Prokineticin 2 (PK2) is a neuropeptide that acts as a signaling molecule regulating circadian rhythms in mammals. We have previously reported PK2 actions on subfornical organ (SFO) neurons, identifying this circumventricular organ as a target at which PK2 acts to influence autonomic control (Cottrell GT, and Ferguson AV. J. Neurosci. 24: 2375–2379, 2004). In this study, we have examined the cellular mechanisms by which PK2 increases the excitability of SFO neurons. Whole cell patch recordings from dissociated rat SFO neurons demonstrated that the mitogen-activated protein (MAP) kinase inhibitor PD-98059 prevented PK2-induced depolarization and decreases in delayed rectifier K+ current. PK2 also increased intracellular Ca2+ concentration ([Ca2+]i) in 39% of dissociated SFO neurons (mean increase = 20.8 ± 5.5%), effects that were maintained in the presence of thapsigargin but abolished by both nifedipine, or the absence of extracellular Ca2+, suggesting that PK2-induced [Ca2+]i transients resulted from Ca2+ entry through voltage-gated Ca2+ channels. Voltage-clamp recordings showed that PK2 was without effects on Ca2+ currents evoked by voltage ramps, suggesting that PK2-induced Ca2+ influx was secondary to PK2-induced increases in action potential frequency, an hypothesis supported by data showing that tetrodotoxin abolished effects of PK2 on [Ca2+]i. These observations suggested PK2 modulation of voltage-gated Na+ currents, a possibility confirmed by voltage-clamp experiments showing that PK2 increased the amplitude of both transient and persistent Na+ currents in 29% of SFO neurons (by 34 and 38%, respectively). These data indicate that PK2 influences SFO neurons through the activation of a MAP kinase cascade, which, in turn, modulates Na+ and K+ conductances.
- circumventricular organs
- potassium current
- sodium current
- mitogen-activated protein kinase
- intracellular Ca2+
prokineticins were originally isolated as human orthologs of secreted peptides from snake (mamba intestinal toxin 1) or frog (Bv8 toxin from Bombina variegata), which possess the ability to cause potent contractions of gastrointestinal smooth muscles (36). Two mammalian peptides, prokineticin 1 (PK1) and prokineticin 2 (PK2), also possessing the ability to modulate gastrointestinal motility, were identified, with the latter showing a circadian pattern of expression (peaks during the day and troughs at night) in the suprachiasmatic nucleus (4). The fact that the oscillation was observed under conditions of 24-h light or dark indicated that PK2 expression is regulated by the endogenous circadian clock (4). Moreover, the observation that, although PK2 was expressed in other brain areas, including the median preoptic nucleus of the hypothalamus and the nucleus tractus solitarius (NTS), circadian changes were only observed in the suprachiasmatic nucleus of the hypothalamus (SCN) (4, 24) indicated that PK2 may play a key role as an output molecule for circadian regulation and autonomic function. Recently, experiments using mice, in which either the gene encoding the PK2 neuropeptide or the gene encoding its cognate receptor, PKR2, have been knocked out, have confirmed the role of PK2 as a potentially important signaling molecule. PK2 knockout mice displayed altered patterns of wheel-running activity, circadian and homeostatic sleep, thermoregulation, corticosterone and glucose concentrations, and feeding (15, 19), while PKR2 knockout mice exhibited disrupted wheel-running activity and impaired nocturnal thermoregulation (27).
PKR2 have been localized to a number of targets in the central nervous system, including nuclei involved in vision (superior colliculus), and the sleep-wake cycle/arousal (dorsal raphe, dorsomedial hypothalamus, lateral hypothalamus, and perifornical region of the hypothalamus) (5). Interestingly, PKR2 have also been localized to areas involved in regulation of food intake and water/salt balance, such as the arcuate nucleus, paraventricular nucleus of the hypothalamus, and the subfornical organ (SFO) (1, 5, 7, 8, 13, 16, 24). Recently, it was demonstrated that intracerebroventricular and arcuate nucleus microinjections of Bv8 resulted in an inhibition of normal circadian feeding behavior (24). In addition, direct microinjection of Bv8 into the SFO, a sensory circumventricular organ that lacks a normal blood-brain barrier, resulted in increases in drinking similar to that observed with the microinjection of angiotensin (ANG) II (24). Neurons in the SFO have been reported to monitor the constituents of both the cerebrospinal fluid and the peripheral circulation and integrate such sensory information. In turn, through their efferent projections to critical hypothalamic autonomic control centers, SFO neurons then utilize this integrated information to control vital autonomic outputs regulating cardiovascular function, the stress response, drinking and feeding behavior, and reproduction (for review, see Ref. 7).
Our laboratory has recently demonstrated that PK2 depolarizes and increases spike frequency in a subpopulation of SFO neurons (6). Although the ligand-receptor interactions underlying this depolarization have yet to be elucidated, we observed specific receptor-mediated actions of PK2 in inhibiting a delayed rectifier K+ current (IK). This effect likely contributes to the increase in spike frequency associated with the membrane depolarization; however, it is unlikely that such modulation of IK is solely responsible for the underlying depolarization observed in response to PK2.
PK2 has been previously demonstrated to activate p44/42 mitogen-activated protein kinase (MAPK) (18, 21–23), an action that has been demonstrated to influence membrane excitability in several neuronal systems (2, 34, 35), identifying this signaling pathway as a likely candidate for the regulation of membrane excitability by PK2. Increases in intracellular Ca2+ concentration ([Ca2+]i) have also been suggested to be important components of the intracellular changes associated with PKR2 activation in both gastrointestinal smooth muscle (20) and dorsal root ganglion (DRG) neurons (25). These effects are associated with the influx of extracellular Ca2+, most likely through L-type Ca2+ channels. Interestingly, SFO neurons have been shown to express L- and N-type Ca2+ currents, the latter of which are potentiated by ANG II (32), suggesting a potential common mechanism through which PK2R and AT1 receptors may stimulate drinking.
In the present study, we have investigated the mechanisms through which PK2 regulates SFO neuron excitability. Specifically, we have used patch-clamp and imaging techniques to determine the intracellular mediators and ion channels through which PK2 modulates the excitability of SFO neurons.
Unless otherwise indicated, chemicals were acquired from Sigma (St. Louis, MO). Recombinant PK2 and the PK2 antagonist A1MPK1 were generously provided to us by Dr. Qun-Yong Zhou, University of California at Irvine. The concentrations of PK2 used in these studies (10–20 nM) were chosen in accordance with those previously demonstrated to be effective in eliciting receptor-mediated responses in SFO neurons and, in all cases, were at least 100-fold higher than the reported EC50 for PK2 effects on these cells (6). A1MPK1 was used at 400 nM, a concentration previously demonstrated to be effective in inhibiting PK2-mediated responses (3, 6). Depletion of extracellular Ca2+ was achieved by bathing cells in Ca2+-Mg2+ free Hanks' balanced salt solution (HBSS; Invitrogen, Grand Island, NY), supplemented with 1 mM EGTA. Depletion of intracellular Ca2+ stores was accomplished by application of 500 μM thapsigargin (TG) (30). Blockade of L-type Ca2+ and voltage-gated sodium channels required treatment with 1 μM nifedipine (32) and 500 nM tetrodotoxin (TTX; Alomone Laboratories, Jerusalem, Israel), respectively. Inhibition of MAPK activity was achieved through the use of the MAPK inhibitor PD-98059 (10 μM) (14). Nystatin stock solution (100 mg/ml in dimethylsulfoxide) for perforated patch recordings was made fresh each day and stored at 4°C until use in experiments, when it was then suspended in pipette solution (see Electrophysiology) at a final concentration of 250 μg/ml. DMSO content in the internal solution was never >0.03%.
Animal experiments were approved by the Queen's University Animal Care Committee and conformed to the standards of the Canadian Council for Animal Care. Dissociated SFO neurons were prepared as previously described (11). Briefly, male Sprague-Dawley rats (125–175 g) were decapitated, and their brains were removed and immediately immersed in ice-cold Ca2+ and Mg2+-free HBSS supplemented with 0.03 M sucrose. A tissue block containing the hippocampal commissure and SFO was dissected free from the brain, and the SFO was separated from the surrounding tissue. The isolated SFO was immersed in 1 mg/ml trypsin and incubated at 37°C in 5% CO2 for 30 min. Cells were then suspended in ice-cold Ca2+ containing HBSS (Invitrogen), supplemented with 4 μg/ml BSA, triturated through a 20-gauge needle, and centrifuged at 900 g for 5 min. The supernatant was removed, cells resuspended, and a second 5-min centrifugation was performed. The supernatant was again removed, and cells were resuspended and plated in glass-bottomed culture dishes for imaging (MatTek, Ashland, MA) or 35-mm plastic culture dishes for electrophysiology (Corning, Corning, NY). Cells were cultured in Neuorobasal A Medium (Invitrogen) at 37°C for a minimum of 24 h before Ca2+ imaging and electrophysiology experiments.
Cell recordings were obtained from SFO cells identified as neurons based on the presence of voltage-gated Na+ currents under voltage-clamp conditions and at least 80-mV action potentials induced at a holding potential below −55 mV in current-clamp recordings. For most electrophysiological experiments, signals were amplified using an Axopatch 200B amplifier (Axon Instruments), filtered at 1 kHz, and digitized using a CED Micro1401 interface (CED, Cambridge, UK) at 8 kHz. These experiments utilized a conventional patch electrode-holder assembly, where an Ag-AgCl wire is in direct contact with the patch electrode filling solution. For experiments investigating PK2 effects on voltage-gated Na+ current, signals were amplified using a Multiclamp 700B amplifier (Axon Instruments), filtered at 10 kHz, and sampled at 40 kHz. These experiments utilized a patch electrode-holder assembly containing a micro-agar salt bridge instead of an Ag-AgCl wire to eliminate junction potentials (28). Voltage-clamp data were collected using the Signal software package (CED, Cambridge, UK).
The effects of PK2 on Ca2+ currents were assessed using whole cell voltage-clamp recordings and were performed as reported previously (32). Patch electrodes were fabricated from borosilicate glass and had resistances of 2–6 MΩ when filled with internal recording solution containing (in mM) 130 potassium-gluconate, 10 KCl, 10 HEPES, 10 EGTA, 1 MgCl2, 0.1 CaCl2, and 2 Na2ATP (290–300 osM, pH = 7.2). Initial recordings from SFO neurons were performed in normal artificial cerebrospinal fluid (aCSF) containing (in mM) 140 NaCl, 5 KCl, 1 MgCl, 10 HEPES, 10 glucose, and 2 CaCl2 at a pH of 7.2 and osmolarity of 290–300 osM. Pharmacological isolation of Ca2+ channel currents was then achieved by bath exchange with aCSF of the following composition (mM): 120 NaCl, 10 BaCl2, 10 tetraethylammonium chloride, 1 4-aminopyridine, 1 MgCl2, 10 HEPES, 10 glucose, and 0.0005 TTX at a pH of 7.2 and osmolarity of 290–300 osM. Calcium channel identity was confirmed at the end of each experiment by blockade with 1 mM CdCl2.
Experiments evaluating the influence of MAPK activity on PK2-mediated depolarization and inhibition of IK were performed as reported previously (5) using perforated patch-clamp techniques (electrodes filled with Nystatin-containing internal recording solution) to minimize current rundown during voltage-clamp experiments. Following pipette/cell contact, a high-resistance seal (∼2–8 GΩ) was obtained, and seal quality and cell access were monitored using −10 mV pulses from a holding potential of −65 mV. Once series resistance was below 30 MΩ, outward currents were monitored over time until cell access stabilized, which usually occurred within 15 min. Changes in series resistance were monitored between recordings, and electrode suction was applied at the end of experiments to ensure that the perforated patch configuration was maintained (rather than whole cell access) throughout the experiments.
Experiments investigating the effect of PK2 on voltage-gated Na+ currents utilized an external and internal recording solution that blocked essentially all currents except the Na+ current. The external recording solution contained (in mM) 25 NaCl, 130 tetraethylammonium chloride, 1 MgCl2, 2 CaCl2, 1 CsCl2, 1 BaCl2, 0.3 CdCl2, 10 HEPES, 10 glucose, (pH 7.2, osmolarity of 290–300 osM). Patch electrodes were filled with internal recording solution containing (in mM) 130 CsMeSO4, 10 CsCl, 1 MgCl2, 10 HEPES, 1 EGTA, 2 Na2ATP (pH 7.2, osmolarity of 290–300 osM). Observed changes in Na+ current were partially reversible after 5 min of washout with external recording solution.
Measurements of [Ca2+]i were achieved by ratiometric imaging of SFO neurons loaded with fura 2 (Molecular Probes, Eugene, OR). Neurons were loaded by incubation (30 min, 37°C) with the acetomethoxy ester form of fura 2 (3 μM), followed by an HBSS wash and a second incubation to permit esterase conversion to free fura 2. Imaging was performed using an InCyt dual-wavelength imaging system (Intracellular Imaging, Cincinnati, OH). [Ca2+]i was calculated from images collected at 0.5 Hz using ratiometric comparison of emissions (505 nm) from excitation at 340-nm and 380-nm wavelengths and comparing these ratios to that obtained from known Ca2+ calibration standards.
The experimental protocol involved a 5-min baseline recording of [Ca2+]i, followed by bath exchange by gravity perfusion of HBSS supplemented with PK2. The PK2 treatment period lasted 8 min and was followed by bath exchange with HBSS and a recovery period of generally 20 min. The experiments were concluded by perfusion with HBSS containing 40 mM KCl to confirm cell viability.
In electrophysiology experiments assessing effects on Ca2+ channels, peak Ba2+ current (IBa) during depolarizing voltage ramps was calculated 3–5 min after treatment initiation [a period previously demonstrated to be associated with peak depolarizing responses (6)] and normalized to peak IBa before treatment. In current-clamp experiments and experiments examining IK, data were analyzed as reported previously (6). In experiments assessing effects on Na+ channels, peak transient Na+ current (INaT; obtained during a 15-ms step from −90 to 0 mV) and peak persistent Na+ current (INaP; obtained during depolarizing voltage ramps) were measured 3–5 min after treatment initiation.
Mean [Ca2+]i was calculated during the 60 s preceding PK2 application and the last minute of PK2 treatment before bath exchange washout (420–480 s), a time point at which most responding cells displayed a peak response. Change in [Ca2+]i was normalized to baseline measures and compared with control HBSS perfusion at the same time points using Student's t-test. A neuron was classified as PK2 responsive if it underwent a minimum [Ca2+]i increase of 20% during the last minute of the treatment period. Neurons that did not display a minimum of 50% increase in [Ca2+]i following 40 mM KCl perfusion at the end of the recording period were excluded from further analysis.
Comparison between mean responses of treatment groups was performed using Student's t-test or one-way ANOVA followed by Tukey's multiple-comparison test. Comparisons between proportions of responsive neurons were carried out using the χ2 test (without Yates correction). All values are presented as means ± SE, with P values set at 0.05.
PK2-mediated membrane depolarization and inhibition of IK requires MAPK.
The regulation of membrane excitability by PK2 activation of MAPK was explored in current-clamp and voltage-clamp studies. Using current-clamp techniques, we first examined the role of MAPK in PK2-mediated depolarizations and increases in action potential firing frequency and were able to show that the effects of PK2 on SFO neurons were abolished in the presence of the specific MAPK inhibitor PD-98059 (Fig. 1A). The mean change in membrane potential following PK2 treatment in the presence of PD-98059 was −3.6 ± 3.3 mV (n = 9), a value not significantly different from a control aCSF treatment (2.8 ± 1.1 mV, n = 14) and significantly lower (P < 0.01) than that observed following PK2 treatment in the presence of aCSF alone (7.5 ± 1.7 mV, n = 45, with 18 cells responding to PK2) (Fig. 1B). In addition, the proportion of neurons influenced by PK2 (18/45) was significantly greater than that observed following PD-98059 (0/9; χ2 test; P < 0.05). Treatment of SFO neurons with PD-98059 alone did not cause a significant change in membrane potential (0.5 ± 0.7 mV, n = 9).
Voltage-clamp recordings using perforated patch techniques to isolate IK also demonstrated that PD-98059 prevented PK2-mediated attenuation of IK in SFO neurons (Fig. 1C). The mean change in IK at the 20-mV voltage step following PK2 treatment in the presence of PD-98059 was 4.7 ± 9% (n = 7), a value not significantly different from that recorded in aCSF alone (10 ± 12% increase, n = 7), but significantly different from the clear decreases in this conductance normally observed in response to PK2 treatment (26 ± 6% decrease, n = 6) (Fig. 1D). These data demonstrate that PK2-mediated changes in membrane excitability require the activation of the MAPK signaling pathway.
PK2 has no effect on nonselective cation conductance.
Although the above observations demonstrate a clear role for MAPK signaling in mediating PK2 effects on SFO neurons, it should be emphasized that the reported effects on IK, while consistent with the increases in action potential frequency observed in response to PK2, are likely not responsible for the accompanying depolarization. Interestingly, it has been suggested that PK2 may regulate a nonselective cation conductance (NSCC) in nociceptors (25, 31), and other neuropeptides have been shown to regulate SFO NSCC (9, 33). The possibility, therefore, exists that PK2 may depolarize SFO neurons through regulation of this current.
We, therefore, examined the effects of PK2 on SFO neuron NSCC using voltage-clamp experiments. Slow depolarizing voltage ramps (12 mV/s, −100–20 mV) were used to assess changes in NSCC following bath application of PK2. Subtracting the ramp currents obtained after application PK2 from those obtained in the control condition revealed that PK2 had no significant effects on the NSCC (Fig. 2A). The mean subtracted currents were linear between −100 and −30 mV and did not have an appreciable slope (n = 7, Fig. 2B), indicating that PK2 has no effect on NSCCs in SFO neurons. There must, therefore, be an alternative mechanism responsible for PK2-induced depolarizations.
PK2 increases [Ca2+]i in SFO neurons via a MAPK cascade.
Previous studies have suggested L-type Ca2+ channels to be a regulatory target through which PK2 modulates other cell types (3, 20–22, 29). Therefore, we next examined whether PK2 also regulated Ca2+ handling in SFO neurons using the method of Ca2+ imaging to assess [Ca2+]i changes in response to PK2. PK2 induced reversible increases in [Ca2+]i in 39% (n = 18) of all of the neurons imaged (n = 46), with these cells being subsequently labeled as “responders” (arbitrarily defined >20% increase in [Ca2+]i) (Fig. 3, A–C). In all cells imaged, PK2 stimulated a mean increase in [Ca2+]i of 20.8 ± 5.5% (n = 46), while similar HBSS wash of SFO neurons was without effects on [Ca2+]i (3.9 ± 5%, n = 20) (Fig. 3D). There were no statistically significant differences between the mean baseline concentrations of [Ca2+]i of PK2 (131 ± 11.4 nM), compared with HBSS-tested (160 ± 18.12 nM) SFO neurons (unpaired t-test, P > 0.1). In the PK2-treated cells identified as responders, the mean increase was 59.2 ± 7.6% (n = 18), and the duration of these changes in [Ca2+]i was variable, with a return toward baseline usually observed within 15 min, but complete recovery often taking longer (time courses similar to depolarizations observed in response to PK2), as illustrated in Figs. 3 and 4. A proportion of SFO neurons showed spontaneous calcium oscillations, and, as illustrated in Fig. 3C, when these cells were responsive to PK2, increases in both the frequency and magnitude of these oscillations were observed. A 10-min pretreatment of SFO neurons with the PKR2 antagonist A1MPK1 resulted in a complete absence of PK2 response with a mean change in [Ca2+]i of −4.5 ± 3.3% (n = 14; P < 0.05 compared with PK2), a value that was not significantly different from the control HBSS wash (Fig. 3D). In addition, the proportion of neurons influenced by PK2 (0/14) was significantly different from that under control conditions (18/46; χ2 test, P < 0.005). Inhibition of MAPK by PD-98059 blocked PK2-mediated increases in [Ca2+]i (mean Ca2+ change 1.0 ± 4.5%, n = 23; P < 0.05 compared with PK2) (Fig. 3D), with, again, a significantly different proportion of neurons influenced by PK2 following PD-98059 (0/23) compared with control conditions (18/46; χ2 test, P < 0.0005). These observations support the conclusion that PK2 causes receptor-mediated increases in [Ca2+]i via a MAPK cascade.
[Ca2+]i increases are dependent on extracellular Ca2+.
To determine whether the increase in [Ca2+]i observed with PK2 treatment was the result of mobilization of intracellular Ca2+ stores, we next performed experiments in the absence of extracellular Ca2+. Replacement of the HBSS bath media with Ca2+- and Mg2+-free HBSS supplemented with EGTA resulted in a gradual decrease in [Ca2+]i (Fig. 4A). PK2 treatment had no effect on [Ca2+]i, although the [Ca2+]i measured following treatment was significantly lower (−20.5 ± 3.9%, n = 16) than baseline levels as a result of progressive loss of intracellular Ca2+ stores (Fig. 4C).
We next examined whether PK2-mediated increases in [Ca2+]i were dependent on mobilization of Ca2+ from intracellular stores in neurons pretreated with TG to deplete these stores. TG pretreatment had no effect on PK2-induced increases in [Ca2+]i, with the number of cells responding (10 of 23) and the magnitude of the response (29.7 ± 10.2%) being similar to PK2 treatment alone (Fig. 4, B and C). The magnitude of [Ca2+]i change was also similar for the neurons labeled as responders (61.1 ± 20%). These data in combination indicate that the PK2-mediated increases in [Ca2+]i occurred through the influx of extracellular Ca2+ rather than release from intracellular stores.
PK2 induces calcium influx through L-type Ca2+ channels.
Gastric smooth muscle contraction in response to PK2 has been demonstrated to be dependent on modulation of L-type Ca2+ channels (20). Our demonstration here that PK2-induced increases in [Ca2+]i result from influx of this ion from extracellular sources suggested there may be a similar role for these channels in SFO neurons. Pretreatment of SFO neurons with nifedipine resulted in the blockade of PK2-mediated increases in [Ca2+]i (mean Ca2+ change −1.7 ± 4.9%) (Fig. 5, A and B), supporting a role for L-type Ca2+ channels in mediating the influx of extracellular Ca2+.
We next examined if PK2 modulated Ca2+ channels using voltage-clamp techniques to assess PK2 effects on pharmacologically isolated Ca2+ channel currents. Under these conditions, IBa elicited by slow depolarizing ramps were not affected by PK2 treatment (Fig. 5C). The normalized mean change in peak IBa was 1.3 ± 7.4% (n = 12) following PK2 treatment and was not significantly different from that measured following a control aCSF wash (3.3 ± 6.8%, n = 16) (Fig. 5D). There were no differences in the kinetics of activation of IBa, as can be observed in Fig. 5C and measured from depolarizing voltage pulses (−70 to 40 mV). These data, in combination, indicate that the increase in [Ca2+]i observed following PK2 treatment, although carried through L-type Ca2+ channels, was not the result of regulation of either the conductance or the voltage activation profile of these channels.
PK2 modulates voltage-gated Na+ currents via a MAPK cascade.
The paradox that nifedipine blocked PK2-mediated increases in [Ca2+]i in imaging experiments, while there were no direct effects of PK2 on Ca2+ currents, suggests that the Ca2+ influx through voltage-gated Ca2+ channels was the result of depolarization-induced opening of these channels. As previous PK2-mediated depolarizations were accompanied by increases in action potential firing frequency (6), Ca2+ imaging experiments were repeated with PK2 responses being evaluated in the presence of TTX. These experiments revealed that blockade of voltage-gated Na+ channels and action potentials abolished PK2-mediated increases in [Ca2+]i with mean changes in [Ca2+]i induced by PK2 under control conditions, which were 20.8 ± 5.5% (n = 46) compared with 3.6 ± 3.1% (n = 24) in the presence of TTX (Fig. 6, A and B).
Therefore, we next examined whether voltage-gated INaT and INaP were modulated by PK2. The INaT was activated using a step to 0 mV from a holding potential of −90 mV. Under these conditions, the peak INaT amplitude was 1.09 ± 0.14 nA (n = 24). Of the neurons tested, 29% exhibited an increase in the amplitude of INaT following PK2 treatment (Fig. 6C), a proportion that is similar to the proportion of neurons that are influenced by PK2 under current-clamp conditions. In the responding neurons, the normalized mean increase in peak INaT was 34.0 ± 9.0% (n = 7), which was significantly greater than the change in amplitude of INaT of 0.5 ± 2.9% (n = 11) in control treatments with aCSF alone (Fig. 6D).
The INaP component of Na+ current is thought to play a key role in regulating electrical activity in SFO neurons (12), so we also directly evaluated the effect of PK2 on INaP. INaP was isolated using a fast depolarizing ramp from −100 to +40 mV (from a holding potential of −90 mV) at a rate of 140 mV/s. The same neurons that exhibited an increase in INaT after treatment with PK2 also exhibited an increase in INaP. Following PK2 treatment, the amplitude of the INaP increased by 37.8 ± 11.4%, which was significantly greater than the change in amplitude of INaP of 1.5 ± 3.3% (n = 10) in control treatments (Fig. 6, E and F). The percentage of increase in amplitude of INaT in sensitive neurons was similar to the increase in amplitude of INaP, indicating that the two aspects of voltage-gated Na+ current are simultaneously regulated by PK2.
To test whether the Na+ currents are modulated by a MAPK signaling cascade, SFO neurons were subjected to pretreatment with PD-98059 before application of the PK2 peptide. In these experiments, pretreatment with PD-98059 abolished the potentiation of Na+ currents by PK2. Specifically, the amplitude of Na+ currents that was observed after application of PK2 was not significantly different from that observed after application of vehicle control solution (P > 0.05) and was significantly smaller than that observed after application of PK2 (P < 0.001) (mean change after treatment with PD-98059 and PK2: INaT = −4.8 ± 1.5%; mean change in INaP = −6.9 ± 3.6%; n = 9) (Fig. 6, B and D). In addition, the proportion of neurons influenced by PK2 (0/9) was significantly different from that under control conditions (7/24; χ2 test; P < 0.05). Treatment of SFO neurons with PD-98059 alone did not cause changes in amplitude of Na+ currents (mean change INaT = −0.5 ± 1.0%; mean change in INaP = −2.8 ± 2.6%; n = 15).
We have shown here that depolarizing effects of PK2 on SFO neurons are mediated through the activation of a MAPK cascade. Depolarizing effects of PK2 on SFO neurons were not the result of effects on NSCC. However, PK2 causes transient increases in [Ca2+]i that were dependent on Ca2+ entry from the extracellular space through L-type Ca2+ channels. These effects were not the result of regulation of these channels. Effects of PK2 on [Ca2+]i were, however, abolished by inhibition of MAPK or inhibition of sodium currents with TTX, suggesting potential modulation of Na+ currents. This suggestion was confirmed by potentiation of both INaT and INaP in SFO neurons by PK2, and that this potentiation was eliminated by inhibiting MAPK. The demonstration that PK2 effects are mediated by MAPK in SFO neurons supports the conclusion that this is a common mechanism that occurs subsequent to PK2 interactions with PKR2. Several studies in different cell systems have described the activation of p44/42 MAPK by PK2 (18, 21–23). In two of these studies, cell expression systems were used to express PKR2, and MAPK activity was measured using Western blot techniques with the phospho-p44/42 MAPK antibody (21, 22). Le Couter et al. (18) showed that MAPK activation by Bv8 stimulated mitogenic activity in endothelial cells. Melchiorri et al. (23) were able to demonstrate that Bv8 induced MAPK activity in cerebellar granule cells, and this action attenuated apoptotic cell death. It is interesting that PKR2 studies show activation of Gq and phosphoinositide pathways (21), although influences of PK2 on PKA, adenylate cyclase, phospholipase A, phospholipase C, and tyrosine kinase activity have not been reported (20, 21, 25). However, in one study, Bv8 activated PKC-ε in a subpopulation of isolated DRG neurons to sensitize transient receptor potential vanilloid type 1 channels (31). Our present observations suggest that MAPK is an important signaling intermediate in the regulation of channel function in the SFO, a hypothesis that was confirmed here by the abrogation of PK2-induced changes in membrane potential and [Ca2+]i by the MAPK pathway inhibitor PD-98059.
Numerous studies have established that PK2, or its ortholog Bv8, has the capacity to stimulate increases in [Ca2+]i (3, 20–22, 29, 31). While Vellani et al. (31) demonstrate that potentiation of the transient receptor potential vanilloid type 1 current may be responsible for PK2-stimulated increases in [Ca2+]i in some DRG neurons, Li et al. (20) provided some evidence that L-type Ca2+ channels could be involved in PK2-mediated Ca2+ changes, while others have shown no link between L-type Ca2+ channels and increases in [Ca2+]i (25). The PKR2 are of the Gq subtype and are capable of increasing intracellular inositol phosphate (3, 21, 22); therefore, PK2 interaction with its receptor also has the capacity to mobilize Ca2+ release from intracellular stores. We demonstrated that, in SFO neurons, PK2-mediated increases in [Ca2+]i required external sources of Ca2+, and this increase required participation of L-type Ca2+ channels. Despite previous evidence suggesting that the PKR2 activated inositol phosphate pathways (3, 21, 22), our TG experiments demonstrated that Ca2+ release from intracellular stores did not contribute to the Ca2+ transients observed in SFO neurons. The differing mechanisms of [Ca2+]i increase demonstrated in these studies suggest that there may be cell-specific regulation of PK2 signaling.
The observations that PK2 did not directly regulate voltage-gated Ca2+ currents, and that nifedipine blocked Ca2+ transients in imaging experiments, suggested that activation of a MAPK cascade by PK2 was affecting other membrane properties, thus resulting in voltage regulation of Ca2+ channels. Imaging experiments with TTX confirmed that Na+ channels were required for PK2-mediated [Ca2+]i changes, and current-clamp experiments demonstrated that the MAPK inhibitor was sufficient to prevent PK2-induced membrane depolarization. The mechanism contributing to these observations appears to be that MAPK activation by PK2 increases INaT and INaP and decreases IK, which, together, results in an increased membrane depolarization and action potential firing frequency. This observation is not unique, for MAPK has been shown in several neuronal systems to modulate membrane excitability by decreasing K+ currents (34, 35).
The prior observation that PK2 plays a dipsogenic role when injected into the SFO is intriguing in that it suggests that, similar to ANG II and vasopressin (VP), PK2 may be a mediator of cardiovascular function (24). ANG II has been shown to depolarize SFO neurons through the inhibition of inactivating K+ current, IA, and the potentiation of N-type Ca2+ currents and NSCC (10, 26, 32). VP also has depolarizing effects on SFO neurons, which may, at least in part, be the result of inhibition of IA (33). Both of these peptides can influence drinking behavior and cardiovascular function through inputs to the paraventricular nucleus of the hypothalamus, supraoptic nucleus, and sympathetic neurons. Similar to the actions of ANG II and VP, PK2 has excitatory effects on SFO neurons, although the effects are via potentiation of INa and inhibition of IK. Interestingly, the failure of ANG II and VP antagonists to inhibit the PK2-mediated dipsogenic response could imply that circadian PK2 modulates a differing neuronal pathway for stimulating thirst and affecting cardiovascular outcomes (24).
The in vivo source of PK2 influence on the SFO is not clear; however, it could potentially arise from multiple locations, including circulating PK2, inputs from the NTS and median preoptic nucleus of the hypothalamus, and humoral release of PK2 into the CSF by the SCN. Thirst-provoking pathways influencing drinking behavior through the SFO have already been established as arising from blood-borne signaling and NTS inputs; however, the circadian differences in PK2-stimulated drinking behavior described by Negri et al. (24) do not rule out a possible influence by the SCN. Intriguingly, Krout et al. (17) identified projections from the SFO to the SCN, suggesting the possibility of a feedback pathway.
Perspectives and significance.
The integrative role of the SFO in regulating autonomic function suggests several possible outcomes for PK2 action in the SFO. Our laboratory's previous observations of PK2-mediated depolarization of SFO neurons (6) are complemented here by the demonstration that this enhanced excitability is MAPK dependent and causes increases in [Ca2+]i that do not rely on mobilization of intracellular Ca2+ stores and are abolished by TTX, suggesting a dependence on the activation of TTX-sensitive voltage-gated sodium channels. These observations led us to experiments in which we have identified MAPK-mediated effects of PK2 in the modulation of both INaT and INaP, effects that contribute to the integrated cellular response to the activation of SFO neurons by this circadian peptide.
This work was supported by a Target Obesity post-doctoral fellowship from Heart and Stroke Foundation of Canada/Canadian Institutes of Health Research/Canadian Diabetes Association (M. Fry), a post-doctoral fellowship from Heart and Stroke Foundation of Canada (G. T. Cottrell), and grants from the Canadian Institutes for Health Research (A. V. Ferguson).
We thank Dr. Qun-Yong Zhou for providing the PK2 peptide and the PKR2 antagonist A1MPK1.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society