(Bmp-4) plays a role in mouse embry

1693Development 122, 1693-1702 (1996)Printed in Great Britain © The Company of Biologists Limited 1996DEV4663

Evidence from normal expression and targeted misexpression that Bone

Morphogenetic Protein-4 (Bmp-4) plays a role in mouse embryonic lung

morphogenesis

Savério Bellusci2,*, Randall Henderson3,*, Glenn Winnier2, Tsuyoshi Oikawa4 and Brigid L. M. Hogan1,2,†

1Howard Hughes Medical Institute, 2Department of Cell Biology, 3Division of Neonatology and 4Division of Pediatric Nephrology,Vanderbilt University Medical Center, Nashville, Tennessee 37232-2175 USA

*These two authors contributed equally to this work†Author for correspondence

Epithelial-mesenchymal interactions are critical for thebranching and differentiation of the lung, but the mecha-nisms involved are still unclear. To investigate this problemin mouse embryonic lung, we have studied the temporaland spatial expression of genes implicated in the morpho-genesis of other organs. At 11.5 days p.c., hepatocytenuclear factor-3β (Hnf-3β) is expressed uniformly through-out the epithelium, while Wnt-2 expression is confined tothe distal mesenchyme. Sonic hedgehog (Shh) trancriptsare found throughout the epithelium, with high levels in thedistal tips of the terminal buds, while bone morphogeneticprotein-4 (Bmp-4) transcripts are localized at high levels inthe distal tips of the epithelium, with lower levels in theadjacent mesenchyme. Epithelial expression is also seen forBmp-7, but transcripts are less dramatically upregulated atthe distal tips. The Type I Bone morphogenetic proteinreceptor gene(Bmpr/Tfr-11/Brk-1) is expressed at low levelsin the epithelium and in the distal mesenchyme. To inves-tigate the role of Bmp-4 in lung development, we have mis-

expressed the gene throughout the distal epithelium oftransgenic lungs using a surfactant protein Cenhancer/promoter. From 15.5 days p.c., transgenic lungsare smaller than normal, with grossly distended terminalbuds and, at birth, contain large air-filled sacs which do notsupport normal lung function. Labeling with BrdU revealsan inhibition of epithelial proliferation in 15.5 days p.c.transgenic lungs. A small but significant stimulation of pro-liferation of mesenchymal cells is also observed, but this isaccompanied by an increase in cell death. In situ hybridiz-ation with riboprobes for the proximal airway marker,CC10, and the distal airway marker, SP-C, shows normaldifferentiation of bronchiolar Clara cells but a reduction inthe number of differentiated Type II cells in transgeniclungs. A model is proposed for the role of BMP4 and othersignalling molecules in embryonic lung morphogenesis.

Key words: Bmp, Shh, Wnt-2, lung development, transgenicmisexpression, mouse embryo

SUMMARY

INTRODUCTION

Morphogenesis of the embryonic lung involves interactionsbetween the epithelium and the surrounding mesenchyme andextracellular matrix (Alescio and Cassini, 1962; Wessels,1970; Roman et al., 1991; Minno and King, 1994). In themouse, these interactions start as early as 9.5 days p.c. with theoutgrowth from the ventral foregut of a lung bud consisting ofendodermal epithelium surrounded by splanchnic mesoderm.During the pseudoglandular stage (9.5 to 16.0 days ofgestation; Ten Have-Opbroek, 1991), the epithelial tissue pro-liferates and branches to form the prospective bronchial andrespiratory systems. By the beginning of the saccular stage at16.0 days p.c., an abrupt demarcation develops between thecolumnar cells of the prospective bronchial system and thecuboidal cells of the future terminal air spaces. Epithelial-mes-enchymal interactions have been studied in the pseudoglandu-lar stage. Transplantation of distal mesenchyme to an area of

tracheal epithelium denuded of most of its mesenchyme leadsto the formation of supernumerary buds that grow, branch anddifferentiate like normal pulmonary epithelium (Alescio andCassini, 1962; Wessels, 1970; Goldin and Wessells, 1979;Shannon, 1994).

Epithelial-mesenchymal interactions and branching mor-phogenesis in the lung are likely to involve polypeptide sig-nalling molecules and transmembrane receptors known to berequired for the development of other organ systems. Poten-tially important secreted growth and differentiation factorsinclude members of the fibroblast growth factor (FGF), trans-forming growth factor-beta (TGF-β), bone morphogeneticprotein (BMP), and epidermal growth factor (EGF) families,as well as those related to mammalian WNT and hedgehog(HH) proteins (for reviews see Carpenter, 1993; Mason, 1994;Kingsley, 1994; Parr and McMahon, 1994; Ingham, 1995;Hogan, 1995).

Evidence for a crucial role for FGFReceptor-2 (FGFR2) in

1694 S. Bellusci and others

branching morphogenesis has recently been obtained usingtransgenic embryos. FGFR2 is expressed in the epitheliumfrom the early embryonic bud stages through late fetal lungdevelopment. Targeted expression of a dominant negative formof the receptor in the developing epithelium using a surfactantprotein C (SP-C) promoter/enhancer (Wert et al., 1993; Peterset al., 1994) leads to abnormal development; transgenic micehave two undifferentiated epithelial tubes that extend from thebifurcation of the trachea down to the diaphragm. In otherstudies, acidic FGF and basement membrane matrix can inducebranching of lung epithelium in mesenchyme-free cultures(Nogawa and Takaaki, 1995). Taken together, these resultsestablish a role for FGFs in branching morphogenesis,although the precise ligand that interacts with FGFR-2 in vivois not yet known.

Several approaches have been used to investigate the role ofTGF-β1 in lung development. Around 15.5 days p.c. tran-scripts are confined to the mesenchyme, whereas TGF-β1protein is localized to the interface between the epithelial cellsand the mesenchyme, in particular around the bronchiolarducts and in the clefts of the epithelial branches, in associationwith matrix components such as collagen, fibronectin and pro-teoglycan (Lehnert and Akhurst, 1988; Heine et al., 1990;Pelton et al., 1991). In an embryonic lung culture system, TGF-β1 inhibits branching morphogenesis (Serra et al., 1994). Morerecently, it has been shown that overexpression of constitu-tively active TGF-β1 in the distal epithelial cells with the SPCpromoter/enhancer delays embryonic lung development. At18.5 days p.c. transgenic lungs are the same size as controlsbut their development is arrested at the pseudoglandular stage(Zhou et al., 1996).

Transcripts for bone morphogenetic protein-5 (BMP5), aTGF-β-related signalling molecule, are expressed throughoutthe mesenchyme of the embryonic mouse lung (but not aroundthe trachea) from 10.5 days p.c. through at least 16.5 days p.c.(King et al., 1994). BMP5 is encoded by the short ear (se)locus and most mice homozygous for a null mutation are viableand have normal lungs. However, on some genetic back-grounds, newborn homozygous and heterozygous mutantshave fluid-filled cysts in the lungs, although the etiology of thisabnormality is not understood (Green, 1968).

To investigate the role of other signalling molecules in lungdevelopment, we have analyzed the temporal and spatialexpression of genes encoding BMP4 and BMP7, a type 1 BMPreceptor (BMPR/TFR-II/BRK-1), WNT2 and Sonic hedgehog(SHH). From 11.5 days p.c., Shh is expressed throughout theepithelium but with highest levels in the distal tips of epithe-lial buds. Similarly, Bmp-4 expression is high in the epitheliumat the tips of the distal buds, but expression is also seen in theadjacent mesenchyme. This very localized expression of Bmp-4 at lung bud sites led us to examine the effect of overexpres-sion of the gene throughout the distal epithelium of transgeniclungs using the SP-C promoter/enhancer. The resulting trans-genic lungs are significantly smaller than normal and showgreatly distended terminal buds with a reduced proportion oftype II alveolar cells at 16.5 and 18.5 days p.c.. These resultssuggest that BMP4 normally plays a key role in lung develop-ment. A model is presented for the potential interaction ofBMP4 and other signalling molecules in regulating embryoniclung growth and differentiation.

MATERIALS AND METHODS

Plasmid constructionThe SP-C-human Bmp-4 vector was constructed using a human Bmp-4 cDNA (gift from John Wozney, Genetics Institute) modified toreduce the amount of 5′ and to eliminate the 3′ noncoding sequences.The 1232 bp cDNA including only 6 bp of 5′ noncoding sequencewas inserted into a vector containing the 3.7 kb human SP-C promoterregion (Korfhagen et al., 1990). An SV40 small T intron and a 0.4 kbsequence containing a poly(A) addition site with stop codons in allthree reading frames were present at the 3′-end of the cDNA. Theexpression cassette was excised with NdeI and NotI.

A construct containing the SP-C enhancer/promoter and bacteriallacZ gene was generated to monitor the sites of transgenic expression.The lacZ gene with a poly(A) addition sequence was removed frompPD1.27 (Fire et al., 1990) with BamHI and NotI and cloned into aSP-C vector modified by BamHI and NotI digestion to remove theSV40 small T intron and the poly(A) addition fragment. Theexpression cassette was excised by HindIII and NotI.

Generation and identification of transgenic miceTransgenic mice were generated as described by Hogan et al. (1994).DNA was purified by low melting point agarose gel electrophoresis(GIBCO, Life Technologies Inc.) followed by gelase digestion(Epicentre Technology), phenol-chloroform extraction, ethanol pre-cipitation and passage through a Qiaquik column (Qiagen). The trans-genes were injected independently into the pronuclei of(C57BL/6×DBA)F2 mouse eggs at a concentration of 3 ng/µl. Noonof the day of injection is 0.5 days p.c.

The genotype of embryos was determined by PCR analysis ofgenomic DNA extracted from embryonic liver. The primers used were5′-AGGAACAAACAGGCTTCAAA-3′ (SP-C-primer for 5′) and 5′-AATGTTTATACGGTGGAAGC-3′ (human BMP-4 for 3′) or lacZinternal primers (5′-TCTGCTTCAATCAGCGTGCC-3′ for 5′ and 5′-GCCGTCTGAATTTGACCTGA-3′ for 3′). The cDNA was amplifiedfor 27 cycles at 61°C. The genotypes of PCR-positive embryosshowing no abnormal phenotypes were verified by Southern blotanalysis.

Of a total of 106 embryos obtained between 11.5 days p.c. and 19.5days of development after SP-C-Bmp-4 injection of eggs, 28 weretransgenic. The number of transgenics obtained at each stage and thenumber showing an abnormal phenotype (respectively in parentheses)were as follows: 11.5 days p.c. (3;0); 12.5 days p.c. (1;0); 15.5 daysp.c. (3;2); 16.5 days p.c. (10;5); 17.5 days p.c. (2;0); 18.5 days p.c.(8;4) and newborn (1;1).

In situ hybridizationThe whole-mount in situ hybridization protocol was based on oneused previously (Winnier et al., 1995). The following murine cDNAswere used as templates for synthesizing digoxigenin-labeled ribo-probes: 1.5 kb HNF3β cDNA (Sasaki and Hogan, 1993); 630 bpWnt-2 (kindly provided by Dr Andrew McMahon); 642 bp Shh (Dr AndrewMcMahon); 654 bp Bmpr (also known as Brk-1 or TFR-II) (Drs YujiMishina and Naoto Ueno); 1.2 kb mouse Bmp-2 (Blessing et al.,1993); a 1.5 kb full-length mouse Bmp-4 (Winnier et al., 1995) or a1.0 kb mouse Bmp-4 cDNA (Jones et al., 1991). Both Bmp-4 probesgave identical results.

The nonradioactive in situ hybridization of tissue sections wasbased on a protocol used previously (Sasaki and Hogan, 1993).Digoxigenin-labeled antisense RNA probes were prepared for mousesurfactant protein C (758 bp) and rat CC-10 (450 bp) (Dr JeffreyWhitsett). Radioactive in situ hybridization of lung sections wascarried out essentially as described by Zhao et al. (1993).

Analysis of lung proliferation in vivoFemales at 16.5 and 18.5 days of pregnancy were injected intraperi-

1695Role of Bmp-4 in lung development

toneally with a mixture of 5-bromo 2′-deoxyuridine (BrdU) and 5-fluoro 2′-deoxyuridine (FUdr; Sigma) (50 and 10 mg per kg bodyweight, respectively) in 0.5 ml of sterile PBS. After 2 hours, lungswere collected, fixed in 4% paraformaldehyde for 2 hours, washed 3times with PBS, dehydrated in 25%, 50%, 75% and 100% ethanol (5minutes each), embedded in paraffin and sectioned (7 µm). Sectionswere then dewaxed and stained as described previously (Sakai et al.,1994). Briefly, staining of trypsinized sections was carried out with arat anti-BrdU antibody (Accurate Chemical, Westbury, New York)diluted 1:400 in PBS. Sections were then treated with peroxidase-coupled rabbit anti-rat IgG (Vector) diluted 1:100 in PBS using 3-amino,9-ethyl carbazole as the chromogen. All cells in 12 photo-micrographs at magnification of ×40 of random portions of a sectionof a control and transgenic lung were counted and scored as labeledor unlabeled. At 16.5 days p.c., it was possible to distinguish clearlybetween epithelial and mesenchymal cells. At 18.5 days, this distinc-tion was not possible, so that only the overall percentages of labellednuclei were compared between control and transgenic. Paired groupsof data were analysed by Student’s t test using the Systat program.Results were determined to be significant if P<0.05. Similar resultswere obtained when single photomicrographs from 12 differentsections of the same transgenic and control lungs were analyzed.

Analysis of cell deathCell death was initially measured using the ApopTag detection kit(Oncor) following the manufacturer’s instructions. Briefly, sectionswere dewaxed and washed in PBS, followed by proteinase K digestionfor 15 minutes at room temperature. Endogenous peroxidase activitywas then quenched in 2.0% hydrogen peroxide, and the sectionsincubated for 1 hour at 37°C with terminal deoxynucleotide trans-ferase (TdT) and digoxigenin-labelled dUTP. After incubation, incor-porated nucleotide was detected using anti-digoxigenin-peroxidasecoupled antibody with diaminobenzidine (Sigma Fast, SigmaImmunochemicals) as substrate. Sections were counterstained withmethyl green. As a control for specificity, samples were incubatedwithout TdT. Sections of adult mouse testis were used as a positivecontrol. Identical results were obtained with the nick-translationmethod (Gold et al., 1993) using DNA polymerase, digoxigenin-labelled dUTP, rabbit anti-digoxigenin-alkaline phosphatase coupledantibody and NBT substrate from Boehringer Mannheim. To quanti-tate results, all of the cells were counted and scored in 12 photo-micrographs of a section of a single transgenic and control lung at16.5 days as described for BrdU labeling above.

RESULTS

Gene expression in embryonic lung Whole-mount in situ hybridization was used to study geneexpression in embryonic mouse lung at 11.5 days p.c., whenthe organ consists of a single tracheal tube and branchingepithelium, surrounded by a layer of mesenchyme. At thisstage, hepatocyte nuclear factor3B (Hnf-3β), which encodes awinged-helix transcription factor (Sasaki and Hogan, 1993), isexpressed uniformly throughout the endoderm-derived epi-thelium (Fig. 1A). In contrast, Wnt-2 expression is confined tothe mesenchyme around the branching epithelium, as shownpreviously at 16.5 to 18.5 days p.c. (Levay-Young and Navre,1992), and is absent from the tracheal mesenchyme (Fig. 1B).This distribution is confirmed by sectioning after whole-mountin situ hybridization (Fig. 1C).

While HNF3β is expressed uniformly throughout the epi-thelium, this is clearly not the case for Shh, which is expressedat high levels in the distal tips and only weakly in the proximalairways (Fig. 1G). This pattern was confirmed histologically

(Fig. 1H,I). A very similar expression pattern is seen for Bmp-4 but, in this case, transcripts are also present in the mes-enchyme adjacent to the most distal epithelium (Fig. 1D-F).This pattern of Bmp-4 expression continues through 12.5 and15.5 days p.c. (Fig. 3A) but declines by 18.5 days p.c. (datanot shown). Bmp-7 transcripts are seen throughout the epi-thelium and are also increased somewhat at the tips of the buds,but not as dramatically as Shh (Fig. 2A). Bmpr is expressed atlow levels throughout the lung epithelium and in the distal tipmesenchyme (Fig. 2C). Finally, at the sensitivity of thistechnique, no expression at all was seen for Bmp-2 (Fig. 2D).

Expression of SP-C-Bmp-4 in transgenic lungsAs expected from the studies of Wert et al. (1993) using an SP-C-CAT transgene, the SP-C enhancer/promoter drivesexpression of Bmp-4 throughout the distal epithelium at 15.5days p.c, but not in the more proximal airways (Fig. 3B anddata not shown). This was confirmed using an SP-C-lacZreporter transgene (data not shown). Levels of Bmp-4 RNA aresignificantly higher in the distal epithelium of transgenic lungscompared with controls (see legend to Fig. 3). At 16.5 days p.c.,Bmp-4 expression is seen in most distal epithelial cells, but witha somewhat more patchy distribution than a day earlier (Fig.3C). Transgene expression at this time was confirmed using atransgene-specific in situ hybridization probe (Fig. 3D).

Abnormal development of SPC-Bmp-4 transgeniclungsAs shown in Figs 1D-F and 3A, there is normally tight local-ization of endogenous Bmp-4 RNA to the distal tips of theepithelial buds and the adjacent mesenchyme. In contrast, theSP-C enhancer/promoter drives transgene expression in epithe-lial cells throughout the distal airways (Wert et al., 1993; Fig.3 and data not shown). We therefore used this promoter toinvestigate the effect of misexpressing Bmp-4 in a largerdomain of the distal epithelium than normal. At 11.5 and 12.5days p.c., embryonic lungs identified as transgenic by PCR hadno altered phenotype, presumably because the level oftransgene expresssion is low at this time (Wert et al., 1993).However, by 15.5 and 16.5 days, the transgenic lungs are dra-matically affected. First, they are significantly smaller thannormal (Figs 4A, 5A-H) and about half the wet weight (forexample, 6.2 mg compared with 12.9 mg at 15.5 days p.c., 7.3mg compared with 16.0 mg at 16.5 days p.c. and 17.2 mgcompared with 30.1 mg at 18.5 days p.c.). Second, the epi-thelium shows less extensive branching with fewer, grosslydilated, terminal buds separated by abundant mesenchyme (Fig.5A-H). Finally, analysis of adjacent sections through the lungsshowed that the dilated buds are not closed cysts, but are con-tinuous with the bronchi and trachea (for example, Fig. 5F,H).

By 18.5 days, the transgenic phenotype is even more pro-nounced and alveolar development is clearly abnormal (Fig. 5I-L). The reduced branching in the transgenic lungs at this timewas confirmed by radial alveolar counts (see legend to Fig. 5I,J).

One transgenic pup was born but soon died. The lungs werecomposed of grossly dilated, air-filled sacs (Fig. 4B,C) and his-tological analysis showed large dilated sacs, and a dramaticabsence of alveolar septation characteristic of normal newbornlung (Fig. 5M-P). Death was probably due, at least in part, toemphysema secondary to the decreased type II alveocytes anddecreased surfactant levels (see later).


Fig. 1. Gene expression in embryonic mouselung. Whole-mount in situ hybridizationfollowed by sectioning was used to examinegene expression. (A) Hnf-3β is expresseduniformly throughout the epithelium of the 11.5days p.c. lung from the trachea (t) to the terminalbuds (tb). (B) Wnt-2 expression is confined tothe mesenchyme around the terminal buds. Inthis sample, a portion of esophagus obscures thetrachea but, in other cases, no expression wasseen in tracheal mesenchyme. (C) Longitudinalsection through a 13.5 days p.c. lungshowingWnt-2 signal in mesenchyme adjacent tothe terminal buds. (D) Bmp-4 expression ishighest in the terminal buds. Arrowhead notes aterminal bud similar to one sectioned in E and F.(E,F) Section at two magnifications showingBmp-4 expression in the distal epithelium of theterminal bud (ep) as well as in the adjacentmesenchyme (m). (G) Shh expression is highestin the terminal buds but extends proximally.Arrowhead notes a terminal bud similar to theone sectioned in H and I. (H,I) Section showingthat the highest level of Shh expression islocalized to the distal epithelium. Scale bar,A,B,D,E,G,H, 200 µm; C, 100 µm; F,I, 50 µm.

Fig. 2. Expression of Bmp-7, Bmpr and Bmp-2 in 11.5 days p.c. embryonicmouse lung. (A) Bmp-7 is expressed uniformly throughout the epitheliumwith increased signal in the terminal buds. (B) Sense Bmp-7 probe showsno signal. (C) Bmpr expression is mostly epithelial with some signal in themesenchyme of the terminal buds. (D) Bmp-2 is not expressed in theembryonic lung at levels that can be detected by whole-mount in situhybridization. Scale bar, A-D, 160 µm.

A B C

D E F

G H I

HNF-3β Wnt-2 Wnt-2

Bmp-4 Bmp-4 Bmp-4

Shh Shh Shh

ep

ep

m

m

F

E

H I

tb

t

B

D

A

CBmp-7 SBmp-7 AS

Bmpr AS Bmp-2 AS

Bmp-4 overexpression inhibits lung epithelial cellproliferation and induces cell death in themesenchymeTransgenic lungs are about half the size and wet weight onormal (Figs 4, 5). Cell proliferation was therefore examineby exposing 16.5 and 18.5 days p.c. embryos in utero to BrdUand FUdr for 2 hours. At 16.5 days p.c., the number of nuclein the transgenic epithelium that had incorporated labellenucleotide is reduced compared to normal (26% versus 34%of cells labelled, P=0.032 by Student’s t-test) (Figs 6A,B, 7AIn contrast, when the mesenchymal cells are compared,there is a small but statistically significant increase in theproportion of nuclei labelled in the transgenic lungcompared to normal (15% versus 11%, P=0.029 byStudent’s t-test) (Fig. 7B).

At 18.5 days p.c., it was not possible to distinguishunambiguously between epithelial and mesenchymalcells, due to the flattened shape of the epithelial cells andtheir close apposition to the underlying mesenchyme(Fig. 6C,D). However, considering the cell population asa whole, there was a statistically significant increase inthe proportion of cells labelled in the transgenic lungscompared with the controls (12% versus 6%, P<0.001 byStudent’s t-test) (Fig. 7C).

Since transgenic lungs are significantly smaller thannormal at 16.5 days p.c., even though BrdU labelling isincreased in the mesenchyme, we investigated cell death,using both the terminal deoxyribonucleotide transferase-mediated digoxigenin-dUTP end labelling (TUNEL)technique, and the in situ nick translation (ISNT) method.Both gave essentially identical results, but the ISNTmethod allowed the number of positive cells to be quanti-tated (compare Fig. 6E,F with G,H). In a control lung at16.5 days p.c., only 3 positive cells were seen in a total of7,983 (approximately 0.04% apoptotic cells). By contrast,

fd

id

).

in a lung from a transgenic littermate, 76/6,823 positive cellswere seen (approximately 1%).

Expression of SP-C and CC-10 in transgenic lungsTo evaluate epithelial cell differentiation, in situ hybridizationwas carried out using riboprobes for two marker genes, SP-Cand CC-10. SP-C is normally expressed in all distal epithelialcells from 10.5 days p.c. (Wert et al., 1993) but as developmentproceeds transcription is gradually switched off in the precur-sors of type I cells and expression is confined to the mature type


Fig. 3. Bmp-4 expression in control andtransgenic lung as revealed by in situhybridization of sections with eitherdigoxigenin-labelled antisense mouse Bmp-4riboprobe (A-C) or antisense small T 3′untranslated region probe (D). (A) In control15.5 days p.c. lung, Bmp-4 is expressed inlocalized regions of the distal epithelium,presumably corresponding to the distal tips.Color reaction time 40 hours. With non-radioactive in situ hybridization, expressionin the mesenchyme cannot be detectedefficiently. (B) High levels of Bmp-4expression are seen throughout the distalepithelium of transgenic 15.5 days p.c. lung.Color reaction time 16 hours. (C) In the 16.5days p.c. transgenic lung, high levels ofBmp-4 RNA are seen in the distalepithelium, with a rather more patchydistribution than in B. Color reaction time 16hours. (D) Expression of transgene specificRNA at 16.5 days p.c. Color reaction time 16hours. Scale bar, A,B, 200 µm; C,D, 50 µm.

A B

C D

15.5 15.5

16.5 16.5

C5 days

wborn

newborn

C

of SP-C-Bmp-4 transgenic and control lungs. (A) Transgenic lung at) is smaller than control, with grossly dilated terminal buds. (B) Thelung (right) has more exaggerated cystic changes than at 15.5 days p.c.enclosed by the red box is magnified in C. (C) Dramatic dilations ofthelium in newborn lung are filled with air.

II cells in the alveolar sacs. CC-10 expression is not apparentuntil 17.5 days p.c (Singh et al., 1993) and is seen only in theepithelial Clara cells that line the terminal bronchioles.

At 16.5 days p.c., SP-C is expressed in most of the epithelialcells of the normal lung buds (Fig. 8A,B). By contrast, in thetransgenic lung fewer cells in the grossly dilated terminal budsexpress SP-C (Fig. 8C,D). At 18.5 days, the same pattern ofreduced SP-C expression is noted in the transgenic lungcompared with the control (Fig. 8 E,F compared with G,H).

Because CC-10 expression is not detected in normal lunguntil 17.5 days p.c., the pattern of expression of CC-10 wasonly compared at 18.5 days p.c. betweentransgenic and control. There appears to beno difference in expression between thetwo samples (Fig. 8I,J).

Expression of HNF3β, Shh, Bmp-7andWnt-2 in transgenic lungsTo determine if misexpression of Bmp-4altered the expression of genes encodingother regulatory molecules, sections ofcontrol and transgenic lungs at 15.5 days p.c.were hybridized with riboprobes for HNF3β,Shh, Wnt-2 and Bmp-7. Initial studies usingdigoxigenin-labeled riboprobes showed nochange in expression of these genes in trans-genic lungs. To increase the sensitivity of theanalysis, 35S-labeled riboprobes were thenused, in conjunction with autoradiography.However, even at the higher sensitivity ofthis technique, no change was seen in thelevel of HNF3β, Shh, Wnt-2 or Bmp-7 tran-scripts in transgenic lungs compared withcontrols (data not shown). As expected,using this technique, transgenic lungs didshow an increase in hybridization with themouse Bmp-4 riboprobe in the distal epi-thelium due to cross hybridization with tran-

A

B

15.

ne

Fig. 4. Comparison 15.5 days p.c. (rightnewborn transgenic The portion of lung distal respiratory epi

scripts from the human Bmp-4 transgene. However, no signifi-cant increase was seen in the level of endogenous RNA in themesenchyme of the transgenic lung.

DISCUSSION

Expression of Hnf-3β, Bmps and Shh during mouselung developmentHere, we report the expression in embryonic mouse lung ofgenes encoding signalling molecules known to play importantroles in the morphogenesis of other organs. Of particular interest


A B C D

E F G H

I J K L

M N O P

15.5 15.5 15.5 15.5

16.5 16.5 16.5 16.5

18.5 18.5 18.5 18.5

newborn newborn newborn newborn

dede

de

de

mm

m

m

tr

br

a

a

a

a

tr

br

brtr

bl bv

bv

bvbl

br

Fig. 5. Comparison of SP-C-Bmp-4 transgenic and control lungs at 15.5, 16.5, 18.5 days p.c. andnewborn. (A) Section of 15.5 days p.c. control lung. (B) Section of 15.5 days p.c. transgenic lung.(C) Magnification of boxed portion of control lung in A shows multiple, small terminal buds and distalepithelium (de) among mesenchymal (m) cells. (D) Magnification of boxed portion of transgenic lung inB shows dilated airways lined by distal epithelium. The mesenchyme appears thicker than control lung.(E) Section of control lung at 16.5 days p.c. reveals portion of trachea and bronchus. (F) Section oftransgenic lung at 16.5 days p.c. shows trachea and its connection with a primary bronchus.(G) Magnification of boxed portion of control lung in E shows terminal buds lined by distal epitheliumand surrounded by mesenchyme. (H) Magnification of boxed portion of transgenic lung in F showsdistal, dilated secondary bronchus connecting with dilated terminal buds, and thicker mesenchyme.(I) Section of 18.5 days p.c. control lung with labelled bronchiole (bl) and blood vessel (bv). Themesenchyme has thinned and septation appears to be progressing. In this section, ten terminalbronchioles are distinguishable. The number of alveoli crossed by a perpendicular line dropped fromeach terminal bronchiole to the nearest septum or parenchymal edge (the radial alveolar count) is7,6,2,5,2,2,6,3,2,3 (average 3.8). (J) Section of 18.5 days p.c. transgenic lung showing continuitybetween trachea and primary bronchus. In this section, five terminal bronchioles are present and theradial alveolar count is 2,3,1,2,4 (average 2.4). (K) Magnification of boxed portion of control lung in Ishows developing alveoli (a). (L) Magnification of boxed portion of transgenic lung in J showsabnormal alveolar development. (M) Section of newborn control lung shows a bronchiole, blood vesseland normal septation. (N) Section of newborn transgenic lung shows abnormal septation.(O) Magnification of control lung in M shows normal septation. (P) Magnification of transgenic lung inN shows abnormal septation. Scale bar, A,B,E,F,I,J,M,N, 500 µm; C,D,G,H,O,P, 100 µm; K,L, 50 µm.

is the finding that Bmp-4 andShh are both expressed at highlevels in the distal epithelium ofthe terminal buds, apparently invery similar or overlappingdomains (Fig. 1). This localiza-tion is not an artifact due toprobe trapping. Under identicalconditions, Hnf-3β antisenseRNA gave uniform labelling ofepithelial cells (Fig. 1A), whileantisense Bmp-2 and senseBmp-7 and Bmp-4 probes gaveno hybridization above back-ground (Fig. 2B,D and data notshown). In addition, sectioningof the lungs after whole-mounthybridization confirmed thelocalization of transcripts seenin the intact organ (Fig. 1F,I).Although Bmp-4 and Shh areboth expressed at high levels inthe distal epithelium, some sig-nificant differences in theoverall patterns were seen. Inparticular, Bmp-4 transcripts arealso present in the adjacent mes-enchyme, while Shh RNAappears to be confined to theepithelial cells.

Bmp-2, which is closelyrelated to Bmp-4, is not tran-scribed in the embryonic lung,at least at a level detectable bywhole-mount in situ hybridiz-ation (Fig. 2D). Bmp-7, incontrast, is expressed through-out the epithelium, with someincrease in the terminal buds(Fig. 2A). This raises the possi-bility that Bmp-4 and Bmp-7 areco-expressed in the distal tips.Bmp-4 null mutant embryos diebefore the formation of lungprimordia (Winnier et al., 1995)but no defects have been seenin the lungs of homozygousnull Bmp-7 embryos, whichsurvive to birth (A. Dudley andE. J. Robertson, personal com-munication) suggesting thatany function of Bmp-7 homo-or heterodimers in lung devel-opment can be compensated forby other proteins.

The effect ofmisexpressing Bmp-4 intransgenic mouse lungsTo explore the role of BMP4 inlung development, we generated


B

F

H

D

epep

m

16.5

16.5

16.5 16.5

16.5

16.5

18.5 18.5

and cell death in transgenic and control lungs. (A-D) Pregnant femalesneally with BrdU and FUdr, and 2 hours later embryonic lungs were

histology, and the proportion of labelled cells determined as described. days p.c.. As shown graphically in Fig. 5A, a higher proportion of theled compared to the transgenic. (B) Transgenic lung at 16.5 days p.c.ith less BrdU labelling of epithelial cells. However, the proportion oflls is increased compared with the control, as shown graphically in Fig.18.5 days p.c. A smaller proportion of cells are labelled compared to thephically in Fig. 5C. (D) Transgenic lung at 18.5 days p.c. with more the control, as shown graphically in Fig. 5C. (E-H) Cell death in

ung at 16.5 days p.c. (E) Section of normal lung showing a single cellethod (arrow). (F) Transgenic lung showing an increase in the number

lls in the mesenchyme (arrows). (G) Normal lung with a single apototicT method (H) Transgenic lung showing individual apoptotic cells. Scale4 µm.

transgenic embryos in which human Bmp-4 is expressed through-out the distal epithelium under the control of the SP-Cpromoter/enhancer (Fig. 3). No abnormalities were seen in trans-genic lungs at 11.5 and 12.5 days p.c. However, at 15.5 and 16.5days p.c., Bmp-4 misexpression leads toa dramatic effect on lung development(Figs 4A, 5A-H). In particular, trans-genic lungs are smaller than normal, areabout half the wet weight (see text) andhave fewer, greatly distended, epithelialterminal buds separated by abundantmesenchyme. By 18.5 days, the lunglobes contain huge, epithelial sacsapparently continuous with the bronchiand trachea and separated by mes-enchyme (Fig. 5I-L). At birth, these sacsfill with air (Fig. 4C), but they show adramatic lack of normal alveolar seg-mentation (Fig. 5M-P) and do notsustain normal lung function.

The fact that transgenic lungs aresmaller than normal raises the possibil-ity that overexpression of Bmp-4 inhibitscell proliferation and/or promotes celldeath. Indeed, both mechanisms doappear to be operating in transgeniclungs, suggesting that the effect of theligand in vivo is also complex. BrdUlabelling experiments at the pseudoglan-dular stage showed a 24% decrease inthe labelling of transgenic epithelium,but a 26% increase in labelling of themesenchyme compared with controls.At 18.5 days, there was a 49% increasein the labelling of cells in transgeniclungs, although it was not possible todistinguish between epithelial and mes-enchymal cells at this stage. At 16.5days, analysis of lung sections usingboth the TUNEL and ISNT techniquesrevealed a very low level of cell death inthe normal lung (approximately 0.04%),but a large increase in the number oflabelled cells in the mesenchyme oftransgenic lungs (to approximately1.0%).

There are several possible, non-exclusive explanations for thesefindings. One is that ectopicallyexpressed BMP4 has different effectson epithelial versus mesenchymalcells, inhibiting the proliferation anddifferentiation of the former, and stim-ulating either proliferation or celldeath in the latter. To account for theeffect on mesenchyme, it is possiblethat there are two different populationsof cells, that respond differently toBMP4. For example, endothelial cellsmay proliferate and interstitial orstromal cells may die in response to

A

C

E

G

m

Fig. 6. Cell proliferationwere injected intraperitocollected, processed for(A) Normal lung at 16.5epithelial cells are labelshows dilated airways wlabelled mesenchyme ce5B. (C) Normal lung at transgenic, as shown gralabelling of cells than innormal and transgenic lstained by the TUNEL mof individual labelled cecell revealed by the ISNbar, A-D, 40 µm; E-H, 2

the ligand. Alternatively, different levels of BMP4 may inducedifferent responses in the mesenchyme, so that cells close tothe epithelium (the source of ectopic BMP4) undergo cell deathand those further away are stimulated to proliferate. Further


studies will be needed to distinguish between these models.Meanwhile, it should be noted that BMP4 has been observedto promote apoptosis in neural crest cells from rhombomeres3 and 5 in the chick embryo hindbrain (Graham et al., 1994).

Does BMP4 affect cell differentiation in the embryonic lung?In situ hybridization reveals a reduction in the number of cellsexpressing SP-C in transgenic lungs versus controls, but nochange in the pattern of expression of CC-10 (Fig. 8). This resultshows that Bmp-4 overexpression has no effect on the differen-tiation of the CC-10-expressing epithelial Clara cells in the bron-chioles and proximal airways, but does appear to affect the differ-entiation of the distal epithelial cells. Again, there are several, notmutually exclusive, mechanisms that may underlie this phenom-

Fig. 7. Graphic representation of BrdU labelling. (A) Total andlabelled epithelial cells counted in photomicrographs of sections ofcontrol and transgenic lungs at 16.5 days p.c. (B) Total and labelledmesenchymal cells at 16.5 days p.c. (C) Total and labelled cells at18.5 days p.c. Samples were analysed as described in Materials andMethods and P values are for the Student’s t-test.

enon. During the transition from the pseudoglandular to saccularstage of development, the SP-C-expressing, cuboidal, airwayprecursor cells begin to differentiate into either type I cells, whichdo not transcribe SP-C, or into mature type II cells, whichupregulate and maintain SP-C expression (Ten Have-Opbroek,1991; Wert et al., 1993). In the transgenic lung, BMP4 may eitheractively promote the differentiation of primordial cells into typeI rather than type II cells, or selectively inhibit the terminal differ-entiation or survival of mature type II cells that maintain highlevels of SP-C expression. Resolution between these models mayrequire the identification of specific markers for type I cells.

Finally, misexpression of Bmp-4 throughout the distal epi-thelium in transgenic lungs has a dramatic effect on branchingmorphogenesis after about 15.5 days, resulting in the formationof fewer, grossly dilated terminal sacs compared with controllungs (Fig. 5). The fact that Bmp-4 overexpression inhibitsepithelial cell proliferation raises the possibility that, duringnormal development, when Bmp-4 is expressed in a veryrestricted domain, high local levels of the protein accumulate atthe tips of branches and gradually suppress distal epithelialgrowth. Any new outgrowth can then only occur from regionsproximal and lateral to the distal tips, resulting in dichotomousbranching. Unfortunately, no phenotype was observed in SP-Ctransgenic lungs at the early pseudoglandular stage (e.g. 11.5-12.5 days), when the pattern of branching morphogenesis canbe observed clearly. This may reflect a low level of BMP4expression from the transgene at this time, so further experi-ments will be needed to test the hypothesis that BMP4 specifi-cally affects dichotomous branching morphogenesis in vivo.Meanwhile, it should be noted that, in the Drosophila larva, dppexpression in the anterior foregut epithelium regulates theformation of the proventriculus. In dpp null mutants, proven-tricular development is abnormal and, in some cases, ectopicoutgrowths appear in the oesophagus, consistent with a normalrole for the protein in suppressing local endodermal cellmovement and morphogenesis (Pankratz and Hoch, 1995).

The phenotypic effect of overexpressing BMP4 in the devel-oping mouse lung is significantly different from that of over-expressing TGF-β1, utilizing the same SP-C vector. Lungstransgenic for SP-C-TGF-β1 are the same size as controls at16.5 days p.c. (Zhou et al., 1996). However, development fromthe pseudoglandular to canalicular and saccular stage isblocked, and in situ hybridization demonstrates a reduction inthe level of both SP-C and CC-10 gene expression.

A model for the interaction between BMP4 and otherpolypeptide signalling molecules in lungmorphogenesisThe patterns of expression of Bmp-4, Shh, Bmpr and Wnt-2(Fig. 1) suggest a model for the role of these genes in lung mor-phogenesis (Fig. 9). First, we propose that SHH proteinsecreted by the distal tip epithelium induces Bmp-4 expressionin the adjacent mesenchyme. This idea is based on the obser-vation that, in Drosophila, the Shh homolog, hedgehog (hh),induces the expression of the Bmp-2/4 homolog, decapenta-plegic (dpp) in adjacent cells (Herberlein et al., 1993; Diaz-Benjumea et al., 1994; Ingham, 1995, for review). The hh geneencodes a secreted glycoprotein which exists in bothmembrane-associated and soluble forms, and it is believed thatthe protein interacts with target cells via a membrane-associ-ated protein, patched (ptc) (Ingham et al., 1991). A mouse ptc


A B C D

E F G H

I J

m

m

br

brbr

de

de

Fig. 8. Expression of SP-C andCC-10 in control and transgeniclungs and lungs. In situhybridization of tissue sectionswas performed with digoxigenin-labeled antisense RNA probes.(A,B) Sections at twomagnifications of control lung at15.5 days p.c. showingexpression for SP-C in mostdistal epithelial cells (de) but notin bronchioles (br) ormesenchyme (m). (C,D) Sectionsat two magnifications oftransgenic lung at 16.5 days p.c.showing grossly distendedterminal airways with fewer SP-C-expressing cells than normal.(E,F) Section at twomagnifications of peripheralregion of a control lung at 18.5days p.c. showing extensiveexpression of SP-C. Arrow marks the edge of the lung. (G,H) Section at two magnifications of a peripheral region of a transgenic lung at 18.5days p.c. showing distorted septation with many fewer SP-C-expressing cells. (I) Section of control lung at 18.5 days p.c. with normalexpression of CC-10 in Clara cells which line the terminal bronchioles. (J) Section of transgenic lung at 18.5 days p.c. also shows CC-10expression in the terminal bronchioles. Scale bar, A,B,C,G, 80 µm; B,F,D,H, 40 µm; I,J, 160 µm.

homolog has recently been cloned and is expressed in the mes-enchyme of the lung (L. V. Goodrich, R. L. Johnson, L.Milenkovic and M. Scott, personal communication). Thissupports the idea that SHH secreted by the distal epitheliumactivates Bmp-4 in the adjacent mesenchyme. However,induction by SHH of Bmp-4 in epithelial cells independent ofPTC cannot be ruled out. While the evidence that hh regulatesdpp expression in Drosophila embryos is very strong, the dataconcerning Bmp expression in response to SHH in vertebrateembryos is still quite minimal. However, in the chick embryo,

Fig. 9. Model for the inter-relationship of Shh, Bmp-4 and Wnt-2 inthe embryonic mouse lung. For details, see text. It is proposed thatSHH protein secreted (red arrow) by distal tip epithelial cells inducesthe expression of Bmp-4 and/or Wnt-2 in the adjacent mesenchyme.Bmp-4 expression in the distal epithelium may also be regulated bySHH. Alternatively, or in addition, expression of BMP4 in theepithelium may be autoinduced by protein made in the mesenchyme(blue arrows). WNT2 made by the mesenchyme (green arrows) mayalso affect gene expression in the epithelium.

Bmp-2 is induced in limb bud anterior mesenchyme in responseto ectopic SHH (Laufer et al., 1994) and ectopic SHH inducesBmp-4 in hindgut mesenchyme (Roberts et al., 1995).

In the Drosophila embryo, hh also regulates the expressionof wingless (wg), which encodes a secreted glycoprotein relatedto the Wnt family of signalling molecules (Diaz-Benjumea etal., 1994). In addition, wg maintains hh expression (Lee et al.,1992). In the developing vertebrate limb bud, simultaneousWnt-7a and Fgf4 expression are required to maintain Shhexpression in the posterior mesenchyme expression (Yang andNiswander, 1995). This raises the possibility that SHH in thedistal epithelium of the lung induces Wnt-2 in the surroundingmesenchyme, and that WNT2 (and possibly an FGF familymember) is required to maintain Shh expression.

Finally, BMP4 protein made in the distal mesenchyme couldact on the distal epithelium to activate Bmp-4 expression. Thiswould establish an autoregulatory circuit similar to thatproposed in the developing mouse tooth bud (Vainio et al.,1993). In order for BMP4 to induce its own expression, it isnecessary that the target cells express appropriate receptors.We have shown here that the lung epithelium and distal mes-enchyme express a Type I transmembrane serine-threoninekinase receptor which can bind both Bmp-2 and Bmp-4(Bmpr/Trk-11/Brk-1) (Koenig et al., 1994). However, it is verylikely that other Type I and Type II receptors that bind BMPsare also expressed. This, and other aspects of the model, arecurrently under investigation.

We thank Dr Jeffrey Whitsett for the SP-C vector, for in situ hybrid-ization probes and for generous advice and encouragement, TerryJohnson, Division of Neonatology, Vanderbilt Medical School, andMartin Offield, Department of Cell Biology, for help with graphics, DrYu Shyr, Department of Preventive Medicine for statistical analysis, andDr Lillian Nanney and the Tissue Analysis Core laboratory (suppportedby AR41943) for analysis of BrdU labelling. We also thank Drs RobertCotton, Christopher Wright, Rosa Serra and laboratory colleagues forcritical comments on the manuscript and Linda Hargett, Julie Blackwell,


Yolanda McClain and Lorene Batts for skilled technical assistance. S.B. acknowledges support from The Philippe Foundation. This work wassupported by NIH grants HD28955 and ST32HL07256. Brigid Hoganis an Investigator of the Howard Hughes Medical Institute.

REFERENCES

Alescio, T. and Cassini, A. (1962). Induction in vitro of tracheal buds bypulmonary mesenchyme grafted on tracheal epithelium. J. Exp. Zool. 150,83-94.

Blessing, M., Nanney, L. B., King, L. E., Jones, C. M. and Hogan, B. L. M.(1993). Transgenic mice as a model to study the role of TGF-beta-relatedmolecules in hair follicles. Genes Dev. 7, 204-215.

Carpenter, G. (1993). EGF : new tricks for an old growth factor. CurrentOpinion in Cell Biol. 5, 261-264.

Diaz-Benjumea, F. J., Cohen, B. and Cohen, S. M. (1994). Cell interactionbetween compartments establishes the proximal-distal axis of Drosophilalegs. Nature 372, 175-179.

Fire, A., White Harrison, S. and Dixon, D. (1990). A modular set of lacZfusion vectors for studying gene expression in C. elegans. Gene 93, 189-198.

Gold, R., Schmied, M., Rothe, G., Zischler, H., Breitschopf, H., Wekerle,H. and Lassmann, H. (1993) Detection of DNA fragmentation in apoptosis:application of in situ nick translation to cell culture systems and tissuesections. J. Histochem. Cytochem. 41, 1023-1030.

Goldin, G. V. and Wessels, N. K. (1979). Mammalian lung development: thepossible role of cell proliferation in the formation of supernumerary trachealbuds and in branching morphogenesis. J. Exp. Zool. 208, 337-346.

Graham, A., Francis-West, P., Brickell, P. and Lumsden, A. (1994). Thesignalling molecule BMP4 mediates apoptosis in the rhomobocephalicneural crest. Nature 372, 684-686.

Green, M. C. (1968). Mechanism of the pleiotropic effects of the short earmutant gene in the mouse. J. Exp. Zool. 167, 129-150.

Heine, U. I., Munoz, E. F., Flanders, K. C., Roberts, A. B. and Sporn, M. B.(1990). Colocalization of TGF-beta 1 and collagen I and III, fibronectin andglycosaminoglycans during lung branching morphogenesis. Development109, 29-36.

Herberlein, U., Wolff, T. and Rubin, G. M. (1993). The TGFβ homolog dppand the segment polarity gene hedgehog are required for propagation of amorphogenetic wave in Drosophila retina. Cell 75, 913-926.

Hogan, B. L. M. (1995). The TGFβ-related signalling system in mousedevelopment. Seminar in Dev Biol 6, 257-265.

Hogan, B. L. M., Beddington, R., Constantini, F. and Lacy, E. (1994).Manipulating the Mouse Embryo: A Laboratory Manual. Cold SpringHarbor, New York: Cold Spring Harbor Publications.

Ingham, P.W. (1995). Signalling by hedgehog family proteins in Drosophilaand vertebrate development Curr. Op. Genetics Dev. 5, 492-498.

Ingham, P. W., Taylor, A. M. and Nakano, Y. (1991). Role of the Drosophilapatched gene in positional signalling. Nature 353, 184-187.

Jones, C.M., Lyons, K.M. and Hogan, B.L.M. (1991). Involvement of BoneMorphogenetic Protein-4 (BMP-4) and Vgr-1 in morphogenesis andneurogenesis in the mouse. Development 111, 531-542.

Korfhagen, T., Glasser, S., Wert, S., Bruno, M., Daugherty, C., McNeish,J., Stock, J., Potter, S. and Whitsett, J. (1990). Cis-acting sequences froma human surfactant protein gene confer pulmonary-specific gene expressionin transgenic mice. Proc. Natl. Acad. Sci. USA 87, 6122-6126.

King, J. A., Marker, P. C., Seung, K. J. and Kingsley, D. M. (1994). BMP5and the molecular, skeletal, and soft-tissue alterations in short ear mice. Dev.Biol.166, 112-122.

Kingsley, D. M. (1994). What do BMPs do in mammals? Clues from the mouseshort ear mutation. Trends in Genetics 10, 16-22.

Koenig, B. B., Cook, J. S., Wosling, D. H., Ting, J., Tiesman, J. P., Correa,P. E., Olson, C. A., Pecquet, A. L., Ventura, F., Grant, R. A., Chen, G.,Wrana, J. L., Massague, J. and Rosenbaum, J. S. (1994). Characterizationand cloning of a receptor for Bmp-2 and Bmp-4 from NIH 3T3 cells. Mol.Cell. Biol. 14, 5961-5974.

Laufer, E., Nelson, C. E., Johnson, R. L., Morgan, B. A. and Tabin, C.(1994). Sonic hedgehog and Fgf-4 act through a signaling cascade andfeedback loop to integrate growth and patterning of the developing limb bud.Cell 79, 993-1003.

Lee, J. J., von Kessler, D. P., Parks, S. and Beachy, P. A. (1992). Secretionand localized transcription suggest a role in positional signaling for productsof the segmentation gene hedgehog. Cell 71, 33-50.

Lehnert, S. A. and Akhurst, R. J. (1988). Embryonic expression pattern of

TGF beta type-1 RNA suggests both paracrine and autocrine mechanisms ofaction. Development 104, 263-273.

Levay-Young, B. K. and Navre, M. (1992). Growth and developmentalregulation of wnt-2 (irp) gene in mesenchymal cells of fetal lung. Am. J.Physiol. 262, 672-683.

Mason, I. J. (1994). The Ins and Outs of fibroblast growth factors. Cell 78, 547-552.

Minno, P. and King, R. J. (1994). Epithelial-mesenchymal interactions in lungdevelopment. Ann. Rev. Physiol. 56, 13-45.

Nogawa, H. and Takaaki, I. (1995). Branching morphogenesis of embryonicmouse lung epithelium in mesenchyme-free culture. Development 121,1015-1022.

Pankratz, M. and Hoch, M. (1995). Control of epithelial morphogenesis bycell signaling and integrin molecules in the Drosophila foregut. Development121, 1885-1898.

Parr, B. A. and McMahon, A. P. (1994). Wnt genes and vertebratedevelopment. Current Opinion Genetics and Development 4, 523-528.

Pelton, R. W., Saxena, B., Jones, M., Moses, H. L. and Gold, L. I. (1991).Immunohistochemical localization of TGF-β1, TGF-β2, and TGF-β3 in themouse embryo: expresion patterns suggest multiple roles during embryonicdevelopment. J. Cell Biol. 115, 1091-1105.

Peters, K., Werner, S., Liao, X., Wert, S., Whitsett, J. and Williams, L.(1994). Targeted expression of a dominant negative FGF receptor blocksbranching morphogenesis and epithelial differentiation of the mouse lung.EMBO J. 13, 3296-3301.

Roberts, D.J., Johnson, R.L., Burke, A.C., Nelson, C.E., Morgan, B.A. andTabin, C. (1995) Sonic hedgehog is an endodermal signal inducing Bmp-4and Hox genes during induction and regionalization of the chick hindgut.Development 121, 3163-3174.

Roman, J., Little, C. W. and McDonald, J. A. (1991). Potential role of RGD-binding integrin in mammalian lung branching morphogenesis. Development112, 551-558.

Sakai, Y., Nelson, K. G., Snedeter, S., Bossert, N. L., Walker, M. P.,McMachon, J. and DiAugustine, R. P. (1994). Expression of epidermalgrowth factor in suprabasal cells of stratified squamous epithelia:implications for a role in differentiation. Cell Growth Diff. 5, 527-535.

Sasaki, H. and Hogan, B. L. M. (1993). Differential expression of multiplefork head related genes during gastrulation and axial pattern formation in themouse embryo. Development 118, 47-59.

Serra, R., Pelton, R. W. and Moses, H. L. (1994). TGFβ1 inhibits branchingmorphogenesis and n-myc expression in lung bud organ cultures.Development 120, 2153-2161.

Shannon, J. M. (1994). Induction of alveolar type II cell differentiation in fetaltracheal epithelium by grafted distal lung mesenchyme. Dev.Biol. 166, 600-614.

Singh, G., Katyal, S. L., Brown, W. E. and Kennedy, A. L. (1993). Mouseclara cell 10-kDa (CC10) protein: cDNA nucleotide sequence and molecularbasis of variation in progesterone binding of CC10 from different species.Exp. Lung Research 19, 67-75.

Ten Have-Opbroek, A. A. W. (1991). Lung development in the mouseembryo. Exp. Lung Research 17, 111-130.

Vanio, S., Karavanova, I., Jowett, A. and Thesleff, I. (1993). Identification ofBMP-4 as a signal mediating secondary induction between epithelial andmesenchymal tissues during early tooth development. Cell 75, 45-58.

Wert, S., Glasser, S. W., Korfhagen, T. R. and Whitsett, J. A. (1993).Transcriptional elements from the human SP-C gene direct expression in theprimordial respiratory epithelium of transgenic mice. Dev. Biol. 156, 426-443.

Wessels, N. K. (1970). Mammalian lung development: interaction in formationand morphogenesis of tracheal buds. J. Exp. Zool. 175, 455-466.

Winnier, G., Blessing, M., Labosky, P. A. and Hogan, B. (1995). BoneMorphogenetic protein-4 (BMP-4) is required for mesoderm formation andpatterning in the mouse. Genes Dev. 9, 2105-2116.

Yang, Y. and Niswander, L. (1995). Interaction between the signalingmolecules Wnt-7a and SHH during vertebrate limb development: dorsalsignals regulate anteroposterior patterning. Cell 80, 939-947.

Zhao, G-Q., Zhou, X., Eberspaecher, H., Solursh, M. and de Crombrugghe,B. (1993). Cartilage homeoprotein 1, a homeoprotein selectively expressed inchondrocytes. Proc. Natl. Acad. Sci. USA 90, 8633-8637.

Zhou, L., Dey, C. R., Wert, S. E. and Whitsett, J. A. (1996) Arrested lungmorphogenesis in transgenic mice bearing an SP-C-TGF-β1 chimeric gene.Dev. Biol. (In press)

(Accepted 6 March 1996)

(Bmp-4) plays a role in mouse embry

Documents

Transcript of (Bmp-4) plays a role in mouse embry