Construction of the hyperoxia induced BPD model
HG significantly decreased body shape and body weight at P7 of experiment compared to NG (HG vs NG: 2.85 ± 0.36g vs 4.52 ± 0.38g, p<0.05) (Fig. 1a,b,c). In this model, expectedly lung morphology analysis revealed disruption of normal alveolar development in HG, characterized by severe impairment of alveolar growth, large airspaces, and incomplete alveolar septation at P7 of experiment (Fig.1d).
Body Weight Gain and Survival Rate
Birth weights were not significantly different between the five experimental groups in P1 (3.18 ± 0.17g and 3.22 ± 0.13g in NG and HG, respectively). However, body weight at P28 in HG was significantly lower compared to NG (10.83 ± 0.52g vs 15.77 ± 0.36g, P<0.001). The reduced body weight gain observed in HG was significantly improved in D1, D2 and D3 (D1 vs HG: 15.773 ± 0.36g vs 10.83 ± 0.52g, p<0.05; D2 vs HG: 16.49 ± 0.34g vs 10.83 ± 0.52g, p<0.05; D3 vs HG: 17.24 ± 0.53g vs 10.83 ± 0.52g, p<0.05, respectively). The seeming gain body weight in D3 and D2 compared to NG did not reach a statistical significance (p < 0.05, vs. NG), but not in D1 (Fig. 2a).
HG significantly increased mortality by the end of experiment (P28) compared to the zero mortality of NG. The increased mortality observed in HG (p < 0.05, vs. NG) was significantly decreased with hUCB-MNCs treatment in D3 (p < 0.05, vs. HG) but not in D1 and D2 (p > 0.05, vs. HG) (Fig. 2b). The seeming lower mortality in D3 and D2 compared to NG did not reach a statistical significance (p < 0.05, vs. NG), but not in D1 (Fig, 2b).
Lung Histopathology
Impaired alveolar growth, as evidenced by microvascular hyperemia and dilatation, adjacent alveolar fusion, alveolar structure destruction and simplification, inflammatory cell exudation, lung tissue structure disorder, was observed in HG compared to NG. After hUCB-MNCs treatment, hyperoxia-induced impairments in alveolar growth and morphological changes were improved, showing significant attenuation in D1, D2 and D3, with the best attenuation in D3 (Fig.3a). In morphometric analyses, the MCL and RAA mean alveolar area, indicating the chord length and size of the alveoli, respectively (Fig.3 b,d), were significantly higher in HG (189.67 ± 33.93μm in MCL and 0.97 ± 0.27mm2 in RAA, p < 0.001) than in NG (60.56 ± 10.02µm in MCL and 0.39 ± 0.08mm2 in RAA, p < 0.001). The RAC means radical alveolar counts, indicating the volume of the alveoli, was significantly less alveoli in HG than in NG (50.22 ± 5.53/mm2 vs 165.78 ± 10.56/mm2, p < 0.001) (Fig. 3c). The data demonstrate significant attenuation of decreased MCL in D3 (64.22 ± 4.52 μm with 66% attenuation vs HG, p < 0.001), in D2 (87.89 ± 6.66 μm with 54% attenuation vs HG, p < 0.001), and in D1 (117.22 ± 23.82 μm with 38% attenuation vs HG, p < 0.05). Decreased RAA in D3 (0.39 ± 0.05mm2 with 60% attenuation vs HG, p < 0.001), in D2 (0.52 ± 0.05mm2 with 46% attenuation vs HG, p < 0.05) and in D1 (0.61 ± 0.13mm2 with 37% attenuation vs HG, p < 0.05)(Fig. 3d). Increased RAC in D3 (136.78 ± 5.69 per mm2 with 172% improvement vs HG, p < 0.001), in D2 (111.89 ± 13.27 per mm2 with 123% improvement vs HG, p < 0.001), and in D1 (84.67 ± 7.07 per mm2 with 69% improvement vs HG, p < 0.001). These findings support the protection of hUCB-MNCs intravenous infusion with the best result in D3.
Pulmonary motion and respiratory blood flow test
In line to the alterations in lung morphology, HG displayed shallow and slow breathing, and the lowest number respiratory rate per minute upon hyperoxia exposure, evidencing by the longer expiratory time with small sawtooth wave compared to those in NG. Interestingly, after hUCB-MNCs treatment, hyperoxia-induced impairments in respiratory motion changes was improved, showing different degrees of improvement in D1, D2 and D3, with the most stable sine wave in D3 (Fig.4a). In pulmonary motion analyses, the MV mean dynamic volume of lung, indicating the ventilation of lung, were significantly lower in HG (8.15 ± 1.20ml) than in NG (13.61 ± 1.74ml, p < 0.001). Our data demonstrate significant improvement of increased MV in D3 (D3 vs HG, 12.53 ± 1.38ml vs 8.15 ± 1.20ml, p < 0.05), but not in D2 (D2 vs HG, 11.26 ± 0.45ml vs 8.15 ± 1.20ml, p > 0.05) and D1 (D1 vs HG, 10.87 ± 0.27ml vs 8.15 ± 1.20ml, p > 0.05) (Fig.4b).
In order to explore the potential impact of hUCB-MNCs towards peripheral pulmonary vascular remodeling under hyperoxia exposure, lung perfusion with Laser Doppler Flowmetry were employed (Fig. 4c). HG displayed lower left or right lung/heart blood flow ratio compared to that in NG (p < 0.05). Our data showed after hUCB-MNCs treatment left or right lung/heart blood flow ratio were improved in P28. In the right lung/heart blood flow ratio was significantly increased in D1 (0.62 ± 0.13), D2 (0.69 ± 0.12) and D3 (0.58 ± 0.07) than that in HG (0.36 ± 0.1, p < 0.05). But in the left lung/heart blood flow ratio, there was no statistical significance in D1 (0.70 ± 0.07), D2 (0.69 ± 0.03) and D3 (0.73 ± 0.05) than in HG (0.44 ± 0.09, p > 0.05) (Fig.4d).
Cytokine expressions analysis of TNF-α, IL-1β, IL-6, IL-2, IL-10
In quantitative RT-PCR, significantly increased mRNA levels of TNF-α (HG vs NG: 2.18 ± 0.58 vs 1.00 ± 0.13, p < 0.05), IL-1β (HG vs NG: 2.80 ± 0.10 vs 1.00 ± 0.00, p < 0.05), and IL-6 (HG vs NG, 1.57 ± 0.09 vs 1.00 ± 0.29, p<0.05), were observed in HG compared to NG. This hyperoxia-induced increase in mRNA expression of these cytokines were significantly attenuated in D2 and D3. TNF-α: D2 vs HG, 0.69 ± 0.13 vs 2.18 ± 0.58, p < 0.001; D3 vs HG, 1.95 ± 0.42 vs 2.18 ± 0.58, p < 0.05 (Fig. 5a), IL-1β : D1 vs HG, 0.52 ± 0.09 vs 2.80 ± 0.10, p < 0.05; D2 vs HG, 2.80 ± 0.10 vs 1.03 ± 0.39, p < 0.05; D3 vs HG, 0.86 ± 0.15 vs 2.80 ± 0.10, p < 0.05 (Fig.5b) and IL-6 :D1 vs HG, 0.81 ± 0.02 vs 1.57 ± 0.09, p < 0.05; D2 vs HG, 0.67 ± 0.32 vs 1.57 ± 0.09, p < 0.05; D3 vs HG, 1.24 ± 0.01 vs 1.57 ± 0.09, p < 0.05), but not in D1 (Fig.5c). Hyperoxia-induced decreased in mRNA expression of IL-10 were significantly increased in D1 and D3. D1 vs HG, 0.24 ± 0.02 vs 0.65 ± 0.15, p < 0.05; D3 vs HG, 2.27 ± 0.15 vs 0.65 ± 0.15, p < 0.05 and IL-2 : D3 vs HG, 2.24 ± 0.41 vs 0.20 ± 0.05, p < 0.05 but not in D2 (Fig. 5d) and IL-10: D2 vs HG, 0.65 ± 0.03 vs 0.65 ± 0.15, p > 0.05) (Fig.5e).
Cytosolic expression of growth factors MMP9, TGF-β and VEGF
Since vascular remodeling and growth reactivation are the main routine to alleviate hyperoxia induced lung injury [7,17]. We systematically examined the protein expression levels of the corresponding cytokines. We observed significantly increased level of MMP9: HG vs NG, 9.14 ±1.16 vs 2.20 ± 1.04, p < 0.05 (Fig. 6a, left panels) and TGF-β (HG vs NG, 19.37 ± 4.08 vs 7.14 ± 0.12, p<0.05 (Fig.6a, middle panels), and decreased level of VEGF were observed in HG: HG vs NG, 8.17 ± 1.07 vs 10.40 ± 0.11, p > 0.05 (Fig.6a, right panels). This hyperoxia-induced increase regarding vascular remodeling of these cytokines were significantly attenuated in D1, D2 and D3 (MMP9, D1 vs HG, 1.18 ± 0.15 vs 9.14 ± 1.16, p < 0.05; D2 vs HG, 1.55 ± 0.74 vs 9.14 ± 1.16, p < 0.05;D3 vs HG, 1.95 ± 0.15 vs 9.14 ±1.16, p < 0.05; TGF-β, D1 vs HG, 2.77 ± 1.35 vs 19.37 ± 4.08, p < 0.05;D2 vs HG, 0.50 ± 0.38 vs 19.37 ± 4.08, p < 0.05; D3 vs HG: 7.19 ± 7.18 vs 19.37 ± 4.08, p < 0.05). Hyperoxia-induced abrogation in VEGF were significantly increased in D3 (VEGF, D3 vs HG, 18.29 ± 1.30 vs 8.17 ± 1.07, p < 0.05), but not in D1 and D2 (VEGF, D1 vs HG, 4.58 ± 1.10 vs 8.17 ± 1.07, p > 0.05; D2 vs HG, 7.06 ± 2.35 vs 8.17 ± 1.07, p > 0.05) (Fig.6b).