Analysis of human articular chondrocyte CD44 isoform expression and function in health and disease.

Interactions between articular chondrocytes and components of the extracellular matrix are of potential importance in the normal function of cartilage and in the pathophysiology of arthritis. Little is known of the basis of these interactions, but cell adhesive molecules such as CD44 are likely to be involved. Immunohistology using six well-characterized anti-CD44 monoclonal antibodies demonstrated standard CD44 isoform (CD44H) expression by all chondrocytes in normal and osteoarthrotic (OA) cartilage but absence of the CD44E variant. Polymerase chain reaction (PCR) of reverse transcribed mRNA from monolayer cultures of normal and OA chondrocytes using primer sequences which span the region containing variably spliced exons produced a predominant band representing the standard form of CD44, which lacks the variable exons 6-15 (v1-v10). No product was seen at the expected size of the epithelial variant of CD44 (CD44v8-10). Use of exon-specific primers, however, showed expression of variant exons resulting in multiple minor isoforms. Standard CD44 was also shown to be the predominantly expressed isoform identified by immunoprecipitation, but human articular chondrocytes did not adhere to hyaluronan in vitro. Chondrocyte CD44 may function as an adhesion receptor for other matrix molecules such as fibronectin or collagen.

Chondrocyte heterogeneity: immunohistologically defined variation of integrin expression at different sites in human fetal knees.

During development and at maturity different forms of cartilage vary in morphology and macromolecular content. This reflects heterogeneity of chondrocyte activity, in part involving differential interactions with the adjacent extracellular matrix via specialized cell surface receptors such as integrins. We undertook an immunohistological study on a series of human fetal knee joints to assess variation in the expression of integrins by chondrocytes and potential matrix ligands in articular, epiphyseal, growth plate, and meniscal cartilage. The results show that articular chondrocytes (beta 1+, beta 5 alpha V+, alpha 1+, alpha 2+/-, alpha 5+, weakly alpha 6+, alpha V+) differed from epiphyseal (beta 1+, beta 5 alpha V+, alpha 1+/-, alpha 2+/-, alpha 5+, alpha 6+, alpha V+) growth plate (beta 1+, beta 5 alpha V+, alpha 1-, alpha 2-, alpha 5+, alpha 6+, alpha V+), and meniscal cells (beta 1+, beta 5 alpha V+, alpha 1+, strongly alpha 2+, alpha 5+, alpha 6+, alpha V+ in expression of integrin subunits. There was no expression of beta 3, beta 4, beta 6, or alpha 3 by chondrocytes. These results differ from previous reports on the expression of integrins by adult articular cartilage, where alpha 2 and alpha 6 are not seen. Variation in distribution of matrix ligands was also seen. Fibronectin, laminin and Type VI collagen were expressed in all cartilages but there was restricted expression of tenascin, ED-A and ED-B fibronectin isoforms (articular cartilage and meniscus), and vitronectin (absent from growth plate cartilage). Regulated expression of integrins by chondrocytes, associated with changes in the pericellular matrix composition, is of potential importance in control of cartilage differentiation and function in health and disease.