Main

The incidence of premature delivery in North America is 12.5% and rising (1). Prematurity is the leading cause of perinatal mortality and morbidity, placing these neonates at high risk for long-term medical impairments such as bronchopulmonary dysplasia (BPD). First documented in 1967 by Northway et al. (2), BPD was described as a chronic lung disease following mechanical ventilation and oxygen therapy for acute respiratory failure at birth. Since then, antenatal steroids and postnatal surfactant have aided in overcoming the biochemical immaturity of the lung. These advances in perinatal care, together with more incremental improvements, enabled neonatologists to push back the limits of viability from the previous 34 wk gestation to the current ~24 wk gestation. Injury to more immature lungs changed the pathology of BPD (3,4). Today, BPD is characterized by impaired alveolar development and dysmorphic pulmonary microvascular growth, along with a lesser degree of inflammation and fibrosis compared with the original BPD (5). Injury at these earlier stages may be more challenging to prevent and increases the risk of long-term consequences, including pulmonary hypertension (PH) and early-onset emphysema (6), which add to the burden of health care (7,8). Thus, therapies that promote both lung repair and lung growth are desirable.

Recent insights into stem cell biology promise the regeneration of damaged organs. Stem cells are capable of self-renewal and differentiation into specialized cell types and thus have the potential to promote organogenesis, tissue regeneration, maintenance, and repair (9). Mesenchymal stromal cells (MSCs) attracted particular interest because of their ease of isolation, characterization, apparent multipotency, and pleiotropic effects. Adult bone marrow–derived MSCs (BMSCs) apparently differentiate into cells of various nonhematopoietic tissues. BMSC studies in various disease models, including cardiovascular and neurodegenerative disorders, demonstrated their efficacy in attenuating organ injury (10,11,12,13). The demonstration that a bone marrow–derived stem cell could differentiate into alveolar epithelial cells ignited stem cell research in the lung (14). Accordingly, preclinical studies suggested that bone marrow–derived stem/progenitor cells were capable of migrating to the injured lung to promote repair (15), and administration of exogenous BMSCs prevented lung injury in various adult lung disease models (reviewed by Weiss et al. (9)). These studies offered substantial promise to mitigate the impaired alveolar growth in experimental models mimicking BPD. The multipotency and self-renewal of stem cells make cell-based therapies appealing for providing both lung injury prevention and lung growth.

Cell-Based Therapies to Prevent Experimental Chronic Neonatal Lung Injury: Proof of Concept

In 2007, Tian et al. (16) showed that intravenous injection of rat BMSCs could ameliorate neonatal lung injury. Shortly after, two simultaneously published articles confirmed the therapeutic potential of BMSCs. Intravenously delivered BMSCs reduced alveolar loss and lung inflammation and prevented PH in hyperoxia-induced mice (17). Likewise, intratracheal delivery of BMSCs increased survival and exercise capacity of hyperoxia-exposed rats while attenuating alveolar and vascular injury and PH (18). Subsequent studies also showed benefits in weight gain (19) and decreased fibrosis (20).

A clinically relevant source of stem cells, especially for the treatment of neonatal diseases, is offered by umbilical cord blood (UCB) (21). UCB contains stem cells that are easily accessible at birth and also capable of differentiating into various cell types (22,23,24), including alveolar epithelial cells (25). Intratracheal and intraperitoneal administration of human UCB-derived MSCs also improved alveolar growth through various mechanisms (26) and in a dose-dependent manner: 5 × 103 cells failed to attenuate both hyperoxia-induced lung injury and inflammation, whereas 5 × 104 and 5 × 105 cells attenuated both hyperoxia-induced injuries and inflammatory responses, but the latter dose was more effective (27). Human cord-derived pericytes and UCB-derived MSCs not only could prevent but also could repair lung injury in neonatal rats when administered 2 wk after established hyperoxia-induced lung injury (28). Long-term assessment at 6 mo showed persistent improvement in lung architecture and exercise capacity, and no adverse effects were observed (28).

While MSCs are affirming their promise in regenerative medicine, other stem and progenitor cells are emerging. Amnion epithelial cells prevent antenatal lipopolysaccharide–induced (29) and ventilation-induced (30) lung injury in fetal sheep. Multipotent amniotic fluid–derived stem cells are capable of differentiating into lineages of all three embryonic germ layers and promote alveolar epithelial cell wound healing and lung growth (31,32). Consistent with the importance of angiogenesis during lung growth, injury, and repair (33), bone marrow–derived angiogenic cells—a novel population of bone marrow myeloid progenitor cells that express angiogenic markers—demonstrated the capacity to restore impaired alveolar and vascular lung growth in hyperoxia-exposed newborn mice (34).

Overall, these observations (summarized in Table 1 ) provide evidence for the therapeutic benefit of bone marrow– and cord blood–derived MSCs in chronic oxygen-induced lung injury in rodents. A recurrent finding, however, is the paucity of cell engraftment, suggesting that stem cell properties such as self-renewal and differentiation are not required for their therapeutic action (35). This finding has led to the hypothesis that MSCs act through a paracrine effect (36), rather than through cell replacement. This realization has expanded the therapeutic options of cell-based therapies.

Table 1 Stem/progenitor cell preclinical trials in experimental neonatal lung diseases
Full size table

MSCs Prevent Lung Injury via a Paracrine Mechanism: It is in the Juice

Several lines of evidence suggest that MSCs act via a paracrine mechanism to protect the developing lung from injury. In vitro, cell-free BMSC-derived conditioned media prevented hyperoxia-induced alveolar epithelial cell apoptosis, accelerated alveolar epithelial cell wound healing, and preserved endothelial cord formation on matrigel during hyperoxia (18).

In vivo, the paracrine effect could also be inferred from the efficacy of intraperitoneal administration of MSCs in preventing oxygen-induced neonatal lung injury (20,26). Accordingly, Aslam et al. (17) provided direct in vivo evidence showing that a single injection of cell-free BMSC-derived conditioned media had a more pronounced effect on alveolar injury and fibrosis prevention than that of BMSCs themselves. In a follow-up study, a single intravenous dose of BMSC-derived conditioned media normalized lung function and reversed alveolar injury and PH (37). A single intratracheal injection of BMSC or BMSC-free conditioned media protected from oxygen-induced alveolar and vascular injury with a persistent benefit followed up to 3 mo (38). Likewise, cell-free conditioned media derived from human UCB-MSCs and pericytes prevented and reversed arrested alveolar growth and lung function in hyperoxia-exposed rats with persistent benefits at 6 mo of age and without adverse effect on lung structure or tumor formation (28). Dose–response studies have not yet been performed.

Although the therapeutic benefit of the conditioned media is undeniable, a potential caveat of this strategy is the lack of cell adaptation to the local injurious environment. In an attempt to overcome this potential limitation, Waszak et al. (39) exposed BMSCs to a “deleterious BPD environment” by priming them ex vivo in hyperoxia for 24 h. Conditioned media collected from these preconditioned cells and injected into hyperoxic rats exerted a more potent therapeutic effect in vivo on lung architecture as compared with nonpreconditioned media.

Thus, rather than replacing injured cells and differentiating into lung cells, MSCs may release factors that protect resident lung cells from injury or modulate the function of inflammatory cells. Tropea et al. recently provided evidence for such a scenario. Bronchioalveolar stem cells are putative epithelial lung stem/progenitor cells at the bronchioalveolar junction, capable of self-renewal and differentiation in culture and also proliferate in response to alveolar injury (40). Both BMSC and BMSC-derived conditioned media increase the number of bronchioalveolar stem cells in neonatal mice exposed to hyperoxia (41). This study also offers new therapeutic perspectives, i.e., the protection of resident lung progenitor cells rather than exogenous supplementation of stem cells. In addition, there is increasing evidence that MSCs interact with inflammatory cells to modulate their response to injury. MSCs can direct macrophages from a M1 (proinflammatory) to a M2 (healer) phenotype in various disease conditions (42,43). Overall, these observations suggest that cell-free conditioned media exerts similar therapeutic benefit to the cell itself. The exciting challenge is how to harness the multiple healing properties of stem cells.

MSCs Prevent Lung Injury via a Paracrine Mechanism: What is in the Juice?

Indeed, the identification of soluble factors in the conditioned media may allow the discovery of novel healing molecules that each by itself or in combination could yield new therapeutic options (44). Besides factors already known to be lung protective, including keratinocyte growth factor (45), vascular endothelial growth factor (46), or adiponectin (47), novel molecules secreted by MSCs have already been identified and shown therapeutic benefit in various disease models, such as stanniocalcin-1 (48)—a potent antioxidant—or tumor necrosis factor-α–stimulated gene/protein 6 (TSG-6) (49)—a potent anti-inflammatory protein.

From a clinical perspective, a relevant question for the design of clinical trials is to determine the most efficacious and safest stem cell–based approach: whole-cell therapy vs. cell-free conditioned media vs. identification of single bioactive molecules vs. identification and determination of the most efficacious combination of molecules. This daunting task could be circumvented by the recent recognition that MSCs release membrane vesicles, exosomes in particular, that act as nanopackages containing a combination of bioactive molecules and microRNAs (50). microRNAs are small noncoding RNA molecules involved in transcriptional regulation of gene expression. In particular, microRNAs could become interesting therapeutic targets in the prevention of BPD (51) by silencing specific genes with deleterious effects during lung injury. Exosomes are 40–100 nm in size and represent a specific subtype of secreted membrane vesicles formed through the fusion of multivesicular endosomes with the plasma membrane. Although known for several decades, membrane vesicles have long been thought of as mere cell debris. Recent evidence, however, suggests that MSC-derived exosomes play important roles in cell communication and mediate the therapeutic benefit of MSCs. For example, MSC-derived exosomes attenuate lung macrophage influx, decrease proinflammatory cytokine levels in the bronchoalveolar lavage, and prevent pulmonary vascular remodeling and hypoxia-induced PH in mice (52). With the exosomes removed, the conditioned media showed no therapeutic effect in this model. Similar therapeutic benefits of MSC-derived exosomes are reported in kidney (53) and cardiac (54) injury. A limitation for the exploitation of exosomes as therapeutic tools remains the process of isolation, characterization, quality control, and large-scale manufacturing. Novel findings continue to uncover the mechanisms by which MSCs protect resident lung cells including the transfer of mitochondria via nanotubes (55). These pleiotropic effects ( Figure 1 ) open exciting avenues, in particular, for multifactorial diseases such as BPD and provide traction for the discoveries of cell-free products.

Figure 1
figure 1

Schematic representation of possible repair mechanisms associated with stem cells. Many therapeutic mechanisms for stem cells have recently emerged. These include microparticle carriers such as microvesicles, exosomes, or multivesicular bodies, which are speculated to be released by stem cells and elicit a therapeutic response. MicroRNA packaged in these vesicles or as a sole effector may also play a therapeutic role. The role of secreted soluble proteins/peptides in neonatal and adult lung injury has been extensively studied. This has lead to the discovery of promising bioactive molecules such as the anti-inflammatory interleukin-10 (IL-10), stanniocalcin-1, tumor necrosis factor-α–stimulated gene/protein 6 (TSG-6), and tumor necrosis factor-α (TNF-α) antagonists, the combination of which may contribute to the pleiotropic effects promoting repair. Recent evidence also unveiled therapeutic mitochondria transfer via nanotubes. These mechanisms can signal endogenous stem cells to amplify or transduce similar repair actions.

PowerPoint slide

Full size image

Considerations for Clinical Trials

A clinical trial testing the safety and efficacy of MSCs in adult patients with chronic obstructive pulmonary disease has been completed (56). Although this clinical trial was predominantly for safety, no substantial evidence of efficacy of the MSCs was recorded. More recently, a phase I clinical trial testing the safety of human UCB-derived MSCs in nine premature infants at risk of developing BPD has been completed (ClinicalTrials.gov: NCT01297205). Although this study upholds the safety of MSC therapy, long-term follow-up is warranted. Further clinical trials are already planned (ClinicalTrials.gov: NCT01207869, NCT01828957).

Preclinical studies have generated proof of concept evidence that cell-based therapies can prevent and restore experimental neonatal lung injury in rodents and sheep. Rather than cell replacement, the therapeutic benefit of stem cells is mediated through a paracrine effect. It is likely that the combination of bioactive molecules contained in the conditioned media provides the compounding pleiotropic effects attributed to MSCs. Administration of the entire cocktail containing unidentified products may conjure unforeseen side effects, and some components may be more beneficial in repair. Thus, further specification of which molecules have reparative properties and/or the isolation of specific micro/nano carriers such as exosomes may lead to pharmacological therapies for BPD.

The cell most suitable for clinical trials appears to be MSCs, likely because of their ease of isolation, characterization, and pleiotropic effects. However, endothelial progenitor cells and other stem/progenitor cells have also proven to be effective in preclinical BPD models. These various cells differ in their roles and respective factors, thereby possibly producing a more pronounced effect when administered in concert (57), although this remains to be proven in the lung.

Likewise, the source of cells is an important consideration. Umbilical cord and cord blood are easily accessible at birth and may have more potent repair capabilities than that of adult BMSCs. Autologous UCB-derived cell therapies may avoid immunological risks and allows the use of minimally manipulated cells. However, given the immunological properties of MSCs, allogeneic cell therapy is feasible and may facilitate the logistics of cell-based therapies.

The timing of the treatment is another factor to be resolved. Recent preclinical evidence showed that UCB-derived MSCs time dependently attenuated hyperoxia-induced injury, eliciting significant protection in the early, but not the late, phase of inflammation (58). In the clinic, the early identification of infants at the most risk of developing BPD through the use of estimators and models may allow for the selection of an appropriate patient population. Patients with early and persistent pulmonary dysfunction have a ~70% risk of developing BPD, as defined by Laughon et al. (59) and may represent an at-risk population of choice for cell-based therapies.

Finally, the safety of each of these cell-based therapies must be investigated thoroughly in well-designed preclinical trials, including large animal models.

In summary, as the incidence of prematurity and chronic neonatal lung disease rises (60), novel therapies are required. Preclinical studies have brought substantial promise in developing an effective clinical therapy that could fulfill the dual role of preventing injury and promoting lung growth. The paracrine effect of cell-based therapies has opened unexpected therapeutic options through the identification of individual molecules or mechanisms including microRNAs, mitochondrial transfer, and microparticles. The promise may not lie in the stem cell itself, but rather in its vast array of bioactive mediators—“it is in the juice.”

Statement of Financial Support

B.T. is supported by the Canadian Institute of Health Research, the Ottawa Hospital Research Institute, and the Children’s Hospital of Eastern Ontario Research Institute. M.E.F. is supported by a summer studentship from Alberta Innovates Health Solutions.

Disclosure:

The authors have nothing to disclose.