|Home | About | Journals | Submit | Contact Us | Français|
Vertebrate Hox genes regulate many aspects of embryonic body plan development and patterning. In particular, Hox genes have been shown to regulate regional patterning of the axial and appendicular skeleton and of the central nervous system. We have identified patterning defects resulting from the targeted mutation of Hoxc10, a member of the Hox10 paralogous family. Hoxc10 mutant mice have skeletal transformations in thoracic, lumbar, and sacral vertebrae and in the pelvis, along with alterations in the bones and ligaments of the hindlimbs. These results suggest that Hoxc10, along with other members of the Hox10 paralogous gene family, regulates vertebral identity at the transition from thoracic to lumbar and lumbar to sacral regions. Our results also suggest a general role for Hoxc10 in regulating chondrogenesis and osteogenesis in the hindlimb, along with a specific role in shaping femoral architecture. In addition, mutant mice have a reduction in lumbar motor neurons and a change in locomotor behavior. These results suggest a role for Hoxc10 in generating or maintaining the normal complement of lumbar motor neurons.
The Hox genes encode evolutionarily conserved transcription factors with Antennepedia class homeodomains. Mammals have 39 Hox genes in four conserved clusters, HoxA, B, C, and D, each located on a different chromosome. The clusters have been divided into 13 paralogous groups according to sequence similarities and position within the complex 1. The Hox genes are expressed in a temporal and spatial order from 3´ paralogous group 1 activated the earliest and with the most rostral limit of expression to 5´ paralogous group 13 activated last and most posterior 2.
Vertebrate Hox genes have broad expression domains in developing embryos encompassing the paraxial mesoderm and neural tube, with more 5´ AbdB-related Hox genes also expressed in the developing limbs. A large number of studies have examined the contributions of these genes to embryonic body plan development and patterning. These studies have shown that Hox genes can play roles in patterning the axial and appendicular skeleton, the nervous system, and the urogenital system 3-9. Most of the Hox genes have been shown to regulate aspects of axial skeletal patterning, with mutations causing homeotic transformations of the vertebrae. These studies have suggested that Hox genes are particularly important in defining the regions of the axial skeleton and regulating the identity of the vertebrae found in these different regions.
The vertebrate axial skeleton is composed of cervical, thoracic, lumbar, sacral and caudal regions, with vertebrae in each region having a characteristic shape and function. In C57Bl/6 mice, there are 30 precaudal vertebrae distributed into the four regions. There are seven cervical vertebrae, thirteen thoracic vertebrae, six lumbar vertebrae, and four sacral vertebrae 10-12. Mutations in Hox genes typically do not affect the total number of precaudal vertebrae, but do affect the distribution of vertebrae in the different regions. Members of the Hox5 and Hox6 paralogous groups appear to regulate the identity of cervical and thoracic vertebrae, members of the Hox9 and Hox10 families regulate the thoracic and lumbar vertebrae, and members of the Hox10 and Hox11 families regulate lumbar and sacral identities 13. Redundancy between members of paralogous gene families is also evident; in combinatorial mutants that eliminate the function of two or more paralogous genes synergistic phenotypes are always observed 8, 13-16).
The Hox10 gene family, comprised of three genes, Hoxa10, Hoxc10, and Hoxd10, is collectively required to pattern the lumbar vertebrae. Mutations in Hoxa10 produce a partial or full transformation of the first lumbar vertebra into a thoracic phenotype, typified by the presence of an extra rib 8, while ectopic expression of Hoxa10 in presomitic mesoderm deletes all ribs 17. Combinatorial mutations in Hoxa10 and Hoxd10 also produce additional ribs, indicative of lumbar to thoracic transformations, and show sacral to lumbar transformations 8. Combined mutations in all three Hox10 genes produce animals in which both lumbar and sacral vertebrae are transformed to a thoracic identity, with vertebrae at thoracic, lumbar, and sacral levels all appearing to have associated ribs 16.
Phenotypes in Hox10 mutants are not restricted to the axial skeleton, however. Hoxa10 mutants have urogenital phenotypes including cryptorchidism in males and homeotic transformations of the uterus in females 4, 18. Both Hoxa10 and Hoxd10 mutants have hindlimb skeletal alterations 4, 5, 7, 19 and Hoxd10 mice have discernible alterations in the central and peripheral nervous system providing hindlimb innervation 7-9, 20. These observations suggest widespread contributions of the Hox10 genes to embryonic patterning.
Despite substantial examination of Hoxa10 and Hoxd10 mutants alone and in combination, there are relatively few studies examining individual phenotypes in Hoxc10 mutants. Combinatorial studies suggest that loss of Hoxc10 contributes to axial skeletal phenotypes 16 and a recent study of the nervous system shows a disruption in lumbar motor pool identity in embryonic Hoxc10/Hoxd10 double mutants 20. In the current study, we examined skeletal and nervous system phenotypes resulting solely from disruption of Hoxc10. Unlike Hoxa10 and Hoxd10 mutant mice, all Hoxc10 mutant mice are viable and fertile. In homozygous mutant mice, skeletal transformations are observed in thoracic, lumbar, and sacral vertebrae and in the pelvis, as well as in the bones and ligaments of the hindlimb. Heterozygous mice show intermediate forms and penetrance of the axial skeletal phenotype, suggesting dosage dependence. In addition, the number of lumbar motor neurons is significantly reduced, resulting in significant changes in locomotor behavior.
A replacement targeting vector was constructed from 129Sv mouse genomic clones for the Hoxc10 gene that were isolated from a CC1.2 embryonic stem (ES) cell-derived library 3, 21. A 157 bp deletion flanked by PlfM1 and BamH1 restriction sites was made in the second exon to remove the last 50 amino acids of the homeodomain, and the deleted sequence was replaced with an in-frame fusion of a lacZ gene with SV40 polyadenylation signal 22 followed by a polII promoter-driven neomycin resistance gene cassette KT3NP4 (neoR, 23; see Figure Figure1).1). The lacZ/neoR sequences were flanked by 8.7 and 3.1 kb of 5´and 3´genomic sequence, respectively, resulting in 11.8 kb of total homology. The replacement sequence was assembled in the KSTK9XbaTK2-2 thymidine kinase gene plasmid vector 24 resulting in a 26.3 kb targeting vector that was linearized for electroporation by digestion with PvuII. Junctions of individually cloned pieces were verified by sequencing across the restriction sites and into the neoR and LacZ genes.
The targeting vector was electroporated into CC1.2 ES cells and the cells were subjected to positive-negative selection and cloned as previously described 25. The cell lines were analyzed for targeting events with the following probes: pro 5, a 1.1 kb EcoRI - BamHI 5´ flanking probe, pro 4, a 2.3 kb BglII - BamHI 3´ flanking probe (see Fig. Fig.1),1), and neoR and LacZ specific internal probes. The primary screen was carried out with a BamHI digestion of DNA from each cell line probed with pro 4, which reveals a 6.5 kb wild-type and an 8.9 kb recombinant fragment. Positive cell lines were confirmed with a XhoI - SphI digest that results in a 18.5 kb wild-type and 13 kb recombinant band with pro 4, and 12.7 kb and 13 kb fragments with neoR probe hybridization. The 5´ flanking probe, pro 5, detects a 11.5 kb wild-type and 18.8 kb mutant fragment in ScaI digested DNA, the latter of which also hybridizes to neoRprobe. Pro 5 also shows a shift from a 25 kb wild-type to a 17 kb fragment in a MluI - SpeI digestion, which, in addition to a 12.3 kb fragment, is detectable with the LacZ probe. Full targeting and analysis strategy is provided in Supplemental Figure 1. Ten out of 106 cell lines had the desired targeting event. One of these also had a random integration detected as an unexpected additional band with internal probes; this line was subsequently discarded. Two of the correctly targeted cell lines were injected into blastocysts that were implanted into pseudopregnant females. The cell line 1d-8 produced 13 male chimeras. These were mated to C57Bl/6J females (Jackson Laboratory) and produced 249 offspring, one of which had the agouti coat color marker for germ-line transmission. Genotypic analysis of tail DNA isolated from this male proved that the mouse carried the targeted mutation, and it was mated to C57Bl/6J females to propagate the line. Additional backcrosses were made to bring the line to a uniform C57Bl/6J background. Once the backcrossing was completed, the line was propagated by heterozygous intercrosses to generate mutants and wild-type littermates for further analysis.
Hoxc10 progeny were genotyped either by the original Southern analysis or by PCR. PCR was performed using the following primers: Hoxc10 forward: 5´-AAGCTGCACCAAGGTCTTGGGTC-3´; Hoxc10 reverse: 5´ AATTGGAGGTCAGTTCCCGGATC 3´ and lacZ-specific: 5´-CGCGCTCGAGATGTGCTGCAAGGCGATTAAG-3´. PCR cycle conditions were 35 seconds at 94oC, 15 seconds at 59oC, and 40 seconds at 72oC for 32 cycles. PCR bands were visualized on a 6% polyacrylamide or 1.5% Trevigel agarose gel (see Figure Figure1)1) and generated a WT band of 327 bp and a mutant band of 228 bp.
Newborn and adult mice were asphyxiated with CO2 and dissected tissues were fixed by immersion in Bouin's fixative or 4% paraformaldehyde for histological analysis. Specimens were embedded in paraffin and 10 μm serial sections were collected and stained with hematoxylin and eosin. Adults of all different genotypes were collected from littermates at 8 weeks (if not otherwise stated) for skeletal analysis and tissue collection. Skeleton preparations and X-gal staining were performed as described 7, 26. Immunohistochemical labeling was carried out according to established protocols 27. Anti-neurofilament antibodies (2H3) and anti-islet-1 antibodies (2D6) were obtained from the Developmental Studies Hybridoma Bank.
Locomotor behavior was assessed using a rotarod. Three month old sex-matched littermates were positioned on a rotarod apparatus rotating at 5 rpm. Once all mice were in position, the rotation was accelerated from 5 to 60 rpm over 180 sec and the rpm at fall recorded. Four trials were performed for each animal and the average rpm at fall calculated for each genotype. Statistical analysis of the results was performed using a Student's 1-tailed t-test assuming unequal variance. Initial analysis was performed separately on male and female mice, but as no differences in rotarod behavior were observed between males and females of the same genotype, data from both sexes were aggregated for the final statistical analysis.
Motor neuron cell counts on newborn animals were performed as previously described 9. Briefly, the position of the lumbar spinal segments was identified by screening serial histological sections for the position of the dorsal root entry zone and spinal nerves. The first lumbar segment was identified as the segment that contained the rostral end of the lateral lumbar motor column; spinal segments were then counted caudally from this reference point with the midpoint between adjacent spinal nerves used as the boundary between adjacent segments. Once segmental position was defined, motor neurons were systematically counted. Lumbar motor neurons were identified by size and position in histological sections; only cell profiles possessing a complete nuclear membrane and at least one distinct nucleolus were counted to minimize double counting. Both medial motor column and lateral motor column motor neurons were counted. Motor neuron counting data was compiled in Microsoft Excel and subjected to statistical analysis using Student's two-sample t-tests assuming unequal variance.
The Hoxc10 targeting vector was constructed such that most of the homeodomain sequence was replaced with an in-frame lacZ gene and a polymerase II promoter-driven neoR cassette (Figure (Figure1).1). This targeting strategy deletes all three alpha-helices of the homeodomain, producing a protein lacking the ability to bind to DNA. All possible genotypes were obtained in Mendelian ratios following heterozygous intercrosses, and Hoxc10 mutant mice are viable and fertile. The embryonic expression pattern of the fusion protein, as detected by X-Gal staining, is nearly identical to the previously described mRNA expression pattern 21, 26. ß-galactosidase expression is evident in caudal regions of the embryo including the prevertebrae, spinal cord, genital tubercule, and kidney, as well as surrounding the cartilage condensations of the pelvis (Figure (Figure2).2). Expression is also found in the proximal hindlimb, particularly in the perichondrium surrounding hyaline cartilage precursors of the tibia and femur (Figure (Figure2C)2C) and in the iliofemoral ligament (Figure (Figure2D-F).2D-F). ß-galactosidase expression is also detected in the developing nervous system, with expression overlapping that of Islet-1 in developing motor neurons in the lumbar spinal cord (Figure (Figure2G,H).2G,H). No overt behavioral phenotypes were observed, although with age, mutant mice show signs of hindlimb wasting and a high incidence of obesity. Both of these phenotypes are presumably caused by the abnormal iliofemoral ligament and the loss of motor neurons (described below), which may cause restricted mobility. In addition, 5 of 60 mutants examined had a single kidney.
The lack of a functional Hoxc10 homeobox causes several homeotic transformations in the axial skeleton (Table (Table1,1, Figures Figures33 and and4).4). Wild-type C57Bl/6 mice typically have 30 precaudal vertebrae, with the more caudal vertebrae organized in T13L6S4 pattern, with thirteen thoracic vertebrae, six lumbar vertebrae and four sacral vertebrae 10, 11. Eighty percent of the wild-type mice examined in this study exhibited this pattern, with the remainder of the animals showing some mild variation in the shape of the L1, L6, or S1 segments; these types of variations are within the range of normal (10. Hoxc10 mutant mice also have 30 precaudal vertebrae, but the patterning of the vertebral column is altered. Most Hoxc10 mutant mice (33 of 34 examined) show a partial to complete transformation of the thirteenth thoracic vertebrae, precaudal vertebra 20 (PC20) into a lumbar identity, typified by the reduction or complete loss of the thirteenth rib. By definition, thoracic vertebrae are those vertebrae with attached ribs; hence loss of the thirteenth rib suggests a posterior transformation of the most caudal thoracic vertebra into a lumbar identity. The most caudal lumbar vertebra (PC25) often undergoes a similar transformation into a sacral (S1) identity (Figures (Figures33 and and4).4). The sacrum in wild-type animals is typically comprised of four vertebrae. The first two vertebrae, S1 and S2, have butterfly-shaped, fused transverse processes. The next two sacral vertebrae, S3 and S4, have transverse processes that are not fused, with S3 having butterfly-shaped transverse processes and S4 having club-shaped transverse processes similar to those seen on more caudal vertebrae (Figures (Figures22 and and3).3). In Hoxc10 mutant animals, the fourth sacral vertebra (S3* in Figures Figures33 and and4)4) exhibits an intermediate shape between a normal S3 and S4, and there are usually five sacral-like vertebrae, three fused and two free. This suggests an anterior transformation of the first caudal vertebra into a sacral identity. This transformation restores the register of the vertebral column and thus there is no overall loss of precaudal vertebrae. In combination, these alterations produce a T12(T13/L1)L5S5 pattern in 94% of mutant mice. Heterozygous mice show several intermediate patterns, most commonly T13L5S5 (31.5%) or T12(T13/L1)L5S5 (38.9%), suggesting dosage compensation. The transitional vertebra, defined as the most anterior vertebra to show a lumbar rather than a thoracic articulation between the pre- and postzygapophyses [28 and references therein] is normally the tenth thoracic vertebra (T10), whereas in the homozygous Hoxc10 mutants the transition occurs at the ninth thoracic vertebra (T9) (Figure (Figure44).
The bones in the hip, the ilium, ischium and pubis, are constructed from independent condensations that grow together after birth. All three bones meet at the acetabulum, but the ischium and pubis also meet at the ventral edge of the pelvis. These two bones are angled 45o from each other on the anterior, acetabular end, and are connected by a cartilaginous bridge where they meet on the posterior side. This bridge undergoes calcification during puberty, and the pelvic apparatus is normally fully fused by eight weeks of age. In Hoxc10 mutants, the cartilaginous bridge forms normally, but the calcification process is delayed, leaving a prominent seam or indentation on the posterior edge of the pelvis, dyssymphysis ischio-pubica. In all mutant adults the bones touch each other but even in the mildest cases, a seam is visible unlike in the controls (Figure (Figure5A,5A, B; Table Table2).2). The C57Bl/6 background strain has been reported to have high incidence of dyssymphysis, especially in females 29, 30, however, somewhat newer reports have claimed only 30% incidence which was lost when C57Bl/6 mice were crossed to another strain 31. To minimize differences attributable to genetic background, wild-type sibling controls were used for all experiments.
The position of the pelvic girdle relative to the spinal column in general does not change in Hoxc10 mutants. At embryonic days 12.5-14.5 (E12.5-14.5), the femoral condensation of the hindlimb buds form at the level of prevertebrae 25 - 26 in both mutants and the wild types. However, the acetabulum is adjacent to the boundary between the third and fourth sacral vertebrae (S3 and S4) in both newborn and adult wildtype mice, while it is seen at the level of the fourth and fifth sacral vertebrae (S3* and S4) in most mutants. However, in both WT and mutant animals, this positioning places the acetabulum at the level of the boundary between the PC29 and PC30 vertebrae. Thus, the overall position of the acetabulum does not change in Hoxc10 mutant animals.
There are several alterations in hindlimbs and hip joints in Hoxc10 mutants. During embryonic and early postnatal development, there are no visible skeletal defects in the hindlimbs. However, by four to six weeks postnatally, a bony ridge develops along the anterior longitudinal line of the shaft of the femur (Figure (Figure5).5). A likely cause for the formation of this femoral ridge is the presence of an ectopic branch of the iliofemoral ligament. The iliofemoral ligament is a sheet-like structure on the anterior side of the femoral neck, acting as a part of the synovial capsule. Normally, the ligament joins the ilium anterior to the acetabulum of the hip joint to the intertrochanteric line of the femur. During development, this ligament expresses Hoxc10 (Figure (Figure2C-F).2C-F). The ischiofemoral ligament connects the dorsal aspect of the ischium to the femoral neck. In the mutant, the ischiofemoral ligament is visibly weaker than in the wild-type mice, and attaches more dorsally to the medial side of the rim. The iliofemoral ligament, on the other hand, has two femoral connections in the mutants: one normal connection that attaches as a part of the synovial capsule into the femoral neck and another that attaches onto the anterior shaft of the femur (Figures (Figures55 and and6).6). This second ectopic branch attaches to and is part of the anterior femoral ridge. The point of attachment varies from just distal to the intertrochanteric line to halfway down the femur, likely affecting the extent of ridge calcification that varies between animals and progresses with age. This phenotypic attachment site variation likely drives the differences in the shapes of the femoral ridges of the mutants from prominent to moderate, reflecting additional stresses on the bone. The femoral ridges are found in all the mutants six weeks and older (Table (Table2).2). Histological study of the extra ligament shows an organized fibrillar structure, indicating no aberrant pathology (Figure (Figure6C).6C). Serial sections of newborns show the ligament attaching to the femoral shaft (Supplemental Figure 2), also seen in some newborn skeletal preparations as Alcian blue-stained material in the ridge area indicating cartilaginous material, likely sulfated proteoglycans, at the femoral end of the abnormal ligament (Figure (Figure5F,5F, G).
Histological studies of the hip joint reveals further age-related changes. In adult mutant mice, the articular hyaline cartilage of the femoral head is thinner than normal, and does not have well-organized maturation layers (Figure (Figure7A,7A, B). In most mutants, no epiphyseal line is found. Transverse sections of adult wild-type thighs show an oval femoral bone, whereas the mutant femur is almost triangular with the anterior ridge forming the tip (Figure (Figure7C,7C, D). The lateral and medial posterior lips are formed by the lesser trochanter and third trochanter on the proximal side continuing as medial and lateral suprecondylar ridges distally by the knee. These structures form the two lips of the posterior femoral prominence, the linea aspara, which mice do not have as a distinguishable structure but which in several other mammals is a major posterior muscle attachment site. In mice, the equivalent muscles attach on an invisible line at the posterior, posteromedial and posterolateral positions of the oval bone. The posterior muscle attachment region cannot be found either in heterozygous or homozygous Hoxc10 mutants.
Hindlimb musculature appeared relatively normal in Hoxc10 mutant mice. Homozygous mutants have substantially shorter muscles on the medial side of the thigh compared to wild type, but the femur is also shorter. All muscles were found in their normal relative positions and their attachment sites were approximately normal on their femoral ends. The psoas muscle attaches to the lesser trochanter on the femur and to the ventral side of the lumbar vertebrae; however, a more anterior vertebral attachment in the mutant animals could not be ruled out. At weaning, the mutant and littermate hindlimbs are indistinguishable, but muscle wasting becomes apparent in older mice. By 14 weeks of age total muscle mass is about 70% of normal by weight of dissected muscles.
Initial open field observations suggested that there were no overt differences in locomotor behavior or motility in Hoxc10 mutants. Hoxc10 mutant mice had normal gaits and were able to grasp a pencil positioned along their ventral midline, a test of adduction. Open field behaviors were also normal, with mutant mice entering the same number of quadrants in an open field as wild-type mice. However, as concurrent observations suggested a significant loss of motor neurons in Hoxc10 mutants (see below), we performed a rotarod analysis as a specific test of locomotor function. A total of 19 mutant and 13 wild-type three-month old animals were subjected to rotarod testing as described in the Methods and Materials. The average rpm at fall was 35.5±11.8 for wild-type animals and 27.1±9.1 for mutant animals (Figure (Figure8).8). A one-tailed t-test, assuming unequal variance demonstrated that the results were significantly different, with p<0.001.
Initial analysis of peripheral nerve projection patterns in Hoxc10 mutant embryos using anti-neurofilament antibody labeling suggested that there were no gross defects in the formation, appearance, and projection of spinal nerves in Hoxc10 mutant embryos. Motor and sensory projections into the developing hindlimb bud appeared normal as well (Figure (Figure2).2). However, in light of alterations in motor neuron positioning in paralogous Hox10 gene mutants 8, 9 we decided to examine these neurons more carefully in serial histological sections collected from newborn animals. Serial 10 µm sections were collected from three wild-type and three Hoxc10 mutant animals. Lumbar segmental position was established as previously described 9 and the number of motor neurons in the medial and lateral motor columns (MMC and LMC, respectively) in the L1-L4 spinal segments was counted. All motor neuron populations showed substantial reductions in Hoxc10 mutants (Figure (Figure8).8). We further distinguished both medial and lateral components of the LMC; both components showed similar reductions in numbers of motor neurons. Despite the reduction in numbers of motor neurons, the distribution of these neurons appeared relatively unchanged, with MMC and LMCM neurons showing a peak in their distribution in segment L2 and LMCL neurons peaking in segments L3/L4. This suggests an absence of an anterior spinal segmental transformation, in contrast to results observed in Hoxd10 mutants and Hoxa10/Hoxd10 double mutants 7, 8. These cell counts appear to reflect an absolute loss of motor neurons, as there is no evidence of compensation in one pool for losses in another.
Hoxa10, Hoxc10, and Hoxd10 constitute a paralogous set of mammalian Hox genes. Previous studies have defined roles for Hoxa10 and Hoxd10 in patterning the axial and appendicular skeleton and the developing nervous system. Our current studies show that Hoxc10 also participates in patterning these structures, supporting the idea that paralogous Hox genes may coordinately regulate the patterning and organization of a set of common structures. In developing embryos, Hoxc10 is expressed in a broad posterior domain encompassing the developing neural tube and paraxial mesoderm. Expression is also evident in the developing kidney, genital tubercle, posterior mesenchyme, and surrounding cartilage condensations that form the pelvic bones. Like other Abd-B-related Hox genes, Hoxc10 also expressed in the developing limb buds; however, unlike their paralogs and other Abd-B-related genes, HoxC expression is restricted to the developing hindlimb 21, in particular to perichondrium and ligaments in the proximal hindlimb. These areas of expression prefigure regions in which tissue alterations are observed in Hoxc10 mutant mice. Alterations in axial and appendicular skeleton were observed, as were changes in nervous system structure and function.
Most Hoxc10 mutant mice show a partial or complete absence of a rib on the 20th precaudal vertebrae. In C57Bl/6 mice, the 20th precaudal vertebra (PC20) is normally T13, the most caudal thoracic vertebra. The loss of the most caudal rib suggests a posterior transformation of the T13 vertebra to an L1 phenotype. As T9 exhibits the phenotype of transitional vertebra instead of T10, we can also conclude that the posterior transformation register extends anteriorly to at least T9. Hoxa10 mutant mice often show a partial anterior transformation of the next most caudal thoracic vertebra, PC21, producing a partial rib and generating a T14 vertebra 8, 19. Both PC20 and PC21 appear unaffected in Hoxd10 mutant mice, however simultaneous inactivation of Hoxa10 and Hoxd10 commonly produces a complete anterior transformation of the PC21 to a full T14 phenotype 8. Combinatorial inactivation of Hoxa10 and Hoxc10 also shows an anterior transformation including additional ribs and loss of lumbar and some sacral segments (Hostikka and Capecchi, unpublished observations). These combinatorial studies suggest that Hoxa10 may be the dominant gene among the Hox10 paralogues, as anterior transformations predominate in combinatorial mutants with Hoxa10 , whereas Hoxd10 and Hoxc10 double mutants show a wide range of milder forms of intermediate lumbosacral identities with the reoccurring theme of an expanded sacral domain (Hostikka and Capecchi, unpublished results). Thus, loss of activity of each of the three individual Hox10 paralogues produces a different axial phenotype. The absence of all three Hox10 genes produces mice with a transformation of all lumbar and sacral segments to a thoracic identity 16, further illustrating the requirement for Hox10 paralogues in establishing lumbar skeletal identity. It should be noted that the Hoxc10 allele used by Wellik and Capecchi is the allele described in this study, so allelic differences arising from lacZ fusion and the retention of neoR in our construct are likely not a factor in producing axial skeletal phenotypes. The phenotypes observed in single and combinatorial mutations in the Hox10 genes suggest a role for these three genes in establishing the division between thoracic, lumbar, and sacral vertebrae. Mutations in two of the three genes individually can alter the position of the boundary between regions by one segment anteriorly or posteriorly, while mutations in all three genes together serve to erase the boundary between regions entirely. These observations suggest a necessary balance in Hox gene activity, with activity of all three genes required to establish the correct rostrocaudal position of the lumbar vertebrae.
Another axial skeletal phenotype was observed in the formation of the pelvis, where evidence of dyssymphysis ischio-pubis was observed in all mutant animals. Dyssymphysis ischio-pubis has been reported as common, with 30-100% penetrance in several inbred strains by researchers in the 1940s - 1960s 10, 11, 31, however our wild-type siblings showed no evidence of this phenotype. Dyssymphysis ischio-pubis is erased in strain hybrids 31; thus its absence in our wild-type controls could reflect some hybrid variability attributable to the small percentage of CC1.2 DNA carried in our parent lines. Alternatively, since the appearance of dyssymphysis ischio-pubis is tied to the presence of the Hoxc10 mutation, this phenotype is likely attributable to the engineered mutation. Pelvic formation and placement appears otherwise unaffected.
Limb phenotypes include the presence of an abnormal branch of the iliofemoral ligament and remodeling of the femur to produce an ectopic anterior ridge and a modified profile in the shaft of the femur. Early muscle cleavage, and differentiation of the pelvic and hindlimb musculature occurs between embryonic day (E) 12.5 and E14.5, although the final muscle structure is not complete until after birth 32. By E14.5 the femur has rotated anteromedially and projects ventrally rather than laterally from the acetabulum. During this process, the muscle anlagen rotate with the femur and the ligament attachments to the pelvis form. In Hoxc10 mutant mice, the femoral attachment extended minimally more anteriorly (or distally) than normal, with the femoral iliofemoral ligament attachment site broad enough to extend distally over the femoral growth plate during development, likely causing it to be torn into two branches by the growth of the femur. Hoxc10 is highly expressed in the developing the ligaments as well as perichondrium of the attachment sites in the cartilage condensations of hip and femoral bones, and may be required for the maturation of the ligament, similar to suggested roles for Hox genes in cartilage differentiation and maturation 33, 34. The absence of Hoxa10 also leads to the production of an abnormal ligament 19, suggesting similar roles for other Hox10 paralogues.
The presence of the ectopic femoral ridge in association with the abnormal iliofemoral ligament suggests a cause-and-effect relationship, whereby the ectopic ligament branch produces additional stresses on the developing femur. The adaptive response to these stresses is thus the formation of the ectopic ridge, similar to bony spurs. The bone formation likely appears through fibrocartilage calcification, but ligament attachment sites have also been suggested to function as growth plates [35, and references therein]. Femoral remodeling in response to load, compression and stress-shielding is seen throughout evolution 35, 36 and in the course of mammalian development. For example, femoral architecture shifts in humans in response to the transition to bipedal locomotion 37 and to changes in body mass 38. The change in profile shape of the shaft of the femur also suggests altered stresses on the bone. The rounded profile and lack of a linea aspera in wild-type femurs suggests roughly equal forces applied via muscle attachments around the shaft of the femur, whereas, the triangular profile evident in mutant animals suggests a shift in the load-bearing of the femur, which would create novel strains on this bone, as the muscle attachment sites did not appear affected. However, as the heterozygotes show an intermediate phenotype, Hoxc10 could be required for the periosteum to retain an oval shape against the stress of the attached muscles. The ability of the femur to undergo remodeling may also be affected by the differential production and placement of osteogenic cells. It has been suggested through archeological evidence that bone formation in stressed ligament attachment sites could be genetically dictated [35, and references therein]. Recently, the expression of Hoxa2 has been shown to impair cartilage maturation 34 and our data could be interpreted as a premature maturation of cartilage and osteogenic cells in the absence of Hoxc10. Hoxc10 mice also appear more obese in adulthood than wild-type littermates. This extra weight may induce further stresses on the limb architecture, with concomitant changes in bone shape and structure, similar to remodeling events observed in obese humans 38.
Observations in Hoxa10 and Hoxd10 mutant mice have suggested that gene inactivation produces alterations in the placement and projection of lumbar motor neurons. Hoxa10 mutants show anterior homeotic transformation of spinal nerves 4 while Hoxd10 mutants have a reduced number of lumbar segments and shifts in the peripheral projections of spinal nerves 7. Our observations in Hoxc10 mice also suggest a nervous system phenotype, with mutant mice showing a significantly reduced number of lumbar motor neurons. The observed motor neuron loss is likely to contribute to the altered locomotor behavior observed in rotarod testing. Hoxa10/Hoxd10 double mutant animals show significant reductions in the number of lumbar motor neurons, along with shifts in their position 9 and Hoxc10/Hoxd10 double mutants show altered differentiation of rostral lumbar motor neurons, with a transformation of rostral lumbar motor neurons into a thoracic identity 20. This combined set of observations suggests that Hox10 gene expression has a significant role in patterning lumbar motor neurons.
In conclusion, our results provide evidence for a contribution of Hoxc10 to the regulation of many aspects of body plan patterning. Significant changes are observed in both skeletal and nervous system patterning and the Hoxc10 mouse appears to provide an excellent model for the study of mechanical stress on bone remodeling. Although deleting the entire HoxC complex appears to have negligible effects on embryonic patterning 39, the phenotype produced by selective deletion of Hoxc10, as well as other individual HoxC genes, demonstrates a clear dependence on the function of these genes to establish normal body plan development.
Supplemental figure 1
Supplemental figure 2
This study was supported by the Sigrid Juselius Foundation (SLH), the Finnish Cultural Foundation (SLH), and an NSF CAREER award to EMC. We wish to thank Mario Capecchi for support and encouragement in the initial phases of this project and Derrick Rancourt for the CC1.2 specific genomic library.