Mesenchymal stromal cell extracellular vesicles improve lung development in mechanically ventilated preterm lambs

Abstract

Novel therapies are needed for bronchopulmonary dysplasia (BPD) because no effective treatment exists. Mesenchymal stromal cell extracellular vesicles (MSC-sEVs) have therapeutic efficacy in a mouse pup neonatal hyperoxia BPD model. We tested the hypothesis that MSC-sEVs will improve lung functional and structural development in mechanically ventilated preterm lambs. Preterm lambs (∼129 days; equivalent to human lung development at ∼28 wk gestation) were exposed to antenatal steroids, surfactant, caffeine, and supported by mechanical ventilation for 6-7 days. Lambs were randomized to blinded treatment with either MSC-sEVs (human bone marrow MSC-derived; 2 × 10¹¹ particles iv; n = 8; 4 F/4 M) or vehicle control (saline iv; 4 F/4 M) at 6 and 78 h post delivery. Physiological targets were pulse oximetry O₂ saturation 90-94% ([Formula: see text] 60-90 mmHg), [Formula: see text] 45-60 mmHg (pH 7.25-7.35), and tidal volume 5-7 mL/kg. MSC-sEVs-treated preterm lambs tolerated enteral feedings compared with vehicle control preterm lambs. Differences in weight patterns were statistically significant. Respiratory severity score, oxygenation index, A-a gradient, distal airspace wall thickness, and smooth muscle thickness around terminal bronchioles and pulmonary arterioles were significantly lower for the MSC-sEVs group. S/F ratio, radial alveolar count, secondary septal volume density, alveolar capillary surface density, and protein abundance of VEGF-R2 were significantly higher for the MSC-sEVs group. MSC-sEVs improved respiratory system physiology and alveolar formation in mechanically ventilated preterm lambs. MSC-sEVs may be an effective and safe therapy for appropriate functional and structural development of the lung in preterm infants who require mechanical ventilation and are at risk of developing BPD.NEW & NOTEWORTHY This study focused on potential treatment of preterm infants at risk of developing bronchopulmonary dysplasia (BPD), for which no effective treatment exists. We tested treatment of mechanically ventilated preterm lambs with human mesenchymal stromal cell extracellular vesicles (MSC-sEVs). The results show improved respiratory gas exchange and parenchymal growth of capillaries and epithelium that are necessary for alveolar formation. Our study provides new mechanistic insight into potential efficacy of MSC-sEVs for preterm infants at risk of developing BPD.

Keywords: alveolar formation; bronchopulmonary dysplasia; chronic lung disease of the neonate; exosomes.

Graphical abstract

Figure 1.

Timeline for the study. MSC-sEVs were given intravenously at hour of life 6 and again at hour of life 72. MSC-sEVs, mesenchymal stromal cell extracellular vesicles.

Figure 2.

Mesenchymal stromal cell (MSC)-derived small extracellular vesicles (MSC-sEVs). A: MSC-sEVs were isolated from medium conditioned by MSCs for 36 h and was subjected to successive differential centrifugation followed by tangential flow filtration. Concentrated medium (50×) was gently floated on an iodixanol (IDX) cushion gradient (B) to further purify and isolate the MSC-sEV population (fraction 9, density ∼1.17 g/mL). C: nanoparticle tracking analysis was used to assess MEx concentration and particle size. Red lines denote SD. D: transmission electron microscopy demonstrates the heterogeneous sEV morphology. E: immunoblots of IDX fractions demonstrate fraction nine to contain ALIX, FLOT1, CD63, CD81, CD9, TSG101, established EV-associated markers, and the absence of GM130. F: immunogold labeling confirms the presence of EV tetraspanin proteins (CD81 and CD63) on MSC-sEVs, as denoted by black arrows.

Figure 3.

MSC-sEVs improved physiological outcomes. A: the MSC-sEVs group tolerated more enteral colostrum/milk volume at days of life (DOL) 3–6 compared with the vehicle control group (*P < 0.05). Within-group comparison showed that each group tolerated larger volume of colostrum/milk on DOL2-7 compared with DOL1 (^P < 0.05). B: the MSC-sEVs group weighed significantly less at preterm delivery than the vehicle control group at preterm delivery (*P < 0.05). The MSC-sEVs group maintained weight daily whereas the vehicle control group lost weight daily during the week-long study (B; ^P < 0.05). The MSC-sEVs group had significantly lower, better, respiratory severity score (D; DOL 4–7), A-a gradient (E; DOL 6 and 7), and oxygenation index (F; DOL 6) than the vehicle control group (*P < 0.05). Within-group comparison showed that the MSC-sEVs had lower respiratory severity score, A-a gradient, and oxygenation index on subsequent DOL compared with DOL1 (^P < 0.05). These parameters were constant or progressively rose for the vehicle control group. The MSC-sEVs group had significantly higher, better, S/F ratio (C; DOL 5–7) compared with the vehicle control group (D; *P < 0.05). Within-group comparison showed that the MSC-sEVs group had higher S/F ratio on DOL2-7 compared with DOL1 (^P < 0.05). Results are shown as means ± SD. Statistical differences were assessed by two-way ANOVA and Tukey’s multiple comparison test (mixed effects p-critical ≤0.05). Asterisks (*) identify differences between the groups for each indicated DOL. Upward carrots (^) identify differences between DOL1 and subsequent DOL within a group. MSC-sEVs, mesenchymal stromal cell extracellular vesicles; S/F ratio, oxygen saturation/$\mathrm{F}{\mathrm{I}}_{{\mathrm{O}}_2}$ratio.

Figure 4.

MSC-sEVs did not affect systemic hemodynamics. Heart rate was not different between the two groups (A). On DOL 1, mean blood pressure (B; *P < 0.05), systolic blood pressure (C; *P < 0.05), and diastolic blood pressure (D; *P < 0.05) were significantly lower in the MSC-sEVs group compared with the vehicle control group. Results are shown as means ± SD. Statistical differences were assessed by two-way ANOVA and Tukey’s multiple comparison test (mixed effects p-critical ≤0.05). Asterisks (*) identify differences between the groups for each indicated DOL. DOL, days of life; MSC-sEVs, mesenchymal stromal cell extracellular vesicles.

Figure 5.

MSC-sEVs did not affect respiratory system mechanics. Respiratory system resistance (A) and reactance (B) were dynamically assessed daily by the forced oscillation technique. Neither respiratory system resistance (A) nor reactance (B) were statistically different between the two groups. Alpha smooth muscle immunohistochemistry (brown color) and morphometry were used to quantify thickness and area of the media surrounding terminal bronchioles (TB). Smooth muscle thickness (C and E) and area (D and F) were significantly less in the MSC-sEVs group (*P ≤ 0.05) compared with the vehicle control group. Results are shown as means ± SD. Statistical differences for respiratory system resistance and reactance were assessed by two-way ANOVA and Tukey’s multiple comparison test (mixed effects p-critical ≤0.05). Statistical differences for terminal bronchiolar smooth muscle accumulation were assessed by unpaired t test (p-critical ≤0.05). MSC-sEVs, mesenchymal stromal cell extracellular vesicles.

Figure 6.

MSC-sEVs improved indices of alveolar formation. The histological appearance (H&E stain) of the lung’s parenchyma was more delicate and lacier across terminal respiratory units (TRU) in the MSC-sEVs group compared with the vehicle control group (A–D). The asterisks (*) in A and B identify regions shown at greater magnification in C and D. The arrowheads in C and D identify secondary septa, which appear longer and thinner in the MSC-sEVs-treated panel relative to in the vehicle-treated panel. The arrows in C and D identify distal airspace walls, which appear thinner in the MSC-sEVs-treated panel relative to the control vehicle panel. Quantitative histological analyses showed that radial alveolar count and secondary septal volume density were significantly greater in the MSC-sEVs group compared with the vehicle control group (E and F; *P < 0.05). Conversely, distal airspace walls were significantly thinner in the MSC-sEVs group compared with the vehicle control group (G; *P < 0.05). Results are shown as means ± SD. Quantitative histological results were assessed by unpaired t test (p-critical ≤0.05). H&E, hematoxylin-eosin; MSC-sEVs, mesenchymal stromal cell extracellular vesicles.

Figure 7.

MSC-sEVs improved indices of alveolar capillary growth. We used immunohistochemistry for PECAM-1 to label capillary endothelial cells (A–D) and stereology to quantify alveolar capillary surface density and blue counterstain for epithelial surface density (E and F). The asterisks (*) in A and B identify the regions shown at higher magnification in C and D, respectively. The arrows in B and D show brown immunostaining that highlights capillary endothelial cells. Capillary surface density was significantly greater in the MSC-sEVs-treated group compared with the vehicle control group (E; *P < 0.05). F shows that surface density of epithelium was comparable (not significantly different) between the two groups, which is important because we used this measurement as the common independent reference space. Results are shown as means ± SD. Quantitative histological results were assessed by unpaired t test (p-critical ≤0.05). MSC-sEVs, mesenchymal stromal cell extracellular vesicles.

Figure 8.

A–L: MSC-sEVs improved lung VEGF-R2 protein abundance. VEGF-R2 normalized protein abundance in lung parenchymal tissue was statistically greater for the MSC-sEVs-treated group compared with the vehicle control group (B; *P < 0.05). VEGF-R2 mRNA level was not different between the two groups. Neither was VEGF protein abundance different between the groups. We quantified other mRNA and protein markers; namely, apoptosis (p53 and caspase 3 mRNA; cleaved caspase 3 protein), cell proliferation (cMyc and TGF-β mRNA; PCNA protein), and surfactant apoproteins (SP-A and SP-B mRNA; SP-B protein). Detected statistically significant differences were limited to lower PCNA relative protein abundance (*P < 0.05) and SP-B relative mRNA level (*P < 0.05) in the MSC-sEVs group compared with the vehicle control group. Results are shown as means ± SD. Semiquantitative results were assessed by Mann–Whitney U test (p-critical ≤0.05). MSC-sEVs, mesenchymal stromal cell extracellular vesicles; PCNA, proliferating cell nuclear antigen; VEGF, vascular endothelial growth factor.

Figure 9.

MSC-sEVs reduced accumulation of vascular smooth muscle in the media of pulmonary arterioles. We used immunohistochemistry for alpha-smooth muscle actin (brown color) and morphometry to quantify vascular smooth muscle abundance in the wall of pulmonary arterioles (PA; opposed black arrows in A and B). We used terminal bronchioles as the independent anatomic landmark to assess the same generation of pulmonary arterioles, which cannot be identified by diameter. Smooth muscle thickness (C) and area (D) in the wall of pulmonary arterioles were significantly less in the MSC-sEVs group (*P < 0.05) compared with the vehicle control group. Results are shown as means ± SD. Quantitative histological results were assessed by unpaired t test (p-critical ≤0.05). MSC-sEVs, mesenchymal stromal cell extracellular vesicles.

Figure 10.

A–I: MSC-sEVs did not affect liver or kidney indicators of dysfunction. No statistical differences were detected for markers of liver dysfunction (alkaline phosphatase, aspartate aminotransferase, alanine aminotransferase, total bilirubin, and direct bilirubin) between the two groups. Likewise, no statistical differences were detected for markers of kidney dysfunction (blood urea nitrogen, creatinine, total protein, albumin) between the two groups. The levels of these parameters were within respective normative range for unventilated, normal fetal lambs (bracket in each graph). Results are shown as means ± SD. Quantitative results were assessed by unpaired t test (p-critical ≤0.05). MSC-sEVs, mesenchymal stromal cell extracellular vesicles.

Figure 11.

A–E: MSC-sEVs did not affect hematological parameters. No statistical differences were detected for hematocrit or total and differential white blood cell counts between the two groups. The levels of these parameters were within respective normative range for unventilated, normal fetal lambs (bracket in each graph). Results are shown as means ± SD. Quantitative results were assessed by unpaired t test (p-critical ≤0.05). MSC-sEVs, mesenchymal stromal cell extracellular vesicles.