Ulnaless (Ul), an X-ray-induced dominant mutation in mice, severely disrupts development of forearms and forelegs. The mutation maps on chromosome 2, tightly linked to the HoxD complex, a cluster of regulatory genes required for proper morphogenesis. In particular, 5′-located (posterior) Hoxd genes are involved in limb development and combined mutations within these genes result in severe alterations in appendicular skeleton. We have used several engineered alleles of the HoxD complex to genetically assess the potential linkage between these two loci. We present evidence indicating that Ulnaless is allelic to Hoxd genes. Important modifications in the expression patterns of the posterior Hoxd-12 and Hoxd-13 genes at the Ul locus suggest that Ul is a regulatory mutation that interferes with a control mechanism shared by multiple genes to coordinate Hoxd function during limb morphogenesis.

In mice, naturally occuring, or randomly induced, mutations have been an invaluable source of informations in the field of developmental genetics. While the ES cell-based approach allows the straightforward inactivation of a given gene, complex mutations such as deletions or rearrangements often illuminate pathogenic mechanisms and developmental regulatory processes of unexpected significance which would have otherwise escaped examination. There are many reported cases in vertebrates where spontaneous mutations did not lead to complete inactivation of a given gene, but rather produced more subtle changes, either in the product of the gene or in its regulation. With respect to limb development, valuable information has been gained from molecular studies of murine mutations such as limb deformity (ld) or Hypodactyly (Hd) (Kuhlman and Niswander, 1997; Haramis et al., 1995; Chan et al., 1995; Mortlock et al., 1996). Likewise, the molecular characterization of human syndromes affecting limbs, such as the Crouzon, SPD or hand-foot-genital syndromes, combined with knowledge acquired on the murine system, has led to further advances in our understanding of limb morphogenesis (Muenke and Schell, 1995; Yamaguchi and Rossant, 1995; Mortlock and Innis, 1997; Muragaki et al., 1996).

The mouse Ulnaless (Ul) mutation was generated some thirty years ago by X-ray irradiation (Morris, 1967). It is a dominant mutation affecting mostly the zeugopod (the intermediate piece of the limbs, forearms and forelegs). In Ul/+ animals, both radius and ulna in the forelimb, or tibia and fibula in the hindlimb, are strongly defective. Length reductions are accompanied by severe distal malformations (Davisson and Cattanach, 1990). The ill-formed articulations in the carpus and tarsus lead to deflected positions of the autopods (hands and feet). No obvious defects are detected in the trunk of these heterozygous animals. Due to these severe alterations of the limbs, Ul/+ males have great difficulties in breeding and homozygous animals could not be obtained by natural matings on the original genetic background. The Ulnaless mutation was mapped genetically to mouse chromosome 2 (Davisson and Cattanach, 1990), at the vicinity of the HoxD complex, a locus containing several genes of importance for limb development (Dollé and Duboule, 1989).

The HoxD complex was mapped to mouse chromosome 2D (Featherstone et al., 1988) and comprises a minimum of eight Hoxd genes that are known to play important functions in the organization of the body plan. In particular, the Hoxd-9 to Hoxd-13 genes, located at the 5′ extremity of the complex and related in sequence to the Drosophila gene AbdB (Izpisua-Belmonte et al., 1991), are essential for proper patterning and development of the limbs, the genitalia and the posterior vertebral column (Dollé et al., 1993; Fromental-Ramain et al., 1996a; Favier et al., 1995; Davis and Capecchi, 1994, 1996; Kondo et al., 1996). With respect to limb development, Hoxd gene knock-outs have revealed their important roles in the growth of both prechondrogenic condensations (Dollé et al., 1993) and bony elements (Davis et al., 1995; Zákány and Duboule, 1996). During limb development, Hoxd genes cooperate with posterior genes from the HoxA complex so that combined inactivations of paralogous members of both complexes lead to very strong phenotypic alterations. While removing the functions of group 13 genes simultaneously (Hoxd-13; Hoxa-13) prevents digit formation (Fromental-Ramain et al., 1996b), mice lacking both Hoxa-11 and Hoxd-11 functions have short and abnormal zeugopods (Davis et al., 1995). The fact that this latter phenotype is clearly reminiscent of the Ul/+ mutation further supports the involvement of Hoxd genes in this abnormal process. In addition, Peichel et al. (1996) recently reported an extensive set of mapping data showing no recombination between the Ulnaless mutation and either extremities of the HoxD complex, strengthening the hypothesis that this mutation affects one or several member(s) of this complex. However, the molecular nature of the mutation and the actual involvement of Hoxd genes in the generation of the phenotype remained to be established.

By using the Ul chromosome as well as a variety of different HoxD alleles in a genetic approach, we show that the Ulnaless mutation affects the regulation of 5′-located Hoxd genes in a complex manner. Expression of Hoxd-13 and Hoxd-12 in limbs and genitalia of Ul animals is perturbed, showing concomitant loss-of-function in digits and ectopic gain-of-function in the zeugopods. We conclude that Ul is allelic to the HoxD complex and propose that the mutation affects an important regulatory element acting upon several Hoxd genes at once. Furthermore, the Ul gain-of-function phenotype lends further support to the proposal that a functional hierarchy exists amongst Hox proteins.

Mutant lines

The Ulnaless mutant line was obtained form the Jackson laboratory (Bar Harbour, Maine), in a B6EiC3H background. The Hoxd-13St line is a null allele of the Hoxd-13 gene, produced by insertion of a selection cassette within the homeobox (Dollé et al., 1993). The HoxdDel allele is a triple loss-of-function of Hoxd-13, Hoxd-12 and Hoxd-11, due to a deletion of the Hoxd-13 to Hoxd-12 genomic locus, plus an insertion of the lacZ reporter gene within the Hoxd-11 gene (Zákány and Duboule, 1996). The HoxDRXI mutation is a small deletion located between Hoxd-13 and Hoxd-12 (Y. H. and D. D., unpublished). These chromosomes are depicted in Fig. 1 and were maintained in a C57Bl/6j×129Sv mixed genetic background. The TgH[d11/lac]Ge mice were produced by transposing a Hoxd-11/lacZ reporter transgene upstream of Hoxd-13 (van der Hoeven et al., 1996).

Fig. 1.

Different alleles from the HoxD complex used in this work. The wild-type HoxD complex is shown on the top line, from Evx-2 to Hoxd-10. The second line corresponds to the Hoxd-13St allele, an insertional mutation within Hoxd-13. The third allele (HoxDDel) is a large deletion covering from Hoxd-13 to Hoxd-12 plus a null mutation within Hoxd-11. The fourth allele (bottom) is a microdeletion between Hoxd-13 and Hoxd-12.

Fig. 1.

Different alleles from the HoxD complex used in this work. The wild-type HoxD complex is shown on the top line, from Evx-2 to Hoxd-10. The second line corresponds to the Hoxd-13St allele, an insertional mutation within Hoxd-13. The third allele (HoxDDel) is a large deletion covering from Hoxd-13 to Hoxd-12 plus a null mutation within Hoxd-11. The fourth allele (bottom) is a microdeletion between Hoxd-13 and Hoxd-12.

Skeletal analysis and whole-mount in situ hybridizations

For skeletal analysis, adult mice were killed, processed and stained with alizarin red S as previously described (Dollé et al., 1993). Mice were derived from crosses between Ul/+ females and males heterozygotes for either the Hoxd-13St or the HoxDDel mutant alleles. The Hoxd-13, Hoxd-12, Hoxd-11 and Hoxd-10 probes used in wholemount in situ hybridizations were as described previously (Dollé et al., 1991a,b; Gérard et al., 1996) and labelled with digoxigenin-11-UTP (Boehringer). Embryos were fixed overnight in 4% paraformaldehyde and hybridizations were performed according to established procedure. The staining was carried out using an alkaline-phosphatase-conjugated anti-digoxigenin antibody.

Ulnaless is a hypomorphic allele of posterior Hoxd genes

The dominant phenotype of mice heterozygous for the Ulnaless mutation was originaly described in Davisson and Cattanach, (1990; see also Peichel et al., 1996). Briefly, the zeugopods are strongly reduced in length and characteristically ill formed (Fig. 2A). In particular, the ulna, radius, tibia and fibula are severely affected. Minor but significant defects were also observed in the shape of the humerus (Fig. 2A). Homozygous animals display the same although somewhat more severe defect (Peichel et al., 1996). In contrast, no obvious alteration was found in either the skull or the axial skeleton. We observed the same set of defects on the C57Bl/6j×129Sv genetic background, with no apparent change either in expressivity or in penetrance.

Fig. 2.

Skeletal preparations of adult limbs.(A) Comparison between a wild-type (left) and Ulnaless (right) arm. Strong reductions and malformations are observed in both radius and ulna of the Ul/+ limb (arrowheads). In the left panel, the thumb (digit I) is hidden behind digit II. (B) Hands of Ul mice when combined with various HoxD alleles. Wild-type and Ul/+ hands look essentially normal whereas Ul/Hoxd-13St and Ul/HoxDDel hands have reduced or absent second phalanges on digits II and V (arrowheads), thus resembling a hypomorphic Hoxd-13−/− phenotype. Size reduction of the P2 in digit III is also visible in Ul/HoxDDel limb animals (small black arrowheads, compare with Ul/+). (C,D) Bone staining of feet from animals of similar genotypes. In Ul/Hoxd-13St and Ul/HoxDDel specimen (D), a specific deformation of digit I (arrowheads) is observed which is identical to that observed in Hoxd-13St/St animals (Dollé et al., 1993). Moreover the P2 of digit II is absent in Ul/Hoxd-13st and Ul/HoxDDel mutant feet. These defects are nevertheless not seen in Ul/+ littermates. u, ulna; r, radius; h, humerus; I, II refer to digit number (from thumb to minimus); M, metacarpus; P1 to P3 are phalanges.

Fig. 2.

Skeletal preparations of adult limbs.(A) Comparison between a wild-type (left) and Ulnaless (right) arm. Strong reductions and malformations are observed in both radius and ulna of the Ul/+ limb (arrowheads). In the left panel, the thumb (digit I) is hidden behind digit II. (B) Hands of Ul mice when combined with various HoxD alleles. Wild-type and Ul/+ hands look essentially normal whereas Ul/Hoxd-13St and Ul/HoxDDel hands have reduced or absent second phalanges on digits II and V (arrowheads), thus resembling a hypomorphic Hoxd-13−/− phenotype. Size reduction of the P2 in digit III is also visible in Ul/HoxDDel limb animals (small black arrowheads, compare with Ul/+). (C,D) Bone staining of feet from animals of similar genotypes. In Ul/Hoxd-13St and Ul/HoxDDel specimen (D), a specific deformation of digit I (arrowheads) is observed which is identical to that observed in Hoxd-13St/St animals (Dollé et al., 1993). Moreover the P2 of digit II is absent in Ul/Hoxd-13st and Ul/HoxDDel mutant feet. These defects are nevertheless not seen in Ul/+ littermates. u, ulna; r, radius; h, humerus; I, II refer to digit number (from thumb to minimus); M, metacarpus; P1 to P3 are phalanges.

The hands of Ul/+ animals did not display major alterations in the number and organisation of digits. However, while mice lacking one dose of Hoxd-13 (Hoxd-13St/+) have close-to-normal hands, mice with the Ul chromosome over the Hoxd-13St allele showed well-defined and severe alterations of the digits (Fig. 2B), similar to that expected from a loss-of-function of posterior Hoxd genes. The second phalanges of digits II and V were reduced or absent, digit I was ill formed, whereas digits III and IV were normal in appearance. This phenotype was further enhanced in mice lacking one functional complement of Hoxd-13, Hoxd-12 and Hoxd-11 (Ul/HoxDDel; Fig. 2B), with an overall stiffer appearance of the hand skeleton. In both cases, the increase in the severity of digit defects observed whenever the Ul chromosome was introduced, was similar to that of stronger Hoxd-related phenotypes. When the Ul mutation was combined with Hoxd alleles, it behaved as a Hoxd-13 hypomorphic allele. This observation was best illustrated with the help of an abnormaly prominent bony mass, located on the metatarsal bone of hindlimb digit I, and which was first observed in Hoxd-13St/St mutant animals (Dollé et al., 1993). While this defect was not detected either in Ul/+ animals, or in Hoxd-13St/+ or HoxDDel/+ specimens (e.g. Fig. 2C), Ul/Hoxd-13St as well as Ul/HoxDDel exhibited this Hoxd-13-specific deformation of digit I (Fig. 2D). Altogether, Hoxd-13 function in digits was altered by the Ul mutation in a way demonstrating that the Ul chromosome carried a hypomorphic Hoxd-13 mutation. No alteration was observed in the vertebral column, even in Ul/HoxDDel animals, indicating that no substantial loss-of-function of posterior Hoxd genes had occurred in the trunk of Ul mutant mice (not shown). In the presence of the HoxDDel allele indeed, such loss-of-function would induce transformations in the lumbosacral region (Zákány et al., 1997).

Extensive mapping of the posterior HoxD complex with a battery of probes failed to detect any genomic rearrangement that could explain this phenotype (Peichel et al., 1996; our unpublished work; see also the accompanying paper by Peichel et al., 1997). Likewise, PCR amplification of selected regions did not reveal the molecular nature of the Ul mutation. In this latter case, DNA was amplified from Ul/HoxDDel animals so that DNA sequences localised within the deficiency could only derive from the Ul chromosome (see Fig. 3). Significantly, among other DNA stretches, the Hoxd-13 coding sequence was found unaffected in Ulnaless (not shown).

Fig. 3.

Crosses used for whole-mount in situ hybridizations. The Ulnaless chromosome was segregated together with three differently labelled chromosomes. Phenotypic alterations and/or molecular typing allowed for an easy recovery of those F1 fetuses with the Ul/HoxdDel genotype. In these embryos, wild-type Hoxd-13 and Hoxd-12 are absent so that in situ hybridization with these probes reveal the hemizygous activity of the two genes from the Ul chromosome exclusively.

Fig. 3.

Crosses used for whole-mount in situ hybridizations. The Ulnaless chromosome was segregated together with three differently labelled chromosomes. Phenotypic alterations and/or molecular typing allowed for an easy recovery of those F1 fetuses with the Ul/HoxdDel genotype. In these embryos, wild-type Hoxd-13 and Hoxd-12 are absent so that in situ hybridization with these probes reveal the hemizygous activity of the two genes from the Ul chromosome exclusively.

Hoxd genes expression at the Ulnaless locus

The difficulty in obtaining Ul/Ul homozygous embryos in our genetic background prevented us from looking at Hoxd gene expression in absence of wild-type copies. To circumvent this problem, we produced Ul/HoxDDel embryos in which the Ul chromosome was brought over the HoxDDel deficiency. These animals were hemizygous for Hoxd-13 and Hoxd-12 present on the Ul chromosome. We could thus examine the expression pattern of Ulnaless Hoxd-13 and Hoxd-12 in the absence of their normal complements. In order to identify embryos of the appropriate genotypes, we crossed Ul/Hoxd-13St females with males heterozygous for two additional alleles of the HoxD complex; HoxDDel and HoxDRXI, this latter one being a small deletion between Hoxd-13 and Hoxd-12 (Y.H., J. Beckers and D.D., unpublished). Therefore, all chromosomes were labelled and could be identified either by Southern blotting or PCR analysis, the Ulnaless chromosome behaving as a wild-type complement using these markers (see Fig. 3).

Whole-mount in situ hybridizations using the Hoxd-13 probe revealed a strikingly abnormal transcript distribution in such Ul/HoxDDel fetuses, even though strong signals were detected in the four developing limbs (Fig. 4A). In addition, the expected robust expression in the genital eminence (Dollé et al., 1991) was hardly, if at all, detected in Ul mutant animals (Fig. 4B). A closer examination of both forelimbs and hindlimbs (Fig. 4C,D; respectively) clearly defined two distinct types of alterations in Hoxd-13 expression. Firstly, a substantial decrease in transcript accumulation was observed over the tips of developing digits (Fig. 4C,D; white arrowheads). However, this decrease was not uniform throughout all digits because clear signals were recovered in the primordium of digits III and IV (Fig. 4C; white arrow). Secondly, a strong ectopic domain was detected proximal to the normal Hoxd-13 expression domain (Fig. 4C,D; black arrowheads). This ectopic patch of expression extended from the proximal future carpus over the entire zeugopods, with a pronounced posterior tendency. The combination of these two traits gave rise to a novel Hoxd-13 expression pattern, resulting from a concomitant loss-of-function in digits and gain-of-function in zeugopods (e.g. Fig. 4D, right panel).

Fig. 4.

Whole-mount in situ hybridizations of Hoxd-13 and Hoxd-12 from the Ulnaless chromosome.(A) Comparison of wild-type (left) and Ul/HoxDel (right) 11.5-day-old fetuses. Hoxd-13 transcripts are produced in both cases but with different distributions in the limbs and genitalia. (B) Magnification of the external genital area of the same two animals showing the distal expression of Hoxd-13 in the genital eminence and the very weak signal obtained from the Ul chromosome. (C,D) Loss- and gain-of-function of Hoxd-13 in developing forelimbs (C) and hindlimbs (D) of Ul animals. The gain-of-function in the zeugopod is already clearly visible on the Ul/+ genotype (black arrowheads) whereas the loss-of-function of this gene in the presumptive digit area (white arrowheads) can only be seen in absence of the wild-type copy, in both hands (C) and feet (D). This loss-of-function affects primarily anterior digits and residual expression can be seen in posterior digits (white arrows in C and E). (E) The expression of Hoxd-12 in the same genotypes reveals similar alterations in the pattern, namely a down-regulation of expression in anterior digits (white arrowhead) as well as a strong ectopic domain in the zeugopod (black arrowheads). In addition, a small but consistent ectopic domain was found in the stylopod as well (small black arrowheads).

Fig. 4.

Whole-mount in situ hybridizations of Hoxd-13 and Hoxd-12 from the Ulnaless chromosome.(A) Comparison of wild-type (left) and Ul/HoxDel (right) 11.5-day-old fetuses. Hoxd-13 transcripts are produced in both cases but with different distributions in the limbs and genitalia. (B) Magnification of the external genital area of the same two animals showing the distal expression of Hoxd-13 in the genital eminence and the very weak signal obtained from the Ul chromosome. (C,D) Loss- and gain-of-function of Hoxd-13 in developing forelimbs (C) and hindlimbs (D) of Ul animals. The gain-of-function in the zeugopod is already clearly visible on the Ul/+ genotype (black arrowheads) whereas the loss-of-function of this gene in the presumptive digit area (white arrowheads) can only be seen in absence of the wild-type copy, in both hands (C) and feet (D). This loss-of-function affects primarily anterior digits and residual expression can be seen in posterior digits (white arrows in C and E). (E) The expression of Hoxd-12 in the same genotypes reveals similar alterations in the pattern, namely a down-regulation of expression in anterior digits (white arrowhead) as well as a strong ectopic domain in the zeugopod (black arrowheads). In addition, a small but consistent ectopic domain was found in the stylopod as well (small black arrowheads).

The analysis of Hoxd-12 expression in the same genetic con-figuration gave a similar picture. While Ul/+ animals already showed a gain-of-function in the zeugopods that overlapped with that of Hoxd-13 (Fig. 4E; black arrowheads), the removal of the wild-type copy of Hoxd-12 (Ul/HoxDDel) indicated that a loss-of-function had occurred in digits (white arrowhead). However, as for Hoxd-13, this decrease in Hoxd-12 transcript accumulation was not complete and primarily concerned digits II and V. Furthermore, an additional ectopic domain was detected more proximal, in the presumptive stylopods (Fig. 4E; short arrowheads), a domain that overlapped the expression domains of more anterior Hoxd genes (see Fig. 5). Altogether, these ectopic Hoxd-13 and Hoxd-12 expression domains perfectly matched the presumptive areas where the Ul mutation generates abnormal skeletal development, i.e. a hypomorphic Hoxd-13 recessive phenotype in the digits, with a preference for digits II and V, together with a fully penetrant dominant phenotype in the zeugopods. Other minor traits such as e.g. the fusion of small bones in the carpus were scored and were also correlated with the ectopic expression of either Hoxd-13 or Hoxd-12.

Fig. 5.

Whole-mount in situ hybridizations of Hoxd-11 (A) and Hoxd-10 (B) on E11.5 fetal forelimbs of different genotypes. (A) Hoxd-11 probe hybridized to either Ul/+ (left) or Ul/HoxDel (right). A robust expression of Hoxd-11 is still observed in both digit and zeugopod domains (large arrowhead). (B) Hoxd-10 probe hybridized with either wild-type (left) or Ul/HoxDel (right) embryos. Expression patterns are rather comparable between the two genotypes. In all cases, a more proximal expression domain is detected (small arrowhead), which is identical to the ectopic domain seen with Hoxd-12 in Ul mutant limbs (see Fig. 4E, same arrowhead). This domain is part of the wild-type Hoxd-10 expression pattern (left panel in B). I to V refer to digit number, from anterior to posterior.

Fig. 5.

Whole-mount in situ hybridizations of Hoxd-11 (A) and Hoxd-10 (B) on E11.5 fetal forelimbs of different genotypes. (A) Hoxd-11 probe hybridized to either Ul/+ (left) or Ul/HoxDel (right). A robust expression of Hoxd-11 is still observed in both digit and zeugopod domains (large arrowhead). (B) Hoxd-10 probe hybridized with either wild-type (left) or Ul/HoxDel (right) embryos. Expression patterns are rather comparable between the two genotypes. In all cases, a more proximal expression domain is detected (small arrowhead), which is identical to the ectopic domain seen with Hoxd-12 in Ul mutant limbs (see Fig. 4E, same arrowhead). This domain is part of the wild-type Hoxd-10 expression pattern (left panel in B). I to V refer to digit number, from anterior to posterior.

Since the combined loss-of-functions of group 11 genes affect the zeugopod (Davis et al., 1995) and because Hoxd-10 is also expressed there, the transcript domains of Hoxd-11 and Hoxd-10 were also analysed in the Ulnaless mice. Expression of Hoxd-11 in Ul/+ limbs was found close to normal (Fig. 5A, left), with perhaps a slight down-regulation. A similar expression was observed in Ul/HoxDDel animals (Fig. 5A, right), which are unable to produce a functional Hoxd-11 protein from the HoxDDel chromosome (Zákány and Duboule, 1996). Expression of Hoxd-10 in Ul/HoxDDel fetal limbs, which thus contain at least one functional copy of the gene, was undistinguishable from that seen in control mice (Fig. 5B, compare left and right panels). In both cases, however, the detection of RNAs transcribed from the HoxDDel chromosome may have obscured the detection of a slight partial loss-of-function. These experiments nevertheless demonstrated that a large amount of both Hoxd-11 and Hoxd-10 transcripts were present in developing Ul/+ limbs.

Is Ulnaless allelic to HoxD?

Three sets of evidence strongly suggest that the Ulnaless mutation is allelic to the HoxD complex: (1) the absence of recombination between the two loci, which places them within an approx. 250 kb interval, i.e. an interval only slightly larger than the HoxD complex itself (Peichel et al., 1996); (2) the recessive digit phenotypic alterations in Ul mice, revealed in complementation studies with loss-of-function alleles of Hoxd genes and (3) the clear ectopic expression of two members of the posterior HoxD complex, when present on the Ulnaless chromosome.

Crosses involving the Ul locus as well as several HoxD alleles, previously produced through the ES cell technology, revealed that Ulnaless mice have a clear hypomorph Hoxd-13 function. This was most evident when looking at hindlimb digit I of transheterozygous Ul/Hoxd-13 mice, which unambigously showed an alteration detected only in other Hoxd-13 mutant alleles (Dollé et al., 1993; Zákány and Duboule, 1996). This was subsequently confirmed by in situ hybridizations showing that a decrease in both Hoxd-13 and Hoxd-12 expression had occurred in the most distal and anterior domains of the autopods. Expression was nevertheless detectable in digits III and IV, i.e. in those digits that have the strongest expression of these genes in wild-type fetuses (unpublished). We thus conclude that the partial loss-of-function affects the entire distal expression domain, but is less visible in its central part. The fact that some expression was left in digits is also consistent with the phenotype observed in homozygous Ul/Ul animals (Peichel et al., 1996), as such mice have a digit phenotype weaker than that of complete Hoxd-13 loss-of-function (Dollé et al., 1993).

The concomitant down-regulation of several Hoxd genes indicates, however, that the Ul mutation may not be allelic to one particular gene. Instead, the mutation may affect a supragenic mechanism that controls the expression in digits of many genes at once. This is in agreement with the proposal that a unique enhancer element might be responsible for the expression of posterior Hoxd genes in presumptive digits (van der Hoeven et al., 1996). In this view, the Ul mutation would be identified as a regulatory mutation and would thus be allelic to an extended part of the HoxD complex as it would interfere with a shared multigenic control mechanism. Interestingly, this putative mechanism was proposed to be involved in the development of the genital eminence as well, due to the coexpression of the same Hoxd genes in both genital and distal limb buds in various transgenic configurations (van der Hoeven et al., 1996). This proposal gains further support after examination of Ulnaless mice, since expression was importantly downregulated in digits and genital eminence, simultaneously. Accordingly, penian bones (baculum) of Ul/HoxDDel mice were significantly smaller than those of wild-type mice (not shown), further suggesting that digits and external genitalia may share some important regulatory controls.

Hoxd-13 gain-of-function

The most dramatic feature of Ul mice was the extreme reduction of their zeugopods which thus resembled those obtained when removing both Hoxd-11 and Hoxa-11 functions (Davis et al., 1995). Upon in situ analyses, a correlation was established between the Ul-induced defect and ectopic expression of Hoxd-13/Hoxd-12 in presumptive zeugopods (Fig. 6C, arrows), suggesting that a causal relationship may exist between a Hoxd-13/Hoxd-12 gain-of-function on the one hand, and a global Hox group 11 loss-of-function (Fig. 6C,D), on the other hand. Interestingly, such a correspondence had previously been noticed in a different context, i.e. when a Hoxd-11 transgene was transposed next to Hoxd-13 (the TgH[d11/lac]Ge mice; van der Hoeven et al., 1996). In such a transposition, the Hoxd-11 transgene was able to up-regulate Hoxd-13 in an ectopic expression domain located within the area of group 11 gene zeugopod domain (Fig. 6B, black domain). In this case, ectopic HOXD-13 protein led to truncation of the distal part of the ulna (Fig. 6B, arrows). There is therefore a robust correlation between the extents of Hoxd-13 ectopic expression domains in both TgH[d11/lac]Ge and Ulnaless mice and the localization and importance of the defects observed in the respective zeugopods (Fig. 6B,C). This strongly suggests that both phenotypes derive from a Hoxd-13 gain-of-function which, in turn, induces a phenocopy of group 11 loss-of-function phenotypes. Furthermore, group 13 Hox gain-of-function approaches in chick have led to similar phenotypic alterations (Goff and Tabin, 1997; Yokouchi et al., 1995).

Fig. 6.

Schematic representation of the relationship between Hoxd-13 gain-of-function in either the TgH[d11/lac]Ge (B) or Ul mice (C) and the phenotypic alterations of the zeugopod bones (right column, in black), and comparison with the wild-type (A) or Hox group 11 loss-of-function phenotype (D). Wherever Hoxd-13 is ectopically expressed, a truncation of the zeugopod is observed which resembles that seen when removing multiple Hox group 11 doses, thus suggesting that ectopic Hoxd-13 antagonizes group 11 functions (see the Discussion).

Fig. 6.

Schematic representation of the relationship between Hoxd-13 gain-of-function in either the TgH[d11/lac]Ge (B) or Ul mice (C) and the phenotypic alterations of the zeugopod bones (right column, in black), and comparison with the wild-type (A) or Hox group 11 loss-of-function phenotype (D). Wherever Hoxd-13 is ectopically expressed, a truncation of the zeugopod is observed which resembles that seen when removing multiple Hox group 11 doses, thus suggesting that ectopic Hoxd-13 antagonizes group 11 functions (see the Discussion).

Another evidence indicating that a Ul-related zeugopod phenotype could derive from the misexpression of Hoxd-13 was obtained when the Hoxd-13 function was further removed from the TgH[d11/lac]Ge chromosome (Fig. 7), through a targeted deletion that thus eliminated both normal and ectopic Hoxd-13 expression domains (Zákány and Duboule, 1996; Fig. 7C; arrow). In the original allele, the Hoxd-11 transgene induced an up-regulation of Hoxd-13 in the forearm (arrow) leading to an ectopic domain and concomitant alteration of the ulna (Fig. 7; black dot between Evx-2 and Hoxd-13). However, when Hoxd-13 was removed, a rescue was observed in the ulna, even though the transgene insertion was still there (data not shown). This result unequivocally demonstrated that the ulna phenotype scored in TgH[d11/lac]Ge mice was due to Hoxd-13 misexpression and hence suggests that a similar mechanism is at work in the Ul mutation.

Fig. 7.

Scheme illustrating that ectopic Hoxd-13 is indeed responsible for the zeugopod phenotype (A). In presence of Hoxd-13, the insertion of the Hoxd-11/lacZ transgene upstream of Hoxd-13 leads to ectopic expression and truncation of the ulna (B). If Hoxd-13 is removed from this chromosome (C), the truncation is lost and the ulna is back to a wild-type morphology, even though both copies of Hoxd-11 are inactivated.

Fig. 7.

Scheme illustrating that ectopic Hoxd-13 is indeed responsible for the zeugopod phenotype (A). In presence of Hoxd-13, the insertion of the Hoxd-11/lacZ transgene upstream of Hoxd-13 leads to ectopic expression and truncation of the ulna (B). If Hoxd-13 is removed from this chromosome (C), the truncation is lost and the ulna is back to a wild-type morphology, even though both copies of Hoxd-11 are inactivated.

Prevalence of posterior Hoxd genes

A striking similarity exists between the Ul/+ phenotype and that of animals double homozygous for both Hoxd-11 and Hoxa-11 inactivations (Fig. 6C,D; Davis et al., 1995). However, the dominant nature of the Ul mutation as well as the presence of at least one dose of normally distributed Hoxd-11 and Hoxd-10 transcripts in Ul/+ mice makes it unlikely that the phenotype entirely derives from a global down-regulation of group 11 gene transcription. This raises the possibility that the Hoxd-13 gain-of-function induces a concurrent loss-of-function of group 11 genes within the Hoxd-13 ectopic domain, without totally switching off their transcriptions, a phenomenon previously referred to as ‘posterior prevalence’ (Duboule, 1991). The presence of the Hoxd-13 protein may, for example, antagonize the function of group 11 proteins through protein-protein interactions or competition for target binding sites (Duboule and Morata, 1994).

An indirect demonstration that ectopic Hoxd-13 acts through preventing group 11 proteins achieving their functions was obtained with the TgH[d-11/lac]Ge allele (Figs 6B, 7B). When present in one copy, this configuration was inactive, i.e. the gain-of-function of Hoxd-13 (arrow) was not sufficient to induce the ulna phenotype. However, two copies generated this phenotype with a full penetrance (Fig. 6B; van der Hoeven et al., 1996). Conversely, Davis et al. (1995) showed that removing two copies of group 11 genes (either from Hoxd-11 or Hoxa-11, or mixed) never produced strong ulna phenotypes, this latter trait appearing only when three copies were removed. To show that both situations involved the same deficiency of group 11 functions (regardless of which proteins were present or absent), we introduced one TgH[d-11/lac]Ge chromosome in mice transheterozygous for both the Hoxd-11 (Favier et al., 1995) and Hoxa-11 (Small and Potter, 1993) null alleles and recovered animals with an altered ulna (F. van der Hoeven, B. Favier, S. Potter and D. D., unpublished). Thus, while neither heterozygous TgH[d-11/lac]Ge nor Hoxd-11+/; Hoxa-11+/ animals have a defective phenotype, the combination of the three alleles induced the expected alteration (though with moderate penetrance).

Therefore, the genetic and molecular analyses of the Ul and TgH[d-11/lac]Ge gain-of-function Hoxd-13 alleles provides additional evidence supporting the existence of a functional hierarchy between Hox gene products. In Ulnaless mice, a similar situation might occur with the clear but restricted gain-of-function of Hoxd-12 in the stylopod. An ectopic domain was detected in foetal stylopods, whereas a reduction in the length of the humerus was observed in Ul/+ adult animals. Interestingly, this proximal domain was shown to be part of the normal transcript domains of more 3′-located genes such as Hoxd-10, which suggests that the overall regulatory control of Hoxd-13 and Hoxd-12, in Ulnaless animals, had been shifted towards a more ‘proximal’ type of regulation. Furthermore, the alteration of the Ul/+ humerus resembled part of the Hoxd-9 homo-zygous mutant phenotype (Fromental-Ramain, 1996a) suggesting that ectopic Hoxd-12 may, in this case, induce a group 9 loss-of-function phenotype. However, the observed slight reduction of Hoxd-11 transcription in Ulnaless mice, as well as a transcriptional down-regulation of Hoxa-11 (see the accompanying paper by Peichel et al., 1997), suggests the possibility that the phenotype results from a combined effect of both posterior prevalence and a cross-regulatory negative transcriptional control by ectopic Hoxd-13 protein.

Is Ulnaless a regulatory mutation?

While this set of data provides a temptative mechanistic explanation for the Ulnaless-related phenotypic alterations, it does not reveal the molecular nature of the mutation. However, Ul is a strong candidate for a regulatory mutation affecting several HoxD genes at once. In contrast to other reported naturally occurring mutations within posterior Hox genes, such as SPD, Hd or HFG (Mortlock and Innis, 1997; Mortlock et al., 1996; Muragaki et al., 1996), Ul probably does not interfere with one particular coding region. A plausible hypothesis involves the alteration (deletion, inversion) of one major regulatory element necessary for limb expression. Hoxd gene expression in limbs is a multiphasic process and is thought to involve discrete regulations, at least for the distal (digits) and proximal (forearm) segments. It is possible that a distance-dependent competition between these two elements impose different patterns to various genes depending upon their positions in the complex. While Hoxd-13 and Hoxd-12 are normally under strong influence of a ‘distal’ element, the rearrangement of this element or deletion thereof may allow a ‘proximal’ element to take over transcriptional controls of these genes turning them into more ‘proximal’ genes and forcing their expressions in zeugopods. In such a case, the molecular characterization of this mutation will help to understand complex regulatory mechanisms involving the coordinate action of several genes.

We would like to thank K. Peichel and T. Vogt for sharing results prior to publication, F. van der Hoeven for his help and colleagues from the laboratory for comments, suggestions and sharing reagents. The laboratory is supported by funds from the Canton de Genève, the Swiss National Research Fund, the Human Frontier Program and the Claraz and Latsis foundations.

Chan
,
D. C. A.
,
Wynshaw-Boris
,
A.
and
Leder
,
P.
(
1995
)
Formin isoforms are differentially expressed in the mouse embryo and are required for normal expression of fgf-4 and shh in the limb bud
.
Development
121
,
3151
3161
.
Davis
,
A. P.
and
Capecchi
,
M. R.
(
1994
).
Axial homeosis and appendicular skeleton defects in mice with a targeted disruption of Hoxd-11
.
Development
120
,
2187
2198
.
Davis
,
A. P.
and
Capecchi
,
M. R.
(
1996
).
A mutational analysis of the 5′ HoxD genes: Dissection of genetic interactions during limb development in the mouse
.
Development
122
,
1175
1185
.
Davis
,
A. P.
,
Witte
,
D. P.
,
Hsieh-Li
,
H. M.
,
Potter
,
S. S.
and
Capecchi
,
M. R.
(
1995
).
Absence of radius and ulna in mice lacking Hoxa-11 and Hoxd-11
.
Nature
375
,
791
795
.
Davisson
,
M. T.
and
Cattanach
,
B. M.
(
1990
).
The mouse mutation Ulnaless on chromosome 2
.
J. Hered
.
81
,
151
153
.
Dollé
,
P.
,
Dierich
,
A.
,
LeMeur
,
M.
,
Schimmang
,
T.
,
Schuhbaur
,
B.
,
Chambon
,
P.
and
Duboule
,
D.
(
1993
).
Disruption of the Hoxd-13 gene induces localized heterochrony leading to mice with neotenic limbs
.
Cell
75
,
431
441
.
Dollé
,
P.
and
Duboule
,
D.
(
1989
).
wo gene members of the murine HOX-5 complex show regional and cell-type specific expression in developing limbs and gonads
.
EMBO J
.
8
,
1507
15
.
Dollé
,
P.
,
Izpisúa-Belmonte
,
J.-C.
,
Tickle
,
C.
,
Brown
,
J.
and
Duboule
,
D.
(
1991a
).
Hox-4 genes and the morphogenesis of mammalian genitalia
.
Genes Dev
.
5
,
1767
1776
.
Dollé
,
P.
,
Izpisúa-Belmonte
,
J.-C.
,
Boncinelli
,
E.
and
Duboule
,
D.
(
1991b
).
The Hox-4.8 gene is localized at the 5′ extremity of the Hox-4 complex and is expressed in the most posterior parts of the body during development
.
Mech. Dev
.
36
,
3
13
.
Duboule
,
D.
(
1991
).
Patterning in the vertebrate limb
.
Curr. Opin. Genet. Dev
.
1
,
211
6
.
Duboule
,
D.
and
Morata
,
G.
(
1994
).
Colinearity and functional hierarchy among genes of the homeotic complexes
.
Trends Genet
.
10
,
358
364
.
Favier
,
B.
,
Le Meur
,
M.
,
Chambon
,
P.
and
Dollé
,
P.
(
1995
).
Axial skeleton homeosis and forelimb malformations in Hoxd-11 mutant mice
.
Proc. Natl. Acad. Sci. USA
92
,
310
314
.
Featherstone
,
M. S.
,
Baron
,
A.
,
Gaunt
,
S. J.
,
Mattei
,
M. G.
and
Duboule
,
D.
(
1988
).
Hox-5.1 defines a homeobox-containing gene locus on mouse chromosome 2
.
Proc. Natl. Acad. Sci. USA
85
,
4760
4764
.
Fromental-Ramain
,
C.
,
Warot
,
X.
,
Lakkaraju
,
S.
,
Favier
,
B.
,
Haack
,
H.
,
Birling
,
C.
,
Dierich
,
A.
,
Dollé
,
P.
and
Chambon
,
P.
(
1996a
).
Specific and redundant functions of the paralogous Hoxa-9 and Hoxd-9 genes in forelimb and axial skeleton patterning
.
Development
122
,
461
472
.
Fromental-Ramain
,
C.
,
Warot
,
X.
,
Messadecq
,
N.
,
Lemeur
,
M.
,
Dollé
,
P.
and
Chambon
,
P.
(
1996b
).
Hoxa-13 and Hoxd-13 play a crucial role in the patterning of the limb autopod
.
Development
122
,
2997
3011
.
Gérard
,
M.
,
Chen
,
J.-Y.
,
Gronemeyer
,
H.
,
Chambon
,
P.
,
Duboule
,
D.
and
Zákány
,
J.
(
1996
)
In vivo targeting of a regulatory element required for positioning the Hoxd-11 and Hoxd-10 expression boundaries
.
Genes Dev
.
18
,
2326
2334
.
Goff
,
D. J.
and
Tabin
,
C. J.
(
1997
).
Analysis of Hoxd-13 and Hoxd-11 misexpression in chicken limb buds reveals that Hox genes affect both bone condensation and growth
.
Development
124
,
627
636
.
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
.
Izpisúa-Belmonte
,
J.-C.
,
Falkenstein
,
H.
,
Dollé
,
P.
,
Renucci
,
A.
and
Duboule
,
D.
(
1991
).
Murine genes related to the Drosophila AbdB homeotic genes are sequentially expressed during development of the posterior part of the body
.
EMBO J
.
10
,
2279
2289
.
Kondo
,
T.
,
Dollé
,
P.
,
Zákány
,
J.
and
Duboule
,
D.
(
1996
).
Function of posterior HoxD genes in the morphogenesis of the anal sphincter
.
Development
122
,
2651
2659
.
Kuhlman
,
J.
and
Niswander
,
L.
(
1997
).
Limb deformity proteins: role in mesodermal induction of the apical ectodermal ridge
.
Development
124
,
133
139
.
Morris
,
T.
(
1967
).
New mutants
.
Mouse News Lett
.
36
,
34
.
Mortlock
,
D. P.
and
Innis
,
J. W.
(
1997
).
Mutation of HOXA13 in hand-foot-genital syndrome
.
Nature Genet
.
15
,
179
180
.
Mortlock
,
D. P.
,
Post
,
L. C.
and
Innis
,
J. W.
(
1996
).
The molecular basis of hypodactyly (Hd): A deletion in Hoxa13 leads to arrest of digital arch formation
.
Nature Genet
.
13
,
284
289
.
Muenke
,
M.
and
Schell
,
U.
(
1995
).
Fibroblast-growth-factor receptor mutations in human skeletal disorders
.
Trends Genet
.
11
,
308
313
.
Muragaki
,
Y.
,
Mundlos
,
S.
,
Upton
,
J.
and
Olsen
,
B. R.
(
1996
).
Altered growth and branching patterns in synpolydactyly caused by mutations in HOXD13
.
Science
272
,
548
551
.
Peichel
,
C. L.
,
Abbott
,
C. M.
and
Vogt
,
T. F.
(
1996
).
Genetic and physical mapping of the mouse Ulnaless locus
.
Genetics
144
,
1757
1767
.
Peichel
,
C. L.
,
Prabhakaran
,
B.
and
Vogt
,
T. F.
(
1997
).
The mouse Ulnaless mutation deregulates posterior Hoxd gene colinearity and alters appendicular patterning
.
Development
124
,
3481
3492
.
Small
,
K. M.
and
Potter
,
S. S.
(
1993
).
Homeotic transformation and limb defects in Hoxa-11 mutant mice
.
Genes Dev
.
7
,
2318
2328
.
van der Hoeven
,
F.
,
Zákány
,
J.
and
Duboule
,
D.
(
1996
).
Gene transpositions in the HoxD complex reveal a hierarchy of regulatory controls
.
Cell
85
,
1025
1035
.
Woychik
,
R. P.
,
Maas
,
R. L.
,
Zeller
,
R.
,
Vogt
,
T. F.
and
Leder
,
P.
(
1990
)
‘Formins’: Proteins deduced from the alternative transcripts of the limb deformity gene
.
Nature
346
,
850
853
.
Yamaguchi
,
T. P.
and
Rossant
,
J.
(
1995
).
Fibroblast growth factors in mammalian development
.
Cur. Opin. Gen. and Dev
.
5
,
485
491
.
Yokouchi
,
Y.
,
Nakazato
,
S.
,
Yamamoto
,
M.
,
Goto
,
Y.
,
Kameda
,
T.
,
Iba
,
H.
and
Kuroiwa
,
A.
(
1995
).
Misexpression of Hoxa-13 induces cartilage homeotic transformation and changes cell adhesiveness in chick limb buds
.
Genes Dev
.
9
,
2509
2522
.
Zákány
,
J.
and
Duboule
,
D.
(
1996
).
Synpolydactyly in mice with a targeted deficiency in the HoxD complex
.
Nature
384
,
69
71
.
Zákány
,
J.
,
Gérard
,
M.
,
Favier
,
B.
and
Duboule
,
D.
(
1997
)
Deletion of a HoxD enhancer induces transcriptional heterochrony leading to transposition of the sacrum
.
EMBO J., in press
.