Detection of age-related changes in the distributions of keratan sulfates and chondroitin sulfates in developing chick limbs: an immunocytochemical study

Keratan sulfate is a glycosaminoglycan that has been identified as a matrix component of two types of connective tissues, corneal stroma and cartilage matrix (Hascall & Hascall, 1981; Hassell et al. 1986). These keratan sulfates are substituted to core proteins that may, particularly in cartilage, also be substituted with chondroitin sulfates (Hassell et al. 1986). More recently, the application of immunochemical and immunocytochemical methods has resulted in the identification of keratan sulfate in a much wider variety of tissues (Hyldahl et al. 1986; Funderburgh et al. 1987; Sorrell et al. 1988b,c). Furthermore, use of anti-keratan sulfate monoclonal antibodies has revealed differences in epitope expression among keratan sulfates (Caterson et al. 1986; Sundarraj et al. 1986; Funderburgh et al. 1987, 1988). Developmental studies, in which more than one anti-keratan sulfate monoclonal antibody was used, have demonstrated both age- and tissue-specific differences in keratan sulfates. SundarRaj et al. (1986) have shown that keratan sulfate epitopes appear in a developmentally defined manner in the rabbit eye. Funderburgh et al. (1988) found that keratan sulfates in scar tissue from corneal wounds contained a structurally different form of keratan sulfate than that found in normal cornea. Differences in epitopes may be related to variations in sulfation patterns within glycosaminoglycans (Mehmet et al. 1986; Tang et al. 1986).

In the present study, four separate anti-keratan sulfate monoclonal antibodies were used to immunocytochemically identify keratan sulfate in developing chick hind limbs. These antibodies were used, not only to identify keratan sulfate, but to determine whether there were spatial-, age-, or epitope-specific differences. Keratan sulfate was identified as early as day 5·5 of incubation. At that age, it was largely confined to joints and articular cartilages, with lesser amounts in epiphyseal cartilages. These results correspond to the identification of keratan sulfates in tendon, ligament, and other noncartilaginous components of joints (Craig et al. 1987b; Vogel & Evanko, 1987; Daniel & Mills, 1988). With advancing age, this distribution of keratan sulfate gradually changed. Extracellular keratan sulfate now appeared in central regions of bones.

The four anti-keratan sulfate monoclonal antibodies differed in their ability to detect keratan sulfate in some regions of developing bones. This was particularly evident in the intracytoplasmically identified keratan sulfate of younger bones. However, differential keratan sulfate epitope expression continued to be present in cartilage of older bones. Thus, this study indicates that keratan sulfate is an early component of developing chick bones and that its expression, like that of chondroitin sulfates (Kimata et al. 1974; Stocum et al. 1979; Shinomura et al. 1984), appears to be developmentally regulated.

All four anti-KS monoclonal antibodies, 5D4, 1B4, 2D3, and 4A1, used in this study were raised against purified mammalian cartilage proteoglycans. The four anti-keratan sulfate antibodies were characterized using a competitive radioimmunoassay in which ¹²⁵I-bovine nasal cartilage-ABC-core was the labelled antigen and keratan sulfate type I (Sigma) and keratan sulfate type II (Martin & Mathews NIH standard) were the unlabelled antigens (Caterson et al. 1983). The data were expressed as nanograms of unlabelled antigen required to give a 50 % inhibition.

The two monoclonal, anti-chondroitin/dermatan sulfate antibodies, 2B6 and 3B3, were raised against chondroitinase ABC-digested cartilage proteoglycans. These antibodies recognized chondroitin-4- and dermatan sulfates and chondroitin-6-sulfate, respectively (Caterson et al. 1985, 1986).

Tissues were taken from outbred lines of White Leghorn chick embryos at the ages indicated in the results section. At earlier stages, three embryos were used at each time point; two each were used for later stages. Tissue processing and indirect immunoperoxidase immunocytochemistry were performed as previously published (Sorrell et al. 1988a). None of the immunostained samples that are present in the text were counterstained. However, other samples were counterstained with hematoxylin to facilitate localization of intracellular epitopes.

Separate controls were performed for each antibody and each tissue. These included sections in which the primary antibody incubation step was omitted and sections in which the primary antibody incubation was performed, but in the presence of I mg ml^-1 of purified corneal keratan sulfate (type I). No immunostain was detected in either situation.

The distribution of keratan sulfate glycosaminoglycans was studied, using an indirect immunoperoxidase method, in embryonic chick hind limbs. Four separated monoclonal antibodies, each antibody recognizing epitopes present in both keratan sulfate type I and type IT, were used. Each antibody varied in its ability to recognize epitopes on these two forms of keratan sulfate (Table 1). This suggests that these antibodies may recognize subtle epitope differences within glycosaminoglycan chains.

All four antibodies identified what appeared, at low magnification, to be discrete foci of immunostain in day 5· 5 and day 6·5 limbs (not shown). At higher magnification, these foci were found to correspond to joints and articular surfaces of the knee (see below), tibiotarsal, and metatarsophalangeal joints (not shown). Lesser, diminishing amounts of keratan sulfate were detected in epiphyseal and diaphyseal regions of both femurs and tibias.

Others have already described the distribution of keratan sulfate in the metatarsophalangeal joint (Craig et al. 1987a); therefore, this study has concentrated on results obtained from the knee joint. The structure of the anterior region of a knee joint obtained from a day 6· 5 embryo is shown in Fig. 1. The joint capsule, articular cartilages, femoral and tibial heads are identified. The joint cavity has not yet formed; consequently, the femur and tibia are still connected by cells that will subsequently degenerate and die to form the joint space. This latter is termed the interzone region (Holder, 1977).

All four anti-keratan sulfate monoclonal antibodies identified high levels of keratan sulfate in the joint region (Fig. 2A). Intense zones of immunostain were confined to articular cartilages and to a band, which apparently connected the femoral and tibial heads, that extended along the apical margin of the joint capsule. This band may represent either ligament or developing patellar tissue (Fell & Canti, 1934). Both the interzone region and the region immediately subjacent to the immunostained apical band lacked detectable keratan sulfate. Also, no keratan sulfate was detectable in the posterior aspect of the joint capsule (not shown). Lesser, but detectable, levels of keratan sulfate were present in epiphyseal cartilages. This nonuniform distribution of keratan sulfate did not reflect the distribution of all glycosaminoglycans, since immunostaining using anti-chondroitin sulfate antibodies 3B3 and 2B6 produced very different patterns (Fig. 2B,C). Chondroitin-6 sulfate (3B3) was intensely detected in joint and articular regions, including regions where there was no detectable keratan sulfate (compare Fig. 2A and B). In contrast, chondroitin-4/dermatan sulfates appeared to be generally absent from these sites (Fig. 2C). These results suggest that some chondroitin sulfate proteoglycans in the joint may be substituted with both chondroitin and keratan sulfate glycosaminoglycans, while others may be substituted with only chondroitin sulfates. It also indicates that a differential distribution of chondroitin and keratan sulfates occurs at this early stage of development.

The abrupt transition between articular and epiphyseal cartilages was due to the amount of extracellular keratan sulfate. In contrast to the articular zone, only a thin halo of pericellular keratan sulfate was seen surrounding many epiphyseal chondrocytes (Fig. 3). In the diaphyseal region, even this pericellular immunostain was not detected by any of the anti-keratan sulfate antibodies (bottom of Fig. 3 and Fig. 4). Based upon studies of counterstained samples (not shown), keratan sulfate epitopes were restricted to a single juxtanuclear site. This was not evident in all cells, possibly due to the fact that this is sectioned tissue and that some juxtanuclear sites may have been out of the plane of section (Fig. 4). Three out of four anti-keratan sulfate antibodies identified this intracellular keratan sulfate (see below). This pattern of keratan sulfate distribution translated, at lower magnifications, to a gradient in keratan sulfate content. As in the joint region, the apparent lack of extracellular keratan sulfate in diaphyseal regions was not due to the lack of proteoglycans, since anti-chondroitin sulfate antibodies readily detected extracellular glycosaminoglycans in this zone (Fig. 5).

The absence of extracellular keratan sulfate in the diaphyseal region appeared to be age-related, since extracellular keratan sulfate could be detected within this region of bones obtained from older embryos. By day 9·5, both pericellular and extracellular keratan sulfate were detected by all four anti-keratan sulfate antibodies (Fig. 6). In fact, by day 11·5, keratan sulfate was intensely detected in diaphyseal regions (Fig. 7). In other words, the high content of diaphyseal keratin sulfate persisted in bones of older embryos, as did the high content of keratan sulfate in articular cartilages and joint tissues, such as the meniscus (Fig. 8). Thus, the distribution of keratan sulfate appears to be both spatially and age-related.

Differences in epitope expression were detected by comparing the results of the four anti-keratan sulfate monoclonal antibodies. In younger embryos, all four monoclonals detected epitopes in joints, articular and epiphyseal regions. However, there were distinct differences in keratan sulfate epitopes detected in middle regions of bones. Antibody 4A1 provided the strongest immunocytochemical detection of intracytoplasmic epitope; while antibodies 2D3 and 5D4 both provided intermediate detection; and, finally, antibody 1B4 barely detected epitope (Fig. 9). These results strongly suggest that a differential expression of keratan sulfate occurs at this site. Similarly, differences in keratan sulfate epitope expression were noted in bones obtained from older embryos. As shown in Fig. 10, antibodies 2D3 and 5D4 differentially identified keratan sulfate in femoral condyle from a day 18 embryo. Both monoclonals detected cell-associated epitope, but only antibody 2D3 detected extracellular epitope, particularly in articular cartilage. Thus, different forms of keratan sulfate likely exist and can be discriminated immunologically. Furthermore, their expression appears to be related to both the location of the chondrocyte producing the epitope and the developmental age of that chondrocyte.

The most striking feature observed in developing hind limbs of chick embryos was the high content of keratan sulfate, as determined by immunocytochemical analyses. Even more interestingly, keratan sulfate was not uniformly distributed throughout cartilage, but was concentrated within joints and developing articular cartilage surfaces. In contrast, chondroitin/dermatan sulfates were more widely distributed throughout all regions of cartilage. The highest extracellular concentrations of keratan sulfate corresponded to regions that also contained a high content of chondroitin-6 sulfate, but low chondroitin-4/dermatan sulfate content. Taken together, these observations indicate that metabolic heterogeneity, in terms of glycosaminoglycan synthesis, occurs, and that these different populations of cells are highly organized into identifiable regions.

Similar conclusions, based upon data obtained from biochemical studies of cartilages, have been reached by others (De Luca et al. 1977 ; Garg et al. 1981; Pacifici et al. 1981; Webber et al. 1987). The relative contents of chondroitin and keratan sulfate were found to vary according to location and to depth within cartilage (Franzen et al. 1981; Zanetti et al. 1985; Williams et al. 1988). Vasan & Lash (1978) found that articular cartilage of embryonic chick tibias contained a higher keratan sulfate content than did epiphyseal cartilage. This was confirmed in this study; furthermore, the present data indicate that equally high contents of keratan sulfate were located in some of the noncartilaginous tissues of the developing joint.

In bones from younger embryos, a gradient in keratan sulfate content was observed that, upon closer inspection, proved to be related to the amount of extracellular keratan sulfate. In bones from younger embryos, extracellular keratan sulfate was detected in articular and epiphyseal regions. Even between these latter two zones remarkable differences in the amounts of keratan sulfate were noted; considerably more extracellular keratan sulfate was present in the articular zone. In the zones containing flattened cells and hypertrophic cells, keratan sulfate was restricted to intracytoplasmic, juxtanuclear regions, presumably corresponding to the Golgi apparatus, as has been previously suggested (Vertel & Barkman, 1984). Vertel & Barkman (1984), in a study of cultured embryonic chick chondrocytes, found that, while all of these cells produced extracellular chondroitin sulfate, only some produced detectable extracellular keratan sulfate. However, keratan sulfate could be detected in some of those cells that did not display this glycosaminoglycan extracellularly. From this they concluded that chondroitin and keratan sulfate synthesis and secretion were independently regulated. Zanetti et al. (1985), in a separate study, reached similar conclusions. In intact bones, a similar independent regulation also appears to exist, but here, regulation seems to depend upon the location of chondrocytes within the developing bone. These results indicate that the differential ability of chondrocytes to store or to secrete keratan sulfate relates to developmental age.

Substantial, age-related increases in keratan sulfate content have been documented in both mammalian and avian cartilages. This increase is accomplished by the substitütion of more, and longer, keratan sulfate chains to chondroitin sulfate proteoglycans (De Luca et al. 1977; Garg et al. 1981; Pacifici et al. 1981; Webber et al. 1988). In the chick embryo, increased keratan sulfate was immunocytochemically detected by day 9·5. At that time, extracellular keratan sulfate was noted in the diaphyseal regions of both femurs and tibias. Extracellular keratan sulfate persisted in all regions of cartilage in older embryos. Thus, the present data indicate that most chondrocytes are capable of synthesizing keratan sulfate, as evidenced by its intracytoplasmic detection. However, the secretion of keratan sulfate, and/or its extracellular metabolism, may be independently regulated.

With the development of anti-keratan sulfate monoclonal antibodies has come the realization that keratan sulfate is much more widely distributed than had previously been expected (Funderburgh et al. 1987). Recent immunocytochemical studies have identified keratan sulfate in such noncartilaginous tissues as tendons, figaments, and other joint tissues (Craig et al. 1987a,b). In addition, the development of significant numbers of different anti-keratan sulfate monoclonal antibodies has revealed that considerable heterogeneity exists within keratan sulfate chains, indicating that keratan sulfate glycosaminoglycan structure is not simply related to keratan sulfate type I or type II (SundarRaj et al. 1986; Funderburgh et al. 1987, 1988). At least some of this heterogeneity is likely related to variations in sulfation patterns that occur within these chains (Mehmet et al. 1986; Tang et al. 1986; Funderburgh et al. 1987). Direct evidence for differences in keratan sulfate epitopes has been documented for antibodies 5D4 and 1B4. Antibody 5D4 appears to recognize a more highly sulfated epitope than does antibody 1B4 (Mehmet et al. 1986). Others have also suggested that sulfation of keratan sulfate is an important factor in tissue and developmental distribution of keratan sulfate (SundarRaj et al. 1986; Funderburgh et al. 1988). The differential detection of keratan sulfate by monoclonal antibodies in chick cartilage suggests that the types of keratan sulfate are developmentally related. Too little is yet known to make predictions about the role that different forms of keratan sulfate might have in influencing the structure of extracellular matrices.

In conclusion, keratan sulfate glycosaminoglycans are present in embryonic chick limbs at very early stages. Their synthesis and distribution, like that of chondroitin/dermatàn sulfate glycosaminoglycans (Shinomura et al. 1984), appears to be both spatially- and age-related. Presumably, this differential keratan sulfate content has important influences on the structure and function of extracellular matrices present in different locations.

This work was supported by NIH grants CA40059, RR 05433, and AR32666.