The thalamic mGluR1-PLC?4 pathway is critical in sleep architecture
The results of our study demonstrated that slow intrathalamic oscillatory activity was significantly enhanced in brain slices from PLC?4?/? mice (Fig. 5). Within the intrathalamic circuit consisting of reciprocal connections between the TC and TRN nuclei, PLC?4 is almost exclusively expressed in TC neurons with tight linkage with the mGluR1, whereas no expression is found in the TRN [24, 25]. Therefore, these results suggested that the enhanced intrathalamic oscillations in the PLC?4?/? slices were caused by the deletion of PLC?4 in the TC neurons. The intrathalamic oscillations that were induced by a single electrical stimulus to the in vitro IC have often been examined in order to better understand the mechanisms underlying the sleep rhythms or spike wave discharges that are generated in the thalamocortical circuit [29–31]. Thus, the findings of enhanced intrathalamic oscillations in the PLC?4?/? thalamic slices were in good agreement with the findings of significant increases in the power density of the ? waves during NREM sleep in PLC?4?/? mice. The essential role of thalamic PLC?4 in control of brain rhythms during NREM sleep was further supported by the TC-restricted PLC?4 knockdown data (Fig. 6). The process underlying the generation of ? waves is unclear [4, 18, 34–37]. A previous study has shown that the generation of ? waves is thalamic-dependent when TC neurons reach a certain level of hyperpolarization [34], which has been questioned due to the view that cortical neurons pace the ? waves because the waves are still observed after large lesions in the thalamus [18]. Our data from both PLC?4?/? and TC-restricted PLC?4 knockdown mice support the view that ? waves are regulated by the thalamic circuit [34]. Taken together, these results suggested that impairments of the mGluR1-PLC?4 pathway in TC neurons enhanced the slow-frequency thalamocortical oscillations and ? power during NREM sleep.
The ? wave appears mainly in the deep-sleep stages during NREM sleep, which is identical to sleep stages 3 and 4 in the human [35]. Concomitant increases in ? power and the duration of NREM sleep have been observed in many studies under sleep debt conditions after sleep deprivation [38, 39]. In contrast, other studies have indicated that the EEG ? power is regulated independently of NREM sleep amount [40]. In this study, we observed enhanced ? power during NREM sleep in parallel with increased NREM sleep amount in PLC?4?/? mice. During the light phase, prolonged NREM sleep episodes in the PLC?4?/? mice was observed with reductions in the NREM to REM sleep transition (Fig. 3). During the dark phase, PLC?4?/? mice showed an increased transition from the awake state to the NREM sleep state resulting in the appearance of short NREM episodes accompanying concomitant disappearance of long lasting awake episodes. Nevertheless, the number of REM sleep episodes in the dark phase was not increased because the transition from NREM to REM sleep was decreased. These results together indicated that, once the mice entered NREM sleep, the transition to progress to the REM sleep state was attenuated in the PLC?4?/? mice. The physiological significance of the reductions in the shift between the NREM-REM sleep states is not yet understood.
In this study, we also observed longer REM sleep episodes in the PLC?4?/? mice compared to the PLC?4+/+ mice (Fig. 3f, l). Some studies have suggested that, after REM sleep deprivation, the time spent in REM sleep is extended during recovery in order to produce a rebound effect [41]. Therefore, the prolonged REM sleep episodes might be homeostatically regulated by abnormally maintained NREM sleep episodes and the reductions in the attempts to enter REM sleep. However, contributions to these results by other brain regions, including the mesopontine nuclei and hypothalamic nuclei [42], that regulate REM sleep cannot be completely excluded because these brain regions express low levels of PLC?4 and send inputs to the thalamus [43]. Therefore, further studies are needed to fully understand the observed changes in the REM sleep states.
Since the brain stem circuitry was first described as an ‘ascending reticular activating system’ that sends inputs to the thalamus and other brain regions [42, 44], many studies have reported that the brain stem circuitry regulates the transition between NREM and REM sleep by turning the REM sleep state on and off [2, 42]. Recently, many studies have focused on the role of these ascending pathways, including the brain stem [1, 45], basal forebrain that receives inputs from the brain stem [5], and hypothalamus [46], in the control of sleep state switching. To date, the function of L6 feedback is far from clear. However, it has been implicated in the shaping of receptive fields by the selective attention, initiation, and termination of thalamocortical oscillations [47], and it acts as a classical modulator [48, 49]. Much indirect in vitro evidence of this modulatory function of L6 inputs is in contrast with the driver functions of L5 inputs in higher-order thalamic nuclei [50, 51]. Definite proof would be changes that are elicited in the thalamocortical network state. Therefore, the current study is novel because it examined the role of the mGluR1- PLC?4 pathway that is specific to L6 inputs in sleep architecture through modulations of thalamic oscillations. The loss of this major excitatory input pathway in a top-down control circuit to thalamus synapses dramatically changed the sleep architecture.
In summary, we found that the deletion of PLC?4, which is specifically expressed postsynaptically to L6 corticothalamic inputs, attenuated the transition from NREM to REM sleep state in PLC?4?/? mice, which subsequently increased the total amount of NREM sleep and enhanced the ?-frequency power in the EEGs. These results, combined with TC restricted-PLC?4 knockdown data, demonstrated that the corticothalamic input to TC neurons through the mGluR1-PLC?4 pathway was critical for sleep architecture and the generation of sleep rhythms.