Does chondroitin sulfate have a role to play in the morphogenesis of the chick primary corneal stroma?

ABSTRACT

This paper makes three points about how the chick corneal epithelium lays down the primary stroma, an orthogonally arranged array of well-spaced, 20-nm-diameter collagen fibrils. (1) Isolated corneal epithelia will, when cultured, lay down de novo stromas whose fibril-diameter distribution, fibril spacing, and proteoglycan profile are similar to those laid down in vivo. They differ from embryonic stromas in two ways first, much of the chondroitin sulfate is released to the medium and, second, there is a relatively small amount of orthogonal organization. Epithelia seem only to lay down such stromas if they are separated from their original stromas with dispase, which leaves an intact basal lamina, and spread out, basal lamina downward, on a Nuclepore filter (poresize, 0.1 micron). (2) Chondroitin sulfate (CS), the predominant proteoglycan (greater than 85%), seems to play no significant role in collagen fibrillogenesis in vitro. Stromas laid down in its absence were indistinguishable from controls as assayed by fibril diameter, organization, and spacing and the amount of collagen synthesized. For these experiments, epithelia were cultured in the presence of hyaluronidase, which degrades CS, and p-nitrophenyl beta-D-xyloside, which inhibits the formation of links between the core protein and glycosaminoglycan side chains in the PG; the absence of intact CS was confirmed by gel filtration. We suggest that, in vivo, CS may facilitate the interfibrillar movement that takes place as the cornea grows. We have also found that keratinase, which degrades the very small amount of keratan sulfate present in the primary stroma, has no effect on stromal deposition. (3) There are substantial amounts of unidentified matrix components in primary stromas laid down both in vivo and in vitro. This conclusion was drawn from SEM observations on both types of stroma after they had been freeze-dried, a process which does not condense hydrated macromolecules. Even after being treated with hyaluronidase to remove the CS, substantial amounts of interfibrillar matrix were still present. Until these components are identified and their interactions with collagen are understood, the mechanisms responsible for stromal morphogenesis are unlikely to be understood.