Research ArticleNEURODEVELOPMENT

A role for corticotropin-releasing factor signaling in the lateral habenula and its modulation by early-life stress

See allHide authors and affiliations

Science Signaling  06 Mar 2018:
Vol. 11, Issue 520, eaan6480
DOI: 10.1126/scisignal.aan6480

Stress, the brain, and behavior

Stress hormones, such as CRF (also known as CRH), inhibit the “reward” signals provided by dopamine signaling in the brain after or in anticipation of a given action. Potentially because of this, severe or chronic stress is associated with depression and can impair decision-making, and early-childhood stress is linked to long-term mental health problems and behavioral disorders. Authement et al. found that CRF exposure in slices or maternal deprivation in pups decreased the abundance of K+ channels, which increased the excitation of neurons in the lateral habenula (LHb), a region of the brain that suppresses dopaminergic circuitry. Maternal deprivation in rats blunted the response of LHb neurons to subsequent, acute stress (CRF exposure), indicating some permanence to the circuitry effects. Blocking the kinase PKA, which mediated these effects, might be therapeutic in patients with abnormal, stress-associated LHb activity.

Abstract

Centrally released corticotropin-releasing factor or hormone (extrahypothalamic CRF or CRH) in the brain is involved in the behavioral and emotional responses to stress. The lateral habenula (LHb) is an epithalamic brain region involved in value-based decision-making and stress evasion. Through its inhibition of dopamine-mediated reward circuitry, the increased activity of the LHb is associated with addiction, depression, schizophrenia, and behavioral disorders. We found that extrahypothalamic CRF neurotransmission increased neuronal excitability in the LHb. Through its receptor CRFR1 and subsequently protein kinase A (PKA), CRF application increased the intrinsic excitability of LHb neurons by affecting changes in small-conductance SK-type and large-conductance BK-type K+ channels. CRF also reduced inhibitory γ-aminobutyric acid–containing (GABAergic) synaptic transmission onto LHb neurons through endocannabinoid-mediated retrograde signaling. Maternal deprivation is a severe early-life stress that alters CRF neural circuitry and is likewise associated with abnormal mental health later in life. LHb neurons from pups deprived of maternal care exhibited increased intrinsic excitability, reduced GABAergic transmission, decreased abundance of SK2 channel protein, and increased activity of PKA, without any substantial changes in Crh or Crhr1 expression. Furthermore, maternal deprivation blunted the response of LHb neurons to subsequent, acute CRF exposure. Activating SK channels or inhibiting postsynaptic PKA activity prevented the effects of both CRF and maternal deprivation on LHb intrinsic excitability, thus identifying potential pharmacological targets to reverse central CRF circuit dysregulation in patients with associated disorders.

INTRODUCTION

Corticotropin-releasing factor [CRF; also known as corticotropin-releasing hormone (CRH)] mediates behavioral responses to stress and is implicated in a multitude of stress- and mood-related disorders. Although widely known for its participation in stress-induced activation of the hypothalamic-pituitary-adrenal (HPA) axis, CRF also acts within other regions of the brain where it directly contributes to stress-responsive behavior. This extrahypothalamic CRF system includes brain regions whose activities are implicated in mental illness (14). Many of these brain regions, including the ventral tegmental area (VTA), extended amygdala, raphe nuclei, the bed nucleus of the stria terminalis (BNST), and medial prefrontal cortex, are under intense investigation with regard to their role in the extrahypothalamic CRF system (58). To date, these studies have demonstrated that the action of CRF is not only region-specific, owing to circuit, cell-type, and pathway specificity, but also dependent on the organism’s previous exposures to stress (911).

To extend our knowledge of the extrahypothalamic CRF system to other brain regions, we investigated the role of CRF in the lateral habenula (LHb). The LHb is a highly stress-responsive brain region because of its ability to convey negative reward information to midbrain monoaminergic systems (12). Dysfunction of the LHb is implicated in stress-related disorders (1316). This area is activated by reward omission and aversive stimuli but inactivated by unexpected reward. The LHb influences behavior through its inhibitory control of midbrain dopamine (DA) neurons. Optogenetic activation of LHb neurons projecting to rostromedial tegmental nucleus (RMTg) γ-aminobutyric acid–containing (GABAergic) neurons is sufficient to elicit behavioral avoidance (17). Other studies have demonstrated the role of fast synaptic transmission mediated by GABA and glutamate, as well as neuromodulatory effects of serotonin, norepinephrine, and DA in the LHb (18); however, no study to date has investigated CRF neurotransmission in the LHb. The CRF receptors CRFR1 and CRFR2 (also known as CRHR1 and CRHR2) are G protein (heterotrimeric guanine nucleotide–binding protein)–coupled receptors (GPCRs). Activation of these receptors has a diverse range of neuronal effects depending on the brain region in which they are expressed. In general, activation of CRFR1 is anxiogenic, whereas CRFR2 activation is anxiolytic (4). Although the LHb shows immunoreactivity and mRNA expression for both CRF and CRFR1 (1921), and receives inputs from the structures that comprise the extrahypothalamic CRF system, its involvement in this system is unexplored.

Here, we investigated the role of CRF signaling on synaptic transmission and neuronal excitability in the rat LHb. We also examined the experience-specific role of CRF signaling in the LHb using an established model of severe early-life stress, maternal deprivation (MD), that is associated with persistent hyperactivity of the HPA axis, CRF signaling, and DA dysfunction (2224). MD impairs GABAergic control of VTA DA neurons through epigenetic impairment of GABAergic plasticity, which in part contributes to MD-induced dysregulation of DA neurotransmission (25). CRF can modulate DA signaling and alter mesolimbic reward function (26, 27). Therefore, an altered CRF signaling within the LHb as an upstream input to the VTA could play a crucial part in DA dysfunction associated with MD stress model. Here, we found that CRF-CRFR1-PKA (protein kinase A) signaling in the LHb increased intrinsic neuronal excitability through a G protein–dependent activation of Ca2+ signaling and changes in afterhyperpolarizations (AHPs) mediated by the Ca2+-activated K+ channels, BK and SK channels, in a PKA-dependent manner. In addition, we found that changes in the abundance of SK2 channels and PKA signaling within the LHb mediated the increases in LHb intrinsic neuronal excitability by an acute episode of MD, which blunted the excitatory action of CRF within the LHb after MD. Although CRF suppressed GABAergic synaptic transmission onto LHb neurons through an endocannabinoid (eCB) retrograde signaling, this effect remained intact after MD. Our study provides evidence for the existence of CRF-CRFR1-PKA signaling within the LHb, which may mediate part of stress-relevant effects of CRF through changes in synaptic transmission and intrinsic excitability of LHb neurons. The recruitment of the CRF stress systems within the LHb may play a key role in the development of stress-, addiction-, and depressive-related behaviors.

RESULTS

CRF increased the excitability of LHb neurons

Extrahypothalamic CRF mediates behavioral responses to stress. The region-specific actions of CRF have been defined in brain regions that are known to be involved in stress-related behaviors; to date, no study has investigated the role of CRF signaling in the LHb. Therefore, we first tested the action of CRF on the excitability of LHb neurons. Neuronal excitability is influenced by both synaptic transmission and neuronal intrinsic properties. Action potentials (APs) generated by depolarizing steps were first measured in LHb neurons with intact fast synaptic transmission. We found that CRF significantly increased the firing of neurons. This may involve changes in fast synaptic transmission and intrinsic excitability by CRF (Fig. 1, A and B). To isolate the action of CRF on the intrinsic excitability of the cell, we repeated the same experiment in the absence of fast synaptic transmission [by blocking GABAA receptors (GABAARs), AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid) receptors (AMPARs), and NMDA (N-methyl-d-aspartate) receptors (NMDARs) in the perfusate] and found that CRF still significantly increased neuronal excitability (Fig. 1, C and D). The excitatory effect of CRF on neuronal excitability of LHb neurons was maximally achieved by 20 to 30 min of bath application, remained stable over time, and persisted even after washout of CRF (fig. S1). With fast synaptic transmission blocked, we then analyzed CRF’s effects on different characteristics of APs and intrinsic neuronal excitability including fast AHPs (fAHPs), medium AHPs (mAHPs), threshold potential, half width and amplitude of AP, and input resistance (Rin). Of these, only fAHP, mAHP, Rin, and AP threshold measurements were significantly different after CRF. AHPs are mainly mediated by multiple Ca2+-dependent K+ channels that tend to be inhibitory (reduce neuronal excitability) (28). Although fAHPs were increased after CRF application, CRF was associated with decreased mAHPs and increased Rin. CRF also lowered AP threshold (Fig. 1, E and F; fig. S1; and table S1). The caveat for the measurements of AHPs in response to sustained membrane depolarization is the inability to produce similar numbers of APs under different conditions because these various AHPs are to varying degrees dependent on intracellular calcium concentration.

Fig. 1 CRF increased the excitability of LHb neurons.

(A) Whole-cell patch-clamp recording of APs in LHb neurons from non-MD rats in response to depolarizing current injections (I) and CRF (250 nM) bath application in slices, with synaptic transmission intact (n = 7 cells from 7 rats; F1,6 = 12.3). (B) Sample AP recordings of LHb neurons in response to a 50-pA depolarizing current step before (baseline, black) and after CRF (red) application (calibration bar, 20 mV/1 s) from data in (A). (C) Whole-cell patch-clamp recordings of APs as described in (A), with fast synaptic transmission blocked (n = 10 cells from 10 rats; F1,9 = 13.901). (D) Sample AP recordings as described in (B) with synaptic blockade. (E) Schematic of the method to measure AP threshold, mAHP, and fAHP in sample AP recordings in response to the lowest depolarizing current step that generated the first AP/s before (black) and after CRF (red) application (calibration bar, 20 mV/1 s). Input resistance (Rin) was calculated from the steady-state voltage deflections generated in response to a 50-pA hyperpolarizing current (calibration bar, 10 mV/1 s). (F) Average amplitude of fAHP, mAHP, and Rin derived from recordings in (C). Data are means ± SEM from cells for each condition (one cell from each rat). *P < 0.05, ***P < 0.001, by two-way repeated-measures analysis of variance (RM-ANOVA) (A and C) or paired Student’s t test (F).

CRF did not affect glutamatergic synaptic transmission but decreased GABAergic synaptic transmission through retrograde eCB signaling

Although CRF increased neuronal excitability in the absence of fast synaptic transmission, it was likely that CRF also altered synaptic transmission to promote LHb hyperexcitability. To test this, we examined the effects of CRF on the frequency and amplitude of AMPAR-mediated miniature excitatory postsynaptic currents (mEPSCs) and GABAAR-mediated miniature inhibitory postsynaptic currents (mIPSCs) recorded from LHb neurons. CRF did not have a significant effect on either the frequency or amplitude of mEPSCs, suggesting that CRF did not affect the presynaptic release of glutamate or the function of postsynaptic AMPARs (Fig. 2A). However, we found that mIPSCs were significantly reduced in the frequency and amplitude; whereas the mean amplitude of mIPSC was not significantly different after CRF, we detected a significant shift in the cumulative probability plot for amplitude by CRF (Fig. 2B), suggesting that CRF reduced GABAergic inhibition in LHb neurons through significant decreases in the release of GABA from presynaptic terminals and perhaps reduction in postsynaptic GABAAR number or function. The suppressing effect of CRF on GABAergic transmission persisted even after washout of CRF (fig. S2). Because CRF depressed GABAergic transmission without altering glutamatergic transmission in control rats that had been deprived of maternal care (non-MD rats), we next recorded evoked EPSCs and IPSCs in LHb neurons by stimulating afferents in the stria medullaris before and after application of CRF to quantify synaptic drive [excitation and inhibition balance; EPSC/IPSC (E/I) ratio] onto LHb neurons and examine whether CRF shifts the excitatory-inhibitory balance toward excitation. However, E/I ratio measurements did not change after CRF application (fig. S3A), suggesting that the balance of excitation to inhibition remains unchanged after CRF.

Fig. 2 CRF depressed GABAergic transmission without affecting glutamatergic synaptic transmission.

(A) Top: Representative AMPAR-mediated mEPSC traces recorded from LHb neurons from non-MD rats before (baseline) and after CRF (250 nM) application (calibration bars, 30 pA/5 s). Graphs below: Average mEPSC amplitude and frequency, and the cumulative probability plots of amplitude and frequency (inter-event interval) for all mEPSCs before and after CRF application (n = 10 cells from 10 non-MD rats). (B) As described in (A) for GABAAR-mediated mIPSCs before and after CRF application (calibration bars, 50 pA/5 s) (n = 6 cells from 6 non-MD rats). Data are means ± SEM from cells for each condition (one cell from each rat). *P < 0.05, ****P < 0.0001 by paired Student’s t tests or Kolmogorov-Smirnov (KS) tests for cumulative distribution curves.

Given the well-known ability of eCB signaling to suppress GABA release from presynaptic GABAergic terminals (29), we investigated whether CRF induced an eCB-mediated GABAergic depression in the LHb. First, we blocked postsynaptic G protein signaling in the neuron by intracellular administration of guanosine 5′-O-(2′-thiodiphosphate) (GDP-β-S), a nonhydrolyzable GDP analog, through the patch pipette and tested the effects of CRF on mIPSCs in the same neurons. Intrapipette GDP-β-S completely prevented CRF-induced decreases in the amplitude and frequency of mIPSCs (Fig. 3A). Consistently, we found that activation of CB1 cannabinoid receptors by WIN55,212-2, a CB1 receptor agonist, depressed GABAergic synaptic transmission (Fig. 3B). AM251, a CB1 receptor antagonist, not only blocked the suppressing effect of CRF on presynaptic GABA release but also led to a leftward shift in the cumulative probability curve of inter-event intervals indicative of an increase in the probability of presynaptic GABA release (Fig. 3C). Notably, we also detected significant reductions in the mean amplitude of mIPSCs and a leftward shift in the cumulative amplitude probability curve of mIPSC amplitude by WIN55,212-2 (Fig. 3B), similar to the effect of CRF (Fig. 2B). Although we assume that significant changes in the probability of GABA release by CRF and WIN55,212-2 affected not only the frequency but also the amplitude distribution of mIPSCs in these recordings (30), it is also possible that CRF affected postsynaptic GABAAR-mediated inhibition independent of eCB-CB1 receptor signaling as the postsynaptic effect of CRF on mIPSC amplitude, which was not blocked by AM251 (Fig. 3C).

Fig. 3 CRF suppressed GABA release from GABAergic terminals through retrograde eCB signaling.

(A to C) Average mIPSC amplitude and frequency, and cumulative probability plots of amplitude and frequency (inter-event interval) for all mIPSCs before and after the application of (A) CRF (250 nM) with GDP-β-S (300 μM) present in the patch pipette (n = 10 cells from 10 non-MD rats), (B) WIN55,212-2 (2 μM; n = 8 cells from 8 non-MD rats), or (C) CRF with AM251 (10 μM) present in the perfusate (n = 7 cells recorded from 7 non-MD rats). Data are means ± SEM from cells for each condition (one cell from each rat). *P < 0.05, ***P < 0.001, ****P < 0.001, by paired Student’s t tests or KS tests for the cumulative distribution curves.

CRF activated CRFR1 and acted through postsynaptic Ca2+ signaling

Given that the LHb is involved in the processing of aversive stimuli and has also been shown to be hyperactive in depressed and stress-related behavioral states, we hypothesized that the ability of CRF to increase LHb neuronal excitability is achieved through binding to postsynaptic CRFR1 rather than CRFR2. In all the following experiments characterizing the mechanisms underlying CRF-induced increases in LHb intrinsic excitability, AP recordings were performed in the presence of synaptic blockers similar to Fig. 1C (AMPARs, NMDARs, and GABAARs were blocked). In the presence of antalarmin, a selective CRFR1 antagonist, we found that CRF’s excitatory action on intrinsic excitability was completely blocked (Fig. 4A). On the other hand, CRF significantly increased the intrinsic excitability of LHb neurons in the presence of antisauvagine as a CRFR2-selective antagonist (Fig. 4B). To further confirm CRF’s action through postsynaptic CRFR1, a GPCR, we blocked postsynaptic G protein signaling in the neuron by intracellular administration of GDP-β-S through the patch pipette. Intrapipette GDP-β-S prevented CRF-induced increase in LHb intrinsic excitability (Fig. 4C). Both intracellular release of Ca2+ from internal stores and extracellular Ca2+ influx have previously been shown to be involved in CRFR1 signaling in other brain areas (31, 32). To further investigate whether an increase in postsynaptic Ca2+ signaling is involved downstream of CRFR1 activation, BAPTA [1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid] (a fast Ca2+ chelator) was used intracellularly through the patch pipette. Because Ca2+ signaling is involved in activation of some of cationic currents including Na+ channels in LHb neurons (33), not surprisingly, we detected that inclusion of BAPTA in the patch pipette blunted neuronal firing before CRF treatment. Nevertheless, intrapipette application of BAPTA blocked CRF’s ability to increase intrinsic excitability of LHb neurons (Fig. 4D). We next sought to determine how exactly Ca2+ signaling through CRFR1 activation could lead to increased intrinsic excitability of LHb neurons.

Fig. 4 CRF increased LHb intrinsic excitability through G protein–dependent CRFR1 signaling and requires intracellular Ca2+.

(A to D) Whole-cell patch-clamp recordings of APs in LHb neurons from non-MD rats in response to depolarizing step currents (left), with representative AP traces (right) in response to a 100-pA depolarizing current step before (baseline, black) and after application of CRF (250 nM, red), the CRFR1 antagonist antalarmin (A) (1 μM; F1,6 = 0.222), the CRFR2 antagonist antisauvagine (B) (25 nM; n = 6 cells from 6 non-MD rats, F1,5 = 5.057), the G protein inhibitor GDP-β-S (C) (300 μM; n = 4 cells from 4 non-MD rats, F1,3 = 0.86), and the calcium chelator BAPTA (D) (30 mM; n = 6 cells from 6 non-MD rats, F1,5 = 0.327), were included in the patch pipette or perfusate. All recordings were performed with fast synaptic transmission blocked. Data are means ± SEM from cells for each condition (one cell from each rat). **P < 0.01 by two-way RM-ANOVA.

CRF affected SK and BK channel functions in LHb neurons in a PKA-dependent manner

Changes in intrinsic conductances through K+ channels, such as SK and BK channels, influence the firing of neurons (34). In our earlier experiments evaluating the effects of CRF on intrinsic excitability of LHb neurons, we observed that CRF altered AHPs coincident with higher input resistance and enhanced excitability but significantly lowered AP threshold in LHb neurons (Fig. 1, E and F, and table S1). The fAHP amplitude was enhanced by CRF, suggesting an increase in BK channel function by CRF. However, the mAHPs measured in response to depolarization was decreased after bath application of CRF, suggesting a decreased function of SK channels. We also detected an increase in Rin by CRF that could support CRF-induced increase in LHb intrinsic excitability. First, we tested whether an SK channel blocker, apamin, could mimic and occlude CRF-induced increase in LHb intrinsic excitability through decrease in mAHP. We found that apamin significantly increased LHb intrinsic excitability and further occluded CRF’s action on LHb neuronal excitability (Fig. 5A). The effect of apamin on intrinsic excitability was also accompanied by an increase in the amplitude of fAHP, a decrease in the amplitude of mAHP (an apamin-sensitive component of mAHP) with a parallel increase in Rin, and a lower AP threshold (Fig. 5A and table S2). To further confirm that increasing SK channel function could prevent CRF-induced increase in intrinsic excitability, we used a positive modulator of SK channels [cyclohexyl-[2-(3,5-dimethyl-pyrazol-1-yl)-6-methyl-pyrimidin-4-yl]-amine (CYPPA)]. CYPPA did not affect intrinsic neuronal excitability by itself but blocked the excitatory effect of CRF on LHb intrinsic excitability (Fig. 5B). CYPPA did not significantly affect fAHP, mAHP, Rin, or AP threshold measurements (Fig. 5B and table S3). Although a decrease in BK-mediated fAHP seems to mediate an increase in neuronal excitability, an increase in BK channel activity can also facilitate high-frequency firing of neurons through inactivation of other K+ channels (35). So, we tested the possibility that CRF may first involve BK channels, which could subsequently inactivate SK channels to promote neuronal intrinsic excitability. The BK channel blocker iberiotoxin slightly increased LHb intrinsic excitability but significantly attenuated CRF’s excitatory effects while blocking both CRF-induced changes in AHPs and Rin (Fig. 5C). No significant changes were detected in AHPs and Rin after CRF application when iberiotoxin was present in the bath application (Fig. 5C). Activation of CRF-CRFR1 signaling initiates PKA activity, so we next examined whether activation of PKA in LHb neurons affected BK and SK channels to increase LHb neuronal excitability. PKA-mediated phosphorylation of the SK2 channel subunit (which contributes to the apamin-sensitive component of mAHP) has been shown to dramatically decrease surface localization of SK2 channels, thereby promoting neuronal excitability and the induction of synaptic plasticity at Schaffer collateral synapses and hippocampal-dependent learning (3638). On the other hand, BK channel activity can be enhanced through PKA-mediated phosphorylation of BKα subunit (the pore-forming subunit of BK channels) (39). Intracellular inclusion of PKI(6-22), a membrane-impermeant PKA inhibitor, significantly blocked CRF-induced enhancement of intrinsic excitability and CRF-induced changes in AHPs and Rin (Fig. 5D and table S4). Although a slight reduction in LHb intrinsic excitability was observed within 15 to 30 min after initiation of the whole-cell recording with PKI(6-22)-filled pipettes, this change in intrinsic excitability was not significantly different from the excitability of LHb neurons recorded within the same time frame from non-MD rats with patch pipettes filled with normal drug-free internal solutions (fig. S4 and table S4).

Fig. 5 Pharmacological modulation of SK and BK channels altered CRF-induced increases in LHb intrinsic excitability in a PKA-dependent manner.

(A to D) AP recordings in response to depolarizing current steps with representative AP traces in LHb neurons from non-MD rats in response to a 100-pA current step before (baseline, black) and after (A) SK channel blocker apamin (100 nM, blue; n = 11 cells from 11 non-MD rats, F2,20 = 9.127), (B) SK channel–positive modulator CYPPA (10 μM, green; n = 9 cells from 9 non-MD rats, F2,16 = 2.058), (C) BK channel blocker iberiotoxin (100 nM, teal blue; n = 7 cells from 7 non-MD rats, F1,6 = 3.039), or (D) PKI(6-22) (10 μM, sky blue; n = 6 cells from 6 non-MD rats, F1,5 = 2.773) alone or with subsequent CRF (250 nM, red) bath application. Right: Amplitudes of fAHP, mAHP, and Rin derived from the AP recordings. All recordings were performed with fast synaptic transmission blocked. Data are means ± SEM from cells for each condition (one cell from each rat). *P < 0.05, ***P < 0.001, by two-way RM-ANOVA (left) or paired Student’s t tests (right).

Early-life stress increased the spontaneous firing and excitability of LHb neurons by decreasing the function and abundance of SK channels

Given that early-life stresses including MD induce long-lasting alterations in CRF-CRFR1 signaling in the brain (24), it was feasible to assume that an adverse early-life experience like MD would affect the function of LHb neurons and alter their responsiveness to acute CRF. MD modifies DA signaling from the VTA through epigenetic modifications of critical signaling pathways involved in trafficking of GABAARs onto VTA DA neurons (25). Considering the potent inhibitory control of VTA DA neurons by the LHb, we chose to use the same model of severe early-life stress in this work to determine whether the response of LHb neurons to extrahypothalamic CRF is affected by exposure to early-life stress. The protocol used was of an acute MD, in which rats were isolated at P9 for 24 hours and then returned to the litter. First, we hypothesized that this period of MD would increase LHb neuronal excitability. To test this hypothesis, we recorded LHb neuronal spontaneous activity using cell-attached recordings in voltage-clamp mode to determine the percentage of neurons that were spontaneously active and, of those, the average spike frequency. In this set of experiments, synaptic transmission was intact. We observed that the portion of spontaneously active LHb neurons was significantly larger in maternally deprived rats (MD rats) compared to control non-MD rats. Moreover, LHb neurons displaying spontaneous activity had remarkably higher spike frequencies in MD rats compared to controls (Fig. 6A).

Fig. 6 MD increased the excitability of LHb neurons.

(A) Top: Representative spontaneous AP recordings of LHb neurons from non-MD and MD rats using cell-attached voltage-clamp recordings. Bottom: Percent of spontaneously active LHb neurons (left) and mean AP frequency (right) from each group (n = 33 cells from 18 rats or 29 cells from 19 rats, respectively). (B) Representative traces (top) and number of APs (bottom) recorded from LHb neurons using whole-cell patch-clamp recording, with synaptic transmission intact, in slices from non-MD and MD rats in response to depolarizing current steps (calibration bar, 20 mV/1 s) [n = 24 cells from 11 rats (non-MD) or 24 cells from 9 rats (MD), F1,460 = 1.69]. (C) As described in (B), with fast synaptic transmission blocked [n = 25 cells from 17 rats (non-MD) or 31 cells from 18 rats (MD), F1,600 = 35.84]. (D) Average amplitude of fAHP, mAHP, and Rin derived from AP recordings in (C). (E) Representative Western blots and quantitation of SK2 abundance (β-actin: loading control) in LHb tissue homogenates from non-MD and MD rats (n = 6 biological replicates). Data are means ± SEM. *P < 0.05, **P < 0.01 by unpaired Student’s t test; ****P < 0.0001 by two-way ANOVA.

In addition to cell-attached recordings of basal neuronal firing, we also tested whether LHb neurons from MD rats exhibit enhanced firing in response to depolarization compared to control non-MD rats in the absence or presence of fast synaptic transmission. Although we did not detect a significant difference in LHb neuronal excitability between non-MD and MD rats in intact synaptic transmission in response to 10- to100-pA depolarizing current steps, MD significantly increased LHb intrinsic excitability in response to depolarization in this range of current injections when fast synaptic transmission was blocked (Fig. 6, B and C). Furthermore, the MD-induced increase in intrinsic excitability was associated with a significant reduction in mAHP amplitude and a significantly higher input resistance without any change in the amplitude of fAHP, suggesting that this increase may be due to a decreased function and/or abundance of SK channels (Fig. 6D). Except mAHP and Rin, all the other measured passive and active properties of LHb neurons including resting membrane potential (RMP) were unaffected after MD (table S5). Consistently, Western blot analysis of LHb tissues from non-MD and MD rats demonstrated that SK2 but not BK channel α subunit protein abundance was significantly decreased in MD rats compared to controls (Fig. 6E and fig. S6). In contrast, we did not observe any significant changes in protein abundance of SK1 and SK3 (fig. S5).

MD potentiated both glutamatergic and GABAergic synaptic transmission but increased E/I ratio

We found that MD-induced changes in glutamatergic synapses supported increases in LHb neuronal excitability, given that LHb neurons from MD rats exhibited significantly higher AMPA/NMDA ratios compared to control non-MD rats (fig. S7A). In support of this postsynaptic glutamatergic potentiation, there was a significant increase in the cumulative probability of AMPAR-mediated mEPSC amplitudes after MD, although the average amplitudes were not significantly different between MD and non-MD rats. Moreover, we detected a marked increase in the average and cumulative probability of frequencies of mEPSCs, suggesting an increase in presynaptic glutamate release by MD (Fig. 7A). We also recorded evoked AMPAR-mediated EPSCs at different holding potentials to measure AMPAR rectification and determine whether MD induced any changes in synaptic AMPAR subunit composition. AMPA EPSC rectification and EPSC IV plots were unchanged after MD (fig. S7B).

Fig. 7 MD potentiated both glutamatergic and GABAergic synaptic transmission onto LHb neurons and increased E/I ratio.

(A) Representative AMPAR-mediated mEPSC traces from non-MD and MD rats (calibration bars, 30 pA/5 s). Average mEPSC amplitude and frequency (left) and cumulative probability plots of amplitude and frequency (inter-event interval) (right) in non-MD and MD rats (n = 20 cells from 20 rats or 17 cells from 17 rats, respectively). (B) As described in (A) for GABAAR-mediated mIPSC traces from non-MD and MD rats (calibration bars, 50 pA/5 s; n = 13 cells from 13 rats or 9 cells from 9 rats, respectively). (C) Summary of E/I ratios obtained from non-MD and MD rats (n = 16 cells from 8 rats or 12 cells from 9 rats, respectively), with representative traces of evoked EPSCs (black, recorded at −55-mV holding potential) and IPSCs (red, recorded at +10-mV holding potential) in response to the stria medullaris stimulation for LHb neurons. *P < 0.05, **P < 0.01, ****P < 0.0001 by unpaired Student’s t tests or KS tests for the cumulative distribution curves.

Although the average amplitudes of GABAAR-mediated mIPSCs were not significantly different between MD and non-MD rats, we found that the cumulative probability of mIPSC amplitudes was significantly increased after MD. Moreover, the average and cumulative probability of frequency of mIPSCs were significantly increased in MD compared to non-MD rats. This suggests that MD potentiated GABAergic synaptic transmission pre- and postsynaptically (Fig. 7B). Although both excitatory and inhibitory synapses onto LHb seemed to be potentiated by MD, we detected an increase in E/I ratio after MD, suggesting a shift in the balance between excitation and inhibition in synaptic inputs toward excitation after MD (Fig. 7C).

The effects of CRF and apamin on LHb intrinsic excitability were blunted after MD because of increased PKA signaling

Depending on the timing and severity of the stress, CRF release can have long-lasting regional effects and change subsequent responses to CRF exposure. Here, we tested whether the responsiveness of LHb neurons to acute exposure to CRF was altered by MD, with the hypothesis being that CRF is excessively released during the episode of MD, leading to dysregulation of central CRF-CRFR1-PKA signaling within the LHb. Although CRF significantly increased neuronal excitability in non-MD neurons in the presence and absence of synaptic blockade (Fig. 1), CRF’s action on neuronal excitability was blunted in MD rats in the presence and absence of fast synaptic transmission, given that there was no significant increase in LHb neuronal excitability in MD rats after adding CRF (Fig. 8, A and B). Except for a significant increase in the amplitude of fAHP accompanied by a lower AP threshold after CRF, no significant changes were detected in mAHP and Rin by CRF in MD rats (table S1). Assuming sustained CRF signaling dysregulation by MD, we next examined the effects of MD on the mRNA abundance of Crh and Crhr1 in LHb tissues. Relative to non-MD rats, no significant changes were detected in Crh or Crhr1 mRNA expression in MD rats (Fig. 8B).

Fig. 8 The effects of CRF and apamin on LHb intrinsic excitability were blunted after MD in a PKA-dependent manner.

(A) Whole-cell patch-clamp AP recordings in LHb neurons in slices from MD rats in response to depolarizing current steps with representative AP traces (in response to a 100-pA current step) before (baseline, black) and after application of CRF (250 nM, red), with fast synaptic transmission intact (top) or blocked (bottom). n = 6 cells from 6 MD rats; F1,5 = 0.38 and F1,5 = 2.335, respectively. (B) Relative expression of Crh and Crhr1 mRNA in the LHb from non-MD and MD rats (Crh, n = 8 and 10 biological replicates, respectively; Crhr1, n = 9 and 12 biological replicates, respectively). Data are normalized to expression in the non-MD group and are means ± SEM. (C and D) AP recordings in LHb neurons from MD rats in response to depolarizing current steps with representative AP traces (in response to a 100-pA current step) before (baseline, black) and after application of (C) apamin (100 nM, blue) or (D) 1-EBIO (300 mM, green), with fast synaptic transmission blocked. n = 12 cells from 12 rats (C) (F1,11 = 1.816) or 7 cells from 7 rats (D) (F1,6 = 5.677). (E) AP recordings in response to depolarizing current steps with representative AP traces (in response to a 100-pA current step) in LHb neurons from MD rats with patch pipettes filled with normal internal solution (black) or along with PKI(6-22) (10 μM; sky blue). Each condition n = 8 cells from 8 rats; F1,140 = 45.02. Data are means ± SEM. *P < 0.05 by two-way RM-ANOVA; ****P < 0.0001 by two-way ANOVA. (F) Representative Western blots and quantitative data of PKA [glyceraldehyde-3-phosphate dehydrogenase (GAPDH): loading control] in LHb tissue homogenates from non-MD and MD rats (n = 11 biological replicates).

Because we had established that CRF signaling affected the activity of postsynaptic SK channels and that the protein abundance of SK2 channels was diminished by MD, we sought to determine whether an impaired function in SK channels in MD rats underlies the blunted effect of CRF on LHb intrinsic excitability (under conditions in which synaptic transmission was blocked). The first piece of evidence to support this hypothesis was that bath application of apamin had a diminished effect on the intrinsic excitability of MD LHb neurons (Fig. 8C). Except for a significant increase in fAHP by apamin, mAHP and Rin were unaltered after exposure in MD slices (table S2). We also tested the effects of another positive SK channel modulator, 1-EBIO, which is similar to CYPPA. Bath application of 1-EBIO significantly decreased intrinsic neuronal excitability of LHb neurons in slices from MD rats despite the lower abundance of SK2 channel protein in MD rats (Fig. 8D). This decrease in intrinsic excitability was associated with a significant increase in mAHP amplitude and a lower Rin, whereas fAHP was unaffected (table S3). PKA-dependent regulation of SK channel activity and trafficking of SK2 plays an important role in neuronal excitability, and we had also observed a complete inhibition of CRF-induced increases in LHb intrinsic excitability by postsynaptic inhibition of PKA through modulation of AHPs (Fig. 5D). Although mRNA expression of Crhr1 was not significantly decreased by MD, we hypothesized that PKA abundance and/or function may have been persistently altered as a biochemical neuroadaptation in response to CRF-CRFR1 activation during MD contributing to increased LHb neuronal excitability. First, we tested whether postsynaptic manipulation of PKA activity in LHb neurons could diminish LHb intrinsic excitability in MD rats through changes in mAHP. We detected a significant decrease in the number of APs triggered in response to depolarization within 15 min after initiation of the whole-cell recording with PKI(6-22)-filled pipettes in LHb neurons recorded from MD rats, which was maximized and then stabilized at 30 min from the start of our recordings. This significant decline in intrinsic excitability was not evident within the same time frame of our recordings in LHb neurons with normal internal solution-filled pipettes from MD rats (Fig. 8E) and in PKI(6-22) experiments performed in non-MD rats (fig. S4). Although postsynaptic inhibition of PKA did not affect fAHPs, the decreased LHb intrinsic excitability of MD rats by PKA inhibition was associated with a significant increase in the amplitude of mAHP parallel with a lower Rin and AP threshold (table S4). Western blot analysis of LHb tissues from non-MD and MD rats revealed no significant difference in the abundance of PKA protein between non-MD and MD rats (Fig. 8F).

We also tested the effects of acute CRF on mEPSCs and mIPSCs recorded from LHb neurons in slices from MD rats. Except for a significant leftward shift in the cumulative probability curve of mEPSC amplitude, CRF did not affect mEPSCs in MD rats (fig. S8A), similar to our observations in non-MD rats. Similarly, acute bath application of CRF significantly reduced GABAergic synaptic transmission onto LHb neurons both pre- and postsynaptically in MD rats, suggesting that the acute effects of CRF on synaptic transmission remained intact after MD (fig. S8B). Consistently, E/I ratio measurements also did not change after CRF application in MD rats (fig. S3B). Together, our data suggest that LHb intrinsic neuronal responsiveness to acute CRF was altered after MD due to changes in SK channel activity and abundance in addition to a persistent increase in PKA activity downstream to CRF-CRFR1 signaling.

DISCUSSION

This study establishes a role for an extrahypothalamic CRF-CRFR1-PKA signaling in the LHb. We demonstrate that the stress neuropeptide CRF increased the excitability of LHb neurons and depressed GABAergic transmission onto LHb neurons through eCB signaling. We found that blockade of CRFR1 and postsynaptic Ca2+ or disruption of postsynaptic G protein signaling prevented the excitatory effect of CRF. Crhr1 and Crhr2 gene expression often overlaps, and Crhr1 and Crhr2 mRNAs have also been detected in the rodent habenula (21, 40). Our findings suggest that postsynaptic G protein–mediated CRFR1 signaling requires intracellular Ca2+ signaling to increase LHb neuronal excitability. The increase in neuronal excitability induced by CRF was associated with a significant reduction in the amplitude of the mAHP with a parallel increase in input resistance (suggesting possible decreases in K+ conductances underlying membrane resistance including leak K+ currents in addition to SK channels by CRF). Consistently, we found that blocking SK channels mimicked and occluded CRF’s effect, whereas positive enhancement of SK channel function blocked CRF’s effect. CRF increases neuronal excitability in several brain regions (4143). Decreases in the amplitude of AHPs and K+ channel function, including SK type, are associated with an increase in neuronal excitability (44, 45). On the other hand, CRF receptor signaling in VTA DA neurons potentiates the function of SK channels (31).

We also found that CRF increased the amplitude of fAHPs (mostly mediated by BK channels) and blocking BK channels prevented the effects of CRF on excitability, AHPs, and Rin. fAHPs may facilitate high-frequency firing of LHb neurons similar to that shown in CA1 hippocampal pyramidal neurons by limiting activation of other K+ channels and Na+ channel inactivation (35, 4649). In addition to an increase in the amplitude of fAHP, CRF significantly lowered AP threshold, suggesting that a hyperpolarized interspike membrane potential during high AP firing rates might be maintained by fAHPs to prevent voltage-gated Na+ channel inactivation and promote suppression of SK channel function by CRF. Activated CRFR1 primarily signals by Gs protein coupling, resulting in the induction of PKA activity and increase in hippocampal neuronal excitability (50), although CRFR1 can also affect protein kinase C (PKC) activity through Gq coupling (51). We found that postsynaptic inhibition of PKA in LHb neurons significantly blocked CRF’s action on LHb intrinsic excitability and AHPs, suggesting that PKA-dependent modulation of AHPs through opposing actions on BK and SK channel activity underlies CRF’s excitatory effects. It has been shown that PKA-mediated phosphorylation of the SK2 channel subunit significantly decreases membrane SK2 channels and SK channel function, whereas BK channel activity is augmented by PKA (3639).

Other studies have shown that the action of CRF is dependent on an organism’s previous exposure to stress (911). CRF modulates DA signaling and mesolimbic reward function (26, 27); thus, an alteration of CRF signaling within the LHb and subsequent changes in LHb function could significantly affect DA release from the VTA. Here, we found that MD increased both basal and depolarization-induced AP firing of LHb neurons. In addition, the increase in LHb intrinsic excitability after MD was accompanied by a decrease in mAHP. Protein abundance of SK2 was lower in MD rats. Positive modulation of SK channels did not reduce firing rate of LHb neurons in non-MD rats, suggesting that the basal activity of these channels may be relatively high. On the other hand, positive enhancement of SK activity normalized intrinsic excitability in LHb neurons of MD rats. A loss of apamin-sensitive SK channels (SK2 channels contribute to apamin-sensitive portion of mAHP) could be related to increased PKA signaling in MD rats. MD was not associated with a significant increase in Crh mRNA expression or a decrease in Crhr1 mRNA expression to suggest hyperactivation of CRF-CRFR1 signaling pathway in the LHb after MD. However, we found that the transduction signaling involving PKA is increased by MD, suggesting that this is an adaptation that persisted into adolescence. PKA inhibition was able to markedly attenuate the excitability of LHb neurons accompanied by a significant increase in mAHPs in MD rats with no significant alterations in BK channel abundance or activity. Because Western blot analysis of LHb tissues from non-MD and MD rats revealed no significant difference in the abundance of PKA protein (Fig. 8F), this suggests a possible sustained rebound increase in PKA activity and signaling in LHb neurons. Therefore, the blunting of CRF’s excitatory effects in the LHb of MD rats is partly due to MD-induced reduction in abundance of SK2 and also PKA-mediated changes in SK channel function and possibly trafficking.

CRF pre- or postsynaptically potentiates GABAergic transmission in some brain areas, such as dorsal raphe, central amygdala, and BNST (5254), although we found that CRF presynaptically depressed GABAergic transmission onto LHb neurons in a CB1 receptor–dependent manner. Inhibition of CB1 receptors not only prevented the suppressing effect of CRF on presynaptic GABA release but also led to an increase in the probability of presynaptic GABA release. These results point to the possible presence of a tonic eCB signaling and persistent CB1 receptor activation in the regulation of GABAergic transmission in the LHb. Glutamatergic synaptic transmission, however, was unaffected by CRF. This indicates that CRF could increase LHb neuronal excitability not only through changes in membrane properties but also by depressing GABAergic transmission. Similar to non-MD rats, CRF did not affect the synaptic excitatory to inhibitory balance of LHb neurons.

Both AMPA/NMDA ratios and the amplitude of mEPSCs were larger in LHb neurons from MD rats, suggesting an induction of a postsynaptic glutamatergic potentiation by MD. This glutamatergic potentiation was not associated with a change in AMPAR subunit composition. We assume that reduction in SK2 protein abundance by MD likely facilitated potentiation of glutamatergic synapses in LHb neurons. Hippocampal CA1 long-term potentiation was accompanied by a PKA-dependent endocytosis of SK2 channels (38). Similarly, potentiation of glutamatergic synapses and reduction of K+ conductance underlie LHb hyperexcitability after repeated injections of cocaine (55). MD also increased presynaptic glutamate release in LHb neurons similar to what was found in LHb neurons projecting to the VTA in learned helplessness animal models of depression (16). Although glutamatergic synaptic modifications induced by MD could support LHb hyperexcitability, MD was also associated with significant increases in GABAergic synaptic transmission both pre- and postsynaptically. Alteration in GABAergic synaptic strength could have been triggered in response to long-term changes in neuronal excitability induced by MD to stabilize neuronal activity (homeostatic synaptic scaling) (56). However, it is also possible that changes in LHb intrinsic excitability are a neuroadaptation triggered by MD-induced changes in GABAergic synaptic inputs to the LHb. Main GABAergic structures projecting to the LHb include the entopeduncular nucleus and the VTA, which are shown to co-release GABA and glutamate onto LHb neurons (15, 57). Therefore, an increase in glutamate and GABA release from such inputs onto LHb neurons after MD may elicit a homeostatic increase in LHb neuronal intrinsic excitability observed here. Yet, we found an imbalance of excitation to inhibition of synaptic inputs onto LHb neurons favoring excitation. The acute suppressing effects of CRF on GABAergic transmission remained intact in MD animals with no subsequent alteration in E/I balance in response to acute CRF, although MD rats still showed higher E/I ratios compared to non-MD rats (fig. S3), promoting LHb excitability.

The role of CRF-CRFR1-PKA regulation of SK channels in MD rats is an important question that merits investigation given increasing interest in the role of the extrahypothalamic CRF system in health and disease. Recent advances in optogenetic interrogation of neuronal circuitries and the availability of promoter-driven Cre mouse and rat lines will help us to elucidate the specific contributions of synaptic inputs to LHb dysfunction after MD and further delineate mechanisms underlying MD-induced dysregulation of CRF signaling in the LHb.

MATERIALS AND METHODS

All experiments were carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Uniformed Services University Institutional Animal Care and Use Committee. All efforts were made to minimize animal suffering and reduce the number of animals used.

MD procedure

Half of the male pups of litter of Sprague-Dawley rats at P9 were isolated from the dam and their siblings at 10:00 a.m. for 24 hours (MD group). The isolated rats were placed in a separate quiet room and kept on a heating pad (34°C) and not disturbed until being returned to their home cage. The remaining nonseparated male rat pups received the same amount of handling and served as the nonmaternally deprived control group (non-MD group). Then, starting at P21, each day through P28, two MD and non-MD (one from each condition) rats (age-matched) of the same litter were sacrificed to obtain LHb tissue and neurons for electrophysiology and Western blotting analyses. With the exception of Figs. 6 (A to D) and 7C, where two or three neurons were recorded per animal (n represents number of recorded cells per rat), one cell per animal was recorded to demonstrate that the MD effects observed were consistent across multiple animals and across the P21 to P28 time frame; therefore, all reported n values in those graphs represent the number of animals.

Slice preparation for electrophysiology

Rats were anesthetized with isoflurane and immediately decapitated. The brains were quickly dissected and placed into ice-cold artificial cerebrospinal fluid (ACSF) containing 126 mM NaCl, 21.4 mM NaHCO3, 2.5 mM KCl, 1.2 mM NaH2PO4, 2.4 mM CaCl2, 1.00 mM MgSO4, 11.1 mM glucose, and 0.4 mM ascorbic acid, saturated with 95% O2 to 5% CO2. Sagittal slices containing the LHb were cut at 250 μm and incubated in ACSF at 34°C for at least 1 hour. Slices were then transferred to a recording chamber and perfused with ascorbic acid–free ACSF at 28°C.

Electrophysiology

Whole-cell and cell-attached recordings were performed on LHb slices using a patch amplifier (MultiClamp 700B) under infrared–differential interference contrast microscopy. Data acquisition and analysis were carried out using DigiData 1440A, pCLAMP 10 (Molecular Devices), Clampfit, and Mini Analysis 6.0.3 (Synaptosoft Inc.). Signals were filtered at 3 kHz and digitized at 10 kHz. The recording ACSF was the same as the cutting solution except that it was ascorbic acid–free. Spontaneous activity was monitored using cell-attached voltage-clamp recordings of APs at I = 0 pA for 5 min. Number of APs was counted over 5 min, and spike frequency was calculated. Neuronal excitability recordings in response to depolarization were performed in whole-cell current-clamp mode. The patch pipettes (3 to 6 megohms) were filled with 130 mM K-gluconate, 15 mM KCl, 4 mM adenosine triphosphate (ATP)–Na+, 0.3 mM guanosine triphosphate (GTP)–Na+, 1 mM EGTA, and 5 mM Hepes (pH adjusted to 7.28 with KOH, osmolarity adjusted to 275 to 280 mOsm). LHb neurons were given increasingly depolarizing current steps at +10-pA intervals ranging from +10 to +100 pA, allowing us to measure AP generation in response to membrane depolarization (5-s duration). Current injections were separated by a 20-s interstimulus interval, and neurons were kept at −65 mV with manual direct current injection between pulses. Synaptic transmission blockade was achieved by adding 6,7-dinitroquinoxaline-2,3-dione (DNQX; 10 μM), picrotoxin (100 μM), and d,l-2-amino-5-phosphonovaleric acid (APV; 50 μM) to block AMPAR-, GABAAR-, and NMDAR-mediated synaptic transmission, respectively. The number of APs induced by depolarization at each intensity was counted and averaged for each experimental group.

Because AHP amplitude increases with increasing number of APs, we measured AP threshold, mAHP, and fAHP amplitudes at the current step that was sufficient to generate the first AP/s (58). AP threshold was measured at the beginning of the upward rise of the AP. mAHP was measured as the difference between AP threshold and the peak negative membrane potential at the end of the current step. fAHPs were calculated as the difference between AP threshold and the peak negative potential following the AP (Fig. 1E). RMP was assessed at the beginning of the recording by quickly switching to I = 0. Input resistance (Rin) was determined by injecting a small (50 pA) hyperpolarizing current pulse (5 s) and calculated by dividing the steady-state voltage response by the current pulse amplitude. Reported AP half width and AP peak amplitude were obtained from measurements of AP characteristics in Clampfit.

Whole-cell recordings of GABAAR-mediated mIPSCs were isolated in ACSF perfused with DNQX (10 μM), strychnine (1 μM), and tetrodotoxin (TTX; 1 μM). The patch pipettes (3 to 6 megohms) were filled with 125 mM KCl, 2.8 mM NaCl, 2 mM MgCl2, 2 mM ATP-Na+, 0.3 mM GTP-Na+, 0.6 mM EGTA, and 10 mM Hepes (pH adjusted to 7.28 with KOH, osmolarity adjusted to 275 to 280 mOsm). AMPAR-mediated mEPSCs were isolated in ACSF perfused with picrotoxin (100 μM), d-APV (50 μM), and TTX (1 μM). Patch pipettes for mEPSC recordings were filled with 117 mM Cs-gluconate, 2.8 mM NaCl, 5 mM MgCl2, 2 mM ATP-Na+, 0.3 mM GTP-Na+, 0.6 mM EGTA, and 20 mM Hepes (pH adjusted to 7.28 with CsOH, osmolarity adjusted to 275 to 280 mOsm). For both mIPSCs and mEPSCs, LHb neurons were voltage-clamped at −70 mV and recorded over 10 sweeps, each lasting 50 s.

EPSCs were evoked with a stimulating electrode placed in the stria medullaris. Evoked EPSCs were recorded in ACSF perfusion containing picrotoxin (100 μM) while the cell was voltage-clamped at +40 mV. Internal solution for patch pipettes was similar to that used for mEPSC recordings (Cs-gluconate–based) but also included intracellular spermine (10 μM). AMPAR-mediated currents were isolated with d-APV (50 μM), a selective NMDAR antagonist. Isolated AMPAR-mediated currents were then subtracted from the combined EPSC to provide the NMDAR-mediated current and thus the AMPA/NMDA ratio. AMPAR EPSCs were also recorded at holding potentials ranging from −65 to +40 mV in the presence of picrotoxin (100 μM), APV (50 μM), and intracellular spermine (10 μM) included in Cs-gluconate–based internal. Normalized current-voltage (I-V) curves were then generated by dividing the AMPAR EPSC peak amplitudes by the mean of AMPAR EPSC peak amplitude recorded at −65 mV. AMPAR EPSC rectification was determined by dividing peak AMPAR EPSCs amplitudes recorded at −65 mV by those recorded at +40 mV.

The excitatory and inhibitory balance (E/I ratio) was recorded with a Cs-gluconate–based internal solution similar to mEPSC recordings. Evoked EPSCs and IPSCs from LHb neurons were recorded in the same neuron in drug-free ACSF using a stimulating electrode placed in the stria medullaris. EPSCs were recorded at the reversal potential for GABAA IPSCs (−55 mV), and IPSCs were recorded at the reversal potential for EPSCs (+10 mV) in the same LHb neuron. The E/I ratio was then calculated as EPSC/IPSC amplitude ratio by dividing the average peak amplitude of 10 consecutive sweeps of EPSCs or IPSCs from the same recording. The cell input resistance and series resistance were monitored through all the experiments, and if these values changed by more than 10%, then data were not included.

Antalarmin, CYPPA, 1-EBIO, WIN55,212-2, and AM251 were prepared as stock solution in dimethyl sulfoxide and diluted (1:10,000) to final concentration in ACSF of 1 μM, 10 μM, 300 mM, 2 μM, and 10 μM, respectively. Stock solutions for apamin, rat CRF, iberiotoxin, and antisauvagine-30 were prepared in distilled water and diluted (1:1000) to final concentration in ACSF of 100 nM, 250 nM, 100 nM, and 25 nM, respectively. The intrapipette concentrations of BAPTA, GDP-β-s, and PKI(6-22) were 30 mM, 300 μM, and 10 μM, respectively. In some of our control interleaved experiments, AP recordings were continued for an hour without addition of drugs and no significant changes were observed over time. For all drug experiments, a baseline depolarization-induced AP recording/mIPSC/mEPSC was recorded in each neuron, and then the appropriate drug was added to the slice by the perfusate and AP generation in response to depolarizing current steps was again tested 25 to 30 min after. For apamin and CYPPA experiments, a second baseline with apamin or CYPPA was obtained before the addition of CRF and AP generation was reevaluated 25 to 30 min after CRF application. CRF washout experiments in supplementary figures were performed after 15 min of bath application of CRF, and the recordings continued at least for 90 min after the initiation of washout of CRF.

Western blotting

The LHb was dissected bilaterally from sagittal slices (300 μM) of non-MD or MD rats in ACSF and then snap-frozen in liquid nitrogen and stored at −80°C. Tissues were thawed, washed in ice-cold phosphate-buffered saline (PBS), and lysed in radioimmunoprecipitation assay buffer containing protease inhibitors (Sigma). Samples were then sonicated, incubated on ice for 30 min, and centrifuged at 10,000g for 20 min at 4°C. Protein concentration in the supernatant was determined by the Pierce BCA Protein Assay Kit (Life Technologies). Equal amounts of protein (20 μg) were combined with loading buffer, boiled for 5 min, and loaded onto 4 to 20% precast polyacrylamide gel (Bio-Rad Laboratories). Separated proteins were transferred onto nitrocellulose membranes, blocked with casein-based blocking reagent (I-Block, Life Technologies) for 60 min at room temperature, and then incubated overnight at 4°C with rabbit antibodies recognizing SKs [antibody against KCa2.1 for SK1, antibody against KCa2.2 for SK2, and antibody against KCa2.3 (C-term) for SK3, 1:300; Alomone Labs], antibody against PKA regulatory β2 subunit (1:5000; Abcam, ab75993), antibody against BK slo-1 (α1 subunit of BK) (1:1000; Millipore, MADN70) and GAPDH (1:1000; Cell Signaling 2118), or antibody against actin (1:10,000; Abcam, ab6276). Secondary antibodies used were horseradish peroxidase (HRP)–linked specific for rabbit (1:2000, Cell Signaling) and mouse (1:2000, Cell Signaling) immunoglobulin G (IgG). After incubation, the membranes were washed with PBS–Tween 20 and exposed to the appropriate HRP-linked secondary antibody (Cell Signaling). Blots were developed with Clarity Western ECL Substrate (Bio-Rad Laboratories) and detected using a Bio-Rad ChemiDoc Touch image acquisition system (Bio-Rad Laboratories). Data were analyzed using ImageJ software. Total abundance of target protein was normalized to appropriate endogenous control. All data were normalized to non-MD group, with summary data reported as fold change.

Quantitative reverse transcription polymerase chain reaction

Brain tissue from the LHb was collected by dissection from both non-MD and MD rats. Total RNA isolation was achieved through TRIzol (Life Technologies)/chloroform-based extraction. Samples were homogenized and then quantified using a NanoDrop spectrophotometer at 260- and 280-nm optical densities. Reverse transcriptase amplification of complementary DNA (cDNA) from total RNA was conducted with the High-Capacity RNA-to-cDNA Kit (Applied Biosystems) and Bio-Rad CFX96 Real-Time System. TaqMan gene expression assays (Thermo Fisher Scientific) with probes for Crh (catalog no. Rn01462137_m1), Crhr1 (catalog no. Rn00578611_m1), and actin (Actb) (catalog no. Rn00667869_m1) as the endogenous control were used to quantify these target genes during amplification. Each sample was run in triplicate. The comparative CT (ΔΔCT) was calculated by subtracting CT averages of the target gene from the respective endogenous control. For each sample calculated, the changes in target gene expression are given as 2−ΔΔCT. All data were normalized to the non-MD group and presented as fold change.

Data analysis

Values are presented as means ± SEM. Statistical significance was determined using unpaired or paired two-tailed Student’s t test, Mann-Whitney test, two-way ANOVA, or RM-ANOVA with Bonferroni post hoc analysis. The threshold for significance was set at *P < 0.05 for all analyses. The peak values of the evoked paired EPSCs were measured relative to the same baseline. A stable baseline value was considered in each sweep of paired pulses starting at 20 to 50 ms right before the emergence of the EPSC current using pCLAMP 10 software. Mini Analysis software was used to detect and measure mIPSCs and mEPSCs using preset detection parameters of mIPSCs and mEPSCs with an amplitude cutoff of 5 pA. The KS test was performed for the statistical analyses of cumulative probability plots of mEPSCs and mIPSCs. All statistical analyses were performed using IBM SPSS Statistics 24, GraphPad Prism 7, or Origin 2016.

SUPPLEMENTARY MATERIALS

www.sciencesignaling.org/cgi/content/full/11/520/eaan6480/DC1

Fig. S1. CRF-induced increases in intrinsic excitability of LHb neurons from non-MD rats persisted up to 1.5 hours after CRF washout.

Fig. S2. CRF-induced suppression of GABAergic transmission onto LHb neurons persisted up to 1.5 hours after CRF washout.

Fig. S3. CRF did not alter the excitatory/inhibitory balance in LHb neurons recorded from non-MD or MD rats.

Fig. S4. Intrapipette PKI(6-22) did not affect LHb neuronal excitability in non-MD rats.

Fig. S5. Western blot analysis of SK1 and SK3 in LHb homogenates from non-MD and MD rats.

Fig. S6. Western blot analysis of BK channel α subunit in LHb homogenates from non-MD and MD rats.

Fig. S7. MD potentiated glutamatergic synapses onto LHb neurons.

Fig. S8. Acute CRF depressed GABAergic transmission onto LHb neurons in MD rats.

Table S1. Summary of changes in passive and active membrane properties and AP characteristics of LHb neurons from non-MD and MD rats before and after CRF.

Table S2. Summary of changes in passive and active membrane properties and AP characteristics of LHb neurons from non-MD and MD rats before and after apamin.

Table S3. Summary of changes in passive and active membrane properties and AP characteristics of LHb neurons from non-MD and MD rats before and after SK channel activators.

Table S4. Summary of changes in passive and active membrane properties and AP characteristics of LHb neurons from non-MD and MD rats in intrapipette PKI(6-22) conditions.

Table S5. Summary of changes in passive and active membrane properties and AP characteristics of LHb neurons from non-MD and MD rats.

REFERENCES AND NOTES

Acknowledgments: The opinions and assertions contained herein are the private opinions of the authors and are not to be construed as official or reflecting the views of the Uniformed Services University of the Health Sciences or the Department of Defense or the government of the United States. We are grateful to D. Lovinger, V. Alvarez, B. Cox, and S. Gouty for their critical discussions for the present work. Funding: This work was supported by the NIH–National Institute of Drugs of Abuse (grant R01 DA039533), Brain and Behavior Research Foundation (formerly known as National Alliance for Research on Schizophrenia and Depression) grant, and Department of Defense intramural grant from the Uniformed Services University to F.S.N. The funding agencies did not contribute to writing this article or deciding to submit it. Author contributions: F.S.N. designed the research. M.E.A., L.D.L., R.D.S., H.K., and F.S.N. performed the experiments. C.A.B. and I.L. helped to design, perform, analyze, and interpret quantitative polymerase chain reaction experiments. M.E.A., L.D.L., R.D.S., and F.S.N. analyzed the data and prepared the figures. M.E.A., L.D.L., R.D.S., and F.S.N. wrote the paper. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper or the Supplementary Materials.
View Abstract

Navigate This Article