Cellular therapies for lung disease: A distant horizon


  • The Authors: Dr Yuben Moodley is Associate Professor of Respiratory Medicine at the University of Western Australia with an interest in lung fibrosis and cellular therapies for lung disease. Dr Ursula Manuelpillai is Senior Research Fellow at the Monash Institute for Medical Research with an interest in placental stem cells. Dr Daniel Weiss is Associate Professor of Medicine at the University of Vermont and a leader in lung stem cell research.


Yuben Moodley, School of Medicine and Pharmacology, University of Western Australia, Level 4, Medical Research Foundation Building, 50 Murray Street, Perth, WA 6000, Australia. Email: yuben.moodley@uwa.edu.au


Lung diseases constitute a major global burden of health and are characterized by inflammation and chronic fibrosis resulting in a loss of gas exchange units. To date there has been no effective treatment to reverse these chronic inflammatory changes and tissue remodelling. Recently, stem cells have been shown to successfully treat animal models of lung disease. In addition, certain cells have demonstrated a capacity to differentiate into lung cells. Based on these preliminary data, there are clinical trials underway to examine the potential for cellular therapies in lung disease. Recently, there have been a variety of cell examined for both their immunomodulatory effects on the lung as well as their potential for differentiation into lung cells. These range from lung progenitor cells, circulating cells, mesenchymal stem cells (MSCs), induced pluripotent stem cells (iPS), placental stem cells and embryonic stem cells (ESCs). Several cell types demonstrate immunomodulatory effects including circulating cells, MSCs and placental stem cells. In addition, iPS, placental cells and ESCs have shown some capacity for differentiation.

Despite these major steps forward cellular therapy for lung diseases still faces challenges. Issues that need to be resolved include bioethical issues, the safety of cell transplantation, ideal routes of delivery, the timing and the specific indications that would make cellular therapy effective.


The wide range of physiological functions and the ability of the pulmonary system to adapt to a variety of different circumstances, stresses and environments is a key reason for the success of air-breathing organisms. In this regard, the human lung is a good example of these characteristics particularly with respect to gas exchange. The human lung is a structure of on average16 generations of branching airways leading to the alveoli, the functional gas exchange units of the lung. The total gas exchange surface area of the lung covers 70 m2 or approximately the size of a tennis court and is the largest surface of the body in direct contact with the environment. In contrast, gas diffusion occurs across 1 µm, the thickness of the alveolar-capillary barrier.1 During normal respiration, the airway transports approximately 10 000 L of air in 24 h that is frequently contaminated with pollutants and infectious agents.2 As a result of regular exposure to both organic and inorganic insults, the lung has evolved multifaceted defence mechanisms that are critical for survival. Nonetheless, despite the barriers of an advanced innate and acquired immune defence system, the respiratory system is subject to injury on a regular basis resulting in damage to the structural cells forming the epithelium and endothelium. As such, the airway has to develop a robust system of replacing injured or damaged cells. Under these circumstances the damaged cell either undergoes repair or death by necrosis or apoptosis. Restitution of dying structural cells of the lung by endogenous progenitor cells is central to recovery. Therefore, the evolution of the respiratory system has involved the development of a complex system of progenitor cells that replace damaged cells.

In prototypical and common adult lung diseases, damage inflicted by the disease overwhelms the normal repair process of the lung resulting in impaired repair and chronic scarring. In an interstitial disease such as idiopathic pulmonary fibrosis (IPF), chronic injury is characterized by ongoing damage to alveolar epithelial cells and the progressive formation of fibroblastic foci and destruction of the lung tissue.3–9

In COPD, an example of an airway disease, there is chronic injury to the epithelium, resulting in chronic inflammation, goblet cell hyperplasia, mucus gland metaplasia and airway remodelling.2 The elevation of matrix metalloproteins (MMPs) results in the breakdown of elastin, an important factor for alveolar regeneration.10 In addition, smoking results in direct damage to the structural cells and their supporting matrix of the lung resulting in widespread apoptosis of endothelial, bronchial and alveolar epithelial cells.2 Furthermore, cigarette smoke impairs epithelial repair by reducing epithelial cell proliferation and migration as well as altering extracellular matrix deposition.11,12 Significantly, the apoptosis of these structural cells is present after smoking cessation.13,14 This in part explains the self-perpetuating cycle of cell death and a fall in lung function in certain subjects with COPD even when smoking has stopped.15 Furthermore, cigarette smoke impairs fibroblast function which is a central mesenchymal cell for lung regeneration.16,17 In addition, to the increased apoptosis, defective clearance of apoptotic cells from the COPD lung also appears to play a role.18

A strategy to treat chronic lung diseases would be to replace damaged epithelial and endothelial cells, thereby supporting the endogenous repair and shifting the balance from impaired repair to normal repair (Fig. 1). Several cells have been shown to be candidates for this function.

Figure 1.

The rationale for cell therapy. When the progenitor cell function of the lung is overwhelmed by injury, there is suboptimum replacement of epithelial and endothelial cells. This would result in impaired repair. If cell therapy can assist in replacing damaged epithelial and endothelial cells following significant injury, then this may restitute normal lung tissue, thus promoting normal repair with reduced fibrosis. The possible mechanisms include immunomodulation, epithelial and endothelial repair and the release of trophic factors that augment repair and prevent fibrosis.


Endogenous progenitor cells

The biology of endogenous progenitor cells is extensive and beyond the scope of this review. Figure 2 demonstrates the distribution of progenitor cells within the lung. The field of endogenous stem cells has undergone rapid advances, and much of our knowledge is based on the pathways during foetal development of the lung, animal models of lung injury as well as the cellular dynamics within explanted tissue.

Figure 2.

The progenitor cells in the lung. The several progenitor cells that have been identified at various parts of the lung. These cells include ciliated and Clara cells in the trachea, basal cells in the main bronchus, neuroendocrine cells in the bronchus as well as bronchoalveolar duct junction cells and alveolar epithelial cells in the distal lung. (Source: Rawlins, EL, Hogan, BL. Epithelial stem cells of the lung: privileged few or opportunities for many? Development 2006; 133: 2455–2465).

The airway is covered by epithelium with a pseudostratified epithelium in the larger airways and columnar as well as cuboidal in the small airways, while the distal lung consists of the alveoli. The main cell types are ciliated, columnar, undifferentiated, secretory and basal (Table 1)

Table 1.  Progenitor cells within the lung
Submucosal glandLarge airwaySource of stem or progenitor cells for cells of the large airway19–21
Trp-63/cytokeratin-5/cytokeratin-14 subpopulation of basal epithelial cellsLarge airwayProgenitor cells for the pseudostratified epithelium22
Type A Clara cellsSmall airwayProliferates and differentiates into mature Clara cells and differentiated ciliated epithelial cells. Following repair returns to a normal quiescent Clara cell state2323
Variant Clara cellsSmall airwayReconstitutes the bronchiolar epithelium2323
Non-ciliated Clara cell secretory protein-expressing cellsNeuroepithelial body, small airwayFacultative progenitor cells of the bronchiole2424
Pulmonary neuroendocrine cell-specific marker – calcitonin gene-related peptide-expressing cellsNeuroepithelial body, small airwayFacultative progenitor cells of the bronchiole2424
Scgb1a1 + Clara cellsNeuroepithelial body, small airwayProliferates in response to injury and constitutes the facultative progenitor cells of the bronchiolar epithelium25
Bronchoalveolar stem cells24,26Bronchoalveolar duct junctionCapacity for bronchiole and alveolar differentiation24,26
Alveolar epithelial type 2 cellsAlveolar epitheliumProgenitor cells for alveolar epithelial type 1 cells27–30
octamer-binding transcription factor 4 (mice)-positive cellsAlveolar epitheliumProgenitor cells in the alveolar region of the lung3132

Progenitor Cell: A collective term used to describe any proliferative cell that has the capacity to differentiate into different cell lineages within a given tissue. Unlike stem cells, progenitor cells have limited or no self-renewal capacity. The term progenitor cell is commonly used to indicate a cell that can expand rapidly, but undergoes senescence after multiple cell doublings.33

Facultative progenitor cell: A cell with the proliferative capacity that has functional properties of differentiated cell types in its quiescent state. Examples include bronchiolar Clara cells. Facultative progenitor cells in their active proliferative state exhibit many properties of a transit amplifying cell.33

Given chronic injury may result in the depletion of endogenous progenitor cells that may lead to chronic fibrosis, it is compelling to suggest that endogenous progenitor cells may be used as cellular treatment for lung disease. Proof of principle does exist as intratracheal instillation of AT2 cells into Bleomycin-injured mice reduced injury and collagen deposition.34 Furthermore, Prominin-positive alveolar progenitor cells also attenuated Bleomycin-induced lung injury in mice.35 However, major challenges do exist in using endogenous progenitor cells for therapy. There may be difficulties isolating cells from diseased individuals with the limited number of cells in the damaged lung.

Haematopoietic stem cells and whole bone marrow cells

There are many studies that have utilized circulating cells as a treatment in animal models of lung disease. These early studies were based predominantly on histological techniques to identify haematopoietic stem cells or whole bone marrow with or without myelo-ablation of recipient bone marrow. The systemic administration of male marrow to female mouse recipients was also employed and these initial studies suggested that circulating cells could engraft and differentiate into mature airway and alveoli epithelial cells. Prior lung injury augmented engraftment but this was not the case in all studies.36 Furthermore, chimerism of airway epithelial cells in the lungs of sex-mismatched clinical bone marrow and lung transplant recipients was found suggesting that similar events might occur in human lungs.37,38

However, the accuracy of these histological techniques has been brought into question as some of these cells may have been superimposed on recipient lung cells. In addition, donor-derived leucocytes were not excluded and autofluorescence was another confounding factor.39,40 Despite these limitations, several studies continue to suggest engraftment of donor-derived cells in the lung.41,42 However, certain variables between these studies need further clarification. To this end, chronic injury seemed to favour engraftment while the age of the mice did not influence engraftment.43–45 The timing of injections also influenced engraftment. The mechanisms by which circulating cells acquire a lung phenotype also remain to be elucidated. Potential mechanisms include cell fusion, soluble factors released by the injured lung or the horizontal transfer of mRNA and protein by membrane-derived microvesicles.46–48 Recently, a study by Wong et al. has identified a bone marrow cell that expresses Clara cell secretory protein (CCSP). In addition, the CCSP-expressing bone marrow cell expressed markers of alveolar epithelial type 1 cells (AT1), AT2 and basal epithelial cells. These cells selectively homed to sites of injury in the naphthalene-damaged murine model of airway injury.49 These findings, while not yet confirmed in other laboratories, continue to suggest that a bone marrow-derived cell might play a major part in the restitution of damaged airway cells.

Endothelial progenitor cells

Another bone marrow-derived cell, endothelial progenitor cells (EPCs) have shown promise in the management of lung disease as levels in the circulation correlate with the pathogenesis of several lung diseases.50 The EPCs are very similar to embryonal angioblasts and differentiate into mature endothelial cells. Based on markers expressed, EPCs are divided into early EPCs (CD34/CD31/CD14) that show growth in vitro, do not form tubes in matrigel-forming assays and release high levels of cytokines. In contrast, late EPCs are CD31-positive, form tubes and do not secrete cytokines.51 As such, the both cell subtypes may have different roles in repair with early EPCs contributing to initial vascular repair while releasing cytokines with paracrine trophic effects, while late EPCs may restore blood vessels.

Circulating EPCs have been identified and isolated from human subjects. Levels in the circulation are reduced pulmonary hypertension, COPD, acute lung injury and lung cancer.52,53 EPCs are raised in ARDS (increased number is associated with survival), non-small cell lung cancer (associated with worse survival) and asthma (no clinical correlation).54–57 Hypoxia causes increases in EPCs while hyperoxia results in lower levels.58

EPCs would benefit lung diseases by inducing repair of blood vessels. Blood vessel injury is a primary event in many lung diseases, and re-endothelialization would improve tissue perfusion as well as preserving the capillary-epithelial barrier, thus preserving functional lung units.

Mesenchymal stem cells

Mesenchymal stem cells (MSCs) are probably the best studied of all the adult stem cell groups. These cells are characteristically of stromal origin; self-renew and can differentiate into several cell lineages. Notably, MSCs may be derived from different sources including the bone marrow, umbilical cord blood, Wharton's jelly of the umbilical cord, placenta, adipose tissue and lungs from lung transplant patients.59,60 Notably, as there is overlap in functions such as transdifferentiation between human bronchial fibroblasts and MSCs. This prompted an international committee to set minimal criteria for the isolation and characterization of MSCs.60 These include:

  • • Plastic adherence in standard tissue culture conditions
  • • Expression of CD73, CD90, CD105
  • • No expression of CD11b, CD14, CD19, CD34, CD45, CD79a or of HLA-DR
  • • Differentiation in vitro to osteoblasts, adipocytes and chondroblasts

MSCs have shown to be effective in treating several animal models of lung disease and is reviewed by Iyer et al..61 In this section we will limit discussion to bone marrow MSCs (bmMSCs) while other sources of MSCs will be discussed under the relevant section. Studies of note include demonstration that intratracheal administration of MSCs decreased pulmonary hypertension in monocrotaline-induced pulmonary hypertension while bone marrow cells decreased elastase-induced emphysema by upregulating hepatocyte growth factor.62,63 Furthermore, MSCs have been shown to reduce inflammation and augment alveolar epithelial survival in animal models of chronic neonatal lung disease.64,65 MSCs possess several characteristics that allow them to repair injured lung. These cells are safe to inject, traffic to sites of injury, are anti-inflammatory and anti-oxidant.

MSCs have low expression of HLA 1 and low HLA II and the absence of co-stimulatory molecules CD40, CD80 and CD86.66,67 Furthermore, MSCs are able to selectively migrate to areas of injury. This is mediated by the wide MSC expression of chemokine and cytokine receptors.68In vitro studies showed that direct contact or separation of injured lung cells from bmMSCs resulted in chemotaxis to the injured cells and reduced the levels of inflammatory cytokines in the supernatant.69 The anti-inflammatory actions of MSCs may be due to the reduction in proliferation of lymphocytes, natural killer cells and dendritic cells.67 Factors implicated in the attenuation of inflammation include secretion of TGF-β, PGE2, HGF, IDO, B7H1, IL-10 and NO70 (Fig. 3).

Figure 3.

Mesenchymal stem cells (MSCs) act on several pathways. The MSCs may act as antigen presenting cells and interactions with natural killer (NK) cells may activate the innate immunity pathways. By the release of several factors, the MSCs exert anti-inflammatory influences on the inflammatory cells. This includes inhibition of cell proliferation, activation and differentiation. (Source: Stapp J. Immune regulation by mesenchymal stem cells: two sides to the coin. Tissue Antigens 2007; 69: 1–9).

A major hurdle to treating lung disease is oxidative stress generated during inflammation that has adverse outcomes on healthy tissue. The reactive oxidant species (ROS) include H2O2, O2 and NO that serve to kill pathogens, remove debris and mediate cellular signalling. However, in pathological states increased levels of ROS are responsible for significant tissue destruction. Notably, adipocyte-derived MSCs demonstrate anti-oxidant levels equivalent to 100 µg of ascorbic acid.71 bmMSCs stimulated by adrenaline or cultured at lower temperature produced reduced levels of ROS, lower levels of NO and markers of oxidant stress such as malondialdehyde which resulted in decreased apoptosis of MSCs exposed to an oxidant state.72 This implies that MSCs possess anti-oxidant properties that may be useful when delivered to sites of lung injury.

Human bmMSCs have shown the ability to express markers of lung epithelial cells. bmMSCs, when cultured with small airway epithelial cells, demonstrated epithelial-like morphology. This occurred through direct differentiation or cell fusion.73 Indeed, cell fusion is promoted by severe tissue injury.46 Given the ability of bmMSCs to differentiate, this may provide a valuable avenue for epithelial restitution by MSC therapy following lung damage.

Gene delivery may be another function of MSCs that would aid in the treatment of lung disease. Human bmMSCs underwent an epithelial-like morphological change and expressed epithelial markers including the CFTR gene when co-cultured with airway epithelial cells (AECs). In addition, MSCs from patients with cystic fibrosis were CFTR gene-corrected and these gene-corrected MSCs were able to contribute to apical chloride on stimulation with cAMP. This suggests that MSCs from cystic fibrosis patients could be harvested, gene-corrected and used for autologous cellular therapy for this disease.74

Little is known about how MSCs mediate the reparative process in animal models of lung injury. However, recently candidate molecules have been identified such as the inhibition of IL-1Ra,75 the expression of Angiopoeitin-1 and production of keratinocyte growth factor.76 There is a strong possibility that there are many more pathways that mediate the actions of MSCs, and further studies elucidating these mechanisms would be vital to future therapeutic strategies in the treatment of lung disease.

Embryonic stem cells

Embryonic stem cells (ESCs) were first derived from the inner cell mass of the mouse embryo in 1981, and are able to differentiate into cells or tissue derived from three germ cell layers.77 As such, they represent a major source of tissue for therapeutic applications that involve the repair and replacement of damaged tissue. ESCs have a high nucleus : cytoplasmic ratio and high levels of telomerase activity that enables these cells to proliferate indefinitely. Surface markers characteristically expressed by undifferentiated ESCs include stage-specific embryonic antigen-4 (SSEA-4), Trafalgar antigen (TRA-1-60, TRA-1-81), octamer-binding transcription factor 4 (Oct4) and alkaline phosphatase.78–80

A major challenge is to keep ESCs undifferentiated in culture in order to maintain a homogenous cell population that could be consistently differentiated into other tissue types. Recently, FGF-2 and members of the TGF super-family were found to fulfil this function.81–83 This is a major advance in standardizing ESC research. The differentiation of ESCs to lung tissue remains a challenge. There has been some success in mouse studies. Based on normal lung development, the formation of endoderm from ESCs is the central step in the formation of lung tissue. To this end, definitive endoderm has been generated down the Nodal pathway and by Activin A in low serum culture medium. Notably the injection of these definitive endodermal cell lines into severe combined immunodeficient (SCID) mice resulted in the formation of tissue derived from endodermal lineages.84

There are very limited studies of the differentiation of human embryonic stem cell (hES) down a respiratory lineage. Culture of human embryoid bodies in small airway growth media resulted in the expression of surfactant protein C (SPC) and lamellar bodies.85 Wang et al. was able to obtain type II cells from embryonic bodies (EB) following culture in conditioned media. The differentiated cells demonstrated markers of type II cells, as well as production of surfactant proteins and complement proteins C3 and C5.86 These findings suggest significant functional differentiation into AT2 cells. Data from the same group further demonstrated that hES-derived AT2 cells abrogated Bleomycin-induced lung injury. Notably, hES-derived AT2 cells persisted in the lung at 300 days post Bleomycin injury and comprise about 5% of the total cell population of the lung. In addition, a proportion of these hES-derived AT2 cells had differentiated into AT1 cells.87 The authors suggest that a denuded epithelial basement membrane is critical for trafficking and engraftment of these cells. The reduction in lung fibrosis may be due to re-epithelialization of damaged alveolar epithelial cells and inhibition of fibroblast ingression into the injured lung. However, the authors suggest that human embryonic cell (hEC)-derived cells may also release trophic factors that are responsible directly for repair or may augment progenitor function within the lung as repair was present in areas not populated by these cells.

Induced pluripotent stem cells

Researchers in 2006 described the generation of pluripotent stem cells by the retroviral transfection of fibroblasts with four transcription factor genes: Oct4, Sox2, Klf4 and c-Myc.88 This experiment was successfully reproduced in human dermal fibroblasts using a set of transcription factors including Oct4, Sox2, Nanog and Lin28.89–91 A comparison of mouse and human induced pluripotent stem cells (iPS) together with ESCs demonstrate that these cells are very similar to ESCs with regards to morphology, growth characteristics and patterns of gene activity.91 Furthermore, iPS differentiated into all three germ layers when injected into SCID mice and formed competent adult chimeras when injected into 2N or tetraploid blastocysts.92 In relation to disease states, iPS has been shown to correct sickle cell anaemia, Parkinson's disease in animal models while skin fibroblasts from an 82-year-old woman with motor neuron disease differentiated into neuronal cells.93,94

Although the development of these cells represents a major step forward in the generation of ‘personalized stem cells’, major limitations remain. These include the low efficacy of the development of iPS cells in culture of around 0.3% due to the cells in large part trapped in a partially reprogrammed state.89 This occurs as a result of repression of differentiation-specifying transcription factors and incomplete chromatin modification.95 Recently, the efficacy of reprogramming of somatic cells to iPS cells has been improved by the introduction of small molecules that inhibit chromatin-modifying enzymes such as DNA methyltransferase inhibitor 5-aza-cytidine or small interfering or short hairpin RNAs against Dnmt1.96 Somers et al. has recently generated iPS for a variety of lung diseases. The researchers transfected fibroblasts with a single, polycistronic excisable lentivirus vector expressing Oct4, KLF4, Sox2 and cMyc.97 The process from skin biopsy to clones took approximately 3–4 months. In addition, they were able to generate definitive endoderm from these clones, the germ cell layer from which the lung is derived, thereby confirming that iPS may have the potential to differentiate into lung. Although the hSTEMCCA-loxP vector was excisable, a ∼200 bp inactive viral insert remains in the host genome which may lead to mutagenesis.

Placental stem cells

The human placenta is also an important source of MSCs and numerous groups worldwide are assessing the potential clinical applications of these cells. The placenta is a foetally derived organ and provides the essential nutrient, gas and waste exchanges required for optimal fetal growth. The abundance of this medical waste product, easy non-invasive access following term delivery, minimal ethical and legal issues makes it an attractive source of stem cells.98 MSCs are readily harvested from placenta, the associated fetal membranes and umbilical cord. Sueblinvong et al. showed that MSCs from umbilical cord blood (ucbMSCs) when cultured in specialized growth media or specific factors including keratinocyte growth factor and retinoic acid can be induced to express CCSP, CFTR, SPC and thyroid transcription factor-1 mRNA as well as CCSP and CFTR protein. In addition, ucbMSCs were easily transduced by recombinant lentivirus vectors to express CFTR. Following transplantation of these cells into NOD-SCID mice, rare cells were detected in the airway epithelium that expressed cytokeratin and human CFTR.99 Interestingly, MSCs derived from the Wharton's jelly—a matrix that supports the blood vessels of the umbilical cord, failed to transdifferentiate into alveolar epithelial cells in vitro. Following transplantation into Bleomycin-injured NOD-SCID mice, Wharton's jelly MSCs were detected at 2 weeks in lungs but were absent by 4 weeks. Importantly, transplantation of these cells was accompanied by reduced inflammation, TGF-β and Smad-2 signalling, reduced collagen content that may be partly due to an enhanced collagen degrading environment due to increases in MMP-2 and a reduction in their inhibitors the tissue inhibitor of metalloproteinases (TIMPs).100 The fetal membranes—amnion and chorion also harbour MSCs. Further, the epithelial layer of the amnion has been shown to have features of pluripotent/multipotent stem cells.101,102 Preparations of human chorion-amnion MSC in combination with amnion epithelial cells have also been tested in immunocompetent mice injured with Bleomycin. Although detection of human DNA 2 weeks post transplantation does not conclusively show the presence of intact cells, the pooled amnion-chorion stem cell transplantation was associated with reduced neutrophil infiltration and fibrosis in Bleomycin-injured lungs.103

In addition to placental tissues, the amniotic fluid is also a source of stem cells. The fluid consists of secretions from the amnion and fetal urine. Amniotic fluid plays a vital role cushioning the fetus and enables symmetrical fetal growth. Cells shed from the fetus and amnion membrane can be obtained through amniocentesis during midpregnancy and expanded in vitro. These cells are used routinely for prenatal genetic testing. Expanded cells that have a fibroblast-like morphology have been shown to have pluripotent/multipotent stem cell characteristics. A recent study using a murine hyperoxia injury model found that amniotic fluid stem cells (AFSCs) engrafted in the lung epithelium with some cells producing surfactant protein-C and a naphthalene injury model showed evidence of Clara cells derived from the transplanted AFSCs.104 Further, human cells derived from the amnion, human amnion epithelial cells (hAECs), reduced Bleomycin-induced lung injury and fibrosis.105 In addition, hAECs adopted a lung phenotype when cultured in small airway growth medium and when injected into injured mouse lung. The hAECs not only stained for surfactant proteins but also secreted surfactant protein D (SPD), thereby suggesting functionality similar to AT2 cells. Possible mechanisms for augmented repair of the injured lung include the hAEC-mediated reduction of inflammatory cytokines. The attenuation of fibrosis by hAECs maybe due to an hAEC-induced fall in TGF-β, re-epithelialization of the alveolar epithelium and the upregulation of MMP-2 and downregulation of TIMPs suggesting a pro-degradative environment that may reduce the collagen content of the lung.

Taken together, the candidates have to date demonstrated differing extents of differentiation and repair as summarized in Figures 4 and 5.

Figure 4.

The levels of differentiation demonstrated by different candidates for lung cellular therapy. This schematic shows the level of differentiation of various cells. Certain cell types adopt just morphological features and express markers of lung cells. These include mesenchymal stem cells and haematopoietic cells. Endothelial progenitor cells, placental stem cells and embryonic stem cells demonstrate some functionality as lung cells.

Figure 5.

The extent of lung repair by different candidates for lung cellular therapy. The candidates for lung cellular therapy have shown in vivo reduction of lung inflammation and fibrosis. These include haematopoietic cells, whole bone marrow cells, embryonic stem cells and placental stem cells. However, only mesenchymal stem cells and endothelial progenitor cells have entered into clinical trials.


Animal model evidence and limitations

The treatment of animal models using stem cells constitutes a vital proof of principle for the utility of cellular therapy in human subjects (Table 2). However, murine models have major limitations. The Bleomycin-induced model of lung injury represents ARDS more than IPF118. This model has been treated successfully by many agents such as angiotensin converting enzyme (ACE) inhibitors, which do not have any impact on the treatment of ARDS or IPF. Therefore, the effective treatment of mouse models by stem cells must be regarded as useful but may not translate into viable and effective therapies in human subjects. This may be due to a greater extent of injury in humans versus mice. Furthermore, Bleomycin-induced lung injury is a self-resolving insult while human disease is generally progressive over many years.118 Therefore, the duration and extent of pathology may be milder in the mouse than humans. The lack of human models of lung disease for the evaluation of cell therapies is a major limitation.119 Recently, an in vitro model of ARDS incorporated a human lung being ventilated and perfused in vitro, was used for the LPS-induced lung injury. Human bmMSCs injected into the pulmonary circulation found this treatment improved the clearance of water from the alveoli spaces. The very short-term viability of explanted human lung tissue is a major limitation for lung disease. However, the model albeit investigated over a few hours, provides valuable insights into the response of a human lung to cellular interventions.113

Table 2.  The uses of stem cells in specific conditions
DiseaseModelsCell therapyOutcomeMechanismsReferences
COPDElastase-treated tight skin miceBone marrow cellsReduced emphysematous changeMinimal engraftment, trophic106
COPDIrradiation + 8% papainBone marrow-derived stromal cellsReduced airspace enlargement and reduced apoptosisMinimal engraftment, trophic effects107,108
Cystic fibrosisCFTR-mice +  irradiationWhole bone marrowLow engraftment. Modest correction of CFTR defectNormal CFTR function in bone marrow cells modestly improves chloride secretion.109
AsthmaOvalbumin sensitizationAdipose derived-stromal cellsReduced airway hyper-responsiveness, eosinophilia and inflammationReduced T-helper cell (Th)-2 response and increased Th1 response110
AsthmaRagweed sensitizationBone marrow stromal cellsReduced airway hyper-responsiveness, inflammationIncreased TGF-β levels and T regulator cells suppress inflammation111
Acute lung injuryLPS sensitizationBone marrow stromal cellsImproves alveolar clearance, reduces inflammationRelease of IL-10, angiopoetin-1, keratinocyte growth factor, activation of RhoGTPase Cdc4276,112,113
Fibrotic lung diseaseBleomycinBone marrow stromal cells, human umbilical cord cells, human amnion epithelial cellsReduces inflammation and fibrosisUpregulation of matrix metalloproteins, epithelial restitution, IL-1α receptor antagonist, immunomodulation75,100,114
Pulmonary vascular diseaseMonocrotaline, hypoxiaMesenchymal stem cells, endothelial progenitor cellsReduces pulmonary vascular resistance and improves right ventricular functionVEGF, eNOS, calcitonin gene-related peptide62,115,116
Lung cancer Endothelial progenitor cells and mesenchymal stem cells home to tumourPotential for delivery of anti-tumour agents 117

Given the wide range of stem cells that has been investigated in the treatment of lung disease, there remains a strong possibility that cellular therapies would form a major limb of therapy for lung diseases. Although major progress has been made, there are serious limitations that need to be addressed before cellular therapy becomes commonplace in conditions affecting the lung. In this section, we discuss the properties that an ‘ideal stem cell’ should have for clinical purposes.

Bioethical controversies

ESCs remain controversial given the derivation of these cells leads to the destruction of an embryo. In addition, the number of embryos that are available from in-vitro fertilization (IVF) limits ESCs. To this end the adult stem cells are more easily available. In terms of MSCs the bmMSCs require invasive biopsy procedures. This also applies to MSCs acquired from organs. Placentally derived cells would be the most readily available. Placental MSCs, cord blood MSCs, umbilical cord MSCs and amnion epithelial cells are all derived from products of delivery following labour. However, recent studies suggest iPS with pluripotent properties may be a replacement for ESCs.120

Genetic manipulation

Genetic manipulation of stem cell may give rise to chromosomal instability and long-term safety issues when injected into human subjects. MSCs and placental stem cells do not require genetic manipulation for the rescue of animal models of lung disease. However, there is compelling rationale to transduce the cells to express therapeutic substances. Differentiated cells from human ESCs also require little genetic manipulation although data for the rescue of lung animal models by differentiated endodermal or lung cells are lacking. However, iPS is generated by retroviral transfection of pluripotent genes and transcription factors.89 There are now ways to remove the retroviral sequences and/or to use non-viral transduction. However, some sequences inserted into the genome of iPS may cause insertional mutagenesis and potential tumour formation. As such, it is unknown how these cells will behave for the duration of the lifetime of the human subject. These important safety issues need to be clarified.

Ease of culture and propagation

The use of stem cells for the treatment of human disease would require massive amounts of cells for this express purpose. Although studies investigating the treatment of animal models use relatively small number of cells, the translation to human beings would necessitate stem cells be easily propagated at commercial quantities. All stem cells reach senescence in culture. However, it is unlikely that stem cells at very late passage number would be utilized clinically. MSCs are cultured without the need for complex culture techniques. Indeed, there are clinical trials utilizing MSCs cultured on a large scale. This applies to MSCs from placental tissue but the epithelial components of placental stem cells may have a more limited passage number. ESCs involve more complex but increasingly standardized protocols. Given the intrinsic properties of ESCs, they may be cultured for many passages. Recent advancements in iPS technology has resulted in a cell line been created in months; however, little is known about long-term culture of these cells for clinical uses. It is unknown whether each cell line, irrespective of stem cell derivation, would maintain consistency with regard to differentiation, markers and chromosomal stability in culture.

Differentiation to lung

The literature contains many studies that claim the stem cells used have differentiated or expressed markers of lung epithelium. We will limit the discussion to human cells. MSCs and placental stem cells have been shown to express markers of lung epithelium following injection into animal models of lung disease.99 ESCs have shown features of lung epithelium in vitro; however, iPS differentiation into lung still remains to be explored.86

The criteria to confirm differentiation of a stem cell into a lung cell needs further clarification. In our opinion, a differentiated cell must display morphological appearances of a lung cell, express the markers and most importantly retain functional properties of a lung cell. Most studies have studied morphology and the expression of markers while studies in functionality are limited. In terms of functional studies, although there are reported data that human stem cells engraft in animal models albeit at very low number, this would not be sufficient to rescue human subjects with significant lung disease.

Trafficking and engraftment

It is becoming more apparent that it may be critical for administered cells to traffic to the lung but efficacy is not dependent on engraftment in tissue but on immunomodulation as in the case of MSCs. Engraftment however was the subject of many studies, and stricter criteria for engraftment have been developed since the vast majority of systemically administered cells would travel to the lung and get trapped in the microvasculature. These cells may be within the blood vessels and not incorporated into the resident tissue. As a result, many studies that were reporting engraftment were reporting false positive images. Cells were trapped in the vasculature, superimposed on resident cells or autofluorescence misinterpreted for GFP positivity. De-convolution and/or confocal microscopy are recommended to avoid the issue of superimposed cells.

Given that therapeutics can be administered via the airway or systemically in the lung, deciding on the route of administration is an important question to answer. From murine models, most studies use systemic instillation. However, the intratracheal route has been used with varying success with intratracheal neonatal mouse lung fibroblasts resulted in alveolar and interstitial engraftment at sites of elastase-induced lung injury, while intratracheal alveolar epithelial progenitor cells resulted in reduced lung injury.35,108 In addition, the intratracheal route has been used successfully in rescuing the naphthalene-induced model of airway injury by CCSP-expressing bone marrow cells. The larger human airways allow greater access to the lung and intuitively the intratracheal route may be valuable for airway disease. However, the barriers of innate immunity, including the mucus biofilm in several airways disease, may prevent adequate re-epithelialization of the damaged airway. The systemic route exploits the chemokine and adhesion molecule signalling in getting to sites of injury and may be preferable to conditions affecting the alveolar epithelium and endothelium.

The major limitation that has been found in the many investigations is the very low levels of engraftment of injected cells in lung tissue. This is in the order of <1% in general. Therefore, the physiological significance of such low rates of engraftment is unlikely to be significant. Extrapolation of animal studies to human disease suggest that if these low rates of engraftment are maintained, then at present cellular therapies may be ineffective in replacing damaged tissue to prompt recovery following significant injury. Therefore, studies should be directed at improving delivery in the human lung as well as improving levels of engraftment and phenotypic conversion.

Survival in an inflammatory milieu

Inflammation is a prominent feature of lung injury. MSCs and placental stem cells have demonstrated anti-inflammatory properties already discussed. It is unknown whether ESCs or iPS-derived lung tissue would demonstrate similar anti-inflammatory properties. Given that MSCs do not have the same differential potential as ESCs and ESCs may not posses the same anti-inflammatory properties as MSCs, it is compelling to postulate that combination cellular therapies may be an effective way of treating lung disease.

Inflammation also results in an oxidant stress and an ostensibly harsh milieu for cells within the damaged lung. It is therefore important for the cell used for therapies to be resilient to these environments. The inability to survive these circumstances would result in cell death of the treating cells and an ineffective outcome. Although adipocyte MSCs possesses anti-oxidant properties, it is unknown whether this property makes them more resilient to the inflammatory environment. Differentiated lung epithelial cells from ESCs or iPS may be just as vulnerable to the ambient environment as endogenous epithelial and endothelial cells thus limiting the ability of these cells to replace injured lung.

Reversal of fibrosis

MSCs and placental cells have been shown to rescue animal models of fibrosis. These were achieved by early injection within 24 h of injury. However, injection at 5 days post injury had no effect on subsequent collagen deposition. Of note, injection of Flk + MSCs at 30 days and 60 days post injury increased fibrosis and were located in areas of scarring.121 This suggests that the timing of MSC injection is critical to the outcome. Early injection may reduce inflammation and preserve lung epithelium and endothelium, thereby reducing downstream fibrosis. Late injection may result in the trafficking of MSCs to areas of fibrosis and an increase in pro-fibrotic mechanisms. Recent studies have shown that MSCs increase fibrosis through the Wnt signalling pathway.122 It is unknown whether the late injection of placental stem cells would augment fibrosis. Differentiated stem cells from ESCs and iPS cells would probably behave in a similar manner to endogenous epithelial and endothelial cells. Alveolar epithelial progenitor cells injected intratracheally reduced Bleomycin-induced fibrosis while endothelial precursor cells reduced pulmonary hypertension.35,123 This supports the concept that alveolar epithelial and endothelial precursors derived from ESCs and iPS may be beneficial in lung injury.

Tumour formation

Tumour formation arising from administered cells is a major consideration in determining the viability of stem cell therapy. There are suggestions that MSCs may play a part in carcinogenesis. Notably, mouse bmMSCs may develop sarcomas while human bmMSCs cultured for 4–7 months developed chromosomal instability.124 Given the data on trafficking of bmMSCs, it is unlikely that bmMSCs would be present in the lung for 4–7 months.125 In a mouse model of gastric cancer, mice were injected with GFP-bmMSCs and then infected with Haemobartonella felis. The bmMSC-derived intestinal epithelium developed cancer suggesting a predisposition to malignancy.126 Furthermore, MSCs may also contribute to myofibroblasts and fibroblasts of tumour stroma.127 However, there is increasing evidence that exogenously administered MSCs home to and is incorporated into lung tumours. For example, there was trophism of CXC3CL1-expressing MSCs which resulted in reduced growth of metastatic tumours in a murine model of lung cancer.128 This opens up a potential new avenue for the treatment of lung cancer, whereby MSCs can deliver therapeutic agents such as tumour necrosis factor-related apoptosis-inducing signal to the tumour.129 There is no data regarding the malignant transformation of placental stem cells. However, it is well known that ESCs but not differentiated tissue derived from ESCs form teratomas in vivo. It would take just one incompletely differentiated ESC to generate teratoma formation in human subjects. Therefore, rigorous screening of ESC-derived cells would be needed. iPS may form tumours such as rhabdosarcomas perhaps related to retroviral transfections.130 Taken together tumour formation needs to be effectively excluded in cells destined for human therapy. Long-term studies in culture and in animal models are needed.

Reduced immunogenicity

All nucleated cells express MHCs on the cell surface. These MHCs are recognized by antigen presenting cells.61 Incompatibilities between antigen presenting cells and donor cells would result in an inflammatory response designed to eliminate the ‘foreign’ cells. Thus rejection is a major limitation to cellular therapy. However, the allogenic and xenogenic use of MSCs and placental stem cells has not resulted in significant rejection, thereby allowing for the wider therapeutic use of these cells. Furthermore, iPS may be generated from the individual patient and would therefore bypass any rejection reactions.


Given the overall shortage of donor lungs, many people awaiting lung transplantation will die before receiving transplant. As such, engineering tissue from stem cells to generate lung tissue for this purpose is a compelling option. Stem cells would be cultured on synthetic or natural matrix or scaffolds that is usually biodegradable to guide differentiation along certain pathways in a three-dimensional environment.131 To this end, natural media including collagen, matrigel and gelfoam have been used. Several studies using natural scaffolds were successful in culturing alveolar epithelial cells. Initial work demonstrated SPC production when primary type 2 pneumocytes were cultured in three-dimensional collagen.132 Furthermore, fetal progenitor cells in collagen gel stimulated by FGF-2/7/10 induced the budding of epithelial structures and formation of endothelial networks.133 The formation of alveolar-capillary membrane connections is critical for a normal functioning lung. Therefore, investigators made efforts in engineering this basic anatomical and physiological unit of the lung. Rabbit fetal lung tissue cultured on Englebreth-Holm-Swarm tumour membrane produced type 2 pneumocytes and resulted in epithelial and endothelial cell development following implantation directly into the lung or abdominal wall but no significant alveolar-capillary connection.134

The use of gelfoam scaffolds has also been encouraging. Gelfoam scaffolds with rat fetal alveolar epithelial cells were implanted directly into the lung parenchyma. New alveoli grew at the border of the tissue and scaffold and not many structures within it.135 Furthermore, labelled fetal cells seeded onto gelfoam scaffolds survived 35 days at low numbers. There was no inflammatory response and lung tissue developed at the border between scaffold and surround lung. However, there was little evidence to support the development of alveoli-capillary connections.135

The natural scaffolds, however, may be immunogenic and may not have consistent degradation rates between them. Thus, synthetic scaffolds such as polyglycolic acid (PGA) have been developed. The first in vivo uses of these PGA scaffolds were carried out in mice and sheep. In both circumstances, the scaffold was well tolerated. In order to mimic the porosity and elasticity of the lung, a scaffold has been developed incorporating inverted colloidal crystal geometry.136 Cortiella et al. plated ovine somatic progenitor cells on PGA or PF-127 scaffolds that resulted in structures similar to the distal lung.136In vivo implantation of somatic lung progenitor cells on PGA/PF-127 and explanted on nude mice developed structures similar to alveoli while constructs into a wedge resection in sheep also developed alveoli-like structures, although no histology was carried out to confirm alveolar epithelial differentiation.136 In a major advance, workers have generated for the first time a fully functional bioartificial rat lung.137 They initially de-cellularized the lung, thus developing an environment similar to a developing lung. They then seeded this scaffold with epithelial and endothelial cells. These cells proliferated and created a functional lung that was transplanted into a rat and provided normal gas exchange up to 6 h post intubation. Notably, parts of the scaffold showed features of incomplete regeneration, a limitation highlighted by the authors.


Two clinical trials have now been registered in North America for the use of cellular therapy in the treatment of COPD or of pulmonary hypertension. The 6-month interim analysis of the MSCs trial in COPD shows a decrease in circulating CRP and a trend towards improved 6-min walk. The trial is still in progress and long-term follow up has not yet been completed. Two trials have demonstrated a positive effect of EPCs on pulmonary hypertension.123,138 A larger clinical trial using EPCs transduced to express eNOS was initiated at the University of Toronto (PHACet trial). Initial data shows the reduction (nearly 50%) in pulmonary vascular resistance over the 3-day delivery period. The study is now moving to inject increasing numbers of EPCs into human subjects (23 million and then 50 million over 3 days).33


There is a growing knowledge of endogenous progenitor cell location and function in the lung. In addition, studies are now addressing the effect of lung disease on progenitor cells. The improved insights into the biology of progenitor cells during disease will enable us to develop cellular therapies that augment and compliment endogenous progenitor cells, thus effectively treating lung disease. There are many studies that demonstrate that stem cell therapies could reduce inflammation and fibrosis in animal models of lung disease. There are candidate cells that have been identified would home to and transdifferentiate into lung tissue, and clinical trials are presently underway to examine the role of stem cells in COPD and pulmonary hypertension. However, the expectations associated with cellular therapy needs to be calibrated. At present, there is no cell type that would create significant new tissue in chronically scarred lung; therefore, the measure of success in clinical trials using stem cells needs to be based on specific, realistic clinical outcomes rather than significant reversibility and regeneration of lung tissue in chronic disease.