mRAGE and sRAGE expression in normal lung development and in response to hyperoxia were examined. Analysis of the developmental regulation of mRAGE and sRAGE in the lung demonstrates that both neonatal rat mRAGE and sRAGE expressions are lower than found in adult lungs and there is an up-regulation of RAGE isoforms that occurs post-natally. In addition, extended hyperoxic exposures result in a loss of sRAGE in the neonatal rat lung.
Several factors may account for the developmental up-regulation of RAGE in the lung. mRAGE has been identified as a marker of type I epithelial cells [11
], and the newborn rat lung is not fully alveolarized. Therefore, the post-natal increase in mRAGE in the neonatal rat lung may reflect ongoing alveolarization characterized by an increase in type I epithelial cells. Furthermore, RAGE expression may be sensitive to oxygen tension [18
]. The increase in RAGE isoforms seen in the lung after birth may therefore also be due, in part, to changes in oxygen tension occurring at birth. While immunochemical examination of RAGE can not distinguish between mRAGE and sRAGE due to the inability of the antibody to distinguish these two isoforms, these immunohistochemical studies confirm that the overall level of RAGE is up-regulated during development (Fig. ). This finding suggests that the increase in RAGE expression does indeed correlate with alveolarization and the increase in type I cells.
A closer look at the soluble and membrane preparations reveals a molecular weight variance between neonatal and adult sRAGE. Interestingly term sRAGE contains a band whose molecular weight does not match that of either the adult sRAGE or neonatal mRAGE. We found that the increase in molecular weight of the neonatal sRAGE is likely due to different glycosylation patterns (see below).
To further assess the expression of mRAGE and sRAGE in the developing lung, additional time-points were explored. Fetal rat lungs express a single RAGE isoform whose molecular weight appears to correspond to the alternatively glycosylated mRAGE expressed in the neonatal lung. This corresponds to its expression in differentiating alveolar epithelial cells as reported previously by Shirasawa et al
] and Katsuoka et al
]. Term, four day and eight-day rat lungs demonstrated multiple RAGE bands, which can correspond to different isoforms or glycosylation patterns. It has been elucidated by Giron et al
] that multiple bands due to alternative splicing and glycosylation patterns have been seen in the rat. There are two possible N-glycosylation sites of RAGE. The two sRAGE and mRAGE bands seen after N-glycosydase F treatment reflect a partial deglycosylation (one sugar removed) and complete deglycosylation (both removed). Notably, when neonatal homogenates were deglycosylated, their molecular weights aligned with that of the adults to confirm that alternative glycosylation of the mRAGE and sRAGE is likely responsible for the different sizes of the RAGE bands seen between the early time-points and older rats. Adult rat RAGE expression consists of two distinct bands; the 50 kDa band of mRAGE and the 45 kDa band of sRAGE, consistent with observations in mice [6
]. We speculate that alternative glycosylation present in neonates is either absent or present at undetectable levels in the normal adult lung. The alternative glycosylation in the neonate may affect the affinity of the receptor for its ligands. Previous studies looking at amphoterin binding revealed a decrease in radio labeled ligand-receptor binding with deglycosylation [14
]. Further work is required to evaluate the whether alternative glycosylation in the neonate affects ligand binding.
Compared to adults, fetal and term lungs contain reduced levels of the sRAGE isoform, which may lead to an increased susceptibility to pathologic insults in the lung. The fact that less sRAGE is present at this stage in development to scavenge detrimental ligands may contribute to the pathologic effects of mRAGE activation, which ultimately leads to NF-kB activation [3
]. Under hyperoxic conditions it is known that NF-kB is activated in the neonate but not in the adult [20
]. The fact that the mRAGE/sRAGE balance is skewed in neonatal lungs may contribute to this difference in response between adult and neonatal lungs. Alternatively, it has recently been shown that mRAGE may be important in spreading of epithelial cells and that sRAGE can inhibit this epithelial spreading [21
]. Thus, the absence of sRAGE at this early time point may promote epithelial spreading of type I cells during alveolar development.
In addition to examining RAGE expression during normal lung development, this study also characterized the effects of hyperoxia on mRAGE and sRAGE expression. One main component of hyperoxic pulmonary injury is inflammation, which can be devastating to the developing lung. Because mRAGE signaling can be highly pro-inflammatory [1
] and mRAGE expression is up-regulated in response to inflammation in other tissues, RAGE isoform expression was examined in a hyperoxic setting. Recent work has shown an increase in lung RAGE abundance in pulmonary inflammation caused by smoke-related damage, and various pneumonias; an increase in its ligands has also been shown [7
], suggesting a role for RAGE in a novel inflammatory pathway in the lung. However, our results do not show an up-regulation of mRAGE (as it appears unchanged even after chronic hyperoxia), but rather a loss of sRAGE in response to hyperoxic injury. The loss of sRAGE may be due to dilution as a result of edema. Notably, when bronchoalveolar lavage fluid (BALF) was analyzed in an adult model of hyperoxia, sRAGE was detectable by Western blot (not shown), while in normal conditions sRAGE was undetectable in the BALF [10
]. Histological examination would be difficult to establish a difference between control and test animals after O2
treatment as there is no change in mRAGE and a loss of sRAGE; RAGE antibody detects the variable N-terminal chain of the receptor making it isoform non-specific. However, the Western blot data shows a significant loss of sRAGE, indicating that after hyperoxia, an imbalance in the ratio of sRAGE/mRAGE occurs that may contribute to the pulmonary inflammation seen in a neonatal model of BPD.
The expression of mRAGE in the lung before and after parturition and its subsequent elevated expression in the adult lung, relative to other tissues, support the hypothesis that this protein may play a homeostatic physiological role in this tissue. These experiments have further shown that sRAGE appears to be absent before birth which may help promote type I cell spreading during alveolar development. In addition, sRAGE is found to decrease in response to hyperoxic injury which may promote mRAGE signaling and contribute to further injury in this pathologic condition.