Perinatal glucocorticoid (GC) treatment induces long-term changes in hypothalamo-pituitary-adrenocortical (HPA) function and anxiety-related behaviour in the rat. These changes have been attributed, in part, to locus-specific alterations in GC-receptor expression in the limbic system and elsewhere in the brain and, hence, in GC signalling (Welberg et al., 2001). Work from our laboratory has shown that annexin 1 (ANXA1) is an important paracrine mediator of GC feedback in the HPA axis (John et al., 2004) and that ANXA1-dependent GC feedback in the adult pituitary is blunted by perinatal GC treatment. The present study extended this work by examining the effects of perinatal dexamethasone treatment on the expression and cellular disposition of ANXA1 at two further sites of GC feedback, namely the hypothalamus and hippocampus.

Dexamethasone (dex) was administered to developing rats via the maternal drinking water (1 µg/ml) on days 16-19 of gestation or days 1-7 of post-natal life. Brain tissue was collected at adulthood (90 days) and ANXA1 expression within the cells and on the cell surface determined by western blot analysis and immunogold electron microscopy (EM, n= 4 rats per group). ANXA1 was readily detectable in both hypothalamus and hippocampus by western blot analysis; fractionation studies showed that the bulk of the protein was contained within cells but that a small fraction was localised to the cell surface. Prenatal dex treatment reduced intracellular and cell surface ANXA1 in the hippocampus and cell surface ANXA1 in the hypothalamus. In contrast, neonatal dex increased cell surface ANXA1 in the hippocampus but had no effect on hypothalamic ANXA1. When incubated in vitro, hippocampal and hypothalamic tissue from control animals responded to dex (100nM, 2h) with a marked increase in cell surface ANXA1. The response of both tissues to the steroid was severely blunted by prenatal or neonatal dex treatment. Within the hypothalamus, strong ANXA1 staining was identified by EM in the ependymal cells lining the third ventricle of control and GC-treated rats. Prenatal dex treatment had no effect on ANXA1 expression in these cells in the male but caused a marked reduction in cytoplasmic (3.2 +/- 0.2 vs. 4.7+/-0.08 grains/ µm², P<0.001) and cell surface (0.3 +/- 0.02 vs. 0.4+/-0.04 grains/ µm², P<0.01) ANXA1 in the female. By contrast, the neonatal treatment had no effect on ependymal cell ANXA1 expression in the female but caused a modest increase in cytoplasmic ANXA1 in the male ( 2.9 +/- 0.2 vs. 2.2+/-0.2 grains/ µm², P<0.05). Both treatment regimes induced a significant reduction in the cytoplasmic area of the ependymal cells in male rats (prenatal: 12.4 +/- 1.3 vs. 17.2 +/- 1.6 µm², P<0.001; neonatal: 8.7 +/- 0.7 vs. 17.2 +/- 1.6 µm², P<0.01). By contrast female rats showed an increase in nuclear size ( 12.0 +/- 0.8 vs. 9.4 +/- 0.8 µm², P<0.05) after prenatal but not neonatal dex treatment, but no change in cytoplasmic area.

These results show for the first time that perinatal dexamethasone treatment causes long-term changes in the expression of ANXA1 in the brain and its modulation by GCs. These changes are accompanied by alterations in ependymal cell morphology and are dependent upon both the time of steroid administration and the sex of the rats. The functional implications of these finding remain to be determined but are likely to include changes in HPA function, ependymal cell function and, possibly, behaviour.

John C.D. et al. (2003) Trends in Endocrinology and Metabolism , 15, 103-109.
Wellberg L.A.M et al. (2001) Neuroscience, 104, 71-79 .