The experimental protocols described herein were reviewed and approved by the Animal Care and Use Committee of Samsung Biomedical Research Institute, Seoul, Korea. This study was also performed in accordance with institutional and National Institutes of Health (NIH) guidelines for laboratory Animal Care. Timed pregnant Sprague-Dawley rats (Orient Co., Seoul, Korea) were housed in individual cages with free access to water and laboratory chow. Rat pups were delivered spontaneously and reared with their dams. The experiment began within 10 h after birth and continued through postnatal day (P) 14.

Rat pups were randomly divided into four groups with six to eight animals in each group: normoxia control group (NC); normoxia with G-CSF treatment group (NG); hyperoxia control group (HC); and hyperoxia with G-CSF treatment group (HG). Rat pups in the NC and NG groups were kept with nursing mothers in standard cages and normal room air throughout the experiment, and pups in the HC and HG groups were maintained with nursing mothers in standard cages within 50 L Plexiglas chambers in which an oxygen concentration of 90% was maintained. The humidity and environmental temperature were maintained at 50% and 24℃, respectively. Nursing mother rats were rotated daily between litters in the normoxia and hyperxoxia groups to avoid oxygen toxicity.

Rat pups in the NG and HG groups were injected intraperitoneally with 20 µg/kg of recombinant human G-CSF (Dong-A Pharmaceutical Co. Ltd., Seoul, Korea) in 5% dextrose solution (2 µg/mL) on P 4, 5 and 6. Pups in the NC and HC groups received the same volume of 5% dextrose solution injected in the same manner. The survived rat pups in each group were weighed daily until sacrifice under deep pentobarbital anesthesia (60 mg/kg, intraperitoneal) at P 14. Morphometric and biochemical analyses of whole lung tissue was done.

For biochemical analyses, transcardiac perfusion using ice-cold phosphate buffered saline (PBS) was done, and then lungs were resected, snap-frozen in liquid nitrogen, and refrigerated at - 80℃.

Lungs were inflated by instillation of 10% buffered formalin with a constant pressure of 20 cm H₂O, and fixed at room temperature through the night for morphometric analyses. Then the fixed lung tissue was processed, and embedded in paraffin. Four-micrometer-thick sections of paraffin blocks were stained with hematoxylin and eosin. Images of each field were captured using a digital camera through an Olympus BX40 microscope (Olympus Optical Co. Ltd., Tokyo, Japan).

MLI and mean alveolar volume were measured to evaluate the level of alveolarization. The mean inter-alveolar distance was measured as MLI, by dividing the total length of the lines drawn across the lung section by the number of intercepts encountered, as described by Thurlbeck.16 The mean alveolar volume was calculated using the method reported by Snyder, et al.17 Briefly, a grid containing equally spaced crosses was placed on a uniformly enlarged photomicrograph of each lung field. The diameters (ℓ) of the alveoli containing crosses were measured along the horizontal axes of the crosses. The cube of the alveolar diameter times π and divided by 3 (ℓ³π/3) was used to estimate the mean alveolar volume. A minimum of two sections per rat and six fields per section were selected randomly and examined for each analysis. Analyses were carried out by two independent observers who were blind to treatment conditions.

Immunofluorescent TUNEL staining (kit S7110 ApopTag, Chemicon, Temecula, CA, USA) was done to determine the extent of apoptosis in the lung. Deparaffinized and rehydrated paraffin section slides were digested for 15 min using proteinase K (20 µg/mL in PBS) (Sigma Co., St. Louis, MO, USA) at room temperature, washed for 10 min with PBS. Sections were incubated using working strength TdT enzyme at 37℃ for 1 h after incubated using equilibration buffer for 1 min, and then immersed in a stop/wash buffer and rinsed. The sections tagged with fluorescein isothiocyanate (FITC)-labeled anti-digoxigenin conjugate were incubated at room temperature for 30 min in the dark, and propidium iodide (0.5 µg/mL, Sigma Co.) was used for nuclear counterstaining. Slides were mounted with a Vectashield mounting solution (Vector Laboratories, Burlingame, CA, USA), and visualized by fluorescent microscopy (Nikon E600 fluorescence microscope, Tokyo, Japan). The number of TUNEL positive cells were counted in ten non-overlapping fields with a magnifying power of ×200.

The MPO activity which is an indicator of neutrophil accumulation, was made by modification of the method reported by Gray, et al.18 The homogenized lung tissues in a phosphate buffer (pH 7.4) were centrifuged at 30,000 g for 30 min, resuspended in another phosphate buffer with 0.5% hexadecyltrimethyl ammonium bromide (50 mM, pH 6.0), and measured absorbance changes at 460 nm, using 0.167 mg/mL of O-dianisidine dihydrochloride and 0.0005% hydrogen peroxide to assay the activity of MPO. One unit of MPO activity equaled the amount of enzyme to degrade 1 µmole of peroxide per minute.

The total RNA in the sample was extracted by using a RNA Trizol kit (Invitrogen Corporation, Carlsbad, CA, USA) according to the manufacturer's protocol. One microgram of RNA was used to produce cDNA with an Protoscript® II RT-PCR kit (New England Biolabs, Ipswich, MA, USA). PCR primers for rat TNF-α, IL-6, TGF-β and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were designed with Primer 3 (Whitehead Institute, Cambridge, MA, USA) and were synthesized by Bioneer Inc. (Bioneer, Daejeon, Korea). The sequence of primers used was as follows: (TNF-α) sense 5'-TACTGAACTTCGGGGTGATCGGTCC-3', antisense 5'-CAGCCTTGTCCCTTGAAGAGAACC-3' (IL-6) sense 5'-CACCAGGAACGAAAGTCAACTC-3', antisense 5'-GGAAGCATCCATCATTTCTTTG-3' (TGF-β) sense 5'-CAACTGTGGAGCAACACGTAGA-3', antisense 5'-CAACCCAGGTCCTTCCTAAAGT-3' (GAPDH) sense 5'-CTCTACCCACGGCAAGTTCAA-3', antisense 5'-GGGATGACCTTGCCCACAGC-3'.

One µL of cDNA was mixed in 19 µL of reaction mixture of PCR Master Mix (Bioneer, Daejeon, Korea) and then 0.5 µmole of each primer (Bioneer) was added for each reactions which were carried out in a T1 thermocycler (Biometra, Goettingen, Niedersachsen, Germany). The cycle profile involved was as follows: 60 s at 95℃, 60 s at X℃, where X is the annealing temperature for each pair of cytokine primers, and 60 s at 68℃. The PCR products were visualized by ethidium bromide after separated using 1.2% agarose gel electrophoresis. Then they were scanned by a Gel Doc 2000 analyzer (Bio-Rad Laboratories, Hercules, CA, USA). The mRNA level of TNF-α, IL-6, and TGF-β were analyzed densitometrically with Quantity One software (Bio-Rad Laboratories, Hercules, CA, USA), which were estimated from the density ratio of cytokines to GAPDH (control).

p67^phox is a subunit of NADPH oxidase, the translocation of which from the cytosol to the plasma membrane generates reactive oxygen species (ROS) and reflects NADPH oxidase activation.19 Lung tissues were fractionated into membrane and cytosolic components for Western blot analysis. Homogenized tissues in ice-cold hypertonic solution were centrifuged at 600 g for 10 min, and the supernatant was centrifuged at 100,000 g for 1.5 h. Then the supernatant containing both the cytosolic and the membrane-particulate pellet was resuspended in hypotonic solution containing 1% Triton X-100. Western blot analysis of these products using antibodies against the p67^phox (1 : 500, BD Biosciences, San Diego, CA, USA), the NADPH oxidase cytosolic subunit was done, and the bands were recognized by horseradish-peroxidase (HRP)-conjugated anti-mouse secondary antibody (1 : 1,000, DAKO, Glostrup, Denmark). Developed Western blots using enhanced luminol-based chemiluminescent (ECL) detection reagents (Amersham Pharmacia, Uppsala, Sweden) were exposed to X-ray film (Fuji, Tokyo, Japan), and re-probed with antibodies against glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1 : 1,000, Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA). The membrane fraction was presented to immunoblotting for calnexin (1 : 500 Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), a membrane marker to determine the relative degree of membrane purification.

The data are expressed as mean ± SEM. Survival curve comparisons were performed using Kaplan-Meier analysis followed by a log rank test. For continuous variables with a normal distribution, ANOVA with a Bonferroni correction was performed. For variables without normal distribution, we performed Wilcoxon rank sum tests with Bonferroni corrections. p values < 0.05 were considered significant.

Exposure to oxygen (HC) reduced the survival rate of the animals to 88% at the end of the experiment (P 14) compared to the 100% survival rate of the normoxia groups (NC and NG), and this reduced survival rate in HC was improved to 92% with G-CSF treatment (HG), however, this improvement was not statistically significant.

Although birth weight was not significantly different between the four experimental groups, body weight at P 14 in HC was significantly lower compared to the normoxia groups, and this decreased body weight gain observed in HC was significantly improved with G-CSF treatment (HG) (Fig. 1).

Representative photomicrographs showing differences between the experimental groups are shown in Fig. 2. Fewer and larger alveoli were observed in HC pups compared to pups in the normoxia groups, indicating hyperoxia-induced impairments in alveolar growth, and these abnormalities seemed to be attenuated with G-CSF treatment (HG). Morphometric data demonstrated significantly increased MLI and mean alveolar volume in HC compared to normoxia groups (NC and NG), and significant attenuation of these abnormalities with G-CSF treatment (HG) support these microscopic findings (Fig. 3).

The numbers of TUNEL positive cells in distal lung samples were markedly increased in HC compared to the normoxia groups (NC and NP), and G-CSF treatment (HG) significantly attenuated this hyperoxia-induced increase in the number of TUNEL positive cells (Fig. 4).

Exposure to oxygen (HC and HG) significantly increased white blood cell counts in the blood in hyperoxia, compared to the normoxia groups (NC and NG), and the G-CSF induced granulopoiesis was not observed at P14 (Fig. 5A).

The significantly increased MPO activity observed in HC, compared to the normoxia groups (NC and NG), was significantly improved with G-CSF treatment (HG) (Fig. 5B).

In semi-quantitative RT-PCR, significantly increased mRNA levels of TNF-α, IL-6 and TGF-β were observed in HC, compared to the normoxia groups (NC and NG) (Fig. 6). The hyperoxia-induced increase in TGF-β mRNA expression was significantly attenuated with G-CSF treatment (HG), however, the attenuation of TNF-α and IL-6 mRNA levels was not statistically significant.

When NADPH oxidase is activated, expression and membrane translocation of a cytosolic subunit of NADPH oxidase p67^phox is increased. Hyperoxia-induced production of ROS was evaluated by NADPH oxidase activation, because NADPH oxidase produces oxygen free radicals both in phagocytic20 and nonphagocytic cells.21 Exposure to oxygen (HC) significantly increased p67^phox both in the cytosolic and membrane fractions, indicating activation of NADPH oxidase, compared to the normoxia groups (NC and NG) in Western blot analyses, and the G-CSF (HG) significantly attenuated this hyperoxia-induced NADPH oxidase activation (Fig. 7).