This study shows that apical culture medium overlying an ATII cell monolayer can be a significant barrier to the rate of oxygen diffusion, and that the differentiation of adult rat ATII cells can be regulated by oxygen tension through HIF signaling in vitro.
The evidence to support this conclusion involves the increase in HIF1α and HIF2α from nuclear extracts and the increases in expression of the HIF-inducible genes VEGF, GLUT1, and PDK1, under the submerged condition (), indicating that cells cultured under submerged conditions are hypoxic. Many previous studies showed that under hypoxic conditions, HIF translocates from the cytoplasm to the nucleus and induces HIF-inducible genes as an adaptive response to match oxygen supply with metabolic and redox demands (17
). Moreover, Mamchaoui and Saumon showed that oxygen tension in the medium above adult rat ATII cells falls under submerged conditions (6
). Moreover, rocking, which intermittently exposes cells to air and mixes the medium to prevent the development of a stagnant layer and an oxygen gradient, partly restored the expression of surfactant proteins under the submerged condition (). Furthermore, supplemental oxygen partly corrected the expression of surfactant proteins under the submerged condition (). Previously, Acarregui and colleagues found that when human fetal lung tissue maintained for 5 days in 1% oxygen was transferred to an environment containing 20% oxygen, rapid morphological development and an induction of SP-A gene expression occurred (7
). We also found that the expression of surfactant proteins decreased when cells were cultured with DMOG, a prolyl hydroxylase inhibitor and HIFα stabilizer, under A/L interface conditions (). Under normoxic conditions, HIFα subunits are hydroxylated at conserved proline residues by prolyl hydroxylase domain–containing enzymes, whose activities are regulated by oxygen availability. When HIFα is hydroxylated in this manner, it is recognized and marked for proteosomal destruction by ubiquinization and the von Hippel–Lindau protein complex. Prolyl hydroxylase inhibitors (PHIs) are commonly used to mimic hypoxia by stabilizing HIFα. The addition of DMOG mimics the effects of submersion in A/L interface cultures. Grek and colleagues also demonstrated similar effects on the expression of pro–SP-B and pro–SP-C with several PHIs, for example, DMOG, ethyl-3,4-dehydroxybenzoate, and L-mimosine, in MLE-15 cells (8
). We also considered the possibility that the effects of A/L interface cultures were attributable to an apical differentiation factor that was concentrated in the absence of added apical fluid. We discounted this hypothesis based on the rocking experiments and the results with supplemental oxygen in which differentiation was maintained but with no reduction of apical fluid. In addition, Dobbs and colleagues used frequent apical washes to test for the presence of a putative apical differentiation factor and were unable to find evidence for its existence (4
). Taken together, these findings support the concept that oxygen tension and HIF signaling can regulate the expression of surfactant proteins.
Several observations require additional comment. Differences are evident in the isoforms of SP-A between freshly isolated and cultured rat type II cells, as shown in and as reported previously (27
). We attribute this change in electromobility to the amount of glycosylation of SP-A. Rat SP-A contains asparagine-linked oligosaccharide units at the C-terminus that confer a triplet with a molecular mass of 26, 32, and 36 kD under reducing conditions (28
). The oligosaccharide moiety of SP-A is involved in the Ca++
-dependent aggregation of phospholipid vesicles, in connection with the formation of tubular myelin (29
). Freshly isolated rat ATII cells on Day 0 contain mainly the 32-kD glycosylated form of SP-A and nonglycosylated SP-A, whereas cultured type II cells contain both 32-kD and 36-kD glycosylated SP-A. As one possible explanation, the 36-kD isoform is rapidly secreted in vivo
, and some problem may occur with the SP-A secretory pathway in cultured type II cells, so that this form accumulates intracellularly. According to another possibility, secreted SP-A is stuck on the surface of the cells in vitro
. However, when we performed immunocytochemistry for SP-A, we did not observe excessive SP-A on the surface of cells (Figure E1).
As shown in , the increase in HIF1 and HIF2 was evident by 18 hours, whereas known target genes such as VEGF and GLUT1 required more than 3 days to demonstrate increased levels of mRNA. The reason for this delay is unknown. Under severe hypoxia, VEGF mRNA is up-regulated rapidly at 2–18 hours (18
). On the other hand, under mild hypoxia (e.g., 10% oxygen), mRNA levels of VEGF do not increase until 18–48 hours (30
). Therefore, the timing of up-regulation of target genes is likely related to the severity of hypoxia. HIF DNA-binding activity is also dependent on oxygen concentrations, with low or undetectable activity in normoxia and a maximum increase for very low oxygen concentrations (e.g., 0.5% oxygen) (21
). Under the submerged condition, the cells are in mild hypoxia (6
). Therefore, we speculate that the lag in response time between observing HIF in the nucleus and the mRNA expression of HIF target genes is primarily caused by HIF DNA-binding activity.
In addition, some observations were not consistent with our hypothesis. The loss of HIF1α early in murine lung development in utero
leads to pups that die within hours of birth, with symptoms similar to those of neonatal respiratory distress syndrome (16
). The lungs of these pups exhibited impaired alveolar epithelial differentiation and an almost complete loss of surfactant protein expression. Asikainen and colleagues found no difference in expression of surfactant proteins in lungs from preterm baboons treated with or without PHI (32
). These observations in fetal and preterm lungs are at variance with our results. This variance may have occurred because those studies used preterm lungs, whereas our study used adult alveolar epithelial cells. Thus the role of HIF may vary in early development compared with the adult state because of the combinatorial nature of the transcription factors necessary for targeted gene expression. Furthermore, the deletion of individual genes may not always provide a complete picture of the physiologic process.
Surprisingly, the expression of surfactant proteins was reversible, and the effect was rapid ( and ). The decrease in expression of surfactant proteins in the switch from the A/L interface condition to the submerged condition occurred within 8 hours. When we switched cultures from the submerged condition to the A/L interface condition, cells were able to restore their expression of surfactant proteins within 48 hours. Hence, their oxygen sensing appears to be quite acute. The observation that the phenotype of type II cells can be dramatically changed by culture conditions is well-established. Previously we and others showed that the ATII cell phenotype and the type I–like cell phenotype is reversible and dependent on soluble factors and the culture substratum or matrix on which cells are plated (24
). In this study, the differentiation of rat ATII cells in terms of the expression of surfactant proteins was dependent on the amount of apical fluid and was reversible.
This study contains several limitations. First, the precise mechanisms whereby oxygen improves and submersion inhibits surfactant protein gene expression are not known. However, there are several possible mechanisms. HIFs may bind to HRE on surfactant proteins and repress the expression of surfactant protein genes. HIFs can inhibit the expression of some genes, while enhancing the expression of others (35
). Potential HRE binding sites in rat surfactant protein regulatory elements were analyzed, using the open-access JASPER database (38
). The rat SP-A promoter was sequenced (39
). Analysis of the promoter for SP-A revealed three potential HRE sites within that region (positions −936 to −929, −1046 to −1039, and −1475 to −1468). Promoters for rat SP-B and rat SP-C have not been sequenced. However, a 2-kb region upstream from the SP-B and SP-C transcripts also revealed similar HRE sites (SP-B, positions −657 to −650, −898 to −891, −915 to −908, and −1825 to −1818; SP-C, positions −1575 to −1568). In addition, the expression of surfactant proteins may be suppressed through the down-regulation of CCAAT/enhancer binding protein α (C/EBPα) by HIF. C/EBPα is expressed in ATII cells, and the expression of C/EBPα coincides temporally with surfactant production during the terminal differentiation of ATII cells (40
). In addition, the expression of SP-A and SP-D increases during the differentiation of type II cells, and their gene promoters contain C/EBP binding sites that comprise important elements regulating their expression (41
). The loss of C/EBPα in the respiratory epithelium leads to an arrest in the type II alveolar cell differentiation program (43
). Seifeddine and colleagues showed that hypoxia down-regulates C/EBPα in breast cancer cells via several mechanisms, including transcriptional repression mediated by HIF-1α binding to an HRE in the C/EBPα promoter and posttranscriptional effects, such as reduced stability of C/EBPα mRNA and altered cellular distribution of C/EBPα protein (36
). Moreover, when HIF1α and C/EBPα interact, they may bring about reciprocal functional changes (44
). Interestingly, the protein levels of C/EBPα were down-regulated under the submerged condition in our system (data not shown). As a third possibility, HIF can induce some repressors of gene expression, as shown for certain lipogenic enzymes in the liver (45
Regarding additional limitations of this study, we did not measure the exact oxygen concentration at the surface of rat ATII cells cultured under the submerged condition because of technical limitations. However, that measurement was performed by Mamchaoui and Saumon (6
). We did not show data from cells cultured under severely reduced oxygen tensions. When we cultured rat ATII cells under A/L interface conditions with 1%, 5%, or 10% oxygen, the results were variable, and significant cytotoxicity occurred, limiting our ability to interpret results. In addition to HIF1α, we found a significant increase in HIF2α, a finding consistent with a previous report (46
). Most genes regulated by HIF1α are also regulated by HIF2α, and thus the individual contributions of the two HIFs are unclear. Importantly, in liver and hepatic cell lines, some important differences are evident between HIF1α and HIF2α. HIF2α regulates erythropoietin, plasminogen activator inhibitor–1, and some enzymes related to lipid metabolism in the liver (45
), whereas HIF1α is more responsible for glucose transport. Hence, additional studies will be required to define the precise role of HIF2α in the regulation of type II cell differentiation. This study was performed with rat type II cells and should be confirmed with human type II cells, because the volume density of type II cell mitochondria is directly related to lung oxygen consumption, and this varies according to species (48
). The mitochondria density in type II cells is higher in small rodents. Finally, the pathways whereby oxygen improves and submersion inhibits surfactant protein gene expression will require considerably more work to define, but may lead to opportunities to learn more about the regulation of type II cell function in the adult lung.
In conclusion, we showed that culture of adult rat type II cells at an A/L interface maintained the expression of surfactant proteins. On the other hand, submersion reduced their expression, but this effect was reversible. Under the submerged condition, the HIF-inducible genes GLUT1, VEGF, and PDK1 were up-regulated, and both HIF1α and HIF2α protein levels were increased in nuclear extracts. The HIF stabilizer, DMOG, also down-regulated the expression of surfactant proteins, whereas it up-regulated GLUT1 mRNA levels. Finally, supplemental oxygen partly corrected the impairment of the expression of surfactant proteins in rat ATII cells cultured under the submerged condition. These results indicate that the expression of surfactant proteins and cell differentiation in rat ATII cells in vitro are regulated by oxygen tension through HIF signaling. These findings suggest that focal alveolar hypoxia in a variety of clinical settings, such as acute lung injury, pneumonia, pulmonary edema, or severe small airway obstruction, may impair the production of surfactant.