Neurosteroid modulation of GABA IPSCs is phosphorylation dependent

Abstract

The neurosteroid 3alpha-hydroxy-5alpha-pregnan-20-one (allopregnanolone) facilitates GABA(A) receptor-mediated ionic currents via allosteric modulation of the GABA(A) receptor. Accordingly, allopregnanolone caused an increase in the slow decay time constant of spontaneous GABA-mediated IPSCs in magnocellular neurons recorded in hypothalamic slices. The allopregnanolone effect on IPSCs was inhibited by a G-protein antagonist as well as by blocking protein kinase C and, to a lesser extent, cAMP-dependent protein kinase activities. G-protein and protein kinase C activation in the absence of the neurosteroid had no effect on spontaneous IPSCs but enhanced the effect of subsequent allopregnanolone application. These findings together suggest that the neurosteroid modulation of GABA-mediated IPSCs requires G-protein and protein kinase activation, although not via a separate G-protein-coupled steroid receptor.

Fig. 1.

Allopregnanolone increases the decay time constant of IPSCs. A,Top, Spontaneous IPSCs recorded in control conditions (control) and after a 15 min bath application of allopregnanolone (ALLO; 1 μm) that caused a substantial increase in the IPSC decay. Bottom, Superimposed averages of IPSCs from the same cell recorded over 3 min periods in control conditions (control), after 15 min of allopregnanolone application (ALLO 15′), and after 15 min of washout of the allopregnanolone (wash 15′). The amplitudes of the mean IPSCs were normalized to control amplitudes. The slow IPSC decay time constant increased by 51.2%, from 31.4 to 44.5 msec, after 15 min in allopregnanolone and increased further after the 15 min washout to stabilize at 272% of the control value (117.2 msec).B, The changes in IPSC decay, amplitude, and frequency measured after a 15 min bath application of allopregnanolone (1 μm;openbars) and a 15 min washout period (filledbars). The average of the slow decay time constant showed a significant increase in allopregnanolone (+67.3 ± 14.8%) and continued to increase during the 15 min washout period (+122.3 ± 25.8%). The average changes in IPSC amplitude (+2.3 ± 6.5%) and frequency (−17.5 ± 7.9%) were not significant. The numbers in parenthesesindicate the number of cells tested. C, Scatter plot of changes in IPSC amplitude and frequency in individual cells. Although some cells showed a change in one or the other of the two parameters after a 15 min application of allopregnanolone (1 μm), there was no correlation between the two in individual cells, nor were the changes consistent across the population.

Fig. 2.

Bath application of isopregnanolone and intracellular application of allopregnanolone had no effect on IPSCs.A,Top, Spontaneous IPSCs were recorded in control conditions (control) and after a 15 min bath application of the physiologically inactive stereoisomer of allopregnanolone, isopregnanolone (ISOP; 1 μm). Bottom, Superimposed averages of IPSCs from the same cells recorded over 3 min periods in control ACSF (control) and after 15 min of isopregnanolone application (ISOP 15′) are shown. B,Top, Spontaneous IPSCs before (control) and after 15 min of intracellular application of allopregnanolone (ALLOinside; 1 μm) are shown.Bottom, Superimposed averages of IPSCs from the same cells before (control) and after a 15 min intracellular application of allopregnanolone (ALLO inside 15′) are shown. C, The slow decay time constant increased 67.3 ± 14.8% after a 15 min allopregnanolone application (ALLO) compared with control values (*p < 0.001, paired t test;n = 21). The changes in IPSC decay were nonsignificant after a 15 min application of isopregnanolone at 1 μm [ISOP (1 μm), 6.6 ± 6.7%; n = 5] and at 10 μm [ISOP (10 μm), 12.9 ± 14.9%; n = 4] and after 15 min of intracellular allopregnanolone application (ALLOinside, 9.6 ± 4.8%; n = 8).

Fig. 3.

G-Protein dependence of the allopregnanolone effect. The G-protein dependence of the allopregnanolone effect on the IPSC slow decay was tested by inhibiting G-protein activity with GDP-β-S (500 μm) and 0 mm ATP and/or GTP applied intracellularly. Averages of IPSCs over 3 min were normalized and superimposed. A, The slow component of the average IPSC decay increased after a 15 min application of allopregnanolone (ALLO) by an average of 51.2% in a control cell. B, Allopregnanolone had no effect after intracellular application of the GDP-β-S solution (GDP-β-S & ALLO). C, The average change in the slow IPSC decay was 67.3 ± 14.8% in control cells and 2.4 ± 6.5% in cells perfused with GDP-β-S (n = 7; *p < 0.01, Wilcoxon rank sum test).reg., Regular.

Fig. 4.

Protein kinase dependence of the allopregnanolone modulation of IPSCs. Averages of IPSCs recorded over 3 min in control and in allopregnanolone (1 μm; 15 min) with and without the addition of protein kinase antagonists. A, The mean IPSCs in control (control) and in allopregnanolone (ALLO) were normalized and superimposed. B–D, The application of allopregnanolone was preceded by the bath application of membrane permeable blockers of PKC (bisindolylmaleimide, 500 nm;B;PKC antagonist & ALLO), PKA (H-89, 1 μm;C; PKA antagonist & ALLO), and PKG (Rp-8-Br-cGMPS, 40 μm; D; PKG antagonist & ALLO). The PKC antagonist completely abolished the effect of allopregnanolone (B), the PKA antagonist partially blocked the effect (C), and the PKG antagonist appeared to have no effect on the allopregnanolone-induced increase in IPSC decay (D). E, The average change in the slow decay of IPSCs caused by allopregnanolone in the absence and presence of protein kinase antagonists is shown. The effect of allopregnanolone (ALLO; control % change, +67.3 ± 14.8%) was significantly inhibited by the PKC blocker(% change, +6.8 ± 5.3%; n = 5; *p < 0.05) but not by the PKA blocker (% change, +28.0 ± 18.9%; n= 5) and was not changed by the PKG blocker (% change, +80.1 ± 20.1%; n = 5) (Kruskal–Wallis one-way ANOVA on ranks; post hoc Dunn's test).

Fig. 5.

Activation of G-proteins or PKC has no effect on IPSCs in the absence of allopregnanolone. A, A 15 min application of allopregnanolone caused a 53.7% change in the slow IPSC (ALLO) in a control cell. B, The slow IPSC decay was unaffected by the intracellular application of GTP-γ-S for 15 min (GTP-γ-S) compared with the IPSC decay at break-in (control). Note thatcontrol and GTP-γ-Straces overlap completely. A subsequent 15 min application of allopregnanolone increased the slow IPSC decay by 90.2% (GTP-γ-S & ALLO). C, Intracellular application of the PKC agonist PMA (40 nm) had very little effect on the slow IPSC decay (+8.1%; PMA) compared with the averaged IPSCs recorded at break-in (control). Subsequent allopregnanolone application (1 μm) caused a 219.4% change in the IPSC decay (PMA & ALLO). D, The average percent changes in the slow IPSC decay with stimulation of G-protein and PKC activity are shown. Allopregnanolone caused a 67.3 ± 14.8% average increase in the slow IPSC decay (Control ALLO). The IPSC decay changed an average of −2.9 ± 1.3% with GTP-γ-S perfusion alone (openbar; GTP-γ-S) and +85.2 ± 21.3% in allopregnanolone (15 min) after intracellular perfusion of GTP-γ-S (15 min;filledbar). The IPSC decay increased by an average of 6.3 ± 6.7% with PMA perfusion alone (openbar; PMA) and by 129.5 ± 62.0% with allopregnanolone application (15 min) after intracellular perfusion of PMA (15 min; filledbar).

Fig. 6.

Models of G-protein and protein kinase dependence of GABA_A receptor modulation by allopregnanolone. Our findings suggest that phosphorylation of GABA_A receptors is necessary for the neurosteroid regulation of synaptic GABA currents, although enhancing G-protein and PKC activity alone is not sufficient to mimic the effect of the neurosteroid. This codependence on receptor phosphorylation and steroid presence can be explained by the following two models. A,Top, The first model is one in which the GABA_A receptor is constitutively phosphorylated (P), allowing the neurosteroid (ALLO; A) to bind and increase channel opening and chloride flux. Bottom, Blocking G-protein and/or PKC activity leads to dephosphorylation of the channel, causing a conformational change in the receptor protein that prevents the neurosteroid from binding. B,Top, In the second model, in the absence of the neurosteroid the phosphorylation site (Ser/Thr) in the GABA_A receptor is hidden from the constitutively active PKC. Bottom, After allopregnanolone binding, a conformational change in the GABA_A receptor occurs, and the Ser/Thr residue(s) (open diamond) becomes exposed for phosphorylation by PKC, increasing channel opening and chloride influx.PLC, Phospholipase C.