The effects of the vitamins dl--tocopherol, ascorbic acid and -carotene, free radical scavengers and lipid peroxidation inhibitors, were analyzed in male Wistar rats made goitrous by feeding a low iodine diet (<20 µg iodine/kg) and perchlorate (1% in drinking water) for 4, 8, 16 and 32 days. Groups of control or goitrous rats received for at least 16 days before killing a diet containing 0·6% vitamin E (as dl--tocopherol acetate), 1·2% vitamin C (ascorbic acid) and 0·48% -carotene, either simultaneously (vitamin cocktail) or separately. This treatment led to a 5-fold increase of vitamin E in the thyroid gland, a 24-fold increase in the liver and a 3-fold increase in the plasma. In control rats, vitamin cocktail administration increased slightly the thyroid weight with little changes in thyroid function parameters. During iodine deficiency, administration of the vitamin cocktail or vitamin E alone reduced significantly the rate of increase in thyroid weight, and DNA and protein contents, as well as the proportion of [ 3 H]thymidine labeled thyroid follicular cells, but not that of labeled endothelial cells. Plasma tri-iodothyronine, thyroxine, TSH levels, thyroid iodine content and concentration as well as relative volumes of glandular compartments were not modified. The proportion of necrotic cells rose from 0·5% in normal animals to about 2% after 16 days of goiter development. No significant protective effect of the vitamins was observed. These results suggest that these vitamins, particularly vitamin E, modulate one of the regulatory cascades involved in the control of thyroid follicular cell growth, without interfering with the proliferation of endothelial cells.

Necrosis and apoptosis coexist in the thyroid during goitre development and involution, but little is known about their respective causes. To test the possible role of free radicals, we analysed separately necrosis and apoptosis in male Wistar rats with depressed or normal antioxidant protection. Vitamin E-deficient and -sufficient rats were made goitrous with perchlorate in drinking water; involution was induced by repeated injection of NaI, without or with methimazole. Increase of thyroid malondialdehyde concentration and decrease of glutathione peroxidase activity confirmed the depressed antioxidant protection in vitamin E-deficient rats. Plasma thyroxine and TSH levels were not modified. Necrosis (swollen cells) and apoptosis (pyknotic cells) were quantified on histological sections. In vitamin E-sufficient rats, dead cells were very rare in control thyroids, increased 3-fold in goitre and still further during involution.Necrotic epithelial cells predominated in the goitre and their number declined after iodide supplementation, without or with methimazole. In contrast, the number of apoptotic cells and the caspase-3 activity were increased in goitre and further increased after involution, with twothirds of pyknotic cells being observed in the interstitium. Apoptosis was prevented by methimazole. Vitamin E deficiency significantly increased total cell death and epithelial cell necrosis and induced the occurrence of much cell debris in the follicular lumen during involution, with no modification of the apoptotic reaction. These results show that the type of cell death is differentially regulated during goitre development and involution: necrosis is related to the oxidative status of the cells, while apoptosis comes with iodine-induced involution.

Nitric oxide (NO) is a well-known mediator of autoimmune processes. In the thyroid gland, it is produced in response to interleukin 1 (IL-1) and may mediate cytokine action at an early stage of autoimmune thyroiditis. In this study, we have investigated whether NO is involved in cytokine-induced cytotoxic effects and epithelial barrier alterations in thyrocytes. Human thyroid epithelial cells were cultured as tight polarised monolayers on a permeable support and exposed or not to IL-1 (100 U/ml), alone or in combination with interferon-(IFN-; 100 U/ml) added to the basal compartment. NO production was not detected in control thyrocytes, but was significantly induced by the combination of IL-1 with IFN-, in a time-dependent fashion. Similarly, expression of the inducible isoform of nitric oxide synthase (NOSII), determined by immunoblot and immunofluorescence confocal microscopy, was not detected in control cells, but was markedly induced after 48-h exposure to both cytokines. This treatment significantly increased the release of cytosolic lactate dehydrogenase (LDH) in the apical and basolateral media and decreased transepithelial electrical resistance. Although IFN-was not sufficient to induce NO production, it could by itself decrease transepithelial resistance and synergised the IL-1 effect on LDH release. The NOS inhibitor, -nitro-argininemethyl ester, suppressed the cytokine-induced NO production and decreased the LDH release, but failed to prevent the loss of transepithelial resistance. These results indicated that human thyrocytes express NOSII and produce NO in response to IL-1 +IFN-and suggest that NO acts as a mediator of cytokine-induced cytotoxicity in the thyroid gland and may promote the exposure of autoantigens to the immune system. In contrast, NO does not appear to mediate the cytokine-induced disruption of the thyroid epithelial barrier.