The skeletal elements are formed by 2 tissues: cartilage and bone. Each of these tissues originates from specific cells: chondrocytes form cartilage, the precursor of bone, whereas osteoblasts form bone. Both cellular types originate from mesenchymal cells (1).

The long bones are formed by endochondral ossification. Mesenchymal cells differentiate into chondrocytes, which later proliferate and become hypertrophic. Eventually the matrix surrounding these cells calcifies, which allows blood vessels to invade. These vessels bring the osteoblasts that produce the bone matrix on this cartilage. The process finishes with the substitution of cartilage tissue for bone tissue.

At birth, the replacement of cartilage by bone has not been completed owing to the entrapment of a layer of cartilage between the epiphysis and the metaphysis. This layer of cartilage is called the growth plate, because this is the last target for the longitudinal growth of the long bones. By the end of puberty, the growth plate is completely replaced by bone, and longitudinal growth ceases (2).

The growth plate is composed of 3 principal layers of chondrocytes: the resting, proliferative, and hypertrophic zones. The resting zone, found under the epiphyseal bone, contains chondrocytes that are relatively quiescent. These cells are spherical, occur singly or in pairs, and are scattered in a bed of abundant cartilage matrix. Distal is the proliferative zone. Here the chondrocytes proliferate constantly. They are flattened and aligned in columns parallel to the long axis of the bone. As they proliferate to the diaphysis, their proliferative capacity is reduced, and they begin to differentiate. Then cell division ceases. Later the chondrocytes undergo hypertrophy. In the hypertrophic zone the chondrocytes are round and secrete matrix vesicles, which are deposited in the extracellular matrix. The vesicles serve as the site for mineralization. The terminally differentiated chondrocytes die, leaving lacunae that later are invaded by blood vessels. Vascularization allows the ingress of stromal cells, chondroclasts, and osteoblasts. Two-thirds of mineralized cartilage matrix is resorbed by chondroclasts; the other one-third serves as a template for deposition of bone matrix by osteoblasts (3).

Alterations in the differentiation and proliferation of chondrocytes, as well as alterations in the extracellular matrix components of the growth plate, can give rise to diverse diseases of the long bones (4,5). Among the diseases more frequently produced by these alterations are osteochondrodysplasias, which are characterized by abnormalities in the growth and development of cartilage or bone or both (6). In humans, there are 26 groups of osteochondrodysplasias, with more than 170 skeletal alterations (7). The most common form of these alterations in humans is achondroplasia, a group of diseases originating from various mutations in theFGFR3 gene and characterized by severe long-bone shortening (8).

In animals, osteochondrodysplasias have been reported mainly in dogs, including the Alaskan Malamute, Scottish deer hound, Norwegian elk hound, Maremmano–Abruzzese shepherd, Great Pyrenees, German shepherd, miniature poodle, Labrador retriever, beagle, and cocker spaniel. Most of the alterations in these dogs result from a de novo mutation, but in some cases they appear to be due to autosomal recessive inheritance (9,10).

Other breeds, such as the bulldog, basset hound, dachshund, Pekingese, and shi tzu, are referred as achondroplastic, since they have a disproportionate rate of dwarfism, and some of them also have a nasal bridge depression and prognathism (11–13). Although the most frequent mutation in achondroplastic humans originates from a G/A transition in nucleotide 1138 of the transmembrane domain of geneFGFR3 (14), a similar mutation seems not to be involved in the bulldog, basset hound, and dachshund with the osteochondrodysplastic phenotype (15).

In veterinary medicine, there is a scarcity of information about the skeletal alterations in animals and no classification like the International Classification of Osteochondrodysplasias in humans. For these reasons, many skeletal dysplasias in animals are named according to their morphologic similarities with the human skeletal dysplasias even though the molecular origins could be different (10,16). The objective of our study was to describe the radiologic characteristics of the long bones and the histopathologic features of their growth plate in the basset hound. For comparison, we also studied Doberman pinschers.

To characterize the form and size of the long bones, we used 3 puppies 2 mo of age and 3 dogs 1 y of age of each breed. Radiographs were made with craniocaudal and medial-lateral views of the humerus, radius, ulna, femur, and tibia. To histologically analyze the growth plate, we euthanized 1 puppy 2 mo of age of each breed by the intravenous administration of sodium pentobarbital, 70 mg/kg. Distal ends of thoracic and pelvic limbs were obtained by cutting 5 to 10 mm proximal to the growth plate on each bone. Samples were cleaned in 1X phosphate-buffered saline and fixed in 4% paraformaldehyde. After 24 h, the samples were decalcified in 5% trichloroacetic acid for 8 d. Longitudinal growth-plate slices 1 × 1 × 3 mm containing epiphyseal and metaphyseal bone ends were obtained. Tissues from the humeral and femoral growth plates were embedded in paraffin and cut in 5-μm-wide slices. After being stained with hematoxylin and eosin, the slices were analyzed by light microscopy.

The 2 breeds showed remarkable radiologic differences in the long bones of both puppies and adults. The basset hound humerus, in comparison with the Doberman humerus, showed a pronounced curvature at the mid-diaphysis, with shortening and widening at the proximal end. The proximal epiphysis was considerably curved, and its caudal end showed a beak-like protrusion. The proximal and distal metaphysis showed considerable widening that diminished towards the medial diaphysis, in a truncated cone fashion. The opacity of the caudal cortical bone was increased; this was most evident in the medial third of the diaphysis (Figures 1A and 1B).

The radius and ulna in the basset hounds were short and wide, with a broad interosseous space owing to the pronounced curvature at the medial third of the radial diaphysis. The distal metaphysis of the radius was severely widened and denser than that of the Dobermans. The ulna was wider than in the Dobermans and had prominent protuberances; its diaphysis was slightly widened in the distal third, as was the proximal part, between the anconeus and olecranon processes (Figures 1C and 1D).

The femur in the basset hounds appeared shortened and widened at the proximal and distal metaphyses, with increased opacity. The trochlear and condylar regions were wider than in the Dobermans, although without other anatomic changes. The neck of the femoral head was shorter, and its epiphysis was cap-like, with a salient rim. The major trocanter was widened at its borders, and the minor trocanter protruded medially. The head of the femur was in good proportion to the acetabulum; thus, the coxofemoral articulation was correct (Figures 2A and 2B).

The tibia in the basset hounds appeared shortened and widened, with notably increased opacity and widening of the proximal metaphyses. The proximal third of the diaphysis was concave, more so at the cranial line. The fibula was not parallel to the tibia; thus, the interbone space was larger in the basset hounds than in the Dobermans (Figures 2C and 2D).

Histologically, the height of the entire growth plate was markedly greater in the Doberman than in the basset hound (Figures 3A and 3B). In the resting zone, the shape of the cells was similar in the 2 breeds, but in the basset hound the zone was larger, and the extracellular matrix was abundant. The proliferating zone in the Doberman was organized in typical symmetric columns formed by single cells stacked on one another. In contrast, in the basset hound this zone contained disorganized columns: some were short and thinly formed by piling up of a single cell, whereas others were formed by piling up of more than 1 chondrocyte; the columns were not symmetric and were separated by wide areas of cartilage matrix. In the Doberman, the growth-plate cells near the diaphysis were hypertrophic and mostly of consistent size, although shrunken cells were also observed; the columns stayed organized and separated by a small amount of cartilage matrix. In contrast, in the growth plate of the basset hound, the hypertrophic zone was greatly diminished and composed of columns of cells of variable size. Next to the hypertrophic chondrocytes in the growth plate of the Doberman, we observed a large quantity of chondrocyte lacunae, with a few shrunken cells and scarce cellular matrix; in the transition zone were osteoblasts, osteoclasts, erythrocytes, and vascular ingrowth, as well as cartilage areas occupied by osteoblasts but no primary bone formation. In the growth plate of the basset hound cellular lacunae were sparse, and in the transition zone we observed early vascular ingrowth, a great quantity of osteoblasts, and abundant primary bone formation on the cartilage matrix. These characteristics in the basset hound suggest an alteration in chondrocyte proliferation and premature ossification.

Our study revealed that basset hounds have curved long bones that shorten and bow the thoracic and pelvic limbs. In addition, the metaphyses are wide and the epiphyses irregular. However, the dogs do not have the craniofacial and vertebral alterations that affect achondroplastic humans. For this reason, the basset hound skeleton is not typical of human achondroplasia. Other osteochondrodysplasias, such as hypochondroplasia, pseudoachondroplasia, and metaphyseal dysplasia, do not present craniofacial alterations. We found, by histopathologic analysis, that the abnormalities of the growth plate in the basset hound are similar to those in humans with achondroplasia and mice with experimentally induced achondroplasia (17). However, these histologic findings are also similar to those observed in humans and mice with metaphyseal dysplasia (18). Therefore, the histologic characteristics of the growth plate alone are insufficient to classify the type of osteochondrodysplasia that basset hounds have.

At present, osteochondrodysplasias in humans are classified on the basis of molecular diagnostics (7). In veterinary medicine, the data are too sparse for a similar classification in animals. We previously reported that basset hounds do not have the mutation usually found in humans with achondroplasia (15). We suggest that the basset hound be identified as osteochondrodysplastic, since osteochondrodysplasia is a term widely used to describe the alterations in cartilage and bone. Studies on the pathological mechanisms in the growth plate of the basset hound and the search for mutations in genes related to skeletal development in these dogs will help us define the type of osteochondrodysplasia. We propose 2 candidate genes that might be associated with the typical osteochondrodysplasia of the basset hound: theFGFR3 gene, whose complete sequence could be obtained to eliminate the possibility of a mutation in a region other than the transmembrane domain; and the collagen X gene, which is expressed exclusively in hypertrophic chondrocytes in the growth plate and is responsible for the Schmid type of metaphyseal dysplasia in humans, in which hypertrophic cartilage is absent (19).

Identifying the gene responsible for canine osteochondrodysplasias and studying its mechanisms of action will provide clues to the development of the skeletal system in mammals.