ポスター
D. 恒常性と神経内分泌システム
D. Homeostatic and Neuroendocrine Systems |
|---|
| 2022年6月30日 13:00～14:00 沖縄コンベンションセンター 展示棟 ポスター会場1
1P-044
グレリン脳室内投与における作用の関係性「摂食、摂水、睡眠・覚醒」
Relationship between effects of intracerebroventricular administration of ghrelin “food- and water-intake, sleep-wakefulness”
*沼田 侑也(1)、四方 裕貴(1)、興津 雄大(1)、前田 一樹(1)、金 主賢(1)
1. 富山大学
*Yuya Numada(1), Yuki Shikata(1), Yudai Okitsu(1), Itsuki Maeda(1), Juhyon Kim(1)
1. University of Toyama
Keyword: Ghrelin, sleep-wakefulness, food intake, water intake
Ghrelin, known as an endogenous ligand for growth hormone secretagogue receptor, is produced in the peripheral organ and some regions of brain. Ghrelin plays significant role in the regulation of growth hormone secretion, feeding behavior and sleep-wakefulness(WAKE). Although feeding behavior and vigilance state are closely related, there are few reports that examined effect of ghrelin on these behaviors simultaneously. Furthermore, previous studies have reported conflicting results on water-intake by ghrelin administration. Therefore, present study was carried out to examine effect of ICV administration of ghrelin on food-, water-intake in parallel with sleep-WAKE using rats with electroencephalogram (EEG) and electromyogram (EMG) monitoring. ICV administration of ghrelin in wide dose range increased both amount of food- and water-intake in accordance with time spent in WAKE, and reduced both non-rapid eye movement sleep (NREMS) and REMS. Furthermore, amount of water-intake was significantly decreased by food deprivation after ghrelin administration. Present results indicated that several dose of ghrelin increased arousal time and food-, water- intake with positive correlation between them. It is also suggested that increase of water-intake by ghrelin administration might be secondary effect due to increase of feeding behavior evoked by ghrelin. | | 2022年6月30日 13:00～14:00 沖縄コンベンションセンター 展示棟 ポスター会場1
1P-045
覚醒状態調節における脳幹部位の機能の解明
Elucidation of the functions of brainstem regions in the vigilance state regulation
*成瀬 文乃(1,2)、宮竹 功一(2,3)、本城 咲季子(2)
1. 筑波大学 医学群医療科学類、2. 筑波大学 国際統合睡眠医科学研究機構、3. 日本学術振興会特別研究員
*Ayano Naruse(1,2), Koichi Miyatake(2,3), Sakiko Honjoh(2)
1. Medical Sciences, Medicine, Univ of Tsukuba, 2. International institute for integrative sleep medicine, 3. Japan society for the promotion of science research fellow
Keyword: SLEEP, BRAINSTEM, WAKEFULNESS
Sleep is essential for any organism with a nervous system. However, the neural mechanisms underlying intrinsic transition between sleep/wake stages remain largely unknown. In this study, we focused on the roles of the deep mesencephalic nucleus (DpMe) in the sleep/wake regulation. DpMe is a midbrain region surrounded by some nuclei, such as the superior colliculus, the substantia nigra compacta, the periaqueductal gray, and the medial geniculate body. Although DpMe makes up a large portion of the midbrain, its physiological functions are still elusive. Previous studies reported that microinjection of anesthetics into the midbrain causes loss of consciousness in rats, suggesting the essential roles of this region in the maintenance of wakefulness. Therefore, we focused on neural populations responsible for maintenance of wakefulness thorough our research. To manipulate neural activity in this region, we employed chemogenetics. We injected AAV-hSyn-DIO-hM3D (Gq) into various midbrain areas including DpMe in Vglut2-Cre transgenic mice, recorded 24h of baseline sleep/wake cycles, and performed chemogenetic activation to see its effects on sleep/wake patterns. We found that chemogenetic activation of glutamatergic neurons in the midbrain resulted in prolonged wakefulness beyond physiological ranges. The results suggested that glutamatergic neurons in the midbrain have a responsibility to maintain the wakefulness in mice. Next, we compared the effects of midbrain activation to those of the cortex. The same amount of AAV-hSyn-DIO-hM3D (Gq) was injected and expressed in the glutamatergic the midbrain or in the cortex. We performed chemogenetic activation and our analyses revealed that activation of the midbrain promoted arousal much more powerfully than cortex does. Taken together, these results suggest that the midbrain DpMe area is one of the most powerful wake-promoting centers. | | 2022年6月30日 13:00～14:00 沖縄コンベンションセンター 展示棟 ポスター会場1
1P-046
再帰型深層学習を用いた頭部固定マウスの瞳孔動態による睡眠段階予測
Recurrent deep learning-based sleep stage estimation by pupil dynamics of a head-fixed mouse
*小林 剛(1)、田中 謙二(1)、高田 則雄(1)
1. 慶應義塾大学医学部
*Goh Kobayashi(1), Kenji F. Tanaka(1), Norio Takata(1)
1. Keio University School of Medicine
Keyword: LSTM, pupil tracking, sleep stage estimation, mouse
The standard procedure for sleep stage classification is thresholding EEG and EMG amplitudes, followed by an expert correction. While the method is widely accepted and routinely used, there are at least two limitations: 1) abnormal EEG or EMG activities of a disease model mouse make it difficult to classify sleep stages with the method, and 2) it is hard to predict sleep stages, e.g., 5 s in the future. Thus, we investigated a method for sleep stage classification without EEG and EMG. We focused on pupil dynamics since the tight correlation between pupil dynamics and cortical activity has been reported recently. Yüzgeç et al. estimated sleep stages from pupil diameter in head-fixed mice using a three-layer perceptron (Curr Biol. 2018). The estimation, however, was only applicable to a continuous single sleep-stage lasting more than 100 s. In this study, we propose a deep learning model of a long short-term memory (LSTM), which is suited for time series data, to estimate and predict sleep stages using pupil dynamics of a shorter period.
After surgery for EEG and EMG recording and habituation for a head-fixed condition using adult mice (n = 4), measurement of EEG, EMG, and left eye pupil dynamics were performed for 6 hours a day for 3–15 days. Pupil dynamics were acquired with a USB camera under infrared light. We used a deep learning toolbox, DeepLabCut, to acquire time-series data of pupil dynamics. Regarding input features for training the LSTM model, we examined pupil diameter (PD), eyelid opening (EO), eye-movement velocity (EV), and pupil location (PL). The performance of the model was assessed with various input time durations of pupil dynamics for 10–45 seconds. We also examined the performance of an LSTM model for sleep stage prediction 1–45 seconds after pupil dynamics.
The multi-class classification metric, micro f1 score, of the sleep stage estimation immediately after the 10-s pupil dynamics was 0.74 ± 0.07 in the test data. The accuracy for each sleep stage was WAKE, 92 ± 2%; NREM, 86 ± 8%; and REM, 64 ± 6%. We are now working on a sleep stage prediction using the model.
Using the LSTM network, we succeeded to classify sleep stages without EEG and EMG. The same level of estimation accuracy as in the previous study using 100-s pupil dynamics was achieved with just a 10-s duration, which was the same time resolution of our supervisory signal of EEG and EMG. | | 2022年6月30日 13:00～14:00 沖縄コンベンションセンター 展示棟 ポスター会場1
1P-047
時計神経dorsal lateral neuronの睡眠制御に対する機能解析
Functional analysis of dorsal lateral neuron on sleep regulation in Drosophila melanogaster
*山本 洵(1)、小林 里帆(1)、冨田 淳(1)、粂 和彦(1)
1. 名古屋市立大学
*Jun Yamamoto(1), Riho Kobayashi(1), Jun Tomita(1), Kazuhiko Kume(1)
1. Nagoya City University
Keyword: clock neuron, sleep , circadian rhythm, Drosophila melanogaster
In Drosophila, sleep studies have revealed various molecular and neural circuits involved in sleep-wake regulation. Most of these studies have been based on "the two-process model of sleep" (Borbély et al., 1982), which states that the sleep pressure consists of two independent components: circadian rhythm (Process C) and sleep homeostasis mechanism (Process S). Process C has been focused on the clock neurons expressing the clock gene period, while Process S has been focused on regions outside of the clock neurons, such as the central complex which is a higher brain structure of flies. On the other hand, there have been several studies on sleep regulation by clock neurons in recent years, and it is necessary to reconsider the functions of both regions in the two-process model.
The dorsal lateral neuron (LNd), which is a clock neuron cluster, consists of six unilateral neurons. The cluster is thought to be responsible for locomotor activity at night under LD conditions and is known as evening cells. It was also shown that LNd exhibited a peak of Ca2+ at dusk using Ca2+ imaging in vivo (Liang et al., 2016). However, it remains unclear whether this neural activity of LNd regulates sleep.
Here we found that activation and inhibition of LNd changed the amount of sleep both acutely and chronically. Furthermore, knockdown of Nmdar1, which we reported to be involved in sleep regulation (Tomita et al., 2014), decreased the amount of sleep. Unexpectedly, flies with knockdown of Nmdar1 in LNd didn’t show any difference in circadian locomotor activity. We also confirmed the same effect with some other GAL4 lines especially expressed in LNd region. These results suggest that neural activity mediated by NMDA receptor in LNd is important for sleep regulation. | |
|
|
