Department of Life Sciences, National Chung Hsing University (C. C. Yen, S. H. Yang, C. Y. Lin, C. M. Chen); Department of Internal Medicine, China Medical University Hospital (C. C. Yen), Taichung, Taiwan.
Prof. Chuan-Mu Chen, PhD, Department of Life Sciences and Biotechnology Center, National Chung Hsing University, No:250, Kuo Kuang Road., Taichung, 402, Taiwan. Tel.: +886-4-2285-6309; fax: +886-4-2287-4740; e-mail: firstname.lastname@example.org
Until recently, it was thought that only embryonic stem cells were pluripotent and that adult stem cells were restricted in their differentiative and regenerative potential to become the tissues in which they reside. However, the discovery that adult stem cells in one tissue can contribute to the formation of other tissues, especially after injury or cell damage, implies that stem cells have developmental plasticity. For example, haematopoietic stem cells (HSCs) and mesenchymal stem cells (MSCs) from bone marrow can be used to regenerate diverse tissues at distant sites, including the lung. This article reviews the character of stem cells in the lung parenchyma and focuses on the potential uses of adult stem cells in research of lung injury and lung disease therapies.
Acute lung injury (ALI) and its severe form, acute respiratory distress syndrome (ARDS), are common causes of morbidity and mortality in the intensive-care unit. Acute lung injury occurs as a result of direct intra-alveolar injury or indirect injury following systemic inflammation . The principal management of ALI currently includes treatment for underlying diseases, adequate haemodynamic support and mechanical ventilation with lung protective strategies. Hastening the repair process and attenuating the inflammatory response are thought to be two of the important directions in the study of improving ALI management . Adequate manipulation of stem cells in the lung parenchyma, including promotion of proliferation and re-epithelization by exogenous growth factors or cytokines, such as keratinocyte growth factor (KGF), hepatocyte growth factor (HGF), and granulocyte colony-stimulating factor (G-CSF), and cellular therapy with stem cells, is a promising treatment for acute lung injury and other diffuse lung diseases [3,4].
The study of stem cells in the lung has progressed relatively slowly because of the anatomical and functional complexities of that organ. The mature lung comprises at least 40 morphologically differentiated cell lineages . The upper airways are lined with ciliated columnar cells and mucus-secreting cells while the lower airways are lined with Clara cells. The alveoli are lined with alveolar type 1 and 2 epithelial cells. The pulmonary interstitium contains several specialized lineages of mesenchymal origin, including fibroblasts and smooth-muscle cells. There are also vascular, lymphatic and neuronal components .
During development of the respiratory system, airway branching is predominantly a prenatal event, while the formation of alveoli extends into post-natal life. Primordial lung buds originate as out-pouching of the primitive fore-gut endoderm. The bronchial tree is generated by repeated budding and branching of these tubules. Simultaneously, vascular components organize distally to form the future alveolar–capillary barrier . It was suggested that development of the conducting airway and alveolar regions are derived from different populations of stem cells; therefore, the discussion of stem cells in the lung should be divided into two parts, namely airway and lung parenchyma [5,8,9]. In this article, we will concentrate on alveolar epithelial stem cells.
Alveolar epithelial cells
The distal airways terminate in nearly 300 million alveolar air-sacs in the adult human lung. The surfaces of the alveoli are continuously lined by alveolar epithelium, which includes alveolar type 1 (AT1) cells and alveolar type 2 (AT2) cells. Although these two epithelial cell types are present in similar numbers, AT1 cells cover approximately 95% of the alveolar surface . The morphology of AT1 cells resembles that of a fried-egg with a centrally located nucleus and an extensive, attenuated cytoplasm forming a large surface area for gas exchange. The synthetic organelles (such as the Golgi apparatus and rough endoplasmic reticulum) are sparse, especially in the far distal region. The AT1 cells are thought to be incapable of cell division and are differentiated from AT2 cells which have normal turnover and can replace injured cells. Also, the AT1 cells are more vulnerable to injury because of their inability to undergo mitosis and because of their extensive cellular surface area, which presents a problem for transportation of proteins and other substances over a long distance .
The typical AT2 cell is a small cuboidal cell with short microvilli on its apical surface. The AT2 cells are made up of polarized epithelial cells and contain diverse organelles common to all cells. The distinguishing morphological feature of AT2 cells is the presence of lamellar bodies, which contain layers of surfactant phospholipids. Proposed as a ‘defender of the alveolus’, AT2 cells are metabolically active and exhibit many functions . It is well-known that AT2 cells are responsible for the synthesis and secretion of surfactant, which regulates alveolar surface tension in the lungs. Other proteins produced by AT2 cells include surfactant-associated proteins that affect surfactant recycling and immuno-modulatory functions, receptors for several growth factors, cytokines, enzymes, matrix proteins, and epithelial mucins . The AT2 cells have the ability to transport sodium from the alveolar to the baso-lateral space, which allows passive chloride and water flow. Fluid clearance from the alveolar space is critical in the resolution of pulmonary oedema and inflammatory lung injury [13,14]. Therefore, AT2 cells play an important role in the maintenance of the alveolar epithelium in lung parenchyma.
Stem cell characteristics
Stem cells are unique cells that have the capacity for limitless self-renewal and differentiation. These small and morphologically simple cells are devoid of differentiated features. Moreover, stem cells are multipotent, meaning that their progeny form at least one and often several lineages of differentiated cells . Although embryonic stem (ES) cells are pluripotent and give rise to every cell type in the animal body [16,17], adult stem cells in many tissues are found to be multipotent and are thought to be important in tissue regeneration and homeostasis .
Relative quiescence is a defining feature of stem cells, while their progeny are often highly prolific. Within the total population of haematopoietic stem cells (HSCs), only the resting cells provide long-term repopulation potential [19,20]. The relative quiescence of stem cells protects them from injury. A cyclin-dependent kinase inhibitor, p21, acts as a molecular switch by governing the entry of stem cells into the cell cycle . However, the detailed mechanisms for regulating the rate of stem cell division and the fate of daughter cells are poorly understood.
The alveolar epithelium can be classified as a continuously renewing tissue. The AT2 cells are thought to be the stem cells of lung parenchyma . In the early 1970's, many investigators found greatly increased numbers of cuboidal granular pneumocytes in the alveoli of lungs injured by hyperoxia and in those pneumonia infected by pathogens. Early studies demonstrated that post-mitotic progeny of AT2 differentiated into AT1 cells during repair. Shortly after injury, the alveolar surface is denuded and only cells with morphological features of AT2 remain. Over the course of several days, AT2 cells proliferate, resulting in repopulation of the alveolar basement membrane. Finally, the cuboidal AT2 cells flatten and acquire morphological features of AT1 cells. A phenotypically intermediate cell can be identified in the process [23,24]. The process of differentiation is also presumed to occur during the normal turnover of type I cells. Under physiological conditions, the turnover time of alveolar epithelium ranges from 28–125 days, while that of bronchial epithelium ranges from 2–10 days . The turnover rate is boosted after acute lung injury. The time course of events depends on the type and severity of injury. However, as early as 48 h after injury, newly labelled AT1 cells can be found .
A niche is a subset of cells and extra-cellular components that can accommodate indefinitely one or more stem cells and control their self-renewal and progeny production in vivo. Although the concept of the niche has been widely accepted, currently their presence is still theoretical in many tissues. Investigation of the niche is experimentally challenging because of the difficulty of identifying and manipulating individual stem cells and their surroundings. Niches in a few organs, such as the basal layer of dermis, the bulge of hair follicle, the crypt of the gut, canal of Hering in the liver and bone marrow have been identified [27–30]. The search for niches in other organs, including the airway and lung parenchyma, is currently underway.
Niches modulate the numbers of stem cells during development and in response to environmental factors. In many tissues, niches multiply to keep pace with growth from the juvenile to adult stages. For example, there are more crypts in the adult gut than in new-born children. Niches also modify crypt regulation in response to changes to ensure that stem cell activity parallels the organism's need for particular differentiated cell types . For example, bulge stem cells of hair follicles produce not only follicle cells but also epidermis in response to a local wound .
Stem cells with potential to contribute to the re-establishment of the normal bronchiolar epithelium have not been definitively demonstrated . Recent reports have suggested that stem cell niches exist in different sites along the airway [5,9,32]. For example, niches in the proximal glandular trachea have been found in the ducts of submucosal glands. In the distal non-glandular trachea and distal bronchial airways, niches are associated with innervated neuroepithelial bodies (NEBs) [9,32]. Similarly, stem cell niches in the lung parenchyma have not been adequately characterized. Although the alveolar epithelial AT2 cell is often referred to as the stem cell of the alveolar surface [23,24,33], it is still not clear that all AT2 cells in the adult lung meet the criteria of stem cells. Also, to date, the question of whether there is a subpopulation of AT2 or AT2-like cells which serve as the source of proliferation and epithelial repair after injury or whether all AT2 cells are capable of proliferating remians unanswered. One candidate of stem cell subpopulation would be the cells located at the bronchio-alveolar duct junction (BADJ) . A recent study discovered that cells at the BADJ carrying Clara cell and alveolar cell markers, Clara cell antigen (CCA) and surfactant protein-C (SP-C), appear refractory to naphthalene and bleomycin-induced damage, resulting in repair of both bronchiolar and alveolar cells. Isolated double-positive cells have been shown to differentiate to both Clara cells and alveolar cells in vitro, indicating that they are progenitors of both cell types . These cells are strategically located and are known to proliferate after injury. Another report showed that after hyperoxic lung injury, AT2 cells could be separated into E-cadherin-positive (CDH1+) and E-cadherin-negative (CDH1−) subpopulations. The E-cadherin-negative subpopulation was damage resistant, proliferative, and had high levels of telomerase activity, whereas the E-cadherin-positive subpopulation was damaged, quiescent, and had lower tolemerase activity .
The traditionally held view is that only embryonic stem cells are pluripotent and that adult stem cells are restricted in their potential; however, this view of adult stem cells has been recently challenged . Plasticity is the ability of cells to cross lineage boundaries . Although there are debates on reproducibility of the plasticity experiment and whether developmental plasticity is in fact cell fusion, recent studies have revealed that the stem cells of some organs have the ability to generate the cells of other lineages and other organs [37–40]. For example, bone marrow-derived cells have been shown to produce blood vessel, bone, muscle, liver, lung, skin and even brain (Fig. 1)[41–51]. Moreover, neural stem cells as well as muscle-derived progenitors can contribute to the production of blood cells [52–54].
Krause et al.  demonstrated that bone marrow-derived HSCs transplanted from an adult male mouse into an adult female mouse not only repopulated all of the haematopoietic cell lines but also engrafted to multiple organs. The Y chromosome containing (Y +) epithelial cells expressing cytokeratins were observed in the bronchi, gut and skin, and constituted a respectable 0·5–4% of the tissue 11 months after the grafting procedure. Over 20% of the epithelial cells in lung parenchyma contained Y chromosome. Functional activity in the engrafted cells was suggested by positive staining of surfactant protein B mRNA, a specific marker of AT2 cells. The highest level of donor engraftment in lung parenchyma may have been owing to the vulnerability of AT1 to injury and to lower cell proliferation of alveolar cells compared with cells in the gut. Several current investigations, using MSCs or HSCs transplanted into different damaged lung tissues in animal model and human clinical studies [56–59], also demonstrated that bone marrow derived cells can convert to AT2 cells via an oblique transdifferentiation pathway as defined in Table 1.
Table 1. Glossary of terms relating to stem cell biology in the lung parenchyma
Process through which unspecialized cells become more complex and specialized in structure and function
AT2 → AT1
Process by which a tissue or lineage-committed cell is reprogrammed to adopt the differentiative and proliferative potential of a more immature cell
AT1 → AT2
Process by which a tissue-specific cell adopts the fate of a cell representative of a different tissue
HSCs → Lung tissue
Process through which multipotent differentiative cells directly become a different tissue's cell with more complex and specialized in structure and function
HSCs → AT1
Process through which multipotent differentiative cells indirectly become a different tissue's cell with more proliferative and immature status
HSCs → AT2
In contrast, Kotton et al. observed that engraftment and transdifferentiation of HSCs were exclusive to AT1 cells in a bleomycin-induced lung injury model . LacZ-labelled cells were intravenously injected into wild-type recipient mice after intratracheal bleomycin treatment caused lung injury. As early as 2·5 days later, donor marrow-derived cells with morphological and molecular phenotypes of AT1 cells had engrafted to the recipient lung parenchyma. No engraftment of AT2 cells was detected during the experiment. Another recent report also demonstrated that AT1 cells and lung fibroblasts can be derived from circulating stem and progenitor cells in a radiation and elastase-induced lung injury model . This pathway is defined as a straight transdifferentiation (Fig. 2).
Despite these findings, some studies have shown that bone marrow-derived cells did not engraft to lung tissue. Wagers et al.  published a study in which a single HSC expressing green fluorescent protein (GFP) KTLS (positive for c-kit, thy-1, and Sca-1; lineage negative) was transplanted into a lethally irradiated host. In contrast to the results reported by Krause et al. , Wagers et al. demonstrated that no GFP positive donor-derived cells engrafted to the gastrointestinal (GI) tract, kidney or lung, although a few GFP cells were observed in the brain and liver. Another report observed that circulating vascular progenitor cells did not contribute to compensatory lung growth in a unilateral pneumonectomy model ; however, these conflicting results may have been caused by significant differences in study designs. For example, none of the studies was consistent in their use of donor cell subpopulations used for transplantation , models and severity of injury , and methods to detect donor cells. The transgene silencing made by target tissue regulation might be another reason for those conflicting results in engrafted cell determination [65–67].
There are a few questions regarding stem cell transdifferentiation which need to be answered. For example, which type of alveolar cells do the bone marrow-derived cells differentiate into first after engraftment? Do the engrafted cells go through the AT2 cells and then differentiate into AT1 or do they differentiate to AT1 cells directly? Are bone marrow-derived cells important for lung repair after injury? Finally, which cellular and molecular signals are involved in inflammation and engraftment?
One recent investigation performed by Yamada et al. showed that inhaled lipopolysaccharide (LPS) induced a rapid mobilization of bone marrow-derived progenitor cells into the circulation; engraftment in the lung was evident 1 week later . Moreover, the suppression of bone marrow by sublethal irradiation before intrapulmonary LPS led to emphysematous changes in the lung . Bone marrow transplantation attenuates the damage.
The mechanisms about which circulating stem cells are recruited into injured organs are not fully understood. The release of chemokines such as stromal cell derived factor-1 (SDF-1) from injured tissue and the interaction with its specific receptor, CXCR4, on HSCs were found to be important in the engraftment [68–70]. On the other hand, endogenous CD26 (DPP IV/dipeptidylpeptidase IV) expression in donor HSCs was found to have a negative effect on homing and engraftment. By inhibiting CD26, it is possible to markedly increase the efficiency of transplantation . This may also explain the negative results reported in some previous investigations [37,62].
Dedifferentiation (Table 1) from AT1 to AT2 cells has been found and AT2 cells were shown to differentiate into AT1 cells with normal morphology during repair and normal turnover . Previous studies have suggested that the AT1 cell phenotype is plastic and capable of reversion to AT2 cells, at least in vitro. The AT2 cells cultured on attached collagen gels began to express AT1 cell markers, but reverted to an AT2 cell phenotype following detachment of the gel [72,73]. A similar phenomenon was found in the central nervous system (CNS) and muscle tissues [74,75]. Purified oligodendrocyte precursor cells (OPCs), which may terminally differentiate into oligodendrocytes after a number of cell divisions, reverted to a status that resembles multipotent CNS stem cells after sequential exposure to foetal calf serum and basic fibroblast growth factors . These results suggest that differentiated cells conserved the ability to de-differentiate into progenitor cells when stimulated by appropriate signals [35,75].
Evolving stem cell model
Recent studies have suggested that the definition of stem cells might need to be revised. Firstly, adult stem cells may not only act locally in specific tissues in which they reside, but may also be recruited by cells of other lineages in distal organs [44,48]. Hence, stem cells can be divided into two types: endogenous and exogenous ones (Fig. 2). Endogenous stem cells contribute locally to replenish cells immediately after injury but have low prolific potential. Exogenous stem cells derived from bone marrow contribute in the later phases of repair and are reserved for more severe injury [68,76]. Secondly, even highly specialized cells may be capable of reversing their differentiated status and contributing to the stem cell pool [35,74,75,77].
A few studies have suggested that the stem cell is not only a discrete entity, but refers to a biological function that can be induced in many distinct types of cells, including differentiated cells [35,77]. The propensity of a cell to initiate stem cell functions is likely to decrease as cells differentiate. The microenvironment, including contact with surrounding cells and the extra-cellular matrix, likely plays a key role in determining the function of stem cells. Tissue with a high cell turnover, such as the skin and blood, requires the presence of local cells with high stem cell propensity. However, in a damaged tissue there are high levels of environmental signals, such as SDF-1, that may recruit cells from the stem cell pool and even trigger the stem cell function of some differentiated cells [69,70].
Stem cells of the alveolar epithelium
It was previously thought that AT2 cells were the only stem cells of the alveolar epithelium, and that AT1 cells were terminally differentiated cells. If either cell type is lost, the nearest AT2 cells will be stimulated to proliferate and, if necessary, differentiate into AT1 cells . All the processes were believed to be linear and irreversible. Division of AT2 gives rise to two daughter cells with different fates. One of the daughter cells transforms into an AT1, while the other remains as an AT2, thereby replenishing the original stem cell population [10,78]. The efficiency of that AT2 alveolar repair is inversely related to the development of lung fibrosis . However, uncontrolled proliferation of alveolar stem cells might result in alveolar cell carcinoma .
A modified working model of alveolar epithelial kinetics in the adult lung was proposed by Uhal . Nearly all AT2 cells in the normal lung are quiescent (G0 phase) and can re-enter the cell cycle after stimulation. In this situation, AT2 cells may be considered to have regained the proliferative function for alveolar epithelial repair [22,72,79]. Cell division leads to two relatively undifferentiated daughter cells with the potential to revert to either differentiated phenotype. It is also possible for AT1 cells to re-enter the cell cycle through dedifferentiation (Table 1) to the AT2 type. Both cell types have the potential for apoptosis, which is important for the maintenance of dynamic balance of the epithelial cell populations . Lung injuries that damage the AT1 cells are generally repairable, whereas injuries that kill both cell types are often fatal, presumably as a result of the destruction of the stem cell population [10,80]. Endogenous and exogenous stem cells may contribute to repopulate the cells of alveolar epithelium, either as AT1 or AT2 cells, and lung regeneration [55–61].
Regulation of cells in the lung parenchyma
The molecular signals that are required to initiate AT2 division are beginning to be identified. Several agents, such as fibroblast growth factor-1 (FGF-1), KGF and HGF, trigger AT2 mitosis in vitro and in vivo. The KGF promotes the proliferation and migration of AT2 cells. Local administration of KGF has been shown to markedly reduce the degree of lung injury from hyperoxia, bleomycin, radiation and acid aspiration [81,82]. Administration of HGF to animals with lung injury has been shown to stimulate alveolar cell proliferation . Transforming growth factor-β (TGF-β) exerts an autocrine negative effect on AT2 cell proliferation and plays a key role in mediating fibrotic tissue remodelling [6,83].
Foetal lung expansion induced the differentiation of AT2 into AT1 cells via an intermediate cell type in vitro. However, KGF promoted the retention of the AT2 cell phenotype and at least partially reversed the transdifferentiation from the AT2 to AT1 cell phenotype in primary culture .
Retinoids, including retinal (vitamin A) and retinoic acid (RA), regulate the morphogenesis, cell proliferation and differentiation in many organs. In the lung, recent studies have shown that retinoic acid stimulated proliferation of AT2 cells and affected a number of processes associated with lung development and maturation, as well as lung repair after injury . Treatment with RA has a positive effect on alveolar structure in elastase and oxygen induced lung injury animal models [87,88]. Previous studies indicated that retinoids exert the primary influence by affecting the genes responsible for lung morphogenesis [89,90]. Furthermore, all-trans retinoic acid (ATRA) and G-CSF were found to increase recruitment of bone marrow-derived cells into the lung .
Prospects for therapy
Regeneration of tissue by stem cells, from endogenous, exogenous and even genetically modified cells, is a promising novel therapy [35,91]. Examples include the generation of different types of neurones for treatment of Alzheimer's disease, spinal cord injury, and Parkinson's disease, the production of cardiac muscle for ischaemic heart disease, the generation of insulin-secreting pancreatic islet cells for diabetes and the generation of derma papilla and hair follicle for certain types of baldness . The understanding of stem cell biology provides three potential clinical applications for ALI and other lung diseases (Fig. 3).
Many growth factors and cytokines influence cell migration, matrix deposition, differentiation and apoptosis. They also play major roles in co-ordinating stem cell proliferation and differentiation in adult tissue during development and regeneration, and engraftment of exogenous stem cells from bone marrow [69,70,92]. It is known that over-expression of certain growth factors can lead to pulmonary fibrosis and even neoplasm [26,93,94]; therefore, much research is needed before this approach can be applied clinically.
Recent animal experiments have indicated that both HGF and KGF stimulated alveolar epithelial cell proliferation [81,82,95]. Furthermore, KGF has also been shown to increase Na+-K+ ATPase expression and up-regulate Na+ transport across the alveolar epithelium in vitro and in vivo[96,97]. Retinoic acid shows promise as a potential stimulator of alveolar regeneration according to two animal studies [86,87]. In the case of ALI, many growth factors and cytokines are involved in the process of inflammation and regeneration, so combination therapies may be more effective than mono-therapy. For example, one drug could be used to block inflammation processes while another one could be used to stimulate stem cells and promote regeneration [98,99].
The purpose of cellular therapy is to restore the function of damaged tissues by the transplantation of cells. Two recent reports, one dealing with embryonic stem cells and the other with adult stem cells, provided evidence that cellular therapy with stem cells is promising. The first report demonstrated that mouse embryonic stem (ES) cells could generate cells which secrete insulin and other pancreatic endocrine hormones . The second report implied that adult stem cells, which invoke few ethical difficulties, can be used for cellular therapy in lung repair .
Autologous stem cell transplantation could be used to overcome the problem of immune rejection in regenerative medicine. Adult stem cells could be transplanted into a new niche where they would be able to repopulate the cells of defective tissue . Stem cells can be purified from the cells of bone marrow, umbilical cord and skin, and then stored before a major operation or when it is predicted that a patient is at risk of acute lung injury. As an agent of combination therapy, stored stem cells would be injected back into patients with acute lung injury to enhance lung repair . Embryonic stem cells derived by somatic nuclear transfer are genetically identical to the donor, and thus potentially useful for cellular therapy [101,102].
It is also possible that stem cells could be used to reconstitute more complex tissues and organs in vitro, and then transplanted to replace failed organs. Tissue engineering is evolving rapidly and no tissue or organ has been excluded from active research; however, only a few stem cell-generated products have entered clinical trials, namely cartilage and skin [103,104].
Gene therapy is the introduction of exogenous genetic material to correct or modify the function of cells [105,106]. However, researchers are still unable to maintain high levels of gene expression without repeated gene transduction. The application of stem cells may resolve that problem.
Gene therapy with stem cells is thought to be an ideal treatment strategy for many genetic diseases and lung injury . In principle, genetic modification of small numbers of stem cells produces a stable population of genetically altered cells and does not require repeated procedures. Another attractive feature of the method is that the genetically modified stem cells can engraft into many tissues where the genetic disease is involved . However, progress has been limited by the low HSC gene transduction rates. Retroviruses are ideal vectors for integrating target genes into the DNA of host cells but it is difficult for them to infect relatively quiescent stem cells [105,107,109]. Human foetal circulation was found to be rich in actively cycling stem cells; furthermore, it was reported that circulating foetal HSCs are more permissive to retroviral transduction. They could serve as novel targets for in utero gene therapy .
AT2 cells are thought to be the stem cells of lung parenchyma. Nearly all AT2 cells in the normal lung are quiescent (G0 phase) but can regain stem cell function after being stimulated to repair alveolar epithelia. It is possible for AT1 cells to re-enter the cell cycle by dedifferentiation to AT2 cells. According to the new definition of stem cells, exogenous stem cells, which come from the stem cell pool, contribute to the population of alveolar epithelial cells and lung regeneration. Furthermore, endogenous stem cells also can derive from AT1 transdifferentiated cells, including lower airway epithelial cells.
Advances in stem cell biology and technology will eventually turn cell transplantation into a useful treatment for patients with a variety of diseases and injuries, including genetic diseases involving lung and acute lung injury. However, much more research on how to purify stem cells,and how to direct the differentiation of stem cells and to stimulate regeneration and functional recovery is needed.
This work was supported in part by Grant NSC92-2311-B-005-012 from the National Science Council, Taiwan and by Grant TCVGH-NCHU-927611 from the Taichung Veterans General Hospital and National Chung Hsing University, Taiwan.