DEVELOPMENTAL DYNAMICS 208:170–177 (1997) Skeletal Development in Transgenic Mice Expressing a Mutation at Gly574Ser of Type II Collagen B. KERRY MADDOX,1,2 SILVIO GAROFALO,1,2 CHAD SMITH,3 DOUGLAS R. KEENE,1,2 AND WILLIAM A. HORTON1,2* 1Research Department, Shriners Hospital for Children, Portland, Oregon 97201 2Department of Molecular and Medical Genetics, Oregon Health Sciences University, Portland, Oregon 97201 3Department of Molecular Genetics, The University of Texas M.D. Anderson Cancer Center, Houston, Texas 77030 ABSTRACT Skeletal development of transgenic mice with a type II collagen mutation was analyzed and compared with wild-type littermates. The single base substitution in Col2a1 resulted in a glycine to serine mutation within the helical domain and corresponded to one previously identified in a patient with the lethal human chondrodysplasia, hypochondrogenesis (Horton et al.  Proc. Natl. Acad. Sci. U.S.A. 89:4583–4587). Skeletal staining of embryos from 14.5 through 18.5 days of gestation demonstrated a dwarf phenotype in the transgenic embryos, most notably short limb bones and vertebral column that was first detected at 15.5 days postcoitus. In addition to the reduced length, the extent of ossification was less in the transgenic mice. The architecture of the long bone growth plate was abnormal in the transgenic tissue, in particular there was no discernible proliferative zone. There were few stacks of characteristically flattened cells and the overall length of the growth plate in the mutant embryos was reduced. At the ultrastructural level, there were fewer collagen fibrils present in the transgenic mouse cartilage compared to that of wild-type littermates. Ultrastructural localization of collagen types II, IX and XI revealed a similar pattern between the transgenic and wild-type pups, suggesting that the collagen fibrils present in the matrix of littermates with both phenotypes had a similar composition. Skeletal analysis and cartilage histochemistry indicated that effect of the type II collagen mutation was to reduce the density of the collagen fibrils within the cartilage matrix which was associated with delayed bone formation and resulted in a short-limbed phenotype. Dev. Dyn. 208: 170–177, 1997. r 1997 Wiley-Liss, Inc. Key words: chondrodysplasia; cartilage; limb development; collagen zone at the outermost area of the growth plate at the epiphysis, the proliferative zone which contains groups of flattened chondrocytes arranged in columns, and the hypertrophic region immediately adjacent to the underlying bone (for a review, see Horton, 1993). The hypertrophic zone is degraded and replaced by bone during the process of endochondral bone formation. Cycles of chondrocyte proliferation, hypertrophy, and replacement by bone result in skeletal growth that continues until the growth plate closes after puberty. By poorly understood mechanisms, this process of bone lengthening is disturbed in a group of dwarfing conditions known as the chondrodysplasias. The histological findings commonly associated with these skeletal disorders include a shortened proliferative zone and sparse extracellular matrix (Horton and Hecht, 1993). In normal growth plate cartilage, a network of collagen fibrils is intertwined with large aggregated proteoglycans. The fibrils are heteropolymers of collagen types II, IX, and XI that are covalently cross-linked in the molar ratio of 8:1:1, respectively (Vaughan et al., 1988; Mendler et al., 1989). From immunolocalization studies type IX collagen appears to be aligned along the fibril surface (Vaughan et al., 1988) where it is speculated to interact with other matrix components through a large globular domain that projects out into the extracellular space. Type XI collagen also contains a globular domain that is exposed to the surface of the fibril and is thought to be responsible for regulation of fibril diameter (Thom and Morris, 1991; Keene et al., 1995). Genes coding for the cartilage matrix components are good candidates for mutations in chondrodysplasias. Indeed, mutations of genes for all three cartilagespecific collagens have been characterized in distinct but related skeletal conditions (Briggs et al., 1994; Spranger et al., 1994; Vikkula et al., 1995; Muragaki et al., 1996). Mutations in COL2A1, the gene for type II collagen, have been identified in chondrodysplasias of the spondyloepiphyseal dysplasia class (Spranger et al., INTRODUCTION Linear growth of long bones occurs within the growth plate, a narrow zone of specialized cartilage that forms the interface of epiphyseal cartilage and developing bone. It is organized into three regions from the resting r 1997 WILEY-LISS, INC. Contract Grant sponsor: National Institutes of Health; Contract Grant number: HD20691; Contract Grant sponsor: Shriners Hospital for Children; Contract Grant number: 15956. *Correspondence to: W.A. Horton, Research Department, Shriners Hospital for Children, Portland, OR 97201. Received 7 May 1996; Accepted 11 September 1996 SKELETAL ANALYSIS OF TRANSGENIC MICE 1994). Heterozygous glycine substitutions have been reported in three cases of hypochondrogenesis, which lies at the severe end of the clinical spectrum (Bogaert et al., 1992; Horton et al., 1992; Freisinger et al., 1994). This lethal disorder is characterized by short limbs, a shortened trunk, cleft palate, and reduced ossification of the vertebral column. We recently reported a heterozygous mutation in the gene for type II collagen in an infant with hypochondrogenesis (Horton et al., 1992). The mutation resulted in an amino acid substitution of serine for glycine at residue 574 (Gly574Ser), slightly carboxy-terminal to the middle of the triple helical domain of the collagen molecule. Towards understanding how it disrupts bone formation, we introduced the Gly574Ser mutation into transgenic mice. The murine Col2a1 promoter was used to produce cartilage-specific expression of this mutation which was expected to act in a dominant negative fashion. In this paper, we examine the temporal aspects of how the phenotype is produced. The data demonstrate that skeletal differences between the wild-type and transgenic littermates are not apparent before endochondral bone formation begins around day 15.5 of embryonic development. In older embryos, histological and ultrastructural analysis revealed an abnormal growth plate architecture and a reduced abundance of fibrils in the cartilage matrix from the transgenic mice in conjunction with a short-limbed phenotype. RESULTS Generation of Transgenic Mice PCR mutagenesis was used to introduce a point mutation in the mouse type II collagen gene that resulted in a Gly574Ser amino acid substitution slightly carboxy-terminal to the middle of the triple helical domain. The genomic clone has been previously described and successfully used to generate other type II collagen transgenic mouse strains (Garofalo et al., 1991; Metsaranta et al., 1992; Garofalo et al., 1993). One male founder was generated using conventional transgenic technology and subsequent breeding produced offspring, 25–50% of which were born dead and had an abnormal phenotype. These pups presented with short limbs, short snouts, cleft palate, and distended abdomen when compared to the wild-type littermates. Skeletal Clearing of Embryos Embryos were removed at days 13.5 through 18.5 post-coitus (p.c.) by Caeserian section. The identification of wild-type or transgenic embryos was verified by Southern analysis (data not shown). They were skinned, fixed, and stained using alcian blue to identify the cartilaginous areas and alizarin red which stains the mineralized bone tissue. The skeleton was revealed following alkali clearing. At least two litters at each stage were analyzed; one representative wild-type and transgenic embryo from each litter at the respective stage of development is shown (Fig. 1). At 13.5 days p.c., 171 there was no discernable difference between the transgenic and the wild-type littermates (data not shown). There was no ossified tissue present at this stage. At 14.5 days there was a small site of ossification above the shoulder, presumably the clavicle, as indicated by the alizarin red stain in both embryos (Fig. 1A, 14.5 arrow). The length of the limbs and overall size appeared equal at this stage of development. By day 15.5, transgenic embryos were distinguished from the wild-type pups based on a difference in the pattern of ossification of the vertebral column, the rib cage, and the limbs while overall size appeared similar. The cranium and mandible, derived from membranous bone, were similarly stained with alizarin red while the limbs and vertebral column, derived by the process of endochondral bone formation, were different between littermates. Ossification of the rib cage proceeds from the vertebral column towards the sternum. All of the ribs in the wild-type embryo had begun to ossify (Fig. 1A, 15.5 arrowheads) and the vertebral bodies of the thoracic region also showed mineral deposition indicating bone formation. Only five of the thoracic and sacral ribs in the transgenic embryo had begun to mineralize and these varied in the extent of alizarin red staining. There was no indication of vertebral ossification and the cartilaginous vertebral bodies appeared smaller and the pedicles less prominent. In the wild-type embryos, bone formation of the hindlimb had begun within the midshaft of the femur, the tibia, and fibula to an approximately equal degree. A similar pattern was seen in the bones of the forelimb. Interestingly, only the bones of the middle segments, the radius and ulna in the forelimb, and likewise the tibia and fibula in the hindlimb, had begun to ossify in the transgenic mouse. There was no visible staining of bone in either the femur or the humerus. Where ossification had occurred in these mice, the bones appeared somewhat wider compared to the normal embryos. At day 16.5, the differences in skeletal development were more noticeable. The wild-type embryos were larger due to the longer vertebral column and the tail, which was quite short in the transgenic embryo. In regards to the rib cage, the distance between ribs, reflecting the height of the vertebral body, looked slightly smaller in the transgenic embryo. In addition, there was a thickening in the ossified region of several ribs (Fig. 1A, 16.5 arrow), reminiscent of the thickening seen in the lower limbs at 15.5 days p.c. The pedicles on the sacral vertebrae had begun to ossify in the wildtype skeleton, unlike the transgenic skeleton where the staining of the vertebral bodies and pedicles indicated only cartilage. The limbs were distinctly shorter as well: the long bones were shorter and paws were underdeveloped in both size and ossification. The femur and humerus had begun the process of endochondral bone formation by this time point, 16.5 days p.c. Profiles of the skulls of the embryos showed differences as well. The mandible, the premaxilla and nasal cavity were not as prominent in the transgenic mouse, result- 172 MADDOX ET AL. Fig. 1. Skeletal analysis of wild-type and transgenic embryos from 14.5 to 18.5 days of gestation. Embryo skeletons were removed by Caesarean sections at time-points indicated; the skeletons were stained for cartilage (alcian blue) and bone (alizarin red). A: Lateral view of intact skeletons from wild-type (WT), upper row, and transgenic (TG) embryos, lower row. At 14.5 days p.c. there is a small ossified site above the shoulder in each embryo (arrow). At 15.5 days p.c. there is less mineralization of the ribs in the TG than WT embryo (arrowheads) and only the tibia and fibula of the hind limb had ossified in the TG (arrows). At 16.5 days p.c. the skulls of the WT have a larger developed occipitus and otic capsule than the TG (arrowheads); the bone-portion of the TG ribs are irregularly thickened (arrow). At 17.5 days p.c. the delay in ossification of the vertebral column of the TG embryo is evident (arrowhead) and persists through day 18.5 p.c. B: The vertebral columns were removed and aligned with the WT above its TG littermate in increasing order of gestational age, from 16.5 to 18.5 days p.c. The proportional difference in the length of the ossified region of the vertebral columns between the WT and TG decreases with increasing days of development (bars delineate the ossified region). C: Hind limbs were removed and aligned as in B. The TG limbs are very short and the reduced ossification in the limb bones and paws is evident Bar 5 5 mm. 173 SKELETAL ANALYSIS OF TRANSGENIC MICE ing in a shortened snout. The posterior region of the skull which included the occipitus and the otic capsule were smaller in the transgenic skeleton so that the parietal and more frontal structures were closer to the cervical spine (Fig. 1A, 16.5 arrowheads). The differences noted earlier persisted through days 17.5 and 18.5 of embryonic development. The transgenic limb length was short and the bone looked thicker than the wild-type littermate due to the shorter length. However, the carpals and metacarpals of the paws had begun to ossify and by 18.5 days p.c., the transgenic paws looked like the paws of wild-type 16.5 or 17.5 days p.c. embryos. There was no ossification of the transgenic lower spine at 17.5 days p.c., although by 18.5 days p.c., there appeared to be a small amount of alizarin red staining in the pedicles. The vertebral bodies of the tail did not ossify during the period of this study and the length of the tail continued to be disproportionally short. In addition, the facial features of the transgenic embryos remained flat and the posterior portion of the skull undeveloped. The abnormal skeletal features noted above, the short limbs, vertebral column and odd-shaped skull were observed at 15.5 days p.c. when endochondral ossification became apparent. Discrepancy in the relative length of the vertebral columns and limbs increased between the wild-type and transgenic embryos as embryonic development proceeded (Fig. 1B,C); however, the proportional difference in the length of the ossified region of the vertebral column between the transgenic and wild-type embryos was lessened as development proceeded (Fig. 1B; bars delineate the ossified vertebral column). The same phenomenon of initial delay of ossification followed by tendency to normalize occurred at other sites, such as the carpal and metacarpal bones of the paws. Unfortunately, the Gly574Ser mutation was lethal so that post-natal development was not studied. the proliferating zone between the transgenic and wild-type cartilage, the cells within the resting zone of the mutant cartilage had a normal rounded morphology (top of TG panel, Fig. 2). Morphological differences were even more apparent between the hypertrophic zones of the respective littermates. The hypertrophic zone of the wild-type growth plate was distinguished by enlarged cells that were organized in clusters and contained large vacuoles. They were much larger than the transgenic cells which did not contain similar vacuoles. In addition, the hypertrophic cells were well-separated by extracellular matrix in contrast to the tightly packed hypertrophic cells of the wild-type growth plate. An interesting observation was made concerning the provisional calcification of the hypertrophic zone. In the TG growth plate there were deposits of calcium scattered irregularly throughout the hypertrophic area (Fig. 2 TG, arrowheads) while they tended to be in longitudinal septa near the interface with bone in the wild-type section. Ultrastructure of the Cartilage Matrix Consistently through the growth plate, the density of collagen fibrils was reduced in transgenic mice (Fig. 3, Tg). This was observed in the matrix adjacent to the cells as well as between cells. The fibrils from the transgenic animals, while few in number, appeared to be the same diameter as wild-type (Fig. 3,F). Antibodies to the three known collagen components of cartilage fibrils were used to assess the composition of normal-appearing fibrils. Immunoelectron microscopy using antibodies specific for collagen types II, IX, and XI demonstrated that the pattern of staining along the individual fibrils was no different between the wild-type and transgenic mouse cartilage, although density of the fibrils differed (Fig. 4, compare A-B, C-D, E-F). DISCUSSION Growth Plate Histology The architecture of the growth plate from the right distal humerus of mice from several different litters was analyzed at 18.5 days p.c. by electron microscopy. A typical growth plate was observed in the wild-type embryos as indicated by the demarcation of the zones represented by distinct cell morphologies (Fig. 2, WT). The space normally occupied by the proliferative zone, i.e., the space between the resting and hypertrophic zones, was shortened in the transgenic tissue and contained fewer cells. Therefore, the same distance from the hypertrophic zone in the transgenic growth plate extended further into the resting zone than the wild-type. The proliferating chondrocytes in the wildtype growth plate were flattened and organized in linear clusters, typical of normal cartilage, while the corresponding cells in the transgenic growth plate were more homogeneously dispersed. The transgenic cells also appeared to be larger and less extended. While there seemed to be a difference in shape of cells within We have described a transgenic mouse strain with a mutation in type II collagen at Gly574Ser that resulted in a short-limbed phenotype. This mutation was identical to one previously characterized in a patient who presented with hypochondrogenesis (Horton et al., 1992), a lethal human chondrodysplasia. There were distinctive similarities in the abnormal skeletal features between the patient and the transgenic mouse, especially the reduced limb length, the delay of ossification of the vertebral column, and the short rib cage. The transgenic mouse model was subsequently used to investigate how defective molecules implicated in these disorders disturb the processes of chondrogenesis and bone formation. From biochemical analyses, it was apparent that the mutation caused a reduced amount of type II collagen to be secreted and assembled into the cartilage matrix. From these data, it appeared that the pathology associated with the mutation was the result of less type II collagen present which accounted for the sparse fibrillar network shown here. 174 MADDOX ET AL. Fig. 2. Growth plate morphology from a WT and TG embryo at 18.5 day p.c. The dark lines to the left of the micrograph denote the regions of the growth plate in the distal humerus as resting (R), proliferating (P) and hypertrophic zone (H); the arrow adjacent to the (R) indicates that the resting zone continues beyond the field shown. The proliferating zone of the WT growth plate contains clusters of flattened, stacked cells (arrow) that are missing in the TG embryos. There are unusual mineral depositions in the hypertrophic zone of the TG section (arrowheads). Bar 5 50 µm. The onset of the abnormal skeletal features coincided with the initiation of ossification of cartilaginous tissues. The molecular events that occur during endochondral bone formation are complex and poorly understood. For example, it is not clear if cartilage acts as a scaffold upon which ossification occurs, if it provides the molecular signals that lead to bone formation, or whether it is simply degraded and replaced by bone in a temporal manner. In terms of a specific mechanism, it is difficult to explain the resulting short bones of the Col2a1 mutation from the sparse fibrillar network. However, it was apparent that the growth plate was short and that the number of cells within this specialized region of cartilage was fewer than normal, possibly reflecting a reduced ability of the chondrocytes to proliferate. The morphology of the hypertrophic cells was abnormal, an observation that raises the possibility of incomplete differentiation of the cells. Both of these findings would affect bone length by reducing the total number of cells turned over in a given period of time in the growth plate and reducing the volume in the hypertrophic cells. There have been a number of other transgenic mouse strains produced with Col2a1 mutations. These include a single base substitution (Garofalo et al., 1991), deletions (Vandenberg et al., 1991; Metsaranta et al., 1992; Helminen et al., 1993), and inactivation of the Col2a1 gene, discussed below (Li et al., 1995). The resulting phenotypes varied, some of which had abnormal skeletal features similar to our dwarf mouse strain with the Gly574Ser mutation while others survived and later developed osteoarthritis. The fact that other Col2a1 mouse mutations resulted in a similar short-limbed dwarfism suggests that they affect bone formation and growth in a common manner which in this case involves reduced secretion of type II collagen. The effect of overexpression of the normal gene product was addressed (Garofalo et al., 1993) in which an increasing level of expression of the transgene was correlated with an increasing number of large diameter fibrils in the cartilage matrix. At 43% expression of transgene to endogenous type II mRNA, thick fibrils were evident. In our mouse strain, the transgene was expressed at approximately this same level and no thick fibrils were present. This provides support that the phenotype of the Gly574Ser strain was the result of the specific mutation rather than the presence of the transgene. We were surprised to discover that the cartilage matrix of the Gly574Ser Col2a1 mutation did not contain any detectable serine at that position because a more severe phenotype is generally thought to result from assembly of the mutant procollagen chains into the extracellular matrix. Milder forms of osteogenesis imperfecta are typically associated with mutations in which the mutant gene products are not incorporated into the extracellular matrix (Byers et al., 1991; Byers, 1993). One notable distinction between the Gly574Ser and the other mouse strains is that the latter contained distended rough endoplasmic reticulum (RER) in the chondrocytes, indicating intracellular storage and presumably poor secretion of newly synthesized proteins. We did not consistently observe enlarged RER in the chondrocytes from the Gly574Ser transgenic mice. A recent publication addressed the role of type II collagen in cartilage and bone formation by inactivating its gene to produce type II collagen-null mice (Li et al., 1995). Comparing the phenotype of the heterozygous and null mice with that of the Gly574Ser transgenic strain described in this report is perplexing. The heterozgous mice showed little difference in skeletal SKELETAL ANALYSIS OF TRANSGENIC MICE 175 Fig. 3. Ultrastructure of cartilage matrix. Electron microscopy of an area within the growth plate resting zone shows fewer fibrils (f) in the extracellular matrix of the transgenic (Tg) matrix to the right of wild-type (Wt). Note the plasma membrane (pm) of a chondrocyte (C) at the lower periphery of the micrograph. Bar 5 50 nm. features compared to their wild-type littermates. On a superficial level, the null mice and our transgenic ones presented with short-limbed chondrodysplasia. The facial features and limb length of the null mouse were consistent with the Gly574Ser mutant phenotype and other previously described transgenic mice mentioned above. However, other subtle differences were noted. For example, the shape of the long bones of the null mouse was quite distinct in that the diameter was abnormally large, reducing the ratio of length to width. The delay of ossification of the vertebral column was much greater in the Gly574Ser mice. Interestingly, there were fewer collagen fibrils in the null mouse cartilage than in our transgenic mouse, and yet more extensive ossification of their skeletons. While it is difficult to explain the respective phenotypes based on the information available, it raises intriguing questions about the biological functions of type II collagen for normal skeletal formation and linear growth of bones. In conclusion, we have analyzed the developing skeleton of a mouse expressing a dominant negative mutation of Col2a1. We speculate that the reduced density of collagen fibrils in the matrix and poorly differentiated chondrocytes contribute to the profound delay of the normal developmental sequence and suggest that this phenomenon accounts for the similar developmental delay seen in human hypochondrogenesis. Our observations raise the possibility that chondrocyte proliferation and differentiation is somehow perturbed by the reduction of collagen fibrils. EXPERIMENTAL PROCEDURES Generation of the Transgenic Mice The original transgene construct, cosmid pWE15 containing the murine wild-type Col2a1, has been previously described (Garofalo et al., 1991) and used successfully for the production of other transgenic mouse strains (Garofalo et al., 1991; Metsaranta et al., 1992; Garofalo et al., 1993). Two mutations were introduced into pWE15 using site-directed mutagenesis to make the Gly574Ser transgene, one of which was a G = A base substitution in exon 33 and coded for the amino acid substitution; the second introduced a silent mutation that destroyed a SacI site. The founder mice were mated with B6D2F1 mice for transgenic F1 litters; founder mice and subsequent offspring were identified and transgene copy number determined by Southern blot analysis after digestion of tail genomic DNA with NcoI as previously described (Garofalo et al., 1991). Preparation of Stained Skeletons The method of staining and clearing the embryo skeletons was based on two previously described protocols (McLeod, 1980; Martin et al., 1995). Briefly, following a Caesarean section, the embryos were removed and eviscerated. The placentas were set aside for DNA extraction and Southern analysis to identify which carried the transgene. The embryos were placed in 176 MADDOX ET AL. Fig. 4. Immunolocalization of collagens to the cartilage fibrils. Antibodies to collagen types II, IX, and XI recognize epitopes along the fibrils in tissue from WT (A,C,E) and TG (B,D,F) growth cartilage. The upper panels have been incubated with antibodies to type II collagen (A,B); the middle panels with antibodies to type IX collagen (C,D); the lower panels with antibodies to type XI collagen (E,F). A secondary antibody conjugated to a 5 nm gold bead particle is directed to the collagen antigens. While there are fewer fibrils present in the Tg tissue, the labeling of the fibrils appears similar to the Wt. Bar 5 200 nm. water overnight. The following day they were immersed in water at 65°C for 30 sec to 1 min (14.5 and 15.5 days p.c. embryos were immersed for 30 sec; the larger embryos for up to 1 min). For the larger embryos, the skin was removed prior to fixation in absolute ethanol for 3 days. The cartilage was stained with an acetic alcian blue solution overnight at 37°C (15 mg alcian blue in 100 ml 75% ethanol containing 20% (V/V) acetic acid). The mice were then placed in 100% ethanol for another overnight period. The following day the skeletons were cleared for 6 hr in 0.5% for the 14.5 and 15.5 days p.c. embryos and 1% for the 16.5, 17.5, and 18.5 days p.c. The bone was stained overnight in alizarin red (50 mg in 1 L of potassium hydroxide, the concentrations consistent for the relative stage of development as above). The clearing was continued in potassium hydroxide until only the skeletons remained. They were stored in 50% glycerol. SKELETAL ANALYSIS OF TRANSGENIC MICE Ultrastructural Analysis For ultrastructural observation, right forearms, including growth plate and bone, were examined following fixation in 1.5% paraformaldehyde/1.5% glutaraldehyde and osmium tetroxide, dehydration in ethanol and embedding in Spurrs epoxy. Ultrastructural immunocytochemistry was performed using the enbloc method previously described (Keene et al., 1995) with some modification. Briefly, left forearms were snap-frozen in hexanes, then thawed in PBS. Following exposure to chondroitinase (0.25 units/ml PBS, pH 7.4) for 2 hr at room temperature, the tissues were rinsed in PBS for 30 min prior to immersion in primary antibody diluted 1:5 in PBS overnight at 4°C. Primary antibodies included a rabbit anti-XI antibody, pAb16123 (Keene et al., 1995) a goat anti-type II antibody (Southern Biotech 1320-01, Birmingham, AL), and a rabbit anti-type IX antibody, pAb9264 (Keene et al., 1995). Following an extensive rinse in PBS, the samples were immersed in 5 nm gold conjugated to appropriate secondary antibody (Amersham) diluted 1:3 in 1.0% BSA buffer (20 mM Tris-HCl, 0.9% NaCl, 1 mg/ml BSA, 20 mM NaN3) pH 8.0 for about 15 hr. The tissues were again rinsed extensively in PBS then fixed, dehydrated, and embedded as described previously. ACKNOWLEDGMENTS The authors gratefully acknowledge the expert technical assistance of Catherine Ridgway and Brian Demings. This work was supported by grants from the National Institutes of Health, HD20691, Shriners Hospital for Children, 15956, Fred Meyer Charitable Trust Foundation, and the R. Blaine Bramble Medical Research Foundation. REFERENCES Bogaert, R., Tiller, G.E., Weis, M.A., Gruber, H.E., Rimoin, D.L., Cohn, D.H., and Eyre, D.R. (1992) An amino acid substitution (Gly853 = Glu) in the collagen a1(II) chain produces hypochondrogenesis. J. Biol. Chem. 267:22522–22526. Briggs, M.D., Choi, H.C., Warman, M.L., Loughlin, J.A., Wordsworth, P., Sykes, B.C., Irven, C., Smith, M., Wynnedavies, R., Lipson, M.H., Biesecker, L.G., Garber, A.P., Lachman, R., Olsen, B.R., Rimoin, D.L., and Cohn, D.H. (1994) Genetic mapping of a locus for multiple epiphyseal dysplasia (EDM2) to a region of chromosome I containing a type IX collagen gene. Am. J. Hum. Genet. 55:678–684. Byers, P.H. (1993) Osteogenesis imperfecta. In: ‘‘Connective Tissue and Its Heritable Disorders,’’ McCusick, V. (ed). New York: WileyLiss, Inc., pp 317–350. Byers, P.H., Wallis, G.A., and Willing, M.C. (1991) Osteogenesis imperfecta: Translation of mutation to phenotype. J. Med. Genet. 28:433–442. Freisinger, P., Alakokko, L., Leguellec, D., Franc, S., Bouvier, R., Ritvaniemi, P., Prockop, D.J., and Bonaventure, J. (1994) Mutation in the Col2A1 gene in a patient with hypochondrogenesis— Expression of mutated Col2A1 gene is accompanied by expression of genes for Type I procollagen in chondrocytes. J. Biol. Chem. 269: 13663–13669. Garofalo, S., Vuorio, E., Metsaranta, M., Rosati, R., Toman, D., Vaughan, J., Lozano, G., Mayne, R., Ellard, J., Horton, W., and de Crombrugghe, B. (1991) Reduced amounts of cartilage collagen fibrils and growth plate anomalies in transgenic mice harboring a 177 glycine-to-cysteine mutation in the mouse type II procollagen a1chain gene. Proc. Natl. Acad. Sci. U.S.A. 88:9648–9652. Garofalo, S., Metsaranta, M., Ellard, J., Smith, C., Horton, W., Vuorio, E., and de Crombrugghe, B. (1993) Assembly of cartilage collagen fibrils is disrupted by overexpression of normal type II collagen in transgenic mice. Proc. Natl. Acad. Sci. U.S.A. 90:3825–3829. Helminen, H.J., Kiraly, K., Pelttari, A., Tammi, M.I., Vandenberg, P., Pereira, R., Dhulipala, R., Khillan, J.S., Ala-Kokko, L., Hume, E.L., Sokolov, B.P., and Prockop, D.J. (1993) An inbred line of transgenic mice expressing an internally deleted gene for type II procollagen (COL2A). J. Clin. Invest. 92:582–595. Horton, W.A. (1993) Morphology of Connective Tissue: Cartilage. In: ‘‘Connective Tissue and Its Heritable Disorders,’’ McCusick, V. (ed). New York: Wiley-Liss, Inc., pp. 73–84. Horton, W.A. and Hecht, J.T. (1993) The chondrodysplasias. In: ‘‘Connective Tissue and Its Heritable Disorders,’’ McCusick, V. (ed). New York: Wiley-Liss, Inc., pp. 641–675. Horton, W.A., Machado, M.A., Ellard, J., Campbell, D., Bartley, J., Ramirez, F., Vitale, E., and Lee, B. (1992) Characterization of a type II collagen gene (COL2A1) mutation identified in cultured chondrocytes from human hypochondrogenesis. Proc. Natl. Acad. Sci. U.S.A. 89:4583–4587. Keene, D.R., Oxford, J.T., and Morris, N.P. (1995) Ultrastructural localization of collagen types II, IX, and XI in the growth plate of human rib and fetal bovine epiphyseal cartilage: type XI collagen is restricted to thin fibrils. J. Histochem. Cytochem. 43:967–979. Li, S.W., Prockop, D.J., Helminen, H., Fassler, R., Lapvetelainen, T., Kiraly, K., Peltarri, A., Arokoski, J., Lui, H., Arita, M., and Khillan, J.S. (1995) Transgenic mice with targeted inactivation of the Col2 alpha 1 gene for collagen II develop a skeleton with membranous and periosteal bone but no endochondral bone. Genes Dev. 9:2821– 2830. Martin, J.F., Bradley, A., and Olson, E.N. (1995) The paired-like homeo box gene MHox is required for early events of skeletogenesis in multiple lineages. Genes Dev. 9:1237–1249. McLeod, M.J. (1980) Differential staining of cartilage and bone in whole mouse fetuses by alcian blue and alizarin red S. Teratology 22:299–301. Mendler, M., Eich-Bender, S.G., Vaughan, L., Winterhalter, K.H., and Bruckner, P. (1989) Cartilage contains mixed fibrils of collagen types II, IX, and XI. J. Cell Biol. 108:191–197. Metsaranta, M., Garofalo, S., Decker, G., Rintala, M., de Crombrugghe, B., and Vuorio, E. (1992) Chondrodysplasia in transgenic mice harboring a 15-amino acid deletion in the triple helical domain of proa1(II) collagen chain. J. Cell Biol. 118:203–212. Muragaki, Y., Mariman, E.C., van Beersum, S.E., Perala, M., van Mourik, J.B., Warman, M.L., Olsen, B.R., and Hamel, B.C. (1996) A mutation in the gene encoding the alpha 2 chain of the fibrilassociated collagen IX, COL9A2, causes multiple epiphyseal dysplasia (EDM2). Nat. Genet. 12:103–105. Spranger, J., Winterpacht, A., and Zabel, B. (1994) The Type II collagenopathies—A spectrum of chondrodysplasias. Eur. J. Pediatr. 153:56–65. Thom, J.R. and Morris, N.P. (1991) Biosynthesis and proteolytic processing of type XI collagen in embryonic chick sterna. J. Biol. Chem. 266:7262–7269. Vandenberg, P., Khillan, J.S., Prockop, D.J., Helminen, H., Kontusaari, S., and Ala-Kokko, L. (1991) Expression of a partially deleted gene of human type II procollagen (COL2A1) in transgenic mice produces a chondrodysplasia. Proc. Natl. Acad. Sci. U.S.A. 88:7640–7644. Vaughan, L., Mendler, M., Huber, S., Bruckner, P., Winterhalter, K.H., Irwin, M.I., and Mayne, R. (1988) D-periodic distribution of collagen type IX along cartilage fibrils. J. Cell Biol. 106:991–997. Vikkula, M., Mariman, E.C.M., Lui, V.C.H., Zhidkova, N.I., Tiller, G.E., Goldring, M.B., Vanbeersum, S.E.C., Malefijt, M.C.D., Vandenhoogen, F.H.J., Ropers, H.H., Mayne, R., Cheah, K.S.E., Olsen, B.R., Warman, M.L., and Brunner, H.G. (1995) Autosomal dominant and recessive osteochondrodysplasias associated with the COL11A2 locus. Cell 80:431–437.