The Authors: Associate Professor Ivan Bertoncello is the head of the Lung Regeneration Laboratory leading research investigating the organization and regulation of adult lung stem and progenitor cells, and their role in lung regeneration and repair. Dr Jonathan McQualter leads the Lung Stem Cells and Microenvironment Group and has a strong interest in understanding how the dynamic interaction of lung epithelial stem cells with their microenvironment regulates their regenerative potential.
SERIES EDITORS: YUBEN MOODLEY AND PHILIP J THOMPSON
Correspondence: Ivan Bertoncello, Lung Health Research Centre, Department of Pharmacology, Lung Health Research Centre, Level 8, Medical Building, The University of Melbourne, Victoria 3010, Australia. Email: firstname.lastname@example.org
Recognition of the potential of stem cell-based therapies for alleviating intractable lung diseases has provided the impetus for research aimed at identifying regenerative cells in the adult lung, understanding how they are organized and regulated, and how they could be harnessed in lung regenerative medicine. In this review, we describe the attributes of adult stem and progenitor cells in adult organs and how they are regulated by the permissive or restrictive microenvironment in which they reside. We describe the power and limitations of experimental models, cell separative strategies and functional assays used to model the organization and regulation of adult airway and alveolar stem cells in the adult lung. The review summarizes recent progress and obstacles in defining endogenous lung epithelial stem and progenitor cells in mouse models and in translational studies.
Adult stem cells: Rare, undifferentiated cells within a tissue that are able to renew and maintain the lifelong production of progenitor cells and the diverse mature functionally differentiated non-dividing cells which comprise that tissue in both the steady state and following perturbation or injury.
Progenitor cells: Relatively undifferentiated transit amplifying cells that arise from stem cells and have a finite lifespan with a more limited proliferative capacity and a more restricted ability to give rise to descendant cell lineages within a tissue.
Facultative stem cells: This term describes a class of normally functional differentiated cells in some adult organs that are able to revert to an undifferentiated stem cell state and regenerate diverse damaged cell lineages in response to severe perturbation or injury.
Regenerative cells: In this review, this is a collective term used to describe any cells in a tissue or organ that are able to proliferate and replenish descendant cell lineages in the steady state or following injury.
Pluripotent/multipotent/unipotent stem cells: These terms describe the relative ability of a stem cell to give rise to diverse differentiated cell lineages. The term pluripotent stem cell is reserved for embryonic stem cells that give rise to differentiated cells of endodermal, mesodermal and ectodermal origin that comprise the organism. Multipotent stem cells give rise to a multiple but limited number of cell lineages, usually of a single germ layer origin. Unipotent stem or progenitor cells are only able to differentiate into a single cell type.
‘Omnis cellula e cellula’
Rudolf Virchow (1858)
The identity of stem cells, their location and organization within tissues, and the regulation of their regenerative potential has been the subject of ongoing debate since articulation of modern cell theory in the mid-19th century, which recognized that cells are the fundamental building blocks of all living organisms and that every cell is derived from a pre-existing cell (Omnis cellula e cellula). The term ‘stem cell’ was originally coined in the late 19th century by embryologists to describe embryonic cells thought to give rise to the specialized tissues of multicellular organisms during ontogeny and by haematologists to describe the hypothesized common precursor of the diverse cell types identified in the bone marrow, peripheral blood and lymphatic system.
Historically, adult organs have been classified as continuously renewing (bone marrow, gut, skin), conditionally renewing (lung, kidney, liver) or essentially non-renewing (nervous tissue, muscle) on the basis of their proliferative capacity and cell turnover in the steady state or following injury.[2, 3] A continuously renewing tissue like the bone marrow has the prodigious capacity to generate approximately 200 × 109 erythrocytes, 50 × 109 white blood cells and 125 × 109 platelets daily in the steady state, and can rapidly augment cell production up to 10-fold in times of stress. The gut also generates approximately 100 × 109 intestinal epithelial cells daily, replacing the entire intestinal epithelium every 5 days.[5, 6]
The lung on the other hand is a conditionally renewing organ. Airway epithelial cell turnover is low (∼1% per day) with replacement of the tracheobronchial epithelium estimated at approximately 4 months. However, like kidney and liver, the lung is able to undergo compensatory growth rapidly increasing its rate of regeneration to replace ablated tissue following severe injury.
The Intrinsic Properties of Adult Stem Cells
The adult stem cell paradigm is largely underpinned by the analysis of cell regeneration in continuously renewing tissues. Accordingly, adult stem cells are defined as rare, morphologically unrecognizable cells endowed with a high proliferative potential and the lifelong ability to: (i) maintain and replenish their own kind; (ii) generate large numbers of functionally differentiated progeny; and (iii) replace senescent and damaged cells in the steady state and following perturbation or injury.
The classical adult stem cell hierarchy comprises an age-structured continuum of stem and transit amplifying progenitor cells of progressively restricted proliferative and differentiative potential that in turn give rise to mature non-dividing functionally differentiated cells (Fig. 1). As a general rule, multipotent stem cells at the head of this hierarchy are highly quiescent. Typically, the vast majority are in a G0 state, and descendant cell lineages are derived from a small number of active clones, thus ensuring the preservation of a stem cell reserve while mitigating the risk of error-prone stem cell replication and transformation. Day-to-day demands for replacement of senescent and damaged cells are met by more abundant rapidly cycling lineage-restricted transit amplifying progenitors of limited proliferative potential. According to this paradigm, stem cells divide asymmetrically to homeostatically regulate stem cell pool size while simultaneously replenishing the more committed transit amplifying progenitor cell compartments, but also retaining the ability to divide symmetrically to expand stem cell numbers in order to meet the increased demand for differentiated cells following tissue injury.[13, 14] Analysis of stem cell compartments in inbred mouse strains and in the BXD set of recombinant mouse strains derived by crossing C57BL/6J and DBA/2J mice[16, 17] has also shown that stem cell variables including pool size, cell kinetic status, sensitivity to cytokines, irradiation and cytotoxic perturbation are genetically determined.
While the organization of regenerative cells in continuously renewing tissues conforms to this classical model, it is not precisely recapitulated in all organs and tissues. Regional differences in the developmental potential and organization of regenerative cells have been documented in different compartments of organs including the prostate, skin and gastrointestinal tract,[18-20] and along the proximal-distal axis of the lung.[7, 21] A recent study argues against the obligate requirement for a long-lived oesophageal epithelial stem cell reserve.[20, 22] There is evidence for and against the existence of regional diversity in neural stem cells in the brain. Subpopulations of differentiated multifunctional cells able to revert to a (facultative) stem cell state have been identified in lung, liver and pancreas, while terminally differentiated cardiomyocytes appear capable of regenerating damaged heart tissue by a process of dedifferentiation and redifferentiation.[26, 27]
However, it is interesting to note that at a time when the delineation and functional analysis of adult stem cells was in its infancy, Till speculated that they would likely be characterized by the expression of multiple markers including differentiation markers characteristic of later stages of development. Likewise, Lajtha questioned the imprecision of the concept that stem cells were necessarily the sole precursors of other cell types and the assumption that they were necessarily undifferentiated. He speculated: (i) that although stem cells may be less differentiated than their descendants, they were nonetheless one of many differentiated cell types; and (ii) that renewal is relative rather than absolute with stem and transit progenitor cells part of a recognizable age-structured continuum in which both stem and transit populations could be precursors of other transit populations. In hindsight, these reflections on the nature and properties of stem cells have proven remarkably insightful, suggesting that classical and non-classical models of stem cell organization may have more in common than might appear at first glance[25, 29, 30] and emphasizing the fact that stem and progenitor cells are defined operationally by what they do rather than how they are decorated.
Stem Cells Are Defined in Context—The Role of the Niche in Stem Cell Specification
Importantly, adult stem cells are not solely defined by their intrinsic properties but also by the permissive or restrictive microenvironments in which they reside. The importance of the microenvironment in stem cell specification had long been appreciated in both embryonic development and adult haematopoiesis, but the concept of the stem cell niche was only articulated by Schofield in 1978 in modelling the regulation of haematopoietic stem cells. He proposed that adult stem cells occupy specific anatomical sites that preserve their developmental potential, regulate their replication, inhibit their differentiation and provide a milieu conducive to the preservation of engrafted cells in a stem cell state. The crosstalk between the stem cell and its niche profoundly affects stem cell behaviour and functionality.[3, 34-37] In the lung, the interdependence of regenerative cells and their niche microenvironment are evident both in the process of lung epithelial regeneration and repair, and in the aberrant proliferation and differentiation of airway and alveolar epithelial cells as a consequence of the pathophysiological remodelling and scarring of the subepithelial microenvironment in lung fibrosis and asthma.[39-42]
Problems and Pitfalls in Assaying Stem Cells and Modelling Stem Cell Hierarchies
Because stem cells are fundamentally defined by what they do rather than by their morphology or biomarker profile, the assignment of ‘stemness’ relies on the precise analysis of their developmental potential and regenerative capacity. For this reason, the modelling of stem and progenitor cell hierarchies has progressed in lockstep with the development and refinement of:
Specific antibodies, reagents and multiparameter cell separative protocols for the prospective isolation of highly purified and defined target cells
Robust in vitro clonogenic assays for enumerating stem and progenitor cells and measuring their proliferative and differentiative potential
Transplant models and assays to measure their regenerative potential in vivo
Lineage-tracing techniques to mark their spatial location in situ and measure their contribution to tissue maintenance, regeneration and repair; and,
Genetically engineered and mutant mouse models to identify specific stem cell regulatory networks
The delineation of haematopoietic stem and progenitor cells is an outstanding example of the power of these multiparameter strategies to: (i) analyse stem and progenitor cell organization and regulation; (ii) validate assays predictive of their regenerative capacity; and (iii) identify stem cell targets around which therapies can be constructed (Fig. 2). The characterization of multipotent mouse mammary stem cells capable of regenerating a functional mammary gland following transplantation of a single cell[44, 45] is another pertinent example. However, because the analysis of the functional attributes of stem cells is context-dependent, there are many experimental variables that can confound the assignment of stemness and compromise the ability of assays to discriminate closely related stem and progenitor cell subsets and measure their ability to give rise to descendant lineages.
Tissue disaggregation disrupts the relationship between stem cells and their regulatory microenvironment potentially affecting their viability, developmental potential and signature profile.[46, 47] Stem cell biomarkers are commonly shared by multiple cell lineages as well as closely related cells of differing developmental potential and regenerative capacity.[48, 49] The proliferation and differentiation of stem and progenitor cells in vitro is determined by colligative properties of the culture system including cell seeding density, choice of medium, oxygen tension, pH and the conditioning of the medium by factors elaborated by the progeny of the cultured stem cells. Discordance between proteome and transcriptome as a result of post-translational regulation of the proteome must also be considered in assessing the fidelity of biomarker signature profiles used to characterize and order stem and progenitor cells and their descendant progeny.[53, 54]
The interpretation of in vivo assays can be similarly confounded. Transplant outcomes are contingent on the number of transplanted cells, their route of delivery and the modality employed to precondition transplant recipients. Evidence from haematopoietic transplant models also shows that stem cell homing and engraftment is influenced by the cell cycling status of transplanted stem cells. The regenerative capacity of resident stem cells in acute lung injury models can be confounded by disruption and remodelling of the stem cell niche microenvironment,[7, 56] and cell lineage tracing in conditional transgenic reporter mouse models can be confounded by the lack of specificity of promoters and the pharmacokinetics of inducers used to elicit gene expression.[57, 58]
Characterization and Organization of Lung Epithelial Stem and Progenitor Cells
The lung is a complex organ comprising at least 40 cell lineages of endodermal, mesoderm and ectodermal origin that arises from an outpouching of the foregut. While progress has been made in delineating stem and progenitor cells in all lung compartments,[21, 60] this review focuses on controversies surrounding the nature, properties and organization of epithelial stem cells in the adult lung.
During development, the primordial lung undergoes a precisely orchestrated program of branching morphogenesis culminating in arborization of the conducting airways and their supporting parenchyma and vasculature, terminating in the gas-exchanging alveolar bed.[59, 61, 62] The trachea and proximal airways are lined by a pseudostratified epithelium comprising basal, secretory, ciliated and neuroendocrine cells. Distal conducting airways are lined by a simple columnar epithelium comprising club (Clara), ciliated and pulmonary neuroendocrine cells, while alveoli are populated by cuboidal surfactant-secreting alveolar epithelial type II cells and squamous gas-exchanging alveolar epithelial type I cells. The cell lineages comprising each of these compartments are maintained by regenerative cells characterized by high proliferative activity during foetal and perinatal life, and slow turnover and the capacity to replace senescent and damaged cells in the adult.[63-65]
The organization of endogenous epithelial stem and progenitor cells in the adult lung has been extensively reviewed in recent years.[7, 21, 60, 62, 66, 67] Briefly, these studies suggest that regional stem and progenitor cells are responsible for the maintenance of specific epithelial cell lineages in the proximal and distal conducting airways and the alveolar bed (Fig. 3). These include p63pos cytokeratin 5-positive (Krt5pos) and/or Krt14pos basal cells, secretoglobin family 1A member 1-positive (Scgb1a1pos) club (Clara) cells and Krt14pos submucosal gland duct cells of the trachea and upper airways;[68-72] naphthalene-resistant (cytochrome P450 negative: CyP450neg) Scgb1a1pos and/or Scgb3a2pos-variant club (Clara) cells and neuroendocrine cells in the bronchiolar airways;[73-75] club (Clara) cell specific protein-positive (CCSPpos) surfactant protein C-positive (SP-Cpos) bronchioalveolar stem cells at the bronchioalveolar duct junction of terminal bronchioles;[76-78] and the type II progenitor of type I cells in the alveolar bed.[79, 80] However, the analysis of lung epithelial cell regeneration in genetically engineered mice and the development of multiparameter cell separative strategies and three-dimensional organotypic clonogenic assays for the identification and characterization of adult lung stem cells in vitro has added another layer of complexity and ambiguity to the understanding of their organization and spatial localization.
Lung injury models confirm that the regenerative potential of candidate stem cells is context-dependent and that different lung epithelial stem and progenitor cell cohorts are recruited to regenerate specific airway epithelial cell lineages in the steady state and following severe perturbation.[81, 82] On the other hand, in vitro clonogenic assay of putative epithelial stem cells and transcription profiling of their progeny provide evidence of the existence of multipotent epithelial stem cells in the adult lung,[21, 83-85] while cell lineage tracing in genetically engineered conditional mutant mice also reveals plasticity in the differentiative potential of specific lung epithelial stem cell cohorts.
Discordant studies notwithstanding, this context-dependent diversity in the proliferative and differentiative potential of epithelial stem and progenitor cells has been observed in all anatomical compartments of the airway tree. In the trachea, Krt14pos basal cells adopt a different fate in the normal lung and following naphthalene injury.[71, 87] Likewise, Hegab et al. have identified multipotent nerve growth factor receptor-positive, α6 integrin-positive (NGFRposα6pos) tracheal submucosal gland duct cells able to regenerate both submucosal gland tubules, and secretory and ciliated airway epithelial cells following severe ischaemic injury. In the distal lung, pulmonary neuroendocrine and alveolar cells (but not club (Clara) and ciliated cells) have been shown to share a common calcitonin gene-related peptide-positive neuroendocrine cell (CGRPpos) lineage origin during development, whereas they function as unipotent neuroendocrine progenitor cells in the normal adult lung but are able to renew and generate both club (Clara) and ciliated cells following naphthalene injury. Other lineage-tracing studies also provide evidence of the derivation of alveolar epithelial cells from Scgb1a1pos airway cells following bleomycin injury and from normally absent p63posKrt5posKrt14pos airway cells following H1N1 virus infection. In the alveolar bed, it appears that alveolar epithelium is not simply repaired by expansion and differentiation of surfactant protein C-positive type II cells following bleomycin injury but also by SP-Cneg α6pos β4 integrin-positive (b4pos) cells sporadically distributed in the alveolar wall.
Modelling Lung Epithelial Stem Cell Hierarchies: Are We There Yet?
The development of powerful animal models and the refinement of cell separative strategies and in vitro clonogenic assays based on approaches proven effective in characterizing haematopoietic stem cells (Fig. 2) have underpinned the significant progress made in identifying putative stem cells in the adult lung, and the critical factors and pathways regulating their regenerative potential. However, the lack of definitive lung epithelial stem cell biomarkers; the co-expression of cellular and molecular biomarkers by cells of differing regenerative potential, cell lineage origin or maturational age; the limitations of flow cytometric sorting and in vitro organotypic assays; and the lack of gold-standard in vivo assays akin to those that have enabled the validation and modelling of the haematopoietic stem cell hierarchy[93, 94] have conspired to blur the identity, spatial location and organization of lung epithelial stem cells in health and disease. In particular, it is still unclear whether the heterogeneous behaviour of seemingly homogeneous target cells is indicative of cofractionation of heterogeneous cells expressing common biomarkers or the stochastic commitment of homogeneous cells to different fates in response to microenvironmental cues. This is especially the case in injury models where modulation of biomarker signature profiles is not necessarily indicative of the ability of regenerative cells to give rise to descendant lineages.
Translational Studies and the Identification of Therapeutic Targets in the Human Lung
As yet, relatively few studies have attempted to isolate and characterize adult epithelial stem cells in the human lung. These have mostly analysed tracheal and proximal airway epithelia utilizing various in vitro assays including air-liquid interface culture and serial propagation of spheroid-forming cells and the generation of mucocilliary epithelium in devitalized tracheal implants xenografted in immune-compromised mice (reviewed in Bertoncello and McQualter). More recent studies also suggest that human homologues of murine lung epithelial stem cells are also amenable to analysis and characterization employing similar experimental strategies to those utilized in mice.
Both immortalized human bronchial epithelial cells and freshly isolated human proximal airway epithelial cells[70, 96-98] generate complex three-dimensional organoids of differentiated airway epithelial cells when cloned in matrigel. Multiparameter flow cytometric analysis shows that these putative human airway epithelial stem cells share a biomarker profile similar to that of their murine counterparts. These include NGFRposα6pos p63pos basal cells that generate cystic organoids of Krt14posp63posbasal cells, Krt8pos luminal cells and ciliated cells; and nerve growth factor receptor-positive α6pos basal and submucosal gland duct cells that can be further resolved on the basis of their aldehyde dehydrogenase, CD166, CD44, and α6 integrin signature profile to identify cells differing in their ability to generate organoids comprising Krt5pos, Krt14pos, and mucous and serous secretory epithelial cells.
Others have propagated dispersed, heterogeneous human lung cell suspensions in liquid culture to pre-enrich lung epithelial stem cells prior to their prospective isolation and characterization by flow cytometric cell sorting and functional analysis.[99, 100] Using this approach, Oeztuerk-Winder et al. have cloned putative E-cadherin-positive, leucine rich repeat-containing G protein-coupled receptor 6-positive (E-CadposLgr6pos) α6pos lung epithelial stem cells that can be serially passaged in liquid culture, and they have shown that both freshly isolated and propagated E-CadposLgr6pos cells are able to generate differentiated human bronchioalveolar epithelial tissue following transplantation of a single cell under the kidney capsule of recipient immune-deprived (CD-1 nude) mice. While lung cells expressing the stem cell-associated antigen c-kit did not exhibit stem cell potential in this study, another study using this approach has described a putative multipotent c-kitpos human lung stem cell with broader developmental potential, seemingly capable of generating lung epithelial, endothelial and mesenchymal cell lineages in different contexts, but its status is as yet uncertain.
This notwithstanding the delineation of human lung epithelial stem and progenitor cell hierarchies is beset with practical difficulties likely to slow progress. Translational studies aimed at identifying stem cell homologues of those in mice will need to consider differences in the distribution of epithelial cell lineages along the proximal-distal axis of the murine and human airway tree. Murine cells also have extremely long telomeres, endowing them with a far more prodigious replicative capacity than human cells. Consequently, human homologues of putative mouse lung epithelial stem cells may have a more restricted proliferative potential and may not be appropriate therapeutic targets in clinical contexts.
Practical issues related to the acquisition of human lung tissue biopsies suitable for stem cell experimentation will impact the development and refinement of optimal strategies for stem cell isolation and characterization. The modelling of lung epithelial stem cell organization and regulation will require access to normal lung tissue from different anatomical regions along the proximal-distal axis of the airway tree and its processing in a timely manner. Regular access to normal lung is rare, and resected ‘normal’ lung tissue and bronchoscopy samples are invariably acquired from patients undergoing procedures for diverse clinical indications. These clinical indications are likely to affect the biomarker profile and attributes of putative stem and progenitor cells, and also amplify the intrinsic biological variability of measured human parameters necessitating larger sample sizes for statistical validation of stem cell attributes.
That said, recent clinical studies demonstrating the therapeutic efficacy of tracheobronchial reconstruction using bioengineered recellularized human tracheal matrix scaffolds are a cause for optimism for the future of stem cell-based therapies in lung regenerative medicine.
The authors are supported by grants from the Australian National Health and Medical Research Council (Grant 1009374) and the Victorian Cancer Agency.