Adult stem cells for acute lung injury: Remaining questions and concerns

Introduction

Acute lung injury (ALI) and its more severe form, acute respiratory distress syndrome (ARDS), are the most common causes of acute respiratory failure in hospitalized patients. Despite extensive research into the pathophysiology, the mortality associated with ALI/ARDS has remained up to 40%, depending on the aetiology.1, 2 Current treatment is primarily supportive with lung-protective ventilation and a fluid conservative strategy.3, 4 Pharmacological therapies that reduce the severity of lung injury in preclinical models have not yet been translated to effective clinical treatment options. Earlier concepts of distinct disease phases underlying the pathophysiology of ALI/ARDS, from an early pro-inflammatory to a later pro-fibrotic phase now appear to be an oversimplification.5 Consequently, strategies that targeted one aspect of the disease process have not been successful in limiting overall morbidity and mortality. Innovative therapies are needed.

Stem cells are undifferentiated precursor cells with the capacity for self-renewal and the ability to differentiate into cells of multiple lineages. They can be broadly classified by their potency (pluri-potent vs multi-potent) and origin (adult vs embryonic). Adult stem cells are multi-potent postnatal stem cells that remain in body tissues throughout life and have the potential to differentiate into a more limited range of mature cell types. In general, they include haematopoietic stem cells (HSC), mesenchymal stem cells (MSC), endothelial progenitor cells (EPC) and organ-specific stem cells such as endogenous lung stem cells. In 1998, Ferrari et al.6 reported in a landmark manuscript that the transplantation of bone marrow-derived adult stem cells into injured muscle tissue mitigated damage and regenerated healthy muscle fibres, generating tremendous enthusiasm for stem cell-based therapy.

Much of the initial interest in stem cell-based therapy for lung injury arose initially from the multi-potent properties of the cells. Krause et al.7 reported up to 20% engraftment of bone marrow (BM)-derived cells in the lung, including epithelial cells, from a single haematopoietic precursor. However, subsequent studies suggested that the major therapeutic effect of adult stem cells in ALI were primarily from their ability to secrete paracrine factors such as growth factors, factors regulating endothelial and epithelial permeability, anti-inflammatory cytokines and antimicrobial peptides/proteins, not from significant engraftment. To date, due to their ease of isolation and extensive preclinical studies, MSC may offer the best hope for clinical trial. However, other progenitor and adult stem cells5, 8-11 have shown some promise as potential therapeutic candidates for ALI/ARDS as well.

Despite these encouraging results, several issues remain which must be addressed.12 (i) The isolation and classification of stem cells need to be defined further, particularly concerning the issue of potency; (ii) a more precise understanding of the mechanisms underlying the therapeutic effect is needed, such as the role of the constituents of the conditioned medium, of cell-contact-dependent or independent effects, and whether the phenotype of the cells changes depending on the alveolar milieu; (iii) the optimal dose and route of cell delivery remains to be determined; (iv) although most of the current focus is on the use of MSC, we still need to determine the optimal ‘stem cells’ for cell-based therapy and (v) despite promising preclinical results and public enthusiasm, clinician-scientists involved in the translation of stem cell research into clinical trials must always keep in mind the lessons learned from the field of gene therapy.13 Above all, we must do no harm.

This review will present the most recent advances in the field of adult stem cell therapy for ALI, focusing specifically on MSC and EPC, and to issues that will arise in preparation for clinical trial. To accomplish this goal, we searched PubMed for relevant studies published up to Dec 2012. We also searched the proceedings of major relevant conferences, trial databases, the reference lists of identified trials and major reviews. We identified 192 studies from electronic databases, of which 36 studies were eligible for inclusion, based on the title and abstract. After independent assessment of the full text, 29 articles were finally considered to be eligible for inclusion in the analyses after we contacted authors for additional data. Figure 1 shows the search leading to the selection of the final 29 articles.

Mesenchymal Stem Cells

Mesenchymal stem or stromal cells are adult non-haematopoietic precursor cells derived from a variety of tissues such as the BM, adipose tissue and placenta which have been used as therapy in multiple diseases and syndromes such as myocardial infarction, renal failure and graft-versus-host disease. The International Society of Cellular Therapy has defined MSC in 2006 by three criteria: (i) MSC must be adherent to plastic under standard tissue culture conditions; (ii) MSC must express certain cell surface markers such as CD73, CD90 and CD105, but must not express other markers including CD45, CD34, CD14 or CD11b; and (iii) MSC must have the capacity to differentiate into mesenchymal lineages including osteoblasts, adipocytes and chrondoblasts under in vitro conditions.14

Mechanisms of action

Although the precise mechanisms of action of MSC remain unclear, a number of important insights from recent preclinical studies have emerged10 (Fig. 2).

Engraftment

Despite initial interest in MSC's multi-potent properties, engraftment in the lung now does not appear to play a major beneficial role.15-17 This perception was initially based on earlier studies that demonstrated that intravenous administration of MSC resulted in high pulmonary engraftment because MSC expressed adhesion molecules, such as vascular cell adhesion molecule 1 and P-selectin, which facilitated their retention in the lung18 and which was enhanced by the presence of hyaluronan degradation products in the inflamed tissue.19, 20 These results were subsequently questioned by multiple groups, who observed only engraftment of leucocyte lineages15 or low engraftment rates in lung injury models with observed rates of <1%.16, 17 Yet, several new publications21-23 have highlighted the potential of in vitro modification of MSC, which may increase lung engraftment and/or regeneration. These authors found specific subpopulations of MSC which expressed functional markers of lung epithelial cells such as Clara cell secretory protein, surfactant protein-C, cystic fibrosis trans-membrane conductance regulator and epithelial sodium channel when cultured in mouse tracheal epithelial cell medium, small airway growth medium or with keratinocyte growth factor (KGF) and retinoic acid respectively. In addition, several groups24-26 have identified adult stem cells within the lung, some with the features of MSC. Tropea et al.27 demonstrated that systemic treatment with MSC or MSC condition medium had a direct effect on stimulating bronchioalveolar stem cells to repair and restore injured lung epithelium.

Immunomodulation

Multiple studies have demonstrated that MSC possess potent immunosuppressive effects by inhibiting the activity of both innate and adaptive immune cells.28-31 This immunosuppression has shown to be mediated by cell-contact-dependent and independent mechanisms through the release of soluble factors such as tumour necrosis factor-inducible gene 6,32 prostaglandin E₂,33 interleukin-1033, 34 and interleukin-1 receptor antagonist35 among others. For example, we36 and other investigators37-39 have demonstrated that MSC may switch the phenotype of alveolar macrophages from a M1 (inflammatory) to a M2 (anti-inflammatory). Also, in an endotoxin-induced ALI model treated with umbilical cord-derived MSC, Sun et al.40 found that MSC suppressed lung injury by upregulating CD4⁺CD25⁺forkhead box P3⁺ regulatory T cells, reduced the levels of the pro-inflammatory cytokines Interferon-γ, macrophage inflammatory protein 2 and tumour necrosis factor-α and increased the anti-inflammatory cytokine, interleukin-10. Treg cells play a central role in the prevention of autoimmunity and in the control of immune responses by downregulating the function of effector CD4⁺ or CD8⁺ T cells;41-43 these findings are consistent with other reports28, 36 that MSC can decrease T-cell responses by shifting from a T-helper1- to T-helper2-type response.

Alveolar fluid clearance

We and other investigators have reported that ALI pulmonary oedema fluid contained high levels of pro-inflammatory cytokines which downregulated alveolar fluid clearance (i.e. the resolution of pulmonary oedema).44, 45 Interestingly, MSC are known to produce several epithelial specific growth factors, such as KGF, the seventh member of the fibroblast growth factor family, which are known to increase alveolar fluid clearance. We46 and other researchers38, 47, 48 have reported that KGF can reduce lung injury in animal and human models of pulmonary oedema whether from direct, indirect or infectious causes. KGF improved alveolar fluid clearance in part by upregulating α epithelial sodium channel gene expression and Na-K-ATPase activity or through increased trafficking of sodium transport proteins to the cell surface.46, 49-51

Lung permeability

The integrity of the lung microvascular endothelium is essential to prevent the influx of protein-rich fluid and inflammatory cells from the plasma, which may aggravate the ability of the lung epithelium to reduce pulmonary oedema. Several MSC paracrine soluble factors, such as angiopoietin-1, are potentially important in these effects. angiopoietin-1, a ligand for the endothelial Tie2 receptor, is a known endothelial survival and vascular stabilization factor that reduces endothelial permeability and inhibits leucocyte-endothelium interactions by modifying cell adhesion molecules and cell junctions.52, 53 We and other investigators have found that allogeneic human MSC secreted a significant amount of angiopoietin-1, which was essential to prevent the increase in lung protein permeability.54-56

Antimicrobial properties

MSC have toll-like and formyl peptide-like receptors and become activated in response to different bacterial products, suggesting that MSC may be directly involved in the innate immune response.36, 57 Recently, we found that human MSC can inhibit bacterial growth directly in part through the secretion of antimicrobial peptides, such as LL-37, which was upregulated upon bacterial stimulation,58 and antibacterial proteins, such as lipocalin-2, which improved bacterial clearance in a mouse model of Escherichia coli pneumonia.59 Mei et al.60 also recently showed that the improvement in bacterial clearance in MSC-treated septic mice following caecal ligation and puncture could be in part explained by enhanced phagocytic activity of host immune cells.

Recent Experimental Results

To date, MSC have been studied extensively in preclinical lung injury models (Table 1). Most of the injury32, 36-38, 40, 46, 55, 56, 64-69 (48%, 14 of 29) involved intra-tracheal endotoxin exposure. Rodents were the predominant small animals studied. Over 58% (17 of 29) of studies32, 35, 36, 38-40, 47, 55, 56, 58, 59, 61, 63-66, 69 used mice while 38% (11 of 29) used rats.34, 37, 48, 62, 67, 68, 70-74 Interestingly, Lee et al.46 developed an E. coli endotoxin-induced ALI model in an ex vivo human lung preparation perfused partially with whole blood. Although the preparation was short term (4 h) and did not include the perfusion of other systemic organs such as the liver or spleen, which may mount a significant inflammatory response, this human model replicated most of the early clinical features of patients with ALI.

The BM remained the most common source of MSC32, 34-38, 46-48, 55, 56, 58, 59, 61-66, 68, 71-74 (83%, 24 of 29), and most current clinical trials used an allogeneic source from the BM. It was believed that MSC are able to evade clearance by the host immune system through a variety of mechanisms including low expression of the MHC I and II proteins and lack of the T-cell costimulatory molecules, CD80 and CD86; this is often referred to as being ‘immunoprivileged’. However, recent studies have shown that MSC can express higher levels of the MHC class proteins than originally thought. In addition, Nauta et al.75 demonstrated that infusion of allogeneic MSC elicited a host response and led to graft rejection. It has now become apparent that MSC have complex interactions with the immune system. However, the alternative approach of harvesting the BM to isolate and culture autologous MSC may be problematic in acute illnesses such as ALI.

Compared with the BM, human umbilical cord-derived MSC have a faster population doubling time than BM-MSC, and such proliferation characteristics do not change even after 30 passages. In addition, umbilical cord-derived MSC showed lower expression of CD106 and human leucocyte antigens in comparison with BM-MSC.76 Therefore, umbilical cord-derived MSC are currently being studied for clinical application due to their accessibility, ease of procurement from donors and lack of ethical concerns.39, 40, 67 Another adult source is fatty tissue. Human orbital fat-derived stem cells are advantageous over BM-MSC because they can be isolated from minimal volume of redundant orbital fat tissue.69, 77 They also exhibit a higher epithelial differentiation potential than adipose-derived stem cells isolated from subcutaneous fatty tissues.70, 77 However, the anti-inflammatory ability of orbital fat-derived stem cells needs to be evaluated in other clinically relevant models such as pneumonia/sepsis. Clearly, further studies in the optimal source of MSC are needed.

Optimal Dosage and Route of Cell Delivery

The dose and route of MSC varies substantially based on different preclinical animal studies, and the optimal treatment remains to be determined. The mean dose32, 35, 36, 38-40, 47, 55, 56, 58, 59, 61, 63, 64, 66, 69 instilled during the early phase of lung injury in mice is (29.9 ± 20.4) × 10⁶ cells/kg, which is slightly higher than that of rats, (20.3 ± 22.5) × 10⁶ cells/kg (Fig. 3a, P = 0.26), suggesting that the effective administration dose is about 20∼30 × 10⁶ cells/kg. One study65 using an ALI model in C57BL/6 mice gave a particularly large dose (888.9 × 10⁶ cells/kg) of MSC just 1 h after exposure. To our knowledge, this is the only study showing the delayed effects of BM-MSC on day 28 because all others therapies were followed up to day 14. Although the administration of MSC was able to repair the lung epithelium and endothelium, reduce the amount of alveolar collapse and led to an improvement in lung mechanics, it is still unknown whether the beneficial effects reported at day 28 persist if the MSC were given later in the course of lung injury. In terms of tissue origin, the average dose of MSC derived from the BM was (28.3 ± 21.8) × 10⁶ cells/kg. No significant differences were found between the dose of MSC from the BM or other tissues such as umbilical cord39, 40, 67 ((17.2 ± 25.4) × 10⁶ cells/kg) (Fig. 3b, P = 0.42) or fat tissue69, 70 ((13.9 ± 0.8) × 10⁶ cells/kg) (Fig. 3b, P = 0.37). For most clinical trials utilizing MSC in lung disease such as for idiopathic pulmonary fibrosis (NCT01385644) or bronchopulmonary dysplasia (NCT01632475), the dose administered is approximately 1∼20 × 10⁶ cells/kg, which appears to be largely based on earlier clinical trials in GVHD, myocardial infarction, etc. Therefore, a current pilot clinical trial conducted by University of California San Francisco (NCT01775774) is underway to assess the safety of intravenous infusion of allogeneic human BM-MSC in patients with ALI/ARDS by using a dose-escalation protocol from 1 to 10 × 10⁶ cells/kg. What remains unclear is whether there is a dose effect or a therapeutic ceiling in the use of MSC for lung diseases or whether the dose is limited by the safety concern of the effect of the infusion on pulmonary vascular resistance. And perhaps more importantly, it is unclear whether a second dose of MSC is needed for the later phase of ALI (the resolution), and whether the functional phenotype of the stem cells is therapeutic at this point. Although many believe that higher doses will give a prolonged response, no actual dose response has been reported in the literature.

Most of the studies39, 58, 59 using an E. coli endotoxin or bacterial pneumonia models of ALI administrated MSC intra-tracheally while those using bleomycin-induced,35, 47, 61-63 ischaemia reperfusion-induced,34, 70, 71 ventilator-induced48, 72 or other73, 74 lung injury models delivered MSC intravenously. Although for practical reasons it may not be feasible to instil stem cells intra-bronchially in ALI patients who are hypoxemic, the optimal route of delivery is unclear. However, for patients with ALI from pneumonia, it is now known that BM-MSC possess direct antimicrobial activity through the secretion of antimicrobial peptides/proteins such as LL-37 or lipocalin-2 as well as the ability to enhance macrophage/monocyte phagocytosis of bacteria. Thus, the intrapulmonary delivery may be the most effective route to enhance bacteria clearance. On the other hand, intravenous delivery of MSC was used by the majority of studies34, 35, 37, 47, 48, 55, 56, 61-67, 69-74 (69%, 20 of 29) and may be more relevant to clinical practice; it may be more practical to give large amount of MSC suspension through intravenously.

Interestingly, Qin et al.68 developed a novel intra-pleural delivery method. They found that MSC delivered by this method can survive at least 1 month in vivo and their distribution was found to be limited to the surface of the pleurae and in the pleural cavity, forming a ‘MSC repository’ in vivo. The advantage of using the pleural cavity for MSC delivery is that it is a potential compartment that can receive larger dose of MSC without restriction. Although promising, the delivery route will need to be investigated further due to the limited number of study.

Timing of MSC Administration

Although preclinical animal models cannot replicate the natural course of ALI, MSC were typically given within 6 h following ALI in these models. In lipopolysaccharides (LPS)-induced ALI models, the beneficial effects of MSC were typically found less than 3 days following intra-tracheal delivery but the effects were prolonged when MSC were given intravenously. What remains to be determined is whether giving MSC once lung injury is firmly established or during the resolution phase where fibrosis may occur have any therapeutic effect.

For bleomycin-induced lung injury, the most optimal time of cell administration and end-point needs to be determined. Existing preclinical studies35, 47, 61 demonstrated that MSC were efficacious in ameliorating the resulting fibrosis, correlating with the early inflammatory steps in the pathogenesis of bleomycin-induced lesions,78 only when administered at the time of injury and not at later time points; it may take 2 to 3 days for MSC to produce soluble factors that modulate inflammation in vivo. For example, Saito et al.63 observed that the highest concentration of 7ND, a dominant-negative inhibitor of chemokine (C-C motif) ligand 2 (an inducer of macrophage recruitment and activation), was obtained on day 2 after MSC, transfected with the 7ND plasmid, administration. However, the serum level of 7ND was undetectable 11 days after MSC administration, suggesting that the number of MSC diminished over time in vivo.63 Ortiz et al.35 also showed that MSC significantly secreted interleukin-1 receptor antagonist, a competitive inhibitor of interleukin-1α, only after 72 h of stimulation by interleukin-1α, which correlated with the development of pulmonary fibrosis exposed to bleomycin. In addition, Aguilar et al.47 demonstrated that the early benefit of MSC in bleomycin-induced lung injury was lost at a later end-point, day 14. Consequently, MSC may reach its therapeutic peak 2 to 3 days after administration and decrease over time regardless of the time of cell delivery.

Summary for MSC in ALI

There are currently more than >300 clinical trials registered with clinicaltrial.gov testing MSC in a variety of disorders, including GVHD, Crohn's disease, acute myocardial infarction and acute kidney failure. More recently, MSC have been tested in clinical trials for lung diseases such as chronic obstructive pulmonary disease, bronchopulmonary dysplasia, pulmonary emphysema, pulmonary hypertension and silicosis.5, 79 However, as previously shown, existing preclinical studies to date have used relatively poorly defined MSC; the potential exists for identifying subpopulations80, 81 of MSC with more efficacy. Although a set of criteria for defining MSC has been developed by International Society of Cellular Therapy in 2006,14 there remains no validated method of measuring MSC bioactivity,82 a potency assay, and the duration of benefit in vivo. It may be time to revise the definition of MSC set forth by International Society of Cellular Therapy to allow better comparisons between preclinical animal studies and efficacy and the subsequent clinical trials which are underway. How potency will be defined, whether through characterization of secretion of soluble factors or through a functional assay, remains to be determined.

Other Adult Stem Cells for ALI

Conclusions

The major findings of this review of adult stem cells for ALI can be summarized as follows: (i) for MSC, the effective administration dose from preclinical studies in animals is approximately 20∼30 × 10⁶ cells/kg, regardless of the origin of the cells. Whereas, from current ongoing clinical trials (>300 for a variety of diseases/syndromes), the average dose of MSC is approximately 5∼10 × 10⁶ cells/kg; (ii) intrapulmonary delivery of MSC may be a more effective route for direct antimicrobial effect, while the intravenous route may be more practical for clinical application where the patients are hypoxic; (iii) in most preclinical studies, MSC were delivered in lung injury models early usually within 6 h following injury. Whereas, in bleomycin-induced ALI, the therapeutic effect of MSC reached its peak 2 to 3 days after administration; (iv) the beneficial effects of MSC in LPS-induced ALI model were typically found less than 3 days after intra-tracheal delivery but the effects were prolonged when MSC were given intravenously; (v) for EPC, intravenous delivery of exogenous EPC is an effective route for pulmonary endothelial-targeted therapeutic strategy, but research into the optimal dose is needed; (vi) EPC were delivered usually within 4 h following injury and (vii) although the engraftment rate was low, multiple studies suggest that EPC appear to home to the injured pulmonary endothelium and repopulate the injured tissue.

Despite some limitations in the characterization of MSC or EPC, especially concerning the issue of potency, a significant amount of preclinical data in both animal models and clinical trials suggest that stem or progenitor cell-based therapies may be beneficial in lung repair and remodelling after ALI. Although questions and concerns still remain, a novel and safe therapy for acute lung diseases might eventually emerge.

Acknowledgements

This work was supported by NHLBI Grant HL-093026 and HL-113022 and the UCSF Hamilton Endowment Funds.