Leveraging Microbial Symbiosis to Modulate Bronchopulmonary Dysplasia

In this issue of the Journal, Freeman and colleagues (pp. 267–278) provide early evidence that Gammaproteobacteria-dominated airway dysbiosis may contribute to the pathogenesis of bronchopulmonary dysplasia (BPD) (1), a chronic lung disease secondary to preterm birth that affects more than a third of extremely preterm infants born at less than 28 weeks’ gestation (2). Previously, the same group had shown that Gammaproteobacteria-dominated airway microbiome correlated with BPD (3), joining several studies linking respiratory dysbiosis and BPD (4). The present study furthers this work by demonstrating a causative effect in a hyperoxia mouse model. In the present study, tracheal aspirates from BPD-diagnosed human infants were delivered to the tracheas of hyperoxia-exposed gnotobiotic neonatal mice; lung pathology was exacerbated by human BPD-associated tracheal aspirates but not by that of non-BPD control tracheal aspirates. A similar effect was seen with monocolonization with the prototypical Gammaproteobacterium Escherichia coli. Knockout mice lacking the antioxidant gene Nrf2 experienced a more severe phenotype after inoculation with E. coli. However, lung pathology in hyperoxia-exposed Nrf2 mice was attenuated by intratracheal delivery of three Lactobacillus spp. cocultures. These findings support a causative relationship, both protective and pathogenic, between respiratory microbiota and redox pathways in hyperoxia-induced neonatal lung disease that could potentially be exploited for much-needed novel microbiota-targeted interventions to prevent and/or treat BPD.

The clinical implications of these findings are severalfold. It has long been accepted that ante- and postnatal infections, including pneumonias, contribute to the development of BPD in preterm infants (5, 6). Growing evidence, however, suggests that antibiotic exposure in the absence of infection, which is common in clinical practice to treat suspected neonatal infection, has an equally harmful impact on respiratory morbidity (7). The present study adds to the evidence that host-microbiome interactions modulate neonatal lung injury. If this paradigm is true, then therapies designed to harness subtle and nuanced host/microbe metabolic interactions could restore tissue homeostasis and attenuate neonatal lung injury. Physicians may then find themselves for the first time in decades armed with a new treatment for BPD.

Clinically, there are two important time points in the course of BPD, and one could imagine using knowledge of airway dysbiosis to intervene at either time. Events in the first weeks of life, such as infection, exposure to antibiotics, hyperoxia, and mechanical ventilation, predispose infants to a pathogenetic cascade that eventually leads to BPD. Interestingly, clinicians often observe a pattern of clinical deterioration at 1–2 weeks of life that predicts BPD (8). Early interventions to establish and maintain a beneficial airway microbiome could serve as a preventive measure. Once the disease is established, infants with BPD demonstrate a unique clinical phenotype that includes not only obstructive-type lung disease but also more systemic morbidity, including dysregulated immunity, restricted growth, cardiovascular disease (e.g., systemic and pulmonary hypertension), abnormal gastrointestinal functioning, and impaired neurodevelopment (9–11). This latter time point is particularly burdensome and difficult to manage. Notably, the tracheal aspirates used by the authors in this study were obtained from infants with established BPD, providing a glimmer of hope that restoration of a healthy microbiome even in advanced stages of BPD may be beneficial.

There are a number of outstanding questions. Although the authors provide some evidence of a cause-and-effect relationship among airway microbial ecology, redox homeostasis, and BPD in an animal model, there is more to learn about the underlying mechanisms before findings can be confidently translated to human trials. For example, what is the risk:benefit ratio for the administration of the “beneficial” Lactobacillus spp. to human neonates? Metagenomic, metabolomic, and rigorous in vitro and in vivo functional approaches comparing various Lactobacillus species, which are known to encode antioxidant-related genes (12), could aid in understanding this potential risk:benefit ratio. Is it safe and feasible to inoculate the airway? Does the introduction of beneficial microbial species require direct inoculation of the airway, or would another route suffice? Is there another way to modify the airway microbiota besides directly introducing new species? Isolated case reports linked gut probiotics to systemic infection in premature neonates, for example, and therefore probiotics are not yet universally accepted among neonatologists as a safe intervention (13). A caveat to consider is that a lack of niche fitness—or the appropriateness of the host’s unique local environment to support a certain microbial community—may preclude the usefulness of attempts to modify the respiratory microbiome (7). Similarly, the systemic derangements in immunity seen in babies with BPD may thwart such efforts. Last, it should be acknowledged that this study did not investigate the potential impact of other components besides microbes present in tracheal aspirates—for instance, cytokines and metabolic by-products—that may contribute to the potentiation of lung injury in their mouse model; it is possible that the microbiome may have played a lesser role than is assumed. Nonetheless, the present study by Freeman and colleagues (1) presents a compelling case to accelerate the field of host–microbiota symbiosis toward novel interventions to prevent and treat BPD.