|Home | About | Journals | Submit | Contact Us | Français|
The mammalian circadian clock in the suprachiasmatic nucleus (SCN) drives and maintains 24-h physiological rhythms, the phases of which are set by the local environmental light-dark cycle. Gastrin releasing peptide (GRP) communicates photic phase setting signals in the SCN by increasing neurophysiological activity of SCN neurons. Here, the ionic basis for persistent GRP-induced changes in neuronal activity was investigated in SCN slice cultures from Per1::GFP reporter mice during the early night. Recordings from Per1-fluorescent neurons in SCN slices several hours after GRP treatment revealed a significantly greater action potential frequency, a significant increase in voltage-activated outward current at depolarized potentials, and a significant increase in 4-aminopyridine (4-AP) sensitive fast delayed rectifier (fDR) potassium currents when compared to vehicle-treated slices. In addition, the persistent increase in spike rate following early night GRP application was blocked in SCN neurons from mice deficient in Kv3 channel proteins. Because fDR currents are regulated by the clock and are elevated in amplitude during the day, the present results support the model that GRP delays the phase of the clock during the early night by prolonging day-like membrane properties of SCN cells. Furthermore, these findings implicate fDR currents in the ionic basis for GRP-mediated entrainment of the primary mammalian circadian pacemaker.
In mammals, daily rhythms in physiology and sleep/wake patterns are driven by molecular clockworks in individual cells that form a network and functional oscillator in the master pacemaker located in the suprachiasmatic nucleus (SCN) of the hypothalamus. While self-sustained oscillators have now been localized in many peripheral tissues, the SCN orchestrates the overall timing of the peripheral oscillators (for review, see Reppert and Weaver 2002). Within the SCN, rhythms in the key gene components of the molecular clock, such as Period1 (Per1) and Period2 (Per2), as well as the spontaneous spike rate of individual neurons, persist ex vivo for many cycles with minimal damping (Kuhlman et al. 2000; Prosser and Gillette 1989; Reppert and Weaver 2001; Yoo et al. 2004). Rhythms of Per1/Per2 expression and spike frequency peak during the day and reach a nadir during the night. In addition, the period of these rhythms are generally close to, but not exactly, 24 h.
It is important that the master clock remains synchronized with the periodicity of the external environment that is 24-h. In fact, misalignment of the clock with the light-dark cycle can lead to compromised human health as evidenced by an increased risk of heart disease, cancer, and gastrointestinal disorders in shift workers (for review, see Foster and Wulff 2005). Entrainment, or synchronization of phase to the environment, occurs primarily through exposure to light such that light given in the early night delays the molecular clock rhythm while light given in the late night advances the molecular clock rhythm (Daan and Pittendrigh 1976; Yan and Silver 2002, 2004). Melanopsin-containing ganglion cells in the retina that are intrinsically photoreceptive transduce this photic signal and transmit it to ventral SCN cells, many of which express gastrin releasing peptide or GRP (Abrahamson and Moore 2001; Berson et al. 2002). In fact, application of GRP to the SCN region mimics light by upregulating Per and inducing phase delays and advances (Aida et al. 2002; Antle et al. 2005; Gamble et al. 2007; Kallingal and Mintz 2006, 2007; McArthur et al. 2000; Piggins et al. 1995).
In addition to upregulation of clock gene expression, photic stimulation also increases neuronal activity for several hours when presented in either the early or late phases of the night when the clock-driven spike rate is declining or rising, respectively (Kuhlman et al. 2003). Acutely, brief pulses of GRP primarily results in excitation in over half of the SCN single-units examined during the late day or early night (Piggins et al. 1994; Tang and Pan 1993), and daytime application of GRP results in depolarization and decreased basal potassium conductance (Reynolds and Pinnock 1997). Interestingly, increased excitability induced by both a phase-resetting light pulse and GRP application during the late night continues to persist even several hours later in Per1-induced neurons (Gamble et al. 2007; Kuhlman et al. 2003). For late-night light pulses, this persistent increase in electrical activity is accompanied by a depolarized membrane potential and an increase in input resistance that is mediated by suppression of K+ currents (Kuhlman et al. 2003). However, during the early night, the underlying mechanism for the GRP-induced phase shifts and persistent increase in neuronal activity remains unknown. Therefore, the current study sought to determine whether early night application of GRP induces a persistent increase in SCN spike rate and sought to identify the ionic basis for long-term, GRP-induced changes in neuronal activity.
Adult, mice (16–30 days of age) homozygous for the mPer1::d2EGFP transgene on a B6C3H hybrid background (Kuhlman et al. 2000), wild type C57BL/6 (colony at University of California, Los Angeles), and Kv3.1–Kv3.2 double knock-out mice (KVDKO mice; also known as Kcnc1−/−/Kcnc2 −/− mice) (Ozaita et al. 2004) on a ICR background (generously provided by Bernardo Rudy, New York University) were housed in a 12:12 light/dark (LD) cycle for a minimum of two weeks before the start of the experiment and had food and water provided ad libitum. Experiments were performed in accordance with the Vanderbilt University and University of California, Los Angeles (UCLA) Institutional Animal Care and Use Committee and National Institutes of Health guidelines.
For electrophysiology experiments, SCN slice cultures from Per1::GFP mice were obtained at zeitgeber time 11 (ZT 11; where ZT 12 refers to lights off) using the same methods as in (Gamble et al. 2007). At projected ZT 15, an 8-μl volume of 37.5 μM GRP (porcine, from Biotrend, Destin, FL; final concentration: 0.3 μM) or vehicle (sterile culture water) was added directly to the culture dish medium. One hour later, at projected ZT 16, slices were transferred to an open recording chamber (Warner Instruments, Hambden, CT) that was continuously perfused at a rate of 2.0 ml/min with extracellular solution bubbled with 5% CO2/95% O2 and heated to 34 ± 0.5 oC (as in Kuhlman et al. 2003) or maintained at room temperature (for KVDKO experiments). All drugs were purchased from Sigma (St Louis, MO) unless stated otherwise.
Neurons were visualized with a fixed-stage, upright microscope equipped for near-IR-DIC and epi-fluorescence. Recordings were made from SCN neurons located in the dorsal and medial regions of the SCN because the majority of BB2 receptors in the SCN lie within this region (Aida et al. 2002; Karatsoreos et al. 2006) as in Gamble et al. (2007). Recordings were made between projected ZT 17 and 19 (2–4 h after GRP application and 1–3 h after washout). Loose patch recordings were generated using the same electrode parameters and solutions as reported previously (Gamble et al. 2007). Firing rate was measured as the average of a 100-sec record. Electrophysiological signals were processed and controlled by an Axopatch 1D/200B or Multiclamp amplifier, and pClamp 8.0 or 10.0 software (Axon Instruments, Union City, CA) in gap-free mode. Recordings were sampled at 10 kHz and filtered at 1 kHz. Solutions for whole-cell conditions were the same as reported in (Kuhlman et al. 2003), and recordings were made within six minutes of membrane rupture. In current clamp mode, a 500-msec square hyperpolarizing current was injected to calculate input resistance (slope of the voltage responses to −5 to −25 pA of current injection), and the membrane potential was defined as the average voltage over 500 msec for non-spiking neurons and the midpoint of the interspike interval for spiking neurons. The reported voltages were corrected for a liquid junction potential of 8 mV. Current traces from voltage clamp recordings (as in Fig. 3) were obtained by using a 180-ms voltage step at progressively depolarized potentials (−80 to +55, 15 mV steps) and a holding potential of −65 mV. For fDR isolation, the conditions and solutions were the same as those reported in (Itri et al. 2005) and contained bicuculline (25 μM) to block GABAA-mediated currents, tetrodotoxin (TTX; 0.5 μM) to block fast voltage-activated sodium channels, 4-aminopyridine (4–AP, 0.5 mM to 5 mM) to block fDR currents, and cadmium (100 μM) to block calcium (Ca2+) channels. Current traces were obtained before and after a 3.5-min application of 4-AP by using a voltage protocol with a hyperpolarizing pre-pulse (100 ms at −90 mV) followed by 900 ms voltage pulses at progressively depolarized potentials (−80 to +40 mV, 10-mV steps). Activation curves were generated by plotting normalized current (I/Imax) over the membrane potential and fit using a charge-voltage Boltzmann function f(Vm) = Imax/(1+exp((Vmid − Vm)/VC)), where Vm is the membrane potential, Vmid is the membrane potential at 0.5 and VC is the slope factor.
Data were analyzed with independent samples t-tests and two-way, repeated measures ANOVA with Fisher’s Least Significant Difference (LSD) post hoc analysis. Significance was ascribed at p < .05.
During the late night, both phase resetting light pulses and GRP treatment increase the frequency of spontaneous action potentials in SCN neurons, even hours later, and this increase is correlated with and requires the induction of Per1 (Gamble et al. 2007; Kuhlman et al. 2003). Although early night GRP application has been shown to induce phase shifts and to increase Per1 and Per2 expression in the SCN (Antle et al. 2005), GRP-induced neuronal activity in the early night has not been demonstrated. Thus, we sought to investigate whether the longer-term effects of GRP associated with Per1 induction and phase resetting during the early night are similar to those observed in the late night. Specifically, we compared the frequency of spontaneous action potentials in SCN neurons 2–4 hours following either a pulse of GRP or of vehicle at projected ZT 15. Hypothalamic slices were prepared during lights on (ZT 11–12) and cultured (Han et al. 2006) until the following day in order to avoid possible phase shifting effects of slice preparation during the night (Gillette 1986). At projected ZT 15, 0.3 μM GRP was applied to the culture medium and washed out one hour later (projected ZT 16) as the slice was transferred to an open recording chamber that was continuously perfused with extracellular solution. One to three hours after the termination of GRP application (projected ZT 17–19) loose patch extracellular recordings were made from Per1-induced SCN neurons, identified by Per1::GFP fluorescence. The mean firing rate of Per1-fluorescent neurons in GRP-treated slices (n = 7) was more than triple the mean firing rate of Per1-fluorescent neurons from vehicle-treated slices (n = 9; mean ± SEM: 3.4 ± 0.8 Hz and 1.0 ± 0.5 Hz for GRP- and vehicle-treated neurons, respectively; t11 = 2.4, p < 0.05, two-tailed; Fig. 1A,B).
In order to characterize the membrane properties of GRP-treated, Per1::GFP-expressing SCN neurons using the same treatment paradigm above, whole-cell current-clamp recordings of GRP-treated, Per1::GFP-expressing neurons (n = 8) were compared with those made from vehicle-treated, Per1::GFP-expressing neurons (n = 9). As shown in Figure 2, there was no significant difference in membrane potential in the two conditions, t(15) = −0.13, p > 0.05 (in mV, mean ± SEM, GRP: −46.0 ± 2.5, vehicle: −45.5 ± 2.6) despite the elevated spike frequency in GRP-treated neurons described above. In order to compare the input resistance of these groups, the slope resistance from a current clamp protocol with 5-pA, square, hyperpolarizing steps was determined. No significant differences between GRP- and vehicle-treated conditions were found, t(12) = 0.98, p > .05 (in GΩ, mean ± SEM, GRP: 1.3 ± 0.3, vehicle: 1.0 ± 0.1).
Although the resting properties of SCN neurons were little altered by early-night application of GRP, we sought to measure any changes in spike-associated voltage-dependent currents that might contribute to the spike frequency increase induced by GRP during the early night by recording and comparing GRP- and vehicle-treated, Per1::GFP-expressing neurons under voltage-clamp conditions. GRP treatment significantly increased the overall steady-state outward current at depolarized membrane potentials ranging from +10mV to +55mV, as indicated by a significant interaction of voltage potential with treatment, F(9,117) = 6.3, p < .01 (see Fig. 3). There were no significant differences in the transient, peak outward current between GRP- and vehicle-treated cells, p > 0.05.
Circadian clock-controlled variation in spontaneous spike frequency can be attributed to both changes in membrane potential and modulation of voltage-dependent ionic currents that underlie generation and recovery from the action potential (for review, see (Kuhlman and McMahon 2006). In general, opening an outward, basal potassium current results in relative hyperpolarization of the neuron, and in fact, this type of current is thought to mediate circadian regulation of membrane potential (Kuhlman and McMahon 2004). In contrast to these potassium “leak” currents, other voltage-gated potassium currents such as the delayed rectifier are active at depolarized membrane potentials and contribute to the repolarization phase of the action potential, thereby decreasing the amount of time occurring between spikes and increasing the overall spike rate (Rudy et al. 1999). In fact, the fast delayed rectifier (fDR) channel has been identified as a current that contributes to circadian regulation in spike rate independent of circadian regulation of membrane potential (Itri et al. 2005). Elevated fDR currents are thought to mediate increased spike rate in SCN neurons by inducing a faster rate of membrane repolarization and a shorter duration of the action potential (Itri et al. 2005; Rudy et al. 1999).
In order to test whether GRP induces fDR currents in SCN neurons, we used the same treatment protocol as above; however, recordings were conducted in the presence of bicuculline (25 μM) to block GABAA-mediated currents, TTX (0.5 μM) to block fast voltage-activated sodium channels, and cadmium (100 μM) to block Ca2+ channels, as in (Itri et al. 2005). Whole-cell, voltage-clamp recordings were made before and after application of 4-AP (0.5 mM and 5 mM for 4 and 11 cells, respectively) in order to isolate fDR current amplitude (Icontrol − I4-AP) using a pulse protocol as indicated in Figure 4. The activation kinetics of isolated fDR currents were similar in GRP- and vehicle-treated cells, as indicated by the slope factor VC (GRP: 13.5 ± 0.9 mV; Veh: 10.0 ± 0.7 mV) and the midpoint potential Vmid (GRP: 13.3 ± 0.7 mV; Veh: 7.0 ± 1.1 mV), consistent with previously published fDR kinetics for SCN neurons (Itri et al. 2005). However, there was a significant interaction between voltage and treatment, such that GRP-treated neurons had significantly greater amplitudes of isolated fDR current as compared with vehicle controls at voltages ranging from +20 to +50 mV (F(13,169) = 3.8, p < .05; post hoc comparisons, p < .05). The I-V relationship was not significantly different between the two concentrations of 4-AP in either GRP- or vehicle-treated samples. In addition, the isolated fDR current (Fig. 4) appears to account for the majority (~60%) of the total voltage-activated, outward current induced by GRP (Fig. 3). For example, at +40 mV, Ivehicle-GRP for IfDR is ~60% of Ivehicle-GRP for Itotal.
The Kv3 family of channel proteins (Kv3.1 and Kv3.2) mediates fDR currents (Rudy et al. 1999). In order to examine whether Kv3.1/2 channel proteins are necessary for the persistent increase in action potential frequency induced by GRP application, we measured the spike rate of SCN neurons from Kv3.1/2 DKO mice following GRP application using the same experimental paradigm as before. As before, GRP significantly increased the spike rate of wild-type SCN neurons recorded 2–3 h after GRP washout (F(2,86) = 22.6, p < 0.05; post hoc comparisons, p < 0.05). Cell-attached loose patch recordings have the advantage of being less-invasive to the cell but limit analyses to spike rate only. The firing rates obtained from vehicle- and GRP-treated WT slices were very similar to those obtained in the Per1::GFP mice (compare Figure 5 to Figure 1), suggesting that the underlying membrane properties may be similar as well. In contrast to the average firing rate of SCN neurons from GRP-treated WT mice, the rate of those from Kv3.1/2 DKO mice was significantly lower (p < 0.05; Fig. 5) and the same as those recorded from wild-type, vehicle-treated slices (p > 0.05).
Circadian variation in spike rate is a well-established feature of the neurons that comprise the master circadian oscillator, with elevated spike rates observed during the day and lower spike rates during the night (Brown and Piggins 2007). During the night, when spontaneous spike frequency is low, phase resetting light pulses and light-mediating stimuli such as GRP significantly increase spiking activity and this effect can persist for several hours following light/GRP treatment (Gamble et al. 2007; Kuhlman et al. 2003). Here, we have shown that early night application of GRP has a persistent excitatory effect similar to what we have shown during the late night (Gamble et al. 2007), in that the spontaneous spike frequency of Per1::GFP neurons continues to be increased even several hours after GRP washout. During the late night, persistent excitatory activity induced by either photic stimulation or GRP application is accompanied by a depolarized membrane potential and increased input resistance (unpublished observations, K.L. Gamble and D.G. McMahon; Kuhlman et al. 2003). However, the underlying mechanism for early-night, light-induced persistent excitation has not been explored. Previous studies have observed that daytime application of GRP to SCN neurons induces an acute (min) depolarization in resting membrane potential that is accompanied by decreased basal potassium conductance (Reynolds and Pinnock 1997). In contrast, the present data revealed that GRP-treated cells in the early night phase showed no long-term (hours) change in membrane potential or input resistance at rest. It is important to note that in the present study, we are examining long-term and persistent effects of GRP which are quite different from the acute effects in the above-mentioned studies. Hours after GRP washout, we did observe an increased voltage-dependent, outward current at depolarized potentials as compared to vehicle-treated cells. In addition, the present study has shown that a large percentage of the GRP-induced current can be accounted for by a 4-AP-sensitive current with kinetics and voltage-dependence similar to previously isolated fDR currents in SCN neurons (Itri et al. 2005). Furthermore, GRP-induced persistent excitatory activity is lost in SCN from mice lacking fDR channel proteins, Kv3.1 and Kv3.2. These results show that fDR currents are involved in the GRP-mediated effects of the primary mammalian pacemaker in addition to mediating clock-regulated rhythms in spike frequency (Itri et al. 2005).
Circadian variation in action potential frequency is a distinguishing feature of SCN clock cells. Several voltage-dependent, ionic currents showing day-night variation in amplitude have been proposed as key modulators of this clock-controlled rhythm, and each contribute to a different aspect of action potential regeneration (for review, see (Kuhlman and McMahon 2006). For example, L-type calcium currents are greater in amplitude during the day than during the night and contribute to the depolarization phase of the action potential by allowing the membrane potential to oscillate at a high frequency near the spike threshold (Pennartz et al. 2002). On the other hand, fDR currents allow a faster repolarization during the falling phase of the action potential, and are also higher during the day than during the night (Itri et al. 2005). A recent study showed that A-type potassium currents in the SCN neurons peak during the day and may underlie circadian variation in firing patterns (Itri et al. 2010). However, this voltage-activated potassium current is unlikely to explain the present data because there was no significant difference in the peak transient current, indicative of an A-type potassium current, following GRP application. Finally, large conductance, calcium-activated potassium currents are higher during the night than during the day and lengthen the inter-spike interval by increasing the after-hyperpolarization at the end of each action potential (Meredith et al. 2006; Pitts et al. 2006). Of these ionic conductances, only increased fDR conductance is a “day”-like membrane property that allows a change in spike rate independently from spike threshold (Rudy et al. 1999) and could potentially explain a GRP-induced increase in spike rate in the absence of membrane potential depolarization. It is also clear that other un-characterized currents are involved because in our data, the 4-AP-sensitive current accounted for the majority of voltage-activated outward current induced by GRP, but did not account for all of it. Although GRP-induced long-term excitability and fDR current activation occur at the same phase as GRP-induced phase shifts, it is also possible that these processes are independently regulated.
One might expect that an increased K+ current would lead to a decrease in the action potential frequency (Kuhlman and McMahon 2004). However, fDR currents are mediated by the Kv3 family of channel proteins (Kv3.1 and Kv3.2) and are a special case of delayed-rectifier potassium channels in that they only activate at depolarized membrane potentials (> −10 mV), and they exhibit steep voltage-dependent activation with very little inactivation allowing for rapid deactivation (since inactivation does not first need to be removed) (Rudy et al. 1999). These properties permit fDR channels to activate at the peak of the action potential, inducing a faster rate of membrane repolarization and a shorter duration of the action potential, thereby decreasing the amount of time necessary for another action potential to be generated by the neuron. In addition, rapid deactivation during the after-hyperpolarization reduces any remaining current during the inter-spike interval. A 4-AP-sensitive current with these same properties has been characterized in the SCN, and furthermore, Kv3.1b and Kv3.2 protein levels show day-night differences in both a light-dark cycle and constant darkness, suggesting that fDR currents could at least partially account for circadian variation in spike rate (Itri et al. 2005).
We now show that these currents may be involved in the entrainment process of the primary mammalian pacemaker and can be induced by the photic-mediating neuropeptide GRP. It is interesting that the Kv3.1 and Kv3.2 genes are known to have cyclic AMP response element (CRE) sites (Gan and Kaczmarek 1998). Because activation of the CRE binding (CREB) pathway is necessary for GRP-induced phase delays (Gamble et al. 2007), it is possible that GRP-induced CREB upregulates Kv3.1 and Kv3.2 gene expression. Another possible mechanism for GRP-induced regulation of fDR currents is post-translational regulation. In the auditory neurons of the medial nucleus of the trapezoid body, increased fDR conductance is associated with dephosphorylation of Kv3.1b channels induced by protein kinase C (PKC) inhibitors (Song and Kaczmarek 2006; Song et al. 2005). Interestingly, PKC inhibition delivered in vivo has also been shown to enhance light-induced phase delays in behavior (Lee et al. 2007), suggesting that inhibition of PKC could result in dephosphorylation of Kv3.1b channels, enhancing fDR current, and potentiating the effect of light on circadian behavior.
In conclusion, night-time phase resetting stimuli such as light and GRP induce persistent increases in SCN neuronal activity during both early and late phases of the night (Gamble et al. 2007; Kuhlman et al. 2003). However, this effect is differentially mediated depending on the phase of the night. During the late night, light-induced neuronal excitation correlates with membrane potential depolarization and decreased basal potassium leak current (Kuhlman et al. 2003). The present study has shown that neuronal activity induced by GRP signaling during the early night is independent of membrane potential and is at least partially mediated by an increase in fDR, voltage-activated potassium conductance. These results provide a link between a circadian clock-regulated ionic conductance and neuropeptide signaling within the SCN night-time phase resetting pathway.
We would like to thank Chris Ciarleglio, Russell Johnson, Richard Penney, Anthony Baucum, Tongrong Zhou, and Dao-Qi Zhang for technical assistance. This work was supported by National Institutes of Health Grants, K99 GM086683 (KLG) and P50 MH078028 (DGM).