R.a.s.

Humans have three sleep and arousal states: waking, asleep (resting or slow-wave sleep), and asleep and dreaming (paradoxical, active, or rapid eye movement sleep). These states are controlled by the reticular activating system located in the mesopons, which interacts with descending reticulospinal and ascending hypothalamic, basal forebrain, and thalamocortical systems. These three states develop and occur in a predictable manner, and we can explain these states according to the firing properties of neurons based on their intrinsic membrane properties, their synaptic and neurochemical connectivity, and their responsiveness to sensory inputs. This article describes the characteristics of mechanisms mediating sleep and arousal, their neurological substrates, and the cellular, neurochemical, and network properties of those substrates.

Reticular Activating System

The reticular activating system spans an extensive portion of the brainstem. Most of the neurons comprising the midbrain reticular formation lie dorsal and lateral to the red nuclei. Complex interactions between multiple neurotransmitters modulate the action of the reticular activating system with both cholinergic and adrenergic neurotransmission having key roles. The reticular activating system's fundamental role is regulating arousal and sleep−wake transitions. The ascending reticular activating system projects to the intralaminar nuclei of the thalami, which projects diffusely to the cerebral cortex. The ascending projections of the reticular activating system enhance the attentive state of the cortex and facilitate conscious perception of sensory stimuli. Additionally, the collective role of the brainstem reticular formation is to regulate autonomic function, muscle reflexes, and tone.

Coherence and Frequency

The RAS controls sleep and waking and fight-or-flight responses. While this system provides signals that modulate our wake-sleep states, it also serves to help us respond to the world around us. For example, strong stimuli simultaneously activate ascending RAS projections to the thalamus and then the cortex and cause arousal and also activate descending projections that influence the spinal cord in the form of postural changes in tone resulting from the startle response, as well as trigger locomotor events in fight-or-flight responses. During sleep, the same system is responsible for the relative lack of ascending sensory awareness during SWS, as well as the descending atonia of REM sleep. This system also modulates the activity of virtually every other system in the brain. Growing evidence suggests that the control of sleep and waking is a fundamental property of neuronal networks and prior activity within each network (Kueger et al., 2008) and that intrinsic properties of neurons in multiple regions modulate sleep autoregulation, that is, suggesting that sleep is neither a passive nor an active phenomenon (Kumar, 2010). As we will see below, it is the RAS that supplies the “context” of sensory experience during waking.

Two major elements determining the activity of large assemblies of neurons such as in the EEG are coherence and frequency. Coherence is the term for how groups of neurons, firing in coordination, can create a signal that is mirrored instantaneously and precisely by other groups of neurons across the brain. These transient episodes of coherence across different parts of the brain may be an electrical signature of thought and actions. Our recent discovery demonstrated the presence of electrical coupling in three nuclei of the RAS, a mechanism that allows groups of neurons to fire synchronously. That mechanism is addressed in Chapter 4, which describes the presence of electrical coupling in the RAS and how that mechanism is modulated by the stimulant modafinil, which increases electrical coupling to drive coherence and disinhibits a number of systems to drive higher frequencies and induce arousal (Garcia-Rill et al., 2008). Briefly, modafinil increases electrical coupling, and since most coupled neurons in the RAS are GABAergic, the coupling decreases input resistance, decreasing activity and GABA release, thus disinhibiting many other cell types. This disinhibition leads to overall higher frequency in activity, that is, during sleep and arousal, in the RAS (Garcia-Rill et al., 2007, 2008; Heister et al., 2007) and thalamocortical systems (Urbano et al., 2007). In other words, because increased coupling in GABAergic neurons will lead to decreased GABA release, the tendency will be to increase coherence and also disinhibit most other transmitter systems, leading to increased excitation, especially during waking. That is, if modafinil increases electrical coupling, it should enable better coherence at all frequencies, during waking and even after its effects are waning, during sleeping. That is why modafinil is also useful in regulating coherence during sleep. Conversely, we will see that the most fast-acting anesthetics known, inhaled halothane and injected propofol, both block gap junctions and put us to sleep very rapidly. That is, the control of gap junctions can determine if we are asleep or awake.

The other face of large-scale activity is the frequency of firing, especially of ensemble activity, which is also essential to the neural encoding process. Chapter 9 is based on another major discovery, the presence of gamma band activity in the same RAS nuclei. Recent data suggest that many, perhaps all, of the neurons in these three RAS regions fire at gamma band frequency when maximally activated, but no higher. These results now suggest that brain stem regions not only can generate but also plateau at such frequencies, which is surprising because gamma band activity was first described in the cortex and is presumably involved in consciousness, learning, and memory. This is less surprising when one considers that gamma band activity has been described in other subcortical regions like the thalamus, hippocampus, basal ganglia, and cerebellum. The goal then becomes one to identify the mechanisms behind gamma band activity in the RAS and determine their function. In Chapter 9, we will first address the classical role of gamma band activity and the presence and mechanisms behind gamma band activity in the subcortical brain regions, then turn to the mechanisms behind gamma band activity in the RAS, and finally speculate on the potential role of such activity appearing at brain stem levels, in a very old, phylogenetically speaking, region such as the RAS.

NEURONAL CIRCUITS AND NUCLEI UNDERLYING SLEEP

The Reticular-Activating Process

The reticular-activating system consists of several distinct but interrelated arousal systems that are differentiated by anatomy, neurotransmitter, and function (Robbins & Everitt, 1995). In this section, I will focus on two arousing and energizing systems: the noradrenergic and the dopaminergic pathways. A large body of evidence shows that acute exercise activates these two monoamine systems and increase the release of noradrenaline (NA) and dopamine (DA) in several brain areas (e.g., Meeusen & De Meirleir, 1995; Meeusen & Piacentini, 2001; Meeusen, Piacentini, & De Meirleir, 2001).

The unique source of NA to hippocampus and neocortex is the locus coeruleus (LC), a brainstem nucleus that widely projects its noradrenergic axons throughout the central nervous system (Berridge & Waterhouse, 2003). NA released by these fibers binds on three families of adrenergic receptors: the α1, the α2, and the β1–3 receptors (Ramos & Arnsten, 2007). It is important to note that LC neurons may fire in two different activity modes: tonic and phasic (Berridge & Waterhouse, 2003). Tonic LC activity is characterized by a sustained and highly regular pattern above 2 Hz during active waking, such as exercising, but lower rates (<1 Hz) during slow-waves sleep (Foote, Aston-Jones, & Bloom, 1980). Such fluctuations in noradrenergic activity can be detected in cortical electroencephalogram (EEG) patterns and event-related potentials (ERP) (Niewenhuis, Aston-Jones, & Cohen, 2005; Pineda, Foote, & Neville, 1989). Phasic LC activity occurs in response to novel, noxious, stressful, or rewarding events and is characterized by a short-latency and brief burst of two to three action potentials followed by a relatively short period (300–700 ms) of silence (Aston-Jones & Bloom, 1981; Grant, Aston-Jones, & Redmond, 1988; Rasmussen, Morilak, & Jacobs, 1986). Phasic LC responses are associated with overt orienting responses and habituate with repeated stimulus presentation (Aston-Jones, Rajkowski, Kubiak, & Alexinsky, 1994). Extrasynaptic brain levels of NA are linearly related to tonic LC activity (Berridge & Abercrombie, 1999), whereas NA levels within the synaptic cleft are more related to phasic firing of LC presynaptic neurons (Berridge & Waterhouse, 2003). The LC noradrenergic system mediates alertness and appears to be involved in detecting sensory signals and maintaining discrimination processes under high levels of arousal and stress (Berridge & Waterhouse, 2003; Pribram & McGuinness, 1975; Ramos & Arnsten, 2007; Robbins & Everitt, 1995). NA released at the cortical level improves the signal-to-noise ratio by reducing “noise” and/or facilitating processing of relevant sensory signals (Hurley, Devilbiss, & Waterhouse, 2004; Moxon, Devilbiss, Chapin, & Waterhouse, 2007; Waterhouse & Woodward, 1980). This enhancement of processing of sensory information takes place at both the single neuron and neuronal network levels and leads to improvement of cognitive function under “noisy” conditions, where irrelevant stimuli could impair performance (Berridge & Waterhouse, 2003). The LC noradrenergic system could have played an important role in the survival of many animal species allowing fleeing preys to detect more easily predators in environments rich in irrelevant stimuli. In this situation, typically associated with high arousal level (threat and exercise), it may be necessary for the prey to scan the environment for rapid detection of multiple stimuli. Several arguments for a facilitating effect of exercise on sensory processing will be presented in section 4. The LC noradrenergic system also innervates prefrontal cortex (PFC) and exerts a potent modulatory influence on executive functions such as inhibiting the processing of irrelevant stimuli (Woods & Knight, 1986) or keeping task-relevant information “online” in working memory (Ramos & Arnsten, 2007). PFC processing is strengthened by moderate levels of NA and α2-receptor stimulation but impaired by higher levels of NA and α1-receptor stimulation (Arnsten & Robbins, 2002). In this way, NA can be considered as a gradual neurochemical swing from anterior cortical regions to more posterior cortical and subcortical processes. Moreover, moderate to vigorous exercise might lead to a high level of brain NA and consequently a progressive shut off of the PFC. This dysregulation of PFC activity by the LC noradrenergic system under highly stressful conditions could be synergistic to the hypofrontality process described later. The role of brain NA in emotional memory consolidation within the amygdala under such high arousal conditions will not be addressed in this chapter. Finally, NA exerts a robust modulatory effect on astrocyte glycogen levels and acts to increase glucose availability, the main energy supply of neuronal activity (Berridge & Waterhouse, 2003). This last action of NA on brain functioning confirms its major role in the energetics of information processing.

The dopaminergic system originates from cell bodies principally located in the substantia nigra pars compacta and from the ventral tegmentum (Grimm, Mueller, Hefti, & Rosenthal, 2004). The nigrostriatal system projects predominantly from the substantia nigra to the corpus striatum, the dorsal putamen, caudate nucleus, and globus pallidus, which, in turn, modulate activity of a large network involving the motor thalamus, supplementary motor area, premotor area, and primary motor cortex (Reeves, Bench, & Howard, 2002). The target sites of the ventral tegmental area are several regions of the limbic system such as the nucleus accumbens, the amygdala, and the anterior cingulate cortex, and widespread regions of the neocortex with higher density of projections to the PFC (Mehta & Riedel, 2006). These two dopaminergic networks have been respectively termed the mesolimbic and the mesocortical systems (Meck, 2006). Microdialysis studies, conducted in animals, showed that acute exercise increases DA level and stays significantly above baseline in the striatum and nucleus accumbens up to 1–2 h after running in both trained and untrained animals (Meeusen, Hasegawa, & Piacentini, 2005; Wilson & Marsden, 1995). This elevation of DA level during and after exercise has not yet been replicated in humans with positron emission tomography (PET) technique (Wang et al., 2014). More studies in humans and animals would be necessary to examine the effects of acute exercise on the three dopaminergic systems (nigrostriatal, mesolimbic, and mesocortical) and on DA level in the striatum and the limbic system, as well as the PFC.

Similar to LC noradrenergic neurons, dopaminergic neurons exhibit two distinct modes of spike firing: tonic and phasic (Grace & Bunney, 1984ab). The tonic mode of firing is dependent on the spontaneous tonic spike activity of DA neurons (Goto, Otani, & Grace, 2007). The DA system is tonically activated by excitatory stimuli via sustained increases in DA neuron firing or via presynaptic stimulation of DA terminals by glutamate (Grace, 2000). In that case, DA mainly escapes from the synaptic cleft and enters the extracellular space (Grace, 2000). Bothe et al. (2013) presented experimental arguments for higher tonic levels of extracellular DA induced by acute exercise in humans. Tonic mode of firing controls the responsiveness to phasic activation of DA neurons in such a way that increases in tonic DA levels cause a potent inhibition of phasic, spike-dependent DA release (Grace, 2000). Higher levels of tonic extracellular DA, after acute exercise, could directly inhibit the magnitude of phasic DA release (Bothe et al., 2013). The phasic mode of firing is defined as the spike-dependent release of DA into the synaptic cleft, primarily responsible for the behaviorally relevant actions of DA systems (Grace, 2000). Several arguments from animal studies as well as neuropsychological studies in Parkinson’s disease patients suggest a role for the nigrostriatal DA system in response preparation and motor readiness (Robbins & Everitt, 1995). In another respect, the mesocortical DA system plays an important role in working memory and cognitive control (Arnsten, 1998; Floresco & Magyar, 2006). Finally, the mesolimbic DA system is implicated in quantification of reward and incentive motivational processes (Robbins & Everitt, 1995). Concerning the mesocortical system, Cools and D’Esposito (2011) showed that the relationship between cognitive performance and PFC baseline dopamine levels follows an inverted-U-shaped function, where both too little and too much DA impairs performance. It would be very interesting to examine this relationship in animals and humans while manipulating baseline tonic DA levels with exercise intensity. In that way, the impairment of PFC processing and executive control, while exercising, could result from three different and synergistic mechanisms induced by acute exercise: (1) a too high level of NA, induced by an activation of the LC system; (2) a too high level of baseline tonic DA, induced by an activation of the mesocortical DA system; and (3) a deactivation of the PFC according to the hypofrontality mechanism described hereafter.

Abstract

The development of the reticular activating system (RAS) reflects changes in wake–sleep patterns from the ­newborn to the adult. The transition from the primitive wakefulness of the ­newborn to the mature wake–sleep pattern was termed “advanced wakefulness” by ­Kleitman. A major milestone during this period is the developmental decrease in REM sleep, during which intrinsic membrane properties and neurotransmitter effects in the RAS change dramatically. Another milestone is puberty and the effects of hormones and gonadal steroids, which describe major functional changes whose dysregulation can result in the postpubertal onset of a number of disorders. The most obvious underlying principle behind the development of the RAS, however, is that there is a dramatic increase in waking as the human develops from birth to adulthood.

Analgesics

Before the advent of ether and subsequent improvements in anesthesia drugs, surgery and other painful procedures were accomplished under the influence of alcohol and toxic plants. These plants included datura (Datura stramonium), gelsemium (Gelsemium sempervirens), henbane (Hyoscyamus niger), poison hemlock (Conium maculatum), and opium poppy (Papaver somniferum). With the exception of gelsemium, most of the plants discussed in this section are less dramatic in their activity and can be used in formulas given over the long term.

Reticular Activating System and Delineation of Arousal Systems

The transection studies just reviewed support the concept of a pontomesencephalic wake-promoting or arousal system. No discovery was historically more significant than the description of the reticular activating system (RAS) by Moruzzi.9 Large lesions of the core of the rostral pontine and mesencephalic tegmentum are followed by persistent somnolence and EEG synchronization, and electrical stimulation of this region induces arousal from sleep. Interruption of sensory pathways does not affect EEG activation. It was hypothesized that cells in the RAS generated forebrain activation and wakefulness.

The concept of the RAS has been superseded by the finding that arousal is facilitated not by a single system but instead by several discrete neuronal groups localized within and adjacent to the pontine and midbrain reticular formation and its extension into the hypothalamus (Fig. 7-1). These discrete neuronal groups are identified and differentiated by their expression of molecular machinery that synthesizes and releases specific neurotransmitters and neuromodulators. These include neuronal groups that synthesize serotonin, noradrenalin, histamine, acetylcholine, and orexin/hypocretin (herein called orexin). Each of these systems has been studied extensively in the context of the control of specific aspects of waking behaviors. Here we will give only a brief overview of each, focusing on their contribution to generalized brain arousal or activation. Before proceeding, we point out certain general properties of these neuronal systems.

Arousal is a global process, characterized by concurrent changes in several physiologic systems, including autonomic, motor, endocrine, and sensory systems, and in EEG tracings. Thus, it is intriguing that most arousal systems share one critical property: the neurons give rise to long, projecting axons with extensive terminal fields that impinge on multiple regions of the brainstem and forebrain. These diffuse projections enable the systems to have multiple actions, as might be expected of arousal systems. In this review, we emphasize the ascending projections—that is, projections from the brainstem and hypothalamus to the diencephalon, limbic system, and neocortex, as these are particularly germane to the generation of cortical arousal. Some arousal systems also give rise to descending projections, which are also likely to play a role in regulating certain properties of sleep–wake states, such as changes in muscle tone.

The release of neurotransmitters and neuromodulators at nerve terminals is initiated by the propagation of action potentials to the terminals. Thus, neurotransmitter release is correlated with the discharge rate of neurons. Most arousal systems have been studied by recording the discharge patterns of neurons in “freely moving” animals, in relationship to spontaneously occurring wake and sleep states. Increased discharge during arousal or wake compared with sleep constitutes part of the evidence for an arousal system.

The actions of a neurotransmitter on a target system are determined primarily by the properties of the receptors in the target. The neurotransmitters and neuromodulators underlying arousal systems each act on several distinct receptor types, with diverse actions. In addition, postsynaptic effects are regulated by transmitter-specific “reuptake” molecules, which transport the neurotransmitter out of the synaptic space, terminating its action. Pharmacologic actions are usually mediated by actions on specific receptor types or transporters (see examples later).

Chronic lesions of individual arousal systems or genetic knockout (KO) of critical molecules have only small or sometimes no effect on sleep–wake patterns (with the exception of serotonin and orexin KOs; see later), even though acute manipulations of these same systems have strong effects on sleep–wake. The absence of chronic lesions or KO effects is probably explained by the redundancy of the arousal systems, such that, over time, deficiency in one system is compensated for by other systems or by changes in receptor sensitivity. Electrophysiologic studies show that the arousal systems are normally activated and deactivated within seconds or minutes. Thus, effects of acute experimental manipulations of particular arousal-related neurotransmitters, as with administration of a drug, may better mimic the normal physiologic pattern and be more informative as to their function.

REM sleep is, on one hand, a sleep state, but, on the other hand, it is associated with neocortical EEG characteristics of wake. In parallel with these two sides of REM, it has been shown that arousal systems can be classified into two types, ones that are “off” in REM, befitting the sleeplike property of REM, and others that are “on” in REM, befitting the wakelike properties of REM. Some arousal-promoting systems (summarized later) also play a role in REM control. Detailed analyses of the control of the role of these systems in REM sleep can be found in Chapters 8 and 9.

A Matter of Time

The regions aside from the reticular activating system (RAS) implicated in the modulation of waking include the hypocretin-containing neurons of the lateral hypothalamus, the cholinergic neurons of the basal forebrain, and the histamine-containing neurons of the tuberomammillary nucleus (TMN). As mentioned in Chapter 2, Jaime Villablanca carried out a series of heroic experiments using brain stem transections and maintaining the animals for prolonged periods (Villablanca, 2004). Midcollicular or mesencephalic transections that separated the RAS from the forebrain produced animals that were awake 75% and 55% of the time, respectively. These animals were sensitive to sensory stimuli, could stand and walk, attempted to climb, and manifested low-amplitude, high-frequency activity below the level of the transection. Nevertheless, as mentioned in Chapter 2, Villablanca ultimately concluded that “true waking behavior depended on the cholinergic reticular core” (Villablanca, 2004). However, the permanently isolated forebrain, after high- or low-brain stem transections, led to alternating episodes of low-amplitude, high-frequency cortical electroencephalogram (EEG) activity (Batsel, 1960; Villablanca, 2004). Presumably, some or all of the regions listed above were responsible for these wake-like episodes above the level of the transections. In the isolated forebrain of the cat, electrical stimulation of the posterior hypothalamus and of the basal forebrain induces fast cortical EEG rhythms (Bakuradze et al., 1975; Berladetti et al., 1977), and cholinergic (Sakai et al., 1990) stimulation of these areas induces arousal, suggesting that these areas indeed modulate waking.

If the cortex and striatum were removed, leaving the thalamus, hypothalamus, and basal forebrain connected to the brain stem, the model was called “diencephalic.” These animals became hyperactive, developing obstinate progression; were hyperreactive to sensory stimuli; and manifested low-amplitude, high-frequency activity in the thalamus. Another model used was the “athalamic” animal in which the thalamus was removed bilaterally. These animals were also hyperactive, reacted to sensory stimuli but could not localize the stimuli, and showed little awareness. Only brief periods of low-amplitude, high-frequency activity were seen in the “athalamic” animal. The implication from these chronic transection studies is that other regions in addition to the RAS can modulate at least some waking. However, it is important to note the amount of time that stimulation of these regions takes to induce waking.

Why is the latency to the induction of a waking EEG important? The implication is that short-latency effects on waking reflect more direct activation of the cortical EEG, whereas long-latency effects reflect a circuitous route for achieving high-frequency EEG activity in the cortex. Typically, stimulation of the RAS, in the region of the pedunculopontine nucleus (PPN) using either electrodes (Moruzzi and Magoun, 1949; Steriade et al., 1991a,b) or optogenetic methods (more on this technology below) activating the locus coeruleus (LC; Carter et al., 2010), will induce high-frequency EEG changes within 1–2 s. (This is a similar latency as that required to induce locomotion on a treadmill following stimulation of the PPN; see Chapter 7.) However, stimulation of the basal forebrain induces high-frequency EEG but only after 15 s of stimulation (Han et al., 2014), and stimulation of the lateral hypothalamus, or optogenetically activated orexin neurons, elicits high-frequency EEG activity only after 20 s of stimulation (Carter et al., 2010; Carter and de Lecea, 2011). Figure 3.1 illustrates these latencies and emphasizes the fact that stimulation of the RAS-thalamic pathway elicits cortical arousal 10 times faster than stimulation of the basal forebrain or lateral hypothalamic/orexin pathways. That is, both regions need to project elsewhere to induce a waking EEG, and, as we will see below, neither region is the final common pathway for arousal.

Wiring Diagram

There are three main nuclei in the reticular activating system (RAS): the locus coeruleus (LC) nucleus, with norepinephrine/noradrenaline (NE/NA)-containing neurons; the dorsal raphe nucleus (RN), with serotonin (5-HT)-containing neurons; and the pedunculopontine nucleus (PPN), with acetylcholine (ACh)- and glutamate (GLU)-containing neurons (Figure 4.1). These nuclei also contain neurons with the inhibitory neurotransmitter gamma-aminobutyric acid (GABA). The dorsal subcoeruleus nucleus (SubCD) is not traditionally considered part of the RAS, but is as critical for generating wake-sleep states as the major nuclei in the RAS. Because LC and RN neurons are most active during waking, have slower firing rates during slow-wave sleep (SWS), and are inhibited during rapid eye movement (REM) sleep, they are sometimes called “REM-off” nuclei. On the other hand, the PPN is most active during waking and REM sleep and the SubCD is most active during REM sleep.