|Home | About | Journals | Submit | Contact Us | Français|
The neuropeptide galanin and its three receptor subtypes (GalR1–3) are expressed in the central amygdala (CeA), a brain region involved in stress- and anxiety-related behaviors, as well as alcohol dependence. Galanin also has been suggested to play a role in alcohol intake and alcohol dependence. We examined the effects of galanin in CeA slices from wild type (WT) and knockout (KO) mice deficient of GalR2 and both GalR1 and GalR2 receptors. Galanin had dual effects on GABAergic transmission, decreasing the amplitudes of pharmacologically-isolated GABAergic inhibitory postsynaptic potentials (IPSPs) in over half of CeA neurons but augmenting IPSPs in the others. The increase in IPSP size was absent after superfusion of the GalR3 antagonist SNAP 37889, whereas the IPSP depression was absent in CeA neurons of GalR1 × GalR2 double KO and GalR2 KO mice. Paired-pulse facilitation studies showed weak or infrequent effects of galanin on GABA release. Thus, galanin may act postsynaptically through GalR3 to augment GABAergic transmission in some CeA neurons, whereas GalR2 receptors likely are involved in the depression of IPSPs. Co-superfusion of ethanol, which augments IPSPs presynaptically, together with galanin caused summated effects of ethanol and galanin in those CeA neurons showing galanin-augmented IPSPs, suggesting the two agents act via different mechanisms in this population. However, in neurons showing IPSP-diminishing galanin effects, galanin blunted the ethanol effects, suggesting a preemptive effect of galanin. These findings may increase understanding of the complex cellular mechanisms that underlie the anxiety-related behavioral effects of galanin and ethanol in CeA.
Galanin is a C-terminally-amidated peptide of 29 amino acids in rodents that binds with high affinity to all three known galanin receptors (GalR1–3) belonging to the category of G-protein coupled receptors (GPCRs). Galanin-like immunoreactivity is widely distributed in the CNS, including the dorsal raphe nucleus (DRN), the locus coeruleus (LC), the CeA, several hypothalamic nuclei and the cholinergic cell bodies of nucleus basalis of Meynert (Melander et al., 1986). Galanin coexists with norepinephrine in CeA and several other brain regions (Melander et al., 1986), and galanin binding sites or expression corresponding to GalR1, GalR2 and GalR3 are observed in CeA (Melander et al., 1988). The CeA is implicated in several behaviors such as those related to fear, stress, anxiety and alcohol dependence (see e.g., (Davis et al., 1994; Pich et al., 1995; Roberto et al., 2010b). This nucleus contains a high proportion of GABA-containing neurons (Cassell et al., 1999; Cassell et al., 1986) and is known to be involved in substance abuse and dependence (see e.g., (Roberto et al., 2010b)).
However, to our knowledge there are no reports of electrophysiological effects of galanin on CeA neurons, despite behavioral studies suggesting that the galanin system in this nucleus could be important for the treatment of anxiety (Khoshbouei et al., 2002; Moller et al., 1999). Effects of galanin have been studied electrophysiologically in several other brain regions. For example, galanin inhibits muscarinic excitatory postsynaptic potentials (EPSPs) in the hippocampus, via presynaptic galanin receptors inhibiting acetylcholine release (Dutar et al., 1989; Fisone et al., 1987). Galanin also excites dorsal root ganglia neurons (Kerekes et al., 2003), but hyperpolarizes noradrenergic neurons in LC (Ma et al., 2001; Pieribone et al., 1995). Galanin 2–11 (Gal 2–11), a GalR2-selective agonist, had no effect on LC neurons (Pieribone et al., 1995; Xu et al., 2001). In a recent study from our laboratory, in most DRN neurons both galanin 2–11 and galanin 1–29 markedly decreased the amplitude of evoked GABAergic IPSPs (Sharkey et al., 2008). Paired-pulse facilitation (PPF) studies suggested that the galanin 1–29 effect was elicited presynaptically by reducing GABA release.
In addition to its suggested role in anxiety and depression (Lu et al., 2007; Ogren et al., 2006; Swanson et al., 2005), galanin and its receptors also have been reported to play some role in food and ethanol consumption (Chang et al., 2007; Karatayev et al., 2009; Karatayev et al., 2010; Leibowitz et al., 2003; Lewis et al., 2004), as well as opiate reinforcement (Picciotto et al., 2005). Thus in rodents, intraventricular injection of galanin or galanin overexpression increases alcohol intake, whereas a galanin knockout decreases ethanol intake (Chang et al., 2007; Karatayev et al., 2009; Karatayev et al., 2010; Lewis et al., 2004); furthermore, ethanol intake increases galanin mRNA in the hypothalamus whereas ethanol withdrawal decreases it (Leibowitz et al., 2003).
Therefore, the present studies were designed to clarify the role of galanin and its receptors in the CeA and increase understanding of the role of CeA in stress, anxiety and ethanol effects. Our results suggest that galanin alters the electrophysiological properties of CeA neurons in a dual manner that may explain the reported mixed effects of intra-CeA galanin microinjection in anxiety tests in rodents (Khoshbouei et al., 2002; Moller et al., 1999). We find no postsynaptic effect of galanin on averaged voltage-current curves, but it significantly alters the amplitude of evoked IPSPs in a dual manner probably involving multiple GalR subtypes. Our PPF studies suggest that these galanin effects may be explained by a postsynaptic action on GABA receptor function in some CeA neurons, but a small presynaptic effect in others. In addition, because we previously showed that superfusion of ethanol and the anxiety/stress-related peptide corticotropin releasing factor (CRF) both elicit the somewhat paradoxical effect of augmenting GABAergic IPSPs in CeA neurons (Nie et al., 2004; Roberto et al., 2010b; Roberto et al., 2003), we also examined the effects of co-applied ethanol on the responses to galanin in CeA neurons and find complex interactions depending upon CeA neuron responses to galanin.
We prepared brain slices containing CeA from male C57Bl/6J mice (25–30 g) that were anesthetized with halothane (3%) and decapitated, as previously described (Nie et al., 2004; Roberto et al., 2010b; Roberto et al., 2003). GalR1 and GalR2 double KO mice were obtained by crossing the GalR1 KO (Mitsukawa et al., 2009) with GalR2 KO (Lu et al., 2008) mice. With the investigator blind to the mouse strain/genetics, the brains were rapidly removed and placed into ice-cold artificial cerebrospinal fluid (ACSF) gassed with 95% O2 and 5% CO2. We cut transverse slices 400 μm thick on a LeicaVT 1000S vibrating slicer (McBain Instruments, Chatsworth, CA), incubated them in an interface configuration for about 30 min, and then completely submerged and continuously superfused (flow rate of 2–4 ml/min) them with warm (31°C), O2/CO2-gassed ACSF of the following composition in mM: NaCl, 130; KCl, 3.5; NaH2PO4, 1.25; MgSO4·7H2O, 1.5; CaCl2, 2.0; NaHCO3, 24; glucose, 10. The inner chamber had a total volume of 0.8 ml; at the 2–4 ml/min superfusion rates used, 90% replacement of the chamber solution could be obtained within 1 min. We added drugs to the ACSF from stock solutions to obtain known concentrations in the superfusate.
We recorded from CeA neurons with sharp micropipettes containing 3 M KCl (65–80 mΩ resistance) using current-clamp mode. Data were acquired with an Axoclamp-2A preamplifier (Axon Instruments (now Molecular Devices, Sunnyvale, CA)) and stored for offline analysis via pClamp software (Molecular Devices). We evoked pharmacologically-isolated GABAAergic IPSPs by stimulating locally within the medial sub-division of the CeA with a bipolar stimulating electrode, while continuously superfusing the glutamate receptor blockers 6,7-dinitroquinoxaline-2,3-dione (DNQX, 20 μM) and DL-2-amino-5-phosphonopentanoic acid (DL-AP5, 30 μM), and the GABAB antagonist CGP 55845A (1 μM).
We held the CeA neurons near their resting membrane potentials (RMPs), that ranged between −64 to −84 mV (mean: −78.6 ± 0.7 mV, n = 73), and applied hyperpolarizing and depolarizing current steps (200 pA increments, 750 ms duration) to generate voltage-current curves. To determine half-maximal IPSP amplitudes, we examined input/output (I/O) curves by measuring evoked IPSP amplitudes at 5 stimulus strengths, threshold to maximum stimulation. Subsequent analyses were done with averages of two IPSPs evoked with half-maximal stimuli. We measured the IPSP amplitudes before (control), during and after (washout) drug application, and we determined the percent change in IPSP amplitude at each stimulus intensity by the equation: (Vdrug / Vcontrol) * 100. The criteria for accepting a galanin or ethanol effect on IPSP amplitudes was a change of ≥ 10% of control at the half-maximal stimulus intensity over the 8–15 min superfusion time-points, and following galanin/ethanol washout (10–25 min).
We examined paired-pulse facilitation (PPF) in multiple neurons, using 100 ms interstimulus intervals and stimulus strength adjusted so that the amplitude of the first IPSP was 50% of maximal determined from the I/O relationship. We calculated PPF as the ratio of the second IPSP amplitude over that of the first IPSP, multiplied by 100. For PPF experiments we took measurements before galanin or ethanol superfusion (control), during (5–15 min) and after galanin/ethanol washout (10–25 min).
We express all values as mean ± SEM. Statistical analysis was performed with GraphPad Prism 5.0 software (GraphPad Software, San Diego, CA). We analyzed the data using a one-way ANOVA followed by a Dunnet's and Bonferroni post hoc test; p < 0.05 was considered statistically significant. In some cases, we also used Student's paired t-test; again, p < 0.05 was considered statistically significant.
CGP 55845A was a gift from Norvartis Pharma. We purchased D-AP5 and DNQX from Tocris Cookson (Holloway Road, MO), and bicuculline from Sigma (St Louis, MO). Galanin (Gal 1–29; GWTLNSAGYLLGPHAIDNHRSFSDKHGLT-amide) was synthesized by Biopeptide (San Diego, CA) and the GalR3 antagonist SNAP 37889 was synthesized by Edward Roberts at The Scripps Research Institute. We obtained ethanol from Remet (La Mirada, CA).
We recorded from a total of 85 mouse CeA neurons and, for some of the ethanol-galanin studies, 4 rat CeA neurons. In our CeA slices all cells were quiescent. The mean resting membrane potential (RMP) was −78.6 ± 0.7 mV. Most neurons in the CeA appear to contain GABA, and may be either interneurons or projection neurons (Cassell et al., 1999; Cassell et al., 1986). Therefore, the effects of galanin were initially assessed on the membrane properties and GABAergic IPSPs in these cells by superfusing a maximal 1 μM concentration of galanin (see Methods).
We fully analyzed 23 mouse CeA neurons for the effects of superfused galanin (1 μM) on voltage/current measurements. Interestingly, in all cells so tested from the medial sub-division of the mouse CeA, and as shown by the virtual superposition of averaged voltage-current (V/I) curves, we found no significant effect of 1 μM galanin on these V/I curves (Fig. 1), whether they were recorded from WT littermate controls or GalR2 KO mice (Fig. 1A), or from GalR1 × GalR2 KO mice or their WT littermate controls after block of GalR3 with SNAP 37889 (SNAP; Fig 1B; see below). SNAP was used to rule out the possibility that actions on GalR3 receptors might counteract actions of GalR1 or 2 on membrane conductance. As a first approximation, these findings suggest that galanin has little or no effect on membrane conductance in CeA neurons over a wide range of membrane potentials.
We assessed the effects of galanin on pharmacologically-isolated GABAA-IPSPs evoked from CeA neurons by a stimulating electrode placed in the medial sub-division of the CeA. In our initial studies on CeA neurons from normal WT C57Bl/6 and WT littermate mice (of GalR2 and GalR1 × GalR2) using a relatively long superfusion time (8–15 min), a maximal concentration of galanin (1 μM) significantly (p < 0.001) decreased the mean evoked IPSP amplitudes, in 11 of the 23 mouse CeA neurons studied (Fig. 2A and C), to a mean of 72% of control. In another sub-population of 5 WT CeA neurons pretreated with SNAP to block GalR3 receptors, galanin still significantly (p = 0.008) decreased IPSP size (Fig. 3).
In many WT CeA neurons galanin (1 μM) superfusion elicited an increase in IPSP amplitudes that was somewhat smaller but developed earlier (5–9 min) than the IPSP decrease seen in the neuronal population described above. The pooled data from 12 of 23 CeA cells from WT and littermate mice revealed that the mean evoked IPSP amplitude at half-maximal stimulus intensity was significantly increased (to 114% of control, n = 12; p < 0.05) by galanin superfused for 5–9 min (Fig. 2B & D). In both populations of neurons, subsequent superfusion of 30 μM bicuculline for 5 min usually completely abolished or greatly reduced the evoked IPSPs (Fig. 2B), demonstrating that they were dependent on GABAA receptor activation.
To initially assess the site of action for galanin in decreasing GABAA-IPSP amplitudes, we examined its effect on PPF of IPSPs in CeA neurons. PPF, as determined from the paired-pulse ratio, has been shown to vary inversely with the presynaptic release of transmitter (Bonci and Williams, 1997; Mennerick and Zorumski, 1995; Salin et al., 1996). In brief, our paired-pulse studies showed highly variable or no effects of galanin on PPF of IPSPs in most CeA neurons of WT, GalR1 × GalR2 KO or their littermate control mice; when averaged according the galanin effect on IPSP increases or decreases, there was no significant effect (data not shown). PPF was significantly changed (increased) by galanin only in CeA neurons of GalR2 KO mice, and only at the early superfusion time-point (data not shown), despite the apparent lack of effect of galanin on IPSP amplitudes in this set of CeA neurons. Because our voltage-current studies showed no effect of galanin on membrane properties, these findings suggest that the changes in IPSP amplitudes might arise in most CeA neurons from modulation of postsynaptic GABAA receptors.
To assess the possible role of GalR subtype-dependent effects in the dual actions of galanin in CeA, we used mice with knockout constructs of either GalR2 receptors alone, or mice with knockouts of both GalR1 and Gal2 receptors (GalR1 × GalR2 KOs). Since mice with knockouts of GalR3 receptors are not yet available, we also used one of the recently developed antagonists of these receptors (SNAP 37889; (Ogren et al., 1992; Swanson et al., 2005)) to block the GalR3 receptors pharmacologically.
Thus, in an initial set of 5 CeA neurons from normal WT mice, we assessed the effect of SNAP 37889 on the two types of responses of evoked IPSP amplitudes to galanin. As shown in figure 3, after superfusion of 200 nM SNAP 37889, subsequent co-perfusion of 1 μM galanin had only depressant effects on IPSP amplitudes, suggesting that the IPSP augmentations but not the IPSP depressions involve GalR3 activation.
In initial studies using GalR knockout mice, of 5 CeA neurons from the double GalR1 × GalR2 KO mice, 4 cells showed a clear galanin-evoked increase of IPSPs, suggesting involvement of GalR3, whereas 1 cell showed little change. No cells showed galanin-induced decreases of IPSP amplitudes. The mean IPSP increase for all 5 cells during galanin superfusion was 114 ± 5% of control, a significant effect (p < 0.05), again suggesting that the IPSP-augmenting galanin effect in WT mice may involve GalR3 receptors in most cells.
In another experiment superfusing SNAP 37889 together with galanin onto CeA neurons of GalR1 × GalR2 KO and wildtype littermate controls (n = 7 and 8, respectively), SNAP prevented the IPSP increase but not the IPSP decrease in the WT littermate CeA neurons (Figure 4). However, there was no significant galanin-induced effect on the IPSPs in the GalR1 × GalR2 KO neurons superfused with SNAP 37889. We conclude that the IPSP depression is absent in the CeA of double KO mice, but not in that of the WT littermates, even in the presence of the SNAP antagonist, suggesting that this effect is due to GalR1 and/or GalR2 receptor activation. The lack of an increase in IPSP amplitudes in the presence of the SNAP antagonist again suggests it is due to activation of GalR3.
To tease out whether GalR1 or GalR2 is involved in the IPSP depression, we tested CeA from mice with deletion only of GalR2 receptors. Thus, in CeA neurons from GalR2 KO mice, although we found some variability from cell to cell, there was no significant decrease of evoked IPSP amplitudes (n = 8; p = 0.15), even after 15 minutes of 1 μM galanin superfusion. However, 1 μM galanin robustly and significantly (n = 6; p < 0.05) decreased mean IPSP amplitudes in CeA neurons from the WT littermate mice (Fig. 5).
Thus by exclusion, our data showing that the GalR3 antagonist greatly reduces the early IPSP increase often seen in response to galanin, and that the more robust IPSP decrease is greatly diminished in CeA neurons from both GalR2 KO and the GalR1 × GalR2 KO mice, suggest that the IPSP decrease arises from activation of GalR2 receptors.
Because of the suggested role of brain galanin in ethanol intake (see Introduction), we also examined possible interactions of ethanol and galanin in mouse and rat CeA neurons. Our previous studies of rat (Roberto et al., 2010b; Roberto et al., 2003) and mouse (Bajo et al., 2008; Nie et al., 2004; Nie et al., 2009) CeA neurons indicate that ethanol reproducibly increases IPSP amplitudes, by increasing GABA release, in most CeA neurons. In the present studies, co-application of maximal concentrations for these two agents (1 μM galanin, 44 mM ethanol) resulted in a summation of effects only in a subpopulation of neurons (cf. figs. 6, ,7),7), suggesting that galanin and ethanol may interact on CeA GABAergic transmission via different cellular mechanisms. Thus, in a set of 12 CeA neurons from WT mice in which galanin decreased IPSP amplitudes, ethanol applied together with galanin only minimally reversed the IPSP amplitude decrease (to 87% of control) elicited by galanin alone (to 80%). Thus, galanin was somewhat preemptive over the ethanol effects (Fig. 6A and B), because ethanol alone generally augments CeA IPSPs to 130% or more of control levels. Interestingly, in this group of CeA neurons, both galanin and galanin with ethanol increased PPF to the same extent in most cells (Fig. 6C), similar to results with the GalR2 KOs, perhaps suggesting decreased GABA release; however this effect was not statistically significant.
In 7 WT mouse CeA neurons showing increases of evoked IPSP amplitudes by galanin alone, 44 mM ethanol together with 1 μM galanin further augmented the galanin effect by increasing IPSP amplitudes to 137% of baseline (from 117%; Fig. 7A, B), suggesting that ethanol and galanin effects involve different mechanisms of action (e.g., pre- versus postsynaptic) in this CeA population, with summation of maximal effects on IPSP size. Although again galanin alone seemed to increase PPF, this effect was not statistically significant; however, the difference between PPF during galanin alone and galanin with ethanol (which decreases PPF) was significant (fig 7C).
Of some interest is the rapid rebound reversal of ethanol effects on IPSP size in both neuron groups (i.e., showing either galanin-induced IPSP increases or decreases) on washout of the two agents. We interpret this rebound as resulting from a more rapid washout of ethanol compared to galanin, with the galanin effect persisting longer.
In a separate study (not shown) the co-perfusion of galanin and ethanol onto CeA neurons of Sprague Dawley rats gave results equivalent to those in mouse CeA: in the pooled data from 4 rat CeA neurons showing galanin depression of IPSPs, ethanol co-superfusion slightly reversed the galanin effect in decreasing mean IPSP amplitudes, but only partially back toward baseline, as in mouse CeA.
The widespread localization of galanin and its various receptors in the amygdala (Jhamandas et al., 1996; Jungnickel and Gundlach, 2005; Kohler et al., 1989a; Kohler et al., 1989b; Lu et al., 2005b; Melander et al., 1988; Mennicken et al., 2002) suggests that there are multiple cellular sites where the neuropeptide may influence CeA neuronal function. For example, the coexistence of galanin with noradrenaline in the LC and CeA (and with serotonin in the DRN; Xu and Hokfelt, 1997) may suggest that galanin is involved in the interactions between these two monoaminergic regions that play key roles in mood regulation. In the case of the CeA, these monoaminergic interactions with galanin may influence the local GABAergic neurons known to comprise most of the CeA population (Cassell et al., 1999; Cassell et al., 1986). Despite the evidence of norepinephrine and galanin coexistence in the CeA, the presence of multiple galanin receptor subtypes there, and behavioral effects elicited by galanin microinjection into CeA (Khoshbouei et al., 2002; Moller et al., 1999), to our knowledge there are no data available on the electrophysiological effects of galanin in CeA neurons. Therefore, in the present study we first examined the effect of the neuropeptide galanin and a GalR3-selective antagonist, in CeA neurons from WT and GalR1 and GalR2 knockout mice, on the electrophysiological properties and GABAergic IPSPs of these neurons.
In contrast to previous reports of hyperpolarizing effects of galanin in DRN and several other neuron types (Hokfelt et al., 1998; Swanson et al., 2005; Xu et al., 1998a; Xu et al., 2001; Xu et al., 1998b), galanin had no significant effect on current/voltage relationships in CeA neurons. However, we found that galanin clearly alters GABAergic neurotransmission in CeA neurons, but in a cell- and receptor-dependent manner. Interestingly, as with the effect of Gal 1–29 and Gal 2–11 in decreasing the amplitudes of isolated GABAA-IPSPs in most DRN neurons (Sharkey et al., 2008), galanin (Gal 1–29) also decreased IPSP amplitudes in many CeA neurons. However, in a large percentage of CeA neurons galanin measurably increased IPSP amplitudes, albeit to a lesser extent than the decrease in IPSPs seen in other cells. It should be noted that the long IPSP depressions seen in some CeA neurons, like those reported for Gal 1–29 decreases of IPSP amplitudes in DRN neurons (Sharkey et al., 2008), persisted and even increased over long application times, suggesting a stronger late effect and a lack of tachyphalaxis in both CeA and DRN neurons.
Our previous work on DRN neurons (Sharkey et al., 2008) suggested that the mechanisms whereby galanin receptor agonists altered GABAergic transmission may have depended on the receptor subtype activated: some GalR receptor subtypes may have been activated in a concentration-dependent manner such that longer exposure to the neuropeptide (and perhaps greater slice penetration) resulted in a greater effect. By contrast in CeA, selective activation of some receptor types (e.g., GalR3) may have a more transitory effect (e.g. increased IPSP amplitudes), peaking at shorter exposure times before desensitizing.
Further differences in the effects of galanin in CeA were seen in the inconsistent or lack of effects on PPF, known to be inversely related to neurotransmitter release (Bonci and Williams, 1997; Mennerick and Zorumski, 1995; Salin et al., 1996), except in CeA of GalR2 KO mice. In most CeA neurons from WT, GalR1 × GalR2 KO and littermate mice, galanin had little effect on PPF in CeA neurons, although CeA neurons of GalR2 KO mice showed galanin-evoked PPF increases, suggesting a possible presynaptic action on GalR1 and/or GalR3 receptors to regulate release of GABA, whereas galanin at GalR2 receptors may act postsynaptically. The localization of the GalR receptor subtypes is not known, due to lack of antibodies selective for receptor subtypes, but our data may suggest that some of the galanin receptors in the CeA may be on GABAergic terminals, whereas many could be on postsynaptic elements (e.g., GPCRs, other signalling pathways) that influence GABA receptor function.
The current study is the first to address the electrophysiological effects of galanin on the GABAergic system in the CeA. We have presented initial evidence for the differential activation of the CeA galanin receptors GalR2 and GalR3 that may have important implications for the development of novel galanin-based antidepressant or anti-anxiety treatments. Previous work showed that chronic antidepressant treatment (fluoxetine, 10mg/kg ip 14 days and electroconvulsive shock treatment 2 days) produced a shift in the GalR1 and GalR2 ratio towards GalR2, as well as an elevation of galanin levels (Lu et al., 2005a), suggesting that the galaninergic system is affected by, and may be involved in, antidepressant effects. Support for this idea is the recent finding that mice with a GalR2 knockout show a more persistent depressive-like phenotype than control WT mice (Lu et al., 2008). Differential activation of GalR2 and GalR3 by endogenous galanin may shed light on the mechanisms by which this antidepressant-induced change in GalR subtype ratio is important in the clinical efficacy of antidepressant treatments. Furthermore, the findings suggest that the development of subtype-specific GalR agonists or antagonists may be useful for novel antidepressant or antianxiety treatments, as recently shown for GalR3 antagonists as anti-anxiety agents (Lu et al., 2007; Ogren et al., 2006; Swanson et al., 2005).
In addition, our past (Nie et al., 2004; Roberto et al., 2010b) and present data may suggest that GalR2 activation or block of GalR3 favors reduction of GABAergic inhibitory function in CeA neurons, perhaps resulting in disinhibition of downstream GABAergic neurons(Bajo et al., 2008), that may provide a mechanism underlying this antidepressant action. Thus, the lack of an early increase in IPSP amplitudes in the presence of the SNAP antagonist suggests the increase is due to activation of GalR3. Notably, these galanin effects are consistent with the anxiogenic/anxiolytic profile we have recently hypothesized for the GABAergic neurons predominant in CeA, based our findings over the last 5–6 years: anxiolytic-like agents such as opioid (Kang-Park et al., 2009; Kang-Park et al., 2007) and orphanin FQ (Roberto and Siggins, 2006) agonists (e.g., enkephalin and nociceptin), and NPY and CB1 agonists (Roberto et al., 2010a; Roberto et al., 2008) (Gilpin et al, 2010, in submission) function in CeA to reduce presynaptic release of GABA, and thus should excite these GABAergic inhibitory neurons by disinhibition. By contrast, the stress- and anxiety-related peptide CRF increases evoked IPSP amplitudes, via presynaptic increases in GABA release (Nie et al., 2004; Roberto et al., 2010b), and thus should counter the effects of the `anxiolytic-like' agonists. Because most rodent CeA neurons contain GABA (with various co-localized peptides) and many of these GABA neurons project outside the nucleus (e.g., to bed nucleus of stria terminalis), local release of the opioids, nociceptin, NPY and cannabinoids (and now galanin acting on GalR2 receptors) in CeA will lead to the release of more GABA in downstream targets, thus inhibiting these areas. As the extended amygdala is known to be involved in stress-related disorders such as anxiety, our combined findings may help to solve the seeming paradox that anxiolytic-like substances reduce rather than enhance, and `anxiogenic' peptides increase, GABAergic transmission within CeA: this CeA profile would have opposite (i.e., more conventional) effects in downstream areas like BNST. The dual effects of galanin on GABAergic IPSPs in CeA might then explain reports of mixed anxiolytic/anxiogenic behavioral effects seen after micro-injection of galanin into CeA (Khoshbouei et al., 2002; Moller et al., 1999) and the anxiolytic-like effects of the GalR3 antagonists (Lu et al., 2007; Ogren et al., 2006; Swanson et al., 2005). The latter may reduce anxiety in part by antagonizing the GalR3 receptor-induced IPSP augmentation in CeA, leaving behind the decrease of IPSPs elicited by GalR2 receptors.
Notably, to date ethanol was the only allegedly anxiolytic agent we tested that increased IPSP amplitudes in CeA (Roberto et al., 2010b; Roberto et al., 2003; Roberto and Siggins, 2006). As noted above, the galanin augmentation of IPSPs found in some CeA neurons suggests the activation of GalR3 receptors, now thought to be involved in anxiety. These considerations and emerging literature suggesting a role for galanin and GalR2 receptors in ethanol intake (Chang et al., 2007; Karatayev et al., 2009; Karatayev et al., 2010; Leibowitz et al., 2003; Lewis et al., 2004) led us to compare and assess possible interactions of ethanol with galanin effects in CeA neurons. We indeed found such interactions between galanin and ethanol. These interactions were suggestive of a summation of like effects only in neurons showing IPSP augmentation by galanin. As no occlusive interactions occured in this population of neurons, we interpret the summated interactions as indicating involvement of different cellular or molecular mechanisms for the two agents in these cells. This interpretation is also supported by the inconsistent or lack of significant alteration of PPF by galanin in CeA of WT mice, in contrast to the PPF decreases always associated with the increased presynaptic GABA release elicited by ethanol (Roberto et al., 2010b; Roberto et al., 2003; Roberto and Siggins, 2006). However, in those CeA cells where galanin diminished IPSPs there was only slight summation of opposing effects by galanin plus ethanol; this interaction (see figure 6) suggests that galanin may have preemptive effects similar to those we previously found with nociceptin (Roberto and Siggins, 2006). The suggestion of PPF increases by galanin in these cells may also indicate that ethanol and galanin act on the same mechanism (e.g., GABA release) in this neuronal population.
Further studies will be required to elucidate the exact cellular sites of the galanin effects, the possible CeA neuron types involved in the two galanin opposing effects on GABAergic transmission reported here, and the possible cellular role of CeA galanin in ethanol effects. With respect to the latter, an important question is the effect of endogenous galanin on discrete Gal receptors in CeA and the effect of ethanol on the endogenous galanin-receptor interactions. The type of study reported by Lu et al (Lu et al., 2010), of the development of a putative GalR2-positive allosteric modulator (CYM2503) that would act on GalR2 receptors primarily via local endogenous galanin, may provide a path to answer such a question for the CeA.
We thank Drs. Scott Moore and Marisa Roberto for helpful comments on the manuscript, Dr. Edward Roberts for providing SNAP 37889, and Novartis Pharma AG for the gift of CGP 55845A. This work has been supported by NIH grants from NIMH (MH074055) and NINDS (NS063560) to TB, and from NIDA (DA03665) and NIAAA (U01 AA013498 under the INIA Consortium) to GRS.