cAMP Response Element Binding Protein1 Is Essential for Activation of Steroyl Co-Enzyme A Desaturase 1 (Scd1) in Mouse Lung Type II Epithelial Cells

Cyclic AMP Response Element-Binding Protein 1 (Creb1) is a transcription factor that mediates cyclic adenosine 3′, 5′-monophosphate (cAMP) signalling in many tissues. Creb1−/− mice die at birth due to respiratory failure and previous genome-wide microarray analysis of E17.5 Creb1−/− fetal mouse lung identified important Creb1-regulated gene targets during lung development. The lipogenic enzymes stearoyl-CoA desaturase 1 (Scd1) and fatty acid synthase (Fasn) showed highly reduced gene expression in Creb1−/− lungs. We therefore hypothesized that Creb1 plays a crucial role in the transcriptional regulation of genes involved in pulmonary lipid biosynthetic pathways during lung development. In this study we confirmed that Scd1 and Fasn mRNA levels were down regulated in the E17.5 Creb1−/− mouse lung while the lipogenic-associated transcription factors SrebpF1, C/ebpα and Pparγ were increased. In vivo studies using germline (Creb1−/−) and lung epithelial-specific (Creb1EpiΔ/Δ) Creb1 knockout mice showed strongly reduced Scd1, but not Fasn gene expression and protein levels in lung epithelial cells. In vitro studies using mouse MLE-15 epithelial cells showed that forskolin-mediated activation of Creb1 increased both Scd1 gene expression and protein synthesis. Additionally, MLE15 cells transfected with a dominant-negative ACreb vector blocked forskolin-mediated stimulation of Scd1 gene expression. Lipid profiling in MLE15 cells showed that dominant-negative ACreb suppressed forskolin-induced desaturation of ether linked lipids to produce plasmalogens, as well as levels of phosphatidylethanolamine, ceramide and lysophosphatidylcholine. Taken together these results demonstrate that Creb1 is essential for the induction and maintenance of Scd1 in developing fetal mouse lung epithelial cells.


Introduction
Prematurity is a major complication for newborn infants and accounts for ,60% of the perinatal morbidity and mortality associated with birth. This is primarily due to an incomplete development of the lung which cannot adequately fulfil the demands of oxygenation for the body. As a result, premature infants often suffer varying degrees of respiratory distress syndrome (RDS) with the severity depending on the degree of lung immaturity. A key event of late lung development is the differentiation and maturation of the type II alveolar epithelial cell (AEC) in the distal lung, which primarily functions to synthesize and secrete surfactant into the airways. Lung surfactant is composed of approximately 90% phospholipids, and 10% surfactant associated proteins [1,2]. This complex mixture reduces the surface tension at the air-liquid interface after birth that prevents alveolar collapse and therefore allows normal lung function. To a large degree, the severity of RDS is closely associated with a profound lack of type-II AEC differentiation and deficiency of lung surfactant (reviewed in [3]). The genetic programs which drive pulmonary morphogenesis, and in particular stimulate epithelial cell differentiation and surfactant production in the lung are controlled by the actions of specific transcription factors, which regulate a complex array of gene expression networks.
Among the many transcription factors identified to have a vital role in the developing lung is the cyclic adenosine 39,59monophosphate (cAMP) response element binding protein (Creb1). Creb1 2/2 mice die shortly after birth due to respiratory distress and show delayed differentiation of both proximal and distal airway epithelial cell populations of the lung [4,5]. Creb1 is a member of the Creb/Atf subfamily of cAMP responsive basic region-leucine zipper (bZIP) transcription factors. The transcriptional activities of Creb1 are primarily activated by phosphorylation at the Serine 133 (Ser133) residue in response to an increase in intracellular cAMP levels. Several hormones, growth factors and cytokines have been shown to induce Ser133 phosphorylation of Creb1 via cAMP stimulation, and activate Creb1 that is normally bound as a dimer to cAMP response elements (CRE) within the promoter regions of target genes. [6,7]. Other members of the family include activating transcription factor 1 (Atf1) and the cAMP response element modulatory protein (Crem), both of which can also heterodimerize with Creb1, and potentially provide an additional degree of diversity in gene regulation [8].
In this study, we have further investigated the potential Creb1mediated regulation of gene targets from our microarray list which may be important for type II AEC lipid biosynthesis, an essential process required for type-II AEC surfactant production. In particular we have examined Creb1-mediated regulation of the key rate limiting lipogenic enzymes; fatty acid synthase (Fasn) and steroyl-CoA desaturase 1 (Scd1), both of which previously showed highly reduced mRNA levels in Creb1 2/2 fetal lungs [4]. The Figure 1. Analysis of mRNA levels by qPCR for Scd1 and Fasn, and lipogenic transcription factor genes during mouse lung development. qPCR confirmed changes in Scd1 and Fasn mRNA levels in the lung of E17.5 Creb1 2/2 fetal mice compared to littermate controls (A). Analysis of mRNA levels for the transcription factors SrebpF1, C/ebpa and Ppparc in the lung of E17.5 Creb1 2/2 fetal mice compared to littermate controls (B). Ontogeny analysis of gene expression in normal fetal mouse lung for Scd1 (1C), Fasn (1D), SrebpF1 (1E) C/ebpa (1F) and Ppparc (1G). All values for qPCR analysis were normalized to levels of 18SrRNA (18S). For Figs. 1C-1G the relative expression was compared with that at E14.5, which is given a value of 1. In the bar graphs, open bar -wildtype (+/+), shaded bars -E17.5 Creb1 2/2 (2/2) lung samples. Data is presented as the mean 6 SEM, n = 4; where 'a' represents p-value ,0.05. doi:10.1371/journal.pone.0059763.g001 Creb1 Activates Scd1 in the Distal Mouse Lung PLOS ONE | www.plosone.org cytosolic Fas enzyme is a multifunctional homodimeric complex which promotes de-novo synthesis of saturated fatty acids [9,10], while Scd1 is an endoplasmic reticulum-based transmembrane enzyme which catalyses the conversion of saturated to monounsaturated fatty acids, which then serve as substrates for synthesis of phospholipids, triacylglycerols (TAGs) and cholesteryl esters (CEs) [9,10,11]. Phospholipid (PC) in particular is an essential component of lung surfactant and accounts for approximately 90% of endogenous surfactant material [12]. Transcriptional and posttranscriptional regulation of both Scd1 and Fasn has been studied extensively in the context of obesity and cancer development in tissues with known roles in lipogenesis such as liver and adipose tissue [11]. However little is known about the regulatory mechanisms for these factors during lung development and their potential role in fetal surfactant biosynthesis.
Therefore, in this study we hypothesized that Creb1 positively regulates gene expression of factors which may be required for type II AEC lipid biosynthesis, in particular Fasn and Scd1. We examined gene and protein expression of Fasn and Scd1 using both in vivo and in vitro models where Creb1 function is either lost or inhibited, and show that in the case of Scd1, Creb1 is essential for normal expression in the lung. We also describe the ontogeny of gene expression for Fasn and Scd1 during late respiratory development, as well as the protein localisation of these factors to epithelial cell subsets within the fetal lung. Finally, we describe the specific lipid classes which show altered levels with the inhibition of Creb1 in a mouse lung type-II AEC-like cell line. Together our results strongly suggest that Creb1 is required for normal expression of Scd1 during lung development.

Immunohistochemistry and immunofluorescence
Mice lung samples were either fixed in optimal cutting temperature (OCT) compound or in 4% paraformaldehyde/ phosphate buffered saline (PBS) solution overnight at 4uC, processed, embedded in paraffin and 5 mm thick sections cut and prepared. For immunohistochemistry antigen retrieval was performed by microwaving in 10 mM sodium citrate buffer for 20 mins followed by blocking of endogenous peroxidases using 3% H 2 O 2 in methanol for 5 mins. Sections were then blocked in 5% goat serum for 1 hr and incubated with primary antibodies: Scd1 (#2794, Cell Signalling, Danvers, MA), Fas (#3180, Cell Signalling), ProSPC (ab3786, Chemicon, Temecula, CA) and Creb1 (#48H2, Cell Signalling) overnight at 4uC or at room temperature for 1hr. Sections were then incubated with a biotinylated goat anti-rabbit secondary antibody (Vector laboratories, Burlingame, CA). Positive staining was either detected using LSABH2 Streptavidin-HRP (Dako, Glostrup, Denmark) and diaminobenzidine (DAB) (Dako) solutions as per manufacturer's protocol for bright field imaging, or using fluorescent-labelled secondary antibodies and/or streptavidin proteins (Life Technologies, Carlsbad, CA). Finally, sections were counterstained with nuclear stains, haematoxylin for bright field imaging and hoechst (Sigma-Aldrich, St. Louis, MO) for fluorescent imaging. Negative controls included sections treated with block buffer and secondary antibody alone.

Transient transfection of MLE15 Cells
MLE-15 cells were transfected using Lipofectamine 2000 (Life Technologies, USA) according to the manufacturer's protocol. Cells were transfected with constructs, an empty pRc/CMV500, a constitutively-expressing mouse Creba, or a dominant-negative isoform of mouse Creb1 (ACreb). Cells were seeded at a density of 0.3610 6 cells per well in a 6 well dish in regular growth medium overnight. After 24 hrs, the medium was replaced with 2 ml of pre-warmed Opti-MEMI (Life Technologies). For transfection, the DNA-lipofectamine complex was prepared and added to the cells using the following DNA/Lipofectamine ratios: 5:1 for pRc/ CMV500 vector and ACreb constructs, 2:1 for the Creba construct. Cell cultures were incubated at 37uC, 5% CO 2 for 9 hrs, then the medium was replaced with MLE-15 culture medium supplemented with 2% charcoal/dextran-treated FBS. Fifteen hours later, cells were treated either with DMSO as vehicle or forskolin (10 mm) for 8 hrs in culture.

Statistical Analysis
Statistical analyses were performed using GraphPad Prism software. Comparisons between two groups were made by a Student's t test. Comparisons between multiple groups were analysed by 1-way ANOVA with post hoc Tukey's test for pair wise comparisons between groups. A p-value of less than 0.05 was considered to be statistically significant. For lipid profiling studies statistical analysis were performed using MatLab R2011b analytical software. Comparisons between two groups were made by Student's t test and corrected, where necessary for multiple comparisons by the Benjamini-Hochberg method [17].

Loss of Creb1 decreases Scd1 and Fasn gene expression in the fetal mouse lung
Microarray analysis of the E17.5 Creb1 2/2 mouse lung and littermate controls identified two lipogenic gene targets: Scd1 (7.8 fold, p,0.05) and Fasn (1.9 fold, p,0.05) to be significantly downregulated in Creb1 2/2 fetal mice (Table 1), that was subsequently verified by qPCR ( Figure 1A). Interestingly the mRNA levels of other transcription factors that regulate pulmonary lipogenesis, such as SrebpF1, C/ebpa and Pparc were also analysed and found to be moderately elevated in total RNA from the Creb1 2/2 E17.5 lung ( Figure 1B). Gene expression levels of these lipogenic genes and transcription factors were found to be developmentally regulated during normal lung development, with very low expression before E16.5, but thereafter there is a marked increase in mRNA levels until after birth ( Figure 1C, D, E, F, G).
The Creb1 2/2 fetal mouse lung shows developmental delay in the induction of Scd1 expression Immunohistochemistry of the mouse lung from E16.5 to E18.5 showed a dramatic delay in Scd1 (Figure 2A-F) protein levels in the Creb1 2/2 fetal mouse lung compared to wildtype controls, however we did not observe any alteration in levels of Fasn protein ( Figure 3A-F). In controls, Scd1 protein was first detected at E17.5 ( Figure 2B), concurrent with the normal differentiation of distal epithelial progenitors into AECs during the cannalicular to saccular stage of mouse lung development. Scd1 protein was restricted in expression to the cytoplasm of type-II AECs as shown by co-localisation with the type-II AEC marker ProSPC ( Figure 2G-I). Conversely, Fas protein was detected as early as E15.5 in the lung (data not shown) and was expressed in the cytoplasm of all distal tubular epithelial cells ( Figure 3A). Upon differentiation of AECs at E17.5, Fas expression similarly became restricted to type-II AECs as shown by co-localisation with ProSPC at E18.5 ( Figure 3G-I), although faint expression was also observed in regions of proximal epithelium ( Figure 3B & C). In the E17.5 Creb1 2/2 fetal lung, Scd1 protein was only detected in a few isolated cells, with a slight increase in levels at E18.5, indicative of a delay in the induction of expression ( Figure 2E & F). Pulmonary Fas expression however was always comparable in staining intensity between wildtype and Creb1 2/2 fetal mice.
Loss of Creb1 in mice is known to cause upregulation of the related bZIP family member Crem in several tissues, including the fetal lung, as a mechanism to potentially compensate for loss of Creb1 function [4,18]. To address the potential of Crem to compensate for the loss of Creb1 in the lung, Scd1 and Fas protein levels were examined by immunohistochemistry in the E17.5 fetal Creb 2/2 mouse lung on a homozygous Crem 2/2 or heterozygous Crem +/2 genetic background. Using immunohistochemistry Scd1 protein levels in Creb1 +/+ fetal lungs on a Crem 2/2 background were similar to E17.5 wildtype fetal mice ( Figure 4A, and compared with Figure 2B). In addition Analysis of Scd1 mRNA levels in total lung RNA from E17.5 showed no statistical difference between Crem wildtype and Crem 2/2 fetal mice ( Figure S1). Loss of Creb1 on a Crem +/2 or Crem 2/2 genetic background showed almost no detection of Scd1 protein identical to an E17.5 Creb1 2/2 fetal lung ( Figure 4B & C and compare with Figure 2E). Fas protein levels on all Crem genetic backgrounds were also similar to the E17.5 Creb1 2/2 lung ( Figure 4D-F, and compared with Figure 3B & E).
Pulmonary Scd1 expression is severely reduced in lung epithelial cell-specific Creb1 EpiD/D fetal mice Systemic loss of Creb1 in mice produces a wide range of physiological defects including myogenic dysfunction [5], disrupted T cell differentiation, impaired brain development [19] and reduced body size [4,19], which may indirectly influence lung development and thus confound analysis of the Creb1 2/2 respiratory phenotype. To evaluate if loss of Creb1 specifically in lung epithelial cells would mimic our finding of disrupted Scd1 expression in the Creb1 2/2 fetal lung. Mice with a lung epithelial cell-specific Creb1 knockout (Creb1 EpiD/D ) were generated using a doxycycline-inducible Cre/LoxP approach whereby Cre recombinase was expressed under the conditional control of the human SFTPC promoter [20]. Using immunohistochemistry, we first tested for the presence or absence of Creb1 protein in the lung epithelial cells. E17.5 control mice showed almost ubiquitous  Figure 2. Immunohistochemical staining for Scd1 in the mouse lung from E16.5 to E18.5 in Creb1 2/2 mice (2/2) and wildtype litter mate controls (+/+). Staining for Scd1 was detected by E17.5 (B) in type II alveolar epithelial cells and increased at E18.5 (2A-C). In lung sections from Creb1 2/2 mice Scd1 was absent at E16.5 and E17.5 with very weak staining detected at E18.5 (2D-F). Scd1 staining was restricted to mature type II alveolar epithelial cells in E18.5 fetal mouse lungs and was predominantly co-localized with the type II epithelial cell marker, pro-surfactant protein C (ProSP-C) (2G-I). All bright field micrographs are taken at the magnification of 206 and fluorescent images at a magnification of 406  (Fig. 5B, C). In E17.5 Creb1 EpiD/D mice Creb1 expression was almost completely absent in the lung epithelial cells (Fig. 5D). Using qPCR, we detected a statistically significant decrease of both Scd1 and Fasn mRNA levels in the lung of E17.5 Creb1 EpiD/D fetal mice relative to controls ( Figure 5A). Furthermore, Scd1 immunostaining was completely absent in lung sections from E17.5 Creb1 EpiD/D transgenic fetal mice ( Figure 5G). Consistent with our previous findings, Fas

Scd1 expression is induced by forskolin in cultured mouse lung epithelial MLE-15 cells
The induction of Scd1 and Fasn was next investigated in vitro using cultured mouse lung epithelial MLE15 cells treated with forskolin to induce levels of intracellular cAMP. MLE15 cells showed increased levels of phosphorylated Creb1 protein after 1 hr, which returned to basal levels 3 hrs after forskolin treatment (data not shown). Creb1-mediated induction of Scd1 levels was examined following treatment with 10 mM forskolin. Short-term treatment did not significantly induce Scd1 mRNA levels ( Figure 6A) and treatments of 4, 8 and 12 hrs ( Figure 6B) were required to achieve statistically significant induction in these cells. Interestingly, treatments of 2 to 3 hrs forskolin increased Scd1 protein levels that were sustained with 4, 8 and 12 hrs of forskolin treatment in culture ( Figure 6C-D).

Expression of dominant-negative ACreb in MLE-15 cells inhibits forskolin-mediated induction of Scd1 gene expression
To determine if Creb1 is required for the forskolin-induced increase in Scd1 mRNA and protein levels observed previously and not via other cAMP regulated pathways, cultured MLE15 lung epithelial cells were transiently transfected with either an empty pRc/CMV500 plasmid or a plasmid that expresses a dominantnegative Creb1 isoform, ACreb. ACreb prevents Creb1-DNA binding via homodimerization with the wildtype Creb1 leucine zipper domain [21]. To increase the sensitivity of MLE15 cells to Creb1-mediated signalling, cells were also transiently transfected with a construct which overexpresses mouse Creba, a Creb1 splice variant which encodes all important functional domains and is one of the most abundant Creb1 isoforms [22]. MLE-15 cells transfected with Creba or the empty plasmid alone showed a 2.1 and 2.2 fold induction in Scd1 mRNA levels, respectively (p,0.05) following 10 mM forskolin treatment after 8 hrs in culture ( Figure 7A). In cells co-transfected with Creba and the dominantnegative ACreb, the statistically significant forskolin induction of Scd1 mRNA levels was abolished ( Figure 7A). In contrast, Fasn mRNA levels were not induced by forskolin treatment in MLE15 cells transfected with either the empty or Creba plasmid alone but interestingly were increased to a statistically significant level upon co-transfection with the ACreb plasmid (p,0.05) ( Figure 7B).

Loss of Creb1 activity reduced ceramide, lysophosphatidylcholine, phosphatidylethanolamine and relative plasmalogen levels in MLE-15 cells
To examine the role that Creb1 plays in regulating lipogenesis and surfactant biogenesis in lung alveolar type II cells, a whole cell lipidomic analysis was performed on MLE15 cells following transient transfection with plasmids expressing the Creba or dominant-negative ACreb1 isoforms. MLE-15 cells were transfected as described above and treated with 10 mM forskolin for 24 hrs. Whole cell lipids were extracted from treated cells and quantitated for lipid components by mass spectrometry. Inhibition of Creb1 by transfection of the ACreb isoform in MLE15 cells abolished the forskolin-induced increase in ceramide levels ( Figure 8A). Forskolin-induced increases in lysophosphatidylcholine and phosphatidylethanolamine levels were also abolished in the ACreb transfected cells (Figure 8B and C). ACreb did not abolish an increase in alkylphosphatidylcholine levels, but did reduce the forskolin-induced increase in alkenylphospatidylcholine levels (data not shown). Furthermore, the relative concentrations of alkenylphospatidylcholine/alkylphosphatidylcholine in cells trans- fected with the ACreb plasmid was decreased ( Figure 8D). There were no changes in any other lipid species measured (data not shown). The forskolin-induced increase in the levels of lipid classes shown here were all significant before correction for multiple comparisons.

Discussion
In this study we have demonstrated that the cAMP signalling pathway via Creb1 has an important function in pulmonary lipogenic pathways via regulation of specific lipogenic-associated genes expressed in the lung epithelial cells. Both microarray and qPCR analyses showed that loss of Creb1 leads to markedly reduced gene expression of the lipogenic enzymes Scd1and Fasn at E17.5 indicating that these factors may be upregulated by Creb1 in the developing lung late in gestation. We further confirmed that Fas and Scd1 proteins are localised to type-II AECs during the canalicular and saccular stages of lung development. AECs first differentiate from distal epithelial progenitors, suggesting that Fas and Scd1 have an important role within type-II AECs close to the time of birth. Given that Ser133 phosphorylation of Creb1 occurs primarily in distal lung epithelial cells, our finding of Creb1mediated regulation of epithelial-expressed factors such as Scd1 and Fas is not surprising. Furthermore, the rapid induction of these genes between E16.5 to E18.5 occurs concurrently with the onset of the respiratory phenotype in Creb1 2/2 lungs, suggesting a temporal link between Creb1 function and the induction of Scd1 and Fasn gene expression [4]. Similar timing of Scd1and Fasn gene induction has also been noted in fetal rat lungs [23].
Our imunohistochemical data did not show any noticeable requirement for Creb1 in regulating Fasn expression in the fetal lung. We suggest this may be due to a lack of sensitivity inherent with immunohistochemical staining which would not detect the relatively low magnitude of change in Fasn mRNA levels observed in E17.5 Creb1 2/2 lungs and the ontogeny analysis. Similarly, Fasn mRNA levels rise only minimally between E19 and E21 in the fetal rat lung [23]. However, Creb1 has been shown to bind putative CREs found in the mouse Fasn promoter and stimulate transcription of a linked luciferase reporter gene in 3T3-L1 preadipocytes [24]. Nevertheless, our results suggest that Creb1 may play only a minor role in promoting Fasn expression during late lung development. We also find Fasn levels are similar at E15.5, well before the observed Creb1 2/2 lung phenotype. It is also likely that levels of Fasn are tightly regulated by a number of transcriptional pathways, as loss of even a single Fasn allele in mice leads to a severe reduction in viability during embryonic development [10].
In contrast, Creb1 function is clearly essential for Scd1 gene expression and protein levels in the developing lung in vivo. Additionally our in vitro studies using MLE15 cells confirm that cAMP-mediated stimulation of Scd1 expression requires Creb1. Arguably the bulk of literature associated with analysis of Scd1 gene regulation has been conducted in mouse 3T3-L1 preadipocytes. In these cells cAMP analogues have been shown to stimulate Scd1 mRNA levels [25], while knockdown of Creb1 function using siRNA prevents the stimulation of Scd1 mRNA levels caused by Antimycin A-mediated inhibition of mitochondrial activity [26]. However, in our studies Scd1 mRNA was only induced to a statistically significant level after long-term forskolin exposure, indicating that Creb1-mediated regulation of Scd1 gene expression is possibly indirect and requires other intermediary factors. In agreement with our findings, maximal cAMP-mediated induction of Scd1 mRNA in mouse 3T3-L1 preadipocytes was also shown to require new protein synthesis [25]. Although a CRE has been found in the mouse Scd1 promoter [25,27], it is uncertain whether this has any biological relevance. Interestingly however we find that Scd1 protein synthesis in MLE15 cells was induced after only 2 hours of forskolin exposure, raising the possibility that Scd1 mRNA can be stored and translated by a Creb1-dependent mechanism to enable rapid protein synthesis. While most studies find Scd1 expression is regulated primarily by gene transcription, post-transcriptional mechanisms have also been described [11,28]. Future studies are therefore necessary to determine whether Scd1 mRNA storage and rapid protein synthesis is a functional mechanism within type-II AECs in the developing lung in vivo.
Although it is clear that Scd1 is highly expressed in type-II AECs during lung development, its precise lipogenic role remains unclear. It is unlikely that Scd1 plays a crucial role in lung surfactant function given that Scd1 2/2 mice do not appear to exhibit any respiratory phenotype [29]. However it is also possible that downregulation of Scd1 in the lung could be compensated for by other Scd isoforms, of which four exist in the mouse genome [11]. Alternatively, Scd1 may be more involved in non-surfactant lipid biosynthesis.
The lipid profiling in MLE15 cells indicate that blockade of Creb1 activity in lung epithelial cells leads to reduced levels of ceramide, phosphatidylethanolamine, lysophosphatidylcholine and a decreased ratio of alkenyphosphatidylcholine to alkylphosphatidylcholine. Creb1 has been shown to induce transcription of the ceramidase gene, ASAH1, which is involved in the catabolism of ceramides in H295R adrenocortical cells [30]. The loss of Scd1 in skeletal muscles of ob/ob mice is also reported to result from reduced ceramide synthesis [31,32]. However, in the human lung epithelial A549 cell line, the reduction of Scd1 leads to reduced cell proliferation with high rate of cell death that is independent of ceramide synthesis [33].
Surfactant lipids such as lysophosphatidyl choline and phosphatidylethanolamine are present in trace amounts in the adult mouse lung and are thought to influence surfactant metabolism [34]. Creb1-mediated induction of Scd1 enzyme in the lung may be an intracellular mechanism to process any excess saturated fatty acids during lung epithelial differentiation late in development. During this stage there is a dramatic increase in de novo lipogenesis for surfactant biosynthesis and secretion. The increase in saturated fatty acids may also relate to the relative decrease in lysophosphatidylcholine. The polyunsaturated fatty acid species typically in the sn2 position of phosphatidylcholine are more prone to phospholipase action than saturated species, thus a decrease in the level of saturation may influence the production of lysophosphatidylcholine. Phosphatidylethanolamine is also biosynthetically related to phosphatidylcholine as a precursor which can undergo multiple methylation steps via the action of phosphatidylethanolamine N-methyltransferase to produce phosphatidylcholine. These results suggest that there may be a decrease in phospholipase A2 activity that would degrade both cell membranes and surfactant to increase levels of these lipid species [35]. High levels of these lipids are known to increase the sensitivity to protein inhibition that decreases surfactant bioactivity and is observed in respiratory distress syndromes and lung injury [34,35,36].
We did not observe any change in the levels of storage lipids such as cholesterol esters, cholesterol and triglycerides with inhibition of Creb1. Hence it is likely that the developing respiratory epithelial cells are preferentially glycolytic in nature rather than lipogenic. Gene deletion of SREBP cleavage-activating protein gene, Scap in mice have shown compensatory upregulation of triglyceride synthesis and the accumulation of resident lipofibroblasts with the loss of lipogenesis in lung epithelial cells [37]. This also enhances type II AEC's ability to synthesize and traffic triglycerides for incorporation into disaturated phosphatidylcholine in surfactant lipids during late gestation [38]. This increase in triglyceride production by lipofibroblasts is thought to be primarily by an increase of cAMP levels that makes these cells more responsive to stimulation by other hormones such as the glucocorticoids and is driven by hormones such as prostaglandin PGE 2 late in gestation [38]. The findings of this study suggest that Creb1-mediated cAMP signalling plays a role in the maintenance of surfactant homeostasis and bioactivity via regulation of lipid classes in mesenchymal lung cell populations such as lipofibroblasts. In summary, both in vivo and in vitro studies presented here demonstrate that Creb1 is an essential transcription factor regulating the induction and levels of Scd1 in type II alveolar epithelial cells during lung development. Figure S1 Analysis of Scd1 mRNA levels in the Crem 2/2 and wild type fetal mouse lung at E17.5 by qPCR. All values were normalized to levels of 18S rRNA and relative expression was compared with that of Crem wildtype (+/+) which was given a value of 1. In the bar graph, open bar -wildtype (+/+), shaded bar -E17.5 Crem1 2/2 (2/2). Data is presented as the mean 6 SEM, n = 4. p = 0.683 and was not considered significant. (TIF)