In this study, we used whole cell patch-clamp techniques to characterize GABAA-mediated sIPSCs from neurons in both the dorsal and ventral subdivisions of the SCN. The sIPSC frequency was consistently higher in the dorsal subdivision compared with the ventral subdivision, although every neuron exhibited sIPSCs. The sIPSC frequency measured in the dSCN varied significantly with the daily cycle and exhibited a peak during ZT 11–15, spanning across late day and early night. In contrast, sIPSC amplitude, rise time, or decay time did not vary between SCN subdivisions or with the daily cycle. The rhythm in sIPSCs was dependent on synaptic activity because it was not seen in the presence of a blocker of voltage-dependent sodium channels (TTX). We used various manipulations of the VIP signaling system and its associated secondary messenger ensemble cAMP/PKA to conclusively show that the rhythm in GABAergic activity was dependent on endogenous VIP release. To determine if the day-night difference in sIPSC frequency was circadian, we examined GABA-mediated sIPSCs in mice housed in DD. Under these conditions, we were unable to detect a significant peak in sIPSC frequency during the early subjective night. However, the frequency of GABA sIPSCs during the late subjective night (CT 20–22) was significantly reduced compared with the other times (CT 4–8 and CT 13–15), indicating that a rhythm in GABA sIPSCs persisted under DD conditions. Additionally, day-night differences in GABA sIPSCs were absent in VIP/PHI-deficient mice housed under both LD and DD cycles.
SCN neurons are well known to exhibit a daily rhythm in electrical activity and membrane conductance with peak electrical activity occurring during the subjective day (Schaap et al. 2003
). The neurotransmitter GABA is frequently thought of as an inhibitory neurotransmitter in the adult mammalian nervous system, activating chloride-permeable GABAA
ion channels that serve to block action potential generation by either hyper-polarizing the membrane or acting as a transmembrane shunt to prevent rapid depolarization. Our observation that all SCN neurons in this study were under tonic GABAergic influence would suggest that this transmitter plays a significant role in regulating spontaneous firing and excitability during all phases of the circadian cycle. Recently, several groups have suggested that GABA may have an excitatory role in the adult SCN (de Jeu and Pennartz 2002
; Wagner et al. 1997
). One possible explanation is that the chloride equilibrium potential shifts with the 24-h cycle and with development in SCN neurons (Shimura et al. 2002
). This issue is as yet unsettled, with one group reporting that GABA is excitatory only during the subjective day (Wagner et al. 1997
), another concluding that GABA is excitatory only during the subjective night (de Jeu and Pennartz 2002
), and yet a third group stating that GABA is inhibitory regardless of the phase (Gribkoff et al. 1999
). In any case, the daily rhythm in GABAergic tone observed in the dSCN undoubtedly impacts both spontaneous electrical activity as well as how these cells respond to excitatory synaptic inputs.
Although the preparation used in this study does not allow us to identify the source of GABAergic input while recording, previous anatomical and electrophysiological studies have described various sources of GABAergic inputs onto SCN neurons. Several electrophysiological studies have observed that SCN neurons receive a tonic level of GABAA
-mediated postsynaptic currents, while focal stimulation in the vicinity of recorded neurons reveal that nearly all SCN neurons receive local or extranuclear GABAergic inputs (de Jeu and Pennartz 2002
; Jiang et al. 1997
; Kim and Dudek 1992
; Strecker et al. 1997
). Other proposed sources of GABAergic input include retinal ganglion cells directly innervating the SCN (Jiao and Rusak 2003
), the contralateral SCN (Buijs et al. 1994
), arcuate nucleus, supraoptic nucleus, and intergeniculate leaflet (Morin and Blanchard 2001
; Saeb-Parsy et al. 2000
). The observed peak in GABA currents during the early night is TTX-dependent, indicating that neuronal activity is driving the circadian rhythm in inhibitory transmission in the dSCN. The fact that the day-night difference persists in an acute slice preparation suggests that GABAergic input to the dSCN is mediated either by active GABAergic neurons within the SCN or at least within the brain slice preparation.
It has been proposed that electrical signals generated by core SCN neurons (vSCN), which are adjusted to the LD cycle by retinal inputs, are transmitted to the shell region of the SCN (dSCN) through monosynaptic connections (van Esseveldt et al. 2000
). Output from the shell region is synchronized by virtue of direct connectivity to the core region and modified by cortical inputs. However, if sIPSCs recorded from dSCN neurons are driven by neuronal activity within the SCN, a seemingly contradictory situation arises in that the highest level of activity-driven GABAergic transmission in the dSCN occurs during the early night, a time period when the spontaneous firing rate of SCN neurons is significantly reduced (Schaap et al. 2003
). However, recent work suggests that the SCN is a heterogeneous cell population (Karatsoreos et al. 2004
) and that some cells exhibit peak activity during the night. A recent study using time-lapse imaging of a green fluorescent protein (GFP) reporter of the clock gene Period 1
) found a positive linear correlation between neuronal spike frequency and Per1
transcription, indicating that Per1
rhythms are representative of physiological activity (Quintero et al. 2003
). Importantly, it was also noted that a small proportion of the sampled neurons cycled in antiphase to the principal phase peak, raising the possibility that a subgroup of SCN neurons acts as inhibitory interneurons within the SCN. Thus we speculate that the dSCN neurons are receiving rhythmic inhibitory input from a distinct population of SCN interneurons that are electrically silent during the day, but highly active during the early night.
We have previously demonstrated that VIP is a potent regulator of sIPSC frequency in SCN neurons (Itri and Colwell 2003
), and in this study, sought to determine the role of the endogenous peptide in driving a circadian rhythm in inhibitory synaptic transmission. An important difference between this study and the previous study (Itri and Colwell 2003
) describing the effects of VIP on GABA-mediated sIPSCs is that, previously, VIP was bath-applied to the slice to demonstrate that VIP is capable of enhancing GABA release in the SCN. In this study, we manipulated the endogenous VIP signal to determine if we could prevent the peak in GABA-mediated sIPSC frequency, providing a physiologically relevant measure of VIPs effect in the SCN. We found that pretreatment of SCN slices with a VPAC1
antagonist prevented the increase of IPSC frequency during the early night in the dSCN, as did pretreatment with the VPAC2
-specific receptor antagonist PG 99-465. Our interpretation is that VIP release is initiated in anticipation of the transition from light to dark (ZT 11-12) and continues to be released during the early night (ZT 13-15), signaling the onset of darkness by presynaptically enhancing GABA release onto dSCN neurons. This is supported by the observation that VIP content in the SCN decreases monotonically in animals maintained in illumination and that light pulses given during the night are more effective at suppressing VIP levels than during the day (Shinohara et al. 1999
). By blocking the receptors for VIP (VPAC2
receptors), we effectively prevent the endogenous VIP signal from reaching GABAergic terminals synapsing on dSCN neurons. Since the activated VPAC2
receptor utilizes the cAMP/PKA second messenger system to enhance GABA release (Itri and Colwell 2003
), we concluded that blocking these intermediates would also prevent the peak in GABA sIPSCs observed during the early night. Pretreatment of SCN slices during the early night with the PKA inhibitor H-89 completely blocked the nocturnal increase in sIPSC frequency. Similar experiments performed in transgenic VIP/PHI-deficient animals housed under an LD cycle revealed that, in the absence of endogenous VIP, the daily rhythm in inhibitory synaptic transmission was nonexistent. Thus using a variety of manipulations, our data strongly suggest that the peak in GABA frequency during the early night in an LD cycle is dependent on the VIP signaling system and associated cAMP/PKA second messenger pathway.
Temporal profiles of VIP release reveal a circadian rhythm that is maintained in organotypic SCN slice culture, suggesting that rhythmic VIP release may drive a circadian rhythm in GABAergic sIPSCs in the dSCN under constant conditions (Shinohara et al. 1994
). We examined sIPSCs in control mice housed in DD, and under these conditions, we were unable to detect a significant peak in sIPSC frequency during the early subjective night. The frequency of GABA sIPSCs between subjective day (CT 4–8) and early subjective night (CT 13–15) were nearly indistinguishable. However, we found that the frequency of GABA sIPSCs was significantly reduced during the late subjective night (CT 20–22) compared with the rest of the cycle in DD, indicating that the rhythm in GABA sIPSCs was sustained in DD but different from what was predicted from the data observed in a LD cycle. The correlation between rhythmic VIP release and rhythmic sIPSC frequency suggests that changes in VIP release underlie the differences in sIPSC frequency observed during the late subjective night compared with subjective day and early subjective night. Based on these observations, we believe that VIP plays a critical role in driving the daily rhythm in GABA frequency under both LD and DD conditions.
It has been suggested that GABA can modulate light- and N
-aspartate (NMDA)-induced phase-shifts by direct regulation of excitability in SCN neurons (Gillespie et al. 1997
; Mintz et al. 2002
). By modulating the frequency of GABA release in the SCN, VIP may be an important mediator in the neuronal pathway responsible for light-induced phase-shifting in mammals. Indeed, behavioral experiments performed in heterozygous and homozygous VIP/PHI-deficient mice indicate that both maintenance of a daily rhythm in wheel-running activity and phase-shifting of this daily rhythm are significantly affected in animals with a reduction or absence of endogenous VIP (Colwell et al. 2003
). VIP has a fundamental functional role in biological clock function (Hannibal and Fahrenkrug 2003
; Harmar 2003
), which may manifest through regulation of GABA release in the dorsal cell population. It has also been shown that daily treatments with GABA are sufficient to synchronize electrical activity in SCN cell populations in culture (Liu and Reppert 2000
; Shirakawa et al. 2000
). In fact, is has been surmised that circadian fluctuations in GABA release within the SCN may be important for synchronizing clock cells with widely different phases and period lengths (Liu and Reppert 2000
). In this study, we have identified a circadian rhythm in GABAergic tone in the dorsal SCN that persists in DD and is driven by the endogenous VIP/VPAC2