Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2004 Apr 30;557(Pt 3):949–960. doi: 10.1113/jphysiol.2004.063818

Autocrine feedback inhibition of plateau potentials terminates phasic bursts in magnocellular neurosecretory cells of the rat supraoptic nucleus

Colin H Brown 1, Charles W Bourque 1
PMCID: PMC1665154  PMID: 15107473

Abstract

Phasic activity in magnocellular neurosecretory cells is characterized by alternating periods of activity (bursts) and silence. During phasic bursts, action potentials are superimposed on plateau potentials that are generated by summation of depolarizing after-potentials. Dynorphin is copackaged in vasopressin neurosecretory vesicles that are exocytosed from magnocellular neurosecretory cell dendrites and terminals, and both peptides have been implicated in the generation of phasic activity. Here we show that somato-dendritic dynorphin release terminates phasic bursts by autocrine inhibition of plateau potentials in magnocellular neurosecretory cells recorded intracellularly from hypothalamic explants using sharp electrodes. Conditioning spike trains caused an activity-dependent reduction of depolarizing after-potential amplitude that was partially reversed by α-latrotoxin (which depletes neurosecretory vesicles) and by nor-binaltorphimine (κ-opioid receptor antagonist), but not by an oxytocin/vasopressin receptor antagonist or a μ-opioid receptor antagonist, indicating that activity-dependent inhibition of depolarizing after-potentials requires exocytosis of an endogenous κ-opioid peptide. κ-Opioid inhibition of depolarizing after-potentials was not mediated by actions on evoked after-hyperpolarizations since these were not affected by κ-opioid receptor agonists or antagonists. Evoked bursts were prolonged by antagonism of κ-opioid receptors with nor-binaltorphimine and by depletion of neurosecretory vesicles by α-latrotoxin, becoming everlasting in ∼50% of cells. Finally, spontaneously active neurones exposed to nor-binaltorphimine switched from phasic to continuous firing as plateau potentials became non-inactivating. Thus, dynorphin coreleased with vasopressin generates phasic activity through activity-dependent feedback inhibition of plateau potentials in magnocellular neurosecretory cells.

Phasic activity in magnocellular neurosecretory cells (MNCs) of the supraoptic nucleus (SON) and paraventricular nucleus is characterized by action potential (spike) discharge in periods of activity (bursts) separated by periods of silence that each last tens of seconds and is highly efficient for secretion of vasopressin (the anti-diuretic hormone) into the circulation(Leng et al. 1999). During phasic bursts, spikes are superimposed on plateau potentials that are essential for the expression of phasic firing (Andrew & Dudek, 1983; Ghamari-Langroudi & Bourque, 1998). In vitro, MNC spikes are followed by non-synaptic depolarizing after-potentials (DAPs) that summate temporally to generate the plateau potential (Andrew & Dudek, 1983; Armstrong et al. 1994; Li et al. 1995; Li & Hatton, 1997; Ghamari-Langroudi & Bourque, 1998) and so sustain spike firing throughout the burst. However, the mechanisms by which individual bursts terminate are unknown.

The κ-opioid peptide dynorphin is copackaged with vasopressin in neurosecretory vesicles (Whitnall et al. 1983) that are exocytosed from the dendrites and terminals of MNCs (Pow & Morris, 1989). Both vasopressin (Richard et al. 1997) and dynorphin (Brown et al. 2000) have been implicated in the generation of phasic activity in vivo since burst duration is increased in MNCs during intra-SON administration of a V1 receptor antagonist (Ludwig & Leng, 1997) or a κ-opioid receptor antagonist (Brown et al. 1998), indicating that endogenous peptides restrain activity under basal conditions. However, vasopressin release restrains spike firing throughout bursts, while endogenous κ-opioid inhibition emerges as bursts progress (Brown et al. 2004). Moreover, vasopressin is generally inhibitory (Brown et al. 2004), whereas κ-opioid inhibition specifically affects postspike excitability (Brown & Leng, 2000). Thus, the organization of the feedback inhibition of activity by dynorphin indicates that this peptide specifically modulates phasic firing.

We have previously shown that activation of SON κ-opioid receptors with the arylacetamide agonist U50,488H inhibits DAP amplitude (Brown et al. 1999). DAPs and plateau potentials are attenuated when evoked immediately following the end of a burst (Andrew & Dudek, 1983), suggesting that DAPs are subject to activity-dependent inhibition. Here, we show that burst termination in MNCs results from an activity-dependent inhibition of DAPs and plateau potentials mediated by dendritic release of the endogenous opioid copeptide, dynorphin.

Methods

Explant preparation

Conscious male Long-Evans rats (ca 150–250 g) were restrained in a soft plastic cone (5–10 s), decapitated and the brains rapidly removed according to a procedure approved by the McGill University Animal Care Committee. A block of tissue 8 mm × 8 mm × 2 mm containing the basal hypothalamus was excised using razor blades and pinned, ventral surface up, to the Sylgard® base of a superfusion chamber. Within 2–3 min, the excised hypothalamic explant was superfused (at 0.5–1.0 ml min−1 at 32–33°C) with carbogenated (95% oxygen; 5% carbon dioxide) artificial cerebrospinal fluid (aCSF; see below) delivered via a Tygon® tube placed over the medial tuberal region. The arachnoid membrane overlying the supraoptic nucleus was removed using fine forceps and a cotton wick was placed at the rostral tip of the explant to facilitate aCSF drainage.

Electrophysiological recording

Intracellular recordings were made using sharp micropipettes prepared from glass capillaries (1.2 mm o.d.; AM Systems, Everett, WA, USA) pulled on a P-87 Flaming-Brown puller (Sutter Instruments, Novato, CA, USA). Micropipettes were filled with 2 m potassium acetate to yield DC resistances of 70–160 mΩ to a Ag–AgCl wire electrode immersed in aCSF. Voltage recordings (DC–3 kHz) were obtained using an Axoclamp 2A amplifier (Axon Instruments, Union City, CA, USA) in continuous current clamp (‘bridge’) mode. Acquired signals were displayed on a chart recorder and digitized (44 kHz; Neurodata, Delaware Water Gap, PA, USA) for storage on videotape. For analyses, signals were digitized (5 or 10 kHz; Digidata 1200 interface, Axon Instruments) and stored on a personal computer running Clampex (Axon Instruments) and analysed offline using Clampfit (Axon Instruments). Current pulses were delivered through a Digitimer DS2 isolated stimulator (Welwyn Garden City, UK) connected to an external pulse generator (Digitimer D-4030). Recordings were made from supraoptic nucleus MNCs impaled with sharp electrodes in superfused hypothalamic explants. These cells had resting membrane potentials more negative than –50 mV, input resistances greater than 150 mΩ and action potential amplitudes exceeding 60 mV when measured from baseline. Each cell displayed frequency-dependent spike broadening and transient outward rectification when depolarized from initial membrane potentials more negative than –75 mV; these combined characteristics are specific to MNCs (Renaud & Bourque, 1991). Depolarizing current injection (100–400 pA for 80 ms) was applied to elicit a train of three to eight action potentials (the number of spikes in the trains was kept constant for each cell) that evoked a clear DAP/plateau potential in each cell under control conditions. Subthreshold DAPs were recorded from cells maintained at a membrane potential approximately 5–10 mV below the action potential threshold (maintained by injection of hyperpolarizing current if necessary) and suprathreshold plateau potentials from cells held approximately 2–4 mV below threshold.

Although we did not positively identify the cells recorded as MNCs using immunochemistry, it is likely that most, if not all, of the cells tested contained vasopressin. Indeed, all of the cells studied here showed DAPs and readily fired afterdischarges during suprathreshold evoked plateaus, a feature found in all vasopressin MNCs (Armstrong et al. 1994). Moreover, almost 70% of the MNCs in the supraoptic nucleus of Long-Evans rats (the strain used here) contain vasopressin (Rhodes et al. 1981). Finally, all of our recordings were obtained from the ventro-caudal part of the nucleus, an area which contains packed vasopressin MNCs (whereas oxytocin MNCs are located more sparsely in the antero-dorsal part of the nucleus).

Data analysis

In the experiments in which a single DAP was evoked every 12.25 s, measurements were made from 10 consecutive DAPs (where necessary DC injection kept the initial baseline membrane potential constant and to within 0.5 mV of the membrane potential from which control measurements were obtained). The 10 consecutive files were averaged using Clampfit and the maximum amplitude of the averaged DAP was measured. In the experiments where ‘paired’ DAPs were evoked 9 s apart every 60 s, DAPs were measured from individual cycles and the second (test) DAP amplitude was expressed as a percentage of the first (control) and then the mean was calculated. Conditioning spike barrages were generated by trains of short depolarizing current injections (100–500 pA for 5 ms each) at 10, 25 and 50 Hz over 1 s or 50 Hz over 2 s (since MNCs do not reliably follow 100 Hz stimulation).

All data are expressed as means ± s.e.m. Statistical analyses were performed using SigmaStat (SPSS, Chicago, IL, USA). Data were analysed using Student's paired or unpaired t tests, or one-way repeated measures (RM) analysis of variance (ANOVA) where appropriate. Where the F-ratio was significant, post hoc analyses were carried out using Student–Newman–Keuls (SNK) tests.

Drugs

The aCSF (pH 7.4; 295 ± 3 mosmol kg−1) was composed of (mm): 120 NaCl, 3 KCl, 1.2 MgCl2, 26 NaHCO2, 2.5 CaCl2, and 10 glucose (Fisher Scientific, Pittsburgh, PA, USA). Stock solutions of 300 nmα-latrotoxin (LTX; Alomone Laboratories, Jerusalem, Israel), 100 μmnor-binaltorphimine (BNI; Tocris Cookson, Ballwin, MO, USA), 1 mm amastatin, d-Phe-Cys-Tyr-d-Trp-Orn-Thr-Pen-Thr amide (CTOP), dynorphin (Sigma-Aldrich, Oakville, ON, USA), captopril, phosphoramidon (ICN Biomedicals, Costa Mesa, CA, USA) and 10 mm Manning compound (MC; [d(CH2)51,O-Me-Tyr2,Arg8]-vasopressin; Bachem Bioscience, King of Prussia, PA, USA) were prepared in deionized water and stored frozen until the day of use. All drugs were bath applied after dilution to the appropriate concentration in aCSF. Bicuculline methochloride (5 μm) and kynurenic acid (0.5 mm; Tocris Cookson) were prepared fresh each day.

Results

Activity-dependent inhibition of DAPs

To determine whether DAPs are subject to activity-dependent inhibition, intracellular recordings were obtained from MNCs in rat hypothalamic explants. Pairs of DAPs were evoked (9 s apart, once per minute) by triggering brief trains of action potentials with depolarizing current pulses. In order to prevent these DAPs reaching threshold and thus generating a plateau potential and regenerative firing, these DAPs were evoked from cells maintained at a membrane potential 5–10 mV below threshold. Under these conditions, no differences were noted between the first (control) and second (test) DAPs (3.2 ± 0.2 and 3.1 ± 0.2 mV, respectively; n = 32; Fig. 1A) and amplitudes were consistent from trial to trial for periods >60 min. However, when barrages of 10–100 conditioning action potentials (a number of spikes corresponding to that which occurs in the first seconds, or tens of seconds, in spontaneous bursts) were interposed between the control and test DAPs, the amplitude of the test DAP was reduced compared to the control, and the degree of inhibition increased exponentially with the number of spikes in the conditioning barrage (r2= 0.93, Fig. 1A and B).

Figure 1. Activity-dependent inhibition of DAPs.

Figure 1

Open in a new tab

A, DAPs (averages of 5) that each follow a 5-spike train (arrowheads) evoked 9 s apart by a 80 ms depolarizing pulse (+150 pA). A conditioning train (arrow) of 0–50 spikes over 1 s each, or 100 spikes over 2 s, was evoked by a corresponding number of 5 ms, +500 pA DC pulses. B, DAP amplitude 4 s after conditioning trains containing 0–100 spikes. ***P < 0.001 versus pre-train, SNK tests. The number of MNCs in each group is shown in parentheses. C, DAP amplitude at various times after a 25-spike conditioning train. A single exponential function with a time constant of 4.7 s was fitted to the data (r2= 0.84, data from 9 cells).

In the absence of firing, the inhibitory effect of a 25-spike conditioning barrage declined exponentially with a time constant (τ) of 4.7 s (Fig. 1C). Moreover, the rate of recovery from activity-dependent inhibition was independent of the number of spikes in the conditioning barrage (τ= 4.3 ± 1.1 s, r2= 0.98 and τ= 4.6 ± 0.1 s, r2= 0.99, after 10- and 50-spike barrages, respectively). Thus, DAPs that sustain plateau potentials and burst firing in MNCs undergo reversible activity-dependent inhibition.

Endogenous peptide release mediates activity-dependent inhibition of DAPs

To determine whether activity-dependent DAP inhibition requires exocytosis, we examined the effects of LTX, a 130 kDa protein component of black widow spider venom that provokes massive exocytosis to deplete stores of neurosecretory vesicles (Ushkaryov, 2002) and stimulates peptide secretion from MNC terminals (Hlubek et al. 2003). In the presence of 0.5 mm kynurenic acid and 5 μm bicuculline (to prevent potential effects of LTX mediated by activation of ionotropic glutamate and GABA receptors by glutamate and GABA released by LTX), single subthreshold DAPs were evoked every 12.25 s, a time interval insufficient to allow full recovery from activity-dependent inhibition (Fig. 1C) and which therefore induced a cumulative inhibition of consecutive DAPs and thereafter maintained a significant level of steady-state inhibition. Upon application of 3 nm LTX, the amplitude of the DAPs declined gradually from a starting value of 2.9 ± 0.3 mV (Figs 2A and B), reaching a minimal amplitude of 2.2 ± 0.3 mV (P = 0.007, SNK test, n = 6) 10–22 min following the onset of the application. This effect is consistent with the expected and progressive exocytotic release of a substance causing inhibition. However, despite the continued presence of LTX, the amplitude of the DAPs eventually recovered (Fig. 2B) and ultimately overshot the control value to 3.4 ± 0.3 mV (P = 0.03, SNK test; Fig. 2B and C), an amplitude equivalent to that of DAPs evoked 1 min apart in these cells (3.5 ± 0.4 mV; P = 0.88, SNK test). The latter observation suggested that the pool of vesicles releasing inhibitory substance was now depleted, and that steady-state inhibition was removed. Remarkably, under these conditions, the ability of a conditioning barrage to inhibit DAP amplitude was strongly attenuated compared to pre-LTX (Fig. 2D and E), suggesting that activity-dependent inhibition of DAPs requires exocytotic release of a substance contained in the neurosecretory vesicles of MNCs.

Figure 2. Dendritic peptide release inhibits DAPs.

Figure 2

Open in a new tab

A, DAPs (averages of 10) that each follow a 3-spike train evoked in a 12.25 s cycle before and during 3 nm LTX superfusion. B, time course of the LTX effects in A. C, mean maximum inhibition (LTX-1) and overshoot (LTX-2) by LTX. *P < 0.05 and **P < 0.01 versus pre-LTX, SNK tests. D, paired DAPs recorded from different cells in the absence (Control) and presence of LTX before (black) and after (grey) a 25-spike train. E, mean DAP inhibition by a 25-spike train before and during LTX. *P < 0.05 and **P < 0.01 versus pre-LTX, Student's t test.

κ-Opioid peptide inhibition of DAPs

To identify the substance responsible for the activity-dependent inhibition of DAPs, we first examined the effects of antagonists of vasopressin and opioid receptors on the degree of inhibition produced by a barrage of conditioning spikes. Blockade of vasopressin receptors with the oxytocin/vasopressin antagonist, MC (Kruszynski et al. 1980) (10 μm, a dose that blocks activity-dependent retrograde inhibition of synaptic inputs in MNCs; Kombian et al. 1997) did not affect the inhibition of DAPs produced by a conditioning barrage of 25 spikes (P = 0.56, paired t test, n = 6; Fig. 3A and B), indicating that vasopressin release is not responsible for this effect. Since dynorphin activates both μ- and κ-opioid receptors, we next discriminated between these possibilities using selective antagonists. Although superfusion of the specific μ-opioid antagonist CTOP (Hawkins et al. 1989) (1 μm) was without effect (P = 0.24, paired t test, n = 3; Fig. 3A and B), application of the specific κ-receptor antagonist BNI (Portoghese et al. 1987) (1 μm, a dose that blocks DAP inhibition by the κ-receptor agonist, U50,488H; Brown et al. 1999) significantly reduced the inhibitory action of the conditioning barrage on DAP amplitude (P = 0.001, paired t test, n = 8; Fig. 3A and B). Thus, inhibition of DAPs induced by a conditioning barrage requires activation of κ-opioid receptors.

Figure 3. Endogenous κ-opioids inhibit DAPs.

Figure 3

Open in a new tab

A, paired DAPs before (black) and after (grey) a 25-spike train in the presence and absence of receptor antagonists. B, mean DAP inhibition by a 25-spike train before (Pre-drug) and during superfusion (Drug) of the vasopressin/oxytocin antagonist MC (10 μm), the μ-opioid receptor antagonist CTOP (1 μm), or the κ-opioid receptor antagonist BNI (1 μm). ***P < 0.001 versus pre-drug, paired t test. C, DAPs that each follow an 8-spike train evoked in a 12.25 s cycle before and during dynorphin (1 μm) and BNI (1 μm) superfusion. D, time course of the opioid effects in C. E, mean DAP amplitude. **P < 0.01 versus pre-drug and ††P < 0.01 and †††P < 0.001 versus DYN, SNK tests.

We next examined if κ-opioid receptors are also involved in modulating the steady-state amplitude of DAPs evoked every 12.25 s. As illustrated in Fig. 3C–E, exogenous dynorphin (1 μm) reversibly inhibited DAP amplitude (similarly to U50,488H; Brown et al. 1999), whereas application of the κ-opioid receptor antagonist BNI reversibly increased DAP amplitude. These observations indicated, respectively, that activation of κ-opioid receptors can inhibit DAPs and that steady-state inhibition of DAPs in this protocol is mediated via κ-opioid receptors activated by an endogenously released agonist. Application of exogenous dynorphin was without effect when tested in the presence of BNI (P = 0.92, SNK test; Fig. 3C–E), confirming that the effects of dynorphin on DAPs are specifically mediated by κ-opioid receptors. In agreement with this observation, neither 1 μm CTOP nor 10 μm MC altered steady-state inhibition of DAPs evoked every 12.25 s (3.6 ± 0.5 and 3.8 ± 0.6 mV before and during CTOP, respectively; P = 0.30, paired t test, n = 7 and 3.0 ± 0.5 and 3.2 ± 0.5 mV before and during MC, respectively; P = 0.62, paired t test, n = 8). Interestingly, in the continued presence of MC a cocktail of peptidase inhibitors that prevents dynorphin breakdown (Hiranuma et al. 1998) (1 μm each of amastatin (aminopeptidase inhibitor), captopril (dipeptidyl carboxypeptidase I inhibitor) and phosphoramidon (endopeptidase-24.11 inhibitor)) reduced the amplitude of DAPs evoked every 12.25 s to 2.5 ± 0.6 mV (P = 0.04 compared to MC, SNK test; n = 8). Thus, peptidases present in the extracellular space normally restrict DAP inhibition mediated by endogenously released dynorphin.

To determine whether inhibition of the DAP by activity-dependent release from neurosecretory vesicles is mediated by activation of κ-opioid receptors, we examined whether the effects of LTX could be prevented by coadministration of the κ-opioid antagonist BNI. In the presence of 1 μm BNI (as well as 0.5 mm kynurenic acid and 5 μm bicuculline), subthreshold DAPs evoked every 12.25 s had an amplitude of 2.1 ± 1.0 mV (n = 4). In the presence of BNI, DAP amplitude was not altered by superfusion of 3 nm LTX for 15 min (2.1 ± 1.1 mV) or 25 min (2.1 ± 1.0 mV) (Fig. 4), the expected times of maximal LTX-induced inhibition and overshoot of DAP amplitude (Fig. 2). These observations indicate that activity-dependent inhibition of DAPs is specifically mediated by the release of dynorphin from MNCs.

Figure 4. Inhibition of DAPs by dendritic peptide release is blocked by κ-opioid receptor antagonism.

Figure 4

Open in a new tab

A, DAPs (averages of 10) that each follow a 3-spike train evoked in a 12.25 s cycle before and during 3 nm LTX superfusion in the presence of 1 μm BNI. B, time course of the LTX effects in A. C, mean DAP amplitude after LTX superfusion for 15 min (LTX-1) and 25 min (LTX-2) in the presence of BNI. P = 0.64, one-way RM ANOVA.

After-hyperpolarizations are not altered by dendritic dynorphin release

MNCs display a prominent after-hyperpolarization (AHP) following short trains of spikes such as those used to evoke DAPs in this study (Armstrong et al. 1994; Ghamari-Langroudi & Bourque, 1998). This AHP peaks early and its decay gives way to the more protracted DAP. Nonetheless, since the AHP and DAP partly overlap temporally, it is possible that the observed activity-dependent inhibition of DAP amplitude could have resulted indirectly from the masking effect of an enhancement of AHP amplitude. However, by contrast to its effects on DAPs, a 25-spike conditioning barrage interposed between a pair of AHPs (evoked 9 s apart by short spikes trains, once per minute) did not induce any activity-dependent modulation of AHP amplitude (Fig. 5B), and AHP amplitude was not altered by 1 μm BNI (Fig. 5C).

Figure 5. AHP amplitude is not modulated by activity-dependent dendritic dynorphin release.

Figure 5

Open in a new tab

A, paired AHPs that each follow a 5-spike train (truncated, arrowheads) evoked 9 s apart by a 80 ms depolarizing pulse (+350 pA) before (black) and after (grey) a 25-spike train in the absence and presence of the κ-opioid receptor antagonist, BNI (1 μm). B, mean amplitude of AHPs evoked 4 s before and 4 s after a conditioning train containing 25 spikes prior to BNI superfusion. C, mean amplitude of AHPs evoked 4 s after a 25-spike conditioning train before and during BNI superfusion. Note that neither the conditioning train nor BNI altered the amplitude of AHPs.

Furthermore, 1 μm BNI did not affect the amplitude of AHPs evoked by short trains of spikes applied every 12.25 s (–5.9 ± 0.7 mV and –5.2 ± 0.7 mV, before and during BNI, P = 0.08, paired t test, n = 10), indicating that the steady state inhibition of DAPs by endogenous κ-opioid peptides was not due to AHP modulation. The failure to observe any effect of BNI on AHP amplitude was not due to ‘masking’ of changes in AHP amplitude by modulation of DAPs since 1 μm BNI also failed to affect AHP amplitude during blockade of DAPs with 3 mm CsCl (Ghamari-Langroudi & Bourque, 1998) (–5.6 ± 0.8 mV and –5.5 ± 0.6 mV, before and during BNI, paired t test, P = 0.71, n = 7). Finally, application of 1 μm dynorphin also failed to alter AHP amplitude (–5.4 ± 1.2 mV and –5.2 ± 1.1 mV, before and during dynorphin, P = 0.65, paired t test, n = 8).

Previous studies have shown that a slower component of the AHP can emerge in response to more prolonged spike trains (e.g. after the conditioning barrages of 25 or more spikes in Fig. 1) in MNCs (Greffrath et al. 1998). We therefore examined whether BNI could affect the slow AHP. In 10 cells, DAPs were blocked by superfusion of 3 mm CsCl (Ghamari-Langroudi & Bourque, 1998) and a 50 Hz spike train was elicited (for 1 s) every 60 s to evoke a slow AHP. The amplitude of the slow AHP (averaged between 4 and 5 s after the 50-spike train; the period when DAP amplitudes were measured in Figs 1–4) was 1.1 ± 0.3 mV and 1.4 ± 0.4 mV before and after superfusion of 1 μm BNI for 50–60 min, respectively (P = 0.77).

Thus, since neither the fast nor the slow component of the AHP was inhibited by κ-opioid receptor antagonism, it is unlikely that the enhancement of DAP amplitude observed under these conditions is due to an ‘unmasking’ of the DAPs through a reduction of AHP amplitude.

κ-Opioid receptor antagonism prolongs plateau potentials and regenerative firing

To determine whether activity-dependent inhibition of DAPs can functionally curtail plateau potentials and regenerative firing, we examined the effects of various drugs on the intensity of the afterdischarge sustained by suprathreshold plateau potentials that ensue when evoked DAPs cross threshold and induce regenerative firing. Blockade of κ-opioid receptors with 1 μm BNI progressively prolonged the afterdischarge (Fig. 6A and C) and increased the number of spikes in the first 10 s of afterdischarge from 3 ± 1 to 27 ± 6 spikes (P = 0.02, paired t test; n = 5) after 14–24 min. In three MNCs, afterdischarges became irreversible after 21–29 min of BNI (Fig. 6A) and the plateau potential and associated firing could only be terminated by hyperpolarizing current injection. Similarly, following depletion of dendritic neurosecretory vesicles with LTX (3 nm, >30 min) the afterdischarges were greatly prolonged (83 ± 24 spikes in the first 10 s of afterdischarge compared to 11 ± 3 spikes under control conditions; P = 0.02, Student's t test; n = 4) and, in two MNCs, became irreversible (Fig. 6B). In contrast to the effects of BNI and LTX, neither CTOP nor MC altered the number of spikes in the afterdischarge (3 ± 1 and 4 ± 3 spikes before and during 1 μm CTOP (P = 0.35, paired t test, n = 4) and 7 ± 2 and 8 ± 3 spikes before and during 10 μm MC (P = 0.73, paired t test, n = 7), respectively). Thus, under physiological conditions the duration of the afterdischarge is specifically limited by an activity-dependent release of dynorphin and the subsequent activation of κ-opioid receptors.

Figure 6. Inhibition of autocrine κ-opioid actions prolongs evoked regenerative firing.

Figure 6

Open in a new tab

A and B, suprathreshold plateau potentials (evoked by a brief spike train: arrowheads) with superimposed afterdischarges recorded before and during 1 μm BNI (A) or 3 nm LTX (B) superfusion. C, the number of spikes in the first 10 s of afterdischarge in suprathreshold plateau potentials evoked >12 s apart for each trial (Pearson product moment correlation coefficient = 0.519; P < 0.001; n = 5 MNCs).

Finally, we examined whether endogenously released dynorphin modulates spontaneous phasic firing in situ. As illustrated in Fig. 7, bath application of 1 μm BNI increased the duration of spontaneous phasic bursts recorded from MNCs impaled in vitro (from 9.5 ± 3.2 to 41.1 ± 12.2 s; P = 0.03, paired t test, n = 6), an effect associated with a delay in repolarization of the plateau potential (Fig. 7A). Indeed two of the six cells eventually adopted continuous activity in BNI that could only be terminated by hyperpolarizing current injection. While these cells could subsequently re-initiate activity, they could no longer spontaneously terminate the activity and thus could not resume phasic firing (Fig. 7B and C).

Figure 7. Inhibition of autocrine κ-opioid actions prolongs spontaneous phasic bursts.

Figure 7

Open in a new tab

A, spontaneous bursts of activity (spikes truncated) before (black) and during 1 μm BNI superfusion, aligned at onset. BNI delayed plateau potential inhibition after 5 min (dark grey) and induced continuous activity after 20 min (pale grey). B, MNC spontaneous firing rate before and during BNI superfusion. BNI progressively increased the active period duration, eventually leading to continuous activity. The final period of activity shown was terminated by hyperpolarizing current injection (see C). C, membrane potential of a spontaneously phasic MNC (top trace) before (left) and during (right) superfusion of 1 μm BNI. Before application of BNI, phasic activity was initiated by reducing the hyperpolarizing current injection to –66 pA (bottom trace). All three Pre-BNI bursts terminated spontaneously without further manipulation of the hyperpolarizing current injection. In the presence of BNI, activity started spontaneously when the hyperpolarizing current injection was at –56 pA. By contrast to activity before BNI application, in the presence of BNI activity did not spontaneously terminate; the first period of activity was terminated by increasing the hyperpolarizing current injection to –64 pA. The MNC subsequently re-started firing without further adjustment of the hyperpolarizing current injection but again did not spontaneously terminate firing. Increasing the hyperpolarizing current injection to –76 pA failed to terminate the second period of firing but a further increase to –90 pA successfully terminated firing. Thus, during blockade of κ-opioid receptors, the MNC could spontaneously initiate activity but could not spontaneously terminate firing (as seen in 2 of 6 MNCs tested).

Discussion

Here, we show that DAPs and associated plateau potentials that sustain phasic bursts in MNCs are subject to powerful activity-dependent inhibition by intrinsic mechanisms; conditioning trains attenuated the amplitude of DAPs that generate plateau potentials as a function of the number of spikes evoked and the resultant inhibition required tens of seconds to fully recover. Thus, the kinetics of this effect are consistent with the rhythmicity of phasically active MNCs in vitro (Andrew & Dudek, 1983; Ghamari-Langroudi & Bourque, 1998; Brown et al. 1999) and in vivo (Harris et al. 1975; Wakerly et al. 1975; Brimble & Dyball, 1977; Brown et al. 1998). Since DAPs could not be fully activated soon after the conditioning spike trains, the mechanisms underlying the generation of DAPs are presumably subject to activity-dependent inactivation.

Our results also show that the activity-dependent inhibition of plateau potentials depends on the local release of dynorphin, an opioid peptide colocalized with vasopressin in the dendritic neurosecretory vesicles of MNCs (Shuster et al. 2000), and on the activation of κ-opioid receptors, but not vasopressin receptors. Thus, when MNC vesicles undergo exocytosis from the dendrites (Pow & Morris, 1989) colocalized dynorphin will also be released and is probably responsible for the activity-dependent inhibition of plateau potentials. Interestingly, protection of endogenously released dynorphin from peptidase degradation enhanced the inhibition of the DAP, indicating that the autocrine actions of dynorphin are limited by extracellular peptidases. Thus, peptidases may effectively re-set the κ-opioid feedback mechanism between bursts since we have recently shown that endogenous κ-opioid inhibition of firing is absent at the onset of spontaneous phasic bursts in vivo but emerges as bursts progress (Brown et al. 2004). While peptidases certainly limit autocrine feedback by dynorphin, a more important role of peptidases may be to limit paracrine actions of dynorphin to prevent synchronization of activity between vasopressin MNCs. Indeed rhythmic bursts in adjacent MNCs occur asynchronously (Cross et al. 1975; Leng & Dyball, 1983), a feature that contributes to the graded non-pulsatile pattern of vasopressin release in vivo.

Effects of coreleased peptides on phasic firing

In vivo, both V1 (Ludwig & Leng, 1997) and κ-opioid (Brown et al. 1998) receptor antagonists prolong burst duration and increase firing rate within bursts. Vasopressin and V1 receptors (Hurbin et al. 2002) and dynorphin and κ-opioid receptors (Shuster et al. 2000) are all colocalized within vasopressin neurosecretory vesicles. Thus, dendritic release of vasopressin and dynorphin will be accompanied by translocation of vasopressin receptors and κ-opioid receptors to the cell surface. Surprisingly, the time courses of the effects of endogenous vasopressin and κ-opioid peptides on phasic firing are different since V1 receptor antagonism induces a uniform elevation of firing rate throughout each burst in vivo (i.e. from burst onset) whereas κ-opioid receptor antagonism does not alter the firing rate at the onset of bursts but progressively increases firing rate over the course of each burst (Brown et al. 2004). Both vasopressin (Burbach et al. 1993) and dynorphin (Hiranuma et al. 1998) are broken down by extracellular peptidases in the brain. Nevertheless, there are measurable concentrations of vasopressin in SON extracellular fluid, even under basal conditions (Ludwig, 1998). The concentration of dynorphin in the SON is not known but vasopressin is estimated to be present in neurosecretory vesicles at much higher concentrations than dynorphin (by orders of magnitude) and we have now shown that plateau potentials fully recover from inhibition by endogenous dynorphin over ∼20 s in vitro. Thus, the simplest explanation of the observed effects of endogenous vasopressin and dynorphin is that dynorphin (due to its lower initial concentrations) is cleared rapidly from the local environment to re-set the κ-opioid receptor system between bursts whereas activation of V1 receptors is maintained between bursts by the persistent presence of vasopressin (Brown et al. 2004).

Opioid receptors belong to the G protein-coupled receptor (GPCR) family and most GPCRs reduce their responsiveness to agonists during prolonged exposure due to a combination of desensitization (over seconds to hours), internalization (over minutes to hours), and down-regulation (over hours to days). Full recovery from activity-dependent κ-opioid inhibition of the DAP occurred over ∼20 s. Thus, desensitization and/or internalization of κ-opioid receptors (Li et al. 1999) could also contribute to the recovery process. However, the recovery from activity-dependent DAP inhibition was best fitted by a single exponential with a time constant of 4–5 s suggesting that functional recovery is dominated by a single process. Since blockade of peptidases enhanced DAP inhibition it would appear that clearance of agonist (presumably by a combination of diffusion and degradation), rather than desensitization/internalization of κ-opioid receptors, is primarily responsible for the recovery from DAP inhibition.

Mechanisms of activity-dependent burst termination

The mechanisms underlying κ-opioid modulation of phasic activity are not known. In MNCs κ-opioid receptor activation modulates voltage-activated K+ currents; U50,488H increases the transient A current, but suppresses the delayed rectifier current (Muller et al. 1999). In addition, exogenous dynorphin inhibits EPSPs, IPSPs (Inenaga et al. 1994) and Ca2+ components of spikes (Mason et al. 1988; Inenaga et al. 1994). However, in our current work κ-opioid receptor antagonism did not affect Ca2+-dependent post-train AHPs or activity-dependent spike broadening during evoked spike trains (data not shown); rather we found that the important action of endogenous dynorphin in the modulation of phasic activity is through inhibition of DAPs and associated plateau potentials.

An alternative mechanism by which bursts could be terminated is through activation of the AHP since AHPs provide an activity-dependent hyperpolarizing influence and so summation of AHPs could terminate bursts by reversal of the plateau potential. However, blockade of the fast component of the AHP with apamin does not prolong bursts; rather it shortens bursts (Kirkpatrick & Bourque, 1996). There is also a slow component of the AHP that can be blocked by charybodotoxin (ChTX) (Greffrath et al. 1998). However, the slow component of the AHP only becomes evident in vitro when spike trains are evoked at > 20 Hz, a firing rate achieved only at the onset of phasic bursts and intracellular Ca2+ concentrations plateau early in bursts (Roper et al. 2004). Hence, the dynamics of slow AHP activation do not appear to match those necessary to terminate spontaneous bursts.

Thus it is probable that the critical modulator of phasic firing is activity-dependent inhibition of DAPs and associated plateau potentials. DAPs are Ca2+ dependent (Bourque, 1986; Andrew, 1987; Li & Hatton, 1997; Li & Hatton, 1997). Reduction of intracellular Ca2+ with calbindin reduces DAP amplitude and converts phasic activity to continuous firing, while immunoneutralization of calbindin converts continuous into phasic firing in MNCs (Li et al. 1995). Thus, it would appear likely that endogenous κ-opioid inhibition of DAPs would result from an inhibition of Ca2+ influx and/or release from intracellular stores. However, we found no effect of κ-opioid receptor antagonism on AHPs, suggesting that Ca2+ dynamics associated with AHPs in MNCs are not altered by endogenous dynorphin. Indeed, a recent study combining electrophysiological recording, Ca2+ imaging and computational modelling of phasic activity has shown that intracellular Ca2+ concentrations rise at burst onset, plateau after ∼5 s of firing (long before burst termination) and only fall once activity has stopped, and has generated a computational model indicating that activity-dependent inhibition may occur by inducing a reduction in the Ca2+ sensitivity of the DAP mechanism rather than by altering Ca2+ dynamics (Roper et al. 2004). Further studies will be required to address the important issue of whether dendritic dynorphin release mediates this proposed reduction in the Ca2+ sensitivity of the DAP mechanism.

Conclusions

In conclusion, we have shown that dynorphin coreleased with vasopressin from the dendrites of MNCs causes activity-dependent inhibition of DAPs and associated plateau potentials to terminate bursts and thus generate the characteristic phasic firing of vasopressin MNCs.

Acknowledgments

This work was supported by The Wellcome Trust (C.H.B.) and The Canadian Institutes of Health Research (C.W.B.).

References

  1. Andrew RD. Endogenous bursting by rat supraoptic neuroendocrine cells is calcium dependent. J Physiol. 1987;384:451–465. doi: 10.1113/jphysiol.1987.sp016463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Andrew RD, Dudek FE. Burst discharge in mammalian neuroendocrine cells involves an intrinsic regenerative mechanism. Science. 1983;221:1050–1052. doi: 10.1126/science.6879204. [DOI] [PubMed] [Google Scholar]
  3. Armstrong WE, Smith BN, Tian M. Electrophysiological characteristics of immunochemically identified rat oxytocin and vasopressin neurones in vitro. J Physiol. 1994;475:115–128. doi: 10.1113/jphysiol.1994.sp020053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bourque CW. Calcium-dependent spike after-current induces burst firing in magnocellular neurosecretory cells. Neurosci Lett. 1986;70:204–209. doi: 10.1016/0304-3940(86)90464-7. [DOI] [PubMed] [Google Scholar]
  5. Brimble MJ, Dyball RE. Characterization of the responses of oxytocin- and vasopressin-secreting neurones in the supraoptic nucleus to osmotic stimulation. J Physiol. 1977;271:253–271. doi: 10.1113/jphysiol.1977.sp011999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brown CH, Ghamari-Langroudi M, Leng G, Bourque CW. Kappa-opioid receptor activation inhibits post-spike depolarizing after-potentials in rat supraoptic nucleus neurones in vitro. J Neuroendocrinol. 1999;11:825–828. doi: 10.1046/j.1365-2826.1999.00419.x. [DOI] [PubMed] [Google Scholar]
  7. Brown CH, Leng G. In vivo modulation of post-spike excitability in vasopressin cells by kappa-opioid receptor activation. J Neuroendocrinol. 2000;12:711–714. doi: 10.1046/j.1365-2826.2000.00547.x. [DOI] [PubMed] [Google Scholar]
  8. Brown CH, Ludwig M, Leng G. Kappa-opioid regulation of neuronal activity in the rat supraoptic nucleus in vivo. J Neurosci. 1998;18:9480–9488. doi: 10.1523/JNEUROSCI.18-22-09480.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Brown CH, Ludwig M, Leng G. Temporal dissociation of the feedback effects of dendritically co-released peptides on rhythmogenesis in vasopressin cells. Neuroscience. 2004;124:105–111. doi: 10.1016/j.neuroscience.2003.11.038. [DOI] [PubMed] [Google Scholar]
  10. Brown CH, Russell JA, Leng G. Opioid modulation of magnocellular neurosecretory cell activity. Neurosci Res. 2000;36:97–120. doi: 10.1016/s0168-0102(99)00121-2. [DOI] [PubMed] [Google Scholar]
  11. Burbach JP, De Bree FM, Terwel D, Tan A, Maskova HP, Van Der Kleij AA. Properties of aminopeptidase activity involved in the conversion of vasopressin by rat brain membranes. Peptides. 1993;14:807–813. doi: 10.1016/0196-9781(93)90118-z. [DOI] [PubMed] [Google Scholar]
  12. Cross BA, Dyball RE, Dyer RG, Jones CW, Lincoln DW, Morris JF, Pickering BT. Endocrine neurons. Recent Prog Horm Res. 1975;31:243–294. doi: 10.1016/b978-0-12-571131-9.50011-6. [DOI] [PubMed] [Google Scholar]
  13. Ghamari-Langroudi M, Bourque CW. Caesium blocks depolarizing after-potentials and phasic firing in rat supraoptic neurones. J Physiol. 1998;510:165–175. doi: 10.1111/j.1469-7793.1998.165bz.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Greffrath W, Martin E, Reuss S, Boehmer G. Components of after-hyperpolarization in magnocellular neurones of the rat supraoptic nucleus in vitro. J Physiol. 1998;513:493–506. doi: 10.1111/j.1469-7793.1998.493bb.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Harris MC, Dreifuss JJ, Legros JJ. Excitation of phasically firing supraoptic neurones during vasopressin release. Nature. 1975;258:80–82. doi: 10.1038/258080b0. [DOI] [PubMed] [Google Scholar]
  16. Hawkins KN, Knapp RJ, Lui GK, Gulya K, Kazmierski W, Wan YP, Pelton JT, Hruby VJ, Yamamura HI. [3H]-[H-D-Phe-Cys-Tyr-D-Trp-Orn-Thr-Pen-Thr-NH2] ([3H]CTOP), a potent and highly selective peptide for mu opioid receptors in rat brain. J Pharmacol Exp Ther. 1989;248:73–80. [PubMed] [Google Scholar]
  17. Hiranuma T, Kitamura K, Taniguchi T, Kanai M, Arai Y, Iwao K, Oka T. Protection against dynorphin-(1–8) hydrolysis in membrane preparations by the combination of amastatin, captopril and phosphoramidon. J Pharmacol Exp Ther. 1998;286:863–869. [PubMed] [Google Scholar]
  18. Hlubek M, Tian D, Stuenkel EL. Mechanism of alpha-latrotoxin action at nerve endings of neurohypophysis. Brain Res. 2003;992:30–42. doi: 10.1016/j.brainres.2003.08.025. [DOI] [PubMed] [Google Scholar]
  19. Hurbin A, Orcel H, Alonso G, Moos F, Rabie A. The vasopressin receptors colocalize with vasopressin in the magnocellular neurons of the rat supraoptic nucleus and are modulated by water balance. Endocrinology. 2002;143:456–466. doi: 10.1210/endo.143.2.8643. [DOI] [PubMed] [Google Scholar]
  20. Inenaga K, Nagatomo T, Nakao K, Yanaihara N, Yamashita H. Kappa-selective agonists decrease postsynaptic potentials and calcium components of action potentials in the supraoptic nucleus of rat hypothalamus in vitro. Neuroscience. 1994;58:331–340. doi: 10.1016/0306-4522(94)90039-6. [DOI] [PubMed] [Google Scholar]
  21. Kirkpatrick K, Bourque CW. Activity dependence and functional role of the apamin-sensitive K+ current in rat supraoptic neurones in vitro. J Physiol. 1996;494:389–398. doi: 10.1113/jphysiol.1996.sp021500. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kombian SB, Mouginot D, Pittman QJ. Dendritically released peptides act as retrograde modulators of afferent excitation in the supraoptic nucleus in vitro. Neuron. 1997;19:903–912. doi: 10.1016/s0896-6273(00)80971-x. [DOI] [PubMed] [Google Scholar]
  23. Kruszynski M, Lammek B, Manning M, Seto J, Haldar J, Sawyer WH. [1-beta-Mercapto-beta, beta-cyclopentamethylenepropionic acid),2-(O-methyl) tyrosine]argine-vasopressin and [1-beta-mercapto-beta, beta-cyclopentamethylenepropionic acid)]argine-vasopressin, two highly potent antagonists of the vasopressor response to arginine-vasopressin. J Med Chem. 1980;23:364–368. doi: 10.1021/jm00178a003. [DOI] [PubMed] [Google Scholar]
  24. Leng G, Brown CH, Russell JA. Physiological pathways regulating the activity of magnocellular neurosecretory cells. Prog Neurobiol. 1999;57:625–655. doi: 10.1016/s0301-0082(98)00072-0. [DOI] [PubMed] [Google Scholar]
  25. Leng G, Dyball RE. Intercommunication in the rat supraoptic nucleus. Q J Exp Physiol. 1983;68:493–504. doi: 10.1113/expphysiol.1983.sp002742. [DOI] [PubMed] [Google Scholar]
  26. Li Z, Decavel C, Hatton GI. Calbindin-D28k: role in determining intrinsically generated firing patterns in rat supraoptic neurones. J Physiol. 1995;488:601–608. doi: 10.1113/jphysiol.1995.sp020993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Li Z, Hatton GI. Ca2+ release from internal stores: role in generating depolarizing after-potentials in rat supraoptic neurones. J Physiol. 1997;498:339–350. doi: 10.1113/jphysiol.1997.sp021862. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Li JG, Luo LY, Krupnick JG, Benovic JL, Liu-Chen LY. U50,488H-induced internalization of the human kappa opioid receptor involves a beta-arrestin- and dynamin-dependent mechanism. Kappa receptor internalization is not required for mitogen-activated protein kinase activation. J Biol Chem. 1999;274:12087–12094. doi: 10.1074/jbc.274.17.12087. [DOI] [PubMed] [Google Scholar]
  29. Ludwig M. Dendritic release of vasopressin and oxytocin. J Neuroendocrinol. 1998;10:881–895. doi: 10.1046/j.1365-2826.1998.00279.x. [DOI] [PubMed] [Google Scholar]
  30. Ludwig M, Leng G. Autoinhibition of supraoptic nucleus vasopressin neurons in vivo: a combined retrodialysis/electrophysiological study in rats. Eur J Neurosci. 1997;9:2532–2540. doi: 10.1111/j.1460-9568.1997.tb01682.x. [DOI] [PubMed] [Google Scholar]
  31. Mason WT, Cobbett P, Inenaga K, Legendre P. Ionic currents in cultured supraoptic neurons: actions of peptides and transmitters. Brain Res Bull. 1988;20:757–764. doi: 10.1016/0361-9230(88)90088-3. [DOI] [PubMed] [Google Scholar]
  32. Muller W, Hallermann S, Swandulla D. Opioidergic modulation of voltage-activated K+ currents in magnocellular neurons of the supraoptic nucleus in rat. J Neurophysiol. 1999;81:1617–1625. doi: 10.1152/jn.1999.81.4.1617. [DOI] [PubMed] [Google Scholar]
  33. Portoghese PS, Lipkowski AW, Takemori AE. Binaltorphimine and nor-binaltorphimine, potent and selective kappa-opioid receptor antagonists. Life Sci. 1987;40:1287–1292. doi: 10.1016/0024-3205(87)90585-6. [DOI] [PubMed] [Google Scholar]
  34. Pow DV, Morris JF. Dendrites of hypothalamic magnocellular neurons release neurohypophysial peptides by exocytosis. Neuroscience. 1989;32:435–439. doi: 10.1016/0306-4522(89)90091-2. [DOI] [PubMed] [Google Scholar]
  35. Renaud LP, Bourque CW. Neurophysiology and neuropharmacology of hypothalamic magnocellular neurons secreting vasopressin and oxytocin. Prog Neurobiol. 1991;36:131–169. doi: 10.1016/0301-0082(91)90020-2. [DOI] [PubMed] [Google Scholar]
  36. Rhodes CH, Morrell JI, Pfaff DW. Immunohistochemical analysis of magnocellular elements in rat hypothalamus: distribution and numbers of cells containing neurophysin, oxytocin, and vasopressin. J Comp Neurol. 1981;198:45–64. doi: 10.1002/cne.901980106. [DOI] [PubMed] [Google Scholar]
  37. Richard P, Moos F, Dayanithi G, Gouzenes L, Sabatier N. Rhythmic activities of hypothalamic magnocellular neurons: autocontrol mechanisms. Biol Cell. 1997;89:555–560. [PubMed] [Google Scholar]
  38. Roper P, Armstrong WE, Callaway J. Burst initiation and termination in phasic vasopressin cells of the rat SON: a combined mathematical, electrical and calcium fluorescence study. J Neurosci. 2004;24:4818–4831. doi: 10.1523/JNEUROSCI.4203-03.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Shuster SJ, Riedl M, Li X, Vulchanova L, Elde R. The kappa opioid receptor and dynorphin co-localize in vasopressin magnocellular neurosecretory neurons in guinea-pig hypothalamus. Neuroscience. 2000;96:373–383. doi: 10.1016/s0306-4522(99)00472-8. [DOI] [PubMed] [Google Scholar]
  40. Ushkaryov Y. Alpha-latrotoxin: from structure to some functions. Toxicon. 2002;40:1–5. doi: 10.1016/s0041-0101(01)00204-5. [DOI] [PubMed] [Google Scholar]
  41. Wakerly JB, Poulain DA, Dyball RE, Cross BA. Activity of phasic neurosecretory cells during haemorrhage. Nature. 1975;258:82–84. doi: 10.1038/258082a0. [DOI] [PubMed] [Google Scholar]
  42. Whitnall MH, Gainer H, Cox BM, Molineaux CJ. Dynorphin-A-(1–8) is contained within vasopressin neurosecretory vesicles in rat pituitary. Science. 1983;222:1137–1139. doi: 10.1126/science.6648526. [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES