Spontaneous spikelets and oscillations in a SubC neuron that were blocked by mefloquine (MEF).¹⁷¹ A. Voltage clamp recording (HP=-50mV) with spontaneous spikelets and excitatory post-synaptic currents (EPSCs). Expanded record at a. depicts both spontaneous EPSCS and underlying spikelets in this cell. Expanded record at b. shows the inhibition of EPSCs following the addition of fast synaptic blockers (CNQX, APV, gabazine, abbreviated as CAG, and the calcium channel blocker cadmium, abbreviated as Cd). Note the continued presence of rhythmic spikelets. Expanded record at c. depicts a complete cessation of PSCs and spikelets after the addition of mefloquine (25 uM). B. Power spectrum analysis showing the rhythmicity present in this cell at baseline (Control, gray diamonds), after CAG+Cd (blue triangles) and after mefloquine (MEF, black squares). Note the decrease in power after the addition of CAG+Cd suggesting that at least some of the previously observed PSCs were also synchronized at this frequency. C. Holding potential necessary to maintain this cell at −50mv. Note the lack of change in holding potential after the addition of MEF, suggesting that this agent does not affect ionic conductances.

The developmental decrease in REM sleep and transmitter changes. A. Adapted from,⁷⁸ to show the decrease in REM sleep from ∼80% of sleep time to ∼10% between 10 and 30 days in the rat, after which point it remains stable. B. Changes in the effects of transmitter systems on the membrane potential of PPN neurons. The y axis shows the mean depolarization above resting membrane potential (RMP, 0 mV) or hyperpolarization below RMP. Between 12 and 21 days, the same concentration of 1) NMDA was found to decrease from +12 mV to + 4 mV (red line)¹¹⁸; 2) KA was found to increase from + 4 mV to + 13 mV (blue line)¹¹⁸; 3) a cholinergic agonist increased hyperpolarization from −8 mV to −10 mV¹²⁷; 4) a α2 adrenergic agonist decreased hyperpolarization from −6 mV to −2 mV¹²¹; 5) a 5-HT1 agonist increased hyperpolarization from −2 mV to −7 mV¹⁰¹; 6) a 5-HT2 agonist did not change the degree of hyperpolarization (-2 mV)¹⁰¹; 7) a GABAa agonist first depolarized (+3 mV) then hyperpolarized (-1 mV)¹³¹; and 8) a GABAb agonist decreased hyperpolarization from −5 mV to −1 mV.¹³¹

Developmental changes in Cx36 protein expres-sion.^170,171 A. Western blots comparing Cx36 protein levels at different ages. 15 ug of protein was loaded in each lane. Each sample contains punches from 2-3 littermates. Blots were stripped and re-probed with anti-β tubulin antibodies to verify equal protein loading. All brain regions showed a decrease in Cx36 protein during development. B. Western blot comparing Cx36 expression in different brain regions. Equal amounts of protein from the samples in A were pooled and run on a single gel to allow comparisons to be made between brain regions. The blot was stripped and re-probed for β tubulin as in A. The PPN and inferior colliculus (IC) had the highest levels of Cx36 in 10 day pups, whereas the ventrobasal thalamus (VB) had the lowest. C. Chart showing quantification of data in B. Cx36 protein was quantified and normalized to β tubulin. At 30 days, Cx36 levels in all three nuclei of the RAS dropped to 25% of the level at 10 days. In contrast, VB dropped to 50% and the IC to 30%. D. Chart showing quantification of data in A and B. Cx36 protein was quantified and normalized to β tubulin. Error bars show standard deviation. Cx36 levels decreased at different rates in different brain regions, IC- inferior colliculus, Pf- parafascicular nucleus, PPN- pedunculopontine nucleus, SubC- subcoeruleus nucleus, VB- ventrobasal thalamus.

System Schematic.^170,171 We hypothesize that the Pf, PPN and SubC are organized similarly to the specific relay and reticular nuclei of the thalamus. Thalamic relay neurons (TRNs) in specific nuclei project to the cortex and send collaterals to inhibitory GABAergic interneurons (which are electrically coupled) in the reticular nucleus that in turn inhibit TRNs and local circuit neurons. Cortical projections return to the TRNs, local circuit cells, and to inhibitory GABAergic interneurons. Ascending cholinergic input from the PPN excites TRNs and inhibits reticular cells. The manner in which this circuit generates thalamocortical oscillations was reviewed recently.¹³⁹ For SubC, our data suggests that descending projections from the PPN excite, possibly via M1 muscarinic receptors, at least some GABAergic, coupled interneurons with LTS (could be REM-on cells), and inhibit, possibly via M2 muscarinic receptors, SubC output neurons (perhaps glutamatergic) and perhaps some LTS neurons (unknown transmitter, may be PGO-burst cells). In turn, GABAergic interneurons may inhibit output neurons (we have little evidence on local circuit cells in SubC). This circuit may be involved in generating oscillations characteristic of REM sleep. In Pf, we speculate that ascending projections from the PPN excite (perhaps via M1 receptors) GABAergic coupled interneurons and inhibit (perhaps via M2 receptors) Pf output neurons (may be Glutamatergic). In turn, GABAergic interneurons inhibit these output neurons (and local circuit). This circuit may help generate thalamocortical oscillations related to arousal. In PPN, we found that cholinergic inputs inhibit type II mostly cholinergic PPN, perhaps output, neurons (via M2 receptors) and excite (via M1 receptors) type I (LTS) and III (IA+LTS) mostly non-cholinergic (may be GABAergic, some may be coupled) neurons. GABAergic interneurons may inhibit output neurons. This circuit may help generate rhythms related to waking and REM sleep. In support of this model, we have considerable evidence showing that virtually all of the coupled neurons in SubC are GABAergic. We also showed that some coupled neurons are depolarized by CAR and most of these are GABAergic, and others are hyperpolarized by CAR and most of these are non-GABAergic. A few non-coupled GABAergic cells are hyperpolarized by CAR, i.e. there may be two populations of GABAergic cells, one coupled, the other not. These suggestions lay the groundwork for the exploration of a novel mechanism for sleep-wake control based not only on known transmitter interactions, but also on superimposed ensemble activation provided by electrical coupling.