The molecular basis of lung morphogenesis

Mechanisms of Development 92 (2000) 55±81 www.elsevier.com/locate/modo

The molecular basis of lung morphogenesis David Warburton a,*, Margaret Schwarz b, Denise Tefft c, Guillermo Flores-Delgado a, Kathryn D. Anderson a, Wellington V. Cardoso d a

Department of Surgery, The Developmental Biology Program, University of Southern California Keck School of Medicine and School of Dentistry, Los Angeles, CA, USA b Department of Cardiothoracic Surgery, Children's Hospital Los Angeles Research Institute, University of Southern California Keck School of Medicine, Los Angeles, CA, USA c The Center for Craniofacial Molecular Biology, University of Southern California, School of Dentistry, Los Angeles, CA, USA d The Pulmonary Center, Boston University School of Medicine, Boston, MA, USA Received 12 July 1999; received in revised form 13 October 1999; accepted 13 October 1999

Abstract To form a diffusible interface large enough to conduct respiratory gas exchange with the circulation, the lung endoderm undergoes extensive branching morphogenesis and alveolization, coupled with angiogenesis and vasculogenesis. It is becoming clear that many of the key factors determining the process of branching morphogenesis, particularly of the respiratory organs, are highly conserved through evolution. Synthesis of information from null mutations in Drosophila and mouse indicates that members of the sonic hedgehog/patched/ smoothened/Gli/FGF/FGFR/sprouty pathway are functionally conserved and extremely important in determining respiratory organogenesis through mesenchymal±epithelial inductive signaling, which induces epithelial proliferation, chemotaxis and organ-speci®c gene expression. Transcriptional factors including Nkx2.1, HNF family forkhead homologues, GATA family zinc ®nger factors, pou and hox, helix-loop-helix (HLH) factors, Id factors, glucocorticoid and retinoic acid receptors mediate and integrate the developmental genetic instruction of lung morphogenesis and cell lineage determination. Signaling by the IGF, EGF and TGF-b/BMP pathways, extracellular matrix components and integrin signaling pathways also directs lung morphogenesis as well as proximo-distal lung epithelial cell lineage differentiation. Soluble factors secreted by lung mesenchyme comprise a `compleat' inducer of lung morphogenesis. In general, peptide growth factors signaling through cognate receptors with tyrosine kinase intracellular signaling domains such as FGFR, EGFR, IGFR, PDGFR and c-met stimulate lung morphogenesis. On the other hand, cognate receptors with serine/threonine kinase intracellular signaling domains, such as the TGF-b receptor family are inhibitory, although BMP4 and BMPR also play key inductive roles. Pulmonary neuroendocrine cells differentiate earliest in gestation from among multipotential lung epithelial cells. MASH1 null mutant mice do not develop PNE cells. Proximal and distal airway epithelial phenotypes differentiate under distinct transcriptional control mechanisms. It is becoming clear that angiogenesis and vasculogenesis of the pulmonary circulation and capillary network are closely linked with and may be necessary for lung epithelial morphogenesis. Like epithelial morphogenesis, pulmonary vascularization is subject to a ®ne balance between positive and negative factors. Angiogenic and vasculogenic factors include VEGF, which signals through cognate receptors ¯k and ¯t, while novel anti-angiogenic factors include EMAP II. q 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Lung branching morphogenesis; Lung cell proliferation and differentiation; Alveolization; Master genes; Peptide growth factor signaling; Extracellular matrix signaling; Mesenchyme induction; Alveolar epithelial cells; Pulmonary neuroendocrine cells; Stem cells

1. Introduction The problem of respiratory gas exchange has been solved in several ingenious ways during evolution. Simple diffusion occurs in unicellular organisms. Aqueous diffusion occurs in gills of ®sh. Delivery of air from the surface of * Corresponding author. Developmental Biology Program, Children's Hospital Los Angeles Research Institute, 4650 Sunset Boulevard MS 35, Los Angeles, CA 90027, USA. Tel.: 11-323-669-5422; fax: 11-323-6713613. E-mail address: [email protected] (D. Warburton)

the body directly to individual cells is accomplished via well-branched respiratory tracheae in Drosophila larvae. In air breathing amphibians and mammals lungs exchange gases with the circulation by diffusion through very thin cells. The metabolic needs of tissue respiration in larger homeotherms such as humans dictate a very large, water tight, readily diffusible interface between air and the circulation. The human lung achieves a ®nal gas diffusion surface of 70 m 2 in area by 0.1 mm in thickness in young adulthood. It is capable of supporting systemic oxygen consumption

0925-4773/00/$ - see front matter q 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0925-477 3(99)00325-1

56

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

Fig. 1. Diagrams showing key events in human lung morphogenesis. (A) The primitive lung anlage emerging as the laryngo-tracheal groove from the ventral surface of the primitive foregut at 5 weeks gestation in human. (B) The primitive trachea separating dorso-ventrally from the primitive esophagus as the two primary bronchial branches arise from the lateral aspects of the laryngo-tracheal groove at 5±6 weeks gestation in human. (C) The embryonic larynx and trachea with the two primary bronchial branches separated dorso-ventrally from the embryonic esophagus at 6 weeks in human. (D) The primitive lobar bronchi branching from the primary bronchi at 7 weeks in human. (E) A schematic rendering of the term fetal airway in human. The stereotypically reproducible, ®rst 16 airway generations are complete by 16 weeks in human. Between 16 and 23 weeks, the branching pattern is random and this is completed by about 24 weeks in human. Alveolization begins after about 28±30 weeks in human and is complete by 7 years of age at the earliest.

ranging between 250 ml/min at rest and 5500 ml/min during maximal exercise. A matching capillary network develops in close apposition to the alveolar surface, which can accommodate pulmonary blood ¯ow rising from 4 l/min at rest to 40 l/min during the transition from rest to maximal exercise (Comroe, 1965). To solve the developmental problem of forming such a large, diffusible interface with the circulation, the embryonic lung undergoes a process termed branching morphogenesis, which culminates postnatally in alveolar saccule formation. In humans, the lungs originate as the laryngo-tracheal groove in the endodermal epithelium lining the ¯oor of the primitive embryonic anterior pharynx at 5 weeks of gestation (Fig. 1). The laryngo-tracheal groove separates

dorso-ventrally from the primitive esophagus to form the tracheal rudiment and at the same time gives rise laterally to two primary bronchial buds, which then begin a reproducible, bilaterally asymmetrical pattern of stereotypic branching, followed by dichotomous branching into the surrounding splanchnic mesenchyme. This process gives rise to bilobed left and trilobed right lungs in humans. Reproducible branching is complete at 16 generations by 16 weeks in humans. The ®nal seven generations of airways (for a total of 23) in human are completed during the latter part of gestation. Alveolization begins around 20 weeks in humans and is completed postnatally. In the mouse, the lungs also arise from the laryngotracheal groove at gestational day 9±9.5 (E9±9.5) (Kauffman, 1992). This forms an anteriorly directed tracheal diverticulum from the ventral side of the pharyngeal region of the primitive foregut, which subsequently forms the primitive larynx and trachea. The ®rst evidence of tracheal bifurcation is seen shortly after the appearance of the laryngo-tracheal groove. At this stage the primitive lungs are fairly symmetrical buds that bulge outwards into the pleuroperitoneal canals. The subsequent branching process in mouse is different from human in that the bronchial buds then give rise to unilobar left and quadrilobed right lungs, which are well established by E12 (Fig. 2). The pulmonary artery also arises as a branch of the paired sixth branchial arch arteries and descends into the embryonic lungs during this time. Histologically, lung development has been divided into four chronological stages in mouse: (1) pseudoglandular stage (E9.5±16.6), the bronchial and respiratory tree develops and an undifferentiated primordial system forms; (2) canalicular stage (E16.6±17.4), terminal sacs and vascularization develop in this period; (3) terminal sac stage (E17.4 to postnatal day 5 (P5)), the number of terminal sacs and vascularization increase and type I and II cells differentiate; and (4) alveolar stage (P5±30), terminal sacs develop into mature alveolar ducts and alveoli.

Fig. 2. Species differences between murine and human pulmonary lobation. (A) Diagram showing a drawing of 7±8 week embryonic human lungs. The mainstem bronchi have already branched into lobar and segmental bronchi. There are three lobes in the right lung and two in the left lung (after Streeter, Contr. Embryol. 32:1948, as cited in Gray's Anatomy by Dai Davies, 1967). (B) Diagram showing a drawing of embryonic day 12 murine lungs. There are four lobes in the right lung and one in the left lung. Drawn by D.W. from a preparation made by Pablo Bringas.

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

The stage of alveolar development at birth varies quite widely between species; in mouse and rat it is largely postnatal, whereas in guinea pig, sheep and humans it begins in utero and is relatively advanced at birth. In the sheep, lung maturation proceeds rostro-caudally, with the apical lobes maturing much earlier than the basal lobes. Epithelial cell lineages are arranged in a distinct proximodistal spatial pattern in the airways. The larynx is lined with squamous epithelium and the upper airways are lined with ciliated columnar cells and mucus secreting cells. The lower airways are lined with Clara cells. The alveoli are lined with alveolar type 1 and 2 epithelial cells (AEC 1 and 2). Pulmonary neuroendocrine (PNE) cells are situated in small foci and are surrounded by the other epithelial cells in the upper airways. The pulmonary interstitium contains several specialized lineages of mesenchymal origin including ®broblasts, myo®broblasts and smooth muscle cells. The vasculature comprises several populations of arterial, venous and capillary endothelial cells, together with smooth muscle and other specialized cells in the vessel walls. There is also a lymphatic system lined by lymphatic endothelial cells. The lung is innervated by specialized sympathetic and parasympathetic neurons innervating the airways and vasculature and also contains ganglia. Specialized pulmonary macrophages can also be identi®ed in the lung interstitium as well as alveolar macrophages in the alveoli. Molecular markers identify gene expression characteristic of several of the pulmonary epithelial cell lineages. In the early murine embryonic lung, the primitive epithelium coexpresses several lineage markers including surfactant associated proteins (SP)-C, SP-A, Clara cell (CC)-10 and calcitonin gene-related peptide (cGRP), which later in gestation are characteristic of distinct AEC 2, Clara cell and PNE cell lineages, respectively (Table 1). Thus, the primitive early embryonic epithelium appears to be multipotential, and can be induced to express speci®c peripheral cell lineages by soluble factors derived from lung mesenchyme. Later in gestation, lung epithelial cell lineages become restricted to speci®c regions of the airway. Additionally,

Table 1 Expression of cell lineage-speci®c markers in epithelial cell lineages in the developing mouse lung a Gestational age (days)

Cell lineage

E12±14 E18

Epithelium AEC 2 Clara cell PNE cell

Marker gene expression SP-A

SP-C

cGRP

1 1 1

1 1

1 1

CC-10 1 1

a At E12±14, the primitive pulmonary epithelium co-expresses SP-A, SPC, cGRP and CC-10 in the same cells. By E18, distinct cell lineages have emerged which express speci®c cell lineage markers (Wuenschell et al., 1996).

57

some potential for transdifferentiation between cell types may still be possible at least in the alveolar epithelium, where transitions from AEC 2 to AEC 1 phenotype and back again have been elicited by culture conditions. Also, following lung injury, AEC regain the capacity to proliferate and repopulate the damaged alveolar epithelium. Whether alveolar epithelial progenitor cells exist and play a role in the proliferative/repair process is currently under study. Our goal in this review is to synthesize a set of coherent, mechanistic insights from what is currently known about the molecular basis of lung morphogenesis. Several other useful reviews on this topic have recently appeared. These emphasize different aspects of the molecular basis of lung morphogenesis, and compare them with other relevant developmental paradigms, such as the Drosophila tracheal system and the limb bud (Whitsett, 1998; Hogan, 1999; Metzger and Krasnow, 1999; Placzek and Skaer, 1999; Warburton and Lee, 1999; Warburton et al., 1999).

2. Lung morphogenesis is determined by functional integration of key transcriptional factors, peptide growth factor receptor-mediated signaling, extracellular matrix, integrin and non-integrin signaling These inputs are integrated during the normal process of embryonic, fetal and postnatal lung morphogenesis. They instruct organized temporo-spatial patterns of cellular proliferation, cell lineage differentiation, cell movement and cell death that determine structure and hence physiological function. Thus, lung development extends in a coordinated manner from the induction of the lung epithelial rudiment, through branching morphogenesis in early embryonic life, through the critical transition from fetal life to air breathing, culminating in the completion of alveolarization, which occurs postnatally (Fig. 1). Pulmonary branching is reproducible in its spatial pattern as far down the respiratory tree as 16 generations in humans (Fig. 1). This suggests that the pattern of respiratory morphogenesis in the ®rst 16 airway generations is genetically predetermined. The term `hard-wiring' has been coined to indicate genetic preprogramming of the temporo-spatial information that determines the highly reproducible orientation of the ®rst 16 airway generations within the rostro-cordal, dorso-ventral and left±right axes of the overall body plan (Hogan, 1999). Between the level of 16 and 23 generations, the last of which lead into the alveoli, branching appears to follow a more random-appearing distribution. It has been suggested that mathematical fractal functions deduced from analysis of branching patterns not only of lungs, but also of other branched structures in biology (breast, bile ducts, pancreatic ducts, etc.) as well as in nature (watersheds, plant morphology), may govern the latter stages of lung morphogenesis.

58

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

3. Genes identi®ed in branching mutants of the respiratory organs (tracheae) of Drosophila are functionally conserved in branching morphogenesis of the lung Drosophila embryos solve the problem of tissue respiration by delivering air directly to individual cells through a series of branched tubes termed tracheae. The tracheae originate during Drosophila embryogenesis from the respiratory placodes, which are placed segmentally along the lateral surface of the embryo. The placode cells proliferate to a limited extent and invaginate inwards to form initial branches. Cytoplasmic projections of individual distal tracheal cells then ramify to form a series of gas delivery tubules that terminate in a predictable fashion at individual internal organs (Fig. 3). At least 30 tracheal enhancer trap markers and several distinct mutations in Drosophila are associated with speci®c defects in a sequential series of morphologically distinct but genetically coupled branching events in the process of tracheal branching morphogenesis (Samakovlis et al., 1996). Primary branching in Drosophila is regulated by branchless (bnl), breathless (btl) and stump; secondary branching is regulated by pointed, rhomboid, sprouty and hypersprouty; terminal branching is regulated by pruned, trimmed, cropped, and misguided. Trachealess (trh) has been identi®ed as necessary to direct the initial steps of tubulogenesis in both the respiratory

organs and salivary glands of Drosophila (Wilk et al., 1996). In trh mutants, tube-forming cells of the trachea, salivary gland and ®lzkoÈrper fail to invaginate to form tubes and remain on the embryo surface (Isaac and Andrew, 1996). Expression of trh is in turn controlled by Sex combs reduced (scr) and forkhead (fkh) transcriptional factors. The human homologue of trh is the hypoxia-inducible factor (HIF)-1a (Andrew et al., 1994). Null mutation of HIF-1a in mice results in severe defects of vasculogenesis and mesenchymal cell death (Ryan et al., 1998; Kotch et al., 1999). Branchless (bnl) encodes a Drosophila ®broblast growth factor (FGF) ligand homologue that controls tracheal cell migration and hence the pattern of branching (Glazer and Shilo, 1991; Sutherland et al., 1996). The latter mutation is complimentary to the breathless (btl) mutation, which is a loss of function mutation in the Drosophila cognate bnl receptor. Fibroblast growth factor receptor (FGFR) is the mammalian homologue of btl in Drosophila. The phenotype of the bnl and btl loss of function mutations suggest that bnl activates btl. This FGF homologue signaling pathway directs the migration of tracheal cells during primary branching and then activates downstream programs of ®ner branching at the ends of the growing branches (Sutherland et al., 1996). Expression of a dominant active btl receptor in Drosophila interfered with tracheal cell migration and led to extra secondary and tertiary branch forming cells. Conversely, reduction in btl signaling enhanced the cell

Fig. 3. Tracheal morphogenesis in Drosophila. (A) Respiratory placodes as paired structures on the lateral surface of Drosophila embryos. Null mutation of breathless (FGFR homologue) and branchless (FGF ligand homologue) prevent progression to the stage shown in the next panel. (B) Primary branches form due to coordinated proliferation and invagination of the cells diagramed as (1) in the respiratory placode. This requires expression of breathless and branchless. (C) Secondary branches arise as cytoplasmic extensions of cells in the base of the primary branch, which will give rise respectively to terminal branches and fusion branches. Expression of the pointed gene is necessary for this step in morphogenesis. (D) Terminal branches arise as cytoplasmic tubules extending from individual cells in the secondary branches. Expression of the terminal-1 gene is necessary for this step in morphogenesis. Pairs of tracheae are linked medially by a fusion branch.

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

migration defects while suppressing the ectopic branching defects (Lee et al., 1996). These results are consistent with a model in which spatially regulated bnl signaling guides tracheal epithelial cell migration while quantitative regulation of btl activity determines the patterns of secondary and tertiary branching (Lee et al., 1996). Supporting the latter model, the Drosophila sprouty null mutation results in a strong gain of function phenotype for tracheal branching. Sprouty functions therefore as an antagonist of bnl activation of btl (Hacohen et al., 1998). In the pointed loss of function Drosophila mutant pnt D88, primary tracheal branching occurs normally, but no secondary branches are ever formed. The pnt gene is also required for activating and repressing marker genes that underlie terminal branching and branch fusion. The pnt gene encodes two ETS-like proteins which are also involved in the development of midline glial cells (Klambt, 1993). The pruned gene product is a Drosophila homologue of serum response factor (SRF), which functions as a growth factor-activated transcription complex together with an ETS domain ternary complex factor (Guillemin et al., 1996). In pruned loss of function mutants, terminal tracheal cells failed to extend cytoplasmic projections, while with expression of either a constitutively activated SRF or ternary complex factor, extra projections grow out in an unregulated manner (Guillemin et al., 1996). Thus, Drosophila SRF may function in a growth factor inducible transcription complex to regulate distal cytoplasmic outgrowths during tracheal epithelial branching morphogenesis. Branching of the respiratory tracheae in Drosophila can therefore be considered as a conserved genetic paradigm for morphogenesis of the lung. Strong evidence is currently emerging that the respective mammalian homologues of Drosophila FGF pathway gene family members, including FGF-10/bnl, FGFR-2/btl and sprouty, play key roles in determining mammalian lung branching morphogenesis (also see Section 6.2 and the excellent review by Metzger and Krasnow, 1999). 4. Initiation of lung morphogenesis and separation from the embryonic anterior foregut is instructed by the interaction of sonic hedgehog (shh)/patched receptor (ptc)/smoothened (smo) and the Gli transcriptional factor family Hedgehog was ®rst discovered in a genetic screen of Drosophila melanogaster (Nusslein-Volhard and Wieschaus, 1980). Subsequently, a family of hedgehog proteins including Sonic, Indian, Desert and Tiggy-Winkle have been discovered. Sonic hedgehog (shh) is the original murine homologue of the Drosophila hedgehog gene, which encodes a secreted protein. Shh regulates decapentaplegic (Dpp) and wingless (Wg) in target cells (Heberlien et al., 1993; Basler and Struhl, 1994; Capdevilla et al., 1994; DiazBenjumea et al., 1994). Low levels of shh are expressed

59

throughout the primitive respiratory epithelium with high levels seen at the tips (Bitgood and McMahon, 1995; Bellusci et al., 1996; Urase et al., 1996). Shh null mutant mice exhibit grossly abnormal foregut development (Fig. 4). The trachea and esophagus remain fused and the lungs form bilateral rudimentary sacs arising

Fig. 4. Diagrams of several selected, informative null mutant phenotypes and a dominant negative misexpression phenotype in early embryonic lung morphogenesis. (A) The HNF3b 2/2 phenotype, which is a severe early embryonic lethal phenotype in which the primitive foregut endoderm fails to close into a tube (Ang and Rossant, 1994). (B) The compound null Gli2 2/2/Gli3 2/2 mutant phenotype in which the primitive lung anlage completely fails to arise from the primitive foregut endoderm (Motoyama et al., 1998). Gli2 2/2 null mutation and Gli2 gene dosage reduction result in abnormalities of lung lobation (see text for some further details). (C) The FGF-10 2/2 phenotype in which the larynx and trachea form and separate dorso-ventrally from the esophagus; but, the primary bronchial branches completely fail to arise from the trachea (Min et al., 1998). (D) The morphologically similar Nkx2.1 and Shh null mutant phenotypes in which the trachea fails to separate dorso-ventrally from the esophagus, forming a tracheo-esophageal tube from the sides of which grossly hypoplastic and dysplastic epithelial bags arise. In the Nkx2.1 null mutant, epithelial differentiation is arrested prior to the expression of peripheral epithelial markers, while in contrast in the Shh null mutant, peripheral lung epithelial markers are expressed (Kimura et al., 1996; Pepicelli et al., 1998; Minoo et al., 1999). (E) The phenotype in transgenic mice that misexpress a dominant negative, tyrosine kinase deleted FGFR under the control of the SP-C promoter/enhancer. Bronchial morphogenesis is abrogated distal to the primary bronchi (Peters et al., 1992).

60

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

from the single tracheo-esophageal tube, due to failure of branching and growth after the formation of the primary lung buds. However, proximo-distal differentiation of the airway epithelium can be observed, indicating that shh is essential for branching but not for proximo-distal epithelial differentiation (Pepicelli et al., 1998). The minimal surfactant protein C (SP-C) promoter is expressed at high levels early on in the primitive embryonic lung epithelium, and eventually becomes restricted to the type 2 alveolar epithelial cell lineage. The SP-C enhancer/ promoter transgene construct invented by Whitsett has been successfully used to drive quite high level ectopic expression of many individual genes of interest into the primitive respiratory epithelium. Misexpression of shh from the SP-C enhancer/promoter in transgenic mice results in increased epithelial and mesenchymal proliferation, giving rise to a neonatal lethal lung phenotype that contains abundant mesenchyme and no functional alveoli (Bellusci et al., 1997a). This is associated with upregulation of ptc expression but not BMP4, Wnt2 or FGF-7. Thus, shh appears to play an important role as a mitogen for primitive respiratory mesenchymal cells near the growing tips of the early embryonic lung, while also inducing ptc expression in the adjacent mesenchyme. Shh pro-protein is cleaved into 19 kDa N-terminal and 26±28 kDa peptides. The N-terminal peptide is modi®ed by the addition of a cholesterol molecule, which acts as a membrane anchor. The C-terminal peptide is freely diffusible and exerts distant effects along concentration gradients. Soluble shh secreted from the embryonic lung epithelium acts on adjacent mesenchymal cells by activating patched receptors. Since Shh targets mesenchymal cells through patched receptors, it is likely that any effects on epithelial proliferation are indirect. Patched (Ptc) is the murine homologue of a segment polarity gene in Drosophila, which encodes a putative multiple pass membrane-spanning receptor that is required for shh signaling. Upon shh binding, ptc receptors release another segmental polarity gene product termed smoothened (smo). Once smo is released, Gli is activated to function as a transcriptional activator. Ectopic expression of hedgehog in anterior imaginal disc structures in Drosophila results in upregulation of ptc and increased growth and duplication of adult structures. Mutations in the human ptc gene are causally associated with hereditary basal cell nevus syndrome, which comprises abnormal proliferation and patterning of many tissues (Johnson et al., 1996). The mouse ptc gene is highly expressed in mesenchyme surrounding terminal lung buds. Murine Gli1, 2 and 3 are related to Drosophila cubitus interruptus. In mice, the three zinc ®nger transcription factors, Gli1, 2 and 3, have been implicated in transduction of the sonic hedgehog (Shh) signal. Partial loss of function mutations in Gli3 result in shape and size reductions in pulmonary segmental branches (Grindley et al., 1997). Gli3 2/2 mice show a similar lung phenotype. However,

Gli2/Gli3 2/2 double null mutants either do not have any lung at all or have complete tracheo-esophageal ®stulas (Fig. 4). Gli2 2/2 mice have unilobar lungs bilaterally (in contrast to the normal phenotype comprising four lobes on the right and one on the left). Gli2 null mutant mice can also exhibit other foregut defects, including stenosis of the esophagus and trachea as well as hypoplasia and lobulation defects of the lung. A 50% reduction of Gli3 gene dosage in the Gli2 null mutant background resulted in esophageal atresia with tracheo-esophageal ®stula and a more severe lung phenotype. Thus, Gli2 and Gli3 appear to have speci®c, overlapping and essential functions during embryonic foregut development, since compound null mutations of both genes completely abrogated lung morphogenesis from the early embryonic foregut (Motoyama et al., 1998). Thus, the shh/ptc/smo/Gli pathway clearly plays an essential role in early pattern formation as the lung arises from the primitive foregut, but the speci®c downstream targets of Gli are as yet unknown. Transcriptional factors play key roles in determining both lung morphogenesis and cytodifferentiation: the contribution of each of the following transcription factors is considered in turn (also see Whitsett, 1998 for a more concise commentary). 4.1. Nkx2.1 Nkx2.1 is a homeodomain protein that for historical reasons is also known as thyroid transcription factor-1 (TTF-1), as well as thyroid enhancer binding protein (T/ ebp). Nkx2.1 is essential for the complete induction of embryonic lung branching morphogenesis (Minoo et al., 1995, 1999; Kimura et al., 1996). Abrogation of Nkx2.1 expression by antisense oligodeoxynucleotide abrogation of expression in culture results in complete interruption of branching morphogenesis and dysplasia of the embryonic mouse lung epithelium (Minoo et al., 1995). Null mutation of Nkx2.1 results in a strikingly similar neonatal lethal lung phenotype in association with absence of the thyroid, pituitary and parts of the brain (Kimura et al., 1996). Nkx2.1 null mutation also results in failure of septation of the esophagotracheal primordium, resulting in a phenotype resembling severe human cases of cleft larynx with an esophago-trachea (Minoo et al., 1999). A pair of lung primordia arise from the sides of the single lumen esophago-trachea (Fig. 5). These mutant lungs fail to undergo branching morphogenesis. In situ hybridization suggests decreased BMP4 expression in the mutant epithelium. Pulmonary-speci®c gene expression such as that of the surfactant proteins (SP-B and -C) and the Clara cell CC-10 protein is also turned off in the epithelium of these dysplastic bag-like structures, which do however express cilia and mucus secreting cells. VEGF expression is also reduced. This demonstrates that the absence of Nkx2.1 arrests both dorso-ventral separation of the trachea from the esophagus, as well as lung branching morphogenesis and epithelial cell lineage determination at an early stage,

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

prior to the speci®cation of peripheral lung cell phenotypes. Recently, deletion of the Nkx2.1 gene has been reported in a human infant with both thyroid dysfunction and respiratory failure (Devriendt et al., 1998). Nkx2.1 consensus recognition sites are found in the 5 0 promoters of several important peripheral lung cell lineage-speci®c genes, including SP-A, -B, -C, -D, CC-10 and Nkx2.1 itself (Bohinski et al., 1993, 1994; Bruno et al., 1995; Ikeda et al., 1995; Venkatesh et al., 1995; Yan et al., 1995; Kelly et al., 1996; Zhang et al., 1997a). Study of the cis-acting regions in the Nkx2.1 promoter itself reveals that Nkx2.1 has multiple promoters, with binding sites for both ubiquitous and speci®c transcription factors including GATA-6, HNF-3, Sp1 and Sp3 as well as Nkx2.1 itself. The latter ®nding suggests that feed-forward mechanisms sustain expression of Nkx2.1 (Toonen et al., 1996; Nakazato et al., 1997). Thus, initiation of Nkx2.1 expression may require initial activation by other factors, but may then be sustained by autoregulation. 4.2. Forkhead homologue hepatocyte nuclear family proteins The hepatocyte nuclear (HNF) family of transcription factors is related to the forkhead family of Drosophila and is known to play a key role in regional speci®cation of epithelial cell fates in the gastrointestinal tract and liver. HNF3-b null mutation results in a severe early embryonic lethal phenotype with complete failure of the primitive foregut to close into a tube (Ang and Rossant, 1994) (Fig. 4). HNF-a is expressed in the gut epithelium anterior to the liver (Fig. 4). Consensus HNF binding sites are found in the 5 0 promoter regions of several peripheral lung-speci®c genes including SP-A, -B, -C, -D, CC-10 and Clara cell secretory protein (CCSP) in close proximity to Nkx2.1 sites (Bohinski et al., 1994; Clevidence et al., 1994). Thus, the HNF family appears to cooperate with the Nkx2.1 family to determine pulmonary epithelial cell lineage fates (Bingle and Gitlin, 1993; Bingle et al., 1995; Zhou et al., 1997). Combinatorial interaction between HNF-3 and Sp family transcriptional factors also regulates the CC-10 promoter (Braun and Suske, 1998). Signi®cantly, HNF-3 also activates transcription of Nkx2.1 in respiratory epithelial cells (Ikeda et al., 1996). 4.3. GATA family zinc ®nger transcriptional factors Members of the GATA family of zinc ®nger transcriptional factors may play important roles in both epithelial and smooth muscle cell lineage diversity in the lung. GATA-6 regulates HNF4 and is required for differentiation of visceral endoderm (Morrisey et al., 1998). GATA-6 induced differentiation of primitive foregut endoderm into respiratory epithelial cell lineages is also directed at least in part by interactions among HNF-3b, various forkhead homologue family members, Nkx2.1 and GATA family members. These genes interact at the level of target gene transcrip-

61

tional regulation, mediating not only expression of surfactant protein genes but also expression of other transcriptional factors and possibly smooth muscle genes. For example, GATA-6 regulates SP-A and -C, Nkx2.1 regulates SP-A, -B, -C, -D and CCSP and HNF-3b regulates SPB and CCSP. GATA-6 is expressed in arterial smooth muscle, the developing bronchi, urogenital ridge and bladder (Morrisey et al., 1996). GATA-5 is also expressed in the lung mesenchyme at early stages and is also expressed in bronchial smooth muscle later in morphogenesis (Morrisey et al., 1997). 4.4. Pou domain proteins Oct-1 is the major POU domain protein expressed in H441 cells and this protein binds two functional sites in the minimal Nkx2.1 promoter (Bingle and Gowan, 1996). This suggests that the ubiquitously expressed POU domain protein Oct-1 may play a role in basal activation or maintenance of transcription of the lung-speci®c transcriptional factor Nkx2.1. 4.5. Hox genes Expression of at least 16 hox genes from clusters A and B is temporo-spatially restricted during murine lung branching morphogenesis. However, the levels of Hox gene expression decline with advancing gestational age. In general, Hox genes within 5 0 clusters have more restricted expression than those within 3 0 clusters (Mollard and Dziadek, 1997). Hoxb4 distribution in the developing mouse lung suggests a role in branching morphogenesis and epithelial cell fate (Volpe et al., 1997). Hoxb5 is negatively regulated by EGF and transforming growth factor-b (TGF-b) in cultured embryonic lung (Chinoy et al., 1998). On the other hand, retinoic acid induces Hoxa5, b5 and b6 gene expression (Bogue et al., 1994). Hoxa5 null mutation results in early postnatal lethality. In Hoxa5 2/2 pups, tracheal occlusion and respiratory distress were associated with a marked decrease in surfactant protein production together with altered Nkx2.1, HNF-3b and N-myc gene expression in the pulmonary epithelium (Aubin et al., 1997). Since Hoxa5 expression is restricted to the lung mesenchyme, the null mutant phenotype strongly supports the inference that Hoxa5 expression is necessary for induction of lung epithelial gene expression by the underlying mesenchyme. 4.6. Pod-1, a mesoderm-speci®c bHLH protein Pod-1 is selectively expressed in mesenchymal cells at sites of mesenchymal±epithelial interaction in the kidney, lung, intestine and pancreas (Quaggin et al., 1998). Inhibition of Pod-1 expression with antisense oligodeoxynucleotides decreased mesenchymal cell condensation around the ureteric bud and decreased ureteric branching by 40% in embryonic kidney culture. Preliminary studies suggest that

62

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

Pod-1 may also play a role in mesenchymal±epithelial interaction in lung morphogenesis. 4.7. Lefty-1 and HFH-4 function in speci®cation of left± right laterality Lefty-1, lefty-2 and nodal are expressed on the left side of mouse embryos and are implicated in determination of left± right laterality. Lefty-1 null mutant mice showed a variety of left±right isomerisms in visceral organs, but the most common feature of lefty-1 null mutation was thoracic left isomerism. The lack of lefty-1 expression resulted in bilateral expression of nodal, lefty-2 and Pitx2 (a homeobox gene normally expressed on the left side). This suggests that lefty-1 normally restricts lefty-2 and nodal expression to the left side, and that lefty-2 or nodal encode a signal for `leftness' in the lung (Supp et al., 1997). Null mutation of murine hepatocyte nuclear factor-4, a forkhead family gene, results in the absence of left±right dynein expression and hence of cilia in association with random left±right laterality of the visceral organs including the lungs (Chen et al., 1998). Cilliary dyskinesia and situs inversus comprise the Kartagener syndrome of congenital abnormalities in humans. 5. Peripheral embryonic lung mesenchyme can act as a `compleat' inducer of lung morphogenesis and cytodifferentiation The linked concepts of morphogens and morphogenetic gradients within multicellular developing organisms have been extant for a century (Morgan, 1897). The hypothesis that morphogenetic signals instruct differential gene expression between cell types during embryonic morphogenesis originates in the classical observation of the `Spemann organizer' in the dorsal lip of the blastopore in Xenopus embryos (Spemann, 1938). A recent re®nement of these concepts demonstrates that morphogens may act both at short and long range within a morphogenetic gradient, depending on whether the morphogen is tethered or diffusible (Zecca et al., 1996). The early embryonic lung develops at E9.5 in the mouse as endodermal epithelial appendages, which arise from the ventral surface of the primitive foregut and extend caudally and ventrally into the surrounding primitive mesenchyme. Rostro-caudally, these structures form the primitive larynx, trachea and mainstem and lobar bronchi by E10±11 in the mouse. Subsequent branching morphogenetic events give rise to the segmental and subsegmental generations of bronchi, respiratory bronchioles, terminal bronchioles and ®nally the alveoli in the left and right lungs (Fig. 1). Both classical and more recent mesenchyme recombination experiments demonstrate that the induction of early embryonic lung branching morphogenesis is determined by soluble factors produced by the primitive peripheral lung mesenchyme. Transplantation of peripheral lung

mesenchyme to the E11 mouse trachea in culture results not only in the induction of supernumary pulmonary branches from the trachea, but also induction of genes which mark the expression of peripheral epithelial cell lineages (Alescio and Cassini, 1962; Wessels, 1970; Shannon, 1994). Grafting peripheral lung mesenchyme to tracheal epithelium induces branching and expression of SP-C within 24 h. However, induction of this supernumary branching only occurs within a restricted window of lung development. Also, neither esophageal nor gut epithelium will respond. Trans®lter recombination experiments have shown that the induction of supernumary branches is mediated by soluble factors. Moreover, deleting individual components of a de®ned medium suggests that FGF-7 is necessary but not suf®cient to mediate these inductive events (Deterding and Shannon, 1995). 6. Peptide growth factor signaling is both inductive and permissive for lung morphogenesis and modulates transcriptional mechanisms Branching morphogenesis and cell lineage differentiation occur spontaneously in E11 mouse early embryonic lung under serumless chemically de®ned conditions (Jaskoll et al., 1988; Warburton et al., 1992; Wuenschell et al., 1996) as well as in zero gravity (Spooner et al., 1994). This ®nding has suggested the hypothesis that autocrine and paracrine factors, produced within the lung anlage, are necessary and suf®cient at least for the embryonic phase of lung branching morphogenesis to occur. Soluble factors released by peripheral lung mesenchyme can induce ectopic branching from the trachea of early mouse embryonic lung explants as well as inducing expression of a complete repertoire of genes speci®c to peripheral lung epithelium including SP-A, -B, -C and CC-10 (Shannon, 1994). Candidate inductive and permissive peptide growth factors, whose signaling peptides and cognate receptors are expressed in the early mouse embryonic lung, include epidermal growth factor (EGF), ®broblast growth factors (FGFs), hepatocyte growth factor (HGF), insulin-like growth factor (IGF), keratinocyte growth factor (KGF), platelet-derived growth factor (PDGF) and transforming growth factor-b3 (TGF-b3). Their inductive and/or permissive in¯uences on lung development have been demonstrated by complementary gain versus loss of function experiments in early embryonic mouse lung organ culture, in transgenic and in null mutant mice (Raaberg et al., 1992, 1995a,b; Warburton et al., 1992, 1993; Minoo and King, 1994; Peters et al., 1994; Kaartinen et al., 1995; Miettinen et al., 1995; Souza et al., 1995; Shiratori et al., 1996). In general, peptide growth factor cognate receptors with tyrosine kinase intracellular signaling domains such as epidermal growth factor receptor (EGFR), FGFR, c-met, insulin-like growth factor receptor (IGFR), KGFR and platelet-derived growth factor receptor (PDGFR) stimulate

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

lung morphogenesis. In contrast, those cognate receptors with serine/threonine kinase intracellular signaling domains, such as the TGF-b family are inhibitory (Warburton et al., 1992; Seth et al., 1993; Serra et al., 1994; Kaartinen et al., 1995; Miettinen et al., 1995; Serra and Moses, 1995; Zhao et al., 1995, 1996). However, evidence is emerging that the BMP signaling system plays a key inductive role in the induction of peripheral lung structure and epithelial lineage differentiation (see Hogan, 1999 for review and commentary). The functional evidence supporting the involvement of each of these classes of peptide growth factors and their cognate receptors will be considered in turn. 6.1. EGF, TGF-a , amphiregulin and the EGF receptor EGF, transforming growth factor-a (TGF-a) and amphiregulin are all EGFR ligands which can positively modulate early mouse embryonic lung branching morphogenesis and cytodifferentiation (Warburton et al., 1992; Seth et al., 1993; Schuger et al., 1996). Stimulation of EGFR signaling with exogenous ligand stimulates early murine embryonic lung branching morphogenesis, epithelial and mesenchymal cell proliferation and Nkx2.1 and SP-C expression as markers of cytodifferentiation (Warburton et al., 1992; Schuger et al., 1996). Abrogation of EGFR signaling with tyrphostins, EGFR antisense oligodeoxynucleotides gene knock-down by arti®cially induced maternal immunity or EGFR null mutation all result in decreased branching morphogenesis in culture and a neonatal pulmonary lethal phenotype in the null mutant associated with decreased branching morphogenesis and Nkx2.1 and SP-C expression (Warburton et al., 1992; Seth et al., 1993; Miettinen et al., 1995; Raaberg et al., 1995a,b). EGF also is expressed in mature alveolar epithelial cells and regulates type 2 cell proliferation through an apparent autocrine mechanism both in culture and in vivo (Raaberg et al., 1992). However, respiratory epithelial cell expression of TGF-a using the SPC promoter in transgenic mice induces postnatal lung ®brosis (Korfhagen et al., 1994). We have found that branching morphogenesis is reduced by 50% both in vivo and in vitro in the lungs of EGFR null mutant mice, while addition of exogenous EGF stimulates branching three-fold and SP-C levels 50-fold in wild type embryonic lungs in culture, whereas the EGFR null mutants do not respond at all to exogenous EGF (Miettinen et al., 1997). 6.2. FGFs, FGFRs and sprouty FGFs are an increasingly large and complex family of growth factors currently consisting of 18 ligands which signal through four cognate tyrosine kinase FGF receptors. Complex ligand±receptor interactions, also involving heparan sulfate proteoglycans, activate a variety of intracellular pathways that regulate cell proliferation, differentiation and pattern formation in several developing systems

63

(Jaye et al., 1992; Reich-Slotsky et al., 1994; Martin, 1998; Hogan, 1999). FGFs are essential components of the regulatory networks between epithelium and mesenchyme in the embryonic lung. The importance of FGF signaling in lung development was ®rst demonstrated by the profound disruption of branching morphogenesis in lungs of transgenic mice expressing dominant negative forms of FGFR-2 (Peters et al., 1994; Celli et al., 1998). Expression of a truncated, dominant negative, kinase de®cient FGFR mutant in the primitive respiratory epithelium of transgenic mice under the control of the SP-C promoter/enhancer resulted in a lethal phenotype comprising complete pulmonary aplasia distal to the mainstem bronchial branches (Peters et al., 1994) (Fig. 4). In these mice, tracheal cell lineages are normally expressed, but peripheral epithelial cell lineages marked by SP-C expression as well as the vasculature are not. Thus, FGFR signaling is clearly essential for pulmonary epithelial branching morphogenesis and differentiation distal to the bronchi as well as for embryonic pulmonary vasculogenesis. This ®nding of pulmonary hypoplasia with abrogation of FGFR signal transduction has recently been con®rmed by generalized overexpression of a dominant negative soluble form of the FGFR extracellular domain (Celli et al., 1998). During postnatal life, FGFR-3 and -4 play an essential role in regulating alveolization (Weinstein et al., 1998). Double null mutation of FGFR-3 and -4 results in a postnatal lethal pulmonary phenotype in which excess elastin is laid down and alveoli fail to form (Weinstein et al., 1998). At early stages of lung development FGFR-2 is expressed at high levels along the entire proximal-distal axis of the respiratory tract epithelium (Peters et al., 1992; Cardoso et al., 1997). The wide temporo-spatial distribution of FGFR-2 expression suggests that spatial control of branching is most likely determined by non-epithelial expression of an FGF ligand patterning the epithelial tubules, as does the Drosophila FGF homologue, branchless (bnl), during tracheal development. Expression of bnl in cells clustered adjacent to tracheal tubules activates the breathless (btl) receptor in the epithelium, inducing epithelial migration and elongation toward the bnl-expressing cells (Sutherland et al., 1996). These cells act as a chemotactic focus for the tracheal epithelium. Recent studies suggest that the chemotactic function of FGFs in respiratory organogenesis might be evolutionarily conserved, and that the mammalian homologue FGF-10 may play a similar role to bnl in epithelial branching. FGF-10 transcripts are present at discrete sites in the mesenchyme of day 11±12 mouse lungs, close to distal epithelial tubules. These sites of expression change dynamically in a pattern that is compatible with the idea that FGF-10 appears in the mesenchyme at prospective sites of bud formation (Bellusci et al., 1997b). The functional similarity to bnl resides in FGF-10's ability to in¯uence budding by exerting a chemotactic effect on lung epithelium, at least

64

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

in culture systems. In mesenchyme-free epithelial cultures, recombinant FGF-10 induces generalized budding when added to the medium (Bellusci et al., 1997b; Park et al., 1998). However, if a protein gradient is established by implantation of an FGF-10-soaked heparin bead, either in mesenchyme-free epithelial cultures or whole lung cultures, bud growth is directed toward the FGF-10 source. In this assay, the presence of high levels of FGF-10, arti®cially ®xed in time and space, likely disrupt natural gradients of resident signaling molecules, leading to distortions of airway pattern formation. Interestingly, in whole lung cultures this manipulation results in formation of a multilayered epithelial structure around the FGF-10 bead (Fig. 5). The chemotactic effect seems to be speci®c for distal lung buds. Whether the results obtained in this model system re¯ect the patterning effects of FGF-10 in the developing lung in vivo is not yet known. Precise spatial-temporal regulation of FGF-10 expression and a ®ne control of proliferation would be required to specify sites of budding and control bud elongation during branching. The most compelling evidence that FGF-10 is an essential regulator of the processes described above in vivo is the complete absence of lungs in FGF-10 null mice (Min et al., 1998; Sekine et al., 1999). The respiratory tract in these animals consists of a blunt ended tracheal tube, truncated at the level of the thymus (Min et al., 1998) (Fig. 4). This phenotype is reminiscent of the tracheal dysgenesis found in bnl mutant ¯ies and reinforces the idea that FGF function in respiratory organogenesis is conserved throughout evolution (Sutherland et al., 1996) (Figs. 3 and 4). Is the ability to induce and guide lung buds unique to FGF-10? Data from lung organ cultures suggest that this does not seem to be the case. When added to the medium, recombinant FGF-1 (aFGF) can elicit a robust generalized budding response in mesenchyme-free epithelial cultures (Nogawa and Ito, 1995). Furthermore, if soaked in beads

Fig. 5. FGF-10 inducing chemotaxis in embryonic lung epithelium. Embryonic mouse lung cultured in the presence of an FGF-10-soaked bead (left) and a PBS-bead (right) for 48 h. FGF-10 exerts a chemotactic effect in the distal lung epithelium. Lung bud growth is deviated toward the bead, forming concentric layers (left) by 48 h, and distorting the tubular architecture. The PBS-bead (right), in contrast, elicits no response either in proximal or distal epithelium. Scale: bead size, 100±150 mm.

and grafted onto whole lung cultures, FGF-1 is also chemotactic for lung buds; the effect is, however, weaker than that seen with FGF-10 (Park et al., 1998). By being able to ef®ciently bind to all FGFRs (Jaye et al., 1992), FGF-1 most likely activates FGFR-2 IIIb, the receptor that transduces FGF-10 responses. However, this effect is not relevant for early budding in vivo, since FGF-1 mRNA is not present in the early lung and its expression pattern differs considerably from that of FGF-10 (Bellusci et al., 1997b; Park et al., 1998). Thus, a role for FGF-1 in lung development is still not clearly established. FGF-7 (KGF) is another family member found in the developing lung mesenchyme; the highest levels are detected at late stages (Post et al., 1996; Park et al., 1998). Yet, despite high sequence homology and utilization of the same receptor, FGF-7 and FGF-10 exert distinct effects on the developing lung. Rather than being chemotactic, FGF-7 in¯uences patterning by promoting epithelial growth. In lung organ cultures, FGF-7 induces formation of cyst-like structures with extensive cell proliferation (Simonet et al., 1995; Shiratori et al., 1996; Cardoso et al., 1997; Park et al., 1998). Similarly, when high levels of FGF-7 are targeted to the distal lung epithelium with an SP-C promoter/enhancer, transgenic mice show pulmonary malformations resembling cystic adenomatoid malformations (Simonet et al., 1995). In this case, dilation of epithelial tubules may be related to decreased expression of a-ENAC, causing a relative increase in non-CFTR-dependent ¯uid secretion in response to FGF-7 (Zhou et al., 1996a). Besides having an effect on patterning, FGF-7 is a differentiation factor for the developing lung. In lung epithelial cultures in the absence of mesenchyme or serum, exogenous FGF-7 can induce a type II cell-like phenotype (Cardoso et al., 1997). Furthermore, FGF-7 is required to induce transdifferentiation of the tracheal epithelium to a distal lung phenotype in mesenchyme-free tracheal cultures (Shannon et al., 1999). The role of FGF-7 in lung development may be shared with other regulators, since FGF-7 knockout mice have apparently no gross abnormalities in the lung (Guo et al., 1996). The involvement of FGFs in such diverse aspects of lung development is likely related to their multiple interactions with other regulatory molecules in the embryonic lung. As presented in other sections of this review, a variety of pattern regulators including Shh, BMPs, TGF-bs and EGF are also present in the lung (Heine et al., 1991; Warburton et al., 1992; Bellusci et al., 1996, 1997a,b). During early branching, it is apparent that FGF-10 occupies a central role in setting up the spatial coordinates for patterning the epithelial tubules. Early events involving the establishment of FGF-10 expression in the primitive mesenchyme are virtually unknown. The idea that transcription factor regulators of body plan, such as Hox genes, could play a role in this process is attractive. Nevertheless, despite the large number of Hox family members expressed in the early lung (Cardoso, 1995; Kappen, 1996), little is known about their precise role in lung morphogenesis. Further-

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

more, Hox±FGF interactions have not been reported in the lung so far. Targets of FGF-10 in the epithelium include the FGFR-2, but the downstream signaling molecules that mediate the chemotactic effect have not yet been identi®ed. Another key aspect of epithelial±mesenchymal interactions is that signals derived from the epithelium control mesenchymal gene expression, and by doing so, in¯uence epithelial patterning. This is illustrated by Shh, which is present at high levels in the distal lung epithelium and has several targets in the mesenchyme (Pepicelli et al., 1998). There is evidence that FGF-10 mRNA levels are negatively controlled by Shh. Low levels of FGF-10 are found in cultured lungs in which Shh expression is upregulated by RA treatment; FGF-10 is downregulated in lungs of transgenic mice overexpressing Shh (Bellusci et al., 1996). Treatment of embryonic lung mesenchymal cells with recombinant Shh markedly inhibits FGF-10 mRNA expression (Lebeche et al., 1999). Moreover, Shh knockout mice show a severe disruption of branching. Curiously, in the absence of Shh, FGF-10 expression is no longer restricted to focal areas, but becomes widespread throughout the distal mesenchyme. This observation raises the possibility that one of the roles for Shh signaling in lung branching is the local control of FGF-10 levels in the mesenchyme surrounding the bud tips. A model proposed by Bellusci et al. (1997b) predicts that, at early stages of lung development, a feedback interaction is established between the distal epithelium and FGF-10 expressing cells. The high levels of Shh at the tips activate signaling in these mesenchymal cells, via ptc and Gli, leading to downregulation of FGF-10 gene expression. As a result, the chemoattractant focus is extinguished, limiting local budding. The appearance of FGF-10-expressing cells in other sites reinitiates the cycle, resulting in formation of a new generation of buds. Another possible complementary mechanism to control lung budding involves antagonists of the FGF pathway. In Drosophila, the sprouty (spry) gene encodes a cysteine-rich protein that is downstream of bnl and btl, and, when spry is inactivated, enhanced tracheal branching is observed. Data from gain and loss of function mutants in ¯ies show that spry behaves genetically as a competitive inhibitor of bnl (Hacohen et al., 1998). Recently, four murine and three human spry homologues have been identi®ed and mspry2 and 4 are expressed in the developing lung (Hacohen et al., 1998; deMaximy et al., 1999; Tefft et al., 1999). Murine spry2 (mspry2) is expressed in the distal epithelium of the embryonic day 12 mouse lung, at a stage when epithelial buds are highly responsive to FGF-10 and when FGF-10 transcripts are present in the mesenchyme. Interestingly, abrogation of mspry2 expression with antisense oligodeoxynucleotides in lung organ cultures leads to a signi®cant enhancement in branching morphogenesis and surfactant protein gene expression (Tefft et al., 1999). Whether this represents a manifestation of a direct or indirect mspry± FGF-10 interaction in the developing lung remains to be

65

established. The presence of another family member, mspry4, widely expressed in the distal lung mesenchyme (deMaximy et al., 1999), suggests that modulation of FGF signaling by sprouties involves the establishment of complex regulatory loops. The exact mechanism of sprouty function is controversial (Fig. 6). Sprouty has been reported to function either as a secreted, extracellular FGF binding protein (Hacohen et al., 1998), or as an intracellular inhibitor of the Ras pathway (Casci et al., 1998) or possibly both (Kramer et al., 1999). Casci et al. (1998) have reported that sprouty is not normally secreted and that it binds, through a myristoylation site in its cysteine-rich domain, to the intracellular surface of the cell membrane, where it acts as an inhibitor of EGFR and FGFR tyrosine kinase pathways in wing morphogenesis. Similarly, Reich et al. (1999) reported that sprouty is a general inhibitor of tyrosine kinase signaling. Ectopic sprouty expression abolished activated MAP kinase expression in Drosophila embryos in response to the Heartless FGF receptor homologue and the EGF receptor. Moreover, sprouty is induced by EGFR activation in Drosophila ovarian follicle cells, restricting the ventral expression pattern of Gurken. Additionally, sprouty rescued embryonic lethality induced by activated Raf, suggesting that it may interact with Raf or downstream of it in the Ras-MAP kinase activation pathway. Sprouty has also been shown to bind in vitro to

Fig. 6. How mspry2, a murine homologue of the Drosophila sprouty gene, may hypothetically negatively regulate FGF-10 signaling in murine lung epithelial morphogenesis. (A) A schematic diagram of an isolated peripheral embryonic epithelial lung bud undergoing chemotactic attraction toward an FGF-10 soaked bead (based on ®gures in Park et al., 1998; Hogan, 1999). The alternative functions of mspry2 to inhibit FGF are shown: extracellular binding of FGF ligand (Hacohen et al., 1998) versus intracellular inhibition of FGF signaling (Casci et al., 1999; Kramer et al., 1999). (B) Schematic representation of the FGF signaling pathway as it passes through the cell membrane. Mpry2 is shown binding FGF-10 in the extracellular space and thus inhibiting ligand from binding FGFR-2 and/or heparan sulfate proteoglycans (HSPG). On the intracellular side of the membrane mspry2 is shown tethered to the cell membrane via its myristoylation site. In the latter location, mspry2 is shown exerting a negative in¯uence on the association between Grb2 and the activated FGFR-2, thus preventing Ras activation.

66

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

members of the Drosophila Ras signaling pathway. Sprouty binding targets include Drk, an SH2-SH3 adapter protein that is homologous to mammalian Grb-2, as well as Gap1, a Ras GTPase activating protein, but sprouty does not bind Sos, Dos, Csw, Ras1, Raf or Leo (Casci et al., 1998). Little is also known about FGF interactions with other growth factors during lung development. HGF and its receptor c-met/HGF, expressed respectively in the mesenchyme and epithelium of the developing lung, are candidate molecules to interact with FGFs during branching. Recombinant HGF potentiates the stimulatory effect of FGFs on budding in lung epithelial cultures (Ohmichi et al., 1998). Members of the TGF-b superfamily, including both the BMP and the TGF-b subfamilies, are also likely to interact with FGFs in the lung (Bellusci et al., 1996; Zhao et al., 1996; Zhou et al., 1996a,b,c; Lebeche et al., 1999). A large number of studies have shown that these factors are negative regulators of lung epithelial proliferation and differentiation. They may play an important balancing role counteracting the bud promoting effects of FGFs. FGF-10 and TGF-bs likely interact during lung morphogenesis, since both regulators are present in the mesenchyme. Both BMP4 and TGF-b appear to negatively regulate FGF-10 expression when applied on beads to cultured murine embryonic lungs (Lebeche et al., 1999). Increasing information originating from culture systems and mouse genetic manipulation will soon allow a better understanding of how FGF interactions with key partners in¯uence lung development. 6.3. HGF and c-met While HGF/scatter factor was originally isolated as a potent mitogen for hepatocytes, it also has potent motogenic, mitogenic and morphogenetic activities on epithelial cells, including respiratory epithelial cells in vitro (Sonnenberg et al., 1993; Shiratori et al., 1995). HGF is expressed in primitive lung mesenchyme, while its receptor, the c-met tyrosine kinase, is expressed in primitive lung epithelium, suggesting the possibility of inductive mesenchymal± epithelial interactions. Expression of HGF in Lx-1 lung cancer cells induces alveolar differentiation (Brinkmann et al., 1995). 6.4. IGF peptides, binding proteins and receptors The insulin-like growth factors (IGFs), their binding proteins and receptors are expressed in both rodent and human fetal lung (Batchelor et al., 1995; Lallemand et al., 1995a,b; Maitre et al., 1995; Schuller et al., 1995; RechtsBogart et al., 1996). There is little apparent change in the distribution or abundance of IGF-1 and -2 and their receptors during gestation. However, the IGFbps 1±6 are differentially regulated; thus, it is possible that the latter may play a key role in mediating temporo-spatial IGF signaling, particularly the regulation of rates of cellular proliferation (Maitre et al., 1995). Mice carrying the null mutation of the insulin-like growth

factor-1 (IGF-1) gene are dwarfed to 60% of normal birth weight (Liu et al., 1993). Depending on the genetic background, some of the IGF-1 null mutants die in the neonatal period, while others survive to reach adulthood. On the other hand, null mutants for the cognate type 1 insulinlike growth factor receptor (IGF1r) gene always die at birth of respiratory failure and exhibit a more severe growth de®ciency (45% of normal birth weight). Dwar®sm is further exacerbated (30% of normal size) in IGF-1 and IGF-2 double null mutants and in IGF1R and Igf-2 double null mutants. There does not appear to be a gross defect in primary branching morphogenesis per se; the lungs merely appear hypoplastic (Liu et al., 1993). However, it also seems likely that IGF signaling may play a key role in facilitating signaling by other peptide growth factor pathways involved in lung morphogenesis, since IGF1r signaling function is required for both the mitogenic and transforming activities of the EGF receptor (Coppola et al., 1994). 6.5. PDGF peptide and cognate receptor isoforms PDGF peptides are dimeric ligands formed from two peptide chains, A and B. The PDGF-AA and PDGF-BB homodimers and PDGFR are present in embryonic mouse lung and are differentially regulated in fetal rat lung epithelial cells and ®broblasts (Buch et al., 1994). PDGF-AA regulates both DNA synthesis and early branching in early mouse embryonic lung epithelium in culture. Abrogation of PDGF-A expression with antisense oligodeoxynucleotides or PDGF-A blocking antibodies both decrease DNA synthesis and hence the size of early embryonic mouse lung in culture as well as interfering with early branch point formation (Souza et al., 1995). On the other hand, abrogation of PDGF-B chain expression with antisense oligodeoxynucleotides reduces the size of the epithelial component of early embryonic mouse lung explants, but does not reduce the number of branches. PDGF-A homozygous null mutant mice are lethal either prenatally before E10, or postnatally. Surviving PDGF-A 2/2 mice develop pulmonary emphysema secondary to failure to septate alveoli (Bostrom et al., 1996). This phenotype is apparently caused by loss of alveolar myo®broblasts and associated elastin ®ber deposition. Since PDGF-a receptors are expressed in the lung at the location of putative alveolar myo®broblasts, and the latter were speci®cally absent in PDGF-A null mutants, it appears that PDGF-A chain expression is essential for the ontogeny of pulmonary alveolar myo®broblasts. Thus, PDGF signaling appears to play a permissive role for epithelial DNA synthesis during embryonic life and an instructive role for the ontogeny of myo®broblasts, elastin synthesis by the latter cell lineage and hence alveolarization in postnatal life. It is also interesting to note that PDGF-B chain expression is essential for the ontogeny of renal mesangial cells. Thus, myo®broblastic cells appear to have key morphogenetic functions in the formation of tubular epithelial appendages both in the lungs and the kidney.

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

6.6. Bombesin-like peptides and GRP-receptors Gastrin-related peptide (GRP) is the mammalian equivalent of bombesin and is produced by pulmonary neuroendocrine (PNE) cells in the adult lung. Expression of GRP and its cognate receptor peaks during fetal life in the rapidly proliferative phase of airway epithelial development (Wang et al., 1996). At earlier stages of development in the mouse, GRP is expressed in undifferentiated epithelial lineage precursor cells (Wuenschell et al., 1996). Blockade of GRP action by immunoperturbation or pharmacological blockers both in vivo and in culture retards lung epithelial development (Aguayo et al., 1994). Thus, GRP-receptor signaling may play a physiological role in the induction of lung morphogenesis, particularly at later fetal stages when pulmonary neuroendocrine cells (PNE cells) have differentiated (Sunday et al., 1990, 1993). CD10/neutral endopeptidase also appears to play a key role in activating bombesinlike peptides (King et al., 1995) (also see Section 8). 6.7. TGF-b family peptides and cognate receptors TGF-b 1, 2 and 3 peptides and the TGF-b type I and type II receptors are expressed and differentially distributed in the embryonic and fetal lung (Pelton et al., 1990; Zhou and Young, 1995; Zhao et al., 1996). TGF-b1 and TGF-b2 both inhibit pulmonary branching morphogenesis in culture, although TGF-b2 is considerably more potent than TGFb1 (Serra et al., 1994; Zhao et al., 1996). The expression of pRb is not necessary for the inhibitory effects of TGF-b on branching to be transduced, since TGF-b1 inhibits branching in pRb 2/2 embryonic lungs to the same extent as in wild type. However, N-myc expression is suppressed by TGF-b1 in wild type lungs, but is not suppressed by TGF-b in the pRb null mutant, indicating that TGF-b is necessary for the inhibitory effect on N-myc expression to occur (Serra and Moses, 1995). Interestingly, gene targeting of the n-myc locus has produced a hypoplastic, neonatal lethal lung phenotype as well as a lethal hypoplasia of sub-endocardial myoblasts (Moens et al., 1993). Perhaps the N-myc knockout prevents the pulmonary epithelium from expanding to a suf®cient surface area to support gas diffusion postnatally. TGF-b3 null mutation also results in a speci®c immatureappearing neonatal lung phenotype which is rapidly fatal in newborn mice (Kaartinen et al., 1995). Unlike the normal neonatal lung phenotype found in TGF-b1 null mutant mice, which has been attributed to maternal transplacental rescue, TGF-b3 null mutation appears to be refractory to maternal transplacental rescue. TGF-b3 gene expression is also strongly induced in response to corticosteroid treatment of fetal lung ®broblasts (Wang et al., 1995), suggesting the hypothesis that the well recognized maturation enhancing effects of glucocorticoids on late fetal lung may in part be mediated by stimulation of temporo-spatially restricted TGF-b3 gene expression. This striking similarity between

67

the neonatal lethal, fetal lung dysplasia phenotypes of TGFb3 2/2 and corticotropin-releasing hormone (CRH) 2/2 mice further suggests intriguing parallels between the TGF-b3 and CRH mediated signaling patterns (Muglia et al., 1999). Both phenotypes suffer from excessive mesenchyme later in gestation, after the mesenchyme should have involuted. However, TGF-b3 null mutation does not block the maturational effects of glucocorticoids (Kaartinen et al., 1995). Evidence from normal and targeted misexpression studies in mice further suggests that BMP4, another TGF-b family peptide, also plays a role in embryonic lung morphogenesis (Bellusci et al., 1996). Misexpression of BMP4 driven by the SP-C promoter in transgenic mice results in lungs that are smaller than normal with grossly distended terminal buds and large air-®lled sacs at birth. There is also reduced expression of alveolar epithelial cell differentiation markers (SP-C), but not of Clara cell markers (CC-10). However, targeted misexpression of TGF-b1 using the same SP-C promoter system also results in a neonatal lethal, hypoplastic pulmonary phenotype with decreased saccule formation and epithelial differentiation (Zhou et al., 1996c). Taken together, these ®ndings suggest that TGF-b family peptide over-expression in vivo merely re¯ects a default negative regulatory effect on morphogenesis similar to that elicited in culture (Serra et al., 1994; Serra and Moses, 1995; Zhao et al., 1996). On the other hand, abrogation of TGF-b type II receptor signaling, either with antisense oligodeoxynucleotides, or with blocking antibodies stimulates lung morphogenesis two- to three-fold and increases expression of TTF-1 and SP-C (Zhao et al., 1996). Thus, endogenous autocrine/paracrine TGF-b signaling through the TGF-b type II receptor appears to negatively regulate lung organogenesis. The negative effect of TGF-b signaling through the TGF-b type II receptor on cell cycle progression in pulmonary epithelial cells probably plays a major role in the latter inhibitory effect by limiting expansion of the epithelial surface area. Recent experiments indicate that TGF-b signaling through the TGF-b type I receptor speci®cally instructs the formation of branch points. Abrogation of TGF-b type I receptor expression with antisense oligodeoxynucleotides signi®cantly reduces the formation of new branch points by E11 early embryonic mouse lung in culture, resulting in a phenotype characterized by long, tubular-appearing airways devoid of new branch points. This effect is associated with decreased ®bronectin and matrix Gla protein gene expression and failure to form condensations of extracellular matrix containing ®bronectin as would be expected to occur at sites where a new branch point is about to form (Heine et al., 1990). Moreover, the growing points of the tubular-appearing airways are virtually devoid of ®bronectin. Thus, the regulation of ®bronectin gene expression by the TGF-b type I receptor apparently plays a key role in the molecular basis of lung morphogenesis by instructing the

68

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

formation of new airway branch points and the localized deposition of extracellular matrix components. TGF-b ligand-interacting proteins and proteoglycans such as betaglycan (TGF-b type III receptor), biglycan, decorin, ®bromodulin and endoglyn may affect pulmonary morphogenesis (Hildebrand et al., 1994). As an example, endoglyn is a dimeric TGF-b1 and -3 binding protein of endothelial cells, modulates cellular responses to TGF-b1 and can form heteromeric complexes with TGF-b signaling receptors (Yamashita et al., 1994; Lastres et al., 1996). Interestingly, endoglyn is the gene for Osler±Weber± Rendu hereditary telangiectasia type 1, a condition which is characterized by large intra-pulmonary arteriovenous malformations (McAllister et al., 1994). It is also interesting to note that massively dilated pulmonary vessels with thin or absent smooth muscle layers were a prominent pathologic feature of the neonatal lethal TGF-b3 null mutant phenotype (Kaartinen et al., 1995). We have recently shown that abrogation of betaglycan expression using antisense oligodeoxynucleotide gene knock-down stimulates lung morphogenesis in culture and strongly inhibits the effectiveness of exogenous ligands, particularly TGF-b2, to inhibit lung morphogenesis in culture (Zhao et al., 1998a,b). The downstream TGF-b receptor signaling pathway is not as yet completely clear. However, several molecules have been found to impart TGF-b signals. Studies have shown that TGF-b effects on DNA synthesis are pertussistoxin sensitive, implicating G-proteins as signal transducers (Murthy et al., 1988; Kataoka et al., 1993). Immunophilin FKBP-12, a target of the macrolides FK506 and rapamycin, was found to interact with type I receptors of the TGF-b family (Wang et al., 1994). A WD-domain protein, TRIP-1 (TGF-b-receptor interacting protein-1), was found to associate speci®cally with type II TGF-b receptor in a kinase-dependent way, which may be analogous to the function of GRB-2 of receptor tyrosine kinases (Chen et al., 1995). Recently, a mouse protein kinase of the mitogen activated protein kinase kinase kinase (MAPKKK) family, TAK1 (TGF-b-activated kinase 1), was found to be speci®cally activated by members of the TGF-b superfamily of ligands and involved in regulation of transcription by TGFb (Yamaguchi et al., 1995). A group of genes homologous to Drosophila Mad (mothers against dpp), a gene that interacts with the TGFb superfamily member encoded by decapentaplegic (dpp), has been found to be a cytoplasmic effector for the TGF-blike signals (Sekelsky et al., 1995; Liu et al., 1996; Newfeld et al., 1996; Wiersdorff et al., 1996). Brie¯y, following ligand binding to the TGF-b type II receptor, TGF-b I receptor is recruited to enter a heteromeric complex with TGF-b IIR, placing the TGF-b IR in position to be transphosphorylated by the constitutively active TGF-b IIR intracellular serine/threonine kinase domain. Smads 2 and 3 can bind the TGF-b receptor heteromeric complex and are phosphorylated by the activated type I receptor serine/threonine kinases. Activated Smads 2 and/or 3 can then form

complexes with Smad 4, separate from the receptors (Zhang et al., 1997b). These complexes between activated Smads 2 or 3 and Smad 4, can translocate to the nucleus, where they can act as transcriptional activation or repression factors at speci®c promoters (Zhang et al., 1996). Smads 6 and 7 are inhibitors of activation of Smads 2 and 3. Smad 7 expression is rapidly induced following TGF-b ligand signaling and thus is considered to function as a negative feedback element in the TGF-b signaling pathway. Abrogation of Smads 2 and 3, or 4 expression, using antisense oligodeoxynucleotide gene knock-down, results in a strong gain of function phenotype for lung branching morphogenesis of early murine embryonic lung in culture. The Smads 2 and 3 or 4 abrogation phenotype is similar to that obtained after abrogation of either TGF-b IIR or IIR signaling (Zhao et al., 1996, 1998a,b). 6.8. BMPs and BMPRs The BMPs comprise a branch of the TGF-b superfamily that also plays a key role in development. Several BMP ligands and BMPRs including BMP3, 4 and 7 as well as type I BMPR are expressed during embryonic lung development (Bellusci et al., 1996; Takahashi and Ikeda, 1996). BMP4 mRNA is localized at high levels in the epithelium of distal tips of terminal buds, with lower levels in the adjacent mesenchyme (Bellusci et al., 1996). These loci of BMP expression overlap with the expression domains of some other important morphogenetic signaling molecules including HNF-3b, Wnt-2, Shh and FGF-10. Misexpression of BMP4 under the control of the SP-C promoter/enhancer in transgenic mice results in lungs that are smaller than normal with grossly dilated terminal sacs, which do not support gas exchange at birth. There is also a reduction in BrDU incorporation into peripheral epithelial cells as well as reductions in the number of differentiated type II epithelial cells (Bellusci et al., 1996). These ®ndings suggest that negative regulation of BMP signaling by homologues of such molecules as DAN and Cerberus must play equally important roles in lung morphogenesis (Weaver et al., 1999). Also, since BMPs and Shh are co-expressed in the same domains, and since Decapentaplegic, the Drosophila BMP homologue, is regulated by the Hedgehog signaling pathway, it seems possible that BMP±Shh interactions may prove to play key roles in lung morphogenesis. Recently published data on ®broblast growth factor interactions suggest that Shh, TGF-b1 and BMP4 all counteract the bud-promoting effects of FGF-10 (Lebeche et al., 1999). 7. Extracellular matrix components during lung development Laminins, entactin/nidogen, type IV collagen, perlecan, SPARC and ®bromodulin are extracellular matrix (ECM) components of the basement membrane which play an important role in the mediation of epithelial±mesenchymal

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

cell matrix interactions during fetal lung morphogenesis. Basement membrane components are differentially expressed, and have a speci®c cell distribution during lung morphogenesis. ECM components may not only be a support of the tissue architecture, but may also play an active role in the modulation of cell proliferation, and cell differentiation (Lwebuga-Mukasa, 1991). Basement membrane components also play a dynamic role as a barrier and reservoir of growth factors, which in turn regulate epithelial and mesenchymal cell proliferation. Absence or inhibition of the interaction of epithelial cells with the basement membrane has a direct consequence in the failure of normal lung development (Minoo and King, 1994; Hilfer, 1996). Similarities exist between the composition of the basement membrane during lung morphogenesis with that found in other tissues (Paulsson, 1992; Dunsmore and Rannels, 1996; Timpl, 1996; Timpl and Brown, 1996). However, temporo-spatial expression of the ECM components and the presence of speci®c isoforms may play a key role in the development of tissue-function speci®city. 7.1. Laminins Laminins (LNs) are a family of extracellular matrix glycoproteins involved in cell adhesion, migration, proliferation and differentiation during tissue development. LNs are composed of three chains, one central (a) and two lateral (b and g) that are linked by disul®de bonds to form a crossshaped molecule (Burgerson et al., 1994). To date ®ve a, three b and three g chain isoforms have been described, which suggests that their combination can lead to approximately 30 variants of LN (Ehrig et al., 1990; Bernier et al., 1994; Vuolteenaho et al., 1994; Galliano et al., 1995; Iivanainen et al., 1995a,b, 1997, 1999; Pierce et al., 1998; Koch et al., 1999). LN1 (a1,b1,g1) was the ®rst LN variant identi®ed and characterized during fetal lung development. However, the recent identi®cation of new chain isoforms in fetal lung indicates that at least 10 LN variants participate in a spatio-temporal manner during lung development (Table 2). Their speci®c roles remain to be determined. LN1 (a1,b1,g1) is crucial in the organization of the basement membrane in the earliest stage of embryonic development. During lung development, LN1 is expressed by epithelial and mesenchymal cells (Klein et al., 1990; Schuger et al., 1990a,b, 1992; Durham and Snyder, 1995, 1996; Lallemand et al., 1995a,b). Studies focusing on the a1 chain have indicated that epithelial±mesenchymal cell interaction is necessary for the expression of this chain isoform (Schuger et al., 1997). The a1 chain has been found principally localized in the basement membrane at the epithelial± mesenchymal interface, with a predominant distribution in speci®c zones. The LN a1 chain has also been identi®ed around a few mesenchymal cells. Characterization of the distinct domains of the a1 chain isoform have shown that

69

Table 2 ECM components in the basement membrane during lung morphogenesis Component a

Laminins LN1 LN2 LN4 LN6 LN7 LN8 LN9 LN10 LN11 Nidogen Collagen IV Proteoglycans (Perlecan)

Chain composition

Mr (kDa)

Heterotrimer (a1,b1,g1) (a2,b1,g1) (a2,b2,g1) (a3,b1,g1) (a3,b2,g1) (a4,b1,g1) (a4,b2,g1) (a5,b1,g1) (a5,b2,g1) Monomer Heterotrimer, a1±6 (IV) Monomer

400±840

150 550 400±500

a The type of laminin variants distributed in the basement membrane during lung morphogenesis is proposed in accordance with the chain isoforms characterized. Only the role of LN1 and LN2 variants during fetal lung development have been reported (see text for details). The recently described g3 chain isoform indicates that at least 10 LN variants are expressed during fetal lung development. The LN variants containing the g3 chain remains to be determined.

a domain in the cross region of the a1 chain is involved in the regulation of lung epithelial cell proliferation (Schuger et al., 1992). Recent studies have also indicated that LN1 may be involved in the maintenance of smooth muscle cell differentiation (Schuger et al., 1997; Yang et al., 1998). The a2 chain isoform found in LN2 (a2,b1,g1) and LN4 (a2,b2,g1) has been described during mouse and human lung development (Bernier et al., 1994; Virtanen et al., 1996; Flores-Delgado et al., 1998). In the pseudoglandular stage of lung development, the a2 chain was observed to have a speci®c distribution around epithelial and smooth muscle cells. In the canalicular stage in human fetal lung, the a2 chain was present only in smooth muscle cells of bronchi (Virtanen et al., 1996). We have reported that mesenchymal cells isolated from fetal mouse lung express the a2 chain isoform. In addition, we have demonstrated that LN2 plays a role in cell adhesion of a subpopulation of embryonic mesenchymal cells bearing a myo®broblast phenotype. The a3 chain, found in LN6 (a3,b1,g1) and LN7 (a3,b2,g1) variants, has been localized principally in fetal lung epithelial cells (Virtanen et al., 1996; Miner et al., 1997; Richard et al., 1998). Mutations of the a3 chain produce abnormal assembly of epithelial basement membranes. The a4 chain, found in LN8 (a4,b1,g1) and LN9 (a4,b2,g1) variants, has been described to be highly expressed in lung and heart tissue during mouse development (Iivanainen et al., 1995a,b, 1997; Richards et al., 1996; Frieser et al., 1997). Preliminary results from our laboratory and others (Miner et al., 1997; Richard et al., 1998) suggest that the LN a4 chain may play a role principally in the organization of lung mesenchyme. The LN a4 chain has

70

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

been localized principally around vessels in fetal lung. The role of LN8 and LN9 variants during fetal lung development remains to be elucidated. The a5 chain, found in LN10 (a5,b1,g1) and LN11 (a5,b2,g1) has been indicated to be abundantly expressed during fetal lung morphogenesis (Miner et al., 1995, 1998; Richard et al., 1998). Mouse embryos bearing mutated LN a5 chain isoform die by E17. Severe fetal defects in internal organs, exencephaly, and syndactyly are observed in these embryos. In particular, fetal lung showed a poor lobe septation and bronchiolar branching, suggesting that the LN a5 chain isoform might be the most indispensable LN variant for lung branching morphogenesis. Expression of the b1 and g1 chains are independent of the a chains. A constant expression of b1 and g1 is observed during fetal lung development (Durham and Snyder, 1995). These two chains also have a role in cell adhesion. The globular domains near the N-terminal of b1 and g1 chains participate in the regulation of cell polarization (Schuger et al., 1995, 1996). In addition the b1 globular domain may play an important role in organotypic rearrangement when isolated mesenchymal and epithelial cells from embryonic mouse lung are cultured together (Schuger et al., 1995). The b2 chain isoform found in LN4 and LN7 variants has been indicated to be highly synthesized at the time of alveolar type II cell differentiation in rabbit fetal lung development (Durham and Snyder, 1996). Immunohistochemistry studies have demonstrated that the LN b2 chain isoform is localized in the basement membrane of prealveolar ducts, airways, and smooth muscle cells of airways and arterial blood vessels. This chain isoform has also been localized beneath the epithelial type II cells. Targeted mutagenesis of the LN b2 chain isoform has been shown to affect both neuromuscular and renal function. Further studies will be necessary to elucidate the biological role of the LN b2 chain isoform during lung morphogenesis. The expression of the LN g1 chain isoform has been localized principally in epithelium in human fetal lung, and mesenchyme and epithelium in murine fetal lung (Thomas and Dziadek, 1994; Lallemand et al., 1995a,b). Recent reports have demonstrated the expression of the g3 chain isoform in lung tissue (Iivanainen et al., 1999; Koch et al., 1999). During fetal lung development this chain has been localized in the lumen of the epithelium. It remains to identify the speci®c LN variants that may contain the g3 chain isoform and its role during lung development. 7.2. Nidogen Nidogen (150 kDa) is a constituent of the basement membranes. Nidogen binds to the g1 and g3 chains of LN, and forms a link between LN and collagen IV (Reinhardt et al., 1993; Dziadek, 1995; Koch et al., 1999). Nidogen is actively synthesized by mesenchymal cells during fetal lung development, which suggests that nidogen has a key role in the organization of the basement membrane

during lung morphogenesis (Senior et al., 1996). Inhibition of the interaction of Nidogen with LN affects the progression of lung development (Ekblom et al., 1994; Dziadek, 1995; Senior et al., 1996). The domain sequences NIDPNAV and NVDPNAV, found in LN g1 and LN g3 chain isoforms, respectively, are involved in the interaction with nidogen. Susceptibility of nidogen to degradation by matrix metalloproteinases may contribute in the remodeling and degradation of the basement membrane (Mayer et al., 1993). 7.3. Collagen IV Collagen IV is a key structural component of all basement membranes. It consists of three chains of 180 kDa bound in a triple helix structure. It is frequently identi®ed in a complex of four molecules, associated in a spider-shaped structure that facilitates the formation of a network structure. Collagen IV chain isoforms have been identi®ed in basement membrane of fetal lung (Chen and Little, 1987; Thomas and Dziadek, 1994). Six collagen IV chain isoforms in human, and ®ve in mouse have been described (Hudson et al., 1993; Miner and Sanes, 1994). Their expression together with nidogen and LNs follows a spatial and temporal distribution that may produce speci®c micro-environments in the basement membrane. These tissue-speci®c environments may play a role in cell proliferation and differentiation. Immunohistochemical studies indicated that a1 and a2 (IV) isoforms of collagen associate with the LN b1 chain isoform in fetal development (Miner and Sanes, 1994), and the globular domain NC1 of collagen IV may interact with nidogen (Dziadek et al., 1985). The speci®c role of collagen IV molecules in the basement membrane during lung morphogenesis remains still to be fully elucidated. 7.4. Proteoglycans Proteoglycans comprise a core protein with sulfated carbohydrate side chains. They function as ¯exible structures in the organization of the basement membrane and may also play an important role as a reservoir for growth factors, water and cations. Perlecan is a predominant proteoglycan in the basement membrane. It is composed of an approximately 450 kDa core protein with three heparan sulfate chains. Interaction of heparan sulfate proteoglycans with a speci®c domain of the b1 chain isoform of LN contributes directly to epithelial cell polarization and organotypic arrangement (Schuger et al., 1996). Perlecan has also been involved in the control of smooth cell proliferation and differentiation. An increase in proliferation of fetal lung smooth muscle cells is accompanied by a highly increased synthesis of perlecan (Belknap et al., 1999). 7.5. Fibromodulin Fibromodulin is an extracellular matrix protein that is

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

associated with elastin ®bers and is believed to serve as a ligand for avb3 integrin. Fibrillin-2 is expressed at the mesenchymal±epithelial interface in embryonic rat lung prior to E13 and later is expressed around the trachea and major airways. Treatment of embryonic rat lung explants with antisense ®brillin-2 resulted in smaller lung explants with loose expanded mesenchyme compared to controls (Yang et al., 1999). The ®brillin-2 loss of function phenotype may be related to perturbation of elastin binding. 7.6. SPARC SPARC (secreted protein, acidic and rich in cysteine) is an anti-adhesive glycoprotein of the extracellular matrix. It has a wide distribution in the basement membrane during lung morphogenesis. Recent studies in which SPARC interactions were blocked by speci®c antibodies to this molecule indicate that it plays a role in lung branching morphogenesis (Strandjord et al., 1995).

8. Pulmonary neuroendocrine cells Pulmonary neuroendocrine (PNE) cells occur both as single cells and as organized clusters of cells called neuroepithelial bodies (NEBs) (Lauweryns and Peuskens, 1972). The PNE cells contain characteristic small, dense-cored granules that cluster at the basal pole of the cell and are secreted by exocytosis in such a way that the contents might act upon nerve ®bers, peribronchial smooth muscle, or capillaries underlying the basement membrane (Becker, 1984). The earliest PNE cells can be identi®ed cytologically in humans at 8 weeks gestation when airway development is just beginning (Cutz et al., 1984), and long before sensing or responding to airway gasses could possibly have any relevance. This observation has lead to speculation that PNE cells play some role in the development of the lung during fetal life. This idea is supported by the fact that PNE cells synthesize and secrete a number of peptides that exhibit growth factor activity. Perhaps the most studied of these is GRP. GRP, or its amphibian homologue bombesin, has been shown to be mitogenic for several cultured cell types (Rozengurt and Sinnett-Smith, 1983; Willey et al., 1984; Speirs et al., 1993). Sunday and co-workers have shown that bombesin stimulates branching morphogenesis and lung maturation in culture and in vivo (Sunday et al., 1990; King et al., 1995). In addition, work by Hoyt et al. (1991) showed that PNE cell clusters in newborn hamster lung are the focal points for regions of increased cell division further supporting the idea that PNE cells produce one or more paracrine growth factors that stimulate proliferation of the surrounding lung epithelium. Despite this evidence, however, it remains unclear whether the proliferative zones surrounding NEBs are a result of some action of the differ-

71

entiated PNE cells on the surrounding epithelium or whether they are part of the epithelial response to whatever local cues are responsible for the differentiation of PNE cells at these sites. Because of their neuroendocrine phenotype, a neural crest/neuroectodermal origin seemed possible for these cells (reviewed by Cutz, 1982). There is, however, no evidence to support a neural crest origin and the available evidence is consistent with the current prevailing view that these cells are true derivatives of the lung epithelium. Early embryonic lung epithelium expresses GRP in a wide spatial distribution, which only becomes restricted to PNE cells late in fetal life (Wuenschell et al., 1996). PNE cells show the same centrifugal pattern of differentiation exhibited by the other elements of the lung epithelium, appearing ®rst in more proximal airways and progressively later in more distal regions (Cutz et al., 1984; Carabba et al., 1985). Also, Hoyt et al. (1990) observed that the [ 3H]thymidine labeling pattern of differentiating PNE cells was consonant with that of the immediately surrounding epithelium and not with that of the intrinsic pulmonary neurons, which are neural crest-derived. Finally, the lung is embryologically derived from the primitive gut and the PNE cells can be regarded as analogous to the neuroendocrine cells found in the intestinal tract. The evidence that at least some of these enteric neuroendocrine cells are derived from endoderm and not from neural crest is unequivocal (reviewed byWeston, 1984). Another intriguing clue to PNE cell origins has come from recent work that has identi®ed an apparent differentiation switch gene for this cell type. The gene is the homologue of achaete-scute, MASH1 (mammalian achaete-scute homologue 1), which encodes a basic helix-loop-helix class transcription factor. The achaete-scute gene was originally identi®ed in Drosophila where it controls differentiation of some types of neurons. The embryonic stem cell/homologous recombination technique has been used to produce mutant mice lacking functional copies of MASH1 and analysis of these animals has shown that the gene controls differentiation of certain subsets of neurons in mice also (Guillemot et al., 1993). Recently, Borges et al. (1997) have examined the lungs of these mutant mice and found that they completely lack PNE cells. Other lung epithelial cell types remained normal and, interestingly, cGRP-immunoreactive autonomic neurons were still present. There is as yet, however, no clue regarding the nature of the local cues that are responsible for turning on MASH1 in some epithelial cells and not others, thereby determining the exact locations and spacing of PNE cells and NEBs within the developing airways. It is interesting that NEBs are located frequently, though not exclusively, at airway branch points (Youngson et al., 1993). One potential cue for PNE cell formation, the presence of neuronal processes, can be ruled out since observations of developing lungs in culture have shown that neurons are not necessary for differentiation of NEBs (Carabba et al., 1985).

72

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

9. Lung neovascularization Lung vascular development involves a delicate balance between positive and negative mediators of vascularization. Endothelial cell differentiation, commitment, and migration are dictated by a diverse group of molecules including extracellular matrix proteins, cell adhesion receptors, growth factors and their receptors (Baldwin, 1996). Blood vessels develop from blood lakes present in the mesenchyme of the embryonic lung at day 9 in the mouse. The isolated lakes then coalesce to form irregular sinusoidal structures. At 12 days gestation, sprouting of arteries and veins from central pulmonary vessels is already occurring. Between 13 and 14 days gestation, connections between the central and peripheral vascular structures can be identi®ed (deMello et al., 1997). The canalicular stage of murine lung epithelial development (E16.5±17.4 days) is also known as the vascular stage. The number of terminal sacs and vascularization increase during the canalicular stage, while type I and II epithelial cells differentiate (Burri, 1974; Kauffman, 1981; Tenhave-Opbroek, 1981; Meyrick and Reid, 1982; Thurlbeck, 1991). During the canalicular or vascular stage, the lung is canalized with multiplication of capillaries. The developing pulmonary arteries and veins follow the development of the branching airways but lag behind it somewhat (deMello et al., 1997). This is associated with ¯attening out of the cuboidal, glycogen-rich peripheral epithelial precursor cells, so that a thin air±blood barrier/interface is formed (Burri, 1997). This stage of transition is pivotal in the normal progression of lung morphologic and neovascular development. Removal from the fetal environment during this crucial time period, as demonstrated by premature delivery of human or baboon, can lead to the marked changes in lung morphologic development, termed bronchopulmonary dysplasia (Reid, 1979). Abnormal lung neovascularization has been noted in infants with bronchopulmonary dysplasia (Coalson et al., 1997). Pulmonary vascular formation is an active process, involving not only proliferation and differentiation of vascular structures, but also regression and stasis of these structures (Scavo et al., 1998). Upon examination of the two processes, termed vasculogenesis and angiogenesis, responsible for lung neovascularization (Poole and Cof®n, 1989; Cof®n et al., 1991) it becomes apparent that direct epithelial±mesenchymal interactions are essential for vessel formation. Vasculogenesis is de®ned as the trans-differentiation and organization of endothelial cells into vascular structures. This occurs from randomly distributed mesodermal cells that trans-differentiate into endothelial cells (EC), proliferate, and organize into multicellular structures arranged in a single layer around a central lumen. As noted above, this occurs in the periphery of the early embryonic lung. Angiogenesis is the process of extension of previously formed vessels into under-vascularized regions, where

differentiated EC proliferate, sprout from previously formed vessels, and form new vascular structures. This process connects the main pulmonary trunks with the peripheral vasculature. Sundell and Roman (1994) originally indicated that it was in the canalicular stage of lung development that there was an increase in the number of (endothelial) cells that express von Willebrand's factor (vWF). These cells coalesced with other vWF-positive cells to form vascular structures by about day 16 of murine gestation (Sundell and Roman, 1994). However, deMello et al. (1997) subsequently demonstrated that peripheral vasculogenesis as well as central angiogenesis both contribute concurrently to form the lung vasculature. Communication between the two respective vascular networks occurs rarely in the mid-pseudoglandular stage, but a gradual increase in communication progresses until a complete vascular circuit is achieved by the beginning of the terminal sac stage. Pulmonary endothelial cell differentiation is poorly understood, but recent studies elucidating the expression pattern of ¯k-1 within the splanchnopleuric mesoderm, from which lung vasculature is derived, suggest that these receptors for vascular endothelial growth factor (VEGF) play a key role in facilitating vessel formation (Gebb and Shannon, 1998). Evidence implicating VEGF (Drake and Little, 1995; Flamme et al., 1995a,b) and its receptors ¯k-1 and ¯t-1 (Yamaguchi et al., 1993; Gebb and Shannon, 1998) as facilitators of vessel formation in the embryo is accumulating. Excess VEGF leads to a massive fusion of vessels and obliteration of normally avascular zones. Studies in the quail by Drake et al. (1998) suggest that the development of the vascular epithelium is dependent on integrin-mediated cell adhesion for normal endothelial tube formation. Further insight into factors in¯uencing neovascular formation is being gained through the formation of blood-island-containing cystic embryoid bodies (Risau and Lemmon, 1988; Risau et al., 1988). However, there is relatively little known regarding the speci®cs of neovascularization in lung morphogenesis. Interactions between the epithelium and mesenchyme are contributing to lung neovascularization. Recent discoveries support the theory that the neovascularization is crucial in normal lung formation. Misexpression of VEGF in a transgenic mouse under the control of the SP-C promoter/enhancer induces gross abnormalities in lung morphogenesis and an increase in peritubular vascularity, with a concomitant decrease in both epithelial acinar tubules and mesenchyme (Zeng et al., 1998). Thus, it is possible that the neovascularization process within the lung contributes to normal lung development. The importance of the balance between angiogenic versus anti-angiogenic proteins was further supported by the recent identi®cation of thalidomide as an anti-angiogenic protein (D'Amato et al., 1994). Thus, it was postulated that the striking phocomelic limb defects that occur in the develop-

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

ing human fetus exposed to thalidomide in utero are secondary to inhibition of blood vessel growth in the developing fetal limb bud. However, thalidomide does not seem to grossly affect human lung morphogenesis, as evidenced by the case of the internationally famous German bass-baritone Thomas Quasthoff, whose intrauterine exposure to thalidomide caused severe phocomelia, but thankfully did not abrogate his amazing vocal artistry (Carriaga, 1999; Hohenadel, 1999). Endothelial monocyte activating polypeptide (EMAP) II (Kao et al., 1992, 1993, 1994) is a potent anti-angiogenic factor in tumor vascular development (unpublished data). Recent ®ndings suggest a functional role for EMAP II as a negative regulator of neovascularization in the developing lung (Schwarz et al., 1999). EMAP II expression is elevated during the early, minimally vascular stage of murine lung development and its expression is localized in a splanchnopleuric distribution, compatible with the region where pulmonary vascular development will subsequently occur. EMAP II expression then rapidly declines during the vascularization stage and is then localized to a perivascular distribution (Schwarz et al., 1999). In addition, new results indicate that EMAP II inhibits fetal lung neovascularization and signi®cantly alters lung epithelial morphology in an in vivo organ explant model. Additionally, EMAP II inhibits epithelial islet formation in murine embryonic lung epithelial±mesenchymal cell cocultures. Thus, it appears that EMAP II functions at the level of epithelial±mesenchymal±endothelial cell interaction to regulate both lung vasculogenesis and epithelial morphogenesis. Therefore, negative modulators of neovascularization appear likely to play a key role in pulmonary vascularization. Furthermore, neovascularization appears to have a direct effect on lung morphogenesis, where signi®cant alterations in vessel formation in the developing lung have a profound effect on morphologic development of the distal lung epithelium. This is supported by recent reports indicating that newborn mouse pups treated with fumagillin and thalidomide underwent an arrest in fetal lung development (Jakkula et al., 1999), suggesting that neovascularization is crucial in lung development. Furthermore, lungs from newborn mice treated with antibodies to ¯t-1 (one of the VEGF receptors) for 5 days were reduced in size and displayed immaturity with a less complex alveolar patterning (Gerber et al., 1999). These ®ndings are consistent with those seen in the VEGF misexpressing transgenic mouse, where gross abnormalities in lung morphogenesis are associated with an increase in peritubular vascularity with a decrease in acinar tubules and mesenchyme (Zeng et al., 1998). Again, these ®ndings strongly support the hypothesis that lung neovascularization exerts a key in¯uence on lung morphologic development. Alteration of the normal balance between angiogenic versus anti-angiogenic proteins certainly appears to have a profound effect on lung development.

73

10. What mediates the speci®city of lung morphogenesis? It is becoming readily apparent that many of the same morphogenetic signaling pathways play key inductive roles in morphogenesis, not only of branched organs such as the lung, kidney and breast, but also in seemingly quite distinct morphogenetic paradigms, such as limb and tooth morphogenesis (Hogan, 1999; Warburton et al., 1999). Thus, the question of what mediates the speci®city of these pathways arises. The concept of `master genes' provides an attractive mechanistic hypothesis as to how speci®city could be imparted by these common factors. Certainly in the case of eye morphogenesis one or a few genes such as eyeless seem to have evolved with the photoreceptor in planaria. Close eyeless homologues are capable of inducing eyes in such diverse species as clams, Drosophila and mice (Gehring, 1996). In the case of the lung, the Gli gene family and Nkx2.1 are currently the best candidates to be considered as `master genes'. However, compound null mutation of two Gli family members is required to ablate lung morphogenesis and there are effects in other organs as well. Null mutation of Nkx2.1 reveals that it does indeed seem to control dorsoventral separation of the lung anlage from the primitive esophagus, but it also induces pituitary and thyroid development as well as controlling the induction of parts of the brain (Kimura et al., 1996; Minoo et al., 1999). Thus, a truly speci®c single pulmonary master gene or genes has not yet been found. An alternative hypothesis involves complex temporospatial registration of lung morphogenesis within the rostro-caudal, dorso-ventral and left±right axes of the body plan (Fig. 1) (see Hogan, 1999 for review). Candidate genes in such a developmental scenario include Shh, FGF10, BMP4 and their respective signaling pathways as positive inductive factors, with TGF-b, Smads and sprouties as negative regulatory factors. However,, precisely how these and other factors are integrated into a developmental algorithm specifying the induction of the lung anlage, as opposed to another organ, together with a branching morphogenesis program that persists in culture, remains to be established. Time lapse photography of embryonic kidney morphogenesis, where the epithelium has been labeled with green ¯uorescent protein, suggests that morphogenesis occurs dipodially at the tips of the growing buds (http:cpmcnet.columbia.edu/dept/genetics/kidney). Observation suggests that, subsequent to the initial asymmetrical, stereotypic branching that lays out the lobes, this is the case in the embryonic lung as well. One conceptual hypothesis that ®ts the known facts is shown in Fig. 7. An `organizer zone' is postulated to exist in distal lung mesenchyme that is chemotactic for adjacent lung epithelium and can induce branch point organizer

74

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

activity in the adjacent epithelium. Candidate positive and negative regulators of this organizing activity are listed in the caption. Both positive and negative regulators of induction are co-expressed in the `organizer zone' and their activities therein presumably balance each other in favor of induction, chemotaxis and progression. Outside the `organizer zone', a `suppression zone' is postulated to exist, where an inhibitory matrix and surrounding myo®broblasts are postulated to suppress branch node formation. It has

been well known for a long time that soluble factors secreted by peripheral lung mesenchyme can induce ectopic branching activity in proximal epithelium, whereas proximal mesenchyme can suppress branching by peripheral epithelium. Epithelial proliferation per se may actually not be necessary to initiate a branching event, as is the case in the salivary gland, although it is almost certainly necessary to provide the additional epithelial surface area needed to complete any signi®cant branching activity and to elongate the airways between branch nodes. It also seems likely that branch nodes may be anchored by speci®c elements in the extracellular matrix and that the epithelium folds around these points of anchorage. During the postnatal alveolization stage the key extracellular matrix component in alveolar nodes is elastin and it is secreted by alveolar myo®broblasts. Which matrix component is key in embryonic epithelial branching is not known for certain, but ®bronectin could be important. It is also likely that matrix proteases play a key role in facilitating internodal extension of the airways, as is the case in breast morphogenesis. Proteases probably also play a key role in activating latent

Fig. 7. The molecular basis of lung morphogenesis: a conceptual hypothesis. (A) Inductive and suppressive mechanisms are balanced so that orderly branching morphogenesis occurs in time and space at the periphery of the branching epithelium. An `organizer zone' is postulated to exist in and immediately around the peripheral lung buds, which involves balanced interactions between FGF-10 and BMP4 expression arising in the mesenchyme, which stimulates the epithelium and Shh expression in the epithelium, which in turn stimulates the mesenchyme. Protease expression also plays a key role as a positive factor, while TGF-b signaling exerts a negative modulatory in¯uence, except that Smads 6 and 7, which are themselves induced by TGF-b signaling, inhibit the latter inhibition. Branch nodes form within the organizer zone, in the formation of which, matrix deposition and myo®broblast contraction probably play key roles. A `suppression zone' is postulated to exist outside the `organizer zone' in which negative in¯uences on branch formation predominate. Matrix elements and myo®broblasts are postulated to play key roles in suppressing new branch formation. (B) A conceptual rendering of how key inductive and suppressive signaling mechanisms may interact. Sonic hedgehog (shh) is secreted from a relatively wide domain in the tip of the growing epithelium. Shh activates the patched (ptc) receptor in the mesenchyme. Ptc activation releases smoothened (smo), which enters the nucleus and activates transcription of Gli2 and 3. The Gli genes negatively regulate the expression of FGF-10 in the mesenchyme. FGF-10 is expressed in the mesenchyme in spatially restricted domains, overlaying the tips of the growing epithelium. FGF-10 secreted from the mesenchyme exerts strong chemotactic and inductive effects on the epithelium by activation of FGFR. FGF ligand binding to the FGFR is facilitated by cell surface heparin sulfate proteoglycans. Activation of FGFR leads to activation of ras, which in turn leads to activation of the cell cycle through the MAP kinase cascade and cyclins. Sprouty family genes mspry2 and mspry4 are respectively expressed in the epithelium and mesenchyme and negatively regulate FGF signaling, probably at a target between FGFR activation and ras. BMP4 is expressed in the epithelium in a narrower domain than shh. BMP4 is secreted from the epithelium, activating BMP receptors on both the epithelium and the mesenchyme. BMPR signaling in the epithelium activates a speci®c set of Smads, which in turn positively regulate genes that negatively regulate the cell cycle. BMP4 signaling also negatively regulates FGF-10 expression in the mesenchyme. Please see the relevant sections of the text for further details and references.

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

growth factors in the matrix and this may in fact direct branching events as may be the case in the breast. What determines the reproducible individual internodal distances between each of the ®rst 16 branch nodes in human embryonic lung and how their speci®c orientation is genetically `hard wired' to the overall body plan is far less certain. FGF-10, Shh and BMP4 are candidate soluble ligands, which may set up distal-proximal morphogenetic concentration gradients and may regulate each other as well (Fig. 7). Hoxa5 null mutation results in abnormalities in alveolization, so it is reasonable to postulate that it as well as other Hox genes may be involved in registration of the lung within the overall body plan. A further key issue that has recently emerged is that correct co-registration of the developing vasculature with the airways appears to be essential for normal lung morphogenesis, at least in the fetal stage later stages. Analysis of the transcriptional factors binding the promoters of lung-speci®c genes, such as SP-A, -B and -C and CC10, is providing some important clues as to how expression of these lung-speci®c genes is regulated in time and space: Nkx2.1, HNF, GATA and AP-1 transcriptional factors play key roles at this level (see Whitsett, 1998). Analysis of the mechanisms specifying temporo-spatial regulation of the candidate morphogenetic gene promoters right before, around and in the lung anlage and subsequent branch points will provide some more of the answers, while null mutation, misexpression and abrogation studies in vivo and in organ explant culture will provide others. Completion of the murine and human genomes will at least delimit the ®nal scope of the problem, in terms of the number of possible morphogenetic genes. However, we predict that multifactorial analysis of the mechanistic interactions between the candidates will take a wee bit longer! Acknowledgements We gratefully acknowledge the support of the work in our laboratories by the National Heart, Lung and Blood Institute, National Institutes of Health, the American Lung Association and the American Heart Association: R01HL 44060, R01HL 44977 and P01HL60231 (D.W.); R29 HL60061, K02HL03981, AHA1131-GI1, ALA RG084-N (M.S.); P01HL60231 MIRS (G.F.D.); P01HL47409 (W.V.C.). The authors would like to thank Cathy Blagg for her attention to the numerous references cited in this work. The authors wish to apologize to those of our colleagues whose work we have failed to cite. A more comprehensive website on lung branching morphogenesis comprising many more citations is available at: http://www.ana.ed.ac.uk/anatomy/database/lungbase/lunghome.html. References Aguayo, S.M., Schuyler, W.E., Murtagh Jr., J.J., Roman, J., 1994. Regula-

75

tion of lung branching morphogenesis by bombesin-like peptides and neutral endopeptidase. Am. J. Resp. Cell Mol. Biol. 10, 635±642. Alescio, T., Cassini, A., 1962. Induction in vitro of tracheal buds by pulmonary mesenchyme grafted onto tracheal epithelium. J. Exp. Zool. 150, 83±94. Andrew, D.J., Horner, M.A., Pettit, M.G., Smolik, S.M., Scott, M.P., 1994. Setting limits on homeotic gene function: restraint of sex combs reduced activity by teashirt and other homeotic genes. EMBO J. 13, 1132±1144. Ang, S.L., Rossant, J., 1994. HNF-3 beta is essential for node and notochord formation in mouse development. Cell 78, 561±574. Aubin, J., Lemieux, M., Tremblay, M., Berard, J., Jeanotte, L., 1997. Early postnatal lethality in Hoxa-5 mutant mice is attributable to respiratory tract defects. Dev. Biol. 192, 432±445. Baldwin, H.S., 1996. Early embryonic vascular development. Cardiovasc. Res., E34±E45. Basler, K., Struhl, G., 1994. Compartment boundaries and the control of Drosophila limb pattern by hedgehog protein. Nature 368, 208±214. Batchelor, D.C., Hutchins, A.M., Klempt, M., et al., 1995. Developmental changes in the expression patterns of IGFs, type 1 IGF receptor and IGF-binding proteins-2 and -4 in perinatal rat lung. J. Mol. Endocrinol. 15, 105±115. Becker, K.L., 1984. Historical perspective on the pulmonary endocrine cell. In: Becker, K.L., Gazdar, A.F. (Eds.). The Endocrine Lung in Health and Disease, W.B. Saunders, Philadelphia, PA, pp. 156±161. Belknap, J.K., Weiser-Evans, M.C., Grieshaber, S.S., Majack, R.A., Stenmark, K.R., 1999. Relationship between perlecan and tropoelastin gene expression and cell replication in the developing rat pulmonary vasculature. Am. J. Resp. Cell Mol. Biol. 20, 24±34. Bellusci, S., Henderson, R., Winnier, G., Oikawa, T., Hogan, B., 1996. Evidence from normal expression and targeted misexpression that bone morphogenetic protein-4 (Bmp-4) plays a role in mouse embryonic lung morphogenesis. Development 122, 1693±1702. Bellusci, S., Furuta, Y., Rush, M.G., Henderson, R., Winnier, G., Hogan, B., 1997a. Involvement of Sonic Hedgehog (Shh) in mouse embryonic lung growth and morphogenesis. Development 124, 53±63. Bellusci, S., Grindley, J., Emoto, H., Itoh, N., Hogan, B.L., 1997b. Fibroblast growth factor 10 (FGF10) and branching morphogenesis in the embryonic mouse lung. Development 124, 4867±4878. Bernier, S.M., Utani, A., Sugiyama, S., Doi, T., Polistinas, C., Yamada, Y., 1994. Cloning and expression of laminin a2 chain (M-chain) in the mouse. Matrix Biol. 14, 447±455. Bingle, C.D., Gitlin, J.D., 1993. Identi®cation of hepatocyte nuclear factor3 binding sited in the Clara cell secretory protein gene. Biochem. J. 295, 227±232. Bingle, C.D., Gowan, S., 1996. Oct-1 interacts with conserved motifs in the human thyroid transcription factor 1 gene minimal promoter. Biochem. J. 319, 669±674. Bingle, C.D., Hackett, B.P., Moxley, M., Longmore, W., Gitlin, J.D., et al., 1995. Role of hepatic nuclear factor-3a and hepatocyte nuclear factor3b in Clara cell secretory protein gene expression in the bronchiolar epithelium. Biochem. J. 308, 197±202. Bitgood, M.J., McMahon, A.P., 1995. Hedgehog and BMP genes are coexpressed at many diverse sites of cell±cell interaction in the mouse embryo. Dev. Biol. 172, 126±138. Bogue, C.W., Groww, I., Vasavada, H., Dynia, D.W., Wilson, C.M., Jacobs, H.C., 1994. Identi®cation of Hox genes in newborn lung and effects of gestational age and retinoic acid on their expression. Am. J. Physiol. 266, L448±L454. Bohinski, R.J., Huffman, J.A., Whitsett, J.A., Lattier, D.L., 1993. Cis-active elements controlling lung cell-speci®c expression of human pulmonary surfactant protein B gene. J. Biol. Chem. 268, 11160±11166. Bohinski, R.J., DiLauro, R., Whitsett, J.A., 1994. The lung surfactant protein B gene promoter is a target for the thyroid transcription factor 1 and hepatocyte nuclear factor 3, indicating common factors for organspeci®c gene expression along the foregut axis. Mol. Cell. Biol. 14, 671±681.

76

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

Borges, M., Linnoila, R.L., van de Velde, H.J.K., Chen, H., Nelkin, B.D., Mabry, M., Baylin, S.B., Ball, D.W., 1997. An achaete-scute homologue essential for neuroendocrine differentiation in the lung. Nature 386, 852±855. Bostrom, H., Willetts, K., Pekny, M., Leveen, P., Lindahl, P., Hedstrand, H., Pekna, M., Hellstrom, M., Gebre-Medhin, S., Schalling, M., Nilsson, M., Kurland, S., Tornell, J., Heath, J.K., Betsholtz, C., 1996. PDGF-A signaling is a critical event in lung alveolar myo®broblast development and alveogenesis. Cell 85, 863±873. Braun, H., Suske, G., 1998. Combinatorial action of HNF3 and Sp family transcription factors in the activation of the rabbit uterglobin/CC10 promoter. J. Biol. Chem. 273, 9821±9828. Brinkmann, V., Fouroutan, H., Sachs, M., Weidner, K.M., Birchmeier, W., 1995. Hepatocyte growth factor/scatter factor induces a variety of tissue-speci®c morphogenetic programs in epithelial cells. J. Cell. Biol. 13, 1573±1586. Bruno, M.D., Bohinski, R.J., Huelsman, K.M., Whitsett, J.A., Korfhagen, T.R., 1995. Lung cell-speci®c expression of the murine surfactant protein A (SP-A) gene is mediated by interactions between the SP-A promoter and thyroid transcription factor-1. J. Biol. Chem. 270, 6531± 6536. Burgerson, R.E., Chiquet, M., Deutzmann, R., Ekblom, P., Engel, J., Kleinman, H., Martin, G.R., Meneguzzi, G., Paulsson, M., Sanes, J., 1994. A new nomenclature for the laminins. Matrix Biol. 14, 209±211. Burri, P.H., 1974. The postnatal growth of the rat lung. 3. Morphology. Anat. Rec. 180, 77±98. Burri, P., 1997. Development and growth of the lung. In: McDonald, J. (Ed.). Lung Growth and Development, 100. Marcel Dekker, New York, pp. 1±35. Capdevilla, J., Pariente, F., Sampedro, J., Alonso, J.L., Guerro, I., 1994. Subcellular localization of the segment polarity gene patched suggests an interaction with the wingless reception complex in Drosophila embryos. Development 120, 987±998. Carabba, V.H., Sorokin, S.P., Hoyt Jr., R.F., 1985. Development of neuroepithelial bodies in intact and cultured lungs of fetal rats. Am. J. Anat. 173, 1±27. Cardoso, W.V., 1995. Transcription factors and pattern formation in the developing lung. Am. J. Physiol. 269, L429±L442. Cardoso, W.V., Ito, A., Nogawa, H., Mason, I., Brody, J.S., 1997. FGF-1 and FGF-7 induce distinct patterns of growth and differentiation in embryonic lung epithelium. Dev. Dyn. 208, 398±405. Carriaga, D., 1999. German baritone Quasthoff resonates in L.A. debut. Music Review. Los Angeles Times, April 13, 1999. Celli, G., LaRocehelle, W.J., Mackem, S., Sharp, R., Merlino, G., 1998. Soluble dominant-negative receptor uncovers essential roles for ®broblast growth factors in multi-organ induction and patterning. EMBO J. 17, 1642±1655. Chen, J., Little, C.D., 1987. Cellular events associated with lung branching morphogenesis including the deposition of collagen type IV. Dev. Biol. 120, 311±321. Chen, J., Knowles, H.J., Hebert, J.L., Hackett, B.P., 1998. Mutation of the mouse hepatocyte nuclear factor/forkhead homologue 4 gene results in an absence of cilia and random left-right asymmetry. J. Clin. Invest. 102, 1077±1082. Chen, R.H., Miettinen, P.J., Maruoka, E.M., Choy, I., Derynck, R., 1995. A WD-domain protein that is associated with and phosphorylated by the type II TGF-b receptor. Nature 377, 548±552. Chinoy, M.R., Volpe, M.V., Cilley, R.E., Zgleszewski, S.E., Vosatka, R.J., Martin, A., Nielsen, H.C., Krummel, T.M., 1998. Growth factors and dexamethasone regulate Hoxb5 protein in cultured murine fetal lungs. Am. J. Physiol. 274, L610±L620. Clevidence, D.E., Overdier, D.G., Peterson, R.S., et al., 1994. Members of the HNF-3/forkhead family of transcription factors exhibit distinct cellular expression patterns in lung and regulate surfactant protein B promoter. Dev. Biol. 166, 195±209. Coalson, J.J., Winter, V., Yoder, B., 1997. Dysmorphic vascular develop-

ment in the premature baboons with bronchopulmonary dysplasia. Am. J. Resp. Crit. Care Med. 155, 262A. Cof®n, J.D., Harrison, J., Schwartz, S., Heimark, R., 1991. Angioblast differentiation and morphogenesis of the vascular endothelium in the mouse embryo. Dev. Biol. 148, 51±62. Comroe, J.H., 1965. Physiology of Respiration, Year Book, Chicago, IL, pp. 11±16. Coppola, D., Ferber, A., Miura, M., Sell, C., D'Ambrosio, C., Rubin, R., Baserga, R., 1994. A functional insulin-like growth factor I receptor is required for the mitogenic and transforming activities of the epidermal growth factor receptor. Mol. Cell. Biol. 14, 4588±4595. Cutz, E., 1982. Neuroendocrine cells of the lung: an overview of morphologic characteristics and development. Exp. Lung Res. 3, 185±208. Cutz, E., Gillan, J.E., Track, N.S., 1984. Pulmonary endocrine cells in the developing human lung and during neonatal adaptation. In: Becker, K.L., Gazdar, A.F. (Eds.). The Endocrine Lung in Health and Disease, W.B. Saunders, Philadelphia, PA, pp. 210±231. D'Amato, R.J., Loughnan, M.S., Flynn, E., Folkman, J., 1994. Thalidomide is an inhibitor of angiogenesis. Proc. Natl. Acad. Sci. USA 91, 4082± 4085. Davies, D.V., 1967. Gray's Anatomy, 34th Edition. Longman's, London, pp. 236±237. deMaximy, A.A., Nakatake, Y., Moncada, S., Itoh, N., Thiery, J.P., Bellusci, S., 1999. Cloning and expressing pattern of a mouse homologue of Drosophila sprouty in the mouse embryo. Mech. Dev. 81, 213± 216. deMello, D.E., Sawyer, D., Galvin, N., Reid, L.M., 1997. Early fetal development of lung vasculature. Am. J. Resp. Cell Mol. Biol. 16, 568±581. Deterding, R.R., Shannon, R.M., 1995. Proliferation and differentiation of fetal rat pulmonary epithelium in the absence of mesenchyme. J. Clin. Invest. 95, 2963±2972. Devriendt, K., Vanhole, C., Matthijs, G., de Zegher, F., 1998. Deletion of thyroid transcription factor-1 gene in an infant with neonatal thyroid dysfunction and respiratory failure. N. Engl. J. Med. 338, 1317±1318. Diaz-Benjumea, F.J., Cohen, B., Cohen, S.M., 1994. Cell interaction between compartments establishes the proximal-distal axis of Drosophila legs. Nature 372, 175±179. Drake, C.J., Little, C.D., 1995. Exogenous vascular endothelial growth factor induces malformed and hyperfused vessels during embryonic neovascularization. Proc. Natl. Acad. Sci. USA 92, 7657±7661. Drake, C.J., Hungerford, J.E., Little, C.D., 1998. Morphogenesis of the ®rst blood vessels. Ann. N. Y. Acad. Sci. 857, 155±179. Dunsmore, S.E., Rannels, D.E., 1996. Extracellular matrix biology in the lung. Am. J. Physiol. Lung Cell Mol. Physiol. 270, L3±L27. Durham, P.L., Snyder, J.M., 1995. Characterization of alpha1, beta1, and gamma1 laminin subunits during rabbit fetal lung development. Dev. Dyn. 203, 408±421. Durham, P.L., Snyder, J.M., 1996. Regulation of the beta 2 subunit chain of laminin in developing rabbit fetal lung tissue. Differentiation 60, 229± 243. Dziadek, M., 1995. Role of laminin-nidogen complexes in basement membrane formation during embryonic development. Experientia 51, 901±913. Dziadek, M., Paulsson, M., Timpl, R., 1985. Identi®cation and interaction repertoire of large forms of the basement membrane protein nidogen. EMBO J. 4, 2513±2518. Ehrig, K., Leivo, I., Argraves, W.S., Ruoslahti, E., Engvall, E., 1990. Merosin, a tissue-speci®c basement membrane protein, is a lamininlike protein. Proc. Natl. Acad. Sci. USA 87, 3264±3268. Ekblom, P., Ekblom, M., Fecker, L., Klein, G., Zhang, H.Y., Kadoya, Y., Chu, M.L., Mayer, U., Timpl, R., 1994. Role of mesenchymal nidogen for epithelial morphogenesis in vitro. Development 120, 2003±2014. Flamme, I., Breier, G., Risau, W., 1995a. Vascular endothelial growth factor (VEGF) and VEGF receptor 2 (¯k-1) are expressed during vasculogenesis and vascular differentiation in the quail embryo. Dev. Biol. 169, 699±712.

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81 Flamme, I., von Reutern, M., Drexler, H.C., Syed-Ali, S., Risau, W., 1995b. Overexpression of vascular endothelial growth factor in the avian embryo induces hypervascularization and increased vascular permeability without alterations of embryonic pattern formation. Dev. Biol. 171, 399±414. Flores-Delgado, G., Bringas, P., Warburton, D., 1998. Laminin 2 attachment selects myo®broblasts from fetal mouse lung. Am. J. Physiol. 275, L622±L630. Frieser, M., Nockel, H., Pausch, F., Roder, C., Hahn, A., Deutznmann, R., Sorokin, L.M., 1997. Cloning of the mouse laminin a4 chain expression in a subset of endothelium. Eur. J. Biochem. 246, 727±735. Galliano, M., Aberdam, D., Aguzzi, A., et al., 1995. Cloning and complete primary structure of the mouse a3 chain. J. Biol. Chem. 270, 21821± 21826. Gebb, S., Shannon, J., 1998. Tissue interactions mediate early events in pulmonary vasculogenesis. Am. J. Resp. Crit. Care Med. 157, A208. Gehring, W.J., 1996. The master control gene for morphogenesis and evolution of the eye. Genes Cells 1, 11±15. Gerber, H.P., Hillan, K.J., Ryan, A.M., Kowalski, J., Keller, G.A., Rangell, L., Wright, B.D., Radtke, F., Aguet, M., Ferrara, N., 1999. VEGF is required for growth and survival in neonatal mice. Development 126, 1149±1159. Glazer, L., Shilo, B.A., 1991. The Drosophila FGF-R homolog is expressed in the embryonic tracheal system and appears to be required for directed tracheal cell extension. Genes Dev. 5, 697±705. Grindley, J.C., Bellusci, S., Perkins, D., Hogan, B.L., 1997. Evidence for the involvement of the Gli gene family in embryonic mouse lung development. Dev. Biol. 188, 337±348. Guillemin, K., Groppe, J., Ducker, K., Treisman, R., Hafen, E., Affolter, M., Krasnow, M.A., 1996. The pruned gene encodes the Drosophila serum response factor and regulates cytoplasmic outgrowth during terminal branching of the tracheal system. Development 122, 1353± 1362. Guillemot, F., Lo, L.-C., Johnson, J.E., Auerbach, A., Anderson, D.J., Joyner, A.L., 1993. Mammalian achaete-scute Homolog 1 is required for the early development of olfactory and autonomic neurons. Cell 75, 463±476. Guo, L., Degestein, L., Fuchs, E., 1996. Keratinocyte growth factor is required for hair development but not wound healing. Genes Dev. 10, 165±175. Hacohen, N., Kramer, S., Sutherland, D., Hiromi, Y., Krasnow, M., 1998. sprouty encodes a novel antagonist of FGF signaling that patterns apical branching of the Drosophila airways. Cell 92, 253±263. Heberlien, U., Wolff, T., Rubin, G.M., 1993. The TGFb homolog Dpp and the segment polarity gene hedgehog are required for propagation of a morphogenetic wave in the Drosophila retina. Cell 75, 913±926. Heine, U.I., Munoz, E.F., Flanders, K.C., Roberts, A.B., Sporn, M.B., 1990. Colocalization of TGF-b1 and collagen I and II, ®bronectin and glycosaminoglycans during lung branching morphogenesis. Development 109, 29±36. Heine, U.I., Burmester, J.K., Flanders, K.C., Danielpour, D., Munoz, E.F., Roberts, A.B., Sporn, M.B., 1991. Localization of transforming growth factor-beta 1 in mitochondria of murine heart and liver. Cell Regul. 2, 467±477. Hildebrand, A., Romaris, M., Rasmussen, L.M., Heinegard, D., Twardzik, D.R., Border, W.A., Ruoslahti, E., 1994. Interaction of the small interstitial proteoglycans biglycan, decorin and ®bromodulin with transforming growth factor-b. Biochem. J. 302, 527±534. Hilfer, S.R., 1996. Morphogenesis of the lung: control of embryonic and fetal branching. Annu. Rev. Physiol. 58, 93±113. Hohenadel, K., 1999. Thomas Quasthoff sings the last note of Brahm's ``German'' Requiem, and the audience stumbles to its feet. Los Angeles Times Calendar Section, April 4, 1999. Hogan, B.L.M., 1999. Morphogenesis. Cell 96, 225±233. Hoyt Jr., R.F., McNelly, N.A., Sorokin, S.P., 1990. Dynamics of neuroepithelial body (NEB) formation in developing hamster lung: light

77

microscopic autoradiography after 3H-thymidine labeling in vivo. Anat. Rec. 227, 340±350. Hoyt Jr., R.F., McNelly, N.A., McDowell, E.M., Sorokin, S.P., 1991. Neuroepithelial bodies stimulate proliferation of airway epithelium in fetal hamster lung. Am. J. Physiol. 260, L234±L240. Hudson, B.G., Reeders, S.T., Tryggvason, K., 1993. Type IV collagen: structure gene organization, and role in human diseases. J. Biol. Chem. 268, 26033±26036. Iivanainen, A., Vuolteenaho, R., Sainio, K., Eddy, R., Shows, T.B., Sariola, H., Tryggvason, K., 1995a. The human laminin b2 chain (S-laminin): structure, expression in fetal lung tissues and chromosomal assignment of the LAMB2 gene. Matrix Biol. 14, 489±497. Iivanainen, A., Sainio, K., Sariola, H., Tryggvason, K., 1995b. Primary structure and expression of a novel human laminin a4 chain. FEBS Lett. 365, 183±188. Iivanainen, A., Kortesmaa, J., Sahlberg, C., Morita, T., Bergmann, U., Thesleff, I., Tryggvason, K., 1997. Primary structure, developmental expression, and immunolocalization of the murine laminin a4 chain. J. Biol. Chem. 272, 27862±27868. Iivanainen, A., Morita, T., Tryggvason, K., 1999. Molecular cloning and tissue-speci®c expression of a novel murine laminin gamma 3 chain. J. Biol. Chem. 274, 14107±14111. Ikeda, K., Clark, J.C., Shaw-White, J.R., Stahlman, M.T., Boutell, C.J., Whitsett, J.A., 1995. Gene structure and expression of human thyroid transcription factor-1 in respiratory epithelial cells. J. Biol. Chem. 270, 8108±8114. Ikeda, K., Shaw-White, J.R., Wert, S.E., Whitsett, J.A., 1996. Hepatocyte nuclear factor 3 activates transcription of thyroid transcription factor 1 in respiratory epithelial cells. Mol. Cell. Biol. 16, 3626±3636. Isaac, D.D., Andrew, D.J., 1996. Tubulogenesis in Drosophila: a requirement for the trachealess gene product. Genes Dev. 10, 103±117. Jakkula, M., Cras, T., Abman, S., 1999. Effects of fumagillin and talidomide treatment on alveolarization in the developing rat lung. Pediatr. Res. 45, 67A. Jaskoll, T.F., Don-Wheeler, G., Johnson, R., Slavkin, H.C., 1988. Embryonic mouse lung morphogenesis and type II cytodifferentiation in serumless, chemically de®ned medium using prolonged in vitro cultures. Cell Diff. 24, 105±117. Jaye, M., Schlessinger, J., Dionne, C.A., 1992. Fibroblast growth factor receptor tyrosine kinases: molecular analysis and signal transduction. Biochim. Biophys. Acta 1135, 185±199. Johnson, R.L., Rothman, A.L., Xie, J., Goodrich, L.V., Bare, J.W., Bonifas, J.M., Quinn, A.G., Myers, R.M., Cox, D.R., Epstein Jr., E.H., Scott, M.P., 1996. Human homolog of patched, a candidate gene for the basal cell nevus syndrome. Science 272, 1668±1671. Kaartinen, V., Voncken, J.W., Shuler, C.A., Warburton, D., Bu, D., Heisterkamp, N., Groffen, J., 1995. Abnormal lung development and cleft palate: defects of mesenchymal±epithelial interaction in mice lacking TGF-b3. Nat. Genet. 11, 415±421. Kao, J., Ryan, J., Brett, G., Chen, J., Shen, H., Fan, Y.G., Godman, G., Familletti, P.C., Wang, F., Pan, Y.C., et al., 1992. Endothelial monocyte-activating polypeptide II. A novel tumor-derived polypeptide that activates host-response mechanisms. J. Biol. Chem. 267, 20239±20247. Kao, J., Fan, Y.G., Haehnel, I., Clauss, M., Stern, D., 1993. Endothelialmonocyte activating polypeptides (EMAPs): tumor derived mediators which activate the host in¯ammatory response. Behring Inst. Mitt. 92, 92±106. Kao, J., Houck, K., Fan, Y., Haehnel, I., Libutti, S.K., Kayton, M.L., Grikscheit, T., Chabot, J., Nowygrod, R., Greenberg, S., et al., 1994. Characterization of a novel tumor-derived cytokine. Endothelial-monocyte activating polypeptide II. J. Biol. Chem. 269, 25106±25119. Kappen, C., 1996. Hox genes in the lung. Am. J. Resp. Cell Mol. Biol. 15, 156±162. Kataoka, R., Sherlock, J., Lanier, S.M., 1993. Signaling events initiated by transforming growth factor-beta 1 that require Gi alpha 1. J. Biol. Chem. 268, 1±19857.

78

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

Kauffman, S.L., 1981. Histogenesis of the papillary Clara cell adenoma. Am. J. Pathol. 103, 174±180. Kauffman, M.H., 1992. The Atlas of Mouse Development, Academic Press, New York, pp. 445±448. Kelly, S.E., Bachurski, C.J., Burhans, M.S., Glasser, S.W., 1996. Transcription of the lung-speci®c surfactant protein C gene is mediated by thyroid transcription factor 1. J. Biol. Chem. 271, 6881±6888. Kimura, S., Hara, Y., Pineau, T., Fernandez-Salguero, P., Fox, C.H., Ward, J.M., Gonzalez, F.J., 1996. The T/ebp null mouse: thyroid-speci®c enhancer binding protein is essential for the organogenesis of the thyroid, lung, ventral forebrain and pituitary. Genes Dev. 10, 60±69. King, K.A., Torday, J.S., Sunday, M.E., 1995. Bombesin and [Leu 8] phyllolitorin promote fetal mouse lung branching morphogenesis via a receptor-mediated mechanism. Proc. Natl. Acad. Sci. USA 92, 4357± 4361. Klambt, C., 1993. The Drosophila gene pointed encodes two ETS-like proteins which are involved in the development of the midline glial cells. Development 117, 163±176. Klein, G., Ekblom, M., Fecker, L., et al., 1990. Differential expression of laminin A and B chains during development of embryonic mouse organs. Development 110, 823±837. Koch, M., Olson, P.F., Albus, A., Jin, W., Hunter, D.D., Brunken, W.J., Burgeson, R.E., Champliaud, M.F., 1999. Characterization and expression of the laminin gamma3 chain: a novel, non-basement membraneassociated, laminin chain. J. Cell Biol. 145, 605±618. Korfhagen, T.R., Swantz, R.J., Wert, S.E., McCarty, J.M., Kerlakian, C.B., Glasser, S.W., Whitsett, J.A., 1994. Respiratory epithelial cell expression of human transforming growth factor-a induces lung ®brosis in transgenic mice. J. Clin. Invest. 93, 1691±1699. Kotch, L.E., Iyer, N.V., Laughner, E., Semenza, G.L., 1999. Defective vascularization of HIF-1alpha-null embryos is not associated with VEGF de®ciency but with mesenchymal cell death. J. Biol. Chem. 274, 17813±17819. Lallemand, A.V., Ruocco, S.M., Joly, P.M., Gaillard, D.A., 1995a. In vivo localization of the insulin-like growth factors I and II (IGF I and IGF II) gene expression during human lung development. Int. J. Dev. Biol. 39, 529±537. Lallemand, A.V., Ruocco, S.M., Gaillard, D.A., 1995b. Synthesis and expression of laminin during human foetal lung development. Anat. Rec. 242, 233±241. Lastres, P., Letamendia, A., Zhang, H., Rius, C., Almendro, N., Raab, U., Lopez, L.A., Lang, C., Fabra, A., Letarte, M., Bernabeu, C., 1996. Endoglin modulates cellular responses to TGF-b1. J. Cell Biol. 133, 1109±1121. Lauweryns, J.M., Peuskens, J.C., 1972. Neuroepithelial bodies (neuroreceptor or secretory organs?) in human infant bronchial and bronchiolar epithelium. Anat. Rec. 172, 471±482. Lebeche, D., Malpel, S., Cardoso, W.V., 1999. Fibroblast growth factor interactions in the developing lung. Mech. Dev. 86, 125±136. Lee, T., Hacohen, N., Krasnow, M., Montell, D.J., 1996. Regulated Breathless receptor tyrosine kinase activity required to pattern cell migration and branching in the Drosophila tracheal system. Genes Dev. 10, 2912± 2921. Liu, J.P., Baker, J., Perkins, S.A., Robertson, E.J., Efstratiadis, A., 1993. Mice carrying null mutations of the genes encoding insulin-like growth factor I (Igf-1) and type 1 IGF receptor (Igf1r). Cell 75, 59±72. Liu, F., Hata, A., Baker, J.C., Doody, J., Carcamo, J., Harland, R.M., Massague, J., 1996. A human Mad protein acting as a BMP-regulated transcriptional activator. Nature 381, 620±623. Lwebuga-Mukasa, J.S., 1991. Matrix-driven pneumocyte differentiation. Am. Rev. Resp. Dis. 144, 452±457. Maitre, B., Clement, A., Williams, M.C., Brody, J.S., 1995. Expression of insulin-like growth factor receptors 1 and 2 in the developing lung and their relation to epithelial cell differentiation. Am. J. Resp. Cell Mol. Biol. 13, 262±270. Martin, G.R., 1998. The roles of FGFs in the early development of vertebrate limbs. Genes Dev. 12, 1571±1586.

Mayer, U., Mann, K., Timpl, R., Murphy, G., 1993. Sites of nidogen cleavage by proteases involved in tissue homeostasis and remodeling. Eur. J. Biochem. 217, 877±884. McAllister, K.A., Grogg, K.M., Johnson, D.W., Gallione, C.J., Baldwin, M.A., Jackson, C.E., Helmbold, E.A., Markel, D.S., McKinnon, W.C., Murrell, J., et al., 1994. Endoglin, a TGF-b binding protein of endothelial cells is the gene for hereditary hemorrhagic telangiectasia type 1. Nat. Genet. 8, 345±351. Metzger, R.J., Krasnow, M.A., 1999. Genetic control of branching morphogenesis. Science 284, 1635±1639. Meyrick, B., Reid, L., 1982. Pulmonary arterial and alveolar development in normal postnatal rat lung. Am. Rev. Resp. Dis. 125, 468±473. Miettinen, P., Berger, J.E., Meneses, J., et al., 1995. Epithelial immaturity and multiorgan failure in mice lacking epidermal growth factor receptor. Nature 376, 337±341. Miettinen, P.J., Warburton, D., Bu, D., Zhao, J.S., Berger, J.E., Minoo, P., Koivisto, T., Allen, L., Dobbs, L., Werb, Z., Derynck, R., 1997. Impaired lung branching morphogenesis in the absence of functional EGF receptor. Dev. Biol. 15, 224±236. Min, H., Danilenko, D.M., Scully, S.A., Bolon, B., Ring, B.D., Tarpley, J.E., DeRose, M., Simonett, W.S., 1998. FGF-10 is required for both limb and lung development and exhibits striking functional similarities to Drosophila branchless. Genes Dev. 12, 3156±3161. Miner, J.H., Sanes, J.R., 1994. Collagen IV a3, a4, and a5 chains in rodent basal laminae: sequence, distribution, association with laminins, and developmental switches. J. Cell Biol. 127, 879±891. Miner, J.H., Lewis, R.M., Sanes, J.R., 1995. Molecular cloning of a novel laminin chain, a5, and widespread expression in adult mouse tissue. J. Biol. Chem. 270, 28523±28526. Miner, J.H., Paton, B.L., Lentz, S.I., Gilbert, D.J., Snider, W.D., Jenkin, N.A., Copeland, N.G., Sanes, J.R., 1997. The laminin a chains: expression, developmental transitions, and chromosomal locations of a1±5, identi®cation of heterotrimeric laminins 8±11, and cloning of a novel a3 isoform. J. Cell Biol. 137, 685±701. Miner, J.H., Cunningham, J., Sanes, J.R., 1998. Roles for laminin in embryogenesis: exencephaly, syndactyly, and placentopathy in mice lacking the laminin a5 chain. J. Cell Biol. 143, 1713±1723. Minoo, P., King, R.J., 1994. Epithelial±mesenchymal interactions in lung development. Annu. Rev. Physiol. 56, 13±45. Minoo, P., Hamdan, H., Bu, D., Warburton, D., Stepanik, P., deLemos, R., 1995. TTF-1 regulates lung epithelial morphogenesis. Dev. Biol. 172, 694±698. Minoo, P., Su, G., Drum, H., Bringas, P., Kimura, S., 1999. Defects in tracheoesophageal and lung morphogenesis in Nkx2.1(2/2) mouse embryos. Dev. Biol. 209, 60±71. Moens, C.B., Stanton, B.R., Parada, L.F., Rossant, J., 1993. Defects in heart and lung development in compound heterozygotes for two different targeted mutations at the N-myc locus. Development 119, 485± 499. Mollard, R., Dziadek, M., 1997. Homeobox genes from clusters A and B demonstrate characteristics of temporal colinearity and differential restrictions in spatial expression domains in the branching mouse lung. Int. J. Dev. Biol. 41, 655±666. Morgan, T.H., 1897. Regeneration in Allolobophora feotida. Roux's Arch Dev. Biol. 5, 570±586. Morrisey, E.E., Ip, H.S., Lu, M.M., Parmacek, M.S., 1996. GATA-6: a zinc ®nger transcription factor that is expressed in multiple cell lineages derived from lateral mesoderm. Dev. Biol. 177, 309±322. Morrisey, E.E., Ip, H.S., Tang, Z., Lu, M.M., Parmacek, M.S., 1997. GATA-5: a transcriptional activator expressed in a novel temporally and spatially-restricted pattern during embryonic development. Dev. Biol. 183, 21±36. Morrisey, E.E., Tang, Z., Sigrist, K., Lu, M.M., Jiang, F., Ip, H.S., Parmacek, M.S., 1998. GATA6 regulates HNF4 and is required for differentiation of visceral endoderm in the mouse embryo. Genes Dev. 12, 3579±3590. Motoyama, J., Liu, J., Mo, R., Ding, Q., Post, M., Hui, C.C., 1998. Essential

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81 function of Gli2 and Gli3 in the formation of lung, trachea and esophagus. Nat. Genet. 20, 54±57. Muglia, L.J., Bae, D.S., Brown, T.T., Vogt, S.K., Alvarez, J.G., Sunday, M.E., Majzoub, J.A., 1999. Proliferation and differentiation defects during lung development in corticotropin-releasing hormone-de®cient mice. Am. J. Resp. Cell Mol. Biol. 20, 181±188. Murthy, U.S., Anzano, M.A., Stadel, J.M., Greig, R., 1988. Coupling of TGF-beta-induced mitogenesis to G-protein activation in AKR-2B cells. Biochem. Biophys. Res. Commun. 152, 1228±1235. Nakazato, M., Endo, T., Saito, T., Harii, N., Onaya, T., 1997. Transcription of the thyroid transcription factor-1 (TTF-1) gene from a newly de®ned start site: positive regulation by TTF-1 in the thyroid. Biochem. Biophys. Res. Commun. 238, 748±752. Newfeld, S.J., Chartoff, E.H., Graff, J.M., Melton, D.A., Gelbart, W.M., 1996. Mothers against dpp encodes a conserved cytoplasmic protein required in DPP/TGR-beta responsive cells. Development 122, 2099± 2108. Nogawa, H., Ito, T., 1995. Branching morphogenesis of embryonic mouse lung epithelium in mesenchyme-free culture. Development 121, 1015± 1022. Nusslein-Volhard, C., Wieschaus, E., 1980. Mutations affecting segment number and polarity in Drosophila. Nature 287, 795±801. Ohmichi, H., Koshimizu, U., Matsumoto, K., Nakamura, T., 1998. Hepatocyte growth factor (HGF) acts as a mesenchyme-derived morphogenic factor during fetal lung development. Development 125, 1315± 1324. Park, W.Y., Miranda, B., Lebeche, D., Hashimoto, G., Cardoso, W.V., 1998. Fibroblast growth factor 10 (FGF-10) is a chemotactic factor for distal epithelial buds during lung development. Dev. Biol. 201, 125±134. Paulsson, M., 1992. Basement membrane proteins: structure, assembly and cellular interactions. Crit. Rev. Biochem. Mol. Biol. 27, 93±127. Pelton, R.W., Dickinson, M.E., Moses, H.L., Hogan, B.L., 1990. In situ hybridization analysis of TGF-b 3 RNA expression during mouse development: comparative studies with TGF-b 1 and 2. Development 110, 609±620. Pepicelli, C.V., Lewis, P., McMahon, A., 1998. Sonic hedgehog regulates branching morphogenesis in the mammalian lung. Curr. Biol. 8, 1083± 1086. Peters, K.G., Chen, W.G., Williams, L.T., 1992. Two FGF receptors are differentially expressed in epithelial and mesenchymal tissues during limb formation and organogenesis. Development 114, 233±243. Peters, K., Werner, S., Liao, X., Wert, S., Whitsett, J., Williams, L., 1994. Targeted expression of a dominant negative FGF receptor blocks branching morphogenesis and epithelial differentiation of mouse lung. EMBO J. 13, 3296±3301. Pierce, R.A., Grif®n, G.L., Mudd, M.S., Moxley, M.A., Longmore, W.J., Sanes, J.R., Miner, J.H., Senior, R.M., 1998. Expression of laminin a3, a4, and a5 chains by alveolar epithelial cells and ®broblasts. Am. J. Resp. Cell Mol. Biol. 19, 237±244. Placzek, M., Skaer, H., 1999. Airway patterning: a paradigm for restricted signaling. Curr. Biol. 9, R506±R510. Poole, T.J., Cof®n, J.D., 1989. Vasculogenesis and angiogenesis: two distinct morphogenetic mechanisms establish embryonic vascular pattern. J. Exp. Zool. 251, 224±231. Post, M., Souza, P., Liu, J., Tseu, I., Wang, J., Kuliszewski, M., Tanswell, K., 1996. Keratinocyte growth factor and its receptor are involved in regulating early lung branching. Development 122, 3107±3115. Quaggin, S.E., Vanden Heuvel, G.B., Irash, P., 1998. Pod-1, a mesodermspeci®c basic helix-loop-helix protein expressed in mesenchymal and glomerular epithelial cells in the developing kidney. Mech. Dev. 71, 37±48. Raaberg, L., Nexo, E., Buckley, S., Luo, W., Snead, M.L., Warburton, D., 1992. Epidermal growth factor transcription, translation and signal transduction in rat type II pneumocytes in culture. Am. J. Resp. Cell Mol. Biol. 6, 44±49. Raaberg, L., Nexo, E., Jorgensen, P.E., Jakab, M., 1995a. Fetal effects of

79

epidermal growth factor de®ciency induced in rats by autoantibodies against epidermal growth factor. Pediatr. Res. 37, 175±181. Raaberg, L., Nexo, E., Poulsen, S.S., Jorgensen, P.E., 1995b. A immunologic approach to induction of epidermal growth factor de®ciency: induction and characterization of autoantibodies to epidermal growth factor in rats. Pediatr. Res. 37, 169±174. Rechts-Bogart, G.Z., Moats-Stats, B.M., Howard, K., D'Ercole, A.J., Stiles, A.D., 1996. Cellular localization of messenger RNAs for insulin-like growth factors (IGFs), their receptors and binding proteins during fetal rat lung development. Am. J. Resp. Cell Mol. Biol. 14, 61±69. Reich, A., Sapir, A., Shilo, B.-Z., 1999. Sprouty is a general inhibitor of tyrosine kinase activity. Development 126, 4139±4147. Reich-Slotsky, R., Bonneh-Barkay, D., Shaoul, E., Bluma, B., Svahn, C.M., Ron, D., 1994. Differential effect of cell-associated heparan sulfates on the binding of keratinocyte growth factor (KGF) and acidic ®broblast growth factor to the KGF receptor. J. Biol. Chem. 269, 32279±32285. Reid, L., 1979. Bronchopulmonary dysplasia ± pathology. J. Pediatr. 95, 836±841. Reinhardt, D., Mann, K., Nischt, R., Fox, J.W., Chu, M.L., Krieg, T., Timpl, R., 1993. Mapping of nidogen binding sites for collagen type IV, heparan sulfate proteoglycan, and zinc. J. Biol. Chem. 268, 10881± 10887. Richard, A.P., Grif®n, G.L., Mudd, M.S., Moxley, M.A., Longmore, W.J., Sanes, J.R., Miner, J.H., Senior, R.M., 1998. Expression of laminin a3, a4, and a5 chains by alveolar epithelial cells and ®broblasts. Am. J. Cell Mol. Biol. 19, 237±244. Richards, A., Al-Imara, L., Pope, F.M., 1996. The complete cDNA sequence of laminin a4 and its relationship to the other human laminin a chains. Eur. J. Biochem. 238, 813±821. Risau, W., Lemmon, V., 1988. Changes in the vascular extracellular matrix during embryonic vasculogenesis and angiogenesis. Dev. Biol. 125, 441±450. Risau, W., Sariola, H., Zerwes, H.G., Sasse, J., Ekblom, P., Kemler, R., Doetschman, T., 1988. Vasculogenesis and angiogenesis in embryonicstem-cell-derived embryoid bodies. Development 102, 471±478. Rozengurt, E., Sinnett-Smith, J., 1983. Bombesin stimulation of DNA synthesis and cell division in cultures of Swiss 3T3 cells. Proc. Natl. Acad. Sci. USA 80, 2936±2940. Ryan, H.E., Lo, J., Johnson, R.S., 1998. HIF-1 alpha is required for solid tumor formation and embryonic vascularization. EMBO J. 17, 3005± 3015. Samakovlis, C., Hacohen, N., Manning, G., Sutherland, D.C., Guillemin, K., Krasnow, M.A., 1996. Development of the Drosophila tracheal system occurs by a series of morphologically distinct but genetically coupled branching events. Development 122, 1395±1407. Scavo, L.M., Ertsey, R., Chapin, C.J., Allen, L., Kitterman, J.A., 1998. Apoptosis in the development of rat and human fetal lungs. Am. J. Resp. Cell Mol. Biol. 18, 21±31. Schuger, L., O'Shea, S., Rheinheimer, J., Varani, J., 1990a. Laminin in lung development: effects of anti-laminin antibody in murine lung morphogenesis. Dev. Biol. 137, 26±32. Schuger, L., O'Shea, K.S., Nelson, B.B., Varani, J., 1990b. Organotypic arrangement of mouse embryonic lung cells on a basement membrane extract: involvement of laminin. Development 110, 1091±1099. Schuger, L., Varani, J., Killen, P.D., Skubitz, A.P., Gilbride, K., 1992. Laminin expression in the mouse lung increases with development and stimulates spontaneous organotypic rearrangement of mixed lung cells. Dev. Dyn. 195, 43±54. Schuger, L., Skubitz, A.P.N., Morenas, A., Gilbride, K., 1995. Two separates domains of laminin promote lung organogenesis by different mechanisms of action. Dev. Biol. 169, 520±532. Schuger, L., Skubitz, A.P.N., Gilbride, K., Mandel, R., He, L., 1996. Laminin and heparan sulfate proteoglycan mediate epithelial cell polarization in organotypic cultures of embryonic lung cells: evidence implicating involvement of the inner globular region of laminin B1 chain and the heparan sulfate groups of heparan sulfate proteoglycan. Dev. Biol. 179, 264±273.

80

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81

Schuger, L., Skubitz, A.P., Zhang, J., Sorokin, L., He, L., 1997. Laminin alpha1 chain synthesis in the mouse developing lung: requirement for epithelial-mesenchymal contact and possible role in bronchial smooth muscle development. J. Cell Biol. 139, 553±562. Schuller, A.G., van Neck, J.W., Beukenholdt, R.W., Zwarthoff, E.C., Drop, S.L., 1995. IGF, type I IGF receptor and IGF-binding protein mRNA expression in the developing mouse lung. J. Mol. Endocrinol. 14, 349± 355. Schwarz, M., Lee, M., Zhang, F., Zhao, J., Jin, Y., Smith, S., Bhuva, J., Stern, D., Warburton, D., Starnes, V., 1999. EMAP II: a modulator of neovascularization in the developing lung. Am. J. Physiol. 276, L365± L375. Sekelsky, J.J., Newfeld, S.J., Raftery, L.A., Chartoff, E.H., Gelbart, W.M., 1995. Genetic characterization and cloning of mothers against dpp, a gene required for decapentaplegic function in Drosophila melanogaster. Genetics 139, 1347±1358. Sekine, K., Ohuchi, H., Fujiwara, M., Yamasaki, M., Yoshizawa, T., Sato, T., Yagishita, N., Matsui, D., Koga, Y., Itoh, N., Kato, S., 1999. Fgf-10 is essential for limb and lung formation. Nat. Genet. 21, 138±141. Senior, R.M., Grif®n, G.L., Mudd, M.S., Moxley, M.A., Longmore, W.J., Pierce, R.A., 1996. Entactin expression by rat lung and rat alveolar epithelial cells. Am. J. Resp. Cell Mol. Biol. 14, 239±247. Serra, R., Moses, H.L., 1995. pRb is necessary for inhibition of N-myc expression by TF-b1 in embryonic lung organ cultures. Development 121, 3057±3066. Serra, R., Pelton, R.W., Moses, H.L., 1994. TGF beta 1 inhibits branching morphogenesis and N-myc expression in lung bud organ cultures. Development 120, 2153±2161. Seth, R., Shum, L., Wu, F., et al., 1993. Role of epidermal growth factor expression in early mouse embryo lung branching morphogenesis in culture: antisense oligodeoxynucleotide strategy. Dev. Biol. 158, 555± 559. Shannon, J.M., 1994. Induction of alveolar type II cell differentiation in fetal tracheal epithelium by grafted distal lung mesenchyme. Dev. Biol. 166, 600±614. Shannon, J.M., Gebb, S., Nielsen, L.D., 1999. Induction of alveolar type II cell differentiation in embryonic tracheal epithelium in mesenchymefree culture. Development 126, 1675±1688. Shiratori, M., Michalopoulos, G., Shinozuka, H., Singh, G., Ogasawara, H., Katyal, S.L., 1995. Hepatocyte growth factor stimulates DNA synthesis in alveolar epithelial type II cells in vitro. Am. J. Resp. Cell Mol. Biol. 12, 171±180. Shiratori, M., Oshika, E., Ung, L.P., Singh, G., Shinozuka, H., Warburton, D., Michalopoulos, G., Katyal, S.L., 1996. Keratinocyte growth factor and lung morphogenesis. Am. J. Resp. Cell Mol. Biol. 15, 328±338. Simonet, W.S., DeRose, M., Bucay, N., Nguyen, H.Q., Wert, S., Zhou, L., Ulich, T.R., Danilenko, D.M., Whitsett, J.A., 1995. Pulmonary malformation in transgenic mice expressing human keratinocyte growth factor in the lung. Proc. Natl. Acad. Sci. USA 29, 12461±12465. Sonnenberg, E., Weidner, K.M., Birchmeier, C., 1993. Scatter factor/hepatocyte growth factor and its receptor, the c-met tyrosine kinase, can mediate a signal exchange between mesenchyme and epithelia during mouse development. J. Cell Biol. 123, 223±235. Souza, P., Kuliszewski, M., Wang, J., Tseu, I., Tanswell, A.K., Post, M., 1995. PDGF-AA and its receptor in¯uence early lung branching via an epithelial±mesenchymal interaction. Development 121, 2559±2567. Speirs, V., Bienkowski, E., Wong, V., Cutz, E., 1993. Paracrine effects of bombesin/gastrin-releasing peptide and other growth factors on pulmonary neuroendocrine cells in vitro. Anat. Rec. 236, 54±61. Spemann, H., 1938. Embryonic Development and Induction, Yale University Press, New Haven, CT. Spooner, B.S., Hardman, P., Paulsen, A., 1994. Gravity in mammalian organ development: differentiation of cultured lung and pancreas rudiments during space¯ight. J. Exp. Zool. 269, 212±222. Strandjord, T.P., Sage, E.H., Clark, J.G., 1995. SPARC participates in the branching morphogenesis of developing fetal rat lung. Am. J. Resp. Cell Mol. Biol. 13, 279±287.

Sunday, M.E., Hua, J., Dai, H.B., Nusrat, A., Torday, J.S., 1990. Bombesin increases fetal lung growth and maturation in utero and in organ culture. Am. J. Resp. Cell Mol. Biol. 3, 199±205. Sunday, M.E., Hua, J., Reyes, B., Masui, H., Torday, J.S., 1993. Antibombesin monoclonal antibodies modulate fetal mouse lung growth and maturation in utero and in organ cultures. Anat. Rec. 236, 25±34. Sundell, C., Roman, J., 1994. Control of lung vasculogenesis by extracellular matrix composition. Mol. Biol. Cell 5, 179a. Supp, D.M., White, D.P., Potter, S.S., Brueckner, M., 1997. Mutation of an axonemal dynein affects left-right asymmetry in inversus viscerum mice. Nature 389, 963±966. Sutherland, D., Samakovlis, C., Krasnow, M.A., 1996. breathless encodes a Drosophila FGF homolog that controls tracheal cell migration and the pattern of branching. Cell 87, 1091±1101. Takahashi, H., Ikeda, T., 1996. Transcripts for two members of the transforming growth factor-beta superfamily BMP-3 and BMP-7 are expressed in developing rat embryos. Dev. Dyn. 207, 439±449. Tefft, D.T., Lee, M., Smith, S., Leinwand, M., Zhao, J., Bringas, P., Crowe, D.L., Warburton, D., 1999. Conserved function of mSpry-2, a murine homolog of Drosophila sprouty, which negatively modulates respiratory organogenesis. Curr. Biol. 9, 219±222. Tenhave-Opbroek, A., 1981. The development of the lung in mammals: an analysis of concepts and ®ndings. Am. J. Anat. 152, 201±219. Thomas, T., Dziadek, M., 1994. Expression of collagen a1(IV), laminin and nidogen genes in the embryonic mouse lung: implications for branching morphogenesis. Mech. Dev. 45, 193±201. Thurlbeck, W., 1991. Pre- and postnatal organ development. In: Chernik, V., Mellins, B. (Eds.). Basic Mechanisms of Pediatric Respiratory Disease: Cellular and Integrative, Decker, Philadelphia, PA, pp. 23± 35. Timpl, R., 1996. Macromolecular organization of basement membranes. Curr. Opin. Cell Biol. 8, 618±624. Timpl, R., Brown, J.C., 1996. Supramolecular assembly of basement membranes. BioEssays 18, 123±132. Toonen, R.F., Gowan, S., Bingle, C.D., 1996. The lung enriched transcription factor TTF-1 and the ubiquitously expressed proteins Sp1 and Sp3 interact with elements located in the minimal promoter of the rat Clara cell secretory protein gene. Biochem. J. 316, 467±473. Urase, K., Mukasa, T., Irigashi, H., Ishii, Y., Yasugi, S., Momoi, M.Y., Momoi, T., 1996. Spatial expression of Sonic hedgehog in the lung epithelium during branching morphogenesis. Biochem. Biophys. Res. Commun. 225, 161±166. Venkatesh, V.C., Planer, B.C., Schwartz, M., Vanderbilt, J.N., White, R.T., Ballard, P.L., 1995. Characterization of the promoter of human pulmonary surfactant protein B. Am. J. Physiol. Lung Cell Mol. Biol. 268, L674±L682. Virtanen, I., Laitinen, A., Tani, T., Paakko, P., Laitinen, L.A., Burgeson, R.E., Lehto, V.P., 1996. Differential expression of laminins and their integrin receptors in developing and adult human lung. Am. J. Resp. Cell Mol. Biol. 15, 184±196. Volpe, M.V., Martin, A., Vosatka, R.J., Mazzoni, C.L., Nielsen, H.C., 1997. Hoxb-5 expression in the developing mouse lung suggests a role in branching morphogenesis and epithelial cell fate. Histochem. Cell Biol. 108, 495±504. Vuolteenaho, R., Nissinen, M., Sainio, K., Byers, M., Eddy, R., Hirvonen, H., Shows, T.B., Sariola, H., Engvall, E., Tryggvason, K., 1994. Human laminin M chain (Merosin): complete primary structure, chromosomal assignment, and expression of the M and A chain in human fetal tissues. J. Cell Biol. 381, 381±394. Wang, T., Donahoe, P.K., Zervos, A.S., 1994. Speci®c interaction of type I receptors of the TGF-b family with the immunophilin FKBP-12. Science 265, 674±676. Wang, J., Kuliszewski, M., Yee, W., Sedlackova, L., Xu, J., Tseu, I., Post, M., 1995. Cloning and expression of glucocorticoid-induced genes in fetal rat lung ®broblasts. Transforming growth factor-b3. J. Biol. Chem. 270, 2722±2728. Wang, D., Yeger, H., Cutz, E., 1996. Expression of gastrin-releasing

D. Warburton et al. / Mechanisms of Development 92 (2000) 55±81 peptide receptor gene in developing lung. Am. J. Resp. Cell Mol. Biol. 14, 409±416. Warburton, D., Lee, M., 1999. Current concepts on lung development. Curr. Opin. Pediat. 11, 188±192. Warburton, D., Seth, R., Shum, P.G., Horcher, P., Hall, F.L., Werb, Z., Slavkin, H., 1992. Epigenetic role of epidermal growth factor expression and signaling in embryonic mouse lung branching morphogenesis. Dev. Biol. 149, 123±133. Warburton, D., Lee, M., Berberich, M.A., 1993. Molecular embryology and the study of lung development. Minisymposium Report, Cell & Developmental Biology Branch, Lung Division, NHLBI, NIH and The Society for Developmental Biology. Am. J. Resp. Cell. Mol. Biol. 9, 5±9. Warburton, D., Zhao, J., Berberich, M.A., Bern®eld, M., 1999. Molecular embryology of the lung: then, now, and in the future. NHLBI Sponsored Workshop Invited Review. NIH, Bethesda, MD, pp. L697±L704. Weaver, M., Yingling, J.M., Dunn, N.R., Bellusci, S., Hogan, B.L.M., 1999. Bmp signaling regulates proximal-distal differentiation of endoderm in mouse lung development. Development 126, 4005±4015. Weinstein, M., Xu, X., Ohyama, K., Deng, C.-X., 1998. FGFR-3 and FGFR-4 function cooperatively to direct alveogenesis in the murine lung. Development 125, 3615±3623. Wessels, N.K., 1970. Mammalian lung development: interactions in formation and morphogenesis of tracheal buds. J. Exp. Zool. 175, 455± 466. Weston, J.A., 1984. The embryonic neural crest: migration and differentiation and possible contributions to the developing lung. In: Becker, K.L., Gazdar, A.F. (Eds.). The Endocrine Lung in Health and Disease, W.B. Saunders, Philadelphia, PA, pp. 210±231. Whitsett, J., 1998. A lungful of transcription factors. Nat. Genet. 20, 7±8. Wiersdorff, T., Lecuit, T., Cohen, S.M., Mlodzik, M., 1996. Mad acts downstream of Dpp receptors, revealing a differential requirement for dpp signaling in initiation and propagation of morphogenesis in the Drosophila eye. Development 122, 2153±2162. Wilk, R., Weizman, I., Shilo, B.-Z., 1996. Trachealess encodes a bHLHPAS protein that is an inducer of tracheal cell fates in Drosophila. Genes Dev. 10, 93±102. Willey, J.C., Lechner, J.F., Harris, C.C., 1984. Bombesin and the C-terminal tetradecapeptide of gastrin-releasing peptide are growth factors for normal human bronchial epithelial cells. Exp. Cell Res. 153, 245±248. Wuenschell, C.W., Sunday, M.E., Singh, G., Minoo, P., Slavkin, H.C., Warburton, D., 1996. Embryonic mouse lung progenitor cells coexpress immunohistochemical markers of diverse mature cell lineages. J. Histochem. Cytochem. 44, 113±123. Yamaguchi, K., Shirakabe, K., Shibuya, H., Irie, K., Oishi, I., Ueno, N., Taniguchi, T., Nishida, E., Matsumoto, K., 1995. Identi®cation of a member of the MAPKKK family as a potential mediator of TGF-b mediated signal transduction. Science 270, 2008±2011. Yamaguchi, T.P., Dumont, D.J., Conlon, R.A., Breitman, M.L., Rossant, J., 1993. ¯k-1, an ¯t-related receptor tyrosine kinase is an early marker for endothelial cell precursors. Development 118, 489±498. Yamashita, H., Ichijo, H., Grimsby, S., Moren, A., ten Dijke, P., Miyazono, K., 1994. Endoglin forms a heteromeric complex with the signaling receptors for transforming growth factor-b. J. Biol. Chem. 269, 1995±2001. Yan, C., Sever, Z., Whitsett, J.A., 1995. Upstream enhancer activity in the

81

surfactant protein B gene is mediated by thyroid transcription factor 1. J. Biol. Chem. 270, 24852±24857. Yang, Y., Palmer, K.C., Relan, N., Diglio, C., Schuger, L., 1998. Role of laminin polymerization at the epithelial mesenchymal interface in bronchial myogenesis. Development 125, 2621±2629. Yang, Q., Ota, K., Tiam, Y., Kumar, A., Wada, J., Kashihara, N., Wallner, E., Kanwar, Y.S., 1999. Cloning of rat ®brillin-2 cDNA and its role in branching morphogenesis of the embryonic lung. Dev. Biol. 212, 229± 242. Youngson, C., Nurse, C., Yeger, H., Cutz, E., 1993. Oxygen sensing in airway chemoreceptors. Nature 365, 153±155. Zecca, M., Basler, K., Struhl, G., 1996. Direct and long-range action of a wingless morphogen gradient. Cell 87, 833±844. Zeng, X., Wert, S.E., Federici, R., Peters, K.G., Whitsett, J.A., 1998. VEGF enhances pulmonary vasculogenesis and disrupts lung morphogenesis in vivo. Dev. Dyn. 211, 215±227. Zhang, Y., Feng, X., We, R., Derynck, R., 1996. Receptor-associated Mad homologues synergize as effectors of the TGF-b response. Nature 383, 168±172. Zhang, L., Whitsett, J.A., Stripp, B.R., 1997. Regulation of Clara cell secretory protein gene transcription by thyroid transcription factor-1. Biochim. Biophys. Acta 1350, 359±367. Zhang, Y., Musci, T., Derynck, R., 1997. The tumor suppressor gene Smad4/DPC 4 as a central mediator of Smad function. Curr. Biol. 7, 270±276. Zhao, J., Bu, D., Lee, M., Slavkin, H.C., Hall, F.L., Warburton, D., 1996. Abrogation of transforming growth factor-beta type II receptor stimulates embryonic mouse lung branching morphogenesis in culture. Dev. Biol. 180, 242±257. Zhao, J., Sime, P.J., Bringas Jr., P., Gauldie, J., Warburton, D., 1998a. Epithelium-speci®c adenoviral transfer of a dominant-negative mutant TGF-beta type II receptor stimulates embryonic lung branching morphogenesis in culture and potentiates EGF and PDGF-AA. Mech. Dev. 72, 89±100. Zhao, J., Tefft, J.D., Lee, M., Smith, S., Warburton, D., 1998b. Abrogation of betaglycan attenuates TFG-beta-mediated inhibition of embryonic murine lung branching morphogenesis in culture. Mech. Dev. 75, 67± 79. Zhou, Y., Young, S.L., 1995. Expression of transforming growth factorbeta type II receptor in rat lung is regulated during development. Am. J. Physiol. Lung Cell Mol. Physiol. 269, L419±L426. Zhou, L., Graeff, R.W., McCray Jr., P.B., Simonet, W.S., Whitsett, J.A., 1996a. Keratinocyte growth factor stimulates CFTR-induced ¯uid secretion in the fetal lung in vitro. Am. J. Physiol. Lung Cell Mol. Physiol. 15, L987±L994. Zhou, L., Lim, L., Costa, R.H., Whitsett, J.A., 1996b. Thyroid transcription factor-1, hepatocyte nuclear factor-3 beta, surfactant protein B, C, and Clara cell secretory protein in developing mouse lung. J. Histochem. Cytochem. 44, 1183±1193. Zhou, L.J., Dey, C., Wert, S.E., Whitsett, J.A., 1996c. Arrested lung morphogenesis in transgenic mice bearing an SP-C-TGF-beta 1 chimeric gene. Dev. Biol. 175, 227±238. Zhou, L., Dey, D.R., Wert, S.E., Yan, C., Costa, R.H., Whitsett, J.A., 1997. Hepatocyte nuclear factor-3beta limits cellular diversity in the developing respiratory epithelium and alters lung morphogenesis in vivo. Dev. Dyn. 210, 305±314.