Further evidence for a membrane receptor that binds glucocorticoids in the rodent hypothalamus

Abstract

In parallel with their well-characterized delayed genomic effects, steroid hormones exhibit rapid, non-genomic effects at molecular, cellular and behavioral levels. We have proposed a model of rapid, non-genomic glucocorticoid inhibition of hypothalamic neuroendocrine cells through a putative membrane-associated glucocorticoid receptor (GR). Here we tested for plasma membrane GR immunoreactivity and binding in the hypothalamic supraoptic and paraventricular nuclei. Selective cross-linking of membrane proteins with membrane-impermeant BS³ and subsequent Western blot analysis with a monoclonal GR antibody revealed a reduction in the intensities of a ∼98kDa immunoreactive band and a ∼64kDa band in the rat paraventricular and supraoptic nuclei, and of a 64kDa band in hippocampal tissue, which suggested that these proteins are associated with the membrane. Saturation binding of [³H]-corticosterone and [³H]-dexamethasone in rat and mouse hypothalamic tissue revealed a K_d 4-24-fold lower and a B_max 4-7-fold lower for the membrane-associated GR compared to the intracellular GR, suggesting a lower affinity and abundance of the glucocorticoid binding sites in the membrane than in the cytosol. Together, these findings suggest the presence of a low-affinity, low-abundance membrane-associated GR in the hypothalamus that shares homology with the intracellular GR, and are consistent with physiological evidence of rapid, non-genomic glucocorticoid actions in hypothalamic neuroendocrine cells that are GR dependent.

Keywords: Corticosteroid; Glucocorticoid receptor; HPA; Hypothalamus; Membrane cross-linking; Stress.

Figure 1. Association of GR immunoreactivity with membrane proteins following BS³ cross-linking

Western blot analysis of rat hypothalamus and hippocampal lysates with a monoclonal mouse anti-GR (BuGR2, Affinity Bioreagent) following BS³ cross-linking of membrane proteins. Following loading of 95 μg total protein, the GR antibody revealed multiple immunoreactive bands. A heavy molecular weight band near the loading wells was present in BS³-treated tissue, but absent in the untreated tissue, from PVN, SON and hippocampus. Each lane contains PVN, SON, or hippocampal tissue collected from a single rat (n=4 rats/treatment); bar graphs represent mean optical density measures ± SEM. (A) In the PVN, immunoreactive bands at ~98 kDa (IR Band 2) and 64 kDa (IR Band 4) were decreased in the BS³-crosslinked samples compared to untreated samples (Student’s t test, p=0.0065 and 0.036, respectively). (B) A similar change in band density was observed in the SON, with decreases in both the 98-kDa (IR Band 2, p=0.0028) and 64-kDa (IR Band 4, p=0.0068) bands in the BS³-crosslinked samples. (C) In the hippocampus, there was a decrease in density of the 64-kDa band (IR Band 4) between the BS³-cross-linked group and untreated group (p=0.0002), but no change in the 98-kDa band (p=0.70). There was no significant difference in the other immunoreactive bands between untreated and crosslinked samples in any of the brain regions. β-actin was used as a loading control for all the samples. *, p < 0.05; **, p < 0.01.

Figure 2. Saturation binding of [³H]-corticosterone to mouse and rat hypothalamic membrane and cytoplasmic fractions

[³H]-corticosterone concentrations up to 40 nM for cytoplasmic and 80 nM for membrane fractions were used. Specific binding was calculated by subtracting nonspecific binding (defined in the presence of 10 μM Cort for both membrane and cytoplasmic fractions) from the total binding (data not shown). (A) Binding of [³H]-Cort to mouse membrane (left) and cytoplasmic (right) fractions. The K_d values are shown in the insets. (B) Binding of [³H]-Cort to rat membrane (left) and cytoplasmic (right) fractions. The K_d values are shown in the insets. Data are representative of two independent experiments. Both total and nonspecific binding were calculated from duplicate points.

Figure 3. Saturation binding of [³H]-dexamethasone to mouse and rat hypothalamic membrane and cytoplasmic fractions

[³H]-dexamethasone concentrations up to 40 nM for cytoplasmic and 80 nM for membrane fractions were used. Specific binding was calculated by subtracting nonspecific binding (defined in the presence of 10 μM Dex for membrane fractions and 100 μM Dex for cytoplasmic fractions) from the total binding (data not shown). (A) Binding of [³H]-Cort to mouse membrane (left) and cytoplasmic (right) fractions. The K_d values are shown in the insets. (B) Binding of [³H]-Cort to rat membrane (left) and cytoplasmic (right) fractions. The K_d values are shown in the insets. Data are representative of two independent experiments. Both total and nonspecific binding were calculated from duplicate points.