A therapeutic role for mesenchymal stem cells in acute lung injury independent of hypoxia-induced mitogenic factor

Abstract Bone marrow mesenchymal stem cells (BM-MSCs) have therapeutic potential in acute lung injury (ALI). Hypoxia-induced mitogenic factor (HIMF) is a lung-specific growth factor that participates in a variety of lung diseases. In this study, we evaluated the therapeutic role of BM-MSC transplantation in lipopolysaccharide (LPS)- induced ALI and assessed the importance of HIMF in MSC transplantation. MSCs were isolated and identified, and untransduced MSCs, MSCs transduced with null vector or MSCs transduced with a vector encoding HIMF were transplanted into mice with LPS-induced ALI. Histopathological changes, cytokine expression and indices of lung inflammation and lung injury were assessed in the various experimental groups. Lentiviral transduction did not influence the biological features of MSCs. In addition, transplantation of BM-MSCs alone had significant therapeutic effects on LPS-induced ALI, although BM-MSCs expressing HIMF failed to improve the histopathological changes observed with lung injury. Unexpectedly, tumour necrosis factor α levels in lung tissues, lung oedema and leucocyte infiltration into lungs were even higher after the transplantation of MSCs expressing HIMF, followed by a significant increase in lung hydroxyproline content and α-smooth muscle actin expression on day 14, as compared to treatment with untransduced MSCs. BM-MSC transplantation improved LPS-induced lung injury independent of HIMF.


Introduction
Acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) are life-threatening conditions characterized by high mor-tality, rapid onset of respiratory failure following extensive inflammation, lung oedema and alveolar haemorrhage [1]. In ALI and ARDS, inflammatory cells migrate into the lung parenchyma and release mediators that destroy bacteria but also cause tissue disruption and lung injury [2]. Although supportive therapies for these conditions exist, no efficient therapies or targeting strategies have been developed. Mesenchymal  down-regulation of nitric oxide metabolites and pro-inflammatory and angiogenic cytokines [3]. In addition, recent work has also demonstrated the derivation of pulmonary progenitor cells such as type II pneumocytes from embryonic stem cells (hESCs) [4], as well as integrin-dependent differentiation and adhesion of stem cells in lung tissue [5].
Hypoxia-induced mitogenic factor (HIMF), an autocrine and paracrine cytokine primarily expressed in airway and alveolar epithelial cells, plays a number of roles in lung disease, including lung cell proliferation and development, in addition to mitogenic, angiogenic and vasoconstrictive effects [6][7][8]. Furthermore, HIMF has anti-apoptotic functions that can protect against the development of lipopolysaccharide (LPS)-induced ALI through the reservation of surfactant protein C production [9], although previous studies have shown that HIMF can promote cell proliferation, migration and production of vascular endothelial growth factor and monocyte chemotactic protein-1 in pulmonary endothelial cells, as well as the production of reactive oxygen species in murine monocyte/macrophage cells. In addition, HIMF up-regulates pro-angiogenic and pro-inflammatory factors, recruits inflammatory cells into the lungs and promotes inflammation in mice with pulmonary hypertension [10].
In the present study we sought to investigate the therapeutic role of bone marrow MSC (BM-MSC) transplantation in LPSinduced ALI and fibrosis and to explore the potential additive and/or synergistic effects of HIMF on BM-MSC transplantation. First, we transplanted BM-MSCs without any manipulation to prevent LPS-induced lung oedema and inflammation from developing. Second, in order to maintain the consistent production of HIMF, BM-MSCs were transfected with virus encoding HIMF or other vectors, then transplanted to treat LPS-induced lung injury and fibrosis.

Isolation, culture and identification of BM-MSCs
Adult BALB/c mice were obtained from Slac Laboratories, Shanghai, China. The experimental protocol was approved by the Committee of Animal Care and Use of Fudan University, according to National Institutes of Health guidelines. BM cells were collected as previously described [11]. Briefly, BM-MSCs were obtained from the femurs and tibias of 6-week-old BALB/c mice and resuspended at 10 6 [11].

Experimental design
The experimental design is outlined in Figure 1.

Histological examination and immunohistochemical staining for ␣-smooth muscle actin (SMA) and hydroxyproline quantification
Tissue sections were embedded in paraffin and prepared for haematoxylin-eosin staining. Lung sections were microscopically examined by a board-certified pathologist blind to the treatment conditions according to a standard, previously described method [13]. A total lung injury score was calculated as the sum of the four criteria at an average of 10 slices. Intra-alveolar septal thickness was measured in different groups. Paraffinembedded lung sections were used for immunocytochemistry via anti-␣-SMA antibody (Sigma-Aldrich) at day 14. The total hydroxyproline content of the lung tissue was measured as an assessment of lung collagen content after homogenization of lung tissue in cold PBS according to the manufacturers' instructions (Nanjing Jiancheng Biological Engineering Institute).

Statistical analysis
Data are represented as means Ϯ S.E.M. Differences between groups were assessed using analysis of variance. A value of P Ͻ 0.05 was considered statistically significant. Analyses were done using SPSS 11.5 software.

MSC transplant improved LPS-induced lung injury independent of HIMF
The severity of LPS-induced lung injury was evaluated using a semiquantitative histopathology scoring system. Lung injury occurred after LPS challenge, with the peak of injury observed on 3 and 7 days after treatment with PBS (Fig. 4A). The severity of LPS-induced lung injury was slightly improved 3 days after MSC transplantation and significantly improved after 7 days (*P Ͻ 0.05). Transplantation of MSCs-null vector resulted in significant effects after 7 days, similar to treatment with MSCs alone (Fig. 4B), whereas treatment with MSCs-HIMF did not result in additional therapeutic effects. Unexpectedly, the level of injury in animals treated with MSCs-HIMF was significantly higher than those treated with PBS, MSCs alone or MSCs-null vector (*P Ͻ 0.05, Fig. 4B).
Histological assessment of lung sections after administration of LPS revealed evidence of marked inflammatory infiltration, interalveolar septal thickening and alveolar congestion and haemorrhage in animals challenged with LPS and treated with PBS. LPS-induced infiltration of inflammatory cells, interstitial oedema and alveolar collapse were significantly alleviated 7 days after transplantation of MSCs alone or MSCs-null vector, but not MSCs-HIMF (Fig. 5). Increased interalveolar septal thickening and fibroblast proliferation were observed 14 days after transplantation of MSCs-HIMF (Fig. 5). Intra-alveolar septal thickness increased 4fold 3 days after LPS exposure (Fig. 4C) 0.05, Fig. 4C and D). No such effect was observed after treatment with MSCs-HIMF. In fact, intra-alveolar septal thickness increased by 74% 14 days after transplantation of MSCs-HIMF, significantly higher than that observed in PBS-treated animals (Fig. 4D, P Ͻ 0.05).

. Transplantation of MSCs or MSCs-null vector prevented LPS-induced increases in intraalveolar septal thickness on day 7 compared to treatment with PBS (P Ͻ
Lung W/D ratios and MPO activity were significantly higher in animals treated with PBS 3 days after LPS challenge. This effect was significantly inhibited by transplantation of MSCs (P Ͻ 0.05 or 0.01, Fig. 6A Fig. 6C and D). The total protein concentration in BAL fluid was significantly higher 3 days after LPS challenge (P Ͻ 0.05 versus day 0, Fig. 6E) 0.05, Fig. 6F).

MSCs treatment increased IL-10 and decreased TNF-␣ expression levels
Levels of TNF-␣ mRNA expression were significantly higher in lung tissue 3 days after LPS challenge in animals treated with PBS when compared to day 0 (P Ͻ 0.05, Fig. 7A). Levels of TNF-␣ mRNA in the lungs of animals treated with MSCs or MSCs-null vector were significantly lower than those treated with PBS on days 3 and 7 after LPS administration, but remained higher than observed on day 0 (P Ͻ 0.05). There was a significant increase  0.05, Fig. 7A  and B). Lung levels of TNF-␣ protein were consistent with the changes observed in TNF-␣ mRNA expression (Fig. 7C and D). Levels of IL-10 mRNA (Fig. 7E and F) and protein (Fig. 7G and H) were significantly higher 7 days after LPS challenge in lung tissue harvested from animals treated with MSCs or MSCs-null vector than those treated with PBS or MSCs-HIMF (P Ͻ 0.05). collagen deposition and fewer ␣-SMA ϩ cells after LPS treatment (Fig. 8A-D). Levels of hydroxyproline in the lungs of animals treated with MSCs alone or MSCs-null vector were significantly lower than those treated with PBS 14 days after LPS challenge ( Fig. 8E and F, P Ͻ 0.05). Fig. 4 The histopathology score of lung injury was evaluated after LPS challenge (A, B). Transplantation of MSCs prevented LPS-induced increase in lung injury score on day 7 compared to treatment with PBS (*P Ͻ 0.05), but treatment with MSCs-HIMF had no effect. The scores increased significantly seven days after transplantation of MSCs-HIMF compared with MSCs-null vector (*P Ͻ 0.05). Intra-alveolar septal thickness in lung sections of various treatment groups was evaluated (C, D). In animals treated with MSCs, intra-alveolar septal thickness was significantly reduced compared with controls on day 7 (*P Ͻ 0.05), whereas transplantation of MSCs-HIMF resulted in an increase in intra-alveolar septal thickness on day 7 and 14 compared with MSCsnull vector (*P Ͻ 0.05). Data are expressed as mean Ϯ S.D. mediators. LPS, a glycolipid of the outermost membrane of gramnegative bacteria, was used in this study to induce ALI and sub-ALI, as measured by hyper-production of the pro-inflammatory factor TNF-␣, neutrophil infiltrates, alveolar congestion, haemorrhage and thickness of the alveolar wall. This was followed by an analysis of the therapeutic potential of MSC treatment, with or without HIMF expression. MSCs are widely used as therapeutic tools in cell and gene therapy due to their pluripotency, easy isolation and culture, and host compatibility [11,15]. HIV-1-based lentiviral vectors are an efficient method for transduction of MSCs, with transgene expression sustained for up to 5 months [16].

ALI is characterized by extensive damage to the barrier of the lung epithelium and endothelium, neutrophil influx into the lung and an imbalance between pro-inflammatory and anti-inflammatory
A recent study demonstrated that intravenous injection of MSCs alone could reduce systemic and pulmonary cytokine levels in animals with abdominal sepsis, prevent the occurrence of ALI and organ dysfunction, as well as promote phagocytosis and bacterial clearance [17]. It was noticed that circulating stem-progenitor cells were elevated along with mobilizing cytokines in neonatal respiratory distress syndrome [18]. This suggests that the body mobilizes its own stem-progenitor cells when ALI/ARDS initially occurs, followed by a decompensation in stem cell function if the disease becomes more serious. This could potentially explain why the transplantation of stem cells into the lung could ameliorate LPS-induced ALI and bleomycin and asbestos-induced chronic lung injury, although the exact mechanism by which stem cells function in this context remains unclear. Previous studies suggested that the use of MSCs may support the injured lung directly or indirectly via stimulation of resident cells in the lung, and that MSC transplantation was a potentially therapeutic strategy for lung disease [19]. In the present study, BM-derived MSCs were cultured as described previously [20]. We found that systemic transplantation of MSCs could reduce LPS-induced lung injury, leucocyte recruitment, over-production of inflammatory HIMF, a hypoxia-driven proliferation-promoting cytokine, plays complex roles in lung injury, such as promotion of vascular adhesion molecule-1 and vascular endothelial growth factor expression and increased inflammatory cell infiltration of the lung parenchyma, which worsens lung injury [21]. Alternatively, HIMF can exert anti-apoptotic functions by regulating surfactant protein C production and improving ALI [9]. We did not observe significant improvement after transplant of MSCs-HIMF in LPS-induced lung inflammation, tissue injury and repair. Not only did treatment with HIMF-transfected MSCs fail to result in significant improvement, but also exacerbated epithelial injury. Specifically, we observed increased intra-alveolar septal thickness and pulmonary vascular leak, and worsened lung tissue oedema and fibrosis in MSCs-HIMF-treated animals compared to treatment with MSCs alone or MSCs-null vector after LPS challenge. Therefore, we conclude that MSCs-HIMF transplantation contributed to increased pro-inflammatory factor expression and lung injury, and that HIMF may serve as an important pro-inflammatory but not anti-inflammatory mediator in ALI. We found that LPS initiated the development of fibrotic proliferation after 5 days, as characterized by the increased proliferation of fibroblasts and the de novo appearance of myofibroblasts with a distinct ␣-SMA-expressing phenotype [22]. HIMF was found to stimulate type I collagen and ␣-SMA expression in lung fibroblasts [23]. The results from the present study demonstrate that the addition of HIMF could increase tissue levels of hydroxyproline and ␣-SMA expression in the lung 14 days after LPS challenge. It seems that HIMF promotes consistent expression of pro-inflammatory factors, development of chronic lung injury and progressive fibrosis.
In conclusion, MSC transplantation had therapeutic effects on LPS-induced overproduction of inflammatory mediators, leucocyte influx and lung injury. However, the combination of HIMF with MSC therapy failed to show any additive effects on the reduction of LPSinduced lung inflammation and injury. Thus, our data indicate that