The reduction in FOXA2 activity during lung development in fetuses from diabetic rat mothers is reversed by Akt inhibition

Hyperglycemia during pregnancy is associated with fetal lung development disorders and surfactant protein (SP) deficiency. Here, we examined the role of FOXA2 and Akt signaling in fetal lung development during diabetic pregnancy. Sprague‐Dawley rats were injected with streptozocin (STZ) during pregnancy to induce diabetes (DM). DM‐exposed fetal lungs exhibited reduced numbers of alveoli, irregularities in the appearance and thickness of the alveolar septum, increased levels of glycogen and lipids in type II alveolar epithelial cells, fewer microvilli and mature lamellar bodies, and swollen mitochondria. SP‐B and SP‐C in DM amniotic fluid and DM lungs were lower than in the control group (P < 0.05). DM lung nuclear FOXA2 was lower compared with the control group (P < 0.05), but p‐FOXA2 was higher (P < 0.05). In murine lung epithelial (MLE) 12 cells, p‐AKT levels were increased by high glucose/insulin, but decreased by the Akt inhibitor MK2206 (P < 0.05). Expression of nuclear FOXA2 was increased by MK2206 compared with the high glucose/insulin group (P < 0.05). These results suggest that maternal diabetes induces fetal lung FOXA2 phosphorylation through the Akt pathway, and also affects the maturation of alveolar epithelial cells and reduces levels of SP‐B and SP‐C in the fetal lungs. An Akt inhibitor reversed the changes in SP expression in vitro.

Hyperglycemia during pregnancy is associated with fetal lung development disorders and surfactant protein (SP) deficiency. Here, we examined the role of FOXA2 and Akt signaling in fetal lung development during diabetic pregnancy. Sprague-Dawley rats were injected with streptozocin (STZ) during pregnancy to induce diabetes (DM). DM-exposed fetal lungs exhibited reduced numbers of alveoli, irregularities in the appearance and thickness of the alveolar septum, increased levels of glycogen and lipids in type II alveolar epithelial cells, fewer microvilli and mature lamellar bodies, and swollen mitochondria. SP-B and SP-C in DM amniotic fluid and DM lungs were lower than in the control group (P < 0.05). DM lung nuclear FOXA2 was lower compared with the control group (P < 0.05), but p-FOXA2 was higher (P < 0.05). In murine lung epithelial (MLE) 12 cells, p-AKT levels were increased by high glucose/insulin, but decreased by the Akt inhibitor MK2206 (P < 0.05). Expression of nuclear FOXA2 was increased by MK2206 compared with the high glucose/insulin group (P < 0.05). These results suggest that maternal diabetes induces fetal lung FOXA2 phosphorylation through the Akt pathway, and also affects the maturation of alveolar epithelial cells and reduces levels of SP-B and SP-C in the fetal lungs. An Akt inhibitor reversed the changes in SP expression in vitro.
Hyperglycemia [either maternal diabetes or gestational diabetes mellitus (DM)] during pregnancy is associated with adverse maternal and fetal outcomes [1,2]. Poor control of blood glucose in mothers may lead to fetal lung development disorders and surfactant protein (SP) deficiency [3]. Deficiency of alveolar SPs along with structural immaturity of the lung may lead to neonatal respiratory distress syndrome (RDS), which is associated with high mortality [4]. The natural course of RDS begins at birth (or shortly after) and increases in severity over the first 2 days of life. Lung development requires the coordination and completion of five developmental stages (embryonic period, pseudoglandular period, canalicular period, original saccular period, and alveolar stages) [5][6][7]. These processes involve both structural maturation and SP expression.
The SPs are components of the surfactant and host defense systems required for postnatal adaptation to air breathing [2,8]. They are part of the surface-active lipoprotein complex formed by type II alveolar cells (AECII) [9] and reduce surface tension, facilitating gas exchange [9]. SP-B is essential for lung function and Abbreviations AECII, type II alveolar cells; DM, diabetes mellitus; MLE 12, murine lung epithelial 12; RDS, neonatal respiratory distress syndrome; SP, surfactant protein; STZ, streptozocin. homeostasis after birth [10]. SP-C is extremely hydrophobic and plays an essential role in the formation of the SP complex for normal lung function [11].
Lung maturation is influenced by many transcription factors, growth factors, and signaling molecules [5][6][7]. The Fox family of transcription factors has wide functions in embryonic development, cell cycle regulation, glucose and lipid metabolism, aging, and immune regulation [12,13]. Forkhead box A2 (FOXA2), also called hepatocyte nuclear factor 3B, regulates lung and pancreas development and glucose and lipid metabolism and is mainly expressed in the liver, pancreas, lung, and adipose tissue [8]. FOXA2 is involved in diabetes and response to hyperglycemia by regulating alpha-cell differentiation and glucagon secretion [12]. In the lung, FOXA2 expression is restricted to specific subsets of respiratory epithelial cells [14] and is limited to AECII late in gestation, where it plays an important role in alveolar epithelial cell differentiation and SP expression [15]. FOXA2 is coexpressed with a number of proteins involved epithelial cell differentiation and lung development, including thyroid transcription factor 1, Clara cell secretory protein, Muc5A/C, E-cadherin, vascular endothelial growth factor A, SP-A, SP-B, SP-C, and SP-D [13][14][15][16][17][18][19][20][21][22]. FOXA2 plays essential roles at multiple important periods of lung morphogenesis [23] and is important for TH2 cell-mediated innate immunity in the developing lungs [24]. FOXA2 can regulate the differentiation of AECII and the synthesis of pulmonary surfactants [2,16].
Energy homeostasis is controlled via the insulin receptor and insulin substrate and involves various downstream pathways such as phosphoinositide-3kinase (PI3K)/Akt and mitogen-activated protein kinase 3/1 (MAPK3/1). The major metabolic effector of insulin is thought to be the PI3K/Akt pathway, which contributes to metabolic control [25]. PI3K/Akt signaling has a very wide range of downstream effects and is also thought to play a pivotal role in mouse lung development [26]. Akt/PKB can phosphorylate FOXA2 directly, resulting in p-FOXA2 being released from the nucleus into the cytoplasm [27], inhibiting the transcriptional activity of FOXA2 [28]. This inactive form of FOXA2 can be regulated by insulin receptors and Akt in the liver of mice with insulin resistance and hyperinsulinemia [29].
As a consequence of diabetic pregnancy, the fetus is exposed to hyperglycemia and shows responsive fetal hyperinsulinemia [30]. We hypothesized that fetal diabetes exposure will promote the phosphorylation of FOXA2, leading to decreased FOXA2 transcription activity in AECII and fetal lung development disorders. Therefore, the aim of the present study was to explore the impact of diabetes exposure on lung development and the events affecting the function of FOXA2 in the lung of rats from mothers with diabetes and to analyze whether the related regulation is mediated by the Akt signaling pathway.

Animals
Specific pathogen-free grade adult Sprague-Dawley rats (female = 20, male = 10, 220-240 g) were purchased from the Hubei Province Center for Disease Control and Prevention and housed at the Animal Experiment Center of Tongji Medical College. Virgin female rats were mated with normal males (female: male ratio of 2 : 1). On the day of conception (day 0 of pregnancy), the presence of sperm in vaginal smears was recorded as pregnancy. Ten pregnant female rats received a single intraperitoneal injection of streptozocin (STZ; 45 mgÁkg À1 ; Sigma, St Louis, MO, USA) on day 2 (DM group). The rats were determined as having diabetes according to fasting blood glucose levels > 6.1 mmolÁL À1 over three consecutive measurements [31]. All 10 rats developed diabetes, and there was no need for replacement. The remaining ten control mothers received saline injection instead of STZ. The fetal lung maturation period was defined based on a previous report, which showed well-differentiated AECII expressing SP-B and SP-C [2]. The study was approved by the Institutional Animal Care and Use Committee, Tongji Medical College, Huazhong University of Science and Technology (No. S607).

Sample collection
Nonfasting blood glucose was measured in the morning from blood collected from the tail vein and using a blood glucose meter. Nonfasting serum insulin levels were measured on day 21 after rats were anesthetized to collect the blood from the heart. Rat fetuses were taken from the uterus after the mothers were sacrificed by cervical dislocation under anesthesia (10% chloral hydrate, 3 mLÁkg À1 , intraperitoneally) on day 21. Fetal blood, amniotic fluid, and fetal lung samples were taken. Parts of the fetal lungs were fixed in 10% formalin, processed, embedded in paraffin, and sectioned. Sections were stained with hematoxylin and eosin for histological examination. Parts of the fetal lungs were fixed in 0.1 mmolÁL À1 glutaraldehyde for electronic microscopy.
Cell culture and treatment MLE12 (mouse lung epithelial type II cells) were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were cultured in Dulbecco's modified Eagle's medium (DMEM; low glucose, Hyclone, Thermo Fisher Scientific, Waltham, MA, USA) containing 10% fetal bovine serum (GIBCO, Invitrogen Inc., Carlsbad, CA, USA) in 25-cm 2 plastic flasks (Corning Inc., Corning, NY, USA). Cells were used at passages three to six. Plastic flasks were kept under 5% CO 2 at 37°C. When the cells were 70-80% confluent, they were trypsinized with 0.05% trypsin/EDTA (Hyclone, Thermo Fisher Scientific) and resuspended in complete media. Resuspended cells were seeded in 6-well plates. In order to study the effects of insulin resistance on FOXA2, an insulin resistance cell model was induced by different conditions of insulin and glucose [no treatment; 30 mM glucose and 25 nM insulin; 30 mM glucose; 25 nM insulin; and 5 lM MK22062Á2HCl (Selleckchem, Houston, TX, USA)]. These conditions have been shown to mimic the condition of insulin resistance in cells [32,33]. The cells were harvested after 48 h of incubation for the subsequent experiments.

Paraffin-embedded sections
Fresh tissue specimens were fixed in 4% paraformaldehyde for > 24 h (the same volume was used for each specimen). Blocks were treated with alcohol gradient for dehydration and embedded in paraffin. Blocks were cut to obtain 4-lm sections. Sections were treated at 60°C and stained with hematoxylin and eosin (HE) to observe the structure of the lungs.

Immunohistochemistry
Samples of fetal lungs were taken from three different litters from each group by random sampling (n = 3/group). Paraffin sections were placed in an oven at 65°C for 2 h, dewaxed, and washed three times with PBS (5 min each time). Antigens were retrieved in EDTA buffer in a lowpower microwave until boiling. After natural cooling, sections were washed thrice with PBS (5 min each time). Peroxidase was eliminated with 3% hydrogen peroxide solution at room temperature for 10 min. Sections were washed in PBS thrice, 5 min each time, and treated with 5% BSA for 20 min. Sections were stained with 100 lL

Real-time reverse-transcription polymerase chain reaction
Total mRNA was extracted from frozen fetal lung tissue and cells of each group (n = 3) using TRIZOL (Invitrogen Inc.) and reverse-transcribed using the RT reagent kit (Takara Bio, Otsu, Japan). Real-time PCR was performed using the Bio-Rad CFX96 Detection System (Bio-Rad, Hercules, CA, USA) with SYBR-green (Bio-Rad). PCR conditions were (a) 95°C for 3 min and (b) 35 cycles at 95°C for 10 s, 60°C for 45 s, 95°C for 39 s, and 60°C for 15 s. Each sample was tested in triplicate, and the mean values were used for quantification. Relative gene expression profiles were analyzed by normalizing to b-actin using the 2 À(DDCt) method. The primers are listed in Table 1.

Western blot
Total proteins, cytoplasmic proteins, and nuclear proteins were extracted from frozen fetal lung tissue and cells using the Total Protein Extraction Kit and the Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime Institute of Biotechnology, Haimen, China), according to the manufacturer's instructions.

Immunoprecipitation
Nuclear proteins from fetal lungs of each group or MLE12 cells from each treatment group were purified using the NE-PER Nuclear and Cytoplasmic Extraction Reagents (#78835; Thermo Fisher Scientific). FOXA2 was immunoprecipitated using an anti-FOXA2 antibody (sc-6554; Santa Cruz Biotechnology) immobilized on protein A/G Agarose (Beyotime Institute of Biotechnology). Phosphorylated FOXA2 (p-FOXA2) was analyzed by western blot using antibody against phospho-threonine (SC-5267; Santa Cruz Biotechnology).

Enzyme-linked immunosorbent assay
Commercial ELISA kits (Elabscience Biotechnology, China; coefficient of variation < 10%) were used to detect secreted SP-B and SP-C in the amniotic fluid and insulin in the serum. All assays were performed with four separate samples in duplicate. Plates were read at 450 nm on a microplate ELISA reader.

Statistical analysis
Statistical analysis was performed using GRAPHPAD PRISM 5 (GraphPad Software Inc., San Diego, CA, USA). Data from representative experiments were presented as means AE standard deviations. Comparisons of two groups were performed using the Student t-test. Multiple groups were compared with ANOVA. P < 0.05 was considered to be significant.

Rat models of maternal diabetes
Female rats were randomized to n = 10 dams/litters per group. Two diabetic (one with miscarriage on day 6 and the other with growth arrest on day 8) and one control (one case of intrauterine fetal death was found in the control group, which might have been due to anesthesia) litters were excluded from analysis due to pregnancy loss, leaving n = 8 dams/73 pups in the DM group and 9 dams/113 pups in the control group. The mothers were monitored for urine output during pregnancy; no differences were observed between the two groups. Litter size was lower in the DM group (9 AE 1 vs. 14 AE 1, P < 0.05 vs. the control group) ( Table 2). All offspring of both genders were included in the analysis, but sex was not recorded.

DM mothers and offspring had higher blood glucose and insulin levels
Diabetes mellitus pregnant rats showed significantly higher nonfasting blood glucose levels on day 21 compared with controls (29.6 AE 0.9 vs. 8.3 AE 0.6 mmolÁL À1 , P < 0.05). In addition, the DM rat fetuses had higher blood glucose levels (17.4 AE 0.6 vs. 2.9 AE 0.2 mmolÁL À1 , P < 0.05 vs. controls). Comparison of blood insulin levels on day 21 showed significantly higher insulin levels in the DM group (1864 AE 149 vs. 676 AE 120 pgÁmL À1 , P < 0.05 vs. controls). The DM offspring also suffered from higher blood insulin levels (729 AE 58 vs. 564 AE 36 pgÁmL À1 , P < 0.05 vs. controls) ( Table 2). It should be noted that fetal blood was sampled after maternal chloral hydrate anesthesia and sacrifice.

Gene
Forward ( DM rat fetuses showed lung pathologic and cellular evidence of immaturity Diabetes mellitus fetal lung tissues showed a reduced number of alveoli, irregular appearance, and irregular thickness of the alveolar septum under light microscopy (Fig. 1A,B). Under electron microscopy, more glycogen and lipid in AECII were observed in DM fetal lungs, with few microvilli, few mature lamellar bodies, and swollen mitochondria (Fig. 1C,D).

SP-B and SP-C levels were lower in the DM amniotic fluid and DM fetal lungs
The expression of SP-B and SP-C in the DM amniotic fluid was lower than in controls (P < 0.05) ( Fig. 2A). SP-B and SP-C were observed in AECII in the control and DM groups by immunohistochemistry (Fig. 2B).
In fetal lung tissues, the protein and mRNA expressions of SP-B and SP-C in the DM fetal lungs were lower than in controls (P < 0.05) (Fig. 2C,D).
These results indicate that SP-B and SP-C protein and mRNA expressions were lower in the DM rat fetal lungs.
DM fetal lungs showed a lower expression of nuclear FOXA2 and a higher expression of p-FOXA2 Forkhead box A2 was expressed in AECII, as detected by immunofluorescence (Fig. 3A). Western blot showed significantly lower nuclear FOXA2 expression in the DM fetal lungs compared with controls (P < 0.05) (Fig. 3B). Notably, the expression of p-FOXA2 was significantly higher in the DM fetal lung tissues (P < 0.05) (Fig. 3C). These findings suggested that higher p-FOXA2 and lower nuclear FOXA2 in DM offspring might be associated with Table 2. Maternal and fetal characteristics of rats with and without diabetes exposed (DM). The DM group had fewer offspring and lower average birth weight than the control group. Maternal and fetal rats in the DM group showed significantly higher levels of blood glucose and blood insulin compared with rats in the control group (*P < 0.05). Glucose and insulin were measured at 21 days. lung dysplasia in rats. While no significant difference in total FOXA2 gene or protein expression was found (Fig. 3B), the activated form of FOXA2 was lower in the DM fetal lungs. Furthermore, the TUNEL assay suggested that the apoptotic ratio was higher in DM offspring than in controls (Fig. 3D).
MK2206 partly reversed lower nuclear FOXA2 and higher p-FOXA2 expression observed under high glucose/insulin conditions Further testing using an in vitro model showed that the Akt pathway was inhibited when an Akt inhibitor was used. The levels of p-AKT were the highest in the high glucose/insulin group and were decreased when cells under high glucose and insulin condition were treated with MK2206 (P < 0.05) (Fig. 4A). Total FOXA2 protein levels remained at the same level among the groups (P > 0.05) (Fig. 4B). Nuclear FOXA2 extracted from the cells of the high glucose and insulin group was lower than in the control group (P < 0.05). Interestingly, the expression of nuclear FOXA2 was increased in the MK2206 group compared with the high glucose and insulin group (P < 0.05) (Fig. 4B).
The expression of p-FOXA2 was the highest in the high glucose and insulin group and was decreased when the cells were treated with MK2206 (P < 0.05) (Fig. 4B). The results suggested that changes in nuclear FOXA2 and p-FOXA2 in cells of the high glucose and insulin group could be partly reversed by MK2206. The analysis of PARP and caspase 3 cleavage suggested that apoptosis was increased compared with controls under high glucose/insulin (P < 0.05), but was decreased by the addition of MK2206 (P < 0.05) (Fig. 4C).

MK2206 partly reversed the changes in SP-B and SP-C expression observed under high glucose and insulin conditions
Western blot and RT-PCR showed that the protein and mRNA expression levels of SP-B and SP-C in the high glucose and insulin group were lower compared with the control group (P < 0.05), but were increased after inhibition of the Akt pathway by MK-2206 (P < 0.05) (Fig. 5A,B). The results suggested that lower expression levels of SP-B and SP-C in the high glucose and insulin group could be partly reversed by MK2206.

Discussion
Diabetes exposure is known to impair lung development and to be associated with RDS, but the exact mechanisms are currently unclear. The present study showed that maternal diabetes induces fetal lung FOXA2 phosphorylation through the Akt pathway and also affects the maturation of alveolar epithelial cells, reducing the levels of SP-B and SP-C in the fetal lungs. An Akt inhibitor reversed the changes in SP expression in vitro.
The lung is affected during fetal development in DM mothers, both in the perinatal period and later in life [34][35][36]. In the present study, DM fetal lungs showed low markers of fetal lung maturation, as supported by a previous study [37]. In addition, the present study suggests that the protein and mRNA expressions of SP-B and SP-C in the DM fetal lungs were lower than in controls. The results also showed immature cells with glycogen and lipid deposition, few villi, and few lamellar bodies in the DM fetal lungs. High blood glucose levels stimulate insulin secretion, resulting in fetal hyperinsulinemia [38].
The present study confirmed the expression of FOXA2 in fetal lung epithelial cells. Some studies showed that FOXA2 activates lipid metabolism and ketogenesis in the liver during fasting and is inhibited by insulin-PI3K-Akt signaling mediated by phosphorylation at Thr156 and nuclear exclusion, not only during feeding, but also in hyperinsulinemic ob/ob or db/db mice and in animals with diet-induced obesity [15,29]. The present study showed that nuclear FOXA2 was significantly lower while p-FOXA2 was higher in the DM fetal lungs compared with controls. Even though there was no change in total FOXA2 expression in fetal lungs, our results suggest that diabetes exposure reduces the activated nuclear FOXA2 through Akt phosphorylation. Supporting these results, a previous study in hyperinsulinemic mice showed that FOXA2 is phosphorylated by Akt via the PI3K-Akt/PKB pathway, resulting in FOXA2 being released from the nucleus and persistently localized in the cytoplasm, losing transcriptional activity and consequently affecting downstream gene expression [17,29]. Therefore, it could be hypothesized that fetal hyperinsulinemia leads to FOXA2 phosphorylation, and then, p-FOXA2 is discharged from the nucleus. p-FOXA2 without transcriptional regulation could then be found in the plasma, affecting the differentiation of AECII and SP synthesis. It is important to note here that the rat diabetes model used in this study induced diabetes early in the pregnancy, and FOXA2 is expressed in AECII late in gestation so there may be some differences between this model and actual gestational diabetes mellitus. There are other differences between short-and long-term activation. Akt has many activities besides regulating glucose metabolism and can become resistant with long-term diabetes exposure and insulin resistance that normally occurs via Akt may shift to the MAPK pathway [25]. In the in vitro experiments, in which short-term high glucose and insulin induction were used, the protein and mRNA expression levels of both SP-B and SP-C in the high glucose and insulin group were lower compared with the control group, but were increased by MK2206. The present study confirms that MK2206 reduces the phosphorylation of Akt, as supported by a previous study [27].
This study involved investigation of maternal hyperglycemia, fetal hyperglycemia, and hyperinsulinemia, but has some limitations, the mouse model used STZ to induce b-cell death and induce diabeteslike symptoms, but gestational diabetes, the most common type of diabetes during pregnancy, is not due to loss of b-cell activity [39], so this model may not completely mimic the clinical situation. Nevertheless, our model simulates the important triad that affects fetal development. The maternal insulin levels were measured after anesthesia, which may account for the high levels. Fetal blood was also sampled after maternal chloral hydrate anesthesia and sacrifice. We did not record the sex of the offspring in this study, and gender differences might have an influence on the development of diabetes. We also did not calculate the lung weight/body weight ratio to evaluate whether smaller lungs were found in smaller pups, or count the pups from the uterine horn, since the first four pups are exposed to the highest glucose levels [40,41], or stratify them by weight, larger pups may be more affected by high glucose levels. These issues should be addressed in future studies.

Conclusion
In summary, maternal diabetes induces fetal lung FOXA2 phosphorylation through the Akt pathway, thereby affecting the maturation of alveolar epithelial cells and reducing the levels of SP-B and SP-C in the fetal lungs. An Akt inhibitor reversed the changes in SP expression in vitro. It remains nevertheless to be proven directly that these post-translational modifications of FOXA2 inactivate its function and cause pulmonary pathologic and cellular evidence of immaturity.