SEARCH

SEARCH BY CITATION

Keywords:

  • lung development;
  • branching morphogenesis;
  • alveolarization;
  • bronchial smooth muscle cells;
  • alveolar myofibroblasts

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

Disrupted lung alveolar myofibroblast and bronchial smooth muscle (BSM) cell development may lead to pulmonary disorders such as bronchopulmonary dysplasia. The molecular mechanisms that regulate BSM and alveolar myofibroblast development are not fully understood. Here we show that mSmile (murine Smile), a novel transmembrane protein with tetratricopeptide repeats, functions in lung alveolar myofibroblast and BSM cell development. mSmile mutant mice exhibit early neonatal lethality with few mice surviving up to 3 weeks. Mutant lungs display both airway branching morphogenesis defect during fetal lung development and alveolarization defect after birth. These defects are associated with reduced numbers of BSM cells in the peribronchial subepithelial region and clefts and myofibroblasts in alveolar septae. Expression of fibroblast growth factor-10 and its down stream target Bmp-4, which are important for BSM formation, is decreased. In vitro, mSmile mutant embryonic fibroblasts show reduced receptor activation and induction of myofibroblast formation in response to Transforming growth factor-β (Tgf-β), indicating that mSmile may mediate myofibroblast development through modulation of Tgf-β signaling. These studies identify mSmile as a novel gene specifying both the BSM and lung alveolar myofibroblast lineages, contributing to our understanding of the biological control of the development of these cells, and may provide insights into the aberrant smooth muscle and alveolar myofibroblast development that occur in pathological conditions. Anat Rec, 2012. © 2011 Wiley Periodicals, Inc.

Epithelial-mesenchymal interactions are essential for the development of many endodermal organs including the lungs. The precise specification of epithelial and mesenchymal cell lineages is vital for the patterning of lung structures. Even though the growth, morphogenesis, and differentiation of lung epithelial cells have been extensively studied (Cardoso,2001), development of the lung mesenchyme into various cell types that make up the lung stroma is poorly understood. Lung stroma consists of endothelial cells, fibroblasts, lipofibroblasts, bronchial and vascular smooth muscle cells, alveolar myofibroblasts, and others (McGowan and Torday,1997). Alveolar myofibroblasts are particularly important for the formation pulmonary alveoli. These are interstitial contractile cells with features intermediate between fibroblasts and smooth muscle cells, and are characterized by the expression of α-smooth muscle actin (Leslie et al.,1990). Alveolar myofibroblasts are essential for secondary septation, a process critical for alveolarization. During alveolar development, secondary crests arise from the prealveolar (primary) septae and elongate to form secondary septae that subdivide the prealveolar sacs into numerous alveoli. Alveolar myofibroblasts are found at the tips of secondary septae and are associated with elastin deposition (Vaccaro and Brody,1978; Noguchi et al.,1989). In the Pdgf-A null mice, development of alveolar myofibroblasts is impaired resulting in the absence of these cells in the lung parenchyma and lack of elastin deposition leading to disrupted secondary septation, alveolar development, and early neonatal death (Boström et al.,1996). Pdgf-A is thought to act through the Pdgfr-α receptor to regulate the growth and spread of alveolar myofibroblast progenitors from the terminal airways to the prealveolar septae (Lindahl et al.,1997). Besides Pdgf-a, notch signaling has also been shown to regulate myofibroblast differentiation during alveolarization (Xu et al.,2010). In mice, retinoic acid can reinitiate alveolar formation in the adults, during which alveolar myofibroblasts are reinduced, a process requiring FGF signaling (Perl and Gale,2009). Besides these, few other genes have been shown to regulate lung alveolar myofibroblast specification.

Smooth muscle cells are normally found around the proximal airways and blood vessels (Jostarndt-Fogen et al.,1998). During airway morphogenesis, bronchial smooth muscle (BSM) cells develop in a tight temporally and spatially coordinated manner with the epithelium. BSM cells are normally found surrounding the proximal bronchioles and at the clefts of the bifurcation points, but are conspicuously absent around the budding epithelial tips (Mitchell et al.,1990; Tollet et al.,2001). Because of this precise positional pattern, these cells have been postulated to play a role in regulating epithelial branching morphogenesis, either by stabilization of the proximal airways, by propelling intraluminal liquid causing mechanical stretch stimulation of the epithelium, or by inducing bifurcation of epithelial tubes by cleft formation (Roman et al.,1998; Schittny et al.,2000). The origin and development of BSM cells are only beginning to be understood. A recent study showed that BSM cells are derived from a population of mesenchymal precursors expressing fibroblast growth factor-10 (Fgf-10) that are present in the lung mesenchyme distal to the growing epithelial buds (Mailleux et al.,2005). In this study, the authors also showed that Fgf-10 is necessary for and acts in a cell non-autonomous manner in the formation of BSM cells. Besides Fgf-10, sonic hedgehog (Shh), a morphogen expressed by lung epithelial cells, is also necessary for BSM development. In the absence of Shh, BSM cells do not develop (Pepicelli et al.,1998).

Besides their essential roles in development, BSM cells and alveolar myofibroblasts have also been implicated in many pathological conditions, including bronchopulmonary dysplasia (BPD), pulmonary fibrosis, and asthma. Premature infants exposed to high level of oxygen and mechanical ventilation develop BPD, a disease of arrested lung development. Pathological specimens show an increased number of BSM cells, suggesting that their over-proliferation may contribute to the impaired lung development (Bland,2005). Asthma is a disease of airway hyper-responsiveness in which there is an increased number of BSM cells whose activity may contribute to airflow limitation (Jeffery,2004). In pulmonary fibrosis, increased population of alveolar myofibroblasts may cause the abnormal tissue remodeling leading to increased collagen and extracellular matrix deposition that is the hallmark of this disease (Thannickal et al.,2004). The critical roles of BSM cells and alveolar myofibroblasts in both developmental and pathological processes underlie the necessity to understand the origin and developmental regulation of these essential lung mesenchymal cells. Yet, very little is known about the regulatory mechanisms that govern the specification of lung mesenchymal cell lineages. In this study, we show that a novel transmembrane protein, mSmile, regulates BSM cell and lung alveolar myofibroblast specification during lung organogenesis. Disruption of mSmile function causes defects in both epithelial branching morphogenesis and alveolarization during murine lung development due to impaired development of these two essential lung mesenchymal cell populations.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

Generation of mSmile Mutant Mice

mSmile gene-trap SV129 embryonic stem cells (JST185 cell line, BayGenomics Consortium, SF, CA) were injected into C57BL/6 blastocysts. Offsprings from germline chimera males were backcrossed to either C57/BL6 or AJ129 mice before intercrosses of heterozygous mutant animals were set up. To genotype animals at weaning, dot blots of DNA prepared from tail biopsies were probed with vector (β-Geo) sequences. Inheritance of the mSmile mutant allele was examined by X-gal staining of yolk sacs and hind limb or dot blots of tail DNA. Mutant embryos were identified by both X-gal staining and polymerase chain reaction (PCR) analysis of tail DNA using primers that specifically amplified either the wild-type or the mutant mSmile alleles. The 5′-RACE PCR performed previously by BayGenomics had indicated that the integration site of the gene trap construct was in intron 8 of the mSmile gene. Thus primers (mSF, 5′-AGAAAGCTTTGAAATACTTC-3′, and mSR, 5′-CACATCTCAAACGAGTCCTT-3′) that flank the integration site in intron 8 were used to amplify a 520 bp fragment from the wild-type allele. β-Geo specific primers (βGeoF: 5′-GCAGCGCATCGCCTTCTATC-3′; βGeoR: 5′-AATTCAGACGGC-AAACGACT-GTC-3′) were used to amplify a 320 bp fragment from the mutated allele. RT-PCR primers used to detect mSmile and KIAA mRNA are shown in Table 1. We thank Drs. Jane Brennan and William C. Skarnes for the mSmile mutant mouse and its initial characterizations.

Table 1. RT-PCR primer sequences
GenesForward primerReverse primer
mSmilleAGAAAGCTTTGAAATACTTCCCAGTATATATTCTTGACAT
Kiaa0373GCTTCTCTTCTAATAGGATAACTATTTTCTTAATTTGTGG
β-ActinTACCACAGGCATTGTGATGGACAACGTCACACTTCATGATGG

X-gal Staining Procedures

For β-galactosidase staining, the samples were fixed for 30°min at ambient temperature in fix buffer and washed in buffer W for 15 min three times as previously described (Skarnes,2000). Samples were then incubated in X-gal staining solution for 6 hr or more at 37°C. After staining, embryos were rinsed in wash solution and stored in ethanol formalin acetic acid (EFA). For Cryostat sections, embryos were sectioned at −20°C and stained with 0.67 mg/mL X-gal in buffer W.

Histological Analysis

Embryos or dissected lungs were fixed in 4% paraformaldehyde (PFA) and dehydrated through a graded series of ethanol solutions and embedded in paraffin. Postnatal lungs were inflated with 4% PFA at 20 cm H2O pressure. Paraffin-embedded sections were stained with Hematoxylin and Eosin. Embryos were sectioned sagitally at 6- to 7-m thickness for immunohistochemistry and in situ hybridization analyses.

Immunohistochemistry

Seven micrometer thick sections were deparaffinized and rehydrated by standard methods, and endogenous peroxidase was quenched with 3% H2O2 in methanol for 15 min. After blocking with 5% normal serum for 1 hr at room temperature, sections were treated with the Vector M.O.M. Basic kit (Vector Laboratories, Burlingame, CA) and then incubated with mouse anti-human α-smooth muscle actin monoclonal antibody (clone 1A4, DAKO, Carpinteria, CA) diluted in phosphate buffered saline (PBS; 1:250 dilution) overnight at 4°C. Sections were then washed in PBS, incubated with biotinylated secondary antibody for 1 hr at ambient temperature, washed in PBS, and incubated with Vectastain Elite ABC reagent (Vector Laboratories, Burlingame, CA) for 30 min at ambient temperature. Antibody binding signals were developed with Fast Red (Sigma, St. Louis, MO).

In Situ Hybridization

In situ hybridization was carried out on deparaffinized sections with digoxigenin-labeled riboprobes. The antisense riboprobes for Pdgfr-α were generated from subcloned DNA fragments obtained by RT-PCR of embryonic mouse lung RNA using gene specific primers as previously described (Lindahl et al.,1997). Positive hybridization signals were detected immunologically with an anti-digoxigenin monoclonal antibody (Roche Applied Science, Indianapolis, IN).

Quantitative Real-Time RT-PCR

Total RNA from individual embryonic lungs was isolated using Trizol reagent (Invitrogen, Carlsbad, CA) and purified on RNeasy quick spin columns (Qiagen, Valencia, CA) and treated with DNase I (Invitrogen, Carlsbad, CA). First strand cDNA was synthesized with 1-μg total RNA using Superscript II and Oligo (dT12–16) primer (Invitrogen, Carlsbad, CA). The resulting templates were subjected to real-time PCR reactions. Primer and probe sequences used are shown in Table 2.

Table 2. Taqman primer and probe sequences
GenesProbeForward primerReverse primer
Fgf-10CCGGGCTGCTGTTGCTGCTTCTTGCCTCAGCCTTTCCCCACCTTGGCAGGTGACAGGGAAC
Bmp-4ACAGATGTTTGGGCTGCGCCGGGACTTCGAGGCGACACTTCTTGCTAGGCTGCGGACG
α-SMACCAGAGACTCTCTTCCAGCCA TCTTTCATTGAAACGAACGCTTCCGCTGATGCCCGCTGACTCCATC
mGapdhCCGCCTGGAGAAACCT GCCAAGTATGTGTGTCCGTCGTGGATCTGACCTGCTTCACCACCTTCTTGAT

The 5′ and 3′ modifications of all the probes are FAM and TAMRA, respectively. The mean number of cycles to threshold (CT) of fluorescence detection was calculated for each sample (n = 3 for each genotype group) and the results were standardized to the mean CT of murine glyceraldehyde 3-phosphate dehydrogenase (Gapdh) for each sample tested. The relative level of cDNA abundance was determined by comparative CT method (user Bulletin #2, Applied Biosystems), and results for mutant samples were expressed as percentage of the level of the corresponding wild-type samples.

Isolation and Culture of Fibroblasts from Mouse Embryos

Mouse embryo fibroblasts (MEFs) were isolated as described (Santiago-Josefat et al.,2001) from individual E13.5 embryos generated by heterozygous crosses. The genotypes of the MEF cells were determined by PCR analysis of genomic DNA. MEFs were cultured in complete Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100-μg/mL streptomycin at 37°C in a humidified atmosphere of 95% air/5% CO2. MEFs from the second to fourth passages were used in all experiments. To induce myofibroblast differentiation, MEFs were treated with 2 ng/mL recombinant human TGF-β1 for 48 hr (R&D Systems, Minneapolis, MN) and then analyzed for α-SMA expression by quantitative real-time RT-PCR.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

Disruption of the mSmile Gene in Mice Resulted in Postnatal Lethality

mSmile is a novel multiple membrane-spanning protein composed of 963 amino acids. Human SMILE was initially identified by the cDNA sequencing consortium of the German Genome project (GenBank accession # NM_181783). Structural analysis predicts nine transmembrane domains and 10 tetratricopeptide repeats (TPRs) (Fig. 1A). TPR repeats are found in proteins with various cellular functions, including cell cycle control, protein transport, and protein dephosphorylation (Blatch and Lassle,1999). TPR domains participate in protein–protein interactions with TPR-containing and SH2 domain-containing proteins (Das et al.,1998).

thumbnail image

Figure 1. Generation and phenotype of mSmile null mutant mice. A: Predicted protein domain structure of the mSmile protein. B: Insertion of the gene trap vector in the mSmile gene occurred in intron 8, thus creating a β-Geo fusion protein with the N-terminal 391 amino acids of mSmile. C: PCR analysis of genomic DNA using mSF, mSR, βGeoF and βGeoR primers showing amplification of a 520-bp fragment from wild-type allele or a 320-bp fragment from the trapped allele. D: RT-PCR analysis of RNA with mSmile and KIAA gene specific primers showing intact expression of KIAA and undetectable expression of mSmile. E: Wild-type and mSmile homozygous mutant mice at birth (p0), and at 3 weeks of age. Homozygous mutant mice often appeared cyanotic and died from respiratory failure at birth. Viable mutants lived to ∼3 weeks and displayed growth retardation.

Download figure to PowerPoint

A mouse embryonic stem cell line with a disruption in the mSmile gene was generated during an insertional mutagenesis screen using a gene trap approach (Skarnes,2000). The gene trap construct containing a strong splice acceptor followed by cDNA encoding β-Geo was introduced into mouse ES cells, which causes insertional mutations with expression of β-Geo under the control of endogenous gene promoters. An ES cell line containing an insertional mutation disrupting the mSmile gene was identified. Mapping of the 5′ end of the chimeric mRNA transcript resulted from the insertion showed that the gene trap construct was integrated in intron 8 of the gene (Fig. 1B). The insertion site was further mapped by genomic PCR analysis, and four primers (mSF, mSR, βGeoF, and βGeoR) were designed to distinguish wild-type and mutant alleles (Fig. 1C). The ES cells carrying the mutated allele were injected into mouse blastocysts and resulted in germ line chimeras that were used to generate heterozygous mutant mice.

Close examination of the mSmile gene in the Ensemble database showed that another gene, KIAA, spans the same chromosomal segment in the opposite direction. This is the only other gene that was present near the mSmile integration site on chromosome 10. We used RT-PCR to determine whether the insertion affected expression of mSmile and KIAA. RT-PCR analysis showed a complete absence of mSmile transcripts in mSmile homozygous embryos (Fig. 1D). In contrast, KIAA transcripts were present in mSmile homozygous embryos (Fig. 1D). These results suggest that only the mSmile gene is disrupted in the mSmile mutant mice, and that the phenotypes observed in these mice are the result of a mutation in the mSmile gene.

Intercrosses of heterozygous mSmile mutant mice showed that mutant mice were born at the expected Mendelian frequency, indicating that this mutation does not result in embryonic lethality. Weight and appearance of homozygous mice were similar to their wild-type littermates at birth (Fig. 1E). However, some of the pups homozygous for the mutation in mSmile became cyanotic and died shortly after birth from respiratory failure. This early postnatal lethality was observed with homozygous mice that were bred on either the C57Bl/6 or 129/AJ strain background. Only a small percentage of homozygous mice survived up to 3 weeks. These displayed growth retardation (Fig. 1E), as had been seen in other mouse mutants with disrupted postnatal lung development (Boström et al.,1996; Weinstein et al.,1998). The early postnatal lethality due to respiratory failure prompted us to examine whether this novel gene functions in lung development.

Temporospatial Expression of mSmile During Embryonic and Lung Development

The expression pattern of mSmile during mouse embryogenesis was examined by staining for β-galactosidase expression in heterozygous mutant embryos. mSmile heterozygous mutant embryos showed no discernable phenotypic abnormalities. In E9.5 embryos, X-gal staining was observed in lateral mesoderm and splanchnic mesoderm (Fig. 2A). In E10.5 embryos, X-gal staining was found in the mesenchyme of the foregut region, intersomitic blood vessels, and hind limb (Fig. 2B and data not shown). In E12.5 embryos, high levels of X-gal staining were found in the craniofacial mesoderm, choroid plexus, intestine, primitive vertebrae, as well as in the mesenchyme of the developing lung (Fig. 2C and data not shown). In the lungs, more intense X-gal staining was evident in the mesenchymal cells juxtaposed to the proximal airway epithelial cells (Fig. 2D). This was more evident in the E14.5 lungs, where X-gal staining was more intense in the cells immediately surrounding the proximal airways (Fig. 2E). The location of these cells suggests that they are bronchiole smooth muscle cells. In situ hybridization with an mSmile antisense riboprobe showed identical expression pattern of mSmile mRNA compared with X-gal staining in the E14.5 lungs (Fig. 2F). Interestingly, whereas mSmile was expressed at high levels during the pseudoglandular stage, by the canalicular stage (E16.5), its expression began to decrease. RT-PCR analysis of endogenous mSmile transcript level showed decreased expression level in the lungs at E16.5 and after birth (Fig. 2G).

thumbnail image

Figure 2. Expression of mSmile during mouse development. A,B: X-gal staining of whole embryos of mSmile heterozygote mutant mice at E9.5 (A), E10.5 (B). In E9.5 embryos, X-gal staining was observed in lateral mesoderm, splanchnic mesoderm. In E10.5 embryos, X-gal staining was observed in the mesenchyme of the foregut region, intersomitic region, and the hind limbs. C: X-gal staining of cryosections of E12.5 heterozygote mutant embryos. X-gal staining was observed in the craniofacial area, the intestine, the primitive vertebrae, and the lungs. Abbreviations: SM, splanchnic mesoderm; LM, lateral mesoderm; FG, foregut; ISV, intersomitic vessels; HL, hind limb; CP, choroid plexus; CFM, craniofacial mesoderm; PV, primitive vertebrae; LM, lung mesenchyme; In, intestines. D,E: X-gal staining of cryosections of E12.5 (D) and E14.5 (E) lungs. X-gal staining was seen diffusely in the mesenchyme of the developing lung at E12.5, with more intense level in cells juxtaposed to the proximal airways (D, arrow). High level of X-gal staining was subsequently restricted to the subepithelial mesenchyme surrounding the proximal airways (E, arrows). Inset in panel E shows higher magnification of an airway with positive lacZ staining in the subepithelial region surrounding the airway (arrow). F: In situ hybridization of E14.5 lung sections with antisense mSmile riboprobe showing mSmile expression at the subepithelial mesenchyme surrounding the proximal airways (F, arrows), consistent with the X-gal staining results. G: RT-PCR analysis of mSmile mRNA transcripts in total RNA isolated from lungs at various stages of development. mSmile mRNA expression decreases from late fetal lung development to postnatally. Scale bar in D-F: 200 μm.

Download figure to PowerPoint

The Lungs of mSmile Mutants Showed Impaired Alveolar Formation and Branching Morphogenesis

At birth, lungs of mSmile mutant mice were of comparable size to those of their wild-type littermates, and showed normal lobation pattern (Fig. 3A,B). However, histological examinations of homozygous mutant lungs at postnatal Day 9 (P9) revealed morphological abnormalities. Lungs from mutant mice showed alveolarization defects, with dilated distal air sacs and reduced number of terminal airspaces (Fig. 3C,D). At this age, alveolarization has been initiated and complex alveolar structures can be seen in wild-type lungs (Fig. 3C). In contrast, mutant lungs showed simplification of distal lung architecture (Fig. 3D). The alveolarization defect persisted, and in 3-week-old homozygous mutant lungs, distal airspace remained enlarged, with few secondary septation (Fig. 3E,F). Heterozygous mutant lungs showed no abnormalities (data not shown).

thumbnail image

Figure 3. Histological analysis of mSmile mutant lungs during postnatal and prenatal development. A,B: Gross morphology of wild-type and mSmile homozygous mutant lungs showing normal lobation of mutant lung compared to wild-type lung. Abbreviations: H, heart; Md, medial lobe, Ac, accessory lobe; Cd, caudal lobe; L, left lung; the cranial lobe is behind the medial lobe. C–N: Histological sections of wild-type (C, E, G, I, K, M) and mSmile homozygous mutant (D, F, H, J, L, N) lungs at day 9 postnatal (C,D), 3 weeks postnatal (E,F), E13.5 (G,H), E14.5 (I,J), E17.5 (K,L), and E18.5 (M,N). In the P9 wild-type lungs, the growth of secondary septa had started to subdivide the lumen of the saccules to form alveoli (C, arrows in inset show secondary septae), whereas in the mutant lungs, distal airspaces were enlarged with few secondary septa (D, inset is image at higher magnification showing smooth primary septae without secondary septation). At 3 weeks of age, in the wild-type lungs, secondary septation has continued resulting in complex alveolar structures (E, inset shows higher magnification), whereas the mutant lungs continued to display enlarged distal airspaces with few secondary septation (F, inset shows higher magnification). In the embryonic lungs, branching morphogenesis in the mutant mice was markedly disrupted. At E13.5 and E14.5, mutant lungs showed dilated epithelial tubes with decreased numbers of terminal buds (H,J, arrows) compared to wild-type littermate lungs (G,I, arrows). During the canalicular stage at E17.5, mutant lungs showed fewer number of branching terminal tubules (L, arrow) compared to wild-type littermate lungs (K, arrow). During the saccular stage at E18.5, the terminal acinar tubules have formed complex distal air sacs structures (saccules) in the wild-type lungs (M). In contrast, lungs from mutant mice showed lower complexity (N). These phenotypes were observed in mice on the either the C57BL/6 or the 129AJ background. Scale bar: C-F, 200 μm; G-H, 400 μm; K-N, 200 μm.

Download figure to PowerPoint

The enlarged distal airspace in the homozygous mSmile mutant lungs may be due to a defect in secondary septation. Alternately, a defect in airway branching morphogenesis, which results in reduced numbers of terminal acini and thereby fewer alveolar units, may also cause enlarged distal airspace, as there are fewer alveolar units to occupy the same amount of space in the thoracic cavity. To determine if mSmile mutation also caused defects in airway branching morphogenesis, we examined mSmile homozygous mutant lungs during fetal lung development. At E13.5, mSmile mutant lungs displayed dilated distal epithelial buds with reduced branching (Fig. 3G,H). This continued at E14.5 (Fig. 3I,J). During the canalicular and saccular stage, there were fewer terminal airspaces, likely due to the reduced airway branching that had occurred previously (Fig. 3K–N). These data indicate that null mutation of the mSmile gene caused disrupted branching morphogenesis during the pseudoglandular stage of lung development and this might contribute to the fewer number of gas exchange units and enlarged distal airspace after birth. Thus, mutant mice are born with reduced number of effective gas exchange units and these defects in lung development most likely contribute to the perinatal lethality of the mutation. However, there is also a defect in secondary septation, as there were also fewer secondary septae in the enlarged distal airspaces in the mice that survived until the alveologenesis period (Fig. 3 C,D).

Bronchial Myogenesis and Alveolar Myofibroblast Development are Defective in mSmile Mutant Lungs

Since alveolar myofibroblasts have been shown to be necessary for secondary septae formation and alveologenesis, we asked whether deficient alveolar myofibroblast development might be the cause of the alveolarization defect in the mSmile mutant mice. Immunostaining with an antibody against α-smooth muscle actin (α-SMA), a marker for alveolar myofibroblasts, showed a significant decrease in the alveolar myofibroblast population in the mSmile mutant lungs.

In wild-type postnatal day 3 lung, α-SMA positive cells are found at prealveolar septae (Fig. 4A). However, α-SMA staining was not observed in the periphery of lungs from the mutant mice (Fig. 4B). At postnatal day 5, there are widespread α-SMA positive cells in the parenchyma of wild-type lung (Fig. 4C), whereas they are not as numerous in the mutant lung (Fig. 4D). These results suggest that alveolar myofibroblast development is impaired in mSmile mutant lungs. To determine whether there is also a lack of myofibroblast progenitors, in situ hybridization with an antisense riboprobe against Pdgfr-α, a marker of alveolar myofibroblast progenitors, was performed on E18.5 lung sections. Pdgfr-α expression was observed in cells around the terminal airways and throughout the mesenchyme in wild-type embryos (Fig. 4E). In contrast, Pdgfr-α expressing cells were few in the parenchyma of mSmile mutant lungs (Fig. 4F). Most of the Pdgfr-α expressing cells in the mSmile mutant lung at this stage were found around terminal airways. These results showed that mSmile mutation resulted in reduced alveolar myofibroblast progenitors and alveolar myofibroblasts, and suggest that this may be the cause of the impaired secondary septae formation in the mutant mice.

thumbnail image

Figure 4. Alveolar myofibroblast and BSM development in the mSmile mutant lungs. A-D: Immunohistochemistry with α-SMA antibody of wild-type (A,C) and mSmile homozygous mutant (B,D) lung sections at P3 (A,B) and P5 (C,D). α-SMA positive cells were found around proximal airways and blood vessels in both wild-type and mutant lungs. However, wild-type lungs also contained a population of α-SMA positive cells in the prealveolar septae (A, arrows) consistent with alveolar myofibroblasts, whereas in the mutant lungs, these were absent. E,F: In situ hybridization of E18.5 wild-type (E) and mutant (F) lung tissue sections with antisense riboprobe for Pdgfr-α, a marker of myofibroblast precursors. Pdgfr-α expression was observed throughout the mesenchyme surrounding the branching epithelium of wild-type lungs, whereas both the intensity and the number of Pdgfr-α positive cells appeared decreased in the mutant lung. G,H: Immunohistochemistry with an α-SMA antibody of E14.5 wild-type (G) and mSmile mutant (H) lungs. In the wild-type lung α-SMA positive cells were present surrounding proximal bronchioles (G, arrow) and at the clefts (G, arrowhead) of branching terminal epithelial buds. However, in mutant lungs, α-SMA staining was reduced surrounding the proximal airways (H, arrow) and was rarely found at the clefts (H, arrowhead). I: Quantitative real time RT-PCR analysis of α-SMA mRNA expression from E14.5 wild-type and mutant lungs confirming decreased α-SMA expression in the mutant lungs. Data were from n = 3 mutant and n = 3 control lungs. J: Quantitative real time RT-PCR analysis of Fgf-10 and Bmp-4 mRNA expression from E14.5 wild-type and mutant lungs showing decreased expression of these genes in the mutant lungs. Data were from n = 3 mutant and n = 3 control lungs. Scale bar: A-D and G-H, 100 μm; E,F, 200 μm.

Download figure to PowerPoint

BSM cells also express Pdgfr-α, albeit at a lower level than alveolar myofibroblast progenitors (Lindahl et al.,1997). The decreased expression of Pdgfr-α in the mSmile mutant lung prompted us to examine whether formation of BSM (BSM) is also disrupted in the mutant lung. Immunostaining for α-SMA in the E14.5 wild-type lungs showed the presence of BSM cells in normal locations surrounding the bronchiolar stalks and in the clefts at branch points of bronchial tubules (Fig. 4G). In contrast, mutant lung displayed relatively weak α-SMA staining at the subepithelial mesenchyme around proximal airway tubules (Fig. 4H). Particularly, mutant lungs rarely displayed α-SMA staining in the subepithelial mesenchyme at the clefts of peripheral branching tubules. To confirm the immunostaining results, the level of α-SMA mRNA expression in wild-type and mutant E14.5 lungs was determined by quantitative real-time RT-PCR analyses, which confirmed reduced expression of α-SMA in the mutant lungs (Fig. 4I). These results show that differentiation along the BSM lineage is also defective in the mSmile mutant lungs, and suggest that loss of α-SMA positive cells at the clefts may contribute to the branching morphogenesis defect in these lungs. Since Fgf-10 and Bmp4 expression has been shown to be important for BSM development, we examined their expression in the mutant lungs. Quantitative RT-PCR analyses showed that both Fgf-10 and Bmp4 mRNA levels are reduced in E14.5 mSmile mutant lungs compared to wild-type lungs (Fig. 4J).

Myofibroblast Development is Impaired in mSmile Mutant Embryonic Fibroblasts

To determine if mSmile function directly contributes to myofibroblast differentiation, we examined if there were differences in the induction of myofibroblast differentiation of wild-type and mSmile mutant mouse embryonic fibroblasts (MEF) in vitro. Transforming growth factor-β (TGF-β) is known to induce the myofibroblast phenotype both in vitro (Desmouliere et al.,1993; Hirschi et al.,2002) and in vivo (Zhou et al.,1996; Sime et al.,1997). We examined the ability of wild-type and mSmile mutant MEF to differentiate into myofibroblasts in response to TGF-β. Untreated wild-type or mSmile mutant MEF cells showed no differences in baseline expression of α-SMA (data not shown). Treatment of wild-type MEF cultures with 2 ng/mL TGF-β for 48 hr increased α-SMA mRNA expression as measured by quantitative real time RT-PCR (Fig. 5A), indicating induction of myofibroblast differentiation. In contrast, mSmile mutant MEF cells treated with TGF-β did not induce α-SMA expression to the same extent (Fig. 5A). These results show that mSmile contributes to TGF-β mediated myofibroblast differentiation. We then determined if this effect of mSmile is due to changes in Tgf-β signaling. As shown in Figure 5B, phosphorylation of the downstream effector Smad-2, which depends on Tgf-β receptor activation, was reduced in the mSmile mutant MEF cells treated with TGF-β compared with control cells. These results indicate that mSmile may mediate myofibroblast induction by TGF-β through modulation of receptor activation.

thumbnail image

Figure 5. TGF-β −induced myofibroblast development was impaired in mSmile mutant embryonic fibroblasts. A: Quantitative real time RT-PCR analysis of α-SMA mRNA expression from wild-type and mSmile mutant mouse embryonic fibroblasts (MEF) following myofibroblast induction with TGF-β treatment showing decreased induction of α-SMA mRNA expression by TGF-β in mutant MEF cultures. The experiments were repeated at least three times using different MEF preparations. B: Western blotting for phospho Smad-2 showing reduced level in mSmile MEF treated with TGF-β compared to control cells.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

In this study, we have investigated the function of a novel gene, mSmile, in lung development. We showed that mutation in the mSmile gene causes disruption of the development of BSM cells and alveolar myofibroblasts, leading to impaired lung epithelial branching morphogenesis and alveologenesis. These findings implicate mSmile function in regulating the specification of lung mesenchyme.

The insertional mutation in the mSmile gene caused by integration of the gene trap vector results in a fusion protein consisting of β-Geo fused to the N-terminal 391 amino acids of mSmile, which include the nine transmembrane domains and 1 TPR repeat. It is likely that this mutation results in a loss of function of mSmile and not a gain of a dominant negative or nonspecific function since the heterozygous mutants do not display any discernable phenotypic abnormalities. However, we cannot completely rule out a dominant negative or nonspecific function of the fusion protein that is gene dose-dependent. Examination of mSmile gene expression in the lungs reveals that mSmile is expressed mostly in the mesenchymal compartment. During branching morphogenesis, the pulmonary mesenchyme differentiates into several different cell types, including smooth muscle cells of the proximal airways, myofibroblast progenitors, and endothelial cells forming the pulmonary vasculature. At E12.5 mSmile is expressed diffusely in the lung mesenchyme. Subsequently, mSmile is highly expressed only in the subepithelial mesenchyme immediately surrounding the proximal airways. The overlapping pattern of mSmile expression with that of α-SMA expression around proximal airways at E14.5 suggests that mSmile positive cells are BSM cells. The widespread expression of mSmile in lung mesenchyme in the early pseudoglandular stage lung and its subsequent restriction to cells of the BSM lineages suggest that mSmile function may be required in multiple cell lineages early, but only in the BSM lineages subsequently.

Disruption of mSmile expression results in impaired lung branching morphogenesis and alveologenesis that is associated with impaired BSM and alveolar myofibroblast development. Previous studies have shown that alveolar myofibroblasts are necessary for the formation of secondary septae, a process critical for alveolar development (Boström et al.,1996; Lindahl et al.,1997). Thus, reduced population of alveolar myofibroblasts in the mSmile mutant lungs may result in impaired septae formation leading to impaired alveologenesis. BSM cells have been suggested to play a role in branching morphogenesis due to their particular location with respect to branching epithelial tubes. The location of these cells around the proximal airways suggests that they may function in airway stabilization, and their location at the clefts suggests that they may function in cleft formation (Heine et al.,1990). Thus, the reduced BSM cell population in the mSmile mutant lungs may lead to impaired branching morphogenesis.

The specification of BSM cells is complex. Fgf-10 expressing cells found in the lung mesenchyme distal to branching epithelial buds are the progenitors of BSM cells (Mailleux et al.,2005). Fgf-10 signaling, through the upregulation of epithelial Bmp-4, is necessary for Fgf-10 positive cells to enter the BSM lineage (Mailleux et al.,2005). We found that the impaired BSM development in the mSmile mutant lungs is associated with decreased expression of Fgf-10 and Bmp-4. Thus, the modulation of expression of Fgf-10 and consequently its downstream target Bmp-4 may be the mechanism whereby mSmile functions to specify BSM lineage. It is of note that mSmile positive cells are present widely in the lung mesenchyme in the early stage of lung development, and their location overlaps with Fgf-10 expressing cells. mSmile expressing cells may interact with Fgf-10 positive cells and modulate their expression of Fgf-10 and their development into BSM cells. Alternately, mSmile may also be expressed in Fgf-10 positive cells and act in a cell-autonomous manner to influence their development into BSM cells.

It is intriguing that both the BSM lineage and the alveolar myofibroblast lineage are affected in the mSmile mutant lungs. It is currently not known whether alveolar myofibroblast progenitors are derived from BSM cells or whether the two lineages are derived from separate populations of lung mesenchymal cells. mSmile may specify the BSM lineage and consequently lung alveolar myofibroblasts, if these cells share the same lineage, or alternately, the alveolar myofibroblast lineage may be separately specified by mSmile. This is the first gene identified that regulates the development of both populations and the fact that one gene specifies both populations suggests that indeed the two populations may share the same lineage or the same progenitor populations.

Our findings of a new gene regulating BSM and alveolar myofibroblast development contribute to our understanding of the biological control of the development of these cells and may provide insights into the aberrant smooth muscle and alveolar myofibroblast development that occur in pathological conditions. We found that TGF-β signaling and TGF-β induced myofibroblast differentiation are reduced in mSmile mutant mouse embryonic fibroblasts, suggesting that mSmile is necessary for TGF-β mediated smooth muscle phenotype induction, perhaps through modulation of TGF-β receptor activation. TGF-β mediated myofibroblastic transformation is thought to underlie the pathogenesis of many fibrotic diseases (Gabbiani,2003). The implication of mSmile in this process makes it a potential therapeutic target. Since myofibroblasts play a key pathogenic role in fibrotic diseases due to their synthesis of the extracellular matrix, delineating the molecular regulation of their development is critical for understanding the pathogenesis of these diseases.

In summary, our data identify a novel gene, mSmile, as a novel factor regulating the specification of BSM cells and alveolar myofibroblasts, whose function is critical for both normal development and diseases, and implicate its function in TGF-β dependent processes. This knowledge may contribute to our understanding of the pathogenesis of diseases such as BPD, asthma, and pulmonary fibrosis. The potential roles of mSmile in these important diseases need to be further investigated.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED