitor cell depletion coupled with lineage tagging to investigate contributions made by basal cells to repair of the injured tracheal epithelium. Naphthalene-induced Clara cell ablation was accompanied by basal cell proliferation and reestablishment of a normal airway epithelium. Basal cell proliferation resulted in hyperplasia of K14-expressing cells, regression of which was accompanied by the appearance of columnar secretory and ciliated cells suggestive of a progenitor-progeny relationship. Lineage tagging using a ubiquitously expressed LacZ reporter revealed that K14-expressing cells exhibited significant clonogenic capacity with either unipotent or multipotent differentiation potential. Tagged clusters exhibiting multipotent differentiation encompassed all cell types normally resident within the steady-state tracheal epithelium including basal, ciliated, and secretory cells. In contrast, tagged clusters of cells showing evidence of unipotent differentiation were composed of cells with the morphological appearance of basal cells. These findings demonstrate that K14-expressing basal cells actively participate in epithelial regeneration after depletion of the secretory cell progenitor pool and have the capacity to reestablish a normal tracheal epithelium.
The differentiation potential of K14-expressing cells revealed in the present study is consistent with a subset of progenitor cells identified among rat and human bronchial epithelia. In a series of elegant studies by Engelhardt and colleagues (5, 6), recombinant retroviruses were used to randomly tag isolated epithelial cells from either rat trachea or human bronchus, allowing analysis of their clonogenic and differentiation potential in a reconstituted tracheal graft. Because of the random process of retroviral gene transfer, it was theoretically possible to introduce lineage tags into every cell type capable of proliferation under the conditions of in vitro culture and to characterize the differentiation potential of these cells in the context of a heterotypic tracheal structure. Caveats of this approach were limitation of the analysis to those progenitor cells that could both proliferate in culture and successfully engraft within a recipient trachea, that differentiation potential of progenitor cells may have been influenced by the new microenvironment in which they reside, and that the random process of retroviral insertion could potentially lead to clone-specific positional effects on reporter gene expression with subsequent errors in assignment of differentiation potential. A further practical limitation of retrovirus-mediated lineage analysis is the inability to retrospectively identify the cell type whose clonogenic and differentiation potential is being evaluated. The bitransgenic mouse model used in the present study overcomes many of these difficulties by allowing in vivo introduction of lineage tags within basal cells and definition of their regenerative potential, in a physiologically relevant model of airway injury and repair. Application of this approach to lineage analysis in the repairing airway epithelium provides the first definitive evidence that K14-expressing basal cells represent a multipotent progenitor population for renewal of the airway epithelium. These data are also suggestive of at least two functionally distinct populations of K14-expressing cells that varied according to differentiation potential.
Caveats with the lineage tagging approach used in this study are that it is impossible to definitively establish clonality of tagged populations of cells. The existence of at least two populations of K14-expressing progenitor cells is based on the spatial separation of clusters of cells in which the lineage tag was localized to cells that uniformly exhibited basal cell morphology (unipotent) vs. clusters in which the lineage tag was distributed among all cell types typical of the tracheal epithelium (multipotent). Moreover, experiments performed herein involved administration of tamoxifen at doses that empirically yielded infrequent recombination within airways, thus increasing the likelihood of monoclonal rather than polyclonal expansion of tagged progenitor cell populations. However, unlike retroviral gene transfer in which the frequency of clonal misassignment can be precisely determined (6), such analysis is not feasible using the in vivo recombinational approach. A second caveat is that introduction of lineage tags in vivo makes it impossible to distinguish between intrinsic vs. extrinsic (i.e., microenvironmental) regulation of progenitor cell differentiation. As such, data generated from the present study and those of Engelhardt and colleagues (5, 6) are complementary, the collective data set arguing strongly that a subpopulation of K14-expressing basal cells exhibit multipotent differentiation potential and have the capacity for appropriate regeneration of the injured tracheal epithelium.
What cannot be determined from the present study is the contribution that other progenitor populations, such as CCSP-expressing secretory cells, make to airway renewal. Multiple cell types of the tracheobronchial epithelium, principally including basal and nonciliated secretory cells, participate in maintenance of the steady-state epithelium (3, 8). However, progenitor cell utilization is altered in the setting of airway injury, with nonciliated secretory cells representing predominant progenitor for renewal of the airway following -induced airway injury (8). In contrast, basal cells rather than secretory cells were the only cell type found to express the proliferation marker Ki67 within the bronchial epithelium of humans with chronic lung disease (1). These data, in addition to those of Engelhardt (5, 6), indicate that goblet cells are nonmitotic and that this differentiation pathway leads to an effective depletion of the secretory cell progenitor pool. These conflicting findings suggest that progenitor pools activated to participate in epithelial regeneration are determined by complex cellular interactions that are greatly influenced by target cells injured by specific agents. The findings of the present study, that basal cells actively participate in epithelial renewal following naphthalene-induced secretory cell depletion, highlight the dynamic nature of the airway epithelium and the redundancy that exists among resident progenitor cell pools. It is unclear, however, whether progenitor cell selection is in any way linked to epithelial remodeling in the chronically injured lung. It is possible, for example, that despite their pluripotent differentiation in the setting of naphthalene-induced airway injury, the differentiation potential of K14-expressing basal cell progenitor pool may be influenced to a greater or lesser extent by the local inflammatory milieu than that of nonciliated secretory cells. This could clearly be the case in airway diseases involving secretory to goblet cell differentiation, such as with chronic bronchitis and asthma, which may alter either the clonogenic or differentiation potential of the secretory cell progenitor pool but not the basal progenitor pool. One outcome of this pattern of biased progenitor cell utilization would be hyperplasia of basal cells, particularly in the event that unipotent rather than pluripotent basal cells were enriched among the proliferative pool, a finding that was observed by Barth and colleagues (1) in the remodeled airway epithelium.
The potential for functionally distinct populations of basal cells, as suggested from variability in the differentiation potential of K14-expressing cells in this study, is not without precedent. Basal cell subsets have been proposed based on either morphological or functional criteria. Basal cells of the hamster airway were characterized as either B1 or B2, depending on the location of the nucleus relative to the basal lamina (3), and similar criteria were used for the classification of basal cells of the normal human airway into either basal or parabasal subsets (2). Engelhardt and coworkers (6) proposed a model to explain lineage relationships in the human bronchial epithelium in which two populations of basal cells exist, B1 and B2, of which only B1 basal cells have the capacity for differentiation into an epithelium lining submucosal glands. In the present study, two populations of basal cells were identified according to their steady-state gene expression profile. All basal cells of the tracheal epithelium were shown to react with the lectin GSI-B4, yet only a fraction of GSI-B4-positive cells expressed K14 in the steady state. In contrast, no K14-expressing cells were observed among GSI-B4-reactive basal cells of the bronchial epithelium (13a). The fact that submucosal glands are restricted to the trachea of mouse airways raises the possibility that GSI-B4-reactive, K14-expressing cells of the tracheal epithelium may represent the murine counterpart of B1 cells described by Engelhardt and colleagues (6) for the human airway. This possibility was not conclusively resolved in the present study. Even though tagged K14-expressing cells were not found within submucosal glands, this may have resulted either from the low frequency with which lineage tags were introduced into K14-expressing cells or the lack of injury among epithelial cells of the submucosal gland.
The results of the present study are consistent with our previous study in which the K14/RS system was applied to intrapulmonary bronchial epithelium to examine the differentiation potential of bronchial basal cells. Our previous study showed that basal cells also play the role of multipotent progenitors in the bronchial epithelium. However, a major difference between these two studies was that in the bronchial epithelium, the majority of β-gal-positive cell clusters at later recovery time points lacked basal cells. This observation may be explained either by basal cells lacking extensive selfrenewal capacity or by the possibility that the tagged cell population is a K14-expressing cell positioned at a higher portion of the progenitor cell hierarchy. This implies that there are inherent differences in properties of progenitors and thus in mechanisms governing maintenance and renewal of the airway epithelium at different airway locations.
The study presented here has important implications in the understanding of mechanisms contributing to epithelial remodeling in chronic airway disease, lung neoplasia, and in the design of delivery approaches to achieve gene therapy. Future studies will involve distinguishing subpopulations among tracheal basal cells and identifying potential stem cell microenvironments within tracheal epithelium.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society