Immunocytochemical observations on the cornification of soft and hard epidermis in the turtle Chrysemys picta

The process of cornification in the shell and non-shelled areas of the epidermis of the turtle Chrysemys picta was analyzed by light and ultrastructural immunohistochemistry for keratins, filaggrin and loricrin. Beta-keratin (hard keratin) was only present in the corneus layer of the plastron and carapace. The use of a beta-keratin antibody, developed against a specific chick scale beta-keratin, demonstrated that avian and reptilian hard keratins share common amino acid sequences. In both, shelled and non-shelled epidermis, acidic alpha keratin (AE1 positive) was limited to tonofilament bundles of the basal and suprabasal layer, while basic keratin (AE3 positive) was present in basal, suprabasal, and less intensely, pre-corneus layers, but tended to disappear in the corneus layer. The AE2 antibody, which in mammalian epidermis recognizes specific keratins of cornification, did not stain turtle shell but only the corneus layer of non-shelled (soft) epidermis. Two and four hours after an injection of tritiated histidine, the labelling was evenly distributed over the whole epidermis of both shelled and non-shelled areas, but was absent from the stratum corneum. In the areas of growth at the margin of the scutes of the shell, the labelling increased in precorneus layers. This suggests that histidine uptake is only related to shell growth and not to the production of a histidine-rich protein involved in keratinization. No filaggrin-like and loricrin-like immunoreactivity was seen in the carapace or plastron epidermis. However, in both proteins, some immunoreactivity was found in the transitional layer and in the lower level of the corneus layer of non-shelled areas. Loricrin- and filaggrin-like labelling was seen in small organelles (0.05-0.3 mum) among keratin bundles, identified with mucous-like granules and vesicular bodies. These organelles, present only in non-shelled epidermis, were more frequent along the border with the corneus layer, and labelling was low to absent in mature keratinocytes. This may be due to epitope masking or degradation. The immunolabelling for filaggrin was seen instead in the extracellular space among mature keratinocytes, over a material previously identified as mucus. The possibility that this labelling identified some epitopes derived from degraded portions of a filaggrin-like molecule is discussed. The present study suggests that proteins with some filaggrin- and loricrin-immunoreactivity are present in alpha-keratinocytes but not in beta-keratin cells of the shell.