Spontaneous sleep in mice with targeted disruptions of neuronal or inducible nitric oxide synthase genes

Abstract

Nitric oxide (NO) affects almost every physiological process, including the regulation of sleep. There is strong evidence that NO plays an important role in rapid eye movement sleep (REMS) regulation. To further investigate the role of NO in sleep, we characterized spontaneous sleep in mice with targeted disruptions (knockout; KO) in the neuronal nitric oxide synthase (nNOS) or inducible (i)NOS genes. REMS in nNOS KO mice was substantially lower than that of their control mice. In contrast, the iNOS KO mice had significantly more REMS than their controls. Inducible NOS KO mice also had less non-REMS (NREMS) during the dark period. Results suggest that nNOS and iNOS play opposite roles in REMS regulation.

Introduction

Nitric oxide (NO) is involved in sleep regulation. NO is derived from arginine via the action of nitric oxide synthase (NOS). There are three types of NOSs: neuronal (nNOS, or NOS-1), inducible (iNOS, or NOS-2), and endothelial (eNOS, or NOS-3). nNOS and eNOS are constitutively expressed and nNOS is normally the major source of brain NO [17]. eNOS is mainly found in vascular endothelial cells and plays a role in blood pressure regulation [47]. In contrast, iNOS is expressed in brain during development but is not normally present in adulthood, although it is induced in glial cells in response to inflammatory mediators [34].

nNOS is present within sleep-related brain structures, such as the pedunculopontine tegmental nuclei (PPT), the laterodorsal tegmental nucleus (LDT) and the dorsal Raphe nucleus (DRN) (e.g., Ref. [30]). There is substantial evidence that NO in these nuclei plays a critical role in rapid eye movement sleep (REMS) regulation. For example, nNOS colocalizes with cholinergic neurons in the LDT/PPT; these neurons project to the medial pontine reticular formation (mPRF) and are crucial in REMS generation [46]. Microinjection of SNAP, a NO donor, into the PPT of cats increases REMS substantially [8]. On the other hand, microinjection of l-NAME, an inhibitor of NOS, into the PPT reduces REMS [8]. In rats, microinjection of l-arginine into the PPT increases the duration of non-REMS and the number of REMS episodes, while microinjection of another NOS inhibitor (l-NAPNA) into this area reduces both NREMS and REMS [14]. Further, local inhibition of NOS within the mPRF of cats reduces acetylcholine (ACh) release and decreases the amount of REMS [32], [33]. Microinjection of 7-NI, a selective nNOS inhibitor, into the DRN of rats decreases REMS [4]. In contrast, microinjection of l-NAME into the anterior diencephalic region increases REMS during the dark period, followed by a negative rebound during the light period [44].

In contrast to these microinjection studies, which indicate a role for NO in REMS regulation, the results from studies using systemic or intracerebroventricular (i.c.v.) routes of NO inhibitor administration are less clear. Systemic or i.c.v. injection of NOS inhibitors or NO donors also affect sleep, although the direction of the effect may depend on the time of the day. For instance, l-NAME given intravenously at light onset suppresses spontaneous NREMS and REMS [23], [37]. In addition, intraperitoneal (i.p.) injection of l-NAME suppresses sleep rebound after sleep deprivation [43]. In contrast, i.p. injection of similar doses of l-NAME at dark onset increases both NREMS and REMS [4]. In addition, the 7-NI given i.p. suppresses sleep when given either at dark onset or light onset [4], [11]. Administration of l-NAME via i.c.v. injection also suppresses both NREMS and REMS in rabbits and rats [21], [23]. Conversely, the i.c.v. administration of the NO donors molsidomine (SIN-1) or S-nitroso-N-acetylpenicillamine (SNAP) at dark onset promotes NREMS after a latency of 9 h in rats [20].

The basal forebrain/preoptic area is another important brain structure involved in the regulation of sleep/wakefulness cycles; it is hypothesized that an interaction between cholinergic and GABAergic systems is involved [50]. Ascending cholinergic projections from the basal forebrain play an essential role in cortical activation. Basal forebrain neurons also send descending projections, probably GABAergic, to key REMS regulating nuclei, such as the LDT/PPT, locus coeruleus, and dorsal raphe nucleus [48]. NO can affect both Ach and GABA release in the basal forebrain [26], [52]. Finally, pretreatment of rats with a GABA receptor antagonist reverses the sleep decrease caused by l-NAME [38].

In rats, soluble NOS activity in the brain has a diurnal variation, with higher levels during the dark period (active period) and lower levels during the light period [2]. Consistent with this finding, Cespuglio et al. [7] found that the highest cortical levels of NO occur during the awake state. Similarly, Williams et al. [54] found that extracellular NO concentrations in the thalamus are high during wakefulness and REMS and significantly lower during NREMS. Superficially, these data are inconsistent with the notion that NO is involved in the induction of sleep. However, these studies do not distinguish the anatomical source of NO or the NOS isoform measured, and it has been shown that NO in one site has different effects than NO in other sites within the brain [52].

Thus, currently it is clear that manipulation of NOS affects sleep. However, specific effects depend upon how drugs are given, the time of day they are given and what specific substance is given. Further, the use of NOS inhibitors is confounded by their relative lack of specificity for individual NOS isoforms. In this study, we use mice with targeted disruptions in the nNOS or iNOS genes, commonly referred to as nNOS or iNOS knockouts (KO) to characterize their spontaneous sleep patterns. We report that mice lacking nNOS have less REMS while those lacking iNOS have more REMS.