Basal IPSC characteristics
Basal IPSC characteristics were examined in 5-HT and non 5-HT DRN neurons. Total spontaneous IPSC (sIPSC) frequency (5-HT cells, 5.7 ± 1.3 Hz, N = 6; non 5-HT cells, 5.0 ± 1.1 Hz, N = 6) was not statistically different from miniature spontaneous IPSC (mIPSC) frequency (5-HT cells, 5.9 ± 1.0 Hz, N = 6; non 5-HT cells, 5.5 ± 1.4 Hz, N = 6). This finding indicates that the majority of spontaneous IPSCs in the DRN are non action potential-dependent, representing spontaneous fusion of GABA vesicles with the presynaptic membrane. As a consequence, sIPSC data are not shown in any of the figures. Basal mIPSC frequency, amplitude, rise time, decay time and membrane holding current were also compared between 5-HT and non 5-HT neurons. For this comparison, baseline data from all experimental groups (oCRF, UCN II, antalarmin, ASVG-30) were pooled since ANOVA demonstrated that there were no significant differences in baseline values for mIPSC frequency, mIPSC amplitude, mIPSC rise time, mIPSC fast or decay times or membrane holding current between these experimental groups (5-HT cells: N = 79; non 5-HT cells: N = 41). The amplitude of mIPSCs was significantly smaller in 5-HT than non 5-HT neurons (5-HT mIPSC amplitude: 17.5 ± 0.7 pA; non 5-HT mIPSC amplitude = 22.3 ± 1.8 pA, p < 0.01). Median rise time was greater in 5-HT than non 5-HT cells (5-HT: 1.21 ± 0.04; non 5-HT: 1.01 ± 0.05; p < 0.01). No other mIPSC characteristics differed between the groups, indicating that resting membrane potential and presynaptic GABA release are similar whereas postsynaptic GABA receptors and receptor kinetics differ between the 5-HT and non-5-HT containing neurons in the basal state.
Ovine CRF, but not UCN II, has presynaptic actions on spontaneous GABA release: increasing mIPSC frequency selectively in 5-HT DRN neurons
Ovine CRF effects on mIPSC frequency were examined in 5-HT and non 5-HT DRN neurons (). The majority of these recordings were in ventromedial DRN neurons. Data (basal synaptic activity and oCRF responses) from the dorsomedial DRN neurons did not differ from ventromedial DRN neurons by unpaired Student’s t-test, therefore, data from both subdivisions were pooled. These effects were examined in 26 5-HT neurons recorded from 23 subjects and in 13 non 5-HT neurons recorded from 13 subjects. Panels A and A’ show mIPSC activity recordings and the effect of oCRF (10 nM) in two neurons that were subsequently identified to be 5-HT-containing (tryptophan hydroxylase-IR positive, panel C) and non 5-HT-containing (tryptophan hydroxylase-IR negative, panel C’) by immunohistochemistry. IPSC frequency was increased from 6.7 to 8.45Hz (26%) in response to oCRF in the 5-HT but not the non 5-HT neuron. Cumulative probability graphs of inter-event interval for each cell (B and B’) illustrate a significant shift to the left in the 5-HT but not the non 5-HT neuron as IPSC frequency was stimulated by oCRF (5-HT: Z = 2.45, p < 0.01; non 5-HT: Z = 0.89, n.s.). IPSCs are mediated by GABAA receptors as mIPSC frequency was completely suppressed by the GABAA receptor antagonist bicuculline (20 μM) under these conditions (data not shown). UCN II did not alter mIPSC frequency in either 5-HT or non 5-HT containing neurons (N = 16 and 14 neurons recorded from 15 and 11 subjects respectively).
Ovine CRF enhanced GABAergic mIPSC frequency selectively in 5-HT DRN neurons
Data for all the cells are summarized in . There was a significant increase in mean mIPSC frequency produced by oCRF selectively in 5-HT neurons (p < 0.01). These effects of oCRF were produced by a 10 nM concentration. Most earlier studies with CRF in brain slices have found effects with this and higher concentrations of CRF (0.01-1 μM; (Liu et al., 2004
;Nie et al., 2004
;Tan et al., 2004
;Kash and Winder, 2006
)). We tested a higher concentration of oCRF (100 nM) in 12 5-HT neurons recorded from 8 subjects and found that it produced a small (4.0 to 5.0 Hz) but non-significant increase in mIPSC frequency. This higher concentration also did not significantly affect any other aspects of GABA synaptic activity examined (data not shown) and may reflect the U-shaped dose-effect function of oCRF on 5-HT neurotransmission that has previously been observed in vivo
. Whereas low concentrations of oCRF significantly inhibit DRN neuronal activity and 5-HT release in forebrain, higher concentrations produce only marginal or slightly excitatory effects (Price et al., 1998
;Kirby et al., 2000
Effect of CRF and CRF receptor-selective agonists and antagonists on mIPSC characteristics and inward current in DRN neurons
Ovine CRF has presynaptic actions on evoked GABA release: increasing eIPSC amplitude and decreasing paired-pulse ratio selectively in 5-HT DRN neurons
The results from mIPSC data indicating an oCRF-induced increase in presynaptic GABA release were confirmed with eIPSC experiments. Ovine CRF effects on eIPSC amplitude and paired-pulse ratio were examined in 5-HT and non 5-HT DRN neurons (). Evoked IPSCs were recorded in 5-HT neurons in the dorsomedial (N = 14) as well as the ventromedial subdivision of the DRN (N = 14). Basal eIPSC amplitude, basal paired pulse ratio (PPR), eIPSC response to oCRF and PPR response to oCRF did not differ between cells in these two DRN subdivisions, as determined by unpaired Student’s t-test. For this reason, data from all 5-HT DRN cells were pooled (N = 28 neurons recorded from 17 subjects) in . Panel A shows eIPSCs averaged from 60 events before (control) and after oCRF administration (29% increase) in a 5-HT DRN neuron. Panel A’ shows that oCRF significantly elevated mean eIPSC amplitude in 5-HT DRN neurons for both the first and second eIPSCs in a pair (p < 0.01). In contrast, oCRF had no significant effect on mean eIPSC amplitude in non 5-HT DRN neurons (N = 7 neurons recorded from 7 subjects; eIPSC1 control values: 157 ± 29 pA, eIPSC1 oCRF values: 179 ± 21 pA; eIPSC2 control values: 121 ± 27 pA, eIPSC2 oCRF values: 136 ± 23 pA). Bicuculline completely suppressed eIPSC amplitude, indicating that eIPSCs were mediated by the GABAA
receptor (panel A). Panel B shows paired pulse data averaged from 60 pairs before (control) and after oCRF administration in a 5-HT DRN neuron. Ovine CRF decreased the PPR by 15% in this neuron. Panel B’ shows that oCRF significantly reduced mean PPR in 5-HT DRN neurons (p < 0.01). In contrast, oCRF had no significant effect on mean PPR in non 5-HT neurons (N = 7; control values: 0.75 ± 0.06, oCRF values: 0.74 ± 0.06). Antagonism experiments (panel A”) demonstrated that while both oCRF and the CRF-R1 antagonist antalarmin individually produced significant elevations of eIPSC amplitude above baseline (p < 0.01), oCRF’s effect was significantly diminished in the presence of antalarmin (p < 0.01). Panel B” also shows that oCRF’s inhibition of eIPSC PPR in 5-HT DRN neurons (p < 0.01) was blocked by antalarmin (p < 0.01). These data indicate that both of oCRF’s effects on eIPSC amplitude and PPR were mediated by the CRF-R1 receptor subtype. We also tested the washout of oCRF effects. In the majority of cells tested, oCRF effects failed to completely reverse, persisting despite waiting as long as 60 min. We further tried to reverse oCRF’s effect with a subsequent application of an antagonist, with only limited success. Similar difficulties have been observed with bath application of CRF ligands onto brain slices in other studies (Rainnie et al., 1992
;Ungless et al., 2003
;Jedema and Grace, 2004
;Kash and Winder, 2006
;Ugolini et al., 2008
). The persistence of this CRF receptor-mediated effect has been suggested to reflect slow kinetics of dissociation of CRF from its receptors in the brain slice preparation (Ugolini et al., 2008
Ovine CRF selectively increased amplitude of evoked IPSCs and decreased paired pulse ratio in 5-HT DRN neurons
Ovine CRF and UCN II have postsynaptic actions on GABA receptors: increasing mIPSC amplitude selectively in 5-HT DRN neurons
Ovine CRF and UCN II (10 nM) effects on mIPSC amplitude were examined in 5-HT and non 5-HT DRN neurons (). Panels A and B are mIPSC traces averaged from 200 events before (control) and at the maximal steady-state effect of oCRF in a 5-HT containing and a non 5-HT containing DRN neuron. In agreement with the eIPSC results, oCRF elevated mIPSC amplitude by 17% in the 5-HT but not the non 5-HT neuron. This effect is further illustrated as a significant shift to the right of the cumulative probability graph of amplitude for the 5-HT (panel A’; Z = 1.70, p < 0.01) but not the non 5-HT neuron (panel B’; Z = 0.89, n.s.). UCN II effects were recorded in 16 5-HT neurons recorded from 15 subjects and in 14 non 5-HT neurons recorded from 11 subjects. Panels C and D are mIPSC traces averaged from 200 events before (control) and at the maximal steady-state effect of UCN II in a 5-HT containing (C) and a non 5-HT containing DRN neuron (D). UCN II elevated mIPSC amplitude by 21% in the 5-HT but not the non 5-HT neuron. This effect is further illustrated as a significant shift to the right of the cumulative probability plot of amplitude for the 5-HT (panel C’; Z = 2.46, p < 0.01) but not the non 5-HT neuron (panel D’; Z = 1.26, n.s.). The data for all of the cells recorded with oCRF and UCN II are summarized in . The increase in amplitude produced by oCRF and UCN II in 5-HT containing neurons was significant (p < 0.01) as shown in . In contrast there was no significant change in mIPSC amplitude in non 5-HT containing neurons.
Ovine CRF and UCN II enhanced GABAergic mIPSC amplitude selectively in 5-HT DRN neurons
A correlation analysis was performed in order to determine if the pre- and postsynaptic actions of CRF agonists were observed in the same cells. There was a significant positive correlation between oCRF effects on mIPSC frequency and amplitude in individual 5-HT DRN neurons (R = 0.59, p < 0.01). Though UCN II did not have a significant effect on mIPSC frequency, there was a significant positive correlation between UCN II effects on mIPSC frequency and amplitude (R = 0.80, p < 0.01).
Ovine CRF and UCN II have direct postsynaptic effects on 5-HT neurons
Ovine CRF and UCN II effects on inward current were examined in 5-HT and non 5-HT DRN neurons (). Panel A shows oCRF (100 nM) increasing the inward current by 15 pA in a 5-HT DRN neuron from an initial holding current of -9 pA to a maximal steady-state effect of oCRF at -24 pA. Ovine CRF (10 nM) significantly increased inward current by 7-10 pA in both 5-HT and non 5-HT neurons (p < 0.01), as shown in and . further shows that the effect in 5-HT neurons is dose-dependent with a higher 100 nM dose producing an increase of inward current (p < 0.05) of greater magnitude. Whereas oCRF produced an inward current in all DRN cells, UCN II was more selective, significantly increasing inward current in 5-HT (p < 0.01) but not non 5-HT neurons (). This is further illustrated in as UCN II produced a 10 pA mean inward current in 5-HT neurons (p < 0.01) as compared to a 3 pA inward current in non 5-HT neurons (n.s.).
Ovine CRF and UCN II enhanced inward current in DRN neurons
Ovine CRF and UCN II alter GABA receptor kinetics: increasing mIPSC fast decay time in 5-HT DRN neurons
Ovine CRF and UCN II effects on mIPSC fast decay time were examined in 5-HT and non 5-HT DRN neurons (). Both ligands produced a similar significant (p < 0.05) increase in fast decay in 5-HT neurons (see examples in ) but not in non 5-HT neurons (see examples in ). Rise time as well as the slow component of the mIPSC decay time were unaffected by any drug treatments in 5-HT and non 5-HT DRN neurons ().
CRF-R1 and CRF-R2 antagonists selectively block the effects of oCRF
Ovine CRF effects on GABA synaptic activity [mIPSC frequency (panel A), amplitude (panel B) and inward current (panel C)] were examined in the presence of the CRF-R1-selective antagonist antalarmin (300 nM) or the CRF-R2-selective antagonist ASVG-30 (100 nM) in 5-HT DRN neurons (). Antagonist effects were also examined on mIPSC rise and decay times (see ). These concentrations of antagonist were approximately 100 times their affinity for their respective receptors (Webster et al., 1996
;Ruhmann et al., 1998
;Higelin et al., 2001
). Antalarmin effects were examined in 20 5-HT neurons recorded from 13 subjects; ASVG-30 effects were examined in 17 5-HT neurons recorded from 15 subjects.
Ovine CRF effects on GABA synaptic activity and membrane characteristics in 5-HT DRN neurons were differentially mediated by CRF-R1 and CRF-R2 receptor subtypes
In the absence of an antagonist, oCRF significantly elevated mIPSC frequency (panel 5A and ; p < 0.01), amplitude (panel 5B and , p < 0.01), fast decay time (, p < 0.05) and inward current (panel 5C and , p < 0.01). shows that antalarmin blocked the effect of oCRF on mIPSC frequency, amplitude and fast decay time, but not on inward current (X2(2)=18.9, p < 0.01). ASVG-30 blocked the oCRF effect on mIPSC amplitude, fast decay time and inward current, but not on mIPSC frequency (X2(2)=8.94, p < 0.05). The data are presented in a different manner as percentage change from baseline in . Antalarmin blocked oCRF’s effect on mIPSC frequency [panel A; significant main effect of antalarmin: F(1,44)=5.47, p < 0.05 and significant interaction of antalarmin x oCRF: F(1,44)=4.40, p < 0.05; post hoc comparisons show oCRF significantly elevated mIPSC frequency in the absence (p < 0.05) but not the presence of antalarmin and oCRF’s effect was significantly reduced in the presence of antalarmin as compared to it’s effect in the presence of ACSF (p < 0.05)]. However, antalarmin did not block oCRF’s effect on inward current [panel C; significant main effects of antalarmin: F(1,44)=10.08, p < 0.01 and oCRF: F(1,44)=8.24, p < 0.01 but no interaction between these factors]. In contrast to antalarmin, ASVG-30 blocked oCRF’s effect on inward current [panel C; significant interaction of ASVG-30 x oCRF: F(1,41)=5.73, p < 0.05; post hoc comparisons show that oCRF significantly elevated inward current in the absence (p < 0.05) but not the presence of ASVG-30 and oCRF’s effect was significantly reduced in the presence of ASVG-30 as compared to it’s effect in the presence of ACSF (p < 0.05)]. However, ASVG-30 did not block oCRF’s effect on mIPSC frequency [panel A; no significant interaction between ASVG-30 and oCRF]. While in the presence of ACSF, oCRF significantly elevated mIPSC amplitude (p < 0.05), in the presence of either antagonist, oCRF was unable to significantly elevate mIPSC amplitude above its baseline, indicating that both antagonists block this effect as well [panel B; significant interaction of antalarmin x oCRF: F(1,44)=5.90, p < 0.05 and significant interaction of ASVG-30 x oCRF: F(1,41) = 7.72, p < 0.01]. Based on these data, we conclude that the actions of oCRF on mIPSC frequency are mediated by CRF-R1, the actions on mIPSC amplitude and decay time by both CRF-R1 and CRF-R2 and the actions on inward current by CRF-R2.
Interestingly, we found that both antagonists had intrinsic effects indicating the possibility that these antagonists had partial agonist activity (see Discussion, and ). Both antalarmin and ASVG-30 increased mIPSC amplitude (, p < 0.05 by post hoc Student-Newman-Keuls test) (: ANOVA, antalarmin: F(2,38)=8.29, p < 0.01 and ASVG-30: F(2,32)=9.18, p < 0.01; p < 0.05 by post-hoc Student-Newman-Keuls test). Antalarmin increased mIPSC fast decay time [, ANOVA: F(2,35)=6.48, p < 0.01, p < 0.05 by post-hoc Student-Newman-Keuls test]. Antalarmin also had an effect on inward current (: (X2(2)=18.9, p < 0.01; p < 0.05 by post-hoc Student-Newman-Keuls test). The one significant effect produced by oCRF in non 5-HT neurons was an increase in inward current (, p < 0.01). In contrast to what was found in 5-HT neurons where the increase in inward current was blocked by ASVG-30 and also produced by UCN II, this effect in non 5-HT neurons was blocked by antalarmin (F(2,10)=0.97, n.s.) and not produced by UCN II, indicating that the effect is mediated by CRF-R1 receptors.