Our general interest is to delineate neural circuitry regulating sleep-wake and motor behaviors as well as neural circuitry underlying neurological sleep-wake and motor disorders such as hypersomnia, coma, cataplexy and Parkinsonism.
Brainstem circuitry regulating REM sleep switch and REM sleep behavior disorder (RBD).
Our previous work (Nature, 2006) demonstrates a flip-flop switch for REM sleep control. The switch is formed by a REM-off GABAergic group in the lateral pontine tegmentum and periaqueductal gray matter inhibit three REM-on cell groups, the glutamatergic parabrachial nucleus controlling cortical arousal and the glutamatergic sublateraodorsal nucleus (SLD) regulating postural muscle atonia, and GABAergic neurons in the SLD, which feedback the REM-off group. One of direct clinical relevance is that lesions of the SLD produce REM without atonia, homolog of human REM sleep behavior disorder (RBD). To demonstrate a specific role for glutamatergic SLD neurons in producing REM without atonia, we recently selectively eliminated glutamate function in the SLD by injecting adeno-associated virus-Cre recombinase (AAV-Cre) into the SLD of mice with lox P sites flanking exon 2 of the vesicular glutamate transporter 2 (VGLUT2) gene. Similarly, we removed GABA function in the SLD and the LPT/PAG in flox-vesicular GABA transporter (VGAT) mice. Consistent with our model that loss of GABA in the REM-off pontine tegmentum doubles REM sleep, loss of GABA in the REM-on SLD reduces REM sleep by half. Removal of glutamate in the SLD produces animal version of RBD.
Since knockout or lesion of the pontine REM-on switch does not eliminate REM sleep, we are looking the other REM sleep switch. Earlier studies (1950-60) showed that transaction caudal to the SLD level eliminates REM sleep, suggesting a potential REM sleep switch in the medullary reticular formation, interaction of which with pontine switch ultimately drive REM sleep. Using cell specific lesions and conditional gene knockout, we found that the GABAergic neurons in the dorsomedial medulla are REM active, loss of VGAT or cells there reduce REM sleep by half. We are now examining effects of double-loss of pontine and medulla REM switches on REM sleep.
There are many paradoxical features in REM sleep motor behaviors that have not been explored. For instance, on the one hand, there is totally atonia in postural muscles (neck and limbs), on the other hand, cranial muscles such as eyes, ears, chins and tongue show prominent phasic (bursting) events during REM sleep. The question is who drives cranial phasic activity. In addition, although we know how the SLD under the normal conditions inhibits motor neurons during REM sleep but we don’t know who drives RBD when the SLD is absent. We are beginning to delineate each of the circuits in the brainstem regulating these motor behaviors. Understanding these circuits will help to battle RBD and even Parkinson’s disease as RBD precedes Parkinson’s disease by about a decade.
Neural substance of the cortical arousal and waking behavior
Although the ascending reticular activating system (ARAS) has been one of the most popular ideas in the sleep field puosed by by Moruzzi and Magoun in 1949, the neural circuitry has never been identified. Upon examining the ascending relay, we found that it was the basal forebrain not the thalamus that is responsible for the cortical arousal. From the basal forebrain we then traced down to the putative source of the ARAS in the parabrachial nuclei. By showing that lesions of these regions caused coma in rats manifested by continuous < 1.0 Hz EEG, we believe that the parabrachia-basal forebrain-neocortical axis is the core of the ARAS. Recently, we ablated glutamate transporter (VGLUT2) gene in the parbarahcial nucleus in mice and found that although it did not produce coma like behavior, it increased 25-30% of sleep, indicating that glutamate is a major neurotransmitter in the parabrachial nucleus mediating arousal. In contrast to general belief, lesions of the thalamus show no significant effects of EEG, Fos and circadian pattern of wakefulness. We hypothesize three parallel pathways all originated in the parabrachial nucleus control cortical arousal and waking behavior. Specifically, the pathway of the parabrachial → basal forebrain → cortex is responsible for cortical arousal; the parabrachial nucleus→ thalamus → cortex is responsible for cognitions of consciousness; finally, we hypothesize that the parabrachial nucleus → posterior lateral hypothalamus (pLH) → cortex controls waking behavior.
Pontine neural circuitry regulating motor behavior, cataplexy and Parkinsonism.
From 1946 to 1966, works from Rhines and Magoun to Shik and colleagues suggest that there is locomotor control region in the pontine tegmentum (Mesencephalic locomotor region, MLR). This region can trigger locomotion when stimulated in decerebrate (midbrain transaction) cats or rats. To this day, the neurons making up the locomotor region and its control mechanism are not known. We did a series of experiments to identify the composition of the locomotor region, their neurotransmitters and projection to the spinal cord and association with cataplexy and orexin receptors by means of combining axonal tracing, in situ hybridization, Fos, chemical stimulation and cell-specific lesions. Our data indicate that (1) the glutamatergic MLR directly activates the spinal central pattern generator to control locomotion and posture. (2) Orexin from the lateral hypothalamus acting on OX2 receptors tonically stimulates the MLR, and loss of orexin signaling in the MLR and emotion triggered dopamine inhibition (via dopamine D2 receptors) of the ventromedial medulla reduce postural muscle tone and produces cataplexy. This hypothesis lead assertion that people with Parkinson’s disease do not prevent cataplexy and further implicate use of D2 antagonist treating cataplexy. We have recently deleted glutamate transporter in the MLR region and confirmed it caused cataplexy.
Circadian and homeostatic control of sleep
Two major mechanisms mainly responsible for sleep-wake cycle are circadian and homeostatic controls. Although the master pacemaker is known in the suprachiasmatic nucleus (SCN), the circuitry how the SCN connects sleep-wake systems still has many gaps to be filled. The circuitry regulating sleep homeostasis is almost unknown. Combing lesion, tracing and gene knockout techniques, we hypothesize that the key circuit for circadian control of sleep-wake behavior is the SCN → ventral subparaventricular zone (vSPZ)→ dordomedial hypothalamic nucleus (DMH) → posterior lateral hypothalamus (pLH)→ cortex; the sleep homeostatic control tested by sleep rebound is via the VLPO, which then project to the pLH, which to the cerebral cortex. Now we are trying to dissect the involved neurotransmitters or neuroppetides in these neural circuits.
|
