Preliminary studies were conducted to determine the best fixation, embedding, and staining methodology to minimize interference with the vitreoretinal interface on LM and on EM. The resultant protocol is summarized in this article. Additional studies were carried out to ensure that the vitreoretinal interface would not undergo any alteration caused by autolysis at the maximal incubation period used in these studies. With fresh porcine globes, it was determined that autolysis becomes significant at 24 hours of incubation at room temperature.
Slow progressive dehydration and the use of whole globes was felt to be critical to avoid artifactual separation of the vitreous from the retina. To stop enzymatic reaction and to facilitate intraocular fixation, a 4- to 6-mm scleral opening was made parallel to the pars plana at least 90° away from the site of injection, and between 0.2 mL and 0.4 mL fixative was injected directly into the eye, toward the midvitreous. The eyes were left in a mixture of 1.25% glutaraldehyde and 1% paraformaldehyde in 0.08 M cacodylate buffer (pH 7.4) at room temperature for 3 days. Then they were slowly dehydrated in progressively higher concentrations (up to 100%) of ethanol over a 4-week period.
After complete dehydration, the eyes were cut in two equal parts with a sharp scalpel blade. The incision was placed in the anteroposterior axis of the injection site. Photographs were taken of the eyecups. Half of each eye was used for LM, and the other half was used for scanning EM (SEM). The specimen for LM was placed in embedding resin (Technovit 7100; Heraeus Kulzer GmbH, Wehrheim/ts, Germany) in accordance with the manufacturer’s instructions. Serial 4-μm thick slices were obtained, placed on a glass slide, and stained with 1% toluidine blue. The specimen for EM was immersion dried in hexamethyldisilazane (Sigma-Aldrich, St. Louis, MO), mounted with carbon glue on special stubs, and coated with ±8 nm platinum. Slides were examined and photographed under an SEM (XL20; Philips, Eindhoven, Netherlands).