Progesterone inhibition of voltage-gated calcium channels is a potential neuroprotective mechanism against excitotoxicity

Abstract

The therapeutic use of progesterone following traumatic brain injury has recently entered phase III clinical trials as a means of neuroprotection. Although it has been hypothesized that progesterone protects against calcium overload following excitotoxic shock, the exact mechanisms underlying the beneficial effects of progesterone have yet to be determined. We found that therapeutic concentrations of progesterone to be neuroprotective against depolarization-induced excitotoxicity in cultured striatal neurons. Through use of calcium imaging, electrophysiology and the measurement of changes in activity-dependent gene expression, progesterone was found to block calcium entry through voltage-gated calcium channels, leading to alterations in the signaling of the activity-dependent transcription factors NFAT and CREB. The effects of progesterone were highly specific to this steroid hormone, although they did not appear to be receptor mediated. In addition, progesterone did not inhibit AMPA or NMDA receptor signaling. This analysis regarding the effect of progesterone on calcium signaling provides both a putative mechanism by which progesterone acts as a neuroprotectant, as well as affords a greater appreciation for its potential far-reaching effects on cellular function.

Figure 1

Progesterone blocks depolarization-induced neuronal death. A and B, Graphical representation of depolarization-induced apoptosis in striatal neurons. (A) Application of 60 mM K⁺ (60K) for 3 h induced a significant increase in apoptotic neurons (t = 4.69 ; p < 0.0001) that was blocked by progesterone (Prog; 50 μM; t = 0.24; p > 0.05). Pregnenolone (Preg; 50 μM; t = 3.32; p < 0.05) or Allopregnanolone (Allo; 50 μM; t = 3.58 ; p < 0.01) did not affect the induction of apoptosis by 60K. (B) Nifedipine (Nif; 5 μM) blocked 60K-induced apoptosis (t = 0.28; p > 0.05). Letters in bars represent statistically similar groups as analyzed by a two-tailed one-way ANOVA with Bonferroni post-hoc analysis.

Figure 2

Progesterone attenuates depolarization-induced L-type calcium channel-mediated increase of intracellular calcium. (A) Increases in intracellular calcium following depolarization with 20 mM K⁺ (20K) in the presence of vehicle (Veh). (B) The 20K-induced increase in intracellular calcium is primarily through activation of L-type calcium channels, as demonstrated following application of Nif (5 μM). (C-E) Progesterone (50 μM) attenuates the 20K-mediated increase in intracellular calcium, whereas Allo (50 μM) or Preg (50 μM) is without effect. (F) Summarized data. 20K-induced calcium signals were largely attributable to activation of L-type calcium channels (t = 12.25; p < 0.05). The actions of 20K were also attenuated by Prog (t = 7.64; p < 0.05) but not Allo (t = 0.72; p > 0.05) or Preg (t = 0.10 ; p > 0.05).

Figure 3

The progesterone receptor antagonist, RU486, does not block the inhibitory effects of progesterone on the L-type calcium channel-mediated increase of intracellular calcium. (A) Progesterone (Prog; 50 μM) inhibited the increase in intracellular calcium induced by 20K in the presence of 100 nM RU486. (B) Summary of the imaging data. At left, the stimulations with 20K + RU486 induced increases of intracellular calcium that were attenuated by application of Prog (t = 7.67; p < 0.0001). At right, the increase in intracellular calcium induced by 20K is not affected by RU486 (t = 0.02; p > 0.05).

Figure 4

Progesterone inhibits voltage-gated calcium currents. (A) Representative traces from a whole-cell voltage clamp recording. Ba²⁺ currents were evoked by a 100 ms step in membrane potential from −80 to 0 mV in the absence or presence of Prog. (B) Time-course of the example experiment in A, illustrating the rapid onset and washout of Prog-mediated inhibition of calcium channels. 200 μM Cd²⁺ was used to block the whole-cell Ba²⁺ current. (C) Concentration-response curve of Prog-mediated inhibition of I_Ba. IC₅₀ = 27 μM. (n ≥ 4 for each point). (D) Summary of results from experiments testing the effects of related compounds on voltage-gated calcium channel activity; Control = 5 ± 2%, Prog = 67 ± 3%, Allo = 7 ± 2%, Preg = 8 ± 1%.

Figure 5

Depolarization-induced calcium signaling is abolished by progesterone. (A–D) Luciferase assay data measuring NFAT- and CRE-dependent transcription. (A,C) 20K produced a significant increase in NFAT- (F = 20.82; t = 6.72; p < 0.05) and CRE- (F = 31.54, t = 7.51; p < 0.05) dependent transcription in the presence of vehicle (Veh). (A–D) These effects were blocked by Prog (50 μM). Preg, Allo, cholesterol (Chol), estradiol (E2) dihydrotestosterone (DHT), or corticosterone (Cort) (all applied at 50 μM) had no effect on 20K-induced gene expression. Letters in bars represent statistically similar groups as analyzed by a one-way ANOVA with Bonferroni post-hoc analysis.

Figure 6

Progesterone does not block glutamate-mediated increases in intracellular calcium. A-E, Example plots of calcium imaging data taken from cultured striatal neurons. (A) Application of 3 μM glutamate (Glu) evoked a reproducible increase in intracellular calcium. (B) The rise in intracellular calcium following brief application of Glu was not sufficient to activate L-type calcium channels. (C and D) Glutamate-induced increases in intracellular calcium were attenuated following application of either CNQX (20 μM) or AP5 (25 μM). (E) 50 μM Prog did not affect the calcium signal induced by Glu. (F) Data summary illustrating the effects of CNQX (t = 10.01; p < 0.05) and AP5 (t = 7.31; p < 0.05) versus Nif (t = 1.24; p > 0.05) Prog (t = 0.89; p > 0.05) on Glu-induced increases in intracellular calcium.

Figure 7

Progesterone does not inhibit ionotropic glutmate receptor signaling. (A and B) Representative traces of whole-cell voltage clamp recordings during a 200 ms picospritzer application of AMPA (100 μM; A) or NMDA (100 μM; B) in the presence or absence of Prog. Progesterone application increased AMPA-mediated currents while having no effect on NMDA-mediated currents. (C) Bar graph summarizing the change in AMPA- and NMDA-mediated peak current size, AMPA = 31 ± 5% (n = 6), NMDA = −1 ± 8% (n = 6).