Chondroitin sulfate proteoglycans are structural renewable constituents of the rabbit vitreous body.

PURPOSE: To characterize the vitreous intrinsic proteoglycans, investigate their dynamics, and examine their role in the supramolecular organization of the vitreous. METHODS: Vitreous from normal rabbits was collected and processed for observation with the transmission electron microscope after treatment with glycosidases. Also, rabbits were injected intravitreally with [35S]-sodium sulfate and sacrificed at several time intervals after the injection. Proteoglycans (PGs) were assayed in the vitreous supernatant or in whole samples extracted with guanidine hydrochloride by polyacrylamide or agarose gel electrophoresis, followed respectively by fluorography or autoradiography, and ion-exchange chromatography and gel-filtration chromatography, combined with glycolytic treatment of the samples. The sulfated glycosaminoglycans (GAGs) were characterized by agarose gel electrophoresis after treating vitreous samples with protease and specific glycosidases. RESULTS: The electron microscopic study revealed a network with hyaluronic acid (HA) as thin threads coating and connecting collagen fibrils. The elimination of the HA coat showed chondroitin sulfate granules (8-25 nm) arranged at regular intervals on the fibril surface. The chondroitinase ABC digestion, besides removing the granules, also caused the formation of thicker bundles of the collagen fibrils. The PG and GAG analysis indicated that there are three renewable PGs in the vitreous (e.g., one heparan- and two chondroitin-sulfate ones). CONCLUSIONS: At least one of the chondroitin sulfate PGs is involved in the interactions that occur in the vitreous structure, mainly by providing adequate spacing between the collagen fibrils, a condition that is probably required for the transparency of the vitreous.

Intravitreal injection of dispase causes retinal hemorrhages in rabbit and human eyes.

PURPOSE: To check the effects of intravitreally injected dispase in the vitreo-retinal region. METHODS: Dispase, 0.05 to 2.5 units dissolved in 100 microl of phosphate-buffered saline, was injected into the midvitreous of rabbits which were killed from 15 to 120 min afterwards. The enzyme was also injected into four human eyes of patients with orbital tumors 15 min before enucleation during orbital exenteration surgery. The eyes were examined in vivo as well as by light and electron microscopy. RESULTS: Hemorrhages were detected by fundus observations and confirmed by microscopical analysis in nearly all rabbits and in half of the human eyes. The red blood cells were observed in the vitreous and retina. Breaches in the inner limiting membrane were visualized in human eyes and ruptures of small blood vessels in rabbit eyes. In spite of that, vitreous detachment was not verified. In fact, the cortical-vitreous collagen-fibril network was conspicuous on scanning electron micrographs. CONCLUSIONS: Retinal hemorrhages were evident as early as 11 min after injection. It is suggested that this enzyme degraded selectively basement membrane components without affecting other proteins involved in the vitreous-retinal junction.