In this work, we investigated the feasibility of extending the preterm rabbit model of BPD from PND7 to PND14, using single, double, and triple hit strategies. We showed that delivering preterm rabbit pups at the saccular phase represents a clear disadvantage by itself, as evidenced by less alveolar development and the inability of preterm pups to catch up regarding lung function with age-matched term animals at PND14. We also combined prematurity with other BPD “hits” like postnatal hyperoxia and intratracheal LPS administration. Hyperoxia at 70% O2 improves survival by supporting the transition to extrauterine life while enhancing the BPD-like phenotype. On the other hand, intratracheal LPS administration was feasible, well tolerated, and enhanced the lung inflammatory phenotype, inducing increased alveolar epithelial permeability that led to higher lung injury scores.
The preterm rabbit model has been widely used in recent years to test pharmacological interventions in the context of BPD[15]. Unlike rodent models, preterm rabbits combine structural and functional prematurity and are more cost-effective than other premature models, such as preterm lambs and baboons, because preterm rabbits do not need the infrastructure of a complete neonatal intensive care unit[15]. Recent studies have shown that preterm birth by itself, without any other insult, impairs the postnatal development of the rabbit lung, resulting in lung function deficits and diminished alveolarization compared to age-matched term pups at PND7[16, 33, 34]. Moreover, longitudinal transcriptomic data comparing the gene expression of preterm and age-matched term rabbits’ lungs has shown a consistent increase in the number of dysregulated genes from PND1 to PND14[34, 35], suggesting that the effect of premature birth on the lung molecular pathways persist beyond the first week of life. Our results confirm this observation, showing that premature delivery at the saccular phase is an important “hit” itself associated with lung function deficits, lower alveolarization, and cardiovascular abnormalities (i.e., significantly lower PAT/PET) than age-matched term animals.
Hyperoxia has been the most widely used postnatal “hit” to mimic BPD in animal models. Hyperoxia disrupts the homeostatic balance of cellular processes due to exacerbated reactive oxygen species and activates inflammation cascades and cytokines, triggering lung injury/remodeling processes that resemble a BPD-like phenotype[36, 37]. Hyperoxia (> 95% O2) has also been used in the 28-day gestation preterm rabbit model for drug testing purposes[17, 19, 20]. Although this model is useful to mimic the initial inflammatory phase of BPD, it has some limitations. For instance, the strong hyperoxic hit might limit the response to therapeutic interventions, which, at most, can just attenuate the effects of hyperoxia[17, 19, 20]. In addition, exposure to 95% O2 for seven days produces acute and apparently irreversible toxicity in preterm rabbits, precluding a longer-term follow-up. Mortality after 95% O2 exposure has been reported to be higher than 85% at PND11[21]. Here, we aimed to extend the preterm rabbit model to PND14 to mimic the evolving phase of BPD and investigate therapeutic interventions at later and more relevant time points. PND14 in preterm rabbits has been proposed to be the equivalent age at which BPD diagnosis is made in premature babies (i.e., 36 weeks post-menstrual age)[38]. Taking the seven-day hyperoxia model as a starting point, we implemented an oxygen-weaning protocol that, unfortunately, yielded a low survival at PND14 (< 20%) in pilot experiments, confirming the toxic effect and irreversible damage of 95% O2 exposure in the first week of life.
Exposure to lower oxygen concentrations, at 50% and 70% O2, achieved significantly higher survival at PND14 compared with the oxygen-weaning strategy. Notably, however, 50% O2 did not show significant differences in any of the cardiopulmonary parameters compared with the preterm pups of the Nox group. In contrast, the Hox-70% group showed an enhanced BPD-like phenotype denoted by less alveolarization, lower lung function, and a higher lung injury score than the Nox group. Interestingly, hyperoxia in the first days of life seemed to improve the survival of preterm pups in a dose-dependent manner. Thus, moderate hyperoxia fulfills two apparently conflicting roles: on the one hand, it supports the transition to extrauterine life of rabbit pups, which usually display mild-to-moderate respiratory distress at delivery[18], and on the other hand, it enhances the BPD-like phenotype at PND14.
A subset of animals in the Nox, Hox-70%, and Term groups were assessed using more sophisticated techniques such as design-based stereology and longitudinal micro-CT imaging. Design-based stereology is the gold standard of lung morphometry, providing minimally biased quantitative data about the volume, surface area, length, and number of certain lung structures[28]. Stereology outcomes were suggestive of arrested development of preterm rabbits compared to age-matched term animals, which was further enhanced by hyperoxia. Preterm pups showed significantly lower lung volumes at PND14, driven by a lower volume of the parenchymal compartment without differences in the non-parenchymal volume. Similarly, the significantly lower parenchymal air volume found in preterm pups was driven by lower alveolar airspace volume without differences in the ductal volume. These findings are indicative of reduced alveolarization and are in line with the RAC.
Micro-CT imaging is a non-invasive technology which enables precise tridimensional lung measurements in the same pup at different time points. We acquired micro-CT scans at PND7 and PND13 using an anesthetic protocol established for spontaneously breathing term rabbits[30], confirming its safety profile and effectiveness also for preterm rabbit pups at the two indicated time points.
Micro-CT derived parameters confirmed a significantly lower total lung volume in preterm groups at PND13 compared to term rabbit pups as detected by stereological analyses. Nevertheless, no differences were observed among groups at PND7 in air/tissue components, suggesting that lung structural and morphological changes occurred during the second week of life, in agreement with the RAC quantification in the histological samples. However, functional CT-derived biomarkers, such as VT and MV, were able to detect substantial differences across the three groups as early as PND7. Preterm animals, particularly those exposed to 70% oxygen, had lower VT and MV than Term group, demonstrating a reduction in pulmonary capacity and validating the role of supplemental oxygen in promoting the BPD-like phenotype. Micro-CT data were also corroborated by invasive respiratory function measurements at PND13, in fact the reduction of FRC, VT, and MV in preterm groups positively correlated with the diminishing of IC and static compliance and negatively with tissue damping and tissue elastance, indicating increased stiffness of the lung and impaired ventilation in normal breathing. This imaging technology enables precise longitudinal monitoring of pulmonary function development in preterm offspring. Since each newborn rabbit develops differently than the others, evaluating lung function longitudinally provides a unique opportunity to significantly minimize intra-experimental variability, thus reducing the number of animals needed to achieve statistical significance and raising the awareness of disease development.
Beside the oxygen supplementation, the intratracheal delivery of LPS was also used as a “hit” to exacerbate the lung inflammatory response. The intratracheal instillation of LPS produces a rapid inflammatory response characterized by the alveolar infiltration of polymorphonuclear neutrophils and the release of mediators that play a fundamental role in inflammation and lung damage[39]. Intratracheal LPS administration has been used to induce pulmonary inflammation in intubated preterm lambs to mimic the inflammatory phase of BPD [40, 41]. Intratracheal LPS cannot be used in newborn rodents because of their small size at delivery. Instead, intraamniotic LPS has been widely used in rodents to induce experimental chorioamnionitis, which leads to alveolar simplification and arrested vascular development at PND14[12]. Unlike in newborn rodents, the intratracheal delivery route is feasible in preterm rabbits[22] and was used to deliver LPS directly to the lungs. LPS was delivered in combination with exogenous surfactant to enhance its lung distribution[42]. Intratracheal LPS injections were well tolerated without evidence of acute mortality. The groups receiving intratracheal LPS achieved the highest lung injury scores. Interestingly, a single LPS administration immediately after birth was associated with a significantly higher injury score at PND14 compared to Nox and Hox-70% groups and increased tissue damping (G), a measure for the air resistance in the small airways and alveoli.
This study shows the feasibility of several strategies to extend the 28-day preterm rabbit model of BPD to PND14. However, the study has some limitations. This study was not randomized; it combines the results from independent experimental sessions conducted sequentially over two years. Thus, the groups included in several experimental sessions (e.g., Nox and Hox-50% groups) have a higher sample size. We acknowledge that we could only investigate a limited number of “hit” combinations. Although preterm birth is the common denominator of all the proposed models, alternative hyperoxia levels (e.g., 60% or 80% O2) and alternative hyperoxia-LPS combinations (e.g., hyperoxia 70% plus LPS) may eventually be explored in future studies. Additionally, antenatal strategies to induce experimental chorioamnionitis, placental insufficiency or intrauterine growth restriction could also be explored in preterm rabbits. Lastly, age-matched term pups were naturally delivered and mother-reared, whereas premature pups were delivered via C-section and fed milk formula, as would be expected in a clinical setting. We acknowledge that these differences may partially influence the overall outcome; however, it has been shown that preterm disadvantage persists even when term rabbit pups are delivered via C-section and fed formula[16].