BMP signaling is essential in neonatal surfactant production during respiratory adaptation

Abstract

Deficiency in pulmonary surfactant results in neonatal respiratory distress, and the known genetic mutations in key components of surfactant only account for a small number of cases. Therefore, determining the regulatory mechanisms of surfactant production and secretion, particularly during the transition from prenatal to neonatal stages, is essential for better understanding of the pathogenesis of human neonatal respiratory distress. We have observed significant increase of bone morphogenetic protein (BMP) signaling in neonatal mouse lungs immediately after birth. Using genetically manipulated mice, we then studied the relationship between BMP signaling and surfactant production in neonates. Blockade of endogenous BMP signaling by deleting Bmpr1a (Alk3) or Smad1 in embryonic day 18.5 in perinatal lung epithelial cells resulted in severe neonatal respiratory distress and death, accompanied by atelectasis in histopathology and significant reductions of surfactant protein B and C, as well as Abca3, whereas prenatal lung development was not significantly affected. We then identified a new BMP-Smad1 downstream target, Nfatc3, which is known as an important transcription activator for surfactant proteins and Abca3. Furthermore, activation of BMP signaling in cultured lung epithelial cells was able to promote endogenous Nfatc3 expression and also stimulate the activity of an Nfatc3 promoter that contains a Smad1-binding site. Therefore, our study suggests that the BMP-Alk3-Smad1-Nfatc3 regulatory loop plays an important role in enhancing surfactant production in neonates, possibly helping neonatal respiratory adaptation from prenatal amniotic fluid environment to neonatal air breathing.

neonatal respiratory distress remains a leading cause of neonatal mortality and morbidity, particularly in preterm infants, despite advances in perinatal care (7). Deficiency in pulmonary surfactant is the cause for the majority of the patients. Surfactant is a surface-active lipid-protein complex, and surfactant proteins B and C (SP-B and SP-C) are key proteins in surfactant for its function of surface tension reduction. ATP-binding cassette subfamily A member 3 (Abca3) plays a key role in surfactant synthesis/lamellar body formation and surfactant secretion. Genetic mutations of surfactant proteins and Abca3 and their deficiency only account for a small number of neonatal respiratory distress cases. Therefore, determining the regulatory mechanisms of surfactant production and secretion is essential for better understanding of the pathogenesis of human neonatal respiratory distress.

Bone morphogenetic proteins (BMPs) are a group of growth factors involved in regulating many fundamental biological processes. BMP ligands bind to cell surface receptor complexes of BMP receptor type II (Bmpr2) and BMP receptor type I, which includes Bmpr1a (Alk3), Bmpr1b (Alk6), or Acvr1 (Alk2). The activated BMP receptor Thr/Ser kinases subsequently phosphorylate and activate downstream R-Smads (Smad1, 5, and 8) by protein serine phosphorylation, which subsequently form complexes with Smad4, and translocate into the nucleus, acting as transcriptional comodulators to regulate BMP-target gene expression (22). Previous studies by us and other groups demonstrate that the BMP4-Alk3-Smad1 pathway is essential for fetal lung morphogenesis (4, 21, 25, 27). However, BMP signaling in regulating neonatal lung growth and maturation has never been investigated. We have found that activation of BMP signaling occurred in neonatal mouse lungs immediately after birth. Using genetically manipulated mouse models, we were able to delete BMP receptor Alk3 or downstream Smad1 in lung epithelial cells around birth. Significant reduction of endogenous Alk3-Smad1-mediated BMP signaling in these perinatal mouse lung epithelia resulted in neonatal respiratory distress and death accompanied by atelectasis although the prenatal lung histological structure appeared unaltered. Significant reduction of SP-B, SP-C, and Abca3 was detected in the neonatal lungs in which Alk3-Smad1-mediated BMP signaling was abrogated. Further studies found that BMP-Alk3-Smad1 signaling directly upregulated Nfatc3, which is known as a key transcription factor promoting surfactant protein and Abca3 gene expression (2, 3). Therefore, disruption of this BMP-Alk3-Smad1-Nfatc3 regulatory loop in neonatal lung epithelial cells may be a new pathogenic mechanism underlying neonatal respiratory distress.

MATERIALS AND METHODS

Mouse strains, breeding, and genotyping.

Wild-type (WT) mice were mated in the morning (9:00 AM to 1:00 PM), and the timed-pregnant mice were closely watched for delivery at embryonic day 19 (E19). The mothers were killed after delivering approximately six pups, so that approximately two undelivered littermates were obtained by C-section and considered harvested at E19. The newborn pups were harvested immediately (0 h) or placed with a foster mother for 1, 4, and 12 h before the harvest of lung tissues.

Floxed-Alk3 (Alk3^fx/fx) mice were generated in Dr. Yuji Mishina's laboratory (14, 25), and floxed-Smad1 (Smad1^fx/fx) mice were provided by Dr. Anita Roberts (8, 27). Inducible lung epithelial cell-specific Cre transgenic mice (SPC-rtTA/TetO-Cre) were provided by Dr. Jeffrey Whitsett (17). Lung epithelial-specific conditional knockout (CKO) of Alk3 or Smad1 was achieved by doxycycline (Dox) induction in the overnight timed-pregnant triple transgenic mice (Alk3^fx/fx/SPC-rtTA/TetO-Cre or Smad1^fx/fx/SPC-rtTA/TetO-Cre). Dox was given by intraperitoneal injection (0.1 mg/g ip) at E18.5 and maintained by feeding the mice with Dox-containing food (625 mg/kg) and water (0.5 mg/ml) to achieve fast genetic deletion. Mice with single allele deletion of Alk3 or Smad1 were referred to as heterozygous CKO (HT), and mice without any genetic deletion of Alk3 or Smad1 were grouped as WT controls. Pups were monitored in the first 24 h after birth for their neonatal survival rate. All mice were bred in a C57BL/6 strain background. Mice used in this study were housed in pathogen-free facilities. All procedures were approved by our Institutional Animal Care and Use Committees.

Histology and immunofluorescence analysis.

Morphological analysis was conducted as previously described (13, 25). Briefly, fetal lung was directly isolated, and neonatal lung was inflated under 25 cmH₂O pressure through intratracheal intubation before lung isolation. The isolated lungs were then fixed with 4% buffered paraformaldehyde at 4°C overnight. five-micron sections were stained with hematoxylin and eosin (H and E). Quantification of air space area in H and E section images was performed using Fiji software. Immunofluorescence staining and confocal imaging were performed following the methods published previously (29). The related antibodies were as follows: goat anti-Alk3 (sc-5676; Santa Cruz Biotechnology) (25), mouse anti-Abca3, rabbit anti-SP-B, and rabbit anti-Pro-SP-C (WMAB-13H257, WRAB-55522, WRAB-9337; Seven Hills Bioreagents) (5, 16, 24).

Real-time PCR analysis and primers.

Total tissue RNAs were isolated from snap-frozen lung tissue using an RNeasy kit (Qiagen). Synthesis of cDNA and quantitative RT-PCR analysis were performed using iScript cDNA synthesis kit and SYBR Green I dye on iCycler-iQ system (Bio-Rad), as reported previously (12). The real-time PCR primers for Gapdh, Abca3, Sftpb, and Sftpc were previously published (25). The primers for mouse Nfatc3 gene are 5′-ACC CTT TAC CTG GAG CAA AC-3′ (sense) and 5′- GGG CTC TAT GGT GAG TTT TAG G-3′ (antisense). Gapdh was used to normalize the amount of template cDNA.

Western blot and ELISA.

Detection of lung proteins by Western blot has been previously described (12). Briefly, fresh lung tissues were lysed on ice in RIPA buffer containing 1% protease inhibitor cocktail (no. 78430; Thermo Fisher Scientific). Protein concentration was measured by the Bradford method using reagents purchased from Bio-Rad. Equal amounts (50 g) of total tissue lysate proteins were separated in mini Tris-glycine-extended precast gels (4–15% gradient, Bio-Rad) and transferred into polyvinylidene difluoride membrane using Bio-Rad Trans-Blot Turbo Transfer System. Proteins of interest were detected using the following specific antibodies: anti-Nfatc3 and Alk3 (sc-8405 and sc-5676; Santa Cruz Biotechnology), anti-Abca3, anti-Pro-SP-B, and rabbit anti-Pro-SP-C (Seven Hills Bioreagents), and anti-Smad1 and phospho-Smad1 (6944 and 9511; Cell Signaling Technology). Protein band intensity was measured as previously published (20) and normalized by Gapdh for protein loading.

To measure secreted SP-B and SP-C, bronchoalveolar lavage (BAL) was performed for neonatal mice as previously published (30). Briefly, lungs were cannulated in situ via the trachea and lavaged with 0.2 ml of PBS, using a 0.5-ml syringe with three consecutive cycles of filling and emptying. The supernatant was then collected by centrifugation at 300 g for 5 min and adjusted to equal volume of each sample (n = 4 per genotype group). The concentrations of secreted SP-B and SP-C in BAL fluid were then measured using ELISA kits (MBS013022 and MBS729752; MyBioSource) following the manufacturer's instructions.

Chromatin immunoprecipitation combined with microarray technology.

Chromatin immunoprecipitation combined with microarray technology (ChIP-chip) assay was performed as previously described (26). Briefly, three E18.5 WT lung tissues were pooled, cross linked with 1% formaldehyde, and then homogenized in cold whole-cell lysing buffer with protease and phosphatase inhibitors. Lysates were sonicated using a Branson 250 Sonifier (30 s on and 2 min off, 20 times on high setting). After removal of the debris, chromatin was immunoprecipitated with 25 μg rabbit anti-phosphorylated Smad1 antibody (no. 9511; Cell Signaling Technology). After a washing series of increasing stringency, the antibody-bound chromatin was eluted and treated with RNase and proteinase K. The released DNA was then purified using phenol:chloroform:isoamyl alcohol (25:24:1, vol/vol). After fill-in with T4 DNA polymerase and ligation with a pair of primers, the sample DNA was amplified by a ligation-mediated 24-cycle PCR. The PCR products were purified using the QIAquick PCR purification kit (Qiagen), labeled with fluorescence, and hybridized to MM8 RefSeq promoter arrays (Roche Nimblegen), which cover ∼2 kb upstream and 0.5 kb downstream of the 5′ transcriptional start site of 19,489 annotated mouse genes. The arrays were scanned, and data were extracted and analyzed by Roche Nimblegen.

ChIP.

Smad1 ChIP of E18.5 WT lung tissue was performed using antibodies against Smad1 (sc7965; Santa Cruz Biotechnology), phosphorylated Smad1 (9511; Cell Signaling Technology), and RNA polymerase II (Covance), as described above (26). Coimmunoprecipitated genomic DNAs were analyzed by PCR with primers (sense 5′- CGGAAAGTTTGCAGTGGAG-3′ and antisense 5′- CGCCACAGTTTGCAGTAGTC-3′) flanking the potential Smad1-binding site of mouse Nfatc3 gene promoter (175–310).

Cell isolation, culture, and treatment.

MLE12 cells (murine type II alveolar epithelial-like cells) were obtained from ATCC (CRL-2110) and cultured in HITES medium with 2% fetal bovine serum, 100 U/ml penicillin, and 0.1 mg/ml streptomycin (11). Cells (3 × 10⁵) were seeded onto a six-well plate and cultured for 24 h, followed by 24 h of starvation before stimulation with 25 ng/ml BMP4 and/or LDN-193189 (30 nM). Recombinant mouse BMP-4 protein was obtained from R&D Systems (5020-BP). Small molecule LDN-193189, which specifically inhibits Alk3/Alk2 activity (28), was provided by Dr. Paul Yu at Harvard Medical School. Cells were harvested after 24 h of treatment for gene expression analysis or reporter assay.

Neonatal lung epithelial cells were isolated using a method published previously (1). Briefly, mouse lung tissues were minced to ∼1-mm³ pieces and digested with 40 mg/ml trypsin and 23 U/ml DNase I (Sigma) for 20 min at 37°C. The dissociated cells were then passed through 100-μm mesh and pelleted at 420 g for 5 min. The cells were further digested with 10 mg/ml collagenase (Sigma) and 23 U/ml DNase I for 15 min at 37°C to achieve single cell suspension. The PBS-washed cells were then plated in a culture flask and incubated for 1 h at 37°C to remove fibroblasts by differential attachment twice. The unattached cells were washed and spun down four times at 120 g for 3 min. The cells were incubated for 24 h at 37°C before cell lysate preparation.

Transient transfection of Nfatc3 promoter reporter plasmids.

A DNA fragment of mouse Nfatc3 gene promoter (−327 to +272) was amplified from mouse genomic DNA, subcloned into the pGL2-Basic vector (pNfatc3-luc), and verified by DNA sequencing. pNfatc3-mut-luc was generated from pNfatc3-luc by deleting a DNA fragment (+80 to +248) of Nfatc3 promoter containing potential Smad1-binding site. pCMV-Smad1 and pCMV-Smurf1 plasmids were constructed previously using parent pCDNA3 plasmid (Clontech), in which expression of the subcloned gene is driven by the CMV promoter (26). The constitutively active Alk3 plasmid was a gift from Dr. Takenobu Katagiri (10).

MLE12 cells were transfected with Nfatc3-luc or cotransfected with pCMV-Smad1, pCMV-caAlk3, or pCMV-Smurf1 using Lipofectamine LTX Reagent following the manufacturer's instructions (Invitrogen). In addition, the cells with Nfatc3-luc transfection were also treated with BMP4 protein and/or LDN-193189. Luciferase activity in cell lysates was detected using the dual-luciferase reporter assay system (E1910; Promega) and normalized to Renilla luciferase activity of cotransfected pRL-CMV. Total amount of plasmid DNA for transfection was equalized using pCR-CMV empty vector plasmid.

Data presentation and statistical analysis.

All experiments were repeated at least three times, and data represent consistent results. The quantitative data are expressed as means ± SD. At least four mice were included in each experimental subgroup. Statistical difference between two independent groups was assessed by an independent-sample t-test. The survival rate was calculated by Kaplan-Meier analysis. P < 0.05 was considered statistically significant.

RESULTS

Endogenous BMP signal activity was increased in neonatal mouse lung immediately after birth, in parallel with increased Abca3 expression.

To understand the molecular mechanisms underlying the transition of lung from prenatal amniotic fluid environment to postnatal air breath, we collected lung tissues from neonatal mice with different time periods after birth (0, 1, 4, and 12 h) and their littermates just before delivery (∼E19). By comparing their growth factor signal activities, dynamic changes in endogenous BMP signaling of neonatal lung tissues were detected by measuring phosphorylation of BMP-specific downstream Smad1, which was significantly increased after birth and reached a peak about 4 h after birth. In parallel with that, Abca3 protein expression was significantly increased 4 h after birth, whereas the levels of SP-B and SP-C were not significantly changed (Fig. 1, A and B). Consistently, increased expression of Abca3, but neither Smad1 nor surfactant protein B (Sftpb) and C (Sftpc), at the mRNA level was detected in neonatal lungs after birth (Fig. 1C).

Fig. 1.

Bone morphogenetic protein (BMP)-Smad1 signaling was upregulated in neonatal lung immediately after birth, accompanied by increased Abca3 expression. A: dynamic profiles of Smad1 phosphorylation and surfactant protein (Sp)/Abca3 expression in neonatal lungs. Lung tissues were harvested from the mice just before delivery (embryonic day 19, E19) and different hours after birth. The related proteins were detected by Western blot, and Gapdh was used as a loading control. B: quantitative analysis of the related protein band intensities by densitometry. Experiments were repeated 3 times, and significant increases of pSmad1 and Abca3 were detected at 4 and 12 h (*P = 0.002, @P = 0.005, #P = 0.007, and $P = 0.004). C: gene expression in lung tissues at the mRNA level was measured by real-time PCR. *P = 0.017, #P = 0.002.

Alk3-mediated BMP signaling in perinatal lung epithelial cells is essential for neonatal respiratory function.

Our previous studies found that Alk3-Smad1-mediated BMP signaling in embryonic/fetal mouse lung epithelial cells is important in regulating airway branching morphogenesis and saccular epithelial cell differentiation at midgestation stage (25, 27). Blockade of BMP signaling specifically in lung epithelial cells by deleting Alk3 or downstream Smad1 in mice from E7.5 disrupts fetal lung formation and subsequently results in neonatal respiratory failure. However, the role of BMP signaling in lung epithelia during perinatal lung development, maturation, and neonatal respiratory adaptation has never been studied. We used the same inducible Tet-On triple transgenic mouse model reported in our previous publication (25) but initiated epithelium-specific Alk3 CKO at E18.5 with intraperitoneal injection of Dox followed by Dox administration in drinking water and food. It has been reported that gene deletion could be detected within 16 h after oral administration of Dox in this SP-C-rtTA/TetO-Cre transgenic system (17). Our neonatal mice were born in <24 h after Dox induction, and Alk3 gene knockout in our neonatal mouse lung tissue was verified by Western blot and immunofluorescence staining (Fig. 2).

Fig. 2.

Alk3 conditional knockout (CKO) in perinatal mouse lung epithelial cells. A: Alk3 protein in total lung tissue lysate was detected by Western blot. Lung samples were isolated from E18.5-induced Alk3 homozygous CKO, Alk3 heterozygous CKO (HT), or wild-type (WT) mice 4 h after birth. B: Alk3 expression in the neonatal lung (4 h after birth) was stained with anti-Alk3 antibody (green). Cell nuclei were counterstained with DAPI (blue).

The neonatal Alk3 CKO mice suffered breathing difficulty and cyanosis within 1 h after birth, the signs of neonatal respiratory distress (Fig. 3A). The majority of Alk3 CKO mice died in <10 h. In contrast, WT and HT Alk3 knockout mice survived and breathed normally (Fig. 3B). By examining the isolated lungs from the neonates with different genotypes, we found that the Alk3 CKO lung was not inflated with air compared with those with WT and HT Alk3 genotypes in the same littermate. Lung structures of the neonatal mice with different Alk3 genotypes were then evaluated for their H and E-stained tissue sections under microscopy. Compared with air-filled saccular structures in WT and HT lungs, the peripheral lung of Alk3 CKO neonate was collapsed (Fig. 3, C and D). In contrast, the lung structure of fetal Alk3 CKO mice before birth, which was harvested about 16 h after Dox induction, appeared comparable to those of WT and HT littermates (Fig. 3, C and D). This suggests that short-term blockade of Alk3-mediated BMP signaling near birth did not affect prenatal saccular structural growth but may cause abnormal neonatal respiratory adaptation and functional maturation, resulting in atelectasis.

Fig. 3.

Neonatal Alk3 CKO mice died from respiratory distress. A: neonatal mice with different Alk3 genotypes. B: altered survival was detected in Alk3 CKO mice compared with the neonates with Alk3 WT and HT genotypes. *P = 0.014. C: hematoxylin and eosin (H and E)-stained tissue sections of the lungs isolated from the prenatal (E19) or postnatal mice (4 h after birth) with the indicated genotypes. D: morphometric analyses of peripheral airspace areas shown in H and E-stained lung tissue sections, measured by percentage of peripheral airspace area [(airspace area in tissue section/tissue section area) × 100]. Comparison was performed between WT and Alk3 HT lungs or WT and Alk3 CKO lungs, *P < 0.01.

Endogenous BMP signaling in neonatal peripheral lung epithelial cells supports neonatal respiratory function through enhancing expression of surfactant proteins and Abca3.

To understand the mechanisms by which perinatal deficiency of lung epithelial BMP signaling caused neonatal respiratory distress, we have examined the related molecules that are critical components in surfactant synthesis and production in the neonatal lungs 4 h after birth. By RT-PCR, expression of Abca3, Sftpb, and Sftpc was significantly reduced in Alk3 CKO lung compared with their WT littermate controls (Fig. 4A). Consistently, the protein levels of Abca3, SP-B, and SP-C (both precursor and mature forms) were also decreased in the Alk3 CKO lungs (Fig. 4, B and C). Secretion of pulmonary surfactant, as measured for SP-B and SP-C in BAL fluid, was also significantly reduced in Alk3 CKO mice (Fig. 4D). By immunofluorescence staining, reductions of these proteins were also detected in peripheral lungs of Alk3 CKO neonates, particularly in lung saccular structures (Fig. 4E). Moreover, these changes were further verified in isolated lung epithelial cells from Alk3 CKO neonatal mice compared with those from WT littermate controls (Fig. 5). A small amount of Alk3 protein detected in lung epithelial cells isolated from Alk3 CKO lungs might be contributed by other airway epithelial cells such as Club cells that were not effectively targeted by the SP-C-rtTA/TetO-Cre driver at late gestation (19).

Fig. 4.

Reduced expression of Abca3 and surfactant proteins in Alk3 CKO lungs 4 h after birth. A: gene expression at the mRNA level in lung tissue was compared by real-time PCR. *P = 0.002, #P = 0.009, @P = 0.003 compared with WT controls. B: related proteins in total lung tissue lysate were compared by Western blot. C: quantitative analyses of protein bands detected in B using densitometry. *P = 0.031, #P = 0.046, @P = 0.040, $P = 0.047, &P = 0.013 compared with WT controls. D: secreted SP-B and SP-C in equal volume of bronchoalveolar lavage (BAL), measured by ELISA, was compared between WT and Alk3 CKO samples. *P = 0.013, #P = 0.012. E: altered distributions of the related proteins in neonatal lung tissues with Alk3 CKO genotypes were also detected by immunofluorescence staining as indicated in the pictures. Cell nuclei were counterstained with DAPI (blue).

Fig. 5.

Altered expression of the related proteins in isolated lung epithelial cells from neonatal Alk3 CKO lungs vs. those from WT littermate controls. A: Western blot detection. B: quantitative analysis of the related protein band intensities in A by densitometry. Experiments were repeated 3 times. Reduced protein expression of Alk3 (*P < 0.001), Abca3 (*P = 0.01), precursors of SP-B (*P = 0.045) and SP-C (*P = 0.043), and Nfatc3 (*P = 0.004) was verified in Alk3 CKO lungs compared with the WT controls.

Smad1 was a key downstream signal transducer that mediates BMP-Alk3-enhanced neonatal surfactant production.

As mentioned above, Smad1, 5, and 8 are well-known intracellular proteins specifically for transducing BMP signaling from cell membrane to nuclei. Although both Smad5 and Smad8 express in fetal lung, genetic deletion of Smad5 in developing lung epithelial cells does not cause malformation of lung structure and neonatal respiratory problems (27), and knockout of Smad8 also has no impacts on fetal and neonatal lung development (9). Smad1 was shown to be a key downstream protein in mediating BMP regulation during fetal lung development (27). Therefore, we also deleted Smad1 in lung epithelial cells during the perinatal stage using the same strategy as for Alk3 deletion described above. Similar to what was seen in Alk3 CKO mice, neonatal Smad1 CKO mice also suffered respiratory distress, and most of them died within 24 h after birth. Histopathology showed that neonatal Smad1 CKO peripheral lung was collapsed, whereas its prenatal lung development (E19) was not altered (Fig. 6A). Similar to the molecular changes observed in the Alk3 CKO lung, expression of Abca3 and surfactant proteins (Sftpb and Sftpc) was significantly reduced in neonatal Smad1 CKO mouse lung tissues at both mRNA and protein levels (Fig. 6, B–D). Moreover, reduced expression of Abca3, SP-B, and SP-C proteins was verified in primary lung epithelial cells isolated from Smad1 CKO neonatal mice compared with those from WT littermate controls (Fig. 7).

Fig. 6.

Neonatal mice with lung epithelium-specific Smad1 CKO induced from E18.5 suffered from atelectasis accompanied by reduced expression of surfactant proteins and Abca3. A: comparison of lung histology between Smad1 CKO and WT littermate control before and after birth [E19 vs. 4 h after birth (P1)]. B: altered expression of surfactant proteins and Abca3 at the mRNA level was compared between neonatal Smad1 CKO lungs and WT littermate controls. *P = 0.002, #P = 0.003, @P = 0.001 (n = 3 for each genotype). C and D: altered protein expression in Smad1 CKO lung tissue was detected by Western blot (C) and quantified by densitometry (D). Experiments were repeated 3 times, and significant reductions of Smad1 (*P = 0.026), Abca3 (*P = 0.024), SP-B (*P = 0.046 and #P = 0.029), and SP-C (*P = 0.013 and #P = 0.019) were detected.

Fig. 7.

Altered expression of the related proteins in isolated lung epithelial cells from neonatal Smad1 CKO lungs vs. those from WT littermate controls. A: Western blot detection. B: quantitative analysis of the related protein band intensities in A by densitometry. Experiments were repeated 3 times. Reduced protein expression of Alk3 (*P < 0.001), Abca3 (*P = 0.001), precursors of SP-B (*P = 0.028) and SP-C (*P = 0.010), and Nfatc3 (*P < 0.001) was verified in Smad1 CKO lungs compared with the WT controls.

BMP-Smad1 directly enhanced expression of Nfatc3, which is a well-known transcription factor upregulating surfactant proteins and Abca3.

To understand the molecular mechanisms by which BMP-Alk3-Smad1 signaling upregulates surfactant proteins and Abca3, we used Smad1-specific antibody to perform ChIP, followed by mouse gene promoter DNA microarray analysis (ChIP-chip) to identify promoter DNAs that interact with Smad1 protein. About 1,000 genes were identified with more than 2.5-fold signal ratio, as published previously (27). One of those was Nfatc3 gene on chromosome 8 (Fig. 8A), suggesting a potential interaction between Smad1 and mouse Nfatc3 promoter DNA. Nfatc3 is a direct activator of surfactant proteins and Abca3 gene transcription (2), and blockade of this transcription factor activation results in neonatal respiratory distress associated with morphological and biochemical immaturity of the lung in mice (3). Therefore, we then examined alteration of Nfatc3 in the Alk3 or Smad1 CKO neonatal lungs induced from E18.5. Consistently, Nfatc3 protein was significantly decreased in both Alk3 and Smad1 CKO neonatal lung tissues compared with the WT littermate controls (Fig. 8, B and C). Reduction of Nfatc3 protein in primary lung epithelial cells isolated from both neonatal Alk3 CKO lung and Smad1 CKO lung was also consistently detected (Figs. 5 and 7). Furthermore, in cultured MLE12 lung epithelial cell line, addition of BMP4 significantly increased Nfatc3 expression at both mRNA and protein levels, accompanied by increased expression of Abca3, Sftpb, and Sftpc (Fig. 8, D and F). Specific inhibition of Alk3 by small molecule inhibitor LDN-193189 (LDN) abrogated the above BMP4-stimulated Nfatc3 expression and the related expressional increases of Abca3, Sftpb, and Sftpc. In cultured primary lung epithelial cells isolated from WT neonatal mice, the changes in response to BMP4 and/or Alk3 inhibitor LDN treatments were similar to those detected in MLE12 cells (Fig. 8, E and G).

Fig. 8.

BMP4-Smad1 signaling directly upregulated Nfatc3 expression. A: chromatin immunoprecipitation combined with microarray technology (ChIP-chip) analysis of E18.5 WT lung tissues suggested that Nfatc3 promoter DNA might interact with Smad1, with peak score 2.5. B: Nfatc3 protein expression in lung tissue, detected by Western blot, was compared between neonatal Alk3 CKO and WT littermates or neonatal Smad1 CKO and WT littermates 4 h after birth. C: quantitative analysis of Nfatc3 protein band intensities detected in B by densitometry. Experiments were repeated 3 times. Significant reduction of Nfatc3 protein expression was verified for Alk3 CKO lungs (*P = 0.026) and Smad1 CKO lungs (*P = 0.046). D and E: in cultured MLE12 cells (D) or primary lung epithelial cells (E), addition of BMP4 (25 ng/ml) significantly stimulated Nfatc3 expression (*P = 0.004 in D, *P = 0.020 in E), as well as expression of Abca3 (*P = 0.049 in D, *P = 0.022 in E), Sftpb (*P = 0.001 in D, *P = 0.007 in E), and Sftpc (*P = 0.017 in D, *P = 0.008 in E) at the mRNA level after 24 h, detected by real-time PCR. Pretreatment of cells with Alk3 inhibitor LDN-193189 (LDN) blocked the above BMP4-stimulated gene expression (Nfatc3: #P = 0.008 in D, #P = 0.013 in E; Abca3: #P = 0.047 in D, #P = 0.008 in E; Sftpb: #P = 0.047 in D, #P = 0.020 in E; Sftpc: #P = 0.028 in D, #P = 0.002 in E). F and G: impacts of BMP4 and/or Alk3 inhibitor LDN on the related protein expression in cultured MLE12 cells (F) and the primary lung epithelial cells (G) were also examined by Western blot.

By analyzing the Nfatc3 promoter DNA sequence, a consensus Smad1-binding DNA sequence was found in the 5′-untranslated region (5′-UTR) of both human and mouse Nfatc3 genes (Fig. 9A). With the use of an anti-Smad1 or anti-pSmad1 antibody to pull down the genomic DNA fragments that bind to Smad1, the related Nfatc3 5′-UTR DNA fragment spanning the Smad1-binding site was presented in these conventional ChIP assays (Fig. 9B). Therefore, a mouse Nfatc3 promoter DNA fragment containing the Smad1-binding sequence (600 bp upstream of ATG initiation codon) was subcloned to a luciferase reporter plasmid (pNfatc3-luc) and analyzed in transfected lung epithelial cells MLE12. Addition of BMP4 (25 ng/ml) to the culture medium significantly enhanced this 600-bp Nfatc3 promoter-driven reporter activity, which could be blocked by pretreatment of Alk3 inhibitor LDN (Fig. 9C). Deletion of a DNA fragment containing the Smad1-binding site in this Nfatc3 promoter region (pNfatc3-mut-luc) abrogated its promoter activity in response to BMP4 stimulation (Fig. 9C). Similarly, increased BMP signaling activity by coexpression of constitutively active Alk3 (CaAlk3) and Smad1 significantly activated this Nfatc3 promoter activity, whereas coexpression of Smurf1, which is a ubiquitin ligase to accelerate Smad1-specific degradation, was able to attenuate Alk3-Smad1-mediated activation of the Nfatc3 promoter activity (Fig. 9D). Therefore, the BMP4-Alk3-Smad1 pathway is able to specifically upregulate Nfatc3 expression in lung epithelial cells through its Smad1-binding promoter region, which in turn activates expression of surfactant proteins and Abca3 and ultimately surfactant production and secretion in airspace.

Fig. 9.

Smad1-binding site on Nfatc3 promoter mediated BMP-Alk3-Smad1 activation of Nfatc3 transcription. A: consensus DNA sequence of Smad1-binding site (bold), which is similar to that reported in mouse Id1 promoter, was found in both human and mouse Nfatc3 promoter by DNA sequence analysis. B: ChIP analysis of E18.5 lung tissue using the indicated antibodies. The coprecipitated Smad1-binding DNA sequence in Nfatc3 promoter was detected by PCR. C: luciferase reporter assay to detect the BMP-Smad1-responsive element in Nfatc3 promoter in MLE12 cells. pNfatc3-luc, a reporter plasmid containing WT Nfatc3 promoter; pNfatc3-mut-luc, a reporter plasmid derived from pNfatc3-luc with a deletion of potential Smad1-binding site; BMP4, 25 ng/ml; LDN, 30 nM. *P < 0.001, #P < 0.001, @P < 0.001. D: activation of BMP signaling by coexpressing CaAlk3 and/or Smad1 significantly enhanced the Nfatc3 reporter activity in MLE12 cells, whereas coexpression with Smurf1, a ubiquitin ligase to degrade Smad1, partially blocked the increase of Nfatc3 reporter activity. *P = 0.001, #P < 0.001, @P < 0.001.

DISCUSSION

Mouse lung development initiates from early gestation (E9.5) with lung bud protruding from foregut endoderm, followed by branching morphogenesis from E10.5 to E16.5 (15). The fetal lung then undergoes structural and functional maturation at late gestation, which is required for adaptation from prenatal amniotic fluid condition to air breathing after birth. This includes dilation of the terminal saccules with thinning of interstitial connective tissue, growth of vascular networks, distal airway epithelial cell differentiation, and maturation into 1) squamous type I epithelial cells to form efficient gas-exchange surface and 2) cuboidal type II epithelial cells for surfactant synthesis and secretion (6). Surfactant plays a critical role in reducing surface tension and also in helping fluid clearance at birth. Disruption of surfactant synthesis and secretion will result in neonatal respiratory distress and respiratory failure. Although genetic mutations of surfactant proteins and ABCA3 have been found in some patients of neonatal respiratory distress, the majority of the cases do not have such detectable mutations. The regulatory mechanisms of surfactant production and secretion, particularly during transition from prenatal to neonatal stages, are not completely understood.

Epithelial BMP signaling in regulating fetal lung branching morphogenesis has been extensively studied. Mice with Alk3 or Smad1 CKO induced in lung epithelial cells from early embryonic stage died immediately after birth with severe respiratory distress accompanied by a deficit in lamellar body formation in type II alveolar epithelial cells (4, 25, 27). In contrast, induction of lung epithelial Alk3 knockout 1 day after birth had no detectable effects on postnatal respiratory function (25). Our present study found that endogenous BMP pathway activity was increased in neonatal mouse lung immediately after birth, in parallel with increased expression of Abca3, which is a key protein in the membranes of lamellar bodies of type II alveolar epithelial cells to facilitate processing and packaging of surfactant-associated lipids and proteins (23). These suggest that lung epithelial BMP signaling in the prenatal stage and/or immediately after birth is critical in promoting neonatal respiratory function. In this study, we have induced Alk3 or Smad1 deletion in lung epithelial cells in <1 day before birth, which has no significantly negative impact on fetal lung branching morphogenesis and prenatal lung structural formation. These CKO neonates still suffer severe neonatal respiratory distress accompanied by significant reduction of SP-B, SP-C, and Abca3, which is similar to what is observed in lung epithelium-specific Alk3 or Smad1 knockout induced from early gestation (4, 25, 27). In contrast, deletion of another BMP downstream signal transducer, Smad5, in fetal lung epithelial cells or global Smad8 does not result in any abnormal respiratory phenotype in neonates (9, 27), suggesting that Smad5 and Smad8 may not be involved in lung development and neonatal respiratory adaptation. Therefore, the BMP-Alk3-Smad1 pathway in perinatal lung epithelial cells is more important than that in an earlier fetal stage in terms of promoting neonatal respiratory adaptation and functional maturation.

Using a ChIP-chip approach, we have identified Nfatc3 as a potential BMP-Alk3-Smad1 pathway downstream target gene, which has been reported as a key transcription factor in promoting surfactant protein and Abca3 gene expression. Further molecular biology study has determined and verified a new regulatory element on Nfatc3 promoter that interacts with Smad1 and mediates BMP regulatory function in promoting surfactant protein and Abca3 expression. Deficiency of Alk3-Smad1 signaling in lung epithelial cells results in downregulation of Nfatc3 in vivo, in parallel with reduced expression of surfactant proteins and Abca3. Furthermore, in the cultured lung epithelial cell line MLE12, stimulation with BMP4 is able to increase Nfatc3 expression accompanied by significantly increased SP-B/C and Abca3. Therefore, BMP-Alk3-Smad1-Nfatc3-surfactant proteins/Abca3 turns out to be a novel regulatory mechanism, which may play an essential role in enhancing surfactant production/secretion for neonatal lung adaption and function.

Although it is known that mutations of BMP signaling components are related to pulmonary hypertension in adults (18), the relationship between mutations in BMP signaling components and human neonatal respiratory distress has never been investigated. Our finding here in mouse models may suggest candidate genes for future clinical genetic study of human neonatal respiratory diseases. Moreover, transient activation of BMP signaling in neonatal lung may be a promising novel strategy in stimulating endogenous surfactant production and secretion, particularly in preterm babies.

GRANTS

This work was supported by National Institute of Health grant HL068597 (W. Shi) and a California Institute of Regenerative Medicine Training grant (Y. Luo).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Y.L., Y.M., and W.S. conception and design of research; Y.L., H.C., S.R., and N.L. performed experiments; Y.L., Y.M., and W.S. analyzed data; Y.L. and W.S. interpreted results of experiments; Y.L. and W.S. prepared figures; Y.L. and W.S. drafted manuscript; Y.M. and W.S. edited and revised manuscript; W.S. approved final version of manuscript.