Pituitary adenylate cyclase-activating polypeptide (PACAP) inhibits the slow afterhyperpolarizing current sIAHP in CA1 pyramidal neurons by activating multiple signaling pathways

The slow afterhyperpolarizing current (sIAHP) is a calcium-dependent potassium current that underlies the late phase of spike frequency adaptation in hippocampal and neocortical neurons. sIAHP is a well-known target of modulation by several neurotransmitters acting via the cyclic AMP (cAMP) and protein kinase A (PKA)-dependent pathway. The neuropeptide pituitary adenylate cyclase activating peptide (PACAP) and its receptors are present in the hippocampal formation. In this study we have investigated the effect of PACAP on the sIAHP and the signal transduction pathway used to modulate intrinsic excitability of hippocampal pyramidal neurons. We show that PACAP inhibits the sIAHP, resulting in a decrease of spike frequency adaptation, in rat CA1 pyramidal cells. The suppression of sIAHP by PACAP is mediated by PAC1 and VPAC1 receptors. Inhibition of PKA reduced the effect of PACAP on sIAHP, suggesting that PACAP exerts part of its inhibitory effect on sIAHP by increasing cAMP and activating PKA. The suppression of sIAHP by PACAP was also strongly hindered by the inhibition of p38 MAP kinase (p38 MAPK). Concomitant inhibition of PKA and p38 MAPK indicates that these two kinases act in a sequential manner in the same pathway leading to the suppression of sIAHP. Conversely, protein kinase C is not part of the signal transduction pathway used by PACAP to inhibit sIAHP in CA1 neurons. Our results show that PACAP enhances the excitability of CA1 pyramidal neurons by inhibiting the sIAHP through the activation of multiple signaling pathways, most prominently cAMP/PKA and p38 MAPK. Our findings disclose a novel modulatory action of p38 MAPK on intrinsic excitability and the sIAHP, underscoring the role of this current as a neuromodulatory hub regulated by multiple protein kinases in cortical neurons. © 2013 The Authors. Hippocampus Published by Wiley Periodicals, Inc.


INTRODUCTION
Action potentials in hippocampal pyramidal neurons are followed by an afterhyperpolarization (AHP) that has three kinetically distinct phases: fast (fAHP), medium (mAHP), and slow AHP (sAHP) (reviewed in Sah and Faber, 2002;Stocker et al., 2004). The sAHP has a timecourse of seconds (1-5 s), is evident after bursts of action potentials, and its amplitude is enhanced with increasing number of spikes in the burst (Alger and Nicoll, 1980;Hotson and Prince, 1980;Schwartzkroin and Stafstrom, 1980;Lancaster and Adams, 1986). The sAHP is mediated by a slow calcium-dependent potassium current, sI AHP , that contributes to the late phase of spike frequency adaptation in hippocampal pyramidal neurons (Madison and Nicoll, 1982;Lancaster and Adams, 1986;Lancaster and Nicoll, 1987). A trademark feature of the sI AHP is its modulation by several neurotransmitters activating various signal transduction pathways. The modulation of sI AHP by monoamine transmitters through the cyclic AMP (cAMP)-protein kinase A (PKA) pathway has been extensively studied (reviewed in Stocker et al., 2004). Beside monoamine transmitters, some neuropeptides expressed in interneurons and released in the hippocampal formation inhibit the sI AHP and reduce spike frequency adaptation in pyramidal neurons. These include the vasoactive intestinal peptide (VIP, Haas and Gahwiler, 1992), the corticotropin releasing factor (CRF, Aldenhoff et al., 1983) and the calcitonin gene-related peptide (CGRP, Haug and Storm, 2000), all exerting their inhibitory action on sI AHP via the cAMP-PKA pathway (Haug and Storm, 2000).
The neuropeptide pituitary adenylate cyclaseactivating polypeptide (PACAP) belongs to a superfamily of structurally related peptide hormones that include glucagon, glucagon-like peptides, secretin and growth hormone-releasing hormone (Harmar, 2001). PACAP consists of two isoforms, PACAP-27 with 27 and PACAP-38 with 38 amino acids, sharing identical sequences for the initial 27 amino acids. Both isoforms were first identified and isolated from ovine hypothalamic tissue (Miyata et al., 1989(Miyata et al., , 1990) and subsequently found to be preserved from protochordates to mammals (Vaudry et al., 2009). PACAP exhibits a wide range of functions and acts as a neurotransmitter and trophic factor in the CNS (Vaudry et al., 2009). The first 28 amino acids of PACAP-38 show 68% sequence identity with VIP (Miyata et al., 1989(Miyata et al., , 1990, suggesting that VIP and PACAP might mediate their biological effects through shared receptors. Two classes of PACAP receptors have been characterized based on their relative affinity for PACAP and VIP. PACAP has a potency >100-fold higher than VIP at the recombinant PAC 1 receptor, whereas it displays a potency similar to VIP at the recombinant VPAC receptors (VPAC 1 and VPAC 2 ) in radioligand binding assays (reviewed in Vaudry et al., 2009;Harmar et al., 2012). PAC 1 and VPAC receptors are coupled to the heterotrimeric G-protein Ga s and increase cAMP levels (reviewed in Harmar, 2001).
PACAP and its receptors are widely expressed in the CNS. PACAP mRNA (Skoglosa et al., 1999;Jaworski and Proctor, 2000;Hannibal, 2002) and immunoreactivity (Koves et al., 1991;Piggins et al., 1996;Hannibal, 2002) have been observed in the rat hippocampus, in the cell body layers of the CA1-CA3 fields, in dentate gyrus granule cells, mossy cells and pyramidal basket cells. Additionally, PACAP immunoreactive fibers have been detected both in the CA1-CA3 fields and in the dentate gyrus. In the hippocampus, PACAP might therefore be released by neurons (pyramidal cells or interneurons) within the hippocampal formation and/or by afferents coming from PACAP-rich regions, such as the hypothalamus (Piggins et al., 1996). Northern blot analysis and in situ hybridization revealed the presence of PAC 1 (Hashimoto et al., 1993;Spengler et al., 1993), VPAC 1 (Ishihara et al., 1992) and VPAC 2 (Lutz et al., 1993) receptor mRNA in the hippocampal formation. Moreover, PACAP binding sites and receptors were detected in the CA1-CA3 regions and dentate gyrus of the hippocampus in autoradiographic studies (Cauvin et al., 1991;Masuo et al., 1992). Immunohistochemistry showed that VPAC 2 receptors were highly expressed in dentate gyrus granule cells and in CA1-CA3 pyramidal cells, whilst PAC 1 and VPAC 1 were expressed in the same cells but at lower levels (Joo et al., 2004).
The presence of PACAP and PACAP receptors in the hippocampus prompted us to hypothesize that PACAP might modulate the sI AHP in hippocampal pyramidal neurons, thereby affecting their excitability and firing pattern. In this study we have investigated the effect of PACAP on the sI AHP and the signal transduction pathway used in rat CA1 pyramidal neurons.

MATERIALS AND METHODS
Male Sprague Dawley rats, 21-to 24-days old, were used to prepare transverse hippocampal brain slices. Rats were anesthe-tized with isoflurane and decapitated according to the UK Home Office Animal Scientific Procedures Act (1986), and protocols were reviewed and approved by the University College London Animal Ethical Committee.
Whole-cell patch clamp recordings were performed from CA1 pyramidal neurons using the blind patching technique (Blanton et al., 1989). All whole-cell patch clamp recordings were conducted with an EPC10 amplifier (HEKA, Germany) and the software Pulse for data acquisition. Slices were perfused with ACSF (in mM: 125 NaCl, 1.25 KCl, 2.5 CaCl 2 , 1.5 MgCl 2 , 1.25 KH 2 PO 4 , 25 NaHCO 3 , and 16 glucose) and the sI AHP was recorded at room temperature with patch pipettes made of borosilicate glass and containing a gluconate-based intracellular solution (in mM: 135 K-gluconate, 10 KCl, 10 HEPES, 1 MgCl 2 , 2 ATP-Na, and 0.4 GTP-Na). The liquid junction potential for the intracellular solution is 211 mV. Voltage values reported here are not corrected for the liquid junction potential. Only cells with a resting membrane potential 250 mV and a series resistance 30 MX were included in this study.
The sI AHP was elicited by stepping to 110-30 mV for 100-150 ms from a holding potential of 250 mV. Recordings were performed in the presence of 0.5 mM tetrodotoxin (TTX) to block voltage-gated sodium channels, and 1 mM tetraethylammonium (TEA) to block a subset of voltage-gated potassium channels to increase the calcium influx and thereby the calcium dependent sI AHP . Traces were filtered at 0.25 kHz and sampled at 1.25 kHz and the stimulus was repeated with a 30 s interval. Bovine serum albumin (BSA) was added to the ACSF (25 mg ml 21 ) when peptides were used extracellularly to minimize unspecific binding. Peptides were applied by bath perfusion. It is worth noticing that the actual concentration of the peptide/s at the relevant receptors might differ from that nominally applied, due to the size and diffusion of the peptide/s to neurons located at different depths in the brain slice and, possibly, to the action of extracellular peptidases. Protein kinase inhibitors were applied intracellularly through the patch pipette. DMSO up to 0.13% did not affect either the sI AHP amplitude or charge transfer, or the effect of PACAP-27 on the sI AHP amplitude or charge transfer when applied intracellularly through the patch pipette as a vehicle to dissolve some protein kinase inhibitors.
Cells were excluded if the series resistance changed significantly (>25%) in the course of experiments. Modulation of sI AHP was determined as changes in amplitude and charge transfer of sI AHP . The sI AHP amplitude was measured by calculating the mean current between two cursors placed 50 ms apart, 1 s after the initiation of the voltage step to avoid contamination by other, faster potassium currents. Charge transfer was determined as the area under the curve starting from the sI AHP peak until full decay had occurred.
Statistical analysis was performed with GraphPad InStat. Data are presented as mean with error bars indicating the standard error of the mean (S.E.M.), when not otherwise specified.

RESULTS
The sI AHP was recorded in the whole-cell patch-clamp configuration from 152 CA1 pyramidal neurons. CA1 neurons had an average resting membrane potential of 259.8 6 3.5 mV (n 5 84; mean 6 SD) and an input resistance of 175.4 6 51.1 MX (n 5 80; mean 6 SD). The sI AHP amplitude was on average 40.6 6 24.2 pA and the charge transfer 101.4 6 63.8 pC (n 5 84; mean 6 SD).
The region responsible for receptor activation of PACAP-38 resides in its N-terminal 27-amino acid sequence, which corresponds to the naturally occurring isoform PACAP-27 (Miyata et al., 1990). We therefore tested the effect of PACAP-27 (500 nM) on sI AHP . PACAP-27 suppressed sI AHP within 15-20 min of bath application (Fig. 1B,D), and no reversibility was observed within 15-35 min of wash-out. PACAP-27 suppressed the sI AHP amplitude by 86.1% 6 4.7% and the charge transfer by 85.8% 6 5.1% (Fig. 1E, n 5 4). No significant difference between PACAP-27 and PACAP-38 was found with regard to the reduction in sI AHP amplitude (P > 0.06, unpaired t test) and charge transfer (P > 0.3, unpaired t test). We therefore decided to use PACAP-27 in the following experiments.
Beside inhibiting sI AHP , PACAP produced an inward shift in the holding current at the holding potential of 250 mV, corresponding to a depolarizing effect. PACAP-38 at 500 nM elicited an inward current of 16.0 6 2.1 pA in 7 out of 7 cells (n 5 7; P 5 0.0002). An inward current of 14 6 3.3 pA (n 5 18; P 5 0.0006) was observed in response to 250 nM PACAP-27 in 16 out of 18 cells. The inward current might be due to the modulatory action of PACAP on background (leak) channels, HCN channels, and/or other nonspecific cationic channels, in analogy to what has been observed for other neurotransmitters inhibiting the sI AHP in hippocampal neurons.
The inhibition of the sI AHP reduces late spike frequency adaptation in hippocampal pyramidal neurons (Madison and Nicoll, 1982;Madison and Nicoll, 1984). Application of 250 nM PACAP-27 reduced spike frequency adaptation in CA1 pyramidal neurons (Fig. 1G, n 5 7), in agreement with its inhibitory effect on the sI AHP .

Activation of PAC-1 Receptors Partially Mimics the Effect of PACAP on sI AHP
Neuronal PACAP receptors comprise PAC 1 , selective for PACAP and coupled to both adenylyl cyclases and phospholipase C, and VPAC 1 and VPAC 2 receptors, activated by both PACAP and VIP and coupled to adenylyl cyclases (Vaudry et al., 2009). All three PACAP receptor subtypes are expressed in CA1 pyramidal cells (Hashimoto et al., 1996;Shioda et al., 1997;Joo et al., 2004). To elucidate the specific contribution of PAC 1 receptors to the PACAP-mediated sI AHP inhibition in CA1 pyramidal cells, we used the PAC 1 selective agonist maxadilan, a peptide structurally unrelated to PACAP (Lerner et al., 1991), which has no activity at VPAC 1 and VPAC 2 receptors (Moro and Lerner, 1997). Maxadilan and PACAP activate PAC 1 receptors with a similar potency in the rat brain (Moro et al., 1996). When tested at 250 nM, the same concentration as used for PACAP-27, maxadilan reduced the sI AHP amplitude (Figs 2A,B) by 32.6% 6 3.9% (n 5 5; P 5 0.001, t test) and charge transfer by 24.4% 6 4.7% (n 5 5; P 5 0.007, t test) (Fig. 2D). This indicates that PAC 1 receptors contribute to the PACAP-induced suppression of sI AHP in CA1 pyramidal neurons.
Surprisingly, the reduction of the sI AHP as a consequence of PAC 1 receptor activation is not supported by the results obtained upon 250 nM PACAP-27 application in the presence of PACAP-(6-38). Under this condition, PACAP-27 reduced the sI AHP amplitude (79.2% 6 11.2%, n 5 5, P 5 0.002, t test) and the charge transfer (86.8% 6 8.5%, n 5 5, P 5 0.0005, t test) to a comparable extent (amplitude: P > 0.5, charge transfer: P > 0.3) as observed in the absence of PACAP-(6-38) (Fig.  1). The suppression of sI AHP by maxadilan but the lack of antagonism of PACAP-27 by PACAP-(6-38) suggests the activation of more than one receptor subtype in the PACAPmediated inhibition of this current.
Max.d.4 is a derivative of maxadilan, in which nineteen amino acids between positions 24 and 42 are deleted. It is a selective PAC 1 receptor antagonist that is more potent than PACAP-(6-38) (Moro et al., 1999). In the presence of 500 nM max.d.4, 250 nM PACAP-27 inhibited the sI AHP amplitude (69.8% 6 10.8%, n 5 4, P < 0.01, t test) and charge transfer (74.5% 6 7.3%, n 5 4, P < 0.01, t test) to an extent that was comparable (amplitude and charge transfer: P > 0.5) to the effect of PACAP-27 alone (Fig. 1). This further indicates that antagonism of PAC 1 receptors alone is not sufficient to prevent the PACAP-27 mediated inhibition of sI AHP , which is instead due to the activation of more than just PAC 1 receptors.

PACAP-27 Suppresses sI AHP Partly Through PKA Activation
In hippocampal neurons monoaminergic neurotransmitters such as noradrenaline, histamine, serotonin and dopamine inhibit the sI AHP through PKA activation Storm, 1993, 1995). Moreover, in CA1 pyramidal neurons CRF, VIP and CGRP have been shown to suppress sI AHP in a PKA-dependent manner (Haug and Storm, 2000). We therefore asked whether PKA was required to mediate the inhibitory effect of PACAP on sI AHP in hippocampal pyramidal neurons.
To address this question, we used the PKA inhibitor Rp-cAMPS added to the intracellular solution (Pedarzani and Storm, 1993). Application of PACAP-27 (250 nM) in the presence of Rp-cAMPS (500 mM) caused an inhibition of the sI AHP that was attenuated when compared to the effect of PACAP-27 alone (Fig. 3A,B). The inhibition of the sI AHP amplitude was reduced from 77.4% 6 6.7% in the absence of Rp-cAMPS (Figs. 1F and 3C) to 56.8% 6 5.9% in the presence of Rp-cAMPS (Fig. 3C). Similarly, the sI AHP inhibition by 500 nM PACAP-27 was reduced from 86.1% 6 4.7% in the absence (Fig. 1F and 3C) to 42.7% 6 11.1% in the presence of Rp-cAMPS (Fig. 3C). An inward shift in the holding current of 16.6 6 6.8 pA (n 5 15; P 5 0.002) was observed upon application of PACAP-27 (250 nM) in the presence of Rp-cAMPS.
To ensure that Rp-cAMPS inhibited PKA under our experimental conditions, we used isoproterenol (1 mM), a b-adrenergic receptor agonist that inhibits the sI AHP by activation of the cAMP/PKA pathway (Pedarzani and Storm, 1993). The sI AHP inhibiton caused by isoproterenol (100% 6 0.02%, n 5 4) was significantly reduced in the presence of Rp-cAMPS (41.2% 6 8.4%, n 5 5, P 5 0.0005, unpaired t test). When isoproterenol was applied after PACAP-27 in the presence of Rp-cAMPS, the sI AHP was inhibited to 43.1% 6 7% (n 5 5) of the amplitude left after PACAP-27 inhibition, showing that Rp-cAMPS inhibited PKA. These results show that PKA contributes to the PACAP-27-mediated inhibition of the sI AHP .

Are MAP Kinases Involved in the PACAP-27-Mediated Inhibition of sI AHP ?
In a Drosophila mutant PACAP modulates a potassium current through the activation of a mitogen-activated protein (MAP) kinase pathway (Zhong, 1995). PACAP has also been shown to activate the MAP kinase kinase (MEK)/extracellularsignal-regulated kinase (ERK) pathway in cultured cerebellar granule cells (Villalba et al., 1997). Furthermore, PACAP-38 activates p38 MAP kinase (p38 MAPK) in mouse cultured cerebellar granule cells (Ster et al., 2007). We therefore asked whether the effect of PACAP on the sI AHP could be mediated by the activation of the ERK or p38 MAPK pathway.
Next, we tested the hypothesis of an involvement of p38 MAPK in the PACAP-mediated inhibition of sI AHP . In the presence of SB 203580 (20 mM), a p38 MAPK inhibitor (Eyers et al., 1998), PACAP-27 (250 nM) partly inhibited the sI AHP (Fig. 4A). Figure 4B shows the time course of the reduction in sI AHP amplitude from the experiment in Figure 4A before, during, and after PACAP-27 application. When compared with experiments performed in the absence of SB 203580, the inhibition of the sI AHP amplitude by PACAP-27 was reduced in the presence of the p38 MAPK inhibitor (31.3% 6 11.2%, n 5 7, P 5 0.006, Mann-Whitney test, Fig.  4C). Similar experiments were performed with an increased concentration of PACAP-27 (500 nM). In this case SB 203580 did not reduce the PACAP-27-induced inhibition of the sI AHP amplitude (70.4% 6 11.6%) when compared with the sI AHP inhibition induced in the absence of the p38 MAPK inhibitor (n 5 5, P > 0.2, unpaired t test). SB 203580 largely prevented the inward shift in the holding current (4.8 6 2.9 pA; n 5 7; P 5 0.14) caused by PACAP-27 (250 nM).
Taken together, these results indicate that p38 MAPK is involved in mediating the inhibitory effect of PACAP on the sI AHP in CA1 pyramidal neurons, while a contribution by the MEK/ERK1/2 pathway seems unlikely.

Does PACAP Activate p38 MAP Kinase and PKA in the Same or in Parallel Pathways Resulting in sI AHP Inhibition?
Both the inhibition of PKA by Rp-cAMPS (Fig. 3) and of p38 MAPK by SB 203580 (Fig. 4) result in a decrease of the PACAP-27-induced sI AHP suppression. This prompts the question as to whether PACAP-27 activates p38 MAPK and PKA as part of the same or two parallel pathways leading to sI AHP suppression. If PKA and p38 MAPK operate sequentially as part of the same pathway, their concomitant inhibition should not have a stronger impact on the effect of PACAP-27 than inhibition of each single kinase. Conversely, if PKA and p38 MAPK act through parallel, independent pathways their combined inhibition should have a stronger impact on the effect of PACAP-27.
The PKA inhibitor Rp-cAMPS and the p38 MAPK inhibitor SB 203580 were included together in the intracellular solu-tion, and 250 nM PACAP-27 was applied for 15 min, causing a partial inhibition of the sI AHP (Fig. 5B). The PACAP-27-induced inhibition of sI AHP was attenuated in the presence of Rp-cAMPS and SB 203580 (Fig. 5A). The reduction of the PACAP-27 effect on the sI AHP upon combined inhibition of PKA and p38 MAPK was significant compared with control conditions (51.8% 6 4.9%; n 5 7; P 5 0.02, Mann-Whitney test; Fig. 5C). However, the PACAP-27mediated suppression of sI AHP amplitude observed in the presence of Rp-cAMPS and SB 203580 was not significantly different from that observed in the presence of Rp-cAMPS alone (compare Figs. 5C and 3C) or SB 203580 alone (compare Figs. 5C and 4C) (Kruskal-Wallis Test-nonparametric ANOVA: P 5 0.105). The observation that the combined inhibition of PKA and p38 MAPK does not prevent the PACAP-27 reduction of the sI AHP to a larger extent than the inhibition of each single kinase suggests that they act sequentially as part of the same signal transduction pathway. cAMP activates not only PKA but also "exchange proteins directly activated by cAMP" (EPAC) (Rehmann et al., 2003), which can lead to the activation of small G-proteins and p38 MAPK (Shi et al., 2006;Ster et al., 2007). Activation of p38 MAPK by EPACs would be in parallel to the one by PKA (Fig. 8). If indeed PKA and p38 MAPK act sequentially in the same pathway to mediate the inhibition of sI AHP by PACAP-27, as suggested by the combined inhibition of PKA and p38 MAPK (Fig. 5), we would expect the EPAC contribution to this pathway to be minor or negligible. To distinguish between PKA and EPAC-mediated effects, we applied the highly specific EPAC superactivator 8CPT-O-Me-cAMP (Rehmann et al., 2003). 8CPT-O-Me-cAMP (5 mM) reduced the sI AHP amplitude by 10.3% 6 5.2% (Fig. 6A-C, n 5 7). This reduction was not significant (P > 0.05). To assess the integrity of the PKA meditated modulation of the sI AHP we applied to the same cells the membrane permeable PKA activator 8CPT-cAMP (Fig. 6A,B). PKA activation by 8CPT-cAMP led to a suppression of the sI AHP by 92.8% 6 3.4% (n 5 7, Fig. 6C), showing the functionality of the PKA pathway. The lack of current inhibition by EPAC activation excludes a contribution by EPACs to the modulation of the sI AHP . This supports the finding that PKA and p38 MAPK are arranged sequentially to mediate the inhibition of the sI AHP by PACAP (Fig. 8).

PKC is not Involved in the sI AHP Inhibition Mediated by PACAP-27
Besides their coupling to the cAMP/PKA and MAP kinase pathways, PACAP receptors have been shown to activate phospholipase C (PLC) linked signaling cascades (Spengler et al., 1993;Harmar, 2001;reviewed in Vaudry et al., 2009). Protein kinase C (PKC) is generally thought to act downstream of PLC (Nishizuka, 1992) and its activation suppresses the sI AHP in hippocampal pyramidal neurons (Malenka et al., 1986). We therefore investigated whether, beside PKA and p38 MAPK, also PKC is involved in the sI AHP suppression induced by PACAP-27.
To obtain maximal PKC inhibition, a combination of two specific inhibitors acting at different sites on PKC, 20 mM chelerythrine (Herbert et al., 1990) and 500 nM bisindolylmaleimide I (BIM-1) (Toullec et al., 1991), was applied intracellularly. Inhibition of sI AHP amplitude (89.3% 6 7.7%, n 5 5) by PACAP-27 (250 nM) was not significantly different compared to recordings performed in the absence of chelerythrine and BIM-1 (Fig. 7A,B, and D; P > 0.3, Mann-Whitney test). An inward shift in the holding current of 29.7 6 9.1 pA  and 8CPT-cAMP (5 mM) on the sI AHP amplitude in the same cell as in A. (C) Bar diagram summarizing the effects of 8CPT-O-Me-cAMP (5 mM; black bar) and subsequently applied 8CPT-cAMP (5-50 mM; white bar) on the sI AHP amplitude in seven cells. The 8CPT-cAMP was used at a concentration of 50 mM in 6 out of 7 cells. The suppression of sI AHP caused by 8CPT-cAMP was significantly larger than that observed in response to 8CPT-O-Me-cAMP (P < 0.0001, n 5 7, unpaired t test). Error bars indicate S.E.M. * indicates statistical significance.
(n 5 5; P 5 0.03) was observed upon application of PACAP-27 (250 nM) in the presence of the PKC inhibitors. To validate PKC inhibition under our experimental conditions, we activated PKC by phorbol-12,13-dibutyrate (PDBu). Under control conditions, PDBu (500 nM) caused a strong inhibition of the sI AHP amplitude (74.4% 6 3.3%; n 5 5), which was significantly reduced (34.7% 6 15.4%, n 5 6, P 5 0.04, unpaired t test) in the presence of chelerythrine (20 mM) and BIM-1 (500 nM). This indicates that PKC was indeed inhibited under these experimental conditions. These experiments show that the PKC signaling cascade does not significantly contribute to the inhibition of the sI AHP by PACAP.
Is a Crosstalk Between the PKA and PKC Pathways Involved in the Inhibition of sI AHP by PACAP?
The lack of PKC involvement in the suppression of the sI AHP by PACAP inferred from the experiments where only PKC was inhibited does not entirely exclude the possibility of a function of PKC in the context of a crosstalk with the PKA/p38 MAPK pathway. To elucidate this hypothesis we applied PACAP-27 in the presence of both PKA and PKC inhibitors. The inhibition of sI AHP by PACAP-27 was reduced to 56.4% 6 5.0% in the presence of Rp-cAMPS, chelerythrine and BIM-1 (Fig. 7C,D; n 5 10, P 5 0.03, Mann-Whitney test). The inhibition of the sI AHP amplitude by PACAP-27 was similar in the presence of Rp-cAMPS alone and the combination of Rp-cAMPS, chelerythrine and BIM-1 ( Fig. 7D; P > 0.8, Mann-Whitney test). By contrast, inhibition of PKC alone by chelerythrine and BIM-1 with no Rp-cAMPS did not reduce the PACAP effect to a similar extent ( Fig. 7D; P 5 0.003 Mann-Whitney test).
These results suggest that the involvement of PKC in the PACAP-mediated inhibition of the sI AHP is unlikely also as part of a crosstalk.

DISCUSSION
This study shows that PACAP increases the excitability of CA1 pyramidal neurons by inhibiting the slow Ca 21 -activated K 1 current sI AHP through the activation of multiple signaling pathways, most prominently cAMP/PKA and p38 MAPK. Our findings reveal a novel modulation of the sI AHP by p38 MAPK in CA1 pyramidal neurons, adding to the role of this current as a convergency point for modulatory inputs mediated by multiple protein kinases in central neurons.
Previous studies have identified the slow afterhyperpolarization in hippocampal neurons as a target for the neuromodulatory actions of several neuropeptides, including CRF (Aldenhoff et al., 1983;Haug and Storm, 2000), VIP (Haas and Gahwiler, 1992;Haug and Storm, 2000), and CGRP (Haug and Storm, 2000). Similar effects on the sI AHP have been reported for CRF, VIP and PACAP in layer II/III and V neocortical pyramidal neurons (Hu et al., 2011). Our study shows that in CA1 pyramidal neurons both PACAP isoforms, PACAP-38 and PACAP-27, suppress the sI AHP , leading to an enhancement of intrinsic excitability and attenuation of spike frequency adaptation. The suppression of the sI AHP by PACAP shown in our study provides a cellular mechanism for the increase in spontaneous firing of CA1 neurons observed in vivo upon application of PACAP-38 (Di Mauro et al., 2003).
CA1 pyramidal neurons express all three types of PACAP receptors, PAC 1 , VPAC 1 and VPAC 2 (Hashimoto et al., 1996; Fig. 1F, reported here for comparison) and in the presence of Chelerythrine 1 BIM-1 (n 5 5; black bar), Rp-cAMPS (n 5 15; grey bar; same data as in Fig. 3C, reported here for comparison), and Chelerythrine 1 BIM-1 1 Rp-cAMPS (n 5 10; stripy bar). Chelerythrine and BIM-1 did not prevent the suppression of sI AHP by PACAP-27 and did not affect the partial inhibition of the PACAP-27 effect by Rp-cAMPS. Error bars indicate S.E.M. * indicates statistical significance. Shioda et al., 1997;Joo et al., 2004). The 30% reduction of the sI AHP amplitude we observed upon application of the selective PAC 1 agonist maxadilan suggests a potential contribution of PAC 1 receptors. The presence of the PAC 1 /VPAC 2 antagonist PACAP-(6-38) (Dickinson et al., 1997;Moro et al., 1999) indeed prevented the effect of maxadilan, supporting a potential role for PAC 1 receptors. Additionally the activation of VPAC 1 , but not VPAC 2 receptors, leads to the suppression of the sI AHP amplitude, because the activation of PAC 1 receptors alone by maxadilan is not sufficient to inhibit the sI AHP to a similar extent as caused by PACAP. The involvement of VPAC 1 receptors was further inferred by the observation that the PAC 1 /VPAC 2 antagonist PACAP-(6-38) did not prevent the inhibition of the sI AHP by PACAP. This is different from what has been observed in the neocortex, where PACAP has been reported to activate the cAMP/PKA pathway by acting mainly through PAC 1 receptors (Hu et al., 2011). Our study concludes that the sI AHP suppression by PACAP can be elicited through the activation of PAC 1 and most likely VPAC 1 receptors in CA1 pyramidal neurons.
PAC 1 and VPAC 1 receptors can couple to multiple signal transduction pathways, often depending on the cell type and developmental stage. Both receptors activate adenylyl cyclases, leading to the production of cAMP in a variety of cells. Additionally, they can activate phospholipase C, leading to the formation of diacylglycerol and inositol trisphosphate, and phospholipase D (Fig. 8, reviewed in Vaudry et al., 2009). Because the closely related neuropeptide VIP suppresses the sI AHP by activating the cAMP/PKA pathway in CA1 pyramidal neurons (Haug and Storm, 2000) and PACAP has been reported to do the same in the neocortex (Hu et al., 2011), we initially expected PACAP to exert its inhibitory action on the sI AHP predominantly through the same pathway. Surprisingly, inhibition of PKA by the phosphodiesterase-resistant cAMP analogue Rp-cAMPS (Botelho et al., 1988) only partially prevented the effect of PACAP on the sI AHP . Interestingly, a similar partial inhibition by Rp-cAMPS was reported for the effect of VIP on the sI AHP (Haug and Storm, 2000). The residual sI AHP inhibition might be a consequence of Rp-cAMPS being a competitive PKA inhibitor that can be partly displaced by high levels of cAMP (Botelho et al., 1988). This would mean that high levels of cAMP, such as those produced by higher concentrations of PACAP, would weaken the inhibitory effect of Rp-cAMPS on PKA. However, our observation that Rp-cAMPS reduced the sI AHP inhibition by higher concentrations of PACAP if anything more than by low concentrations makes this an unlikely explanation. Given the coupling of PACAP receptors to multiple signaling pathways and the modulation of sI AHP by various protein kinases, including PKA, PKC, and calcium-calmodulin protein kinase II (reviewed in Stocker et al., 2004), it is plausible that PACAP suppresses this current by activating also other pathways beside cAMP/PKA. In neurons PACAP activates two different MAP kinases, ERK1/2 and p38 (Sakai et al., 2001;Shi et al., 2006;Monaghan et al., 2008;Stetler et al., 2010). Our data support a significant contribution of p38 MAPK to the PACAP effect on the sI AHP .
Interestingly, the contribution of p38 MAPK is more prominent than that of PKA for the low concentration of PACAP, while PKA seems to prevail at the high PACAP concentration. The different impact of p38 MAPK in response to different PACAP concentrations might reflect the recruitment of different signaling pathways by signals of different intensities (Dumaz and Marais, 2005). Conversely, inhibition of MEK/ ERK1/2 did not affect the inhibition of the sI AHP by PACAP-27 at any concentration tested. On the basis of our data, we propose a novel role for p38-MAPK in the modulation of the sI AHP and intrinsic excitability of CA1 pyramidal neurons.
If PACAP activated the PKA and p38 MAPK pathways in parallel, their concomitant inhibition should lead to an additive reduction of the sI AHP suppression. The lack of additive effect of PKA and p38 MAPK inhibitors in our study suggests that these two kinases act as part of the same signaling pathway, rather than being engaged by two separate pathways activated in parallel by PACAP and converging on the sAHP channels (Fig. 8). Evidence for a sequential action of PKA and p38 MAPK within the same pathway has been found in various cell types, where PKA has been proposed to activate p38 MAPK through an as yet unknown mechanism (Zheng et al., 2000;Delghandi et al., 2005;Shi et al., 2006). A similar signaling pathway might indeed underlie the effect of PACAP on the sI AHP that we have observed in CA1 pyramidal neurons. In view of the results obtained with PACAP in this study, it will be interesting to see whether p38 MAPK might contribute to Signal transduction pathway mediating the suppression of sI AHP by PACAP in CA1 pyramidal neurons. Schematic drawing illustrating the signal transduction pathway mediating the effect of PACAP on the sI AHP in CA1 pyramidal neurons. Pathway components supported by the results obtained in our study are depicted in black; other potential contributors, whose role was discounted by the results of our experiments, are shown in grey. AC, adenylyl cyclase; G s , G-alpha-s heterotrimeric G-protein; G q , G-alpha-q heterotrimeric G-protein; PLC, phospholipase C; DAG, diacylglycerol; IP 3 , inositol 1,4,5 trisphosphate; MKK, mitogenactivated protein kinase kinase.
the neuromodulatory effects also of other neuropeptides and neurotransmitters. Indeed, p38 MAPK has been shown to mediate the persistent sAHP suppression caused by prolonged stimulation of type 5 metabotropic glutamate receptors in CA3 pyramidal neurons (Young et al., 2008).
In other studies, p38 MAPK is a downstream substrate of the cAMP signaling pathway independent of PKA, through the activation of exchange proteins directly activated by cAMP (EPACs) and small G-proteins (Fig. 8, Shi et al., 2006;Ster et al., 2007). Our results, obtained by using a highly specific EPAC superactivator (Rehmann et al., 2003), do not support an involvement of EPACs in the modulation of sI AHP .
PAC 1 and VPAC 1 receptors are coupled to the phospholipase C (PLC) pathway, leading to the activation of PKC (Fig. 8, reviewed in Vaudry et al., 2009). PKC suppresses the sI AHP in hippocampal pyramidal neurons (Malenka et al., 1986). However, our data show that PACAP does not suppress sI AHP by activating PKC in CA1 pyramidal neurons. This argues in favor of selective coupling between PACAP receptors and specific signal transduction pathways in hippocampal neurons. The basis for this selectivity might lie in the existence of signaling microdomains defining the spatio-temporal dynamics of PACAP signaling, a topic that will be addressed in future studies.
In conclusion, our study shows that PACAP effectively modulates the intrinsic excitability and firing behavior of CA1 pyramidal neurons through the suppression of sI AHP . It is well established that a reduction in the sAHP in CA1 and CA3 neurons accompanies acquisition of hippocampus-dependent learning, whereas increases in the sAHP are correlated with cognitive impairment (reviewed in Disterhoft et al., 2004). The impact of the sI AHP on learning processes reflects its multiple roles in controlling firing and intrinsic excitability (Zhang and Linden, 2003) and integration of synaptic signals (Sah and Bekkers, 1996;Borde et al., 1999;Lancaster et al., 2001;Wu et al., 2004;Fernandez de Sevilla et al., 2007). Drosophila harboring a mutation in the PACAP-related gene amnesiac display deficits in associative learning (Quinn et al., 1979;Feany and Quinn, 1995). In the rat, PACAP intracerebroventricular administration enhances learning, possibly by affecting memory consolidation (Sacchetti et al., 2001). Mice lacking PAC 1 receptors display impaired long-term potentiation (LTP) of synaptic transmission at several synapses in the hippocampal formation (Otto et al., 2001;Matsuyama et al., 2003), correlating with a deficit in contextual fear conditioning, a hippocampus-dependent learning task (Sauvage et al., 2000;Otto et al., 2001;Matsuyama et al., 2003). PACAP regulates bidirectional synaptic plasticity at the Schaffer collateral CA1 synapse in a concentration-dependent manner, whereby subnanomolar concentrations of the peptide induce long-lasting facilitation of excitatory synaptic potentials , while concentrations similar to those used in this study (250-500 nM) cause either an initial decrease followed by an enhancement , or a long-lasting depression of excitatory synaptic transmission (Kondo et al., 1997;Ster et al., 2009) that depends on EPAC and p38 MAPK acti-vation (Ster et al., 2009). Taken together, our results suggest that modulation of the sI AHP via PAC 1 and VPAC 1 receptors might contribute to the complex effects of PACAP on synaptic plasticity in the hippocampal formation and ultimately on learning and memory consolidation.