Liu, Guo Jun and Barry W. Madsen. PACAP38 modulates activity of NMDA receptors in cultured chick cortical neurons. J. Neurophysiol. 78: 2231–2234, 1997. The outside-out recording mode of the patch-clamp technique was used to study modulatory effects of pituitary adenylate cyclase-activating polypeptide (PACAP38) on N-methyl-d-aspartate (NMDA) receptor activity in cultured chick cortical neurons. Biphasic concentration-dependent effects of PACAP38 on channel opening frequency induced by NMDA (20 μM) and glycine (1 μM) were found, with low concentrations (0.5–2 nM) of PACAP38 increasing activity and higher concentrations (10–1,000 nM) causing inhibition. These effects were reversible, reduced with higher concentrations of glycine (2–10 μM) but not by 200 μM NMDA, and inhibited by 10 μM 7-chlorokynurenic acid. In addition, 1 μM PACAP6–38 (a PACAP antagonist) inhibited channel activity due to 20 μM NMDA and 1 μM glycine by 66%, and this inhibition was reduced to 13% in the additional presence of 2 nM PACAP38. These observations suggest thatPACAP38 has a direct modulatory effect on the NMDA receptor that is independent of intracellular second messengers and probably mediated through the glycine coagonist site(s).
Pituitary adenylate cyclase-activating polypeptide (PACAP)is a recently identified member of a neuropeptide family that includes vasoactive intestinal peptide (VIP), glucagon, secretin and growth hormone releasing hormone (Miyata et al. 1989). Two forms have been found in many species, PACAP27 and PACAP38 consisting of 27 and 38 amino acids, respectively, with widespread distribution throughout the central nervous system and various peripheral tissues (Arimura 1992; Zhong and Pena 1995). Physiological, pharmacological, and developmental effects are usually mediated by G protein-coupled PACAP receptors, which in turn regulate intracellular adenosine 3′,5′-cyclic monophosphate (cAMP) (Arimura 1992; Leech et al. 1995) and inositol phosphate levels (Rawlings et al. 1994). So far, however, there have been relatively few studies concerned with the electrophysiological aspects of PACAP action. One such study by Zhong and Pena (1995) in Drosophila larval muscle cells found that local application of PACAP38 at the neuromuscular junction triggered two temporally distinct responses—an immediate depolarization caused by a slow inward current, followed some minutes thereafter by marked enhancement of a voltage-activated potassium current. High-frequency stimulation of motor nerve fibers evoked a postsynaptic response that mimicked the effect of exogenous PACAP38, and desensitization of this response was observed after preincubation of the preparation with PACAP38. In another study (Leech et al. 1995), it was found that both forms of PACAP induce depolarization and elevate intracellular free calcium in insulinoma (HIT-T15) cells. It was concluded that depolarization in these cells was due to an increased inward sodium current combined with a reduction in overall membrane conductance through closure of potassium channels.
With this background, the aim of the present study was to characterize further the electrophysiological effects ofPACAP38, focusing particularly on neuronal glutamatergic transmission involving NMDA receptors given the possibility of interacting effects through the common second messenger calcium. A preliminary report has been presented (Liu and Madsen 1996a).
Cortices from 10-day-old chick embryos were dissected out under sterile conditions and placed in Puck's solution, which contained (in g/L) 80 NaCl, 4.0 KCl, 0.6 KH2PO4, 0.9 Na2HPO4⋅7H2O, and 10 glucose, pH 7.4; meningeal material then was removed with fine forceps. The cleaned tissue was transferred to fresh Puck's solution containing 0.25% trypsin II, teased apart manually and then incubated in a water bath at 37°C for 20 min. Dissociated cells were collected by centrifugation at 100 g for 5 min at 4°C. The pellet was resuspended in Eagle's minimum essential medium (EMEM; pH 7.35) supplemented with 5% fetal calf serum, 5% horse serum, 1.5% embryo extract (prepared as described in Le Dain et al. 1991), 2 mM glutamine, 0.5 mM sodium pyruvate, 5 μg/ml insulin, 10 mM glucose, 75 U/ml penicillin, and 45 μg/ml streptomycin. The cell suspension then was aspirated gently several times through a flamed Pasteur pipette to thoroughly disperse cells, and these then were filtered through 50 μm nylon mesh. Cells were diluted with supplemented EMEM to 106 cells/ml and 5 ml placed into a 55 mm Petri-dish containing a coverslip pre-coated with 25 μg/ml poly-l-lysine for 24 h. Coverslips were washed twice with double-distilled water to remove excess, unbound poly-l-lysine. Cells were incubated for 7–30 days at 37°C in a humidified atmosphere containing 3.5% CO2. One-half the culture medium was replaced every 3 days with supplemented EMEM as described above, except that 5% fetal calf serum and 5% horse serum were replaced with 10% horse serum, and 2 μM cytarabine was added to inhibit nonneuronal cell growth in vitro.
Methods for single-channel recording and analysis were as described (Le Dain et al. 1991; Liu and Madsen 1996b) with the following modifications: cells were removed from the culture medium, washed, and placed in recording bath solution, which contained (in mM) 140 NaCl, 3 KCl, 1.0 CaCl2, 10 N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES), and 10 glucose and 0.1 μM tetrodotoxin, pH 7.4. Pipettes were filled with an intracellular solution containing (in mM) 140 KCl, 2 MgCl2, 10 HEPES, and 10 ethylene glycol-bis(β-aminoethyl ether)-N,N,N′,N′-tetraacetic acid, pH 7.3, and drugs then were applied by microperfusion in various combinations and sequences to membrane patches containing NMDA receptors. Single-channel currents were recorded at room temperature (22–24°C), and pipette potential was held at −55 to −60 mV. Cell-free patches were moved ≥100 μm from cultured cells in the recording bath before data were collected. The number of openings used to calculate channel opening frequency was usually >1,000, and the critical closed time for analysis of burst behavior was set at 20 ms through use of an equal proportion misclassification approach. Errors given in the text are SE, with n = 3 unless otherwise indicated. Trypsin, poly-l-lysine, cytarabine, NMDA, glycine, MK-801, 7-chlorokynurenic acid, and ifenprodil were supplied by Sigma; PACAP38 and PACAP6–38 by Auspep (Australia), with all other chemicals AR grade.
Figure 1 shows typical recordings from outside-out membrane patches. These inward currents were blocked by extracellular Mg2+ and 10 μM MK-801, had a mean conductance of 58 ± 3.6 pS (n = 7), and could be induced repeatedly or eliminated within individual patches in direct response to application or removal of 20 μM NMDA. Figure 1 A shows that channel activity due to NMDA alone (top) was increased in the same patch by NMDA plus 2 nM PACAP38 (middle) and that this stimulatory effect by PACAP38 was reversible when receptors were again exposed to only 20 μM NMDA (bottom). The mean increase in channel opening frequency compared with control in three such experiments was 3.2 ± 0.58. An even greater effect of PACAP38 on channel activation was seen when this experiment was repeated in the continued presence of 1 μM glycine (Fig. 1 B), with the relative frequency of openings in seven such experiments being 15 ± 9.8. Figure 2 summarizes results for several experiments over a range of PACAP38 and glycine concentrations. The positive modulatory effect of 2 nMPACAP38 in the presence of 1 μM glycine (Fig. 2 A) became inhibitory as PACAP38 concentration was increased beyond 10 nM, such that at 1,000 nM PACAP38, channel activity was reduced to <5% of control. When glycine concentration was increased to 2 μM, the stimulatory and inhibitory effects of PACAP38 were still apparent but reduced in magnitude. For PACAP38 concentrations ≤10 nM, these modulatory effects were rapidly and fully reversible, whereas higher concentrations required >5–10 min of washout to achieve partial reversibility. PACAP38 had no consistent effect on single-channel current amplitude, and by itself was devoid of agonist activity.
Figure 2 B presents complementary information on the interacting effects of glycine and PACAP38 in the continued presence of 20 μM NMDA. PACAP38 (2 nM) stimulated opening frequency induced by 20 μM NMDA in the absence of glycine, and then gradual increase in glycine concentration from 1 to 5 μM resulted in further enhancement followed by a reduction back to control level. Hence in the presence of 5 μM glycine, the additional presence of 2 nM PACAP38 was of little consequence in terms of receptor activation. With an inhibitory concentration of PACAP38 (100 nM) in the absence of glycine, there was insufficient activity by NMDA alone to reliably measure channel opening frequency; however, this strong inhibition was reversed gradually with increasing concentrations of glycine. In addition to these interacting effects of PACAP38 and glycine, it was found that receptor activation by 20 μM NMDA, 1 μM glycine, and 2 nM PACAP38 was not affected when the NMDA concentration was increased from 20 to 200 μM; MK-801 (10 μM) reduced channel activity due to 20 μM NMDA, 1 μM glycine, and 2 nM PACAP38 to <3% (0.026 ± 0.004) of control; the glycine site antagonist 7-chlorokynurenic acid (10 μM) reduced receptor activity induced by 20 μM NMDA and 1 μM glycine to 5 ± 3% of control, whereas this inhibition was slightly less (0.14 ± 0.06) in the presence of 20 μM NMDA and 2 nM PACAP38; ifenprodil (10 μM) had no effect on activation by 20 μM NMDA plus 2 nM PACAP38 (99 ± 28% of control); and PACAP6–38 (1 μM, a PACAP antagonist) reduced activity due to 20 μM NMDA and 1 μM glycine to 34 ± 1% of control, but when 2 nM PACAP38 also was present, there was far less inhibition (87 ± 6% of control).
Table 1 shows results of kinetic analyses of single channel data from a series of paired outside-out experiments where individual patches were first exposed to control (20 μM NMDA + 1 μM glycine) and then test (20 μM NMDA + 1 μM glycine + PACAP38) solutions. Channel openings consisted of brief events with a mean duration of about a millisecond and longer duration openings with a mean of 4–6 ms. Bursting activity was apparent, with an average of two to three openings per burst and openings usually separated by fast closures (time constants 0.6–2.0 ms) and occasionally by intermediate duration closures. Although, in principle, the increased channel opening frequency illustrated in Figs. 1 and 2 in the presence of 2 nM PACAP38 could be due to an increase in the occurrence of bursts, an increase in the number of openings per burst or a combination of both of these effects, these results demonstrate that the stimulatory effect was due primarily to an increase in the frequency of occurrence of bursts. At 2 nM PACAP38, the slow component of the burst duration (the component containing multiple openings) increased only marginally from 15 ± 4.6 to 20 ± 0.9 ms, whereas the intermediate and slow components of the closed distribution were reduced significantly (72 ± 38 to 3.4 ± 1.4 ms for the intermediate component and 823 ± 572 to 55 ± 9.1 ms for the slow component).
These results using cell-free membrane patches indicate that PACAP38 can act on NMDA receptors independently of intracellular mediators such as cAMP, calcium, and inositol phosphates. Although this was an unexpected finding given that PACAP receptors belong to the G protein-coupled family of receptors (Arimura 1992), there have been earlier reports of cAMP-independent effects of VIP on control of inwardly rectifying K+ channels in a bovine endothelial cell line (Pasyk et al. 1992) and neuronal nicotinic receptors in rat intracardiac neurons (Cuevas and Adams 1996). Stimulatory glutamatergic effects of PACAP and other members of the peptide family also have been reported (Martin et al. 1995; Stella and Magistretti 1996; Wu and Dun 1996).
Single channel data analysis indicated that the stimulatory effect of PACAP38 on channel opening frequency was due mainly to an increase in the probability of channel opening. Similar conclusions were made regarding the potentiation of nicotinic acetylcholine receptor channel activity by VIP (Cuevas and Adams 1996). Furthermore, in that study, VIP had no intrinsic agonist activity, gating behavior was only affected in the presence of acetylcholine, there was no effect on single channel conductance, and potentiation still was seen with cell dialysis or in cell-free membrane patches. However, although it was suggested that G proteins (but not cAMP) were involved in the VIP-nicotinic receptor interaction, this seems unlikely for the PACAP38-NMDA interaction given the absence, in the present experiments, of nucleotides to activate G proteins in cell-free recording mode. The PACAP6–38 results, with inhibition in the absence of a PACAP agonist and only marginal inhibition (13%) when excess antagonist (1 μM) was present with agonist (2 nM), also suggest a more direct interaction.
The data in Figure 2 showing the interdependence of PACAP38 and glycine concentrations on NMDA receptor modulation, together with blockade by 7-chlorokynurenic acid, indicate the importance of the glycine coagonist site. Direct modulation of NMDA receptors through this site by a number of peptides has been suggested previously (Chen et al. 1995; Rusin and Randic 1991). Brauneis et al. (1996) found that the opioid dynorphin had both stimulatory and inhibitory effects on NMDA receptor activity that were independent of opioid receptors but dependent on glycine concentration, and that modified forms of dynorphin where the glycines normally present at N-terminal amino acid positions 2 and 3 of the peptide were removed lost their potentiating effects. Interestingly, PACAP6–38 does not contain the glycine normally present at position 4 in PACAP38, and it has (weak) antagonist action without any stimulatory effect in comparison to the dual action of the full peptide.
G. J. Liu was supported financially as a University of Western Australia Postdoctoral Research Fellow.
Address for reprint requests: B. W. Madsen, Dept. of Pharmacology, University of Western Australia, Nedlands WA 6907, Australia.