Morphometry of cupromeronic blue-stained proteoglycan molecules in animal corneas, versus that of purified proteoglycans stained in vitro, implies that tertiary structures contribute to corneal ultrastructure.

Isolated, purified small chondroitin (dermatan) sulphate proteoglycans from corneas of cow and rabbit and cow sclera were stained with Cupromeronic blue in 'model' experiments. The lengths and thicknesses of the images were compared with those of the same proteoglycans stained in the tissue, using the critical electrolyte concentration principle to give specificity for sulphated proteoglycans, and keratanase 1 or chondroitinase ABC digestion to distinguish between chondroitin and keratan sulphate. Corrections for orientation of the stained glycan filaments within the section plane were made to convert the observed lengths to true average lengths. Observed lengths of stained chondroitin (dermatan) sulphate were greater than those of keratan sulphate, both in models and tissues, in agreement with published data from biochemical and rotary-shadowing studies, in both species. Corrected (true) average lengths of stained isolated chondroitin (dermatan) sulphate proteoglycans were slightly, but not significantly, longer than expected from rotary shadowing or biochemical measurements. Keratan sulphate lengths were similarly somewhat longer. The data support the idea that Cupromeronic blue acts as a scaffold that helps maintain polyanion shape against distortion on staining. Stained filaments in tissues were sometimes over twice the length of isolated stained proteoglycans, suggesting that 2 glycan chains were aligned end-to-end. Thicknesses of proteoglycan filaments suggested that at least 2 glycan chains were aligned side-by-side, both in models and in tissues. A scheme for proteoglycan tertiary structure in cornea is proposed, in which glycan chains may bridge collagen fibrils in duplexed forms similar to those observed in rotary shadowed preparations.