Umbilical cords (n=10, normal pregnancies, approved by the human research Ethics Committee of the Qilu Hospital, Shandong University) were excised and washed in a 0.1 mol/L phosphate buffer (pH 7.4) to remove excess blood. The cords were dissected and the blood vessels were removed. The remaining tissue was cut into small pieces (1–2 mm³) and placed in plates with low-glucose Dulbecco-modified Eagle medium (L-DMEM) (Gibco-BRL, Grand Island, NY, USA), supplemented with 10% fetal bovine serum (FBS; Gibco-BRL), 2 ng/mL of epidermal growth factor (R&D Systems, Minneapolis, MN, USA), 2 ng/mL of fibroblast growth factor (R&D Systems), 2 ng/mL of (VEGF; R&D Systems), and 100 µg/mL of streptomycin (Gibco-BRL), 100 U/mL penicillin. Cultures were maintained with 5% CO₂ at 37℃ in a humidified atmosphere. The media were changed every 3–4 days. Adherent cells proliferated from individual explanted tissues 7–12 d after initiating incubation. At this time, the small tissue pieces were removed from the culture, and the adherent fibroblast-like cells were cultured to confluence, which subsequently took 2–3 weeks in culture. The cells were then trypsinized using 0.25% trypsin (Gibco-BRL) and passaged at 1×10⁴ cells/cm² in the medium described above. The cells were used after four passages.

P4 cells were cultured under conditions appropriate for inducing the differentiation of each lineage to investigate the differentiation potential of the fibroblast-like cells. Cells were seeded at a density of 3×10³ cells/cm² and the differentiation media were changed every 3–4 days. Osteogenic differentiation of MSCs assays were conducted in L-DMEM supplemented with 0.1 µM dexamethasone, 10% FBS, 0.2 mM ascorbic acid, and 10 mM β-glycerol phosphate (Sigma-Aldrich, St. Louis, MO, USA). Adipogenic assay was performed with high-glucose DMEM supplemented with 1 µM hydrocortisone, 0.5 mM 3-isobutyl-1-methylxanthine, 10% FBS, and 0.1 mM indomethacin (Sigma-Aldrich). Chondrogenic inductive medium consisted of 10% FBS, 3 ng/mL of TGF-β₁, 0.2 mM ascorbic acid, 2 µg/mL of insulin, 30 ug/mL of sodium pyruvate, 0.1 umol/L of dexamethasone, 5 mM β-glycerin sodium phosphate, 0.4 mg/mL of bovine serum albumin, 2 ug/mlL of linoleic acid, and 2 ug/mL of alizarinic acid (Sigma-Aldrich).

Cells cultured without differentiation media served as the control group. UCMSCs were labeled with antibodies CD₂₉, CD₃₄, CD₄₄, CD₄₅, CD₇₃, CD₃₁, CD₉₀, CD₁₀₅, CD₁₃₃, and corresponding isotype controls (from BD Biosciences), and were analyzed by flow cytometry (Cytometer 1.0, CytomicsTMFC500, Beckman Coulter Company, Brea, CA, USA).

Human 293T cells were maintained in DMEM with 10% FBS. Lentiviral vectors expressing short hairpin RNAs (shRNAs) against Ang1 (NM_001146) were synthesized as the following: forward: 5'-GAGGATCCCCGGGTACCGGTCGCC ACCATGA CAGTTTTCCTTTCCTTTG-3', and reverse: 5'-TCCTTGTAGTC CATA CCAAAATCTAAAGGTCGAATCATC-3'. The non-silencing shRNA sequences were used as negative controls and synthesized as follows: forward, 5'-CCGGTTCT CCGAACGTGT CACGTCTCGAGACGTGACACGTTCGGAGAATTTTTG-3', and reverse, 5'-AATTCAAAAATTCTCCGAACGTGTCACGTCTCGA GACGTGACAC GTTCGGAGAA-3', obtained from Gene Chem Co., Ltd (Shanghai, China). The coding sequence of human Ang1 (1497bp) was checked from NCBI. For over- expression of Ang1, the open reading frame of Ang1 (NM-001146) was cloned into the lentiviral vector GV287 (Ubi-MCS-3FLAG-SV40-EGFP; Gene-Chem Co., Ltd). Information on vector GV287 is provided in Fig. 3A.

For in vivo study, P4 MSCs (1×10⁵/mL) were transduced with Ang1 plasmid or empty vector plasmid at a multiplicity of infection of 8. The expression of Ang1 in UCMSCs was observed on a fluorescence microscope after 48 hours of lentiviral infection via fiuorescent microscopy and flow cytometry, and cells were harvested for western-blotting analysis to detect the protein expression of Ang1.

Sprague-Dawley rats (male, eight weeks old, and weight 200±20 g) were purchased from the Beijing Hua-Fukang Laboratory. All experimental protocols were approved by the animal Ethics Committee of Shandong University. The rats were randomized into a control group with normal saline (NS), LPS group, fibroblasts group, UCMSCs group (LPS+UCMSCs), and UCMSCs-Ang1 (LPS+UCMSCs-Ang1) treated group (n=25 for each group). Except the control group, ALI model groups were induced by the injection of LPS from E. coli 055:B5 (Sigma-Aldrich) (10 mg/kg of intra-peritoneal). and after 1 hour, the four groups were given 300 µL of NS , adult human lung fibroblasts (MRC-5) (5×10⁵ cells in 300 µL of NS), UCMSCs-green fluorescent protein (GFP) (5×10⁵ cells in 300 µL of NS) or UCMSCs-Ang1 (5×10⁵ cells in 300 µL of NS) via injection into the tail vein respectively. GFP was used as a report gene to trace the UCMSCs. Five rats from each group were sacrificed at different points (6 hours, 24 hours, 48 hours, 8 days, and 15 days post-injection of LPS) and serum, lung tissue, and bronchoalveolar lavage fluid were harvested for cytokine measurements, as well as the assessment of lung histology and lung injury.

The right lower lung lobes of animals were used for histological study. Five-µm thick sections were obtained from paraffin-embedded lungs and stained with hematoxylin and eosin for histological study. Lung injuries were scored using the following categories: neutrophil infiltration into the airspace or vessel wall, alveolar congestion, vascular dilatation, hemorrhage, and thickness of alveolar wall/hyaline membrane formation. The category was graded on a 0 to 4 point scale: 0=no injury; 1=injury up to 25% of the field; 2=injury up to 50% of the field; 3=injury up to 75% of the field; and 4=diffuse injury.9

The trachea was isolated firstly, and then, the right bronchial tube was ligated after the rat was euthanized. The left lung was lavaged three times in situ with 3 mL of sterile NS at room temperature, and the recovered fluid was pooled. Cells, including neutrophils, were counted with a hemocytometer, and the BALF was centrifuged (3 kg, 20 min). The supernatant was frozen (-80℃) until ready for further analysis.

Myeloperoxidase (MPO) activities in the homogenized lung tissues were evaluated to quantify the neutrophil infiltrations.10 The lung tissues were homogenized in a phosphate buffer (20 mM, pH 7.4) and then centrifuged (at 30 kg for 30 min) after thawing. The pellet was resuspended in a potassium phosphate buffer (50 mM, pH 6.0) with 0.5% hexadecyltrimethyl ammonium bromide. The samples were centrifuged (at 20 kg for 15 min at 4℃). The absorbances were measured by spectrophotometry at 460 nm 0.0005% hydrogen peroxide and 0.167 mg/mL of O-dianisidine hydrochloride were added to each sample. The results were expressed as units of MPO per gram of wet tissue.

The right upper lobes were used for quantifying the magnitude of pulmonary edema. They were placed into previously-weighed microcentrifuge tubes and weighed. The lungs were desiccated under a vacuum at 80℃ for 24 hours and weighed again. The wet lung mass was divided by the dry lung mass to give the wet-dry ratio.

The levels of IL-10, TGF-β₁, TNF-α, and IL-6 (ebioscience, USA) in serum (from blood drawn by cardiac puncture, and centrifuged for 10 min at 2 kg) were measured by ELISA according to the manufacturer's instructions.

Five groups of rats (n=10 per group) were used for survival study. LPS, fibroblast cells, UCMSCs, and UCMSCs-Ang1 were administered, respectively. Then, the rats were allowed to recover and the mortality was recorded till to 15 days after the treatment.

The lung tissues were cut into 5-µm thick sections and then observed by Laser confocal microscope. The distribution of UC-derived MSCs-Ang1 was analyzed by tracing GFP-positive cells. Twenty fields were randomly selected at ×40 magnification, and the GFP-positive and GFP-negative cells were counted. RT-PCR was used to detect of exogenous GFP mRNA expression in receptor lung tissue (Takara, Japan).

Analyses were done using SPSS 7.0 software (SPSS Inc., Chicago, IL, USA) and Graph-Pad Prism software (La Jolla, CA, USA). Data were shown as mean±SD. We used a one way ANOVA adjusted for multiple comparisons for statistical analysis. We used Kaplan-Meier survival curves and log rank to compare survival rates. In all instances, we considered values of p<0.05 to be statistically significant.

Most of the cells exhibited a spindle or irregular polygon shape that is consistent with MSCs (Fig. 1A and B). They expressed typical MSCs cell surface antigens, such as CD₂₉, CD₄₄, CD₉₀, and CD₁₀₅. Non-hematopoietic or endothelial cell markers, such as CD₃₁, CD₃₄, CD₁₃₃, and CD₄₅ could not be detected (Fig. 1C) The adherent cells induced to differentiate in culture exhibited multipotent differentiation potential, as shown by their ability to form osteoblasts and adipocytes. Osteoblasts were confirmed by alizarin red staining and alkaline-phosphate staining by the end of 14 days (Fig. 2A and B). In the normal growth medium, no mineralized matrix was observed in the control cells. Almost all cells contained numerous oil red-O-staining lipid droplets by the end of the second week (Fig. 2C and D). Cartilage structures were observed by immunohistochemical detection of cartilage specific type II collagen expression and toluidine blue staining after 2 weeks (Fig. 2E and F).

Results from functional differentiation experiments further indicated that the MSCs at passage 3 had strong adipogenesis, osteogenesis, and cartilaginous, respectively, under a given medium culture.

After obtaining enriched UCMSCs (P4), the cells were transduced with lentivirus vector carrying the GFP report gene and the Ang1 gene. The transduction efficiency was assessed by fluorescence microscopy of lentiviral infection. The transduction efficiency was about 95% (GFP-positive cells in 200 MSCs) at 72 h after UCMSCs-Ang1 tranduction. The photomicrographs of UCMSCs-Ang1 were obtained using a fluorescence microscope in the BX model at 72 h after transfection and prior to injection into the rats (Fig. 3B). At the same time, overexpression of Ang1 mediated by lentiviral transduction was confirmed in GFP cells by FACS sorting (Fig. 3C). Furthermore, the effectiveness of the lentiviral infection of LV-Ang1 was also confirmed by western-blot analysis. The protein expression of Ang1 in UCMSCs was significantly higher in the Ang1 overexpression group than that in the control group at 72 h after transfection (Fig. 3D).