The Dlx5 gene encodes a Distal-less-related DNA-binding homeobox protein first expressed during early embryonic development in anterior regions of the mouse embryo. In later developmental stages, it appears in the branchial arches, the otic and olfactory placodes and their derivatives, in restricted brain regions, in all extending appendages and in all developing bones. We have created a null allele of the mouse Dlx5 gene by replacing exons I and II with the E. coli lacZ gene. Heterozygous mice appear normal. β-galactosidase activity in Dlx5+/ embryos and newborn animals reproduces the known pattern of expression of the gene. Homozygous mutants die shortly after birth with a swollen abdomen. They present a complex phenotype characterised by craniofacial abnormalities affecting derivatives of the first four branchial arches, severe malformations of the vestibular organ, a delayed ossification of the roof of the skull and abnormal osteogenesis. No obvious defect was observed in the patterning of limbs and other appendages. The defects observed in Dlx5/ mutant animals suggest multiple and independent roles of this gene in the patterning of the branchial arches, in the morphogenesis of the vestibular organ and in osteoblast differentiation.

The Distal-less (Dll) gene of Drosophila is required for correct morphogenesis of the distal portion of the legs, antennae and mouth parts (Cohen et al., 1989; O’Hara et al., 1993). Several homologues of Dll (called Dlx in the mouse) have been isolated from vertebrate species ranging from zebrafish to human (Stock et al., 1996). These genes constitute a highly conserved family of homeobox genes, which are thought to act as transcription factors; however, their mode of action as regulatory molecules might be more complex as it has been shown that members of the Dlx family can form dimeric complexes with Msx homeoproteins mutually affecting their DNA-binding properties (Zhang et al., 1997). In the mouse, there are at least six Dlx genes arranged as pairs facing each other through the 3′ end and located near Hox clusters (Dlx1 and Dlx2 near HoxD; Dlx3 and Dlx7 near HoxB; Dlx5 and Dlx6 near HoxC) (Simeone et al., 1994a,b; McGuinness et al., 1996; Nakamura et al., 1996; Liu et al., 1997).

Dlx genes are all expressed in spatially and temporally restricted patterns in craniofacial primordia, basal telencephalon and diencephalon, and in distal regions of extending appendages including the limb and the genital tubercle. The pattern of expression of Dlx5 differs from that of the other members of the family in two respects. First, it has been shown recently (Yang et al., 1998) that Dlx5 is expressed much earlier than other Dlx genes during development in territories that define the rostral and lateral border of the neural plate when these regions have organizing activities that pattern the adjacent rostral prosencephalon. Second, Dlx5 and Dlx6 have the unique property to be expressed in all developing bones from the time of initial cartilage formation onward (Simeone et al., 1994a,b; Zhao et al., 1994). A further indication of the possible importance of Dlx5 in the control of bone differentiation comes from a recent study (Ryoo et al., 1997) in which it has been shown, that this gene is expressed at specific stages of osteoblast differentiation and could repress osteocalcin gene expression by interacting with a single homeodomain-binding site in its promoter.

In situ hybridization studies on 9.5 and 10.5 dpc mice embryos have shown that Dlx1 and Dlx2 are expressed in the mesenchyme of both the proximal and the distal domain of the first two branchial arches while Dlx3, Dlx5 and Dlx6 appear to be expressed predominantly in the distal arch mesenchyme (Simeone et al., 1994a,b; Qiu et al., 1997). A mediolateral patterning of Dlx genes within the branchial arches has also been suggested (Robinson and Mahon, 1994).

So far only the functions of Dlx1 and Dlx2 have been analyzed by targeted inactivation in the mouse. Mice homozygous for Dlx1 or Dlx2 deletion shows that both genes are essential for correct craniofacial development in particular for derivatives of the proximal regions of the branchial arches (Qiu et al., 1995, 1997). These observations lead to the hypothesis that Dlx genes contribute to the readout of a proximodistal pattern within the branchial arches. Furthermore, as mice lacking both Dlx1 and Dlx2 have unique abnormalities, such as the absence of maxillary molars, a partial functional redundancy of Dlx genes has been suggested (Qui et al., 1997). Similarly, as Dlx1 and Dlx2 are expressed together with Dlx3, Dlx5 and Dlx6 in the distal part of the first and second branchial arch, the concept of functional redundancy has been used by the same authors to explain the lack of defects affecting distal arch derivatives in Dlx1 and Dlx2 mutant mice. The functional inactivation of other members of the Dlx family is critical to clarify these points.

We set out to inactivate the Dlx5 gene in the mouse by homologous recombination. The DNA construct for gene targeting was engineered so to associate the disruption of Dlx5 with a positional insertion of the E. coli β-galactosidase reporter gene, to provide a marker for Dlx5-expressing cell lineages in developing structures.

Targeting vector

A 400 bp HaeIII genomic fragment comprising exon III of murine Dlx5 was used to screen a mouse (strain 129) genomic library cloned in the HindIII site of pBelo BAC-11 vector (Genome System, St Louis, MO USA). Of the four positive clones identified, one was further analyzed and found to contain the complete Dlx5 and Dlx6 genes. The Dlx5 targeting DNA construct, designated pGN-2lox X5, was prepared as follows. A 4.5 kb PstI-SmaI fragment spanning the promoter region of the Dlx5 gene and the entire 5′ untranslated sequence was cloned upstream of the E. coli lacZ coding sequence. The 3′ end of this fragment corresponds to the position −6 relative to the ATG Start codon of the Dlx5 open reading frame (ORF). Downstream of the lacZ reporter, we introduced a RSV-neoR expression cassette flanked by two Lox P sites. In the presence of Cre recombinase activity, this construction allows for removal of the RSV-neoR expression cassette. This control was devised for ruling out a possible interference of RSV-neoR with the expression of other genes. The downstream region of homology was cloned 3′ to the second Lox P site and consisted of a 1 kb fragment spanning part of the third exon. Homologous recombination of this construct results in the deletion of exons I and II of Dlx5, including the Start Codon and the positional insertion of the lacZ ORF. Thus, Dlx5 protein synthesis should be abolished and replaced by the synthesis of bacterial β-galactosidase.

ES cell culture and generation of chimeric mice

For electroporation, 20 μg of pGN-2lox X5 KOB DNA linearized at the unique KpnI site was added to 1.2×107 HM-1 cells at passage 18 in a volume of 0.8 ml of PBS containing 0.1% β-mercaptoethanol. Electroporation was performed in a 0.4 cm width cuvette at 200 V and using a capacitance of 1050 μF. After electroporation, the cells were transferred to non-selective medium supplemented with 1000 U/ml Leukaemia Inhibitory Factor (LIF, Gibco-BRL). G418 was added to concentrations varying from 350 to 450 μg/ml after 24 hours of culture and the selection was continued for 10 days. G418-resistant ES cell clones were picked and transferred into 1.5 cm diameter wells and subsequently expanded in 35 mm diameter Petri dishes for freezing and for DNA analysis. Chimeric mice were generated from recombinant ES cells as described (Robertson et al., 1986).

DNA analysis

DNA was extracted from ES cell clones, embryos or mouse tails using standard methods. Polymerase chain reaction (PCR) amplification was used for a primary screening of homologous recombinants and to determine the genotype of embryos and newborn animals. PCR reactions were carried out under standard conditions using SuperTherm thermostable polymerase (Eppendorf). The reaction profile included a denaturation step at 94°C 5 minutes, 30 cycles of 94°C 1 minute, 65°C 1 minute, 72°C 1 minute and a final elongation step at 72°C 5 minutes. The upstream and the downstream primers were chosen, respectively, in the neoR sequence (primer 4: TGCTGTGTTCCAGAAGTGTT) and immediately 3′ to the downstream recombination boundary, i.e. external to the targeted sequence (primer 3: GCCCATCTAATAAAGCGTCCCGG). A 1,100 bp fragment is expected from a correct homologous recombination event. To confirm correct targeting, Southern blotting was also performed on BamHI-digested DNA using a 350 bp EcoRI fragment, external to the homology regions, as probe. For the typing of mice offspring, a PCR reaction was utilized that included four oligonucleotides: primers 3 and 4 for the mutated allele, and primers 1 (GACAGGAGTGTTTGACAGAAGAGTCCC) and 2 (GTAGTCG-GCATAAGCCTTGGC) for the wild-type allele. The latter set of primers yields a 280 bp fragment.

The presence of Dlx5 mRNA was determined on total RNA samples prepared from genotyped individual 12.5 dpc embryos generated by crossing heterozygous partners. The total mRNA was prepared using Trizol Reagent (Gibco-BRL). 1 μg of RNA sample from two of each +/+, +/− or −/− embryos was subjected to RT-PCR amplification of the complete murine Dlx5 cDNA with the Titan One Tube RT-PCR kit (Boehringher) according to the manufacture’s instructions. A set of oligonucleotides was designed to flank the ORF (sense: TCATGACAGGAGTGTTTGACAG; antisense: GGGCTA-AACCAGCACAACACTGTAG) and to yield a 820 bp fragment. The identity of the fragment was confirmed by complete sequencing. β-actin cDNA was amplified from the same samples and under the same conditions to control for the quality of RNA preparation.

Mice carrying the Cre recombinase under the control of the CMV promoter were obtained from Dr P. Chambon (Strasbourg, France) and were typed using published primer sequences. Mice obtained from the breeding of CMV-Cre animals with Dlx5/lacZ heterozygous were typed for Dlx5 by PCR replacing primer 4 with primer 5 (TCCCTCGAAAAGGTTCACTA). Correct Cre-mediated removal of the PGK-neoR cassette was confirmed by PCR replacing primer 4 with primer 6 (TACAAATAAAGCAATAGCATC), and later by Southern blotting of PstI-digested genomic DNA.

β-gal staining and histological analyses

Postimplantation embryos were collected at the desired stages, considering the day of the plug as 0.5 dpc. For lacZ expression analysis, 8.5, 9.5 and 10.5 dpc embryos were fixed for 5-10 minutes in 2% paraformaldehyde (PFA) in PBS, while 12.5 dpc and later stages embryos were fixed for 10 minutes in 2% PFA + 0.2% glutaraldehyde. X-gal staining was performed as described (Sham et al., 1993). Newborn and adult animals were stained by perfusion. Stained specimens were photographed in toto, then either embedded in paraffin and used to obtain 10 μm serial sections, or dehydrated and clarified in a benzyl-benzoate/benzyl alcohol 2:1 mixture to reveal the staining of internal structures. For conventional histology, sections were counterstained with eosin-haematoxylin.

Immunohistochemistry for the detection of osteocalcin in tissue sections was performed on paraffin sections of newborn mice previously stained with X-gal by perfusion using anti-mouse-osteocalcin polyclonal antibodies (a gift of Dr Caren Gundberg, Yale University, Department of Orthopedics).

Alcian blue/Alizarin red staining of newborn skeletons was performed as described (Wallin et al., 1994).

Probes and in situ hybridization

In situ hybridization experiments were performed as described (Simeone, 1998). The Shh, Tbrain1 and Gbx2 probes were PCR products spanning the regions between amino acids 109 and 177 (Echelard et al., 1993), amino acids 599 and 680 (Bulfone et al., 1995) and amino acids 6 and 294 (Chapman et al., 1997), respectively. The Dlx1, Dlx6 and Otp probes have been previously described (Simeone et al., 1994a,b). The Ihh probe corresponded to the 600 bp 3′end of murine Ihh cDNA. Cbfa1/Osf2 probe corresponded to 630 bp in the 3′end of the mouse ORF, not including the runt domain. Both Ihh and Cbfa1 probes were obtained by PCR.

BrdU labeling and detection of apoptotic cells

Pregnant mice at E10.5 and E11.5 were injected intraperitoneally with BrdU solution (50 mg/kg body weight) and killed after 1 hour. After embryo genotyping, BrdU detection was performed according to Xuan et al. (1995). Three embryos for each genotype were scored. Four comparable sections for each embryo were analyzed. The fraction of BrdU-positive cells was determined by dividing the number of BrdU-positive nuclei by the total number of nuclei identified in units of tissue corresponding to a square of 100 μm/side (Xuan et al., 1995). The proportion of BrdU-positive cells in wild-type embryos was considered 100%.

To detect apoptotic cells, the sections were processed by the TUNEL method as described (Gavrieli et al., 1992).

Targeted disruption of the Dlx5 gene in ES cells and mice

The Dlx5 gene was disrupted in the plasmid pGN-2lox-X5 by inserting the E. coli lacZ coding sequence and a neoR expression cassette in replacement of the Dlx5 exon I and II (Fig. 1A).

Fig. 1.

Disruption of the mouse Dlx5 gene by homologous recombination. (A) Exon-intron organization and restriction map of murine Dlx5 gene, targeting vector and mutated allele following homologous recombination. Exons are indicated in roman numbers (I, II and III), oligonucleotides are indicated by solid arrowheads in Arabic numbers (1-6). The probe used for Southern blot hybridization is indicated by a thick line. Open boxes, protein coding sequences; striped boxes, non-coding exon sequences; open triangles, Lox P sequences. Restriction sites are: B, BamHI; E, EcoRI; H, HindIII; K, KpnI; P, PstI; S, SmaI. (B) Southern blot analysis of BamHI-digested genomic DNA from wild-type (+/+) and recombinant (+/−) ES cells (lanes on the left) and mice (lanes on the right). The wild-type and mutant alleles yielded, respectively, a 6.2 and a 4.2 kb fragment, as expected (arrows). (C) Allele-specific PCR on genomic DNA from +/+, +/− and −/− mice. The wild-type (primers 1-2) and the mutant (primers 3-4) alleles are amplified and yield a 280 and 1,100 bp fragment, respectively (arrows). (D) RT-PCR analysis of Dlx5 (top) and β-actin (bottom) expression in +/+, +/− and −/− total RNA samples from 12.5 dpc embryos. The PCR product of the expected sizes were obtained (indicated by arrows). No expression of Dlx5 was seen in homozygous mutant embryos. (E) Deletion of the PGK-neoR cassette from the Dlx5-lacZ allele, to generate the Dlx5-lacZCRE allele using the Cre recombinase/Lox P system in vivo. PCR amplification of genomic DNA from two individual mice (#30 and #32) for the mutated Dlx5 allele (left), the CRE transgene (middle) and the modified mutated Dlx5 allele (right panel), using primers 3-5, CRE-amplimer and 3-6, respectively (A). The expected fragments were always obtained (arrows). The 3-6 fragment was detected only in +/− animals in the presence of the CRE transgene.

Fig. 1.

Disruption of the mouse Dlx5 gene by homologous recombination. (A) Exon-intron organization and restriction map of murine Dlx5 gene, targeting vector and mutated allele following homologous recombination. Exons are indicated in roman numbers (I, II and III), oligonucleotides are indicated by solid arrowheads in Arabic numbers (1-6). The probe used for Southern blot hybridization is indicated by a thick line. Open boxes, protein coding sequences; striped boxes, non-coding exon sequences; open triangles, Lox P sequences. Restriction sites are: B, BamHI; E, EcoRI; H, HindIII; K, KpnI; P, PstI; S, SmaI. (B) Southern blot analysis of BamHI-digested genomic DNA from wild-type (+/+) and recombinant (+/−) ES cells (lanes on the left) and mice (lanes on the right). The wild-type and mutant alleles yielded, respectively, a 6.2 and a 4.2 kb fragment, as expected (arrows). (C) Allele-specific PCR on genomic DNA from +/+, +/− and −/− mice. The wild-type (primers 1-2) and the mutant (primers 3-4) alleles are amplified and yield a 280 and 1,100 bp fragment, respectively (arrows). (D) RT-PCR analysis of Dlx5 (top) and β-actin (bottom) expression in +/+, +/− and −/− total RNA samples from 12.5 dpc embryos. The PCR product of the expected sizes were obtained (indicated by arrows). No expression of Dlx5 was seen in homozygous mutant embryos. (E) Deletion of the PGK-neoR cassette from the Dlx5-lacZ allele, to generate the Dlx5-lacZCRE allele using the Cre recombinase/Lox P system in vivo. PCR amplification of genomic DNA from two individual mice (#30 and #32) for the mutated Dlx5 allele (left), the CRE transgene (middle) and the modified mutated Dlx5 allele (right panel), using primers 3-5, CRE-amplimer and 3-6, respectively (A). The expected fragments were always obtained (arrows). The 3-6 fragment was detected only in +/− animals in the presence of the CRE transgene.

Heterozygous disruption of the Dlx5 gene in ES cells was obtained by homologous recombination after electroporation with pGN-2lox X5. Recombination events at the Dlx5 locus were identified among G418-resistant colonies initially by PCR amplification using primers 3 and 4 (Fig. 1A), and subsequently confirmed by Southern blot analysis of BamHI-digested genomic DNA hybridized with the indicated probe (Fig. 1A). The expected 6.2 and 4.2 kb fragments, corresponding respectively to the wild-type and the mutated allele, were observed in the DNA from recombinant cells (Fig. 1B). Among 300 G418-resistant clones analyzed, 7 were found heterozygous for the Dlx5/lacZ allele. Southern blot analysis of DNA from mice generated with these ES cells showed the same restriction fragments (Fig. 1B) and no subsequent modification was observed.

One ES cell clone corresponding to a correct replacement event led to germline transmission after injection into C57Bl/6 blastocysts. The chimeras obtained were used to establish a family of heterozygous carriers by crossing with C57Bl/6 × DBA/2 F1 females. The presence of the wild type or the Dlx5/lacZ alleles was determined by PCR amplification of genomic DNA using the oligonucleotides reported in Fig. 1A: primers 1 and 2 for the wild-type allele, and 3 and 4 for the mutated allele (Fig. 1C).

The Dlx5/lacZ mutant allele is expected to be unable to transcribe Dlx5 exon I and II sequences, leading to the absence of full-length Dlx5 mRNA. To confirm this, we checked for transcription of Dlx5 mRNA in wild-type, heterozygous and homozygous mutant embryos by RT-PCR, using primers flanking the Dlx5 ORF. The results show absence of full-length Dlx5 mRNA in samples from homozygous embryos, while transcripts were detected in +/+ and +/-embryos at comparable levels (Fig. 1D). The identity of the amplified cDNA was determined by sequencing. Amplification of murine β-actin mRNA was used to control for the quality of the RNA preparations (Fig. 1D). Exon III sequences, although present, should not be transcribed due to the presence of a polyadenylation-addition signal downstream from the lacZ sequence and upstream of the neoR cassette (Fig. 1A).

Cre-mediated removal of the PGK-neoR cassette was obtained by crossing Dlx5/lacZ heterozygous animals with CMV-Cre transgenic partners. The offspring were genotyped for the presence of the Cre transgene, the presence of the Dlx5/lacZ allele and the correct modification of the latter to yield the Dlx5/lacZCre allele, by PCR (Fig. 1E) and Southern blotting. The analysis was carried out using the primers indicated in Fig. 1A: 3 and 5 for Dlx5/lacZ, and 3 and 6 for the Dlx5/lacZCre allele. Correct removal of the sequence flanked by Lox P sites was observed in every animal with the genotype Cre+, Dlx5/lacZ+/−, as demonstrated by PCR analysis (Fig. 1E). Southern blot analysis of PstI-digested genomic DNAs from mice carrying the Dlx5/lacZCre allele confirmed correct Cre-Lox P-mediated deletion (data not shown).

Heterozygous mutant mice are viable, fertile and do not exhibit any obvious abnormality. Genotype analysis of over 500 individual mice representing the F1 of Dlx5/lacZ heterozygous parents show genotype frequencies not significantly different from those expected by Mendelian segregation of the mutated allele (+/+ 29%; +/− 49%; −/− 22%) indicating that lack of Dlx5 does not result in embryonic lethality. In contrast, homozygous Dlx5/lacZ animals are easily recognized at birth, in that they show decreased motility, gasping respiration, do not suckle, do not have milk in their stomach and develop a bloated abdomen accumulating air in their stomach and intestine. No mutant survived longer than 24 hours after birth. Exencephaly was observed in about 12% homozygous mutant embryos and newborns, in addition to all other phenotypes. We shall refer to these animals as ‘exencephalic phenotype’, to distinguish them from the majority of mutant embryos and newborn.

To our surprise, mice carrying one copy of the Dlx5/lacZCre allele, derived from crossings between Dlx5/lacZ heterozygous animals (either males or females) and CMV-Cre transgenic partners, die 18 days after birth without presenting any obvious defect or skeletal deformity. As both heterozygous parents are normal and fertile and no defect is seen in the expression pattern of either Dlx5 (Fig. 2J,K) or Dlx6 (data not shown) in Dlx5/lacZCre heterozygous embryos, their death remains unexplained. As Dlx5/lacZCre heterozygous animals did not reach sexual maturity, we could not analyze the phenotype of Dlx5/lacZCre homozygous animals.

Fig. 2.

Expression of Dlx5/lacZ during embryogenesis, and phenotypes observed in homozygous mutants. (A,B) Whole-mount X-gal staining of 8.0 dpc embryos in lateral (A) and dorsal (B) view. (C,D) 8.5 dpc-stained heterozygous (C) and homozygous mutant (D) embryos. In D, the first branchial arch (1) is indicated. Arrowheads in A-D indicate the otic placode. (E,F) 9.5 dpc embryos stained with X-gal showing expression of lacZ in the otic vesicle, branchial arches (indicated by numbers 1-4, G), olfactory placode and limb buds. An heterozygous (E) and an homozygous mutant (F) embryo are shown. (G) Detail of the branchial arches (1-4) to show expression of Dlx5/lacZ. In the second and third arch staining is observed in an anterior and a posterior streak of cells. (H,I) Expression of lacZ in normal (H) and homozygous mutant (I) 11.5 dpc embryos. (J-M) Expression of lacZ in 12.5 dpc embryos. Expression in Cre+ (J) and Cre (K) heterozygous embryos is compared. Clarified heterozygous (L) and homozygous mutant (M) embryos are shown. The arrow in H indicates the genital bud. (N-S) Expression pattern of lacZ in the developing limbs of 9.5 (N), 10.5 (O,P), 11.5 (Q), 13.5 (R) and 14.5 (S) dpc heterozygous embryos, in lateral view, except in P where a frontal view is shown to highlight the apical ectodermal ridge (AER). Limbs of homozygous mutant mice showed no difference in the β-gal-staining pattern at any stage of development. Arrowheads in N indicate the AER. White asterisks indicate a region of expression located at the anterior margin of the limb bud. At 13.5 (R) and 14.5 (S) dpc, expression in limb skeletal elements is also observed. For each pair of matched embryos, the same magnification is used. Abbreviations: de, diencephalon; sc, semicircular canals; tb, tail bud; te, telencephalon; vce, ventral cephalic epithelium.

Fig. 2.

Expression of Dlx5/lacZ during embryogenesis, and phenotypes observed in homozygous mutants. (A,B) Whole-mount X-gal staining of 8.0 dpc embryos in lateral (A) and dorsal (B) view. (C,D) 8.5 dpc-stained heterozygous (C) and homozygous mutant (D) embryos. In D, the first branchial arch (1) is indicated. Arrowheads in A-D indicate the otic placode. (E,F) 9.5 dpc embryos stained with X-gal showing expression of lacZ in the otic vesicle, branchial arches (indicated by numbers 1-4, G), olfactory placode and limb buds. An heterozygous (E) and an homozygous mutant (F) embryo are shown. (G) Detail of the branchial arches (1-4) to show expression of Dlx5/lacZ. In the second and third arch staining is observed in an anterior and a posterior streak of cells. (H,I) Expression of lacZ in normal (H) and homozygous mutant (I) 11.5 dpc embryos. (J-M) Expression of lacZ in 12.5 dpc embryos. Expression in Cre+ (J) and Cre (K) heterozygous embryos is compared. Clarified heterozygous (L) and homozygous mutant (M) embryos are shown. The arrow in H indicates the genital bud. (N-S) Expression pattern of lacZ in the developing limbs of 9.5 (N), 10.5 (O,P), 11.5 (Q), 13.5 (R) and 14.5 (S) dpc heterozygous embryos, in lateral view, except in P where a frontal view is shown to highlight the apical ectodermal ridge (AER). Limbs of homozygous mutant mice showed no difference in the β-gal-staining pattern at any stage of development. Arrowheads in N indicate the AER. White asterisks indicate a region of expression located at the anterior margin of the limb bud. At 13.5 (R) and 14.5 (S) dpc, expression in limb skeletal elements is also observed. For each pair of matched embryos, the same magnification is used. Abbreviations: de, diencephalon; sc, semicircular canals; tb, tail bud; te, telencephalon; vce, ventral cephalic epithelium.

Expression of the Dlx5/lacZ mutant allele during development

The normal development of heterozygous Dlx5 mutant mice allowed us to use the lacZ gene as a sensitive reporter of Dlx5 promoter activity during mouse embryogenesis. Furthermore comparison of lacZ expression in heterozygous and homozygous mutant embryos helped us to identify the lesions induced by gene inactivation. In heterozygous embryos, the expression pattern of lacZ reproduced in all cases the known profile of expression of Dlx5 obtained previously by RNA in situ hybridization (Yang et al., 1998; Simeone et al., 1994a,b). In 8.0 and 8.5 dpc embryos (Fig. 2A-D), lacZ activity is present in the ventral cephalic epithelium, in the otic and olfactory placodes, in a narrow strip of cells at the neural-non-neural boundary along the length of the neural plate and in the tail bud. The signal detected at the border of the neuroepithelium was more intense in regions rostral to the otic pit suggesting a possible expression of Dlx5 in presumptive premigratory cephalic neural crest cells (Yang et al., 1998). No obvious changes in the lacZ expression pattern nor any evident lesions were observed in homozygous embryos at these stages (Fig. 2D).

After anterior neuropore closure (8.5 dpc) and in all subsequent stages, the distal portion of the first (mandibular) branchial arch is stained by X-gal. At 9.5 dpc expression of the reporter gene can already be seen in the maxillary branch of the first branchial arch (Fig. 2E,F) where it rapidly increases at later stages (Fig. 2J-L). The second (hyoid) and, more weakly, the third and fourth branchial arches show expression of the reporter gene (Fig. 2E-G). The expression in the second and third arch is characterized by two streaks of positive cells at the anterior and posterior border of the arch (Fig. 2G). The expression in arch-derived structures persists later in development. In homozygous mutant embryos at 9.5 dpc, no obvious difference could be observed in lacZ expression in the branchial arches or in any other lacZ-expressing structures (Fig. 2E,F, see also Fig. 8I,K,M).

Expression in the brain begins around 9.5-10 dpc and, by 10.5 dpc, is confined to the basal anterior telencephalon, the ganglionic eminence and the thalamic region of the diencephalon. A detailed account of the effects of the mutation in these regions is given in a later section.

The first phenotypic differences induced by the inactivation of Dlx5 can be observed at 11.5 dpc (Fig. 2H,I) and becomes more evident at 12.5 dpc (Fig. 2L,M), affecting derivatives of the branchial arches and of the otic and olfactory placodes (see below). Starting at around 12 dpc, the reporter gene is expressed surrounding all endochondral bones (Fig. 2L,M,R,S). The pattern of lacZ expression in heterozygous embryos did not change after removal of the PGK-neoR cassette by crossing with CMV-Cre transgenic partners (Fig. 2J,K) indicating that the activity of the Dlx5 promoter was not affected by the presence of the PGK-neoR cassette.

Dlx5, as all other Dlx genes, is strongly expressed in all extending appendages including the limb bud, the genital tubercle and external ear. Expression in the limb bud is observed beginning at 9.5 dpc and in all subsequent stages (Fig. 2N-S). Initially, it is confined to the apical ectodermal ridge and to a wider region at the anteroproximal border of the limb bud. In later stages, expression marks the epithelium and underlying mesenchyme of the tip of all growing appendages, including finger and toe tips, the genital bud and the ear lobes.

Craniofacial defects during embryogenesis

Craniofacial malformations in Dlx5/lacZ homozygous embryos were seen at 14.5 dpc (Fig. 3A-C). Mutant embryos (Fig. 3B) could be recognized by their shorter snout and open fontanelle; in the few exencephalic embryos, the craniofacial defects were more conspicuous (Fig. 3C). Limb development was normal in all embryos as already observed for Dlx1 and Dlx2 mutants (Qiu et al., 1995, 1997). Cartilage skeleton preparations from 14.5 dpc embryos (Fig. 3D-F) confirmed the normal development of limbs and axial skeleton and the defective development of head structures.

Fig. 3.

Craniofacial defects in 14.5 dpc Dlx5/ mutant embryos. (A-C) Frontal view of normal (A), homozygous mutant (B) and exencephalic phenotype (C) embryos. Note the shorter snout and the open fontanelle of the mutant embryo (B) versus the normal (A), and the dramatically abnormal head and brain development of the exencephalic case (C). (D,E) Lateral view of cartilage skeleton at 14 dpc of normal (D), homozygous mutant (E) and exencephalic phenotype (F), stained with Alcian blue and clarified. Note the defects of the middle and inner ear structures (detailed in Fig. 4), the shortening and rotation of the Meckel’s cartilage, and the profound skull alterations in the exencephalic animal. The body and limb skeleton appears normal in all cases.

Fig. 3.

Craniofacial defects in 14.5 dpc Dlx5/ mutant embryos. (A-C) Frontal view of normal (A), homozygous mutant (B) and exencephalic phenotype (C) embryos. Note the shorter snout and the open fontanelle of the mutant embryo (B) versus the normal (A), and the dramatically abnormal head and brain development of the exencephalic case (C). (D,E) Lateral view of cartilage skeleton at 14 dpc of normal (D), homozygous mutant (E) and exencephalic phenotype (F), stained with Alcian blue and clarified. Note the defects of the middle and inner ear structures (detailed in Fig. 4), the shortening and rotation of the Meckel’s cartilage, and the profound skull alterations in the exencephalic animal. The body and limb skeleton appears normal in all cases.

The most obvious cranial alteration in Dlx5−/− 14.5 dpc embryos affects Meckel’s cartilage, which appears shorter and fuses to the contralateral partner with a wider angle compared to normal embryos (Figs 3E,F, 4A,B). In addition it is interrupted near its proximal end generating an additional joint and leaving the malleus primordium as an individual cartilage element (Fig. 4D,E). One to three additional cartilage elements, form ectopically at the site of the interruption in Meckel’s cartilage, at the position where the gonial and tympanic bones will form. In some cases, one of these elements is fused with the distalmost portion of the malleus. These additional cartilages contribute to the development of part of the ectopic bones observed at birth (see below).

Fig. 4.

Defects in the head cartilage skeleton at 14.5 dpc. (A,B) Ventral view of the base of the skull of 14.5 dpc normal (A) and homozygous mutant (B) embryos, stained with Alcian blue. (C-E) Lateral view of the proximal-most portion of the Meckel’s cartilage and the middle ear ossicles, dissected from normal (C), homozygous mutant (D), or exencephalic phenotype (E) embryos. In the inserts is shown a detail of the stapes. In mutant animals, the Meckel’s cartilage is interrupted and disconnected from the malleus primordium, and gives rise to additional skeletal elements directed toward the midline (arrows). The position and articulation of the ossicles appears normal (arrowheads), except for the stapes which is missing in the exencephalic phenotype. Abbreviations: at, auditory tube; co, cochlea; i, Incus; m, malleus; mc, Meckel’s cartilage; ph, pharynx; s, stapes; sty, styloid process; th, thyroid cartilage; ts, tympanic space; ve, vestibulum.

Fig. 4.

Defects in the head cartilage skeleton at 14.5 dpc. (A,B) Ventral view of the base of the skull of 14.5 dpc normal (A) and homozygous mutant (B) embryos, stained with Alcian blue. (C-E) Lateral view of the proximal-most portion of the Meckel’s cartilage and the middle ear ossicles, dissected from normal (C), homozygous mutant (D), or exencephalic phenotype (E) embryos. In the inserts is shown a detail of the stapes. In mutant animals, the Meckel’s cartilage is interrupted and disconnected from the malleus primordium, and gives rise to additional skeletal elements directed toward the midline (arrows). The position and articulation of the ossicles appears normal (arrowheads), except for the stapes which is missing in the exencephalic phenotype. Abbreviations: at, auditory tube; co, cochlea; i, Incus; m, malleus; mc, Meckel’s cartilage; ph, pharynx; s, stapes; sty, styloid process; th, thyroid cartilage; ts, tympanic space; ve, vestibulum.

At 14.5 dpc, the middle ear ossicles hammer, incus and stapes appear normally shaped and positioned, and engage in proper articulations with one another (Fig. 4C-E), except for a few exencephalic embryos in which the stapes was absent (Fig. 4E). Likewise, at birth, the middle ear ossicles appear normal. This was confirmed by serial histology of the ear region of normal and mutant animals, which indicated that, in Dlx5-deficient animals, the malleus contacts the tympanic membrane and the stapes contacts the vestibular window of the sacculus correctly (data not shown). The styloid process appears shorter than normal and is rotated toward the midline (Fig. 4A,B).

Craniofacial defects at birth

As expected from the analysis of the mutant embryos, bones derived from branchial arches forming the base and sides of the skull are severely defective in newborn homozygous mutant mice (Fig. 5); while minor defects appear in neurocranial bones. Dlx5/ animals present a cleft secondary palate; the horizontal laminae of the palatine bones are missing. This causes the presphenoid bone to become visible in ventral view. The nasal and maxillary bones are shorter, resulting in a general reduction of the length of the snout. The palatine processes of the maxilla are reduced especially with respect to their posterior development and they fail to form proper connections with the palatine bones (Fig. 5A,B). We observe a deformation of the anterior part of the pterygoids, which appear to have changed their angles with respect to the basisphenoid bone (Fig. 5A,B,E,F). The tympanic ring is always reduced in length, although with considerable variation among individuals (Fig. 5A,B,H,I). The region of the alisphenoid comprising the foramen rotundum and ovalis is misshapen, with the foramina fused to form one large irregular opening (Fig. 5E,F). The remaining anterolateral portion of the lamina obturans and the ali-cochlear commissure, connecting the ala temporalis with the otic capsule, appears normal although slightly mispositioned due to the rotation of the otic region. Finally, the squamous and jugal bones are not obviously affected.

Fig. 5.

Craniofacial defects in newborn Dlx5/ mutant animals. (A,B) Ventral views of skulls of normal (A) and homozygous (B) mutant animals after removal of the jaw. (A′,B′) Drawings derived from images A and B of the cranial bones affected in Dlx5 mutant mice: green, unaffected bones; yellow, affected (reduced, rotated or deformed, see text) bones; red, the ectopic bone. (C,D) Dorsal view of normal (C) and homozygous mutant (D) skulls. (E,F) Details of the region of the Pterygoid-Lamina Obturans bones of the base of the skull, in normal (E) and homozygous mutant animals (F). White asterisk indicates the ectopic bone. (G) Dissected jaws of normal (left) and homozygous mutant (right) animals. Note that the coronoid process (arrow) is missing in the mutant jaw. (H,I) Dissected middle and inner ear region of normal (H) and homozygous mutant (I) newborn animals, stained with Alcian blue-Alizarin. Note the position of the ectopic bone (asterisk) and the malformation of the labyrinth. (J) Dissected hyoid bone and laryngeal cartilages from normal (right) and homozygous mutant (left) newborn animals, stained with Alcian blue-Alizarin. Note the reduction of the small horns of the hyoid bone (black arrowheads) and the missing superior horns of the thyroid cartilage (black arrow). Abbreviations: bo, basioccipital; bs, basisphenoid; fo, foramen ovalis; fr, foramen rotundum; ft, frontal; g, gonial; hy, hyoid; ip, interparietal; lo, lamina obturans; mc, Meckel’s cartilage; mx, maxilla; n, nasal; p, parietal; pt, pterygoid; pl, palatine; ps, presphenoid; tr, tympanic ring.

Fig. 5.

Craniofacial defects in newborn Dlx5/ mutant animals. (A,B) Ventral views of skulls of normal (A) and homozygous (B) mutant animals after removal of the jaw. (A′,B′) Drawings derived from images A and B of the cranial bones affected in Dlx5 mutant mice: green, unaffected bones; yellow, affected (reduced, rotated or deformed, see text) bones; red, the ectopic bone. (C,D) Dorsal view of normal (C) and homozygous mutant (D) skulls. (E,F) Details of the region of the Pterygoid-Lamina Obturans bones of the base of the skull, in normal (E) and homozygous mutant animals (F). White asterisk indicates the ectopic bone. (G) Dissected jaws of normal (left) and homozygous mutant (right) animals. Note that the coronoid process (arrow) is missing in the mutant jaw. (H,I) Dissected middle and inner ear region of normal (H) and homozygous mutant (I) newborn animals, stained with Alcian blue-Alizarin. Note the position of the ectopic bone (asterisk) and the malformation of the labyrinth. (J) Dissected hyoid bone and laryngeal cartilages from normal (right) and homozygous mutant (left) newborn animals, stained with Alcian blue-Alizarin. Note the reduction of the small horns of the hyoid bone (black arrowheads) and the missing superior horns of the thyroid cartilage (black arrow). Abbreviations: bo, basioccipital; bs, basisphenoid; fo, foramen ovalis; fr, foramen rotundum; ft, frontal; g, gonial; hy, hyoid; ip, interparietal; lo, lamina obturans; mc, Meckel’s cartilage; mx, maxilla; n, nasal; p, parietal; pt, pterygoid; pl, palatine; ps, presphenoid; tr, tympanic ring.

The gonial bone forms in a normal position but is misshapen and in several cases is fused to one or two ectopic bones not present in normal skulls (Fig. 5H,I). These novel bones extend from the gonial in the direction of the pterygoid, have an extremely irregular and variable shape, and do not resemble any other structure. The variability in size and shape of this ectopic bones is seen between individuals and within the same individual between the left and the right side. Examination of the skeleton of mutant embryos at 16.5 dpc indicated a continuity between the proximal end of Meckel’s cartilage (which is disconnected from the malleus) and the lateral-most portion of the ectopic bone. Thus Meckel’s cartilage and the extranumerary cartilage elements seen at 14.5 dpc contribute to the ectopic bone formation.

At the base of the skull, along the midline, we observe an elongation of the basioccipital bone (Fig. 5A,B), while the basisphenoid and the presphenoid appear normal. Laterally, the cochlear part of the cartilaginous otic capsule is normal while the vestibular part is always reduced, deformed and rotated. This alteration is likely to be a consequence of the deformation affecting the membranous labyrinth hosted within (see below). The mandible is shorter, deformed and lacking the coronoid process (Fig. 5G).

Turning to the dermatocranial bones forming the roof of the skull, we observe delayed ossification affecting the parietal, interparietal and superoccipital bones, resulting in an open anterior and posterior fontanellae as well as wide cranial sutures with multiple Wormian bones; the nasal bones are always reduced in length and size (Fig. 5C,D). It is important to note that Dlx5 (and lacZ in the mutant mice) is expressed in all differentiating bone tissue (see later). Defects of dermatocranial bones might derive from a generalized defect in osteogenesis. In the mice with exencephaly, the roof bones are missing altogether. The lesser horns of the hyoid are reduced and rotated from their normal angle of attachment with the central bony structure, the bigger horns are less affected, but form a different angle with the body of the hyoid bone. Finally, the thyroid cartilage lacks its superior horns (Fig. 5J).

The skeleton of the trunk and limbs of Dlx5/ mice was found to be morphologically normal at all stages of development and at birth (Fig. 3D-F).

Analysis of Dlx5/ developing brains

Dlx5 is expressed in several areas of the forebrain including the lateral ganglionic eminence, the septal area, the ventral thalamus and restricted regions within the hypothalamus (Simeone et al., 1994a,b).

In all these areas, Dlx5 is coexpressed with other members of the Dlx gene family (Simeone et al., 1994a; Bulfone et al., 1993; Rubenstein et al., 1994). In contrast, a number of genes are known to be transcribed in regions adjacent and/or complementary to those expressing Dlx5. To evaluate possible abnormalities in Dlx5/ developing brains, we analysed at 12.5 and 15 dpc the expression pattern of genes transcribed within the Dlx5 territory such as Dlx1 and Dlx6 (Simeone et al., 1994a) or bordering the Dlx5 expression domain such as Tbrain1, Otx2, Orthopedia (Otp) and Sonic hedgehog (Shh) (Bulfone et al., 1995; Simeone et al., 1993, 1994b; Echelard et al., 1993). In Dlx5+/ embryos, the expression pattern of the lacZ gene fully overlapped that of the normal allele as deduced in adjacent sections (data not shown), thus indicating that lacZ expression only identified Dlx5 transcribing cells. We, therefore, compared the expression pattern of lacZ, Dlx1, Dlx6, Tbrain1, Otx2, Otp and Shh in Dlx5+/ and Dlx5/ brains at 12.5 and 15 dpc.

At 12.5 dpc no obvious difference was detected in the expression pattern of all these genes, either within the territory expressing lacZ (Dlx5) or in the adjacent and complementary areas (Fig. 6A-P′). In particular, no major difference was observed anteriorly, in sections through the dorsal telencephalon, the medial and lateral ganglionic eminence and the septal area (Fig. 6A-E′), medially, in sections through the dorsal thalamus, the ventral thalamus and the hypothalamic anterior preoptic area (Fig. 6F-L′) and posteriorly, in sections through the ventral thalamus, the posterior entopeduncolar area and the suprachiasmatic area (Fig. 6M-P′). The minor differences observed were due to a non-perfect orientation of the two embryos. To gain insight into the possibility that abnormalities were generated in the brain of Dlx5/ embryos later than 12.5 dpc, the expression pattern of lacZ, Dlx1, Tbrain1, Otx2 and Otp genes was studied at 15 dpc. Also at this stage, no obvious abnormality was detected in Dlx5 mutant brain (Fig. 7A-H′). Although these findings fail to highlight any abnormality either in brain territories expressing Dlx5 or in adjacent regions, the possibility that more subtle phenotypes could be detected in more restricted areas or in specific cell lineages remains still open.

Fig. 6.

Expression pattern of forebrain markers in frontal sections of Dlx5 mutant embryos at 12.5 dpc. (A-P′) Dlx5+/− (A-P) and Dlx5/ (A′-P′) embryos hybridized with lacZ (A,A′,F,F′,M,M′), Dlx1 (B,B′,G,G′,N,N′), Dlx6 (C,C′,H,H′,O,O′), Tbrain1 (D,D′,I,I′), Otx2 (E,E′,J,J′), Otp (K,K′,P,P′) and Shh (L,L′). Abbreviations: se, septal region; mge, medial ganglionic eminence; lge, lateral ganglionic eminence; dTe, dorsal telencephalon; dt, dorsal thalamus; vt, ventral thalamus; poa, preoptic anterior area; emt, eminentia thalami; ZLI, zona limitans intrathalamica; mri, massa cellularis reuniens pars inferior; pep, posterior entopeduncolar area; sch, suprachiasmatic area.

Fig. 6.

Expression pattern of forebrain markers in frontal sections of Dlx5 mutant embryos at 12.5 dpc. (A-P′) Dlx5+/− (A-P) and Dlx5/ (A′-P′) embryos hybridized with lacZ (A,A′,F,F′,M,M′), Dlx1 (B,B′,G,G′,N,N′), Dlx6 (C,C′,H,H′,O,O′), Tbrain1 (D,D′,I,I′), Otx2 (E,E′,J,J′), Otp (K,K′,P,P′) and Shh (L,L′). Abbreviations: se, septal region; mge, medial ganglionic eminence; lge, lateral ganglionic eminence; dTe, dorsal telencephalon; dt, dorsal thalamus; vt, ventral thalamus; poa, preoptic anterior area; emt, eminentia thalami; ZLI, zona limitans intrathalamica; mri, massa cellularis reuniens pars inferior; pep, posterior entopeduncolar area; sch, suprachiasmatic area.

Fig. 7.

Expression pattern of forebrain markers in frontal sections of Dlx5 mutant embryos at 15 dpc. (A-H′) Dlx5+/− (A-H) and Dlx5/ (A′-H′) embryos hybridized with lacZ (A,A′,E,E′), Dlx1 (B,B′,F,F′), Tbrain1 (C,C′,G,G′), Otx2 (D,D′) and Otp (H,H′). Abbreviations stand as in the previous figure plus: st, striatum; spv, supraoptic/paraventricular area; ah, anterior hypothalamus.

Fig. 7.

Expression pattern of forebrain markers in frontal sections of Dlx5 mutant embryos at 15 dpc. (A-H′) Dlx5+/− (A-H) and Dlx5/ (A′-H′) embryos hybridized with lacZ (A,A′,E,E′), Dlx1 (B,B′,F,F′), Tbrain1 (C,C′,G,G′), Otx2 (D,D′) and Otp (H,H′). Abbreviations stand as in the previous figure plus: st, striatum; spv, supraoptic/paraventricular area; ah, anterior hypothalamus.

Altered morphogenesis of the semicircular canals

The Dlx5/lacZ allele is expressed in the otic pit and later in the otic vesicle starting 8.0 dpc (Fig. 2A). We have therefore analyzed the development of the inner ear in mutant animals.

In heterozygous animals, Dlx5/lacZ is initially expressed on the dorsoposterior region of the otic vesicle and subsequently in the semicircular canals and in the endolymphatic duct and vesicle of the vestibular organ (Fig. 8). In Dlx5/ embryos, the vestibulum is smaller in size and heavily deformed; the three canals fail to form properly, the anterior and posterior canals do not develop and are fused into one single large vesicle, and the horizontal canal is also reduced (Fig. 8B,D,F). The development of the endolymphatic duct is much less affected by the mutation. The morphology of the sacculus and the cochlea appears essentially normal. Although the lesion was present with complete penetrance and was similar between individuals, the severity of the dysgenesis varied among individual mutant animals and between both ears within individuals. The inner ear epithelium of mutant embryos and newborn animals appears much thinner compared to the normal and is composed of large flat cells (Fig. 8G,H). Within this mutant epithelium, thickened regions are observed that may represent remnants of the cristae ampullaris, which could not be recognized. In contrast, the maculae of the utriculus and sacculus appear normal (not shown).

Similar vestibular defects have been observed in mice mutant for Nkx-5.1/Hmx3, an NK-related homeobox gene (Hadrys et al., 1998; Wang et al., 1998). In order to evaluate whether Dlx5 could regulate or be regulated by Nkx-5.1, we have analyzed the expression of both genes in Dlx5/ and in Nkx-5.1/ mice (Fig. 8I-N). Inactivation of either gene did not abrogate or profoundly modify expression of the other, excluding the possibility that they control reciprocally their expression. Furthermore, while Dlx5 was strongly expressed in the endolymphatic duct at all stages of development, Nkx-5.1 was never expressed in this structure indicating a different regulation (Fig. 8I-N and data not shown).

Proliferation and apoptosis in Dlx5/ embryos

Some features of the Dlx5/ phenotype may suggest that proliferation and/or cell survival might be affected in tissues and structures where Dlx5 is expressed. In order to address this issue, we studied cell proliferation and survival in Dlx5+/ and Dlx5/ embryos at 10.5 and 11.5 dpc.

Cell proliferation was visualized by a short pulse of bromodeoxyuridine (BrdU) incorporation and subsequent detection of BrdU-positive cells. Apoptotic cell death was studied by the TUNEL method (Gravieli et al., 1992).

At 10.5 dpc, we concentrated our attention on the branchial arches (Fig. 9A-C′, G-I′), the primordium of the ganglionic eminence (Fig. 9D-F′) and the otic vesicle (Fig. 9J-L′). In particular, cell proliferation and survival were studied in adjacent sections and compared to the lacZ expression domain. No significant difference was detected in the number of apoptotic cells labelled in the branchial arches (Fig. 9C,C′,I,I′), in the primordium of the ganglionic eminence (Fig. 9F,F′) and in the otic vesicle (Fig. 9L,L′). A similar result was obtained in other embryonic districts expressing the Dlx5 gene (data not shown). At 11.5 dpc, as compared to heterozygous or wild-type embryos, the pattern of apoptosis also remained unaltered (data not shown). The number of BrdU-positive cells in Dlx5+/ embryos was compared to that of Dlx5/ embryos in the same territory. As revealed in serial adjacent sections through the branchial arches of three embryos for each genotype, the percentage of proliferating cells detected per unit of surface (see Material and Methods) was unaltered (Fig. 9B,B′,H,H′ and data not shown). In more medial sections, the lacZ expression domain included the rostral third of the first branchial arch of Dlx5+/ embryos (Fig. 9G). Interestingly, in the same territory, an increased density of proliferating cells was detected (compare Fig. 9G to H) while, in the rest of the branchial arch, BrdU-positive cells appeared lower in number (Fig. 9H). As compared to Dlx5+/ embryos, in Dlx5/ mutants, the lacZ expression domain was expanded into an area including about half or more of the first branchial arch (compare Fig. 9G′ to G). Noteworthy, the territory with the highest density of proliferating cells also resulted expanded and roughly coincident with the lacZ expression domain (compare Fig. 9H′ to G′).

Fig. 8.

Defects of the inner ear in Dlx5-deficient mice. (A-F) X-gal stained embryos and newborn in whole-mount clarified preparations. Semicircular canals and endolymphatic duct of the labyrinth of heterozygous (A,C,E) or homozygous Dlx5 mutant (B,D,F) mice. Embryos at 10.5 (A,B) and 12.5 (C,D) dpc, or newborn animals (E,F) are shown. Note the reduction in size and deformation of the organ. (G,H). Histological appearance of the epithelium lining the semicircular canals of the labyrinth of normal (G) and homozygous mutant (H) newborn mice. Note the flattening of the vestibular epithelial cells. (I-N) Whole-mount in situ hybridization on 9.75 dpc embryos with Dlx5 (I,K,M) or Nkx-5.1 (J,L,N) probes. The genotype of each embryo is indicated on the corresponding panel. Abbreviation: ac, anterior canal; ed, endolymphatic duct; hc, horizontal canal; pc, posterior canal; tr, tympanic ring.

Fig. 8.

Defects of the inner ear in Dlx5-deficient mice. (A-F) X-gal stained embryos and newborn in whole-mount clarified preparations. Semicircular canals and endolymphatic duct of the labyrinth of heterozygous (A,C,E) or homozygous Dlx5 mutant (B,D,F) mice. Embryos at 10.5 (A,B) and 12.5 (C,D) dpc, or newborn animals (E,F) are shown. Note the reduction in size and deformation of the organ. (G,H). Histological appearance of the epithelium lining the semicircular canals of the labyrinth of normal (G) and homozygous mutant (H) newborn mice. Note the flattening of the vestibular epithelial cells. (I-N) Whole-mount in situ hybridization on 9.75 dpc embryos with Dlx5 (I,K,M) or Nkx-5.1 (J,L,N) probes. The genotype of each embryo is indicated on the corresponding panel. Abbreviation: ac, anterior canal; ed, endolymphatic duct; hc, horizontal canal; pc, posterior canal; tr, tympanic ring.

Fig. 9.

Cell proliferation and apoptosis in Dlx5 mutant embryos at 10.5 dpc. (A-L′) Adjacent sagittal sections of Dlx5+/− (A-L) and Dlx5/ (A′-L′) embryos showing lacZ expression domains (A,A′,G,G′,D,D′,J,J′) BrdU-positive cells (B,B′,H,H′,E,E′,K,K′) and TUNEL-positive cells (C,C′,I,I′,F,F′,L,L′) throughout branchial arches (A-C′,G-I′), primordium of ganglionic eminence (D-F′) and otic vesicle (J-L′). The arrows in C,C′,I,I′,L,L′ point to single or groups of apoptotic cells that are identified in both Dlx5+/ and Dlx5/ embryos. Dots in H,H′ define the boundary between lacZ-positive and -negative areas; Dots in C,C′ define the border of branchial arches. Abbreviations: fg, foregut; ba, bulbus anteriosus; I and II, first and second branchial arches. Note that K′ is not exactly adjacent to J′ and L′.

Fig. 9.

Cell proliferation and apoptosis in Dlx5 mutant embryos at 10.5 dpc. (A-L′) Adjacent sagittal sections of Dlx5+/− (A-L) and Dlx5/ (A′-L′) embryos showing lacZ expression domains (A,A′,G,G′,D,D′,J,J′) BrdU-positive cells (B,B′,H,H′,E,E′,K,K′) and TUNEL-positive cells (C,C′,I,I′,F,F′,L,L′) throughout branchial arches (A-C′,G-I′), primordium of ganglionic eminence (D-F′) and otic vesicle (J-L′). The arrows in C,C′,I,I′,L,L′ point to single or groups of apoptotic cells that are identified in both Dlx5+/ and Dlx5/ embryos. Dots in H,H′ define the boundary between lacZ-positive and -negative areas; Dots in C,C′ define the border of branchial arches. Abbreviations: fg, foregut; ba, bulbus anteriosus; I and II, first and second branchial arches. Note that K′ is not exactly adjacent to J′ and L′.

This finding, therefore, suggests that, rather than controlling cell proliferation, Dlx5 is required to define a territory where the highest density of proliferating cells is detected. In the absence of Dlx5, this territory expands and a new boundary is defined. Cell proliferation was also studied in other districts such as the primordium of the ganglionic eminence (Fig. 9E,E′) and the otic vesicle (Fig. 9K,K′). No relevant difference was detected in these structures.

The effect of Dlx5 mutation on osteoblast differentiation

During development, Dlx5/lacZ is strongly expressed in all sites of perichondral bone formation (Figs 2R,S, 10A-D). In the skeleton, β-galactosidase activity was first detected in the peripheral region of the cartilage model of long bones around 13.5 dpc (Fig. 10A). At this stage, expression was confined to a population of osteoblasts located at the periphery of the diaphyseal part of the cartilage model with no labelling of the more centrally located chondrocytes. At later stages, β-galactosidase activity appeared progressively in all bones. Up to 16.5 dpc it could only be detected in periosteal preosteoblasts and osteoblasts. Perinataly, staining of some endosteal cells, presumably endosteal osteoblasts was also detected (Fig. 10C). During embryonic development, the cartilage was negative.

Fig. 10.

Expression of Dlx5 in developing bones and bone defects in Dlx5-deficient mice. (A) Dlx5/lacZ pattern of expression in a section through the rib of a 13.5 Dlx5+/− embryo. Staining is present only in the perichondral region. (B) Section through the diaphysis of the tibia at 14.5 dpc. Only preosteoblasts and osteoblasts within the periosteum express Dlx5/lacZ. (C) Expression of Dlx5/lacZ in a section through the growth plate of the tibia of a newborn Dlx5+/− animal. Activity is seen in the periosteum and in few endosteal osteoblasts (arrow). (D) Whole-mount X-gal staining of the tibia of a heterozygous animal at birth. (E,F) Histological analysis (Mallory trichromic) of ribs from normal (E) or homozygous mutant (F) animals. (G,H) Immunohistochemistry with anti-osteocalcin antibody on the femur periosteal region from heterozygous (G) and homozygous (H) mutant mice. Arrows in H indicate periosteal osteoblasts positive both for Dlx5/lacZ activity and anti-osteocalcin immunoreactivity. Before anti-osteocalcin immunohistochemistry sections G, H were treated with X-gal. (I-M) In situ hybridization on sections through the ribs of Dlx5+/− (I,J,L) or Dlx5/ (K,M) 16.5 dpc embryos. Probes are indicated on the panel. Care was taken to choose matching sections corresponding to the same rib, cut at the same level. Abbreviation: c, cartilage; hc, hypertrophic cartilage; po, periosteum; tb, trabecular bone. Magnification A-C, ×250; D, ×8; in E,F, ×80; G,H ×1000.

Fig. 10.

Expression of Dlx5 in developing bones and bone defects in Dlx5-deficient mice. (A) Dlx5/lacZ pattern of expression in a section through the rib of a 13.5 Dlx5+/− embryo. Staining is present only in the perichondral region. (B) Section through the diaphysis of the tibia at 14.5 dpc. Only preosteoblasts and osteoblasts within the periosteum express Dlx5/lacZ. (C) Expression of Dlx5/lacZ in a section through the growth plate of the tibia of a newborn Dlx5+/− animal. Activity is seen in the periosteum and in few endosteal osteoblasts (arrow). (D) Whole-mount X-gal staining of the tibia of a heterozygous animal at birth. (E,F) Histological analysis (Mallory trichromic) of ribs from normal (E) or homozygous mutant (F) animals. (G,H) Immunohistochemistry with anti-osteocalcin antibody on the femur periosteal region from heterozygous (G) and homozygous (H) mutant mice. Arrows in H indicate periosteal osteoblasts positive both for Dlx5/lacZ activity and anti-osteocalcin immunoreactivity. Before anti-osteocalcin immunohistochemistry sections G, H were treated with X-gal. (I-M) In situ hybridization on sections through the ribs of Dlx5+/− (I,J,L) or Dlx5/ (K,M) 16.5 dpc embryos. Probes are indicated on the panel. Care was taken to choose matching sections corresponding to the same rib, cut at the same level. Abbreviation: c, cartilage; hc, hypertrophic cartilage; po, periosteum; tb, trabecular bone. Magnification A-C, ×250; D, ×8; in E,F, ×80; G,H ×1000.

Although at a first macroscopic examination bones in mutant embryos appeared normal, histological analysis revealed a lesion characterised by the presence of a more complex structure of the endosteal component of the diaphysis, which forms an elaborate mesh of woven bone, and by the reduction of the periosteal bone lamina. This defect is more obvious in certain bones, e.g. the ribs, where the endosteal space is strongly reduced (Fig. 10E,F).

In heterozygous animals, we never detected osteocalcin expression in the periosteum at birth; periosteal cells of Dlx5/ animals were simultaneously stained for Dlx5/lacZ and osteocalcin. This might suggest a role for Dlx5 as a repressor of osteocalcin expression in vivo (Fig. 10G,H).

To check whether any major alteration occurred in genes involved in the control of cartilage and bone differentiation, the expression of Cbfa1 and Ihh was analyzed by in situ hybridization on matching sections of 16.5 dpc embryos without finding any major difference (Fig. 10I-M).

We have demonstrated that the homeobox gene Dlx5 plays a major role in the control of craniofacial structures morphogenesis, in the development of the vestibular organ and in bone formation. Newborn homozygous Dlx5/ animals die at birth accumulating air in their stomach and intestine. Perinatal death with similar symptoms has been observed in several other mutants with a cleft secondary palate including, between others, the two other Distal-less related genes Dlx1 and Dlx2 (Qiu et al., 1995, 1997).

Development of craniofacial structures requires complex interactions involving multiple embryonic tissues (Hanken and Hall, 1993). Lineage studies mainly carried out in the chick embryo show that specific subpopulations of cranial neural crest (CNC) cells participate in craniofacial morphogenesis. Some of these cells are responsible for direct bone tissue deposition within the cranial mesenchyme, which then develop into bones of the dermatocranium. Other CNC cells migrate from the midbrain-hindbrain region into the pharyngeal arches (Lumsden et al., 1991; Serbedzija et al., 1992; Kontges and Lumsden, 1996) to give rise to specific cranial structures, either directly (arch-derived dermatocranium) or through a cartilage intermediate (splanchnocranium) (Trainor and Tam, 1995; Kontges and Lumsden, 1996).

The genetic code responsible for the determination of craniofacial patterning is now beginning to be elucidated thanks to gene targeting technologies (for review see Francis-West et al., 1998).

Our data confirm the observation that Dlx5 is expressed very early on in embryonic development in anterior regions of the embryo (Yang et al., 1998). During early somite stages, Dlx5 expression is found in a stripe of cells at the periphery of the neural plate, which might correspond to the position of the presumptive premigratory neural crest. Later expression continues in cranial neural crest derivatives that form the mesenchyme of the branchial arches.

The expression territories of Dlx genes within the branchial arches have been determined by in situ hybridization on 10.5 dpc embryos (Qiu et al., 1997). At this stage of development, nested expression pattern of Dlx3, Dlx5 and Dlx6 characterize the distal part of the arch, while Dlx1 and Dlx2 have a larger expression territory. These data have led to the hypothesis that Dlx genes would have partially redundant functions in the distal part of the arch. According to this model, Dlx5 inactivation should affect, if anything, skeletal elements distal to those observed in the Dlx1 and Dlx2 mutants. In the light of our results, such model seems only partly true.

Dlx5 inactivation causes defects in skeletal derivatives of the first four branchial arches (Table 1). The craniofacial defects that we observe are in a sense more ‘distal’ than those produced by Dlx1 and Dlx2 inactivation. In particular, we have not observed defects of the incus, the jugal and the squamous bones of the maxillary (proximal) portion of the first arch, seen in the Dlx2 mutant. Conversely, the Meckel’s cartilage, the mandible, the tympanic and gonial bones, from the mandibular (distal) part of the first branchial arch and the hyoid lesser horn derived from the distal part of the second arch are deformed in Dlx5/ mice and are unaffected in other Dlx mutants. The presence of defects in derivatives of the mandibular arch of Dlx5/ animals, where also Dlx1, Dlx2, Dlx3 and Dlx6 are expressed at 10.5 dpc suggests that redundancy between Dlx genes is not generalized, but occurs only in specific cases (e.g. the absence of molars in Dlx1 and Dlx2 double mutants). Some craniofacial defects observed in Dlx5/ mice cannot be simply explained by a proximodistal patterning of arch organization. For example, in both Dlx2/ and Dlx5/ mice, defects of the maxilla and the palatine bones are observed that derive from proximal position of the maxillary arch where Dlx5 is not expressed at 10.5 dpc (but where is expressed at later stages). In our view, the origin of the molecular patterning within the branchial arches should be seen in a more dynamic perspective. Our data show (Figs 2 and 8) that the territory of expression of Dlx5 in the first branchial arch changes during development. At 9.5 dpc, Dlx5 is expressed in the distal part of the mandibular portion of the first arch with low incipient expression in the maxillary arch. In subsequent stages, the territory of expression rapidly extends so that, at 10.5 dpc, Dlx5 is expressed along most of the mandibular arch and strongly in the maxillary arch. Precise timing in the expression of sets of genes interacting in a complex spatiotemporal manner is required to assure correct embryonic development. Patterning within the branchial arches does not constitute an exception. To better understand the mode of action of Dlx genes in controlling craniofacial development, we studied cell proliferation and survival in the branchial arches of Dlx5+/ and Dlx5/ embryos at E10.5 and E11.5. We found that inactivation of Dlx5 alters the territory with the highest density of proliferating cells within the first branchial arch. In the absence of Dlx5, this territory results expanded and a new boundary is defined. Whole-mount in situ analysis of the pattern of expression of Msx1 and Msx2, two potential modulators of Dlx5 activity, did not show any major difference between normal and mutant embryos (data not shown). As Dlx genes are most probably acting as transcriptional regulators, the identification and characterization of their targets remains critical for the elucidation of their function.

Table 1.

Defects caused by Dlx5 inactivation in skeletal derivatives of the pharyngeal arches

Defects caused by Dlx5 inactivation in skeletal derivatives of the pharyngeal arches
Defects caused by Dlx5 inactivation in skeletal derivatives of the pharyngeal arches

In the mutant mice, we have also observed a deformation of bones located along the midline of the base of the cranium (e.g. the basoccipital bone was more elongated). These bones are not crest derived, the change in their shape might derive by the change in boundary topological restrains imposed by the surrounding crest-derived structures. However, one should consider that Dlx5 is expressed in these bones, as in any other bone, during later phases of osteoblast differentiation (see later), it is possible that defects observed in midline-located bones and in neurocranial bone derive in part from a delay in osteoblast differentiation.

We did not find any obvious malformations in the limbs of Dlx5 mutant animals. In man, DLX5 and DLX6 genes are considered as candidate genes for certain types of split hand/foot malformations (SHFM) (for a review see Buss, 1994) since they are expressed in the developing limb and map to the critical interval of SHFM1 on chromosome 7q21.1. Furthermore, in some families, SHFM with complete penetrance is correlated to deletions, inversions or translocations of the chromosomal region 7q21.3-q21.1 (Scherer et al., 1994). We have recently found (Pfeffer et al., unpublished data) that the first exon of human and mouse DLX6 genes contain a CAG/CCG (poly-glutamine/poly-proline) repeat region with high homology to the trinucleotide repeat present in the Huntington’s disease gene. This CAG repeat is polymorphic in the normal human population suggesting that DLX6 could have a role in the control of limb patterning. Mutation analysis of Dlx6 will possibly contribute to answer this question.

Dlx5 is one of the first genes to be expressed in the otic pit and later in the lateral wall of the otocyst and in the epithelium of the vestibular apparatus. Our data show that inactivation of Dlx5 leads to severe malformations of the semicircular canals of the inner ear. This lesion is in agreement with fate map studies that have traced the origin of vestibular cells to the lateral part of the otocyst (Li et al., 1978). Similar vestibular defects have been observed in mice mutant for Nkx-5.1/Hmx3 (Hadrys et al., 1998; Wang et al., 1998). We have excluded the possibility of a direct cross-regulation between these two genes. It is, however, still possible that Dlx5 and Nkx-5.1 have a synergistic effect in the control of vestibular morphogenesis, possibly acting on different cellular compartments.

Our data indicate that Dlx5 expression is elevated during osteoblast differentiation and disappears in fully differentiated osteocytes. Its expression is more evident in periosteal bone, but is also seen in cells of the endosteal compartment, which might represent osteoblasts at a specific stage of differentiation. Dlx5/ mice show a delayed ossification of dermatocranial bones, which closely resemble that observed in mice in which one copy of the Cbfa1 gene is inactivated (Otto et al., 1997). The defect in osteogenesis that we observed in Dlx5−/− mice suggests that this gene plays a role in osteoblast differentiation and in bone formation; our data show an increased complexity of the structure of woven bone and a reduction of the periosteal bone lamina. The expression of Cbfa1, a key regulator of osteoblast differentiation, was not affected in Dlx5 mutants. However, we observed an increased osteocalcin expression in the periosteum suggesting a lesion in osteoblast differentiation. Unfortunately, as the mice died at birth, we could not follow the effect of Dlx5 in later phases of mineralization and during the formation of compact bone. The increased osteocalcin expression in the periosteum observed in mutant animals could corroborate the notion that Dlx5 can act as a repressor of the osteocalcin gene (Ryoo et al., 1997); however other more complex pathways of regulation cannot be ruled out.

In conclusion, we have shown that Dlx5 has multiple and independent functions in the patterning of the branchial arches, in the morphogenesis of the vestibular organ and in osteoblast differentiation. The next challenge will be to unravel the network of regulations that links this mutation to each of these phenotypes.

G. L. was supported by grants from Association Pour la Recherche sour le Cancer, ARSEP, AISM, Consiglio Nazionale delle Ricerche (Progetto Finalizzato ‘Biotecnologie’) and Ministero della Sanità. The support from Telethon (Italy) for the project: ‘Use of transgenic mutant mice as a model to study the molecular control of bone development and peripheral myelination and to develop new gene therapy strategies in the embryo’ (Project D76) is gratefully acknowledged. A. S. was supported by the Associazione Italiana per la Ricerca sul Cancro (AIRC), the Telethon program (project n. D37), the ‘CNR target project on Biotechnology’, the ‘Murst CNR Biotechnology Programme Legge 95/95’ and the EC Biotech Programme. L. P. is the recipient of a fellowship from FIRC (Fondazione Italiana Ricerca sul Cancro). We would like to thank Ms Barbara Pesce and Dr Maja Adamska for excellent technical assistance and Dr Caren Gundberg, Yale University, Department of Orthopaedics for generous gift of anti-osteocalcin antibodies.

Bulfone
,
A.
,
Puelles
,
L.
,
Porteus
,
M. H.
,
Frohman
,
M. A.
,
Martin
,
G. R.
and
Rubenstein
,
J. L. R.
(
1993
).
Spatially restricted expression of Dlx-1, Dlx-2 (Tes-1), Gbx-2, and Wnt-3 in the embryonic day 12. 5 mouse forebrain defines potential transverse and longitudinal segmental boundaries
.
J. Neurosci
.
13
,
3155
3172
.
Bulfone
,
A.
,
Smiga
,
S. M.
,
Shimamura
,
K.
,
Peterson
,
A.
,
Puelles
,
L.
and
Rubenstein
,
J. L. R.
(
1995
).
T-brain-1: A homolog of Brachyury whose expression defines molecularly distinct domains within the cerebral cortex
.
Neuron
15
,
63
78
.
Buss
,
P. W.
(
1994
).
Cleft hand/foot: clinical and developmental aspects
.
J. Med. Genet
.
31
,
726
730
.
Chapman
,
G.
,
Remiszeski
,
J. L.
,
Webb
,
G. C.
,
Schultz
,
T. C.
,
Bottema
,
C. D.
and
Rathjen
,
P. D.
(
1997
).
The mouse homeobox gene, Gbx2: genomic organization and expression in pluripotent cells in vitro and in vivo
.
Genomics
46
,
223
233
.
Cohen
,
S. M.
,
Brönner
,
G.
,
Küttner
,
F.
,
Jürgens
,
G.
and
Jäckle
,
H.
(
1989
).
Distal-less encodes a homeodomain protein required for limb development in Drosophyla
.
Nature
338
,
432
434
.
Echelard
,
Y.
,
Epstein
,
D. J.
,
St-Jacques
,
B.
,
Shen
,
L.
,
Mohler
,
J.
,
McMahon
,
J. A.
and
McMahon
,
A. P.
(
1993
).
Sonic hedgehog, a member of a family of putative signaling molecules, is implicated in the regulation of CNS polarity
.
Cell
75
,
1417
1430
.
Francis-West
,
P.
,
Ladher
,
R.
,
Barlow
,
A.
and
Graveson
,
A.
(
1998
).
Signalling interactions during facial development
.
Mech. Dev
.
75
,
3
28
.
Gavrieli
,
Y.
,
Sherman
,
Y.
and
Ben-Sasson
,
S. A.
(
1992
).
Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation
.
J. Cell Biol
.
119
,
493
501
.
Hanken
,
J.
, and
Hall
,
B. K.
, (
1993
).
The Skull. Vol. 1: Development
.
Chicago
:
University of Chigago Press
.
Hadrys
,
T.
,
Braun
,
T.
,
Rinkwitz-Brandt
,
S.
,
Arnold
H.-H.
and
Bober
E.
(
1998
).
Nkx5-1 controls semicircular canal formation in the mouse inner ear
.
Development
125
,
33
39
.
Kontges
,
G.
and
Lumsden
,
A.
(
1996
).
Rhombencephalic neural crest segmentation is preserved throughout craniofacial ontogeny
.
Development
122
,
3229
3242
.
Li
,
C. W.
,
Van de Water
,
T. R.
and
Ruben
,
R. J.
(
1978
).
The fate mapping of the eleventh and twelfth day mouse otocyst
.
J. Morph
.
157
,
249
268
.
Liu
,
J. K.
,
Ghattas
,
I.
,
Liu
,
S.
,
Chen
,
S.
and
Rubenstein
,
J. L. R.
(
1997
).
Dlx genes encode DNA-binding proteins that are expressed in an overlapping and sequential pattern during basal ganglia differentiation
.
Dev. Dynamics
210
,
498
512
.
Lumsden
,
A.
,
Sprawson
,
N.
and
Graham
,
A.
(
1991
).
Segmental origin and migration of neural crest cells in the hindbrain region of the chick embryo
.
Development
113
,
1281
1291
.
McGuinness
,
T.
,
Porteus
,
M. H.
,
Smiga
,
S.
,
Bulfone
,
A.
,
Kingsley
,
C.
,
Qiu
,
M.
,
Liu
,
J. K.
,
Long
,
J. E.
,
Xu
,
D.
and
Rubenstrin
,
J. L. R.
(
1996
).
Sequence organisation and transcription of the Dlx-1 and Dlx-2 locus
.
Genomics
35
,
473
485
.
Nakamura
,
S.
,
Stock
,
D. W.
,
Wydner
,
K.
,
Zhao
,
Z.
,
Minowada
,
J.
,
Lawrence
,
J. B.
,
Weiss
,
K. M.
and
Ruddle
,
F. H.
(
1996
).
Genomic analysis of a new mammalian distal-less gene: Dlx 7
.
Genomics
38
,
314
324
.
O’Hara
,
E.
,
Cohen
,
B.
,
Cohen
,
S. M.
and
McGinnis
,
W.
(
1993
).
Distal-less ia a downstream gene of Deformed required for ventral maxillary identity
.
Development
117
,
847
856
.
Otto
,
F.
,
Thornell
,
A. P.
,
Crompton
,
T.
,
Denzel
,
A.
,
Gilmour
,
K. C.
,
Rosewell
,
I. R.
,
Stamp
,
G. W.
,
Beddington
,
R. S.
,
Mundlos
,
S
,
Olsen
,
B. R.
and others
(
1997
).
Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development
.
Cell
89
,
765
771
.
Qiu
,
M.
,
Bulfone
,
A.
,
Martinez
,
S.
,
Meneses
,
J. J.
,
Shimamura
,
K.
,
Pedersen
,
R. A.
and
Rubenstein
,
J. L. R.
(
1995
).
Null mutation of the Dlx-2 results in abnormal morphogenesis of proximal first and second branchial arch derivatives and abnormal differentiation in the forebrain
.
Genes Dev
.
9
,
2523
2538
.
Qiu
,
M.
,
Bulfone
,
A.
,
Ghattas
,
I.
,
Meneses
,
J. J.
,
Christensen
,
L.
,
Sharpe
,
P. T.
,
Presley
,
R.
,
Pedersen
,
R. A.
and
Rubenstein
,
J. L. R.
(
1997
).
Role of the Dlx homeobox genes in proximodistal patterning of the branchial arches: mutations of Dlx-1, Dlx-2, and Dlx-1 and Dlx-2 alter morphogenesis of proximal skeletal and soft tissue structures derived from the first and second arches
.
Dev. Biol
.
185
,
165
186
.
Robertson
,
E.
,
Bradley
,
A.
,
Kuehn
,
M.
and
Evans
,
M.
(
1986
).
Germ-line transmission of genes introduced into cultured pluripotent cells by retroviral vector
.
Nature
323
,
445
448
.
Robinson
,
G. W.
and
Mahon
,
K. A.
(
1994
).
Differential and overlapping expression domains of Dlx-2 and Dlx-3 suggest distinc roles for Distal-less homeobox genes in craniofacial development
.
Mech. Dev
.
48
,
199
215
.
Rubenstein
,
J. L. R.
,
Martinez
,
S.
,
Shimamura
,
K.
and
Puelles
,
L.
(
1994
).
The embryonic vertebrate forebrain: the prosomeric model
.
Science
266
,
578
580
.
Ryoo
,
H. M.
,
Hoffmann
,
H. M.
,
Beumer
,
T.
,
Frenkel
,
B.
,
Towler
,
D. A.
,
Stein
,
G. S.
,
Stein
,
J. L.
,
van Wijnen
,
A. J.
and
Lian
,
J. B.
(
1997
).
Stage-specific expression of Dlx-5 during osteoblast differentiation: involvement in regulation of osteocalcin gene expression
.
Mol. Endocrinol
.
11
,
1681
1694
.
Scherer
,
W. S.
,
Poorkaj
,
P.
,
Massa
,
H.
,
Soder
,
S.
,
Allen
,
T.
,
Nunes
,
M.
,
Geshuri
,
D.
,
Wong
,
E.
,
Belloni
,
E.
,
Little
,
S.
and others
(
1994
).
Physical mapping of the split hand/split foot locus on chromosome 7 and implication in syndromic ectrodactyly
.
Hum. Mol. Genet
.
3
,
1345
1354
.
Serbedzija
,
G. N.
,
Bronner-Fraser
,
M.
and
Fraser
,
S. E.
(
1992
).
Vital dye analysis of cranial neural crest cell migration in the mouse embryo
.
Development
116
,
297
307
.
Sham
,
M. H.
,
Vesque
,
C.
,
Nonchev
,
S.
,
Marshall
,
H.
,
Frain
,
M.
,
Gupta
,
R. D.
,
Whiting
,
J.
,
Wilkinson
,
D.
,
Charnay
,
P.
and
Krumlauf
,
R.
(
1993
).
The zinc finger gene Krox20 regulates HoxB2 (Hox2
.
8) during hindbrain segmentation. Cell
72
,
183
196
.
Simeone
,
A.
,
Acampora
,
D.
,
Mallamaci
,
A.
,
Stornaiuolo
,
A.
,
D’Apice
,
M. R.
,
Nigro
,
V.
and
Boncinelli
,
E.
(
1993
).
A vertebrate gene related to orthodenticle contains a homeodomain of the bicoid class and demarcates anterior neuroectoderm in the gastrulating mouse embryo
.
EMBO J
.
12
,
2735
2747
.
Simeone
,
A.
,
Acampora
,
D.
,
Pannese
,
M.
,
D’Esposito
,
M.
,
Stornaiuolo
,
A.
,
Gulisano
,
M.
,
Mallamaci
,
A.
,
Kastury
,
K.
,
Druck
,
T.
,
Huebner
,
K.
and
Boncinelli
,
E.
(
1994a
).
Cloning and characterization of two members of the vertebrate Dlx gene family
.
Proc. Natl. Acad. Sci. USA
91
,
2250
2254
.
Simeone
,
A.
,
D’Apice
,
M. R.
,
Nigro
,
V.
,
Casanova
,
J.
,
Graziani
,
F.
,
Acampora
,
D.
and
Avantaggiato
,
V.
(
1994b
).
Orthopedia, a novel homeobox-containing gene expressed in the developing central nervous system of both mouse and Drosophila
.
Neuron
13
,
83
101
.
Simeone
,
A.
(
1998
).
Detection of mRNA in tissue sections with radiolabelled riboprobes
. In:
In situ hybridization. A practical approach. D. G. Wilkinson (ed
.), 2nd edition, pp
69
86
,
Oxford University Press
.
Smith
,
K. K.
and
Schneider
,
R. A.
(
1998
).
Have gene knockouts caused evolutionary reversals in the mammalian first arch ?
BioEssays
20
,
245
255
.
Stock
,
D. W.
,
Ellies
,
D. L.
,
Zhao
,
Z.
,
Ekker
,
M.
,
Ruddle
,
F. H.
and
Weiss
,
K.
(
1996
).
The evolution of the vertebrate Dlx gene family. Proc. Natl. Acad. Sci. USA
93
,
10858
10863
.
Trainor
,
P. A.
and
Tam
,
P. P. L.
(
1995
).
Cranial paraxial mesoderm and neural crest cells of the mouse embryo: Co-distribution in the craniafacial mesenchyme but distinct segregation in branchial arches
.
Development
121
,
2569
2582
.
Wallin
,
J.
,
Wilting
,
H.
,
Koseki
,
R.
,
Fistsk
,
B.
,
Christ
,
B.
and
Balling
,
R.
(
1994
).
The role of Pax-1 in axial skeleton development
.
Development
120
,
1109
1121
.
Wang
,
W.
,
Van De Water
,
T.
and
Lufkin
,
T.
(
1998
).
Inner ear and maternal reproductive defects in mice lacking the Hmx3 homeobox gene
.
Development
125
,
621
634
.
Xuan
,
S.
,
Baptista
,
C. A.
,
Balas
,
G.
,
Tao
,
W.
,
Soares
,
V. C.
and
Lai
,
E.
(
1995
).
Winged helix transcription factor BF-1 is essential for the development of the vertebral hemispheres
.
Neuron
14
,
1141
1152
.
Yang
,
L.
,
Zhang
,
H.
,
Hu
,
G.
,
Wang
,
H.
,
Abateé-Shen
,
C.
and
Shen
,
M. M.
(
1998
).
An early phase of embryonic Dlx5 expression defines the rostral boundary of the neural plate
.
J. Neurosci
.
18
,
8322
8330
.
Zhao
,
G. Q.
,
Zhao
,
S.
,
Zhou
,
X.
,
Eberspaecher
,
H.
,
Solursh
,
M.
and
de Crombrugghe
,
B.
(
1994
).
rDLX, a novel distal-less-like homeoprotein, is expressed in developing cartilages and discrete neuronal tissues
.
Dev. Biol
.
164
,
37
51
.
Zhang
,
H.
,
Hu
,
G.
,
Wang
,
H.
,
Sciavolino
,
P.
,
Iler
,
N.
,
Shen
,
M. M.
and
Abate-Shen
,
C.
(
1997
).
Heterodimerization of Msx and Dlx homeoproteins results in funtional antagonism
.
Mol. Cell. Biol
.
17
,
2920
2932
.