3957
Development 126, 3957-3968 (1999) Printed in Great Britain © The Company of Biologists Limited 1999 DEV2432
Synergistic activities of α3 and α6 integrins are required during apical ectodermal ridge formation and organogenesis in the mouse Adèle De Arcangelis1, Manuel Mark1, Jordan Kreidberg2, Lydia Sorokin3 and Elisabeth Georges-Labouesse1,* 1Institut de Génétique et de Biologie Moléculaire et Cellulaire, CNRS/INSERM/ULP, BP 163, 67404 Illkirch, CU de Strasbourg, France 2Department of Medicine, Children’s Hospital, and Department of Pediatrics, Harvard Medical School, Boston, MA 02115, USA 3IZKF, Nikolaus Fiebiger Center, Friedrich Alexander Universität, Erlangen-Nuremberg, Germany
*Author for correspondence (e-mail:
[email protected])
Accepted 4 June; published on WWW 5 August 1999
SUMMARY β1 and α6β β4 are cell surface receptors for Integrins α6β laminins. Integrin α6-null mice die at birth with severe skin blistering and defects in the cerebral cortex and in the β1 can associate with laminins and other retina. Integrin α3β ligands. Integrin α3-null mice also die at birth, with kidney and lung defects at late stages of development, and moderate skin blistering. To investigate possible overlapping functions between α3 and α6 integrins, we α6−/−− mutant analyzed the phenotype of compound α3−/−−/α embryos. Double homozygous mutant embryos were growth-retarded and displayed several developmental defects not observed in the single mutant animals. First, limb abnormalities characterized by an absence of digit separation and the fusion of preskeletal elements were observed. Further analyses indicated a defect in the apical ectodermal ridge, an essential limb organizing center. In the double mutant, the ridge appeared flattened, and ridge cells did not show a columnar morphology. A strong
reduction in ridge cell proliferation and alterations of the basal lamina underlying the ectoderm were observed. These results suggest that α3 and α6 integrins are required for the organization or compaction of presumptive apical ectodermal ridge cells into a distinct differentiated structure. Additional defects were present: an absence of neural tube closure, bilateral lung hypoplasia, and several abnormalities in the urogenital tract. Finally, an aggravation of brain and eye lamination defects was observed. The presence of novel phenotypes in double mutant embryos demonstrates the synergism between α3 and α6 integrins and their essential roles in multiple processes during embryogenesis.
INTRODUCTION
being involved in cytoskeleton rearrangement as well as signal transduction (Howe et al., 1998). Thus, integrins participate in the control of cell behaviour, proliferation, survival and differentiation. To understand the role of integrins and the ECM in the organism, animal models have proved to be very valuable. The targeted inactivation of several ECM molecules and their receptors lead to embryonic lethality (Fässler et al., 1996; Hynes, 1996). This approach has allowed the definition of developmental processes that require integrins. For example, it has been shown that integrin α5β1, a major fibronectin receptor, is necessary for development of mesodermal derivatives in the posterior part of the mouse embryo, and for the survival of a subpopulation of neural crest cells (Yang et al., 1993; Goh et al., 1997). The integrin α6 chain associates with either β1 or β4 subunits, resulting in integrin heterodimers which recognize several laminin isoforms (Mercurio and Shaw, 1991; Delwel and Sonnenberg, 1996). As described previously, mice lacking
Interactions of cells with their environment and with neighbouring cells are processes essential to embryonic development. Integrins represent the main cell surface receptors that mediate cell-matrix and cell-cell interactions (Hynes, 1992). They are heterodimeric transmembrane receptors composed of noncovalently associated α and β chains. To date, 22 integrin heterodimers have been characterized, which can recognize extracellular matrix (ECM) components such as the laminins, collagens and fibronectins (Mercurio and Shaw, 1991; Hynes, 1992). Generally, one ligand is recognized by several integrin heterodimers, and, conversely, many integrins can bind several ligands. Numerous functions have been attributed to integrin receptors: attachment to and organization of ECM, and control of cell shape and cell motility (Hynes, 1992, and Howe et al., 1998 for recent review). It has been well documented that integrins provide an essential link between the outside and the inside of the cell,
Key words: α3 integrin, α6 integrin, Laminins, Basal lamina, Gene knock-out, Limb, Apical ectodermal ridge, Lung, Urogenital tract, Brain, Eye, Mouse
3958 De Arcangelis and others the α6 integrin chain die at birth with severe skin blistering and abnormal cerebral cortical and retinal lamination (GeorgesLabouesse et al., 1996, 1998). Even though these defects are severe, given the early expression of α6 during embryogenesis (Sutherland et al., 1993) it is unexpected that integrin α6-null embryos can develop to birth. Although a function for α6 integrins in branching morphogenesis has been inferred by its tissue-specific distribution pattern and function-blocking experiments (Sorokin et al., 1990), no obvious abnormalities of branching morphogenesis were observed in α6-null fetuses. This suggested that other molecules may have overlapping functions with α6 integrins. Indeed, integrin α3β1, present in epithelia and the nervous system, can also bind various laminins, among other ligands (Carter et al., 1991; DiPersio et al., 1995, 1997). In contrast to α6 integrins, it shows a weak binding to laminin-1, but like α6 integrins, it is a prominent receptor for laminin-5, the major laminin isoform in the subepidermal basal lamina (BL) (Carter et al., 1991; DiPersio et al., 1995, 1997). Integrin α3-null mice die at birth and exhibit kidney and lungs defects associated with disorganization of the BLs, and notably, moderate skin blistering (Kreidberg et al., 1996; DiPersio et al., 1997). The similarities between α6 and α3 integrins raised the possibility that they may have overlapping functions. To investigate such overlapping functions, we have produced mice carrying null mutations in both the integrin α3 and α6 genes and have analyzed their phenotypes. This analysis has revealed their synergistic activities in limbs, lung and the urogenital system, and provides new insight into the functions of integrin/laminin interactions during embryonic development. In particular, a role for these two integrins in the formation of the apical ectodermal ridge (AER), a limb organizing center, has been identified.
Immunofluorescence Embryos were collected in PBS and frozen directly or fixed (E10.5E11.5) in 4% paraformaldehyde for 3-4 hours at 4°C, and placed in 20% sucrose in PBS overnight at 4°C before freezing. Frozen sections (10-12 µm) were processed as described previously (GeorgesLabouesse et al., 1996). Antibodies included a polyclonal anti-integrin α3 (8-4; DiPersio et al., 1995), a rat monoclonal anti-integrin α6 (GoH3; Serotec), a polyclonal anti-integrin β1 (gift from S. Johansson), a rat monoclonal anti-integrin β4 (346-11A; Pharmingen), a polyclonal anti-Engelbreth-Holm-Swarm (EHS) laminin (Sigma), rat monoclonal antibodies to laminin α1 (198) (Sorokin et al., 1992) and α5 (4G6) chains (Sorokin et al., 1997b), a polyclonal anti-type IV collagen (Chemicon), a rat monoclonal antiheparan sulphate proteoglycan (HSPG) (Chemicon), and a rat monoclonal anti-E-cadherin (Sigma). Secondary antibodies were Cy3-conjugated donkey anti-rabbit and tetramethylrhodamine isothiocyanate-conjugated rabbit anti-rat (Sigma). Signals were analyzed by confocal microscopy.
MATERIALS AND METHODS
Whole-mount in situ hybridizations Whole-mount in situ hybridizations were performed as described by Décimo et al. (1995). Proteinase K treatment of embryos were for only four minutes at 5 µg/ml to preserve surface ectoderm and the AER (Haramis et al., 1995). FGF-4 (Hebert et al., 1990) and FGF-8 (Crossley and Martin, 1995) cDNA probes were obtained from G. Martin, and Msx-1 cDNA probe from B. Robert (Robert et al., 1991).
Mice and genotyping Integrin α6+/− mice (Georges-Labouesse et al., 1996) and α3+/− mice (Kreidberg et al., 1996) were intercrossed to generate the double heterozygotes used for double null homozygote production. Genotypes of embryos were determined by PCR from yolk sac DNAs, in conditions already established for the α3 mutation (Kreidberg et al., 1996). For the α6 mutation, the following primers were used: wild-type allele, 5′-GTGATAACTCCAGCTTGTGTGTCAAG-3′ and 5′-CCTCTGCAGCGGGAGT-GCTTC-3′, mutant allele, 5′-TCAGAGCAGCCGATTGTC-3′ and 5′-GCAGATGTGATCCCCTTC-3′. PCR was performed in the presence of 10% dimethylsulfoxide. After 1 minute of denaturation at 94°C, 35 cycles of amplification were performed at 94°C (30 seconds), 61°C (30 seconds), and 72°C (1 minute). The amplified fragment lengths were 0.5 kb for the wild-type allele and 0.8 kb for the mutant allele. Fetuses illustrated as controls were littermates of wild-type or heterozygote genotype. Nile blue sulphate, skeletal and histological analyses Embryonic day 13.5 (E13.5) limbs were stained with Nile blue sulphate (3 µg/ml) in phosphate-buffered saline (PBS) for 30 minutes, rinsed in PBS, and photographed immediately after. Developing cartilage of E12.5-E16.5 embryos was analyzed by whole-mount Alcian blue staining as described by Jegalian and De Robertis (1992). Processing of embryos for histological analyses was as described by Georges-Labouesse et al. (1996).
Electron microscopy For transmission and scanning electron microscopy, E10.5 embryos and E11.5 limb buds were fixed in 2.5% glutaraldehyde in PBS overnight at 4°C and processed using standard procedures. Bromodeoxyuridine (BrdU) labelling BrdU (50 mg/kg body weight) was injected intraperitoneally into pregnant females 2 hours before being killed. Embryos were fixed in Bouin’s fixative, dehydrated and paraffin embedded. Limbs were processed separately. Transverse sections (7 µm) through the AER were prepared. Sections were stained with a mouse anti-BrdU monoclonal antibody (Boehringer Mannheim), and a peroxidaseconjugated secondary antibody (Vector Laboratories). Following detection with the 3,3′-diaminobenzidine substrate kit (Vector Laboratories), sections were counterstained with hematoxylin. The AER was defined by cell and tissue morphology. Total cells and BrdUpositive cells were counted in 12 sections displaying a distinct AER and in a defined area within 0.17 mm in the mesenchyme underlying the AER.
RESULTS α6ⴚ/ⴚ compound Embryonic defects in integrin α3ⴚ/ⴚ/α mutant embryos Double heterozygotes obtained from intercrosses between α6+/− and α3+/− mice appeared normal and were fertile. Both α3-null and α6-null mutations are recessively lethal at birth. To look for possible genetic interactions between the two mutations, embryos obtained from intercrosses between α3+/−/α6+/− mice were analyzed at different stages of development. From E10.5 to E16.5, the expected (1:16) Mendelian ratio of α3−/−/α6−/− embryos (47 out of 611 analyzed) was recovered, thus providing no evidence for early embryonic lethality. However, double mutant embryos were growth-retarded, in particular at E16.5, and had multiple defects. From E13.5 on, all compound α3−/−/α6−/− embryos were
α3 and α6 integrins compound mutant mice 3959
Fig. 2. Absence of a delimitating epithelium in the distal mutant limbs. (a,b) Semithin sections of E13.5 control (a) and α3−/−/α6−/− mutant (b) limbs. (c,d) EHS laminin immunostaining of distal E13.5 control (c) and mutant (d) limbs. Arrowheads in b and d point to the ectopic epithelial sheet in the mesenchymal compartment. A strong laminin staining delineates the folded epithelium (d) while the signal is restricted to the epithelial/mesenchymal interface in the wild-type limb (c). Endothelial BLs are similarly labelled in control and mutant limbs. e, epithelium; m, mesenchyme. Magnification: ×100.
observed in 43% of the compound mutant animals. Limb and neural tube defects were never observed in single α3−/− and α6−/− mutants or in α3−/−/α6+/− or α3+/−/α6−/− embryos.
Fig. 1. Failure of neural tube closure and limb defects in α3−/−/α6−/− double mutant embryos. (a) Control and (b) α3−/−/α6−/− embryos at E13.5. White arrowheads in b indicate the absence of neural tube closure from the midbrain/hindbrain boundary to the abnormally kinked tail. Note the abnormal shape of the mutant limbs, with no visible digital rays. (c,d) Dorsal view of E13.5 control (c) and double mutant (d) hindlimbs stained with Nile blue sulphate. Black arrowheads indicate the labelled interdigital apoptotic areas. (d) No staining and no distinct digits are observed in the central region of the mutant limb. (e-h) Whole-mount Alcian blue staining of cartilaginous condensations in E13.5 limbs. Wild-type (e,g) and mutant (f,h) forelimbs (e,f) and hindlimbs (g,h), dorsal views. Black arrows show the close apposition of the terminal phalanges between digits II-IV and the fusion of the first phalanges between digits II and III in the mutant forelimb (f) and hindlimb (h). c, carpals; mc, metacarpals; ph, phalanges; t, tarsals; mt, metatarsals.
readily identified by characteristic abnormalities affecting both fore- and hindlimbs (Fig. 1a,b). Mutant limbs were abnormally shaped and lacked digit separation as illustrated in Fig. 1b,d. Moreover, varying degrees of failure of neural tube closure (failure of neural folds apposition at the midbrain-hindbrain boundary, exencephaly and a kinked tail) (Fig. 1b) were
Analysis of interdigital cell death One explanation for the limb abnormalities in the double mutant embryos would be that the process of interdigital cell death normally associated with digit separation was impaired. Vital staining using Nile Blue sulphate was therefore performed at E13.5 to visualize apoptotic cells in the limbs. In the double mutants, no apoptotic cells were observed in the central region corresponding to digits 2, 3 and 4 (Fig. 1d), whereas stained cells were clearly present in the interdigital zones of the corresponding wild-type limbs (Fig. 1c). However, apoptotic cells were observed in the regions between digits 1 and 2, and digits 4 and 5 in double mutant embryos, consistent with the fact that individualisation of digits 1 and 5 was observed at later stages in some double mutant limbs (data not shown). These results indicate that the apoptotic process, although impaired, is not completely abolished in double mutant limbs. Limb skeletal abnormalities To investigate whether the absence of digit separation was associated with skeletal abnormalities, whole-mount Alcian blue stainings were performed at E12.5, E13.5 and E16.5. At E12.5, the fore- and hindlimb footplates were irregularly shaped, as compared to the rounded wild-type footplates (data not shown). In all double mutant limbs examined at E13.5 (n=12) and E16.5 (n=8), complete fusion or close apposition of the chondrogenic elements of the central two or three digits were observed (Fig. 1f,h). These fusions involved terminal phalanges at E13.5 (Fig. 1e-h) and phalanges 1 and/or 2 at E16.5 (data not shown). In some cases, additional defects were present, such as loss or atrophy of digits, or abnormalities of
3960 De Arcangelis and others carpal and metacarpal, or of tarsal and metatarsal cartilages (data not shown). At the stages examined, no abnormalities of more proximal skeletal elements were apparent. Disruption of the surface ectoderm Histological analysis at E13.5 revealed that the most distal part of the double mutant limb, corresponding to the tip of the fused digits lacked a surface epithelium (Fig. 2a,b). Intriguingly, ectopic epithelial cell layers were found within the mesenchymal tissue (Fig. 2a,b) from which they were separated by the BL components laminin (Fig. 2c,d) and type IV collagen (data not shown). At the ultrastructural level a BL, although discontinuous, was detected between the epithelium and the mesenchyme (data not shown). This may be explained by rupture of the surface epithelium in the distal part of the limbs, followed by protrusion of the mesenchymal cells. No such rupture was observed in the double mutant limbs at E12.5 (not shown). Integrin α3 and α6 chains are present in the apical ectodermal ridge BL alterations, which may result in disruption of the epithelial sheet and outgrowth of mesenchymal cells, could explain some of the defects observed in the double mutant limbs. However, mutant limbs already appeared abnormal before the rupture of the epithelium, suggesting that earlier defects may account for the phenotype. Fusion of digits has been correlated with alterations in the properties or function of the apical ectodermal ridge (AER), a specialized epithelium located at the dorsoventral limb interface and an essential signalling center involved in limb outgrowth. For instance, mice carrying mutations in a member of the notch signalling pathway (serrate-2 or jagged-2) (Sidow et al., 1997; Jiang et al., 1998) exhibit fusion of distal skeletal elements linked to a hyperplastic AER. Another mouse mutant for which the responsible gene has not been identified (dactylaplasia) also shows severe limb defects (loss and fusion of digits,
Fig. 4. Morphological alterations of the AER in the α3−/−/α6−/− mutant limb buds. (a,b) Scanning electron micrographs of E10.5 control (a) and double mutant (b) forelimb buds. Control limb buds exhibit a distinct, thickened ridge whereas mutant limbs display a broad and flattened ridge, barely detectable in this sample. (c,d) Semithin sections of E11.5 control (c) and mutant (d) forelimb buds. Cell shape and arrangement in the mutant AER (d) is altered as compared to the densely packed columnar cells in the control AER (c). (e,f) E-cadherin immunostaining of E11.5 control (e) and mutant (f) limb buds. aer, apical ectodermal ridge; e, non-ridge epithelium; m, mesenchyme; p, periderm. Magnifications: (a,b) ×80, (c-f) ×270.
Fig. 3. α3 and α6 integrins are expressed in the AER. Immunostaining of α3 (a), α6 (b), β1 (c) and β4 (d) integrin chains in wild-type E10.5 forelimb buds; confocal images. Note the enrichment of α6 chain in the AER (b). aer, apical ectodermal ridge; e, non-ridge epithelium; m, mesenchyme. Magnification: ×285.
monodactyly) associated with degeneration and a reduction in AER activity at E11.5 (Crackower et al., 1998). Limb deformity (ld) mice also display skeletal abnormalities associated with defects in the AER (Zeller et al.,1989; Haramis et al., 1995; Kuhlman and Niswander, 1997). To analyze if an AER defect could be responsible for the phenotype observed in the α3−/−/α6−/− mice, it was necessary
α3 and α6 integrins compound mutant mice 3961
Fig. 5. Altered composition and structure of the BL in the α3−/−/α6−/− limbs. Immunostaining of EHS laminin (a,e), HSPG (b,f), α1 laminin chain (c,g), and α5 laminin chain (d,h) in control (a-d) and mutant (e-h) E10.5 forelimb buds, visualized by confocal microscopy. Arrows indicate the epithelial/mesenchymal interface. Note the broader and irregular EHS laminin and HSPG signals in mutant limbs, as compared to the ones in control buds (e, f versus a, b). (c,d,g,h) Reduced and discontinuous α1 and α5 laminin chain signals as compared to the strong linear staining in the wild-type buds (g, h versus c, d). Note the presence of abnormal positive material within the mutant AER. (i-k) Ultrastructural analysis of the BL underlying the AER in control (i) and mutant (j,k) E11.5 limbs. Arrows point to the continuous dense BL in the control (i) and to near-normal BL portions in the mutant (j,k). The arrowheads in j show abnormal deposits of BL material. Asterisk in k indicates the detachment of the mutant BL from the epithelial compartment. aer, apical ectodermal ridge; e, epithelium; m, mesenchyme; LN, laminin. Magnifications: (a-h) ×115, (i-k) ×18300.
to determine the localization of these integrins in early limb buds. Immunohistochemistry was therefore performed on wildtype E10.5 (Fig. 3) and E11.5 limb buds (data not shown). Integrin α3 chain was mainly detected along the entire ectodermal surface, delineating the basolateral plasma membrane of AER and non-ridge epithelial cells (Fig. 3a). Interestingly, α6 integrin chain was more strongly expressed in the AER columnar cells than in the non-ridge epithelium (Fig. 3b). Integrin β1 chain was expressed throughout the AER (Fig. 3c), and at the epithelial-mesenchymal interface. By contrast, only a faint signal was observed for the β4 chain at the basal surface of epithelial cells, with no enrichment in the AER (Fig. 3d). The distribution of α3 and α6 integrins therefore suggests that both integrins play a role in the formation and/or maintenance of the AER, and that their absence could lead to alterations of AER properties. Morphological defects of the apical ectodermal α6ⴚ/ⴚ embryos ridge in α3ⴚ/ⴚ/α To assess alterations in the AER, scanning electron microscopy was performed on early limb buds at E10.5 (Fig. 4a,b) and E11.5 (data not shown). Compared to controls, the AER of α3−/−/α6−/− limbs was barely detectable, and flattened. These morphological differences between wild-type and double mutants were also apparent under the light microscope (Fig. 4c,d). The AER of double mutants appeared enlarged and
seemed to contain a higher number of cells compared to controls. In addition, cells of the AER in the double mutants looked smaller, and were less columnar in shape than cells of the AER of wild-type embryos (Fig. 4c,d). To better visualize these morphological differences, an antibody against Ecadherin, a specific epithelial cell marker, was used on sections from E11.5 wild-type and double mutant limbs. This staining confirmed the striking differences between wild-type and double mutants both in the structure of the AER itself, and in the shape of individual AER cells (Fig. 4e,f). Integrins as receptors for BL components Interactions between epithelial cells and the BL are known to be essential for epithelial morphogenesis, differentiation, and establishment of cell polarity (Ashkenas et al., 1996; De Arcangelis et al., 1996). Laminins are the main ligands for integrins α3β1, α6β1 and α6β4, and are present in the BL separating the limb epithelial and mesenchymal compartments (Critchlow and Hinchliffe, 1994; Miner et al., 1998). To determine whether alterations in the BL were associated with changes in the morphology of the AER, immunohistochemistry with antibodies against laminins, type IV collagen and HSPG were performed at E10.5 (Fig. 5) and E11.5 (data not shown). In the wild-type limb bud, the anti-EHS laminin antibody gave a strong signal along the BL underlying the AER and non-ridge epithelium, in the BL surrounding blood vessels and, at a lower
3962 De Arcangelis and others
Fig. 6. Reduction of the cell proliferation rate in the mutant AER. (a,b) Analysis of BrdU incorporation on sections from control (a) and mutant (b) E10.5 forelimb buds showing the strong reduction of BrdU-labelled cells (brown nuclei) in the mutant AER. aer: apical ectodermal ridge; e: non-ridge epithelium; m: mesenchyme; p: periderm. (c,d) Percentage of BrdU-labelled cells (BrdU-positive cells per total cell number, mean value for 12 sections) in control and double mutant AERs of E10.5 forelimb and E11.5 hindlimb buds (c) and in the corresponding underlying mesenchyme (d) +/− standard error measurement. Magnification: ×240.
level, in the mesenchyme (Fig. 5a). A similar pattern was observed with antibodies against HSPG (Fig. 5b) and type IV collagen (data not shown). In the double mutants, however, altered staining patterns were observed for both anti-EHS laminin (Fig. 5e) and anti-HSPG (Fig. 5f). In both cases, the immunofluorescence signal appeared broader (not as defined) and discontinuous in the BL separating the epithelium from the mesenchyme, and underneath the AER. To more specifically analyze the BL alterations, the distributions of laminin α1 and α5 chains, both of which are expressed in the ectodermal BL (Miner et al., 1998; L. S., unpublished data) were determined. In wild-type E10.5 (Fig. 5) and E11.5 (data not shown) limb buds, both laminin α1 and α5 chains were detected in the BL underlying the non-ridge epithelium and the AER (Fig. 5c,d). Laminin α1 signal was also detected in limb buds of α3−/−/α6−/− embryos, but the signal was less intense and irregular, compared to wild-type animals. In addition, abnormal deposits of laminin α1 were present within the epithelium (Fig. 5g). Similarly, laminin α5 was still detectable in double mutant limbs, however, a strong reduction of the signal was apparent and areas of discontinuities in the BL of non-ridge and AER epithelium were noted at E10.5 (Fig. 5h) and E11.5 (data not shown). Areas of BL disorganization were also identified by electron microscopy. At E11.5, a well-defined BL was present in the wild-type limb at the interface of the AER and underlying mesenchyme (Fig. 5i). By contrast, although a BL was present in the double mutant limbs (n=3), ultrastructural alterations were visible (Fig. 5j,k). The BL appeared generally thinner than in the wild-type limb (Fig. 5k), and in some areas was not tightly apposed to the plasma membrane of epithelial cells (Fig. 5k). Interruptions in the BL and a marked disorganization
Fig. 7. Expression of FGF-8, FGF-4 and Msx-1 in double mutant AERs. Whole-mount in situ hybridizations of control (a,c,e) and α3−/−/α6−/− limbs (b,d,f). (a,b) E11.5 forelimb buds. FGF-8 signal in the mutant AER (b) is decreased, and has an altered aspect compared to the control (a). (c,d) FGF-4 signal in (c) control or (d) mutant E10.5 forelimbs. Note the irregular aspect of the FGF-4 signal in (d). (e,f) E10.5 hindlimbs. Msx-1 is expressed in the mesenchyme in both control (e) and mutant (f) limbs. A slight decrease in the posterior mesenchyme of the mutant limb (f) is observed.
resulting in an accumulation of extracellular material and debris within the extracellular space between the AER and the mesenchyme were also observed (Fig. 5j). It should be noted that even in areas where the BL was present, cells did not have the typical columnar morphology of wild-type AER cells, but had a more flattened and less cohesive appearance (not shown). These alterations were only observed in compound α3−/−/α6−/− limbs, and not in α3−/− (n=3) or in α6−/− limbs (n=4). α3 and α6 integrins are required for cell proliferation in the AER The formation of the AER involves several steps of migration and compaction and a balance between cell proliferation and apoptosis. To better understand the origin of morphological differences in double mutant AERs, cell proliferation was assayed by BrdU labelling in utero. A dramatic decrease in cell proliferation was consistently observed in the AER from double mutant limbs at E10.5 (n=3) and E11.5 (n=2) (Fig. 6ac), and not in α3−/− (n=6) or α6−/− (n=4) limbs (data not shown). As an example, the proportion of proliferating cells within the AER fell from control levels of 28.5% to 4% in a E10.5 double mutant forelimb, and from 24% to 10.5% in a E11.5 double mutant hindlimb (Fig. 6c). This decrease in proliferation was restricted to the AER, since it was not observed in other tissues (data not shown). This suggests that α3 and α6 are required for the proliferation of AER cells.
α3 and α6 integrins compound mutant mice 3963
Fig. 8. Pulmonary and arterial defects in α3−/−/α6−/− mutants. Serial frontal histological sections through the lungs of a E13.5 wild type (WT) and of a α3−/−/α6−/− (α3/α6) fetus. (a, c) Anterior aspects, at the level where the right common carotid (rC) originates from the aorta. (b,d) More caudal aspects at the level of the right atrium (rAT). (e,f) Level of the left subclavian arteries (S). AA, arch of the aorta; aAO and AO, ascending and descending aortas, respectively; rAT, right atrium; B, terminal lung bud; C and rC, left and right carotid arteries, respectively; H, hemiazygos vein; L, left lung; crL, mL, cL and aL, cranial, middle, caudal and accessory lobes of the right lung, respectively; rL, single-lobed right mutant lung; O, oesophagus; P, pulmonary trunk; rROS, right retroesophageal subclavian artery; S, and rS, orthotopic left and right subclavian arteries, respectively; T, thymic lobe; V, left superior vena cava; VC, vertebral column. Magnifications: ×45.
However, as already mentioned, the total number of nuclei was higher in the double mutant as compared to the wild-type AER. From a few TUNEL assays, this could not be correlated to a consistent decrease in the extent of cell death in the mutant AER (data not shown). It is well established that AER function is crucial for stimulating the growth of mesenchymal cells. The alteration of AER properties observed here could thus lead to a reduction in the proliferation of the underlying mesenchymal cells. By counting the BrdU-labelled cells in the mesenchymal compartment of the double mutant limb, a slight decrease in their percentage (from 46% to 42.5% at E10.5, and from 47% to 40% at E11.5) was found (Fig. 6d). Mutant AER and fibroblast growth factors (FGFs) FGFs play a crucial role both in the formation and proliferation of the AER, and in mediating the AER influence on the underlying mesenchyme (Martin, 1998). Since morphological as well as proliferative defects occurred in double mutant limbs, it was important to verify FGF expression. By wholemount in situ hybridizations, an FGF-8 signal was found in the AER of mutant limbs at E10.5 (n=8) and E11.5 (n=2) (Fig. 7a,b). However, the signal was less intense in some limbs, and always presented alterations, in that the domain of expression was broader, patchy and irregular (Fig. 7b). In some limbs the signal was discontinuous. Since expression of FGF-4 has been shown to be altered in ld mutant mice which have defects in AER differentiation (Haramis et al., 1995; Kuhlman and Niswander, 1997), it was thus interesting to check for the expression of this FGF in α3−/−/α6−/− mutant limbs. In two double mutant animals that were examined at E10.5, a signal for FGF-4 was still present in fore- and hindlimb buds (Fig.
7c,d). At least in the forelimbs, the signal appeared irregular and sometimes interrupted. The expression of the Msx-1 homeogene has been reported to be directly dependent on AER function (Robert et al., 1991), and could thus be affected by the AER defects. In three double mutants, a Msx-1 signal was still found, although the domain of expression was slightly narrower in the posterior mesenchyme (Fig. 7e,f). Multiple organ defects In addition to limb and neural tube defects, histological analysis of four α3−/−/α6−/− fetuses at E13.5, E14.5, E15.5 and E16.5 revealed a pulmonary and pancreatic hypoplasia, as well as abnormalities in the urogenital tract, in the terminal portion of the digestive tract and in the pattern of cephalic arteries. These defects were not observed in α3−/− (n=3) and α6−/− (n=3) null mutants. The terminal lung buds (B) of E13.5 and E14.5 α3−/−/α6−/− mutants were fewer and larger than their wild-type counterparts (compare Fig. 8b and d). Moreover, at all developmental stages, the right lungs displayed a single lobe instead of the normal four (compare rL with crL, mL, cL and aL in Fig. 8a-f) and in the E15.5 fetus, the left lung was almost absent (not shown). The α3−/−/α6−/− pancreas was also consistently hypoplastic (data not shown). It is noteworthy that branching morphogenesis in α3−/−/α6−/− lungs and pancreas was not abolished, since terminal bronchi and pancreatic acini were present in mutants analyzed at E15.5 and E16.5. Branching morphogenesis in the α3−/−/α6−/− kidneys and salivary glands appeared only slightly delayed. The caudal portions of the Wolffian and Müllerian ducts were absent or interrupted in all α3−/−/α6−/− fetuses (compare
3964 De Arcangelis and others Fig. 9a, b and g with c, d and h; Fig. 9j). Moreover, in one fetus, the Wollfian duct was markedly dilated (W in Fig. 9i and j). These data suggest that α3 and/or α6 integrins are involved in the maintenance of the genital ducts during the ambisexual phase of their development. All α3−/−/α6−/− mutants also displayed partial or complete absence of the ureters (compare Fig. 9a with c and Fig. 9g with h; Fig. 9j). Thus, these integrins are also required for the maintenance of the ureters but dispensible for the formation of the ureteric buds: the near normality of all α3−/−/α6−/− kidneys indicates that during early embryogenesis the ureteric bud reached the renal blastema in order to trigger kidney tubule formation. The vesical portion of the primitive urogenital sinus and its derivative, the urinary bladder, were missing in all α3−/−/α6−/− fetuses (compare Fig. 9a and b with c and d; Fig. 9j) and the genital folds were never fused resulting in agenesis of the phallic portion of the urethra (compare Fig. 9e with f). In addition, three of the double homozygotes displayed an absence of caudal portion of the rectum (R, compare Fig. 9a and b with c and d). These three abnormalities may reflect a requirement of α3 and/or α6 integrins for the correct specification of the hindgut endoderm from which the bladder, urethra and rectum are normally derived. An abnormal origin, from the descending aorta, of the artery destined to the right forelimb (i.e retroesophageal right subclavian artery, rROS in Fig. 8f) was observed in two α3−/−/α6−/− fetuses. Note that as the origin of the normal right subclavian artery (rS in Fig. 8e) is localized at a more cranial level of the aorta, it is not in the plane of figure 8c. Retroesophageal subclavian arteries can be generated in chick embryos by ablation of the post-otic neural crest cell precursors (Kirby and Waldo, 1990). The occurrence of this abnormality in α3−/−/α6−/− fetuses therefore suggests the involvement of the corresponding integrins in the ontogenesis of neural crestderived arterial smooth muscle cells. Interestingly, of the three α3+/−/α6−/− fetuses analyzed histologically, one showed an absence of the rectum and another bilateral ectopic openings of the ureters into the Wollfian ducts. Thus, the removal of one allele of the integrin α3 gene from an α6 null genetic background occasionally recapitulates a subset of the abnormalities observed in the double null mutants. In contrast, morphological analysis of three α3−/−/α6+/− fetuses did not reveal any obvious defect. Brain and eye defects Previous studies have shown that α6 integrins are required for the proper organization of layers in the cortex and retina. In α6-null fetuses, disorganization of the cortical plate in the brain and of the retinal ganglion cell layer in the eye lead to ectopic neuroblastic outgrowths (Georges-Labouesse et al., 1998). Similarly, a recent analysis of brains of α3-null mice has shown that α3β1 is also required for laminar organization of the cerebral cortex (Anton et al., 1999). In four double α3−/−/α6−/− fetuses that we analyzed from E13.5 to E16.5, a more severe disorganization of the cortex was observed (Fig. 10a,b), possibly reflecting additive effects of the two mutations. In addition, the choroid plexus was abnormal (Fig. 10a,b). In the retina of α3−/−/α6−/− animals, a drastic disorganization of the retinal ganglion cell and nerve fiber layers, and massive ectopias into the vitreous body (Fig. 10c,d) were observed. In the eyes of double mutant embryos, a novel phenotype absent
from each of the single mutants was noticed (Fig. 10c,d): the anterior epithelium limiting the lens showed large breaches, which allowed the extrusion of lens cells and material. DISCUSSION Several studies have revealed essential functions of integrin receptors during developmental processes. By using gene knock-out strategies, it has been possible to define specific roles for some integrins at different stages of development. While inactivation of the β1 subunit leads to a very early embryonic lethality (Fässler and Meyer, 1995; Stephens et al., 1995), none of the α subunit knock-outs result in such an early phenotype, suggesting that more than one integrin receptor is involved in a given process. Our aim was to determine whether α3 and α6 integrins have overlapping functions during embryogenesis. We have produced and analyzed compound mutants for the integrin α3 and α6 chains and show that the simultaneous loss of these receptors leads to several new phenotypes, thus revealing synergistic activities for these integrins in the mouse embryo. Limb development requires α3 and α6 integrins All α3−/−/α6−/− mutant embryos have recognizable limb defects characterized by an abnormal shape, absence of digit separation and fusion of distal preskeletal elements. An impairment of the interdigital mesenchymal cell death is observed in mutant limbs, but is unlikely to be the primary cause of the limb defects, which are apparent before interdigital apoptosis starts. Another process which could involve α3 and α6 integrins is the formation of prechondrogenic masses and cartilage condensations. Integrin α3β1 is also a receptor for fibronectin and collagens, which are present within the limb bud mesenchyme (Critchlow and Hinchliffe, 1994). However, the predominant epithelial distribution of α3 and α6 chains in the early limb bud suggests that the defects originate from a dysfunction of the epithelium. Indeed, a rupture of the surface ectoderm followed by a leak of mesenchymal cells has been observed in the E13.5 mutant limbs. The rapid growth of the limb at this stage probably exerts a force on the epithelium, whose properties may be altered due to the absence of BL receptors. This may contribute to the abnormal shape and to fusion of digits in the mutant limbs, by provoking a deficit of mesenchyme. A similar phenomenon, epithelial rupture and digit fusion, was observed in embryos lacking the laminin α5 chain (Miner et al., 1998). A role for α3 and α6 integrins during AER formation Interestingly, limb bud defects in α3−/−/α6−/− embryos were already visible at early stages of development, before the epithelial rupture, thus raising the possibility that they originate from defects in the AER. By several approaches we showed that the AER does not differentiate properly in double mutant embryos. First, the typical protruding structure at the distal margin of the limb, although present, was barely detectable and flattened in all mutant limbs. Second, cells within the AER were abnormally shaped. They were smaller, less densely packed, and did not present the typical columnar morphology of AER cells. Furthermore, the mutant AER was broader than
α3 and α6 integrins compound mutant mice 3965 the wild-type AER even at early stages of development, which may explain why a higher number of cells was present in the mutant AER. The formation of the AER involves several steps which in the mouse embryo take place at E9.5-E10.5. The allocation of cells to this structure is thought to involve cell migration and compaction (Martin, 1998, and references therein). A final step to establish the definitive AER structure is the compaction of a broad domain and change of cell shape from flat to columnar. In contrast to the chick embryo where the AER is composed of one layer of pseudostratified epithelium covered by the periderm, the mouse AER is a true stratified epithelium (Martin, 1998). Based on the morphological differences between AERs of wild-type and mutant embryos, we propose that the steps that require α3 and α6 integrins are the compaction of a broad domain of presumptive AER cells to form the final structure, and the change of epithelial cell morphology from flat to columnar. One mechanism whereby integrins could play a role in these processes is by mediating anchorage to the BL. Indeed, interactions with the ECM are essential for epithelial morphogenesis. In contrast to α6-null and α3-null fetuses, where detachment of the epidermis occurs late in embryogenesis, no detachment of AER cells was observed in the double mutant suggesting that mechanisms other than attachment to the BL are involved. It should be noted that even in the areas where a normal BL is present, AER cells have an altered morphology, suggesting that the compound mutations affect the AER as a whole three-dimensional structure. Functional integrins are required for AER cell proliferation An additional and important result of this study is that α3 and α6 integrins are necessary for AER cell proliferation. As expected, AER defects also result in a moderate decrease in mesenchymal cell proliferation, which may account for limb abnormalities, as proposed Fig. 9. Defects in the urogenital system of α3−/−/α6−/− mutants. Serial transverse histological sections at comparable levels of E14.5 wild type (WT) and α3−/−/α6−/− (α3/α6) mutants. (a,c) Level of the vesical part of the urogenital sinus (B; missing in the mutant). (c,d) Level of the definitive urogenital sinus (UG). (e,f) Level of the genital tubercle (GT). (g-i) Transverse sections through the middle portion of the genital ducts at E14.5. Note that h and i correspond to the left and right sides of the same section, respectively. (j) Diagrammatic representations of the urogenital system. AO, descending aorta; A and rA, left and right umbilical arteries, respectively; B, vesical portion of the primitive urogenital sinus (E13.5 and E14.5) or urinary bladder (E15.5 and E16.5); CP, caudal-most portion of the peritoneal cavity. F, genital folds; GT, genital tubercle; K, kidney; M, Müllerian duct; ME, urogenital mesentery; N, neural tube; P, pelvic portion of the peritoneal cavity; R, rectum; U and rU, left and right ureters, respectively; UG, definitive urogenital sinus (E13.5 and E14.5) or pelvic urethra (E15.5 and E16.5) UT, urethra; W, Wolffian duct. The broken line marks the level of the sections in h and i. Magnifications: ×35.
for other mutations (Crackower et al., 1998; Jiang et al., 1998). There are several ways by which integrins could regulate AER cell proliferation. Numerous in vitro studies have demonstrated the involvement of integrins in signal transduction pathways
3966 De Arcangelis and others the fusion of digits may be due to an imbalance in the amount of FGFs, as evidenced by the alterations of the signals.
Fig. 10. Brain and eye defects in α3−/−/α6−/− fetuses. (a,b) Coronal sections through the cerebral cortex in E15.5 control (a) and double mutant (b) fetuses, showing a disorganized cortical plate, with numerous ectopias (asterisk), and an abnormal aspect of the choroid plexus. (c,d) sections of E14.5 control (c) and α3−/−/α6−/− (d) eyes. Note the severe disorganization of the retinal ganglion cell layer (arrow), and the abnormal lens (arrowhead). ch, choroid plexus; cp, cortical plate; le,lens vesicle; n, nerve fibers; v, vitreous body; vz, ventricular zone. Magnifications: (a,b) ×25, (c,d) ×50
which control cell proliferation such as the ras/MAPK pathway (Howe et al., 1998). In such cases, integrins are believed to communicate information coming from the extracellular environment to the inside of the cells. The absence of integrins could thus lead to a block in pathways critical for cycling of AER cells. Another explanation is that the dosage of FGFs in the double mutants is not sufficient to sustain AER cell proliferation. It is also possible that the defects are contributed by a disorganization of the BL, and the loss of its properties, in particular its ability to store and deliver growth factors. Indeed, FGFs have been shown to interact with the ECM (reviewed by Martin, 1998). As in other mutants (Zeller et al., 1989; Kuhlman and Niswander, 1997; Sidow et al., 1997; Crackower et al., 1998; Jiang et al., 1998) the growth of double mutant limbs indicates that the AER is partially functional at initial stages. Consistent with this is the expression of FGF-8 in the AER and of Msx1, an AER-dependent gene, in the mesenchymal compartment. Similarly, FGF-8 was found to be expressed, with some alterations, in ld mutants which also present AER defects. In contrast to ld mutant limbs, α3−/−/α6−/− mutants still express FGF-4, which may relate to the fact that α3−/−/α6−/− limb defects are less severe than those observed in ld mutant limbs (Zeller et al., 1989; Haramis et al., 1995; Kuhlman and Niswander, 1997). As in the mutant mice already mentioned,
Multiple roles of integrins in epithelia In addition to the AER defects, other novel phenotypes were observed in the compound mutants, which further illustrate the synergism of α6 and α3 integrins during mouse development. One striking abnormality was the bilateral hypoplasia of the lungs. Lung defects are present in α3-null fetuses (Kreidberg et al., 1996), however, at much later stages of development than those observed in the double mutant. The onset of the lung phenotype in compound mutants seems to be at the very first steps of lung development when two lung buds start to separate, since in a few embryos, the left lung bud is absent at early (E11.5) stages. An early phenotype is consistent with the fact that both α3 and α6 integrin chains are present in the bud epithelial cells at these stages (Wu and Santoro, 1996, and our own results). This confirms that cell-matrix interactions mediated by integrins guide bud growth during lung development (Schuger et al., 1991; Mollard and Dziadek, 1998). However, the lung defect is not clearly related to a disorganization of the BL, since in mutant lungs staining patterns for laminin isoforms, collagen and HSPG were not obviously modified (data not shown). While there is a marked reduction in lung size, branching morphogenesis is not abolished, as attested by the presence of secondary buds. In the pancreas too, branching morphogenesis was decreased but not abolished. Integrins α3 and α6 are expressed throughout kidney development in epithelial cells adjacent to laminin-containing BL. Studies using function-blocking antibodies in organ culture systems have suggested a role for α6β1 in kidney tubulogenesis (Sorokin et al., 1990). The near-normality of the kidneys in the compound α3−/−/α6−/− embryos at early stages of embryonic development indicates that other BL receptors are sufficient for kidney morphogenesis to occur, such as integrin α2β1 or α-dystroglycan, for which a role in the initial steps of BL organization has been recently proposed (Henry and Campbell, 1998). Integrins and laminins Besides revealing new roles for α3 and α6 integrins in embryonic development, our analysis has also revealed their potential binding partners in vivo. Past results of knock-out studies have shown that α3β1 is probably a receptor for laminin-5 (α3β3γ2), and that α6 integrins are receptors of laminin-1 (α1β1γ1) and laminin-5, in vivo. The presence of limb and neural tube defects in α3−/−/α6−/− compound mutants and in laminin α5 chain mutants (Miner et al., 1998), and the absence of such abnormalities in integrin α3-null or α6-null mutants suggest that both α3 and α6 integrins binding to α5containing laminins (laminins 10 and 11, Miner et al., 1997) are essential in vivo in some tissues. A recent study has suggested that integrin α3β1 in lung carcinoma cells binds α5containing laminins in vitro (Kikkawa et al., 1998). Alternatively, the appearance of the phenotype in compound mutants may be explained by the fact that several integrin/laminin interactions are disrupted. Conclusion This study has identified novel roles for integrins in multiple
α3 and α6 integrins compound mutant mice 3967 developmental processes. New phenotypes were uncovered in compound mutants, reflecting the fact that several receptors must be active in common pathways. The requirement of integrin receptors for proper AER organization and cell proliferation illustrates that the regulation of tissue morphogenesis and cell proliferation are integrated processes, and suggests a link with signalling pathways. It should be noted that an association of defects in the respiratory, cardiovascular and urogenital systems has also been observed in retinoic acid receptor compound mutants (Ghyselinck et al., 1997, and references therein). Similarities revealed by genetic approaches may help to understand the relationship between integrin pathways and other signalling pathways. We thank M. Choné and R. Lorentz for excellent technical assistance and M. Digelmann and P. Goetz-Reiner for histological analysis. We are very grateful to A. Gansmüller and N. Messaddeq for transmission and scanning electron microscopy. We also thank M. DiPersio, R. Hynes, S. Johansson, G. Martin and B. Robert for gift of probes and antibodies, and P. Dollé, M. Labouesse and R. Mollard for helpful discussions. This work was supported by institutional funds from CNRS, INSERM, HUS, and grants to E. G.-L. from ARC, AFM, and the CNRS program ‘Biologie Cellulaire’. A. D. is a recipient of a fellowship from the Société de Secours des Amis des Sciences.
REFERENCES Anton, E. S., Kreidberg, J. A. and Rakic, P. (1999). Distinct functions of alpha3 and alpha(v) integrin receptors in neuronal migration and laminar organization of the cerebral cortex. Neuron. 22, 277-289. Ashkenas, J., Muschler, J. and Bissell, M. J. (1996). The extracellular matrix in epithelial biology: shared molecules and common themes in distant phyla. Dev.Biol. 180, 433-444. Carter, W. G., Ryan, M. C. and Gahr, P. J. (1991). Epiligrin, a new cell adhesion ligand for integrin alpha 3 beta 1 in epithelial basement membranes. Cell 65, 599-610. Crackower, M. A., Motoyama, J. and Tsui, L. C. (1998). Defect in the maintenance of the apical ectodermal ridge in the Dactylaplasia mouse. Dev. Biol. 201, 78-89. Critchlow, M. A. and Hinchliffe, J. R. (1994). Immunolocalization of basement membrane components and beta 1 integrin in the chick wing bud identifies specialized properties of the apical ectodermal ridge. Dev. Biol. 163, 253-269. Crossley, P. H. and Martin, G. R. (1995). The mouse Fgf8 gene encodes a family of polypeptides and is expressed in regions that direct outgrowth and patterning in the developing embryo. Development 121, 439-451. De Arcangelis, A., Neuville, P., Boukamel, R., Lefebvre, O., Kedinger, M. and Simon-Assmann, P. (1996). Inhibition of laminin alpha 1-chain expression leads to alteration of basement membrane assembly and cell differentiation. J. Cell Biol. 133, 417-430. Décimo, D., Georges-Labouesse, E, and Dollé, P. (1995). In situ hybridization to cellular RNA. In Gene Probes 2: A Practical Approach, (ed. B. D. Hames and S. J. Higgins), p183-210. Oxford: IRL Press. Delwel, G. O. and Sonnenberg, A. (1996) Laminin isoforms and their integrin receptors. in Adhesion Receptors as Therapeutic Target (ed. Michael Horton), pp. 9-36. CRC Press. DiPersio, C. M., Hodivala-Dilke, K. M., Jaenisch, R., Kreidberg, J. A. and Hynes, R. O. (1997). alpha3beta1 integrin is required for normal development of the epidermal basement membrane. J. Cell Biol. 137, 729742. DiPersio, C. M., Shah, S. and Hynes, R. O. (1995). alpha 3A beta 1 integrin localizes to focal contacts in response to diverse extracellular matrix proteins. J. Cell Sci. 108, 2321-2336. Fässler, R. and Meyer, M. (1995). Consequences of lack of beta 1 integrin gene expression in mice. Genes. Dev. 9, 1896-1908. Fässler, R., Georges-Labouesse, E. and Hirsch, E. (1996). Genetic analyses of integrin function in mice. Curr. Opin. Cell Biol. 8, 641-646. Georges-Labouesse, E., Mark, M., Messaddeq, N. and Gansmuller, A.
(1998). Essential role of alpha 6 integrins in cortical and retinal lamination. Curr. Biol. 8, 983-986. Georges-Labouesse, E., Messaddeq, N., Yehia, G., Cadalbert, L., Dierich, A. and Le Meur, M. (1996). Absence of integrin alpha 6 leads to epidermolysis bullosa and neonatal death in mice. Nat. Genet. 13, 370-373. Ghyselinck, N. B., Dupe, V., Dierich, A., Messaddeq, N., Garnier, J. M., Rochette-Egly, C., Chambon, P. and Mark, M. (1997). Role of the retinoic acid receptor beta (RARbeta) during mouse development. Int. J. Dev. Biol. 41, 425-447. Goh, K. L., Yang, J. T. and Hynes, R. O. (1997). Mesodermal defects and cranial neural crest apoptosis in alpha5 integrin-null embryos. Development 124, 4309-4319. Haramis, A. G., Brown, J. M. and Zeller, R. (1995). The limb deformity mutation disrupts the SHH/FGF-4 feedback loop and regulation of 5′ HoxD genes during limb pattern formation. Development 121, 4237-4245. Hebert, J. M., Basilico, C., Goldfarb, M., Haub, O. and Martin, G. R. (1990). Isolation of cDNAs encoding four mouse FGF family members and characterization of their expression patterns during embryogenesis. Dev. Biol. 138, 454-463. Henry, M. D. and Campbell, K. P. (1998). A role for dystroglycan in basement membrane assembly. Cell 95, 859-870. Howe, A., Aplin, A. E., Alahari, S. K. and Juliano, R. L. (1998). Integrin signaling and cell growth control. Curr. Opin. Cell Biol. 10, 220-231. Hynes, R. O. (1992). Integrins: versatility, modulation, and signaling in cell adhesion. Cell 69, 11-25. Hynes, R. O. (1996). Targeted mutations in cell adhesion genes: what have we learned from them? Dev. Biol. 180, 402-412. Jegalian, B. G. and De Robertis, E. M. (1992). Homeotic transformations in the mouse induced by overexpression of a human Hox3.3 transgene. Cell 71, 901-910. Jiang, R., Lan, Y., Chapman, H. D., Shawber, C., Norton, C. R., Serreze, D. V., Weinmaster, G. and Gridley, T. (1998). Defects in limb, craniofacial, and thymic development in Jagged2 mutant mice. Genes Dev. 12, 1046-1057. Kikkawa, Y., Sanzen, N. and Sekiguchi, K. (1998). Isolation and characterization of laminin-10/11 secreted by human lung carcinoma cells. Laminin-10/11 mediates cell adhesion through integrin alpha3 beta1. J. Biol. Chem. 273, 15854-15859. Kirby, M. L. and Waldo, K. L. (1990). Role of neural crest in congenital heart disease. Circulation 82, 332-340. Kreidberg, J. A., Donovan, M. J., Goldstein, S. L., Rennke, H., Shepherd, K., Jones, R. C. and Jaenisch, R. (1996). Alpha 3 beta 1 integrin has a crucial role in kidney and lung organogenesis. Development 122, 35373547. Kuhlman, J. and Niswander, L. (1997). Limb deformity proteins: role in mesodermal induction of the apical ectodermal ridge. Development 124, 133-139. Martin, G. R. (1998). The roles of FGFs in the early development of vertebrate limbs. Genes Dev. 12, 1571-1586. Mercurio, A. M. and Shaw, L. M. (1991). Laminin binding proteins. BioEssays 13, 469-473. Miner, J. H., Cunningham, J. and Sanes, J. R. (1998). Roles for laminin in embryogenesis: exencephaly, syndactyly, and placentopathy in mice lacking the laminin alpha5 chain. J. Cell Biol. 143, 1713-1723. Miner, J. H., Patton, B. L., Lentz, S. I., Gilbert, D. J., Snider, W. D., Jenkins, N. A., Copeland, N. G. and Sanes, J. R. (1997). The laminin alpha chains: expression, developmental transitions, and chromosomal locations of alpha1-5, identification of heterotrimeric laminins 8-11, and cloning of a novel alpha3 isoform. J. Cell Biol. 137, 685-701. Mollard, R. and Dziadek, M. (1998). A correlation between epithelial proliferation rates, basement membrane component localization patterns, and morphogenetic potential in the embryonic mouse lung. Am. J. Respir. Cell Mol. Biol. 19, 71-82. Robert, B., Lyons, G., Simandl, B. K., Kuroiwa, A. and Buckingham, M. (1991). The apical ectodermal ridge regulates Hox-7 and Hox-8 gene expression in developing chick limb buds. Genes Dev. 5, 2363-2374. Schuger, L., Skubitz, A. P., O’Shea, K. S., Chang, J. F. and Varani, J. (1991). Identification of laminin domains involved in branching morphogenesis: effects of anti-laminin monoclonal antibodies on mouse embryonic lung development. Dev. Biol. 146, 531-541. Sidow, A., Bulotsky, M. S., Kerrebrock, A. W., Bronson, R. T., Daly, M. J., Reeve, M. P., Hawkins, T. L., Birren, B. W., Jaenisch, R. and Lander, E. S. (1997). Serrate2 is disrupted in the mouse limb-development mutant syndactylism. Nature 389, 722-725. Sorokin, L., Sonnenberg, A., Aumailley, M., Timpl, R. and Ekblom, P.
3968 De Arcangelis and others (1990). Recognition of the laminin E8 cell-binding site by an integrin possessing the alpha 6 subunit is essential for epithelial polarization in developing kidney tubules. J. Cell Biol. 111, 1265-1273. Sorokin, L. M., Conzelmann, S., Ekblom, P., Battaglia, C., Aumailley, M. and Timpl, R. (1992). Monoclonal antibodies against laminin A chain fragment E3 and their effects on binding to cells and proteoglycan and on kidney development. Exp. Cell Res. 201, 137-144. Sorokin, L. M., Pausch, F., Frieser, M., Kroger, S., Ohage, E. and Deutzmann, R. (1997). Developmental regulation of the laminin alpha5 chain suggests a role in epithelial and endothelial cell maturation. Dev. Biol. 189, 285-300. Stephens, L. E., Sutherland, A. E., Klimanskaya, I. V., Andrieux, A., Meneses, J., Pedersen, R. A. and Damsky, C. H. (1995). Deletion of beta
1 integrins in mice results in inner cell mass failure and peri-implantation lethality. Genes Dev. 9, 1883-1895. Sutherland, A. E., Calarco, P. G. and Damsky, C. H. (1993). Developmental regulation of integrin expression at the time of implantation in the mouse embryo. Development 119, 1175-1186. Wu, J. E. and Santoro, S. A. (1996). Differential expression of integrin alpha subunits supports distinct roles during lung branching morphogenesis. Dev. Dyn. 206, 169-181. Yang, J. T., Rayburn, H. and Hynes, R. O. (1993). Embryonic mesodermal defects in alpha 5 integrin-deficient mice. Development 119, 1093-1105. Zeller, R., Jackson-Grusby, L. and Leder, P. (1989). The limb deformity gene is required for apical ectodermal ridge differentiation and anteroposterior limb pattern formation. Genes Dev. 3, 1481-1492.
