The experimental methods were approved by the animal research committee of Korea University and the ethics committee of Korea University Medical Center. Five-week-old specific pathogen-free male C57BL/6 mice, each weighing 20 to 25 g, were randomly divided into the following four experimental groups: (a) sham tracheostomized group (sham group, n = 18); (b) LPV group (n = 18), in which the mice were ventilated with low tidal volume (V_T) and positive end-expiratory pressure (PEEP); (c) VILI group (n = 18), in which the mice were ventilated with high V_T without PEEP; and (d) VILI with PJ34 pretreatment group (PJ34+VILI group, n = 18), in which the mice were pretreated with the PARP inhibitor PJ34 and ventilated with the same settings as in the VILI group. Each group was subdivided into three experimental subgroups: (a) tissue subgroup (n = 6) for histopathologic examination and measurements of wet-to-dry (W/D) weight ratio and PARP activity assay; (b) bronchoalveolar lavage (BAL) subgroup (n = 6) for myeloperoxidase (MPO) activity assay and measurements of inflammatory cytokine concentration and nitric oxide (NO) metabolites in BAL fluid (BALF); and (c) tissue homogenate subgroup (n = 6) for measurement of NF-κB activity in lung tissue homogenates.

Each mouse was anesthetized with an intraperitoneal injection of 65 mg/kg of pentobarbital sodium and intubated via tracheostomy. MV was performed with a rodent ventilator (Harvard Apparatus, Holliston, MA, USA). The mice in the LPV group were ventilated with a peak inspiratory pressure (PIP) of 15 cm H₂O, a PEEP of 3 cm H₂O, and a respiratory rate of 90 breaths per minute. Adequate setting for the VILI model has been determined by preliminary studies using various MV settings. Histopathologic examination of the lung tissues every 30 minutes allowed determination of the time and setting that yielded typical pathological findings of VILI 17. The typical indications developed under the following setting: PIP 40 cm H₂O + PEEP 0 cm H₂O + 90 breaths per minute. These changes were most prominent after about 2 hours of MV. Therefore, the VILI and PJ34+VILI groups were ventilated at this setting for 2 hours, and the LPV group mice were also ventilated for 2 hours. PIP and V_T were measured and monitored using a linear pneumotach (series 8430; Hans Rudolph, Inc., Shawnee, KS, USA) and research pneumotach system (model RSS 100 HR; Hans Rudolph, Inc.). Changes in dynamic compliance (C_D) between the beginning and after 2 hours of MV were calculated from the V_T, PIP, and PEEP: C_D = V_T/(PIP - PEEP). To maintain deep anesthesia, half of the initial dose of pentobarbital sodium was administered after 1 hour of MV.

After MV, the tissue subgroup mice were rapidly exsanguinated by dissecting the abdominal aorta. The heart and lungs were removed en bloc through a midsternal incision. After ligation of the left main and right upper bronchi, the left lung was excised, embedded in optimal cutting temperature compound (Tissue-Tek^®; Sakura Finetechnical Co., Ltd., Tokyo, Japan) in a cryomold, and stored at -70°C for PARP activity assay. Excised right upper lobe was weighed in a tared container and dried in an oven until a constant weight was obtained, and the W/D weight ratio was calculated. The remnant of the right lung was immediately instilled with 4% paraformaldehyde through the right main bronchus at a hydrostatic pressure of 15 cm H₂O and fixed in 4% paraformaldehyde for 48 hours. Paraffin blocks were prepared by dehydration with ethanol and embedding in paraffin.

For the BAL subgroup mice, the thorax was opened following euthanasia by exsanguination, and three BAL procedures were performed, each with 1 mL of phosphate-buffered saline (PBS). The retrieval fluid was centrifuged (2,000 g at 4°C) for 10 minutes and the supernatants were divided into aliquots and stored at -70°C until analysis for MPO activity and measurements of inflammatory cytokine concentration and NO.

The posterior portions of the right lower lobe were sectioned at a thickness of 5 μm, placed on glass slides, and stained with hematoxylin-eosin. A pathologist blinded to the protocol and experimental groups examined the degree of lung injury and graded the specimens by ALI score based on (a) alveolar capillary congestion, (b) hemorrhage, (c) infiltration or aggregation of neutrophils in the airspace or the vessel wall, and (d) thickness of the alveolar wall/hyaline membrane formation. Each item was graded according to the following five-point scale: 0, minimal damage; 1, mild damage; 2, moderate damage; 3, severe damage; and 4, maximal damage 18. The degree of VILI was assessed by the sum of scores for items 0 to 16 in five randomly selected high-power fields (HPFs) (×400). The average of the sum of each field score was compared among groups.

PARP activity in lung tissues was measured by using an immunohistochemical method of PARP activity using biotinylated NAD⁺, the substrate of the PARP 1219. Briefly, cryosections of 10 μm were fixed for 10 minutes in 95% ethanol at -20°C and then rinsed in PBS. Sections were permeabilized by incubation for 15 minutes at room temperature with 1% Triton X-100 in 100 mM Tris (pH 8.0). A reaction mixture consisting of 10 mM MgCl₂, 1 mM dithiothreitol, and 30 μM biotinylated NAD⁺ in 100 mM Tris (pH 8.0) was then applied to the sections for 30 minutes at 37°C. Reaction mixtures containing PJ34 or without biotinylated NAD⁺ were used as controls. After three washes in PBS, incorporated biotin was detected with peroxidase-conjugated streptavidin (1:100 for 30 minutes at room temperature). After three 10-minute washes in PBS, color was developed with cobalt-enhanced nickel-DAB substrate. Sections were counterstained in Nuclear Fast Red (Vector Laboratories, Burlingame, CA, USA), dehydrated, and mounted in Vectamount (Vector Laboratories). PARP activity was quantified by summing the numbers of cells positive for PARP activity in five HPFs.

The mice in the PJ34+VILI group were intraperitoneally pretreated with 20 mg/kg of PJ34 [N-(6-oxo-5,6-dihydrophenanthridin-2-yl)-N,N-dimethylacetamide, hydrochloride] (Calbiochem, Darmstadt, Germany), which is not antioxidant and does not directly interfere with the reactivity of peroxynitrite 20, and there are no reports that it has an independent inhibitory effect on NF-κB. The dose was proven to be effective in lipopolysaccharide (LPS)-induced acute lung inflammation 12. To determine optimal pretreatment time, PJ34 was administered intraperitoneally at each of 24, 12, 6, 4, 3, 2, 1, and 0.5 hours before MV to six mice at each time. The lowest PARP activity was observed after 2-hour pretreatment with PJ34. Thereafter, PARP activity increased at 1 and 0.5 hours. Therefore, the VILI+PJ34 group mice were pretreated at 2 hours before MV and the mice in the sham, LPV, and VILI groups were pretreated with 200 μL of PBS 2 hours before tracheostomy or MV.

As an indicator of activated neutrophil accumulation, a major source of reactive oxygen species (ROS), the activity of MPO was determined directly in cell-free BALF according to the method described previously 21, with minor modifications. Tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6) in BALF were measured by enzyme-linked immunosorbent assay (ELISA) (R&D Systems, Minneapolis, MN, USA). Pulmonary production of NO was determined by measuring nitrate and nitrite (NO_X), the stable end products of NO metabolism, in the BALF using an NO (NO₂^-/NO₃^-) assay kit (Assay Designs, Inc., Ann Arbor, MI, USA). Nuclear proteins from the tissue homogenate subgroup mice were prepared with a nuclear extract kit (Active Motif, Carlsbad, CA, USA). Activation of the NF-κB p65 subunit in 5 μg of nuclear extracts was measured using an NF-κB p65 ELISA-based transcription factor assay kit (TransAMTM NF-κB p65 Transcriptional Factor Assay Kit; Active Motif) 2223.

All data are expressed as mean ± standard error of the mean. Statistical analysis was performed using SPSS for Windows^® (Release 11.0.1; SPSS Inc., Chicago, IL, USA). Intergroup differences were determined by nonparametric Mann-Whitney U and Kruskal-Wallis tests. Statistical significance was defined as a P value of less than 0.05. Spearman rank correlation coefficient was used to determine the correlations between PARP activity in the tissues and the other parameters examined.

Histopathologic examination of the VILI group indicated high levels of ALI parameters (Figure 1c). These findings of lung injury were markedly reduced in the PJ34+VILI group (Figure 1d). In quantitative comparison by ALI score (Figure 1e), the VILI group (12.0 ± 0.87) showed a significantly higher score than the sham and LPV groups (1.20 ± 0.58 and 2.40 ± 0.6, respectively) (P < 0.05). The score of the PJ34+VILI group (2.67 ± 0.67) was significantly lower than that of the VILI group (P = 0.001) and was not different from those of the sham and LPV groups (P > 0.05). W/D weight ratio (Figure 2a) was also higher in the VILI group (6.28 ± 0.26) than in the sham and LPV groups (4.60 ± 0.21 and 4.33 ± 0.11, respectively) (P < 0.05). In the PJ34+VILI group, the ratio (5.05 ± 0.32) was significantly decreased relative to that of the VILI group (P = 0.012) and was similar to those of the sham and LPV groups (P > 0.05). There were no differences in C_Ds (Figure 2b) at the beginning of MV among the LPV, VILI, and PJ34+VILI groups (0.0314 ± 0.0009, 0.0307 ± 0.0012, and 0.0327 ± 0.0005 mL/cm H₂O, respectively) (P = 0.368, Kruskal-Wallis test). After 2 hours of MV, however, there were statistically significant differences between the three groups (P = 0.007, Kruskal-Wallis test); the C_D of the PJ34+VILI group (0.0284 ± 0.0006 mL/cm H₂O) was significantly higher than that of the VILI group (0.0244 ± 0.0004 mL/cm H₂O) (P = 0.021) and lower than that of the LPV group (0.0316 ± 0.0004 mL/cm H₂O) (P = 0.020).

The PARP activity assay showed large numbers of positively stained cells in the VILI group (Figure 3c). However, in the PJ34+VILI group (Figure 3d), the number of cells was decreased markedly to the levels of the sham (Figure 3a) and LPV (Figure 3b) groups, in which positively labeled cells were almost completely absent. The number of cells with PARP activity in five HPFs (Figure 3e) in the VILI group (108.75 ± 13.185) was greater than in the sham and LPV groups (19.75 ± 2.287 and 17.00 ± 7.638, respectively) (P < 0.05). The number of cells in the PJ34+VILI group (23.50 ± 3.704) was lower than in the VILI group (P = 0.002), but there were no statistically significant differences among the sham, LPV, and PJ34+VILI groups (P > 0.05). In Spearman correlation analysis, PARP activity was positively correlated with the ALI score (r = 0.950, P < 0.0001) and W/D weight ratio (r = 0.680, P = 0.015) in significance. The C_D showed negative correlation with PARP activity (r = -0.820, P = 0.002).

The optical densities (ODs) of the MPO activities in the BALF (Figure 4a) were significantly higher in the VILI group (0.109 ± 0.006 OD) than the sham (0.076 ± 0.003 OD), LPV (0.076 ± 0.001 OD), and PJ34+VILI (0.089 ± 0.004 OD) groups (P < 0.05). The PJ34+VILI group showed lower activity than the VILI group (P = 0.035) but higher activity than those of the sham and LPV groups (P < 0.05). Spearman correlation analysis showed this activity to be significantly correlated with PARP activity (r = 0.631, P = 0.004). The concentrations of NO metabolites nitrate and nitrite (NO_X) in BALF (Figure 4b) were also higher in the VILI group (7.18 ± 0.9 μM) as compared with the other three groups (P < 0.05). The PJ34+VILI group (3.76 ± 0.76 ìM) showed lower levels than the VILI group (P = 0.017) and higher levels than the sham and LPV groups (1.84 ± 0.04 and 1.98 ± 0.31 μM, respectively) (P < 0.05). NO_X level was also closely correlated with PARP activity (r = 0.523, P = 0.026).

TNF-α was not detected in BALF of the sham group, and IL-6 was not detected in the sham and LPV groups. TNF-α concentration (Figure 5a) in the VILI group (14.16 ± 2.533 pg/mL) was higher than those in the LPV and PJ34+VILI groups (3.24 ± 0.416 and 3.58 ± 0.325 pg/mL, respectively) (P < 0.05). IL-6 concentration (Figure 5b) in the PJ34+VILI group (57.85 ± 19.499 pg/mL) was lower than that of the VILI group (204.01 ± 41.846 pg/mL) (P = 0.015). The concentrations of inflammatory cytokines were correlated with PARP activity (r = 0.691, P = 0.039 for TNF-α; r = 0.699, P = 0.011 for IL-6). NF-κB DNA-binding activity measured in lung tissue homogenates (Figure 5c) was higher in the VILI group (1.51 ± 0.088 OD) than the sham and LPV groups (0.28 ± 0.056 and 0.17 ± 0.014 OD, respectively) (P < 0.05). However, NF-κB DNA binding in the PJ34+VILI group (0.91 ± 0.189 OD) was lower than that in the VILI group (P = 0.009) and higher than those in the sham and LPV groups (P < 0.05). NF-κB activity was positively correlated with those of PARP (r = 0.734, P = 0.001) and the inflammatory cytokines (r = 0.668, P = 0.035 for TNF-α; r = 0.806, P = 0.005 for IL-6).