Histamine: from flop to flip-flop

Abstract

When you eliminate the impossible, whatever remains, no matter how improbable, must be the truth. It's when the entirely possible is mistakenly eliminated that the problems start. A longstanding puzzle in the field of sleep research has been put to bed by Liu and colleagues in this issue of The Journal of Physiology (Liu et al. 2010). They re-evaluate the actions of histamine in the GABAergic, sleep-active ventrolateral preoptic area (VLPO) and find that the most attractive answer was the right one after all. The VLPO is one of several sleep-promoting regions that innervate arousal nuclei. Sleep-active neurons begin firing prior to behavioural sleep and, in the case of VLPO, increase their firing with sleep depth. Through the release of GABA (and probably galanin) they are thought to quieten neural activity in brain regions associated with wakefulness, thereby allowing the transition into sleep. Saper and colleagues have advanced this theory of sleep–wake control with their ‘flip-flop’ proposal (Saper et al. 2001). This describes the neuronal pathways of sleep and wakefulness as opposing sides of a bi-stable switch. A key element of this model is that these pathways reciprocally inhibit each other: a high level of activity on one side is self-reinforcing, and pushes the switch in that direction. In keeping with this theory, VLPO and other sleep-active regions receive projections from brain arousal nuclei: for example, the histaminergic tuberomammillary nucleus, the noradrenergic locus coeruleus and serotonergic dorsal raphe nucleus. However, simply demonstrating innervation of VLPO by arousal regions does not fully address whether reciprocal inhibition occurs. A previous electrophysiological characterisation of sleep-active VLPO neurons (Gallopin et al. 2000) examined the effects of wake-active neurotransmitters, and convincingly showed inhibitory actions of noradrenaline, acetylcholine and serotonin on the firing rates of these neurons. However, histamine was reported (data not shown) to have had no effect. This was surprising, given the well-established role of histamine in the maintenance of arousal, the dense histaminergic innervation of VLPO and the present confirmation (Liu et al. 2010) that histamine infusion into this region causes an increase in wakefulness. Liu et al. revisit this experiment and, by using longer application times, show clear inhibitory effects of histamine on the firing of sleep-active VLPO neurons. The requirement for longer perfusions may be because histamine exerts its effects indirectly by increasing the frequency of GABAergic IPSCs. Sleep-active VLPO neurons show a direct hyperpolarising response to noradrenaline, but another smaller population of bipolar VLPO neurons are excited by this neurotransmitter. As Liu and colleagues show, this subpopulation is also directly excited by histamine. Their working model is that histamine (acting via H1 and H2 receptors) excites bipolar VLPO neurons causing increased GABA release onto sleep-active VLPO neurons, thereby inhibiting them indirectly. In the present study, Liu et al. have mainly focused on GABAA receptor-mediated inhibition, showing that gabazine largely reverses the effects of histamine on firing rate. However, GABAB receptors may also play a role because a GABAB antagonist ({type:entrez-protein,attrs:{text:CGP52432,term_id:875421701,term_text:CGP52432}}CGP52432) partly reduces the effects of histamine. The authors suggest the effect is postsynaptic, but the possibility of presynaptic actions should be explored. The physiological impact of GABAB receptor-mediated inhibition of VLPO neurons may be significant. This slow form of GABAergic inhibition is critical in orexinergic neurons (Matsuki et al. 2009), which are implicated in the sleep disorder narcolepsy, but has not yet been explored in many other regions associated with sleep and arousal. Whether other wake-active neurotransmitters may also use this indirect GABAergic pathway is unknown. Low concentrations of noradrenaline actually reduce GABA release onto sleep-active VLPO neurons via presynaptic α2 adrenoceptors (Matsuo et al. 2003), which appears contradictory. However, this study examined the responses of mechanically dissociated neurons. Given that Matsuo et al. were also able to show excitatory actions of noradrenaline on bipolar VLPO neurons, this neurotransmitter may be activating the same pathway as histamine. Examination of dissociated neurons means that a network action, if present, would not have been detected and also means the significance of the observed presynaptic inhibition is unclear. Alternatively, noradrenergic and histaminergic modulation of these neurons may indeed be non-overlapping. Clearly, there are many unsolved issues in the neuronal control of sleep and wakefulness. But in demonstrating that histamine does in fact inhibit sleep-active VLPO neurons, Liu et al. have added further weight to the ‘flip-flop’ theory and made sense of a previously confusing observation.