The present study is a comprehensive report on the early responses of VEGF and its receptors in an animal model of ALI. We observed an increase of VEGF in the BAL, a decreased expression of VEGF and VEGFR-1, and an altered expression pattern of VEGFR-2 in the lung tissue. The VEGF levels in BAL correlated with pulmonary permeability. Decreased expression of VEGF and VEGFR-1 in the lung tissue negatively correlated with death of alveolar epithelial cells. Using cell culture as a model system, we further demonstrated that VEGF and/or VEGFR-1 may play an important role in lung epithelial cell survival.
The VEGF levels in the BAL were increased in both sham and IIR groups, which suggests that these changes may be related to hyperoxia and/or MV, applied to animals in both groups [32
]. Although it is well known that hypoxia is the most potent regulator of VEGF gene expression and protein production [35
], an oxygen-independent up-regulation of VEGF and vascular barrier dysfunction has been observed [36
]. A rise in VEGF levels in the BAL in a chronic hyperoxia model in piglets has also been reported [32
]. This could be explained at least partially by the release of VEGF from extracellular matrix through hyperoxia-induced proteolytic cleavage [33
]. Although animals in this study were ventilated with low tidal volume, we cannot exclude the contribution of mechanical factors to the release of VEGF [37
], or an addictive effect between MV and hyperoxia.
Alveolar macrophages represent a potential source of VEGF in ALI [16
]. We found a positive correlation between the VEGF levels and percentage of macrophages in the BAL. The granules in the neutrophils also contain VEGF and may represent an additional source of VEGF [38
]; however, proteases released by these cells may cleave VEGF [39
], which could explain the negative correlation between VEGF levels and the number or percentage of neutrophils in the BAL. The numbers of observations in these correlation studies are small; thus, these results should be interpreted with caution.
We noted a significant correlation of the VEGF concentrations with the total protein concentrations and with the total cell counts in the BAL. High concentrations of VEGF within the lung may contribute to the development of pulmonary edema by alternating the state of the adherens junction complexes on the endothelium [40
]. An alternative explanation for this correlation is that the increased VEGF is simply the reflection of increased protein leakage in the lung. In clinical studies, increased VEGF in plasma [16
], and decreased VEGF in epithelial lining fluid [17
], or BAL [18
], were noted in ARDS patients. The present study was limited to the first four hours of observation, while these clinical studies were performed within the first couple of days after ARDS developed. It is known that C57BL6 mice are very susceptible to lung hyperoxic stress [41
]. These confounding factors may explain the differences between our observation and those of others. Further investigation is required to address these questions.
Despite the increased levels of VEGF in the BAL, a decreased expression of VEGF in the lung tissue, as revealed by the IHC staining, was observed specifically in the IIR group. Factors other than hyperoxia, such as IIR-induced acute inflammatory response, should be responsible for this drop in VEGF. Down-regulation of VEGF has been observed in the rat lungs after four hours of lipopolysaccharide challenge [18
]. A down-regulation of VEGF, as well as VEGF receptors, was also found at 24 hours and 72 hours after lipopolysaccharide injection in the mouse lungs [42
Cell death is a common feature of ALI and ARDS, contributing to the dysfunction of the alveolar-capillary barrier [6
]. The role of VEGF as a survival factor for endothelial cells is already well established [43
]. A correlation between the reduced VEGF levels and endothelial cell death has been found in the lungs of ARDS patients [45
]. The function of VEGF in epithelial cells, however, is largely unknown. Recent evidence suggests that VEGF could also be a survival factor for epithelial cells. VEGF stimulated growth of fetal airway epithelial cells [46
] and the proliferation of renal epithelial cells [47
]. In a rat model of obliterative bronchiolitis, Krebs and colleagues [48
] observed that VEGF either directly promoted epithelial regeneration or inhibited epithelial cell death. Tang and co-workers [49
] observed that a transient ablation of the gene encoding VEGF in the lung was associated with an increased number of TUNEL-positive cells in the alveolar walls. In the present study, we found a negative correlation between the number of VEGF-positive cells and TUNEL-positive epithelial cells. To further determine the role of VEGF in lung epithelial cell survival, we used siRNA to knock-down VEGF expression in A549 cells. This technique has been successfully used to effectively and specifically reduce the expression of other signal transduction proteins in lung epithelial A549 cells [30
] and other cell types [29
]. Our data show that the cell viability was significantly reduced by siRNA for VEGF. Therefore, VEGF could be a survival factor for alveolar epithelial cells. On the other hand, these cells are one of the main sources of VEGF in the lung [11
]. Thus, the death of alveolar epithelial cells could be partially responsible for the decreased expression of VEGF [50
VEGFR-1 is normally expressed on epithelial and endothelial cells in the lung [23
]. Compared with VEGFR-2, the function of VEGFR-1 in the lung is less determined. In the present study, IHC showed a significant decrease in the expression of VEGFR-1 in both the sham and the IIR groups, suggesting that hyperoxia and/or MV may suppress its expression. The decreased expression level of VEGFR-1 was more significant in the IIR group (p
< 0.01). The significant negative correlation between VEGFR-1 positive cells and epithelial cell death suggests that down-regulation of VEGFR-1 may be due to epithelial cell death. However, when we knocked down the VEGFR-1 protein level with siRNA, the viability A549 of cells was significantly reduced. This observation is intriguing. It suggests that VEGF and VEGFR-1 may modulate survival of lung epithelial cells in an autocrine fashion.
VEGFR-2, mainly located on endothelial cells, but also on epithelial type II cells [23
], is responsible for most of the known functions of VEGF in the lung [12
] and, in particular, is involved in the anti-apoptotic properties of VEGF on endothelial cells [43
]. In the control and sham-operated animals, the strong staining of VEGFR-2 was found mainly on cells located at the corners of the alveolar space with the appearance of type II pneumocytes and alveolar macrophages. In the IIR group, the number of VEGFR-2 positive cells did not change significantly. The morphological features of cells that expressed VEGFR-2, however, suggest that some of them could be the interstitial monocytes. The accumulation of VEGFR-2-positive inflammatory cells was also noted in a model of Bleomycin-induced lung injury [23
]. Moreover, in the IIR group, the expression of VEGFR-2 was localized mainly in the cytoplasm, which was not noted in the other two groups. This phenomenon indicates that the function of VEGFR-2 may be altered by the inflammatory responses in the lung tissue. Further investigation with double immunostaining and confocal microscopy may provide more convincing evidence.
When using western blotting and real-time quantitative RT-PCR to measure the protein and mRNA levels of VEGF and its receptors, the results are not statistically significant. The expression and distribution of these molecules are scattered in the lung tissue. When total tissue lysates were used as samples for immunoblotting, or as sources for total RNA extraction, the background from surrounding cells and tissues may mask the changes of VEGF, or its receptors.