In this manuscript, we characterised term and preterm hAEC-EVs and assessed their efficacy in an experimentally induced BPD mouse model. This study provides the first evidence that hAEC-EVs from term and preterm pregnancy bore exosome characteristics based on size distribution, morphology and surface markers. We did, however, observe that EV cargo and functional potency between term hAEC-EVs and preterm hAEC-EVs were significantly different. Only term but not preterm hAEC-EVs were beneficial in the setting of BPD-like lung disease, which was comparable to the efficacy of term hAECs.
Both term and preterm hAEC-EVs appeared to be of the size range most commonly associated with exosomes, displaying their typical cup-shaped morphology with no significant difference in EV yield. In order to verify their identity, we sought to determine the presence of typical EV markers CD9, CD63, CD81, and ALIX. In current study, we observed EV yield from hAECs with 2 to 9µg EVs from 1x106 cells, or 4x1010 to 24x1010 particles between different hAEC donors, which may be due to the human biological variability. However, there was no significant difference in protein yield, particle concentration and particle size distribution between term and preterm hAEC-EVs. It is worth noting that others have reported yields of 10µg EVs from 2.5x105 to 1x107 human bone marrow derived MSCs26, 27.
Further, when we examined exosomal surface epitopes, we have shown that term hAEC-EVs had higher level of CD142 and CD133. CD142 is also called tissue factor (TF), which belongs to the cytokine receptor class II family, and it is known to exist with membrane vesicles28. CD142 was proven to induce pro-angiogenic matrix for blood vessel infiltration, and the activation of TF/VIIa (coagulation factor VIIa) signaling pathway promoted angiogenesis29. CD133 is one of the key biomarkers for isolation and characterization of stem cells, and it can be found in a few cells including epithelial cell membrane30. CD133+ cells are known to have regenerative properties including promoting cell differentiation and enhancing angiogenesis31, 32 and have been used in a few clinical trials such as ischemia and hepatic fibrosis33, 34. Similarly found, MSC-derived EV containing CD133 epitope promoted recovery in a moue model of ischemia reperfusion renal injury, proving that EV therapeutics might be superior to cell-based therapy in terms of safety and versatility35.
We demonstrated that the term hAEC-EVs, but not preterm hAEC-EVs, were efficacious in mitigating lung injury in a mouse model of experimental BPD. Specifically, we observed an increased tissue-to-air space ratio, secondary septal crest density, and small pulmonary vessel numbers suggesting that term hAEC-EVs may protect against the classical alveolar simplification associated with postnatal hyperoxia. This improvement in lung structure was accompanied by a reduction in lung inflammation with specific changes in IL-1β, TNF-α, LIF, MCP-1, MIP-2 and GM-CSF levels. Similar to term hAEC treatment, term hAEC-EV administration prevented peripheral pulmonary artery muscularization, and further prevented pulmonary hypertension and right ventricle hypertrophy, and improved lung function as long-term outcome. Additionally, term hAEC-EV treatment was associated with increased numbers of AT2 cells, but without affecting the BASC population. This suggests that term hAEC-EVs may improve lung tissue regeneration either through stimulating the local lung stem cell AT2 niche or promoting the maturation of BASCs to AT2 cells.
On PND 7 and 14, the tissue-to-air space ratio and secondary septal crest percentage were significantly decreased in the injured group compared to the control group, and we observed that both term hAEC and term hAEC-EV treatment brought the percentages back to healthy levels. As angiogenesis promotes normal lung alveolar development and contributes to the maintenance of alveolar structure36, BPD animals with simplified lung structure had fewer pulmonary vessels. Our previous studies showed that only term but not preterm hAECs derived conditioned media supported vascular tubule formation25. This supports the findings in our current study where only term hAECs and term hAEC-EVs protected against the loss of small pulmonary vessels.
When observing another characteristic in BPD, lung inflammation, we found that term hAEC-EVs exerted anti-inflammatory effects by PND7. The elevated levels of IL-1β and TNF-α caused by the injury were diminished by either term hAECs, term hAEC-EVs or preterm hAEC-EVs. IL-1β and TNF-α are mainly secreted from activated macrophages and monocytes37, which we had previously shown to be reduced by hAEC23. It is thus possible that term EVs may also modulate macrophage function. Along similar lines, MSC-derived EVs reportedly polarized macrophages from M1 to M2 phenotype in vitro38, suppressed macrophage pulmonary infiltration in response to hypoxic lung injury39, and attenuated burn-induced inflammation through reducing the expression of IL-1β and TNF-α in mice40. In addition, we also found in vitro that preterm hAEC-EVs suppressed T cell proliferation more significantly than term hAEC-EVs and that term hAEC-EVs improved macrophage phagocytosis more significantly than preterm hAEC-EVs. Similarly, MSC-derived EVs have also been shown to restore T cell function disrupted by neonatal hypoxia41. This data suggests that impact may be due to direct effect of EV cargo on macrophages.
The cytokine levels of GM-CSF, MCP-1, MIP-2 and LIF were significantly higher in injured animals by PND14, as previously reported23. Macrophages and monocytes are the major cell types secreting MIP-2 (IL-8 in human) and MCP-1 during immune response. MCP-1 regulates monocyte migration and infiltration, and recruit and direct macrophage movement42, and MIP-2 is a neutrophil chemoattractant43. It has been reported that alveolar macrophages secrete MIP-2 and MCP-1 during lung inflammation44, 45, and elevated levels of MCP-1 and IL-8 were reported in lavage fluid of BPD-affected infants46. Additionally, GM-CSF is a known macrophage activator47, and LIF is known to potentiate macrophage aggregation and activation48. In this study we found that similar to term hAEC treatment, term hAEC-EVs reduced the above cytokines to control levels. It is likely that term hAEC-EVs reduced lung inflammation through modulating macrophage and neutrophil populations as we49 and others50, 51 have shown previously. However, preterm hAEC- EVs only reduced GM-CSF level to normal, leaving IL-1β, MCP-1, MIP-2 and LIF at higher levels than control and treated animals.
In this study, we found that the percentage of BASC or AT2 cells were unchanged in the injured group. This correlates with previous studies that have shown that hyperoxia at 75% oxygen had no effect on BASC proliferation52. Surprisingly, unlike hAECs, term hAEC-EVs had no effect on BASC numbers, but significantly increased the percentage of AT2 on PND7 and PND14. The increase in AT2 correlated with the improvement in lung structure, which may have resulted in activation, proliferation and differentiation into AT1 to improve tissue-to-air space ratio. Some researchers have suggested that there are different mechanisms involved in AT2 and BASC proliferation and differentiation53–56. The observation that term hAEC-EVs increased AT2 numbers while term hAECs increased BASC numbers could therefore be attributed to differential activation of signaling pathways. However, neither BASC nor AT2 proliferation were observed in preterm hAEC-EV treated group.
In our experimentally induced BPD animals, we observed thickened and muscularized pulmonary vessels on PND14, pulmonary hypertension, right ventricle hypertrophy, and declined lung function with increased resistance and increased compliance. Similar to the term hAEC administration, term hAEC-EVs prevented early pulmonary vascular muscularization and later on secondary pulmonary hypertension. Kourembanas and her colleagues induced BPD injury in mice by exposing neonates to 75% oxygen for a week, they also observed pulmonary vascular remodeling, increased lung capacity, and elevated right ventricular systolic pressure, and found that both bone marrow derived MSCs (BM-MSC) and Wharton’s jelly MSCs (WJ-MSCs) derived EVs ameliorated vascular remodeling and improved lung function57, 58. It is worth noting that although EVs do not package all the factors that cells contain, this study implies that term hAEC-EVs may carry sufficient bioactive material to exert reparative effects. This is supported by studies that showed MSC-EVs had reparative effects, but EV-depleted MSC conditioned media had no effect in a few preclinical models including hypoxic-induced pulmonary hypertension and cardiotoxin induced muscle injury59, 60. From what mentioned above, Kourembanas and her colleagues treated BPD mice with EVs derived from different sources of MSCs (BM-MSCs and WJ-MSCs) and found both achieved the same therapeutic outcomes. This suggested that EVs from different cell sources could transport effective cargo to exert efficacy. Future experiments targeting specific cargo like microRNA (miRNA) have shown enormous promise for the development of therapeutic agents for human diseases like BPD61, 62.
Proteomic analysis revealed that EV cargo from term and preterm hAECs were significantly different. Proteins that were more highly expressed in term hAEC-EVs were enriched in several pathways associated with functional characteristics observed in our current study. The extracellular matrix (ECM) organization pathway is crucial to stem cell lineage specification, and cell migration and proliferation through its dynamic regulation of the microenvironment63. Cell junction organization pathway and adherents junction interactions pathway are cell-cell or cell-ECM contacts that are required for cell survival, differentiation, and migration64. The ECM homeostasis is important in normal lung development, and collagen is one of the major compositions of the ECM65, 66. Increased gene expression level of collagen alpha-1 (cola1) was detected in neonatal mice and aided the process of alveolar development67. Integrin-mediated cell-ECM interactions are known to play an important role in normal lung development too, and higher expression levels of integrin alpha1, 2, 6 and beta1 were reported in the bronchial and alveolar epithelium during the alveolar stage of lung development68, 69. Indeed, cola1 and integrin alpha6 were highly expressed in term hAEC-EVs compared to preterm hAEC-EVs. Among the proteins that were significantly higher in term hAEC-EVs, 15 proteins were associated with extracellular matrix organisation, 9 proteins were associated with cell junction organisation, and 5 proteins were associated with integrin cell surface interactions. This indicated that term hAEC-EVs may support normal lung development through these pathways. In the pathways associated with proteins that were highly expressed in the preterm hAEC-EVs, the top three pathways are involved in the immune system, innate immune system and developmental biology which are associated with proteins 19, 13 and 12, respectively. This suggests that preterm hAEC-EVs may play a role in inflammation. What we found interesting is that Axon guidance protein was highly expressed in both term and preterm EVs, axon guidance proteins guide growing axons during development and control structural plasticity of synaptic connections. Changes in expression or function of these proteins induces pathological changes in neural circuits that predispose to, or cause, neurological diseases. This is an important finding in that EV therapy would be beneficial for preterm babies born with BPD who also develop cerebral palsy70, as axonal injury is one of the hallmarks of cerebral palsy. Taken together, the significant differences in EVs protein cargo and the pathways that they are enriched in likely explain the functional differences between preterm and term hAEC-EVs.
We have previously shown that preterm hAECs had limited reparative effect in bleomycin induced lung injury compared to term hAECs71. Hence, we did not assess the efficacy of preterm hAECs in treating BPD and expectedly, we show that preterm hAEC-EVs did not have significant therapeutic effect in this BPD disease setting. We think that prematurity and complications associated with preterm birth may have influenced hAEC function and consequently the EV cargo and their functional potency. When bone marrow derived MSCs were preconditioned with 95% oxygen, the conditioned media contained higher levels of antioxidant stanniocalcin-1 and was more benefit for hyperoxia induced lung injury rats compared to normal cultured MSC conditioned media72, which implies the changes of cargo in cell secretomes impacts their functional potency. However, Bhandari et al reported the MSC EVs that derived from umbilical cords of 25–30 weeks of gestational age, exerted reparative effect in 95% hyperoxia induced BDP mouse model by improving lung structure and function73. Another study showed that the placental EVs derived from maternal peripheral plasma, had declined bioactivity (endothelial cell migration area) in late pregnancy compared to early pregnancy74. However, this may not reflect the property of fetal stem/stem-like cells derived EVs, as EVs derived from maternal peripheral plasma is a mixture of EVs from maternal and fetal origin, and factors such as secretion of EVs and placental perfusion need to be considered.