Mesenchymal stem cells from different sources show distinct therapeutic effects in hyperoxia‐induced bronchopulmonary dysplasia in rats

Abstract Mesenchymal stem cells (MSCs) have been shown as an effective medicinal means to treat bronchopulmonary dysplasia (BPD). The widely used MSCs were from Wharton's jelly of umbilical cord (UC‐MSCs) and bone marrow (BM‐MSCs). Amniotic fluid MSCs (AF‐MSCs) may be produced before an individual is born to treat foetal diseases by autoplastic transplantation. We evaluated intratracheal (IT) MSCs as an approach to treat an hyperoxia‐induced BPD animal model and compared the therapeutic effects between AF‐, UC‐ and BM‐MSCs. A BPD animal model was generated by exposing newborn rats to 95% O2. The continued stress lasted 21 days, and the treatment of IT MSCs was conducted for 4 days. The therapeutic effects were analysed, including lung histology, level of inflammatory cytokines, cell death ratio and state of angiogenesis, by sacrificing the experimental animal at day 21. The lasting hyperoxia stress induced BPD similar to the biological phenotype. The treatment of IT MSCs was safe without deaths and normal organ histopathology. Specifically, the treatment was effective by inhibiting the alveolar dilatation, reducing inflammatory cytokines, inducing angiogenesis and lowering the cell death ratio. AF‐MSCs had better therapeutic effects compared with UC‐MSCs in relieving the pulmonary alveoli histological changes and promoting neovascularization, and UC‐MSCs had the best immunosuppressive effect in plasma and lung lysis compared with AF‐MSCs and BM‐MSCs. This study demonstrated the therapeutic effects of AF‐, UC‐ and BM‐MSCs in BPD model. Superior treatment effect was provided by antenatal MSCs compared to BM‐MSC in a statistical comparison.


| INTRODUC TI ON
Bronchopulmonary dysplasia (BPD) is a chronic lung disease major in preterm infants characterized by arrest of alveolation, fibroblast activation and inflammation. Some patients have fibrosis of the lungs caused by prolonged mechanical ventilation and oxygen exposure. 1 The incidence rate of BPD was up over 30% in premature infant in the United States 2 and Europe, 3 regardless of race. For antenatal, the risk factors of BPD included maternal smoking and intrauterine growth restriction (IUGR). For postnatal cases, the risk factors were hyperoxia and mechanical ventilation. 2 BPD increases mortality in neonates, and it is the leading cause of chronic lung disease in children. 4 Adult survivors of BPD also presented with pulmonary impairment-associated diseases 5 and also had long-term cardiopulmonary morbidities. 2 The clinical treatments for BPD were dependent on their manifestation. Old BPD, characterized by fibrosis and inflammation, were cured by the discovery of effective biomaterial and ventilation devices. However, new BPD, marked by tissue simplification and arrest of alveolarization, still affect premature infants. 6,7 Cytotherapy, namely, mesenchymal stem cell (MSC)-based therapy, provides a new possibility to effectively treat BPD.
MSCs are cultured primary cells, which are defined with specific bio-characteristics 8 in vitro and bio-functions in vivo, namely, tissue protection, inflammation regulation and promotion of angiogenesis. 9 Likewise, MSC-derived conditioned media conferred therapeutic benefit for alveolarization, pulmonary artery remodelling and angiogenesis. 10 These cells may be derived at the prenatal stage, 11,12 such as amniotic fluid MSCs (AF-MSCs) and umbilical cord MSCs Treatment with MSCs in preclinical hyperoxic models of BPD in rodents resulted in statistically significant improvement in lung injury. 10 MSCs may be infused for treatment by intravenous, 2 intraperitoneal (IP) or intratracheal (IT) administration. The latter is the first choice for MSCs for the treatment of BPD by being safe and effective as showed in animal studies 18 and clinical treatment, 19 even used in some clinical phase one study. 20,21 However, the therapeutic effect of MSCs from different sources has not been investigated.
In this study, we generated a hyperoxia-induced neonatal lung injury model to mimic BPD. The modelling rat got the physiological manifestation correspond with BPD pathology, such as simplification of lung histology, inflammation and cell death increase and decrease of neovascularization. Then, we IT administration of AF-, UC-and BM-MSCs and tested their therapeutic effect in our BPD model. We showed that the hyperoxia-induced pathological changes were relieved by all treatments; however, physiological and biochemical difference between the groups were observed. Collectively, we demonstrated the safety and effectiveness in improving hyperoxia-

| BPD model and MSC cytotherapy
The Sprague Dawley (SD) rats (Guangdong Medical Laboratory Animal Center, Guangzhou, China) on the first day after birth were used to generate BPD rats by exposing them to hyperoxic conditions (95% O2), in which animals were raised in a Hypoxia/Hyperoxia incubator (PH-A1, Puhe bio, Wuxi, China). The BPD rats were fed by lactation by one rat for one day and changed to other lactating rats kept in normoxia (21% O2). In addition, non-specific control (NC) neonatal rats were raised in normoxia (21% O2). Neonatal rats transferred to hyperoxic conditions regarded as day one, cytotherapy using MSCs was performed on day 4 via a different MSC intratracheal injection (n = 5 for each group, 5 × 10 5 cells in 50 μl of PBS per animal). As a control, NC (n = 5) and BPD rats (n = 5) were given 50 μl of PBS without cells. All of the experimental animals were sacrificed on day 21 for further study ( Figure 1A). This study was approved by the Academic Committee of the Third Affiliated Hospital of Guangzhou Medical University.
Superior treatment effect was provided by antenatal MSCs compared to BM-MSC in a statistical comparison.

K E Y W O R D S
bronchopulmonary dysplasia, different source, inflammation, mesenchymal stem cells, neovascularization F I G U R E 1 Intratracheal administration of MSCs relieves hyperoxia-induced lung histological changes in neonatal rats. Schematic representation of the time course of BPD modelling and MSC therapy (A). Non-specific control (NC) was set up in normoxia (21% O2), bronchopulmonary dysplasia (BPD) was modelling by hyperoxia atmosphere (95% O2). 50 μl PBS or 50 μl PBS with 5 × 10 5 MSCs from different sources was intratracheal administration on day 4, and all animals were sacrificed on day 21 for further study. The histology of the lung in different groups was identified using HE staining (B). IT for MSCs ameliorated hyperoxia-induced alveolar expansion, according to the mean linear intercept (C) and mean alveolar volume (D). n = 5; *, p < 0.05 compared with the BPD group, #, p < 0.05 compared with the BPD+UC-MSC group. Scale bar=100 μm

| Cell culture
The human amniotic fluid MSCs (AF-MSCs) and umbilical cord MSCs (UC-MSCs) were isolated and characterized as previously described. 11 Human bone marrow MSCs were a gift from the Center for Stem Cell Biology and Tissue Engineering of Sun Yat-sen University.
The derivation and characterization of these cells was previously reported. 22 The maintenance and expansion of MSCs was performed using Yu's protocols. 22

| Histology and tissue staining
After animals were sacrificed, several organ histology was completed on liver, kidney, heart, spleen, brain, lymph, thymus and lung. Organs

| Biochemical assay
The concentration of secreted cytokines in plasma and lung lysates were determined by ELISA (enzyme-linked immunosorbent assay).
Plasma was separated by centrifuging whole blood without anticoagulation which was held at room temperature for 2 h after drawing.

| Statistical analysis
Quantitative results for Western blot was expressed as the mean ± standard deviation. The other data were demonstrated in scatter plot. The data of multiple groups were statistically analysed by ANOVA. Two-tailed Student's t test was applied to compare the differences between the two groups. A value of p < 0.05 was considered to indicate a statistically significant difference.

| MSCs rescued hyperoxia-induced BPD associated with pathological changes in neonatal rat lung
A BPD model was established on neonatal Sprague Dawley rats by exposing them to hyperoxia (95% O 2 ). In addition, non-specific control (NC) neonatal SD rats were raised in normoxia (21% O 2 ). Cytotherapy using MSCs was performed on day 4, which included IT administration of 50 μl PBS with 5 × 10 5 AF-MSCs, UC-MSCs or BM-MSCs. As a control, NC and BPD rats were given 50 μl of PBS without cells.
Furthermore, biological analysis was performed on day 21 by sacrificing five experimental animals from each group (n = 5) ( Figure 1A).
During the experimental treatments, no deaths of infant rats were observed. Morphological changes were observed by HE staining.
Hyperoxia did not induce significant pathological changes in main organs, including liver, kidney, heart, spleen, brain, lymph and thymus and so did MSC cytotherapy ( Figure S1). For lung histological analysis, hyperoxia induced larger and fewer alveoli in the BPD group compared with the NC normoxia group. Cytotherapy, using MSCs, ameliorated the lung pathological changes ( Figure 1B). The mean linear intercept ( Figure 1C) and mean alveolar volume ( Figure 1D) were used to perform statistical analysis of alveoli. All MSC therapies reduced the hyperoxia-induced mean linear intercept and mean alveolar volume increased. The UC-MSC group showed a significant reduction of the histological indexes of alveoli compared with BM-MSC groups.

| MSCs ameliorated hyperoxia-induced secreted cytokine changes in circulatory system and lung tissue
The secreted cytokines from rats were assayed using ELISAs. For the circulatory system, three plasma cytokines were assayed, including proinflammatory factor TNFα (Figure 2A), angiogenesis-associated factor VEGFA and ET1 ( Figure 2B

| MVD and cell death in lung tissue were improved by MSC cytotherapy
Immunohistochemistry with the CD34 antibody was used to discriminate vessels, and TUNEL staining labelled the apoptotic cells in lung tissue ( Figure 3A). In the NC group, most of alveoli were covered by For cell death statistical analysis, hyperoxia-induced apoptosis was attenuated by all MSC cytotherapies, and there was no significant difference between MSC groups ( Figure 3C).

| Hyperoxia-induced angiogenesis and epithelialization-associated protein degradation in lung was ameliorated by MSC cytotherapies
The protein level in lung tissue was determined by Western blotting for angiogenesis (CD34, CD31 and VEGFA) and epithelialization (CDH5) markers ( Figure 4A). There was significant decrease in all markers in the BPD group compared with the NC group. For the angiogenesis-associated markers, a similar trend was present, as there were significant increases in the CD34, CD31 and VEGFA levels in MSC cytotherapy groups compared with the BPD group The transplantation of cultured MSCs was used for medicinal purposes to regenerate the microenvironment during high levels of inflammatory cytokines. 34 MSCs inhibit immunoactivation-induced cytotoxicity 34 and promote endothelial cells forming functional structures. 35 The mechanism of MSC cytotherapy was to reduce inflammation, prompt angiogenesis and avoid cell death, as it is difficult to be stripped for single factor in vivo study, while those bio-functions have to be proved by independent experiments in vitro and observed in vivo. 10 For the hyperoxia-induced BPD animal model, as shown in a clinical setting, the usage of IT MSC treatment is an effective solution to attenuate its pathological changes, 10 while the engrafted MSCs share a less than ten percentage of lung after 14 and 21 days after transplanting and adopt a type II alveolar epithelial cell (AEC2) phenotype. 18 As safety and effectiveness of MSCs treat BPD were taken by some independent study group, and MSCs were already translated in to clinical use for curing BPDs. 4 In this study, the effectiveness of IT MSC treatments were confirmed by the improved BPD-associated physiology and biochemical changes in all three MSC treatment groups, but there were differences in groups that need to elucidate clearly before clinical trials.
MSCs from different source present different biological functions and therapeutic effects. As MSCs may be derived from various biological sampling, the difference of them was cared by some independent research groups. The most widely used MSCs were BM-MSCs and UC-MSCs, which is derived from Wharton's jelly. The difference between the MSCs involved the WNT pathway growth and differentiation potential. 36 In the comparison of BM-MSCs and UC-MSCs, there were differences in mRNA expression and protein secretion level 37,38 and UC-MSCs had enhanced proliferation and immunosuppression ability, 37,39 which had high expression of IL6, IL8, PDGF, HGF and TGF-β2. 37 AF-MSCs were a mixture of cells derived from lung, kidney and foetal membrane, which is considered as potential for regenerative medicine 40 and immunomodulatory. 41 In our previous study, AF-MSCs were derived from second-trimester amniotic fluid for cytotoxicity 42  The funding sources and sponsors did not participate in the design of the study; the collection, analysis, and interpretation of data; or in writing the manuscript.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated or analysed during this study are included in this published article.