To evaluate whether human mesenchymal stem cells (hMSCs) administration alter the clinical course of hyperoxia-induced lung injury. hMSCs were obtained from bone marrow aspirates from healthy donors after informed consent was signed, hMSCs were separated, cultured, amplified, identified and labeled with BrdU. For BrdU labeling, a sterile stock solution was added to the culture medium 48 h before the end of culture, at a final concentration of 10 micromol/L. Thirty-two 3-day old SD rats from four litters were randomly divided into four groups, as hyperoxia exposed + hMSC group (A), air-exposed + hMSC group (B), hyperoxia exposed group (C), and air-exposed group (D). The rats from the group A and the group C were placed in a sealed Plexiglas chamber with a minimal in- and outflow, providing six to seven exchanges per hour of the chamber volume and maintaining O2 levels above 95%, while the rats in the group B and the group D were only exposed to room air. Seven days later, all of them were taken out of the chamber, rats in the group A and B were injected intraperitoneally with hMSCs (1 x 10(5) in 50 microl of PBS) immediately, while the rats in the group C and D were only treated with 50 microl of PBS 3 days later. All the animals were sacrificed by an injection of sodium pentobarbital (120 mg/kg), perfused with cold 0.9% NaCl, and the left lungs were removed, the upper lobes of which were ground as tissue homogenates and used for ELISA, while the inferior lobes were stored at -70 degrees C until use for RT-PCR. The right lungs were fixed in situ for 2 h by the intratracheal instillation with 10% neutral formalin and then postfixed for 24 h. Sagittal sections (4-microm) of paraffin-embedded middle lobe and upper lobe of the right lung were used for immunohistochemistry and histology, respectively. (1) There was a significant difference in the value of RAC (raditive alveoli coant) among the 4 groups (11.145 +/- 1.331, 13.941 +/- 0.985, 9.595 +/- 0.672, 14.819 +/- 1.080, F = 43.234, P = 0.000). RAC in group A and C were significantly reduced compared with subjects in group D (P < 0.05, P < 0.05); and there was also a significant difference between group A and group C (P < 0.05), but not between group B and D subjects (P > 0.05). (2) There were significant differences in the levels of both TNFalpha and TGFbeta(1) in the homogenate of lungs among the 4 groups (142.933 +/- 24.017, 79.033 +/- 11.573, 224.088 +/- 41.915, 76.500 +/- 10.373, F = 59.970, P = 0.000; 1726.484 +/- 91.086, 1530.359 +/- 173.441, 2047.717 +/- 152.057, 1515.777 +/- 131.049, F = 24.977, P = 0.000). The levels of TNFalpha and TGFbeta1 were significantly elevated in both group A and group C when compared with subjects in group D (P < 0.05 for both). Concentrations of TNFalpha and TGFbeta1 were both significantly decreased in group A versus group C (P < 0.05 for both). There was no significant difference between group B and D subjects in the fields of TNFalpha and TGFbeta(1) (P > 0.05 for both). (3) BrdU-labelled cells were observed at alveolar wall and bronchioles in both group A and group B, and there was a significant difference in BrdU-labeled cells between two groups (0.230 +/- 0.026, 0.190 +/- 0.015; t = 3.769, P = 0.002), but none was found in group C and group D. Electrophoresis of the PCR products showed a 224 bp band, specific for Alu mRNA, in 7 of 8 rats of group A and 5 of 8 rats of group B, respectively, but no such band was found in group C and group D. hMSCs administered by intraperitoneal injection could be implanted in the lungs of newborn rats, and they could effectively protect the rats against damage to the lungs caused by hyperoxia.