Release of prostaglandins in ocular inflammation in the rabbit.

Abstract

VARIOUS studies have suggested that prostaglandins (PGs) may be involved in the ocular response to acute inflammation. PGs are present in ocular tissues1–4 and can reproduce many of the characteristic changes associated with ocular trauma in the rabbit5–7, cat8 and monkey9. We therefore examined the possibility that PGs are involved in the response of the rabbit eye to an acute immunological inflammatory reaction. Eighty-four adult albino rabbits of both sexes were used in 8 groups of 6 to 12, but 11 were discarded because uveitis did not develop and 1 because of infection. Uveitis was induced by a single intravitreal injection of a sterile solution of crystallized bovine serum albumin (British Drug Houses). The left eye was anaesthetized with topical 0·4% oxybuprocaine hydrochloride and bovine serum albumin (10 mg in 0·1 ml. normal saline) was then injected slowly into the anterior part of the vitreous. On subsequent days the eyes were examined with a slit-lamp. Immediately after the injection, small amounts of protein and cells appeared in the aqueous and a slight flare was seen. This was probably due to trauma and the eyes usually appeared normal by about the fifth day. On the 9th-15th day signs of uveitis appeared, characterized by a pronounced flare (due to liberation of large quantities of protein and cells into the anterior chamber) and by marked dilatation of iris and ciliary blood vessels. Occasionally, the pupillary margin was seen to adhere to the anterior surface of the lens (posterior synechiae). The contralateral control eye was usually unaffected, but in about 20% of cases fine filaments (probably strands of fibrin) were seen in the anterior chamber when uveitis occurred in the test eye. At the height of the inflammatory response, the animals were anaesthetized with pentobarbitone (25 mg/kg i.V.). The aqueous humour (approximately 0·2 ml./eye) was then withdrawn through a 25G needle and stored immediately at −20°C. In each group, the aqueous humour samples from the test (left) and control (right) eyes were pooled in separate containers. Biological activity was assayed on the rat fundus preparation10 in Krebs solution containing atropine, methy-sergide and mepyramine (all 2 × 10−7 g/ml.). Unextracted samples were assayed in some experiments, whereas in others the PG-like material was first extracted (chloroform method of Unger et al.11). Alkaline hydrolysis (0·2 M NaOH, 45 min, 37° C) was used to distinguish between PGE and PGF compounds and thin layer chromatography was used to separate the PGs (a II solvent system of Green and Samuelsson12; silica gel plates or impregnated paper13, treated with 3% w/v silver nitrate). Cuts of the chromatograms at and between the RF values of PGE1, PGE2 and PGF2α were eluted with Krebs solution containing excess (0.2 M) NaCl to remove the silver ions prior to extraction for PGs11,13. Samples of unextracted aqueous humour from the test eyes caused contractions (preceded by a relaxation in 1 of the 5 experiments) of the rat fundus preparation. Since atropine, methysergide and mepyramine were present the spasmogen was not acetylcholine, 5-hydroxytryptamine or histamine. The control aqueous had no effect. The material was PG-like since it extracted into chloroform11 when the aqueous humour was acidified and it caused contractions of the rat fundus which were reduced in parallel with PGE2 by the selective PG antagonist SC-1922014,15 (Fig. 1). The biological activity was completely destroyed by alkaline hydrolysis in 3 experiments, indicating E-type prostaglandin (Fig. 1) and approximately halved in another (samples from 2 batches of aqueous humour before chloroform extraction and 2 batches after extraction). PGE2 incubated at the same time was destroyed whereas PGF2α was unaffected.