0446 A Novel System for Enabling High-Density EEG Recordings in a Mouse

Introduction

Recent advances in micro-electromechanical system (MEMS) technology have promoted the development of microelectrode arrays (MEA) that allow high resolution recordings in neuroscience research. However, applying MEA in studies in freely moving mice remains very challenging due to the large number of electrical connections required in this type of studies. The use of commutators for a large number of connections is not practical, and headmounts/loggers placed on the animal head are too heavy for small animals such as mice. Therefore, there is a need for a better compact system for using MEA in mice. Herein, we designed such a system and successfully recorded high-density-EEG in freely moving mice.

Methods

We designed a system in which forty flexible ultrathin wires are connected to the headstage enclosed in a container held close to the mouse. The container also houses a logger and battery connected to the headstage. This recording system allows minimizing weighted pressure on the animal using a counterbalance, so that the animal can freely move in the cage.

Results

We tested the system using a signal generator and mouse EEG arrays (NeuroNexus). When potentials produced by the signal generator were recorded via the wires, recorded traces were indistinguishable from the traces that were recorded when the signal generator was connected directly to the logger. We then implanted mice with EEG electrode arrays under surgical anesthesia. The high-density EEG recordings were performed one and four weeks after the surgery. High-quality EEG signals were observed in all the channels of the 32-channel logger (SpikeGadgets) in freely moving mice.

Conclusion

We successfully developed and tested a novel system for enabling high-density EEG recordings in freely moving mice. We expect that this system will be useful for recording biopotentials from different types of MEA in freely moving mice.

Support

NIH 1R43OD023231 (LG), NIH 1RF1AG061774 (DG), and NIH 5R21NS106406 (DG)

Substance P and the neurokinin-1 receptor regulate electroencephalogram non-rapid eye movement sleep slow-wave activity locally

The neuropeptide substance P is an excitatory neurotransmitter produced by various cells including neurons and microglia that is involved in regulating inflammation and cerebral blood flow--functions that affect sleep and slow-wave activity (SWA). Substance P is the major ligand for the neurokinin-1 receptor (NK-1R), which is found throughout the brain including the cortex. The NK-1R is found on sleep-active cortical neurons expressing neuronal nitric oxide synthase whose activity is associated with SWA. We determined the effects of local cortical administration of a NK-1R agonist (substance P-fragment 1, 7) and a NK-1R antagonist (CP96345) on sleep and SWA in mice. The NK-1R agonist significantly enhanced SWA for several hours when applied locally to the cortex of the ipsilateral hemisphere as the electroencephalogram (EEG) electrode but not after application to the contralateral hemisphere when compared to saline vehicle control injections. In addition, a significant compensatory reduction in SWA was found after the NK-1R agonist-induced enhancements in SWA. Conversely, injections of the NK-1R antagonist into the cortex of the ipsilateral hemisphere of the EEG electrode attenuated SWA compared to vehicle injections but this effect was not found after injections of the NK-1R antagonist into contralateral hemisphere as the EEG electrode. Non-rapid eye movement sleep and rapid eye movement sleep duration responses after NK-1R agonist and antagonist injections were not significantly different from the responses to the vehicle. Our findings indicate that the substance P and the NK-1R are involved in regulating SWA locally.

Further characterization of sleep-active neuronal nitric oxide synthase neurons in the mouse brain

We recently demonstrated that Fos is induced in a subpopulation of cortical neuronal nitric oxide synthase (nNOS)-immunoreactive neurons in three rodent species both during spontaneous sleep (SS) and recovery sleep (RS) after a period of sleep deprivation (SD); the proportion of cortical Fos(+)/nNOS neurons was significantly correlated with non-REM (NREM) sleep delta energy. The present study was undertaken to evaluate the specificity of this state-dependent activation of cortical nNOS cells. The percentage of nNOS neurons that expressed Fos during SD and RS was determined in nine subcortical brain regions and the cortex of the mouse brain; a significantly greater proportion of Fos(+)/nNOS neurons was observed during RS only in the cortex and in none of the nine subcortical regions. The proportion of calretinin-, calbindin- and parvalbumin-immunoreactive cortical interneurons that expressed Fos during SD and RS was also determined. In contrast to cortical nNOS neurons, a higher percentage of Fos(+)/calbindin neurons was found during SD than RS; there were no differences in the proportions of Fos-expressing parvalbumin or calretinin neurons between these conditions. Since the nNOS and calretinin cortical interneuron populations overlap extensively in the mouse brain, triple-labeling with these two phenotypic markers and Fos was undertaken in mice from the RS group to determine which combination of markers could best identify the rare "sleep-active" cortical interneuron population. The proportions of both Fos(+)/nNOS neurons and Fos(+)/nNOS/calretinin neurons far exceeded the proportion of Fos(+)/calretinin neurons during RS, but the proportions of these two cell types were not significantly different during RS. Thus, functional activation of nNOS neurons during sleep appears to be restricted to the cerebral cortex and cortical nNOS cells and nNOS/calretinin cells collectively define a cortical interneuron population that is activated during sleep.

Insomnia following hypocretin2-saporin lesions of the substantia nigra

The neuropeptide hypocretin, also known as orexin, has been implicated in waking since its deletion leads to the sleep disorder narcolepsy. Hypocretin neurons project to major arousal areas, and in an effort to determine which region is responsible for the changes in sleep-wake architecture we have developed the neurotoxin hypocretin2-saporin, which lesions hypocretin receptor bearing neurons. Here, in rats, we investigate the effects of hypocretin2-saporin lesions of the substantia nigra and ventral tegmental area in the regulation of sleep and wakefulness. Bilateral injection of hypocretin2-sap into both the ventral tegmental area and substantia nigra (92 and 184 ng/microl, 0.25 microl in the ventral tegmental area and 0.5 microl in the substantia nigra) or into the substantia nigra alone (184 ng/microl, 0.5 microl) produced insomnia. The insomnia seemed to be associated with a large increase in locomotion on days 4 and 6 postinjection, as hyperactivity and stereotypic movements were consistently observed on the video recordings in all lesioned rats. In these rats, a nearly complete loss of both tyrosine hydroxylase and neuron-specific nuclear protein (neuronal nuclei) immunoreactive cells in the substantia nigra as well as diminution of tyrosine hydroxylase-immunoreactive fibers in the caudate putamen was found. Following bilateral injection of hypocretin2-sap at a lower concentration (46 ng/microl, 0.25 microl in the ventral tegmental area and 0.5 microl in the substantia nigra), very little reduction in the number of tyrosine hydroxylase- and neuronal nuclei-immunoreactive neurons and only a temporary increase in wakefulness (17.4% increase during light-off period on day 6 postinjection) were observed. Ventral tegmental area lesions (184 ng/mul of hypocretin2-sap, 0.25 microl, bilateral injections) did not produce significant changes in sleep, although most of the tyrosine hydroxylase- and neuronal nuclei-immunoreactive neurons in the ventral tegmental area were destroyed. Insomnia following hypocretin2-sap lesions of the substantia nigra could be secondary to increased motor activity resulting from reduction of tonic inhibitory control by the substantia nigra.

The diurnal rhythm of adenosine levels in the basal forebrain of young and old rats

There are significant decrements in sleep with age. These include fragmentation of sleep, increased wake time, decrease in the length of sleep bouts, decrease in the amplitude of the diurnal rhythm of sleep, decrease in rapid eye movement sleep and a profound decrease in electroencephalogram Delta power (0.3-4 Hz). Old rats also have less sleep in response to 12 h-prolonged wakefulness (W) indicating a reduction in sleep drive with age. The mechanism contributing to the decline in sleep with aging is not known but cannot be attributed to loss of neurons implicated in sleep since the numbers of neurons in the ventral lateral preoptic area, a region implicated in generating sleep, is similar between young (3.5 months) and old (21.5 months) rats. One possibility for the reduced sleep drive with age is that sleep-wake active neurons may be stimulated less as a result of a decline in endogenous sleep factors. Here, we test this hypothesis by focusing on the purine, adenosine (AD), one such sleep factor that increases after prolonged W. In experiment 1, microdialysis measurements of AD in the basal forebrain at 1 h intervals reveal that old (21.5 months) rats have more extracellular levels of AD compared with young rats across the 24 h diurnal cycle. In experiment 2, old rats kept awake for 6 h (first half of lights-on period) accumulated more AD compared with young rats. If old rats have more AD then why do they sleep less? To investigate whether changes in sensitivity of the AD receptor contribute to the decline in sleep, experiments 3 and 4 determined that for the same concentration of AD or the AD receptor 1 agonist, cyclohexyladenosine, old rats have less sleep compared with young rats. We conclude that even though old rats have more AD, a reduction in the sensitivity of the AD receptor to the ligand does not transduce the AD signal at the same strength as in young rats and may be a contributing factor to the decline in sleep drive in the elderly.

Effects of lateral hypothalamic lesion with the neurotoxin hypocretin-2-saporin on sleep in Long-Evans rats

Narcolepsy, a disabling neurological disorder characterized by excessive daytime sleepiness, sleep attacks, sleep fragmentation, cataplexy, sleep-onset rapid eye movement sleep periods and hypnagogic hallucinations was recently linked to a loss of neurons containing the neuropeptide hypocretin. There is considerable variability in the severity of symptoms between narcoleptic patients, which could be related to the extent of neuronal loss in the lateral hypothalamus. To investigate this possibility, we administered two concentrations (90 ng or 490 ng in a volume of 0.5 microl) of the neurotoxin hypocretin-2-saporin, unconjugated saporin or saline directly to the lateral hypothalamus and monitored sleep, the entrained and free-running rhythm of core body temperature and activity. Neurons stained for hypocretin or for the neuronal specific marker were counted in the perifornical area, dorsomedial and ventromedial nucleus of the hypothalamus. More neuronal nuclei (NeuN) cells were destroyed by the higher concentration of hypocretin-2-saporin (-55%) compared with the lower concentration (-34%) in the perifornical area, although both concentrations lesioned the hypocretin neurons almost equally well (high concentration=91%; low concentration=88%). The high concentration of hypocretin-2-saporin also lesioned neurons in the dorsomedial nucleus of the hypothalamus and ventromedial nucleus of the hypothalamus. Narcoleptic-like sleep behavior was produced by both concentrations of the hypocretin-2-saporin. The high concentration produced a larger increase in non-rapid eye movement sleep amounts during the normally active night cycle than low concentration. Neither concentration of hypocretin-2-saporin disrupted the phase or period of the core temperature or activity rhythms. The low concentration of unconjugated saporin did not significantly lesion hypocretin or neurons and did not alter sleep. The high concentration of unconjugated saporin produced some loss of neuronal nuclei-immunoreactive (NeuN-ir) neurons and hypocretin immunoreactive neurons, but only a transient increase in non-rapid eye movement sleep. These results led us to conclude that the extent of hypocretin neuronal loss together with an accompanying loss of cells in the lateral hypothalamus may explain the differences in severity of symptoms seen in human narcolepsy.

Hypothalamic regulation of sleep

The recent discovery linking narcolepsy, a sleep disorder characterized by very short REM sleep latency, with a neuropeptide that regulates feeding and energy metabolism, provides a way to understand how several behaviors may be disrupted as a result of a defect in this peptide. In this chapter we review the evidence linking hypocretin and sleep, including our own studies, and propose that a defect in the lateral hypothalamus that also involves the hypocretin neurons is likely to produce a disturbance in sleep, mood, appetite, and rhythms.

Hypocretin-2-saporin lesions of the lateral hypothalamus produce narcoleptic-like sleep behavior in the rat

Hypocretins (Hcrts) are recently discovered peptides linked to the human sleep disorder narcolepsy. Humans with narcolepsy have decreased numbers of Hcrt neurons and Hcrt-null mice also have narcoleptic symptoms. Hcrt neurons are located only in the lateral hypothalamus (LH) but neither electrolytic nor pharmacological lesions of this or any other brain region have produced narcoleptic-like sleep, suggesting that specific neurons need to be destroyed. Hcrt neurons express the Hcrt receptor, and to facilitate lesioning these neurons, the endogenous ligand hypocretin-2/orexin B (Hcrt2) was conjugated to the ribosome-inactivating protein saporin (SAP). In vitro binding studies indicated specificity of the Hcrt2-SAP because it preferentially bound to Chinese hamster ovary cells containing the Hcrt/orexin receptor 2 (HcrtR2/OX(2)R) or the Hcrt/orexin receptor 1 (HcrtR1/OX(1)R) but not to Kirsten murine sarcoma virus transformed rat kidney epithelial (KNRK) cells stably transfected with the substance P (neurokinin-1) receptor. Administration of the toxin to the LH, in which the receptor is known to be present, eliminated some neurons (Hcrt, melanin-concentrating hormone, and adenosine deaminase-containing neurons) but not others (a-melanocyte-stimulating hormone), indicating specificity of the toxin in vivo. When the toxin was administered to the LH, rats had increased slow-wave sleep, rapid-eye movement (REM) sleep, and sleep-onset REM sleep periods. These behavioral changes were negatively correlated with the loss of Hcrt-containing neurons but not with the loss of adenosine deaminase-immunoreactive neurons. These findings indicate that damage to the LH that also causes a substantial loss of Hcrt neurons is likely to produce the multiple sleep disturbances that occur in narcolepsy.

Effects of hypocretin-saporin injections into the medial septum on sleep and hippocampal theta

Neurons containing the peptide hypocretin, also known as orexin, were recently implicated in the human sleep disorder narcolepsy. Hypocretin neurons are located only in the lateral hypothalamus from where they innervate virtually the entire brain and spinal cord. This peptide is believed to be involved in regulating feeding and wakefulness. However, to fully understand what other behaviors are regulated by this peptide it is necessary to investigate each hypocretin target site. In the present study, we focus on one hypocretin target site, the medial septum, where there is a dense collection of hypocretin-2 receptor-containing cells, and degenerating axons are present here in canines with narcolepsy [J. Neurosci. 19 (1999) 248]. We utilize a saporin toxin conjugated to the hypocretin receptor binding ligand, hypocretin-2, and find that when this toxin is injected into the medial septum, it lesions the parvalbumin and cholinergic neurons. We contrast the effects of the hypocretin-saporin with another saporin conjugated toxin, 192 IgG-saporin, that lesions only the cholinergic neurons in the basal forebrain. 192 IgG-saporin reduced theta activity, a finding consistent with previous reports [J. Neurophysiol. 79 (1998) 1633; Neurodegeneration 4 (1995) 61; Neuroscience 62 (1994) 1033]. However, hypocretin-saporin completely eliminated hippocampal theta activity by day 12, indicating that parvalbumin-containing cells in the medial septum generate theta. The daily amount of sleep and wakefulness were not different between hypocretin-saporin, 192 IgG-saporin, or saline-treated rats. The homeostatic response to 12 h prolonged wakefulness was also not affected in hypocretin-saporin lesioned rats. These findings suggest that hypocretin neurons could facilitate theta generation during episodes of purposeful behavior by activating GABAergic neurons in the MS/VDB. In this way, hypocretin, which is implicated in feeding, energy metabolism and wakefulness, serves to influence cognitive processes critical for the animal's survival.

Cellular localization of lipocalin-type prostaglandin D synthase (beta-trace) in the central nervous system of the adult rat

We applied high-resolution laser-scanning microscopy, electron microscopy, and non-radioactive in situ hybridization histochemistry to determine the cellular and intracellular localization of lipocalin-type prostaglandin D synthase, the major brain-derived protein component of cerebrospinal fluid, and its mRNA in leptomeninges, choroid plexus, and parenchyma of the adult rat brain. Both immunoreactivity and mRNA for prostaglandin D synthase were located in arachnoid barrier cells, arachnoid trabecular cells, and arachnoid pia mater cells. Furthermore, meningeal macrophages and perivascular microglial cells, identified by use of ED2 antibody, were immunopositive for prostaglandin D synthase. In the arachnoid trabecular cells, the immunoreactivity for prostaglandin D synthase was located in the nuclear envelope, Golgi apparatus, and secretory vesicles, indicating the active production and secretion of prostaglandin D synthase. In the meningeal macrophages, prostaglandin D synthase was not found around the nucleus but in lysosomes in the cytoplasm, pointing to an uptake of the protein from the cerebrospinal fluid. Furthermore, the existence of meningeal cyclooxygenase (COX) -1 and COX-2 was investigated by Western blot, Northern blot, and reverse transcriptase-polymerase chain reaction (RT-PCR), and the colocalization of COX-2 and prostaglandin D synthase was demonstrated in virtually all cells of the leptomeninges, choroid plexus epithelial cells, and perivascular microglial cells, suggesting that these cells synthesize prostaglandin D(2) actively. Alternatively, oligodendrocytes showed prostaglandin D synthase immunoreactivity without detectable COX-2. The localization of lipocalin-type prostaglandin D synthase in meningeal cells and its colocalization with COX-2 provide evidence for its function as a prostaglandin D(2)-producing enzyme.

Strong rebound of wakefulness follows prostaglandin D2- or adenosine A2a receptor agonist-induced sleep

We studied the effect of sleep excess on the sleep-wakefulness pattern of rats. Subarachnoid infusion of prostaglandin D2 or the adenosine A2a receptor agonist CGS21680 effectively induced slow wave sleep (SWS) for the first 12 h of the night-time period, whereas they did not induce sleep during the following 24 h of infusion. An increase in the amount of wakefulness was seen during the last 12 h of prostaglandin D2 infusion. The amounts of wakefulness strongly increased during the following 36-h recovery period. Rebound wakefulness was extraordinarily strong after the cessation of CGS21680 infusion, reaching almost complete insomnia during the night-time. Treatment of animals with prostaglandin D2 overnight, following by treatment with CGS21680 on the next night, resulted in the strongest induction of wakefulness rebound. During the rebound period, the amount of wakefulness reached up to 50 min per hour in the daytime. Rebound of wakefulness depended on the amounts of preceding SWS induced by infusion of prostaglandin D2 for 6 or 12 h and of CGS21680 for 12 h. The larger the amount of SWS, the larger the amount of the following rebound of wakefulness. Rebounds of wakefulness occurred as a result of decrease in SWS amounts, whereas paradoxical sleep amounts did not change. Desensitization of adenosine A2a receptors and accumulation of prostaglandin E2 may be involved in the production of strong wakefulness rebound following relatively long treatments (more than 12 h) with prostaglandin D2 or CGS21680.

Dominant expression of rat prostanoid DP receptor mRNA in leptomeninges, inner segments of photoreceptor cells, iris epithelium, and ciliary processes

Prostaglandin (PG) D2 is one of the major prostanoids in the mammalian brain and eye tissues. Its function is mediated by the prostanoid DP receptor, which is specific for PGD2 among the various prostanoids. In this study, we cloned the full-length cDNA for the rat DP receptor and used it for detection of DP receptor mRNA in various rat tissues. Northern blotting and RT-PCR analyses revealed that this DP receptor was expressed most intensely in the eye tissues, moderately in the leptomeninges and oviduct, and weakly in the epididymis. The tissue distribution profile of the mRNA for the rat DP receptor is overlapped with those of hematopoietic and lipocalin-type PGD synthases. Among rat eye tissues, the expression was the highest in the iris. In situ hybridization and in situ RT-PCR revealed DP receptor mRNA to be localized in the epithelium of the iris and ciliary body and in photoreceptor cells of the retina, suggesting the involvement of the receptor in the physiological regulation of intraocular pressure and the vision process. In the brain, DP receptor mRNA was dominantly expressed in the leptomeninges and was not detected in the brain parenchyma including the ventral rostral forebrain, the surface area of which is reportedly involved in sleep induction by PGD2.

Activation of ventrolateral preoptic neurons by the somnogen prostaglandin D2

Prostaglandin D2 (PGD2) is an extensively studied sleep-promoting substance, but the neuroanatomical basis of PGD2-induced sleep is only partially understood. To determine potential regions involved in this response, we used Fos immunohistochemistry to identify neurons activated by infusion of PGD2 into the subarachnoid space below the rostral basal forebrain. PGD2 increased nonrapid eye movement sleep and induced striking expression of Fos in the ventrolateral preoptic area (VLPO), a cluster of neurons that may promote sleep by inhibiting the tuberomammillary nucleus, the source of the ascending histaminergic arousal system. Fos expression in the VLPO was positively correlated with the preceding amount of sleep and negatively correlated with Fos expression in the tuberomammillary nucleus. PGD2 also increased Fos immunoreactivity in the basal leptomeninges and several regions implicated in autonomic regulation. These observations suggest that PGD2 may induce sleep via leptomeningeal PGD2 receptors with subsequent activation of the VLPO.

Continuous recordings of brain regional circulation during sleep/wake state transitions in rats

Continuous measurement of regional blood flow (RBF) in the brain of a freely behaving rat was attained by a combination of laser-Doppler (LD) flowmetry and our originally devised apparatus, which had been developed for the automatic releasing of the twisting of lines connected between experimental apparatus and the freely behaving animal. RBF changes were studied in a ventral region of the rostral basal forebrain along with sleep-wake states. When compared with the RBF level during slow-wave sleep (SWS), levels of RBF during paradoxical sleep (PS) and wakefulness were higher by 24 (P = 0.0001) and 9% (P < 0.05), respectively. The LD signals suggested that the RBF elevation during PS was produced by dilation of both the large brain arteries and small vessels, whereas the elevation during wakefulness was caused by dilation of small vessels that was counteracted by contraction of large arteries. It was noticed that the original circulation tended to begin changing before the onset of SWS. A circadian rhythm was also demonstrated for the RBF, which largely decreased around the onset of the light period and returned to the high level before the beginning of the dark period. Thus continuous and real-time recordings of regional circulation were performed with satisfactorily precision in freely behaving rats.

Influence of pyruvate, threonine and phosphoethanolamine on activities of some acetaldehyde-producing enzymes

Threonine (50 mg/100 g, i.p.) leads to increased hepatic threonine aldolase activity in rats, although endogenous ethanol concentrations remain stable. After pyruvate administration (50 mg/100 g, i.p.), endogenous blood ethanol levels are raised within 30 min, but return to normal at 60 min. The activity of threonine aldolase is decreased in the liver, whereas phosphoethanolamine lyase and pyruvate dehydrogenase activities remain unchanged. Phosphoethanolamine administration (23 mg/100 g, i.p.) did not change the endogenous ethanol concentration or pyruvate dehydrogenase, threonine aldolase and phosphoethanolamine lyase activities. Pyruvate appears to be a better precursor of acetaldehyde than threonine or phosphoethanolamine.