Vibrissal thalamic modes
|Manuel Castro-Alamancos (2010), Scholarpedia, 5(7):7278.||doi:10.4249/scholarpedia.7278||revision #150522 [link to/cite this article]|
A thalamic mode (or thalamocortical mode) refers to a particular arrangement of the response properties of the thalamocortical network that gives rise to a distinct input-output function. Since the thalamus controls the flow of information to the neocortex, a thalamic mode adjusts how thalamocortical cells relay sensory and corticothalamic information. Thalamic modes can change on a moment-to-moment basis due to the actions of neuromodulators and/or excitatory and inhibitory inputs. In the vibrissal system, thalamic modes lead to robust changes on how sensory information from the vibrissa is processed temporally, integrated spatially and relayed to the neocortex.
Components of the vibrissa thalamocortical network in the VPM thalamus
The vibrissa thalamocortical network consists of two distinct nuclei. The main thalamocortical nucleus is the ventroposterior medial thalamus (VPM). In addition, cells in the medial sector of the posterior complex (POm) also project to the cortex. Here we will focus on VPM. Thalamocortical cells are at the center of a neuronal network that involves sensory, cortical and modulatory inputs. In VPM, thalamocortical cells form clusters, called barreloids, that project to clusters of cells in layer 4 of somatosensory cortex, called barrels. Thalamocortical cells in VPM receive signals from four main sources (see Figure 1):
- Lemniscal sensory fibers originating from clusters of cells, called barrelettes, in the principal trigeminal nucleus (PR5) provide sensory signals.
- Corticothalamic fibers originating in layer 6 of barrel cortex provide cortical feedback and top-down influences.
- Fibers from nucleus reticularis thalamic (NRT) cells provide the main inhibitory control of thalamocortical cells because there are no inhibitory interneurons in VPM.
- Finally, a variety of fibers that originate mainly in several brainstem nuclei provide neuromodulator inputs. These influences can directly affect thalamocortical cells or they can affect the other sources of inputs (NRT, corticothalamic, and Pr5 cells), which will indirectly affect thalamocortical cells.
As discussed later, vibrissa thalamic modes are defined by the properties of these network inputs, which are modified on a moment to moment basis by neuromodulators acting locally or in afferent structures (Castro-Alamancos, 2004a). Considering their role in setting network modes, we first discuss neuromodulators.
Neuromodulators set network modes
Neurotransmitters often act through ionotropic receptors (ligand-gated channels), while neuromodulators act through metabotropic receptors (G-protein coupled). The effects of neurotransmitters acting on ionotropic receptors are usually phasic, lasting only 10’s of milliseconds. The effects of neuromodulators acting on metabotropic receptors are usually slower and longer lasting, in the range of 100's of milliseconds to seconds or more. However, the distinction between a neurotransmitter and a neuromodulator can be rather arbitrary and most neuroactive substances can function as both. For instance, a substance acting as a neuromodulator can alter the properties of ion channels that are activated by the same substance acting as a neurotransmitter (e.g. by affecting channel opening probabilities, receptor desensitization, release). Substances acting on ionotropic receptors may also (appear to) act as neuromodulators if the presynaptic neuron fires continuously in a sustained manner.
A number of substances are well-known neuromodulators and some of these have significant actions in the thalamocortical network:
- Glutamate acts on ionotropic receptors (AMPA, NMDA, kainate) and on metabotropic receptors (mGluR1-8). Glutamate is released by lemniscal and corticothalamic synapses. It appears that metabotropic receptors are selectively activated by corticothalamic synapses innervating NRT and thalamocortical cells (McCormick and von Krosigk, 1992), and not by lemniscal synapses (Castro-Alamancos, 2002a).
- GABA acts on ionotropic receptors (GABAA) and metabotropic receptors (GABAB). GABA is released by NRT cells.
- Norepinephrine is a catecholamine acting on α and β type metabotropic receptors. Noradrenergic neurons are found in the locus coeruleus in the brainstem reticular formation, from where they project throughout the brain, including the thalamus. Noradrenergic neurons discharge robustly during high levels of vigilance and attention, reduce their firing during slow-wave sleep and stop firing during paradoxical (also called, “rapid eye movement”; REM ) sleep (Foote et al., 1980).
- Dopamine is a catecholamine acting on D1 and D2 type metabotropic receptors. So far, there is little evidence of any role of dopamine in the vibrissa thalamocortical network.
- Histamine acts on H1-H4 type metabotropic receptors. Histamine neurons are found in the posterior hypothalamus, in the tuberomammillary complex, from where they project throughout the brain, including the thalamus. Histaminergic neurons discharge robustly during wakefulness (Brown et al., 2001).
- Serotonin acts on ionotropic (5-HT3) and metabotropic (5-HT1, 5-HT2 5-HT4, 5-HT5, 5-HT6, 5-HT7) receptors. Serotonin neurons are found in the raphe nuclei in the brainstem reticular formation, from where they project throughout the brain, including the thalamus. Similar to noradrenergic neurons, 5-HT neurons fire tonically during wakefulness, decrease their activity in slow-wave sleep, and are nearly quiet during paradoxical sleep (McGinty and Harper, 1976; Trulson and Jacobs, 1979).
- Acetylcholine acts on ionotropic (nicotinic) and metabotropic (muscarinic) receptors. Acetylcholine neurons projecting to the thalamus are found in the pedunculopontine nuclei (PPT) and in the dorsolateral tegmental nuclei (LDT) in the brainstem. Acetylcholine neurons are also found in the basal forebrain from where they project to the neocortex. Cholinergic neurons in the LDT/PPT complex discharge vigorously during paradoxical sleep and also during wakefulness (el Mansari et al., 1989), and the levels of acetylcholine increase in the thalamus during those states (Williams et al., 1994).
- Neuropeptides usually act at metabotropic receptors. Neurons very often make both a conventional neurotransmitter (glutamate, GABA) and one or more neuropeptides. Examples include opioids (endorphins, enkephalins, dynorphins), substance P, etc.
- Hormones are chemicals released by cells that affect cells in other parts of the organism generally through the bloodstream. For example, epinephrine (adrenaline) is a catecholamine that is released by the adrenal gland.
- Other intrinsic neuroactive substances released within the thalamus may include adenosine, cannabinoids, growth factors, cytokines, etc.
- Extrinsic neuroactive substances that reach the thalamus may also affect thalamic modes. For example, nicotine from tobacco, caffeine from coffee, etc.
Trigeminal complex cells provide sensory inputs and sensory (lemniscal) synapses act as low-pass filters
Trigeminal complex cells projecting to VPM give rise to lemniscal synapses that form specialized glomeruli (for a review see (Castro-Alamancos, 2004a)). Lemniscal synapses release the neurotransmitter glutamate and trigger EPSPs in thalamocortical cells by activating primarily AMPA receptors and may also activate NMDA receptors to some extent, but they do not seem to activate mGLUR receptors. Lemniscal EPSPs are highly specialized for the effective transmission of information. They have short latencies, fast rise times, large amplitudes, are highly reliable, and depress at frequencies above 2 Hz (Castro-Alamancos, 2002a). The short latencies and fast rise times reflect the thickness and myelination of lemniscal fibers, and the fact that the synapses they form are electrotonically close to the soma. The large amplitude and security at low frequencies and the depression at high frequencies reflect a large number of release sites at synaptic glomeruli, and a high release probability (i.e. chance that a release site will fuse a vesicle and excrete neurotransmitter to the synaptic cleft) at those sites during low frequency inputs. Release probability is likely reduced during high frequency presynaptic activity. The depression of lemniscal synapses acts as a low-pass filter, enabling the relay of low-frequency sensory inputs under most circumstances, while the relay of high-frequency sensory inputs is subject to the membrane potential (Vm) of thalamocortical cells (Castro-Alamancos, 2002b, a) (see Figure 1). The efficacy (synaptic strength) of lemniscal synapses per se is not affected by certain neuromodulators, such as acetylcholine and norepinephrine (Castro-Alamancos, 2002a). But these neuromodulators can impact the transmission of lemniscal inputs by affecting the membrane potential of thalamocortical cells postsynaptically.
Layer 6 cells provide corticothalamic feedback and corticothalamic synapses act as high-pass filters
Upper layer 6 cells located in barrel cortex project to VPM and leave a collateral fiber in NRT. Thalamocortical cells in VPM and layer 6 corticothalamic cells form closed-loops for the flow of information between a thalamic barreloid and a cortical barrel column (Bourassa et al., 1995). For each thalamocortical fiber ascending to barrel cortex there are many more corticothalamic fibers coming back to thalamus. Corticothalamic fibers form corticothalamic synapses that release glutamate and trigger EPSPs in thalamocortical and NRT cells by activating AMPA, NMDA and mGLUR receptors (Golshani et al., 2001). Corticothalamic EPSPs mediated by ionotropic receptors are very different compared to lemniscal EPSPs. They have long latencies, slow rise times, small amplitudes, are unreliable, and facilitate at frequencies above 2 Hz (for a review see (Castro-Alamancos, 2004a)). The long latencies and slow rise times reflect the thinness and sparse myelination of corticothalamic fibers, and the fact that the synapses they form are located in distal dendrites; electrotonically far from the soma. The small amplitude and low security at low frequencies and the facilitation at high frequencies reflect a small number of release sites per synapse (estimated to be 1), and a low release probability at those sites during low frequency inputs that sharply increases during high frequencies inputs (see Figure 2). Corticothalamic synapses display a robust form of LTP when stimulated repetitively at relatively high frequencies (10 Hz and above), while LTD is also induced when repetitive stimulation occurs at low frequencies (1 Hz), thereby providing mechanisms for bidirectional changes in synaptic efficacy (Castro-Alamancos and Calcagnotto, 1999). Corticothalamic EPSPs mediated by mGLUR receptors are triggered by high-frequency stimulation and produce a long-lasting slow depolarization (McCormick and von Krosigk, 1992). The efficacy of corticothalamic synapses is suppressed by some neuromodulators, such as acetylcholine and norepinephrine (Castro-Alamancos and Calcagnotto, 2001). The amplitude of EPSPs evoked in NRT neurons by stimulating single corticothalamic fibers is several times larger than those evoked in thalamocortical neurons, and the number of GluR4-receptor subunits at these synapses may provide a basis for the differential synaptic strength (Golshani et al., 2001). The stronger corticothalamic EPSPs on NRT cells assures that low-frequency corticothalamic activity drives NRT cells and triggers robust feedforward inhibition in VPM thalamocortical cells.
NRT cells provide inhibitory inputs
NRT cells project to thalamocortical cells in VPM and are also thought to influence each other through inhibitory collateral fibers (Shu and McCormick, 2002; Sohal and Huguenard, 2003) and gap junctions (Landisman et al., 2002). NRT synapses are inhibitory, release GABA and trigger IPSPs in thalamocortical cells by activating GABAA and GABAB receptors. GABAA receptors are activated by the amount of GABA released by a single action potential in an NRT fiber, while GABAB receptor activation appears to require more GABA, usually released by bursts of action potentials (Kim and McCormick, 1998). Similar to thalamocortical cells, NRT cells have two intrinsic firing modes: burst and tonic firing (Steriade et al., 1997; Hartings et al., 2003). The relay of sensory information to the neocortex by thalamocortical cells is strongly influenced by the TRN. Thus, TRN is probably the main gatekeeper of the neocortex (Crick, 1984). The reciprocal synaptic connectivity between TRN and thalamocortical cells in VPM is critical for generating normal and abnormal rhythmic activities, such as spindle oscillations (see movie below) and absence seizures (Beenhakker and Huguenard, 2009).
Spontaneous spindle oscillations recorded from a thalamocortical cell in a VPM slice
The movie below shows an intracellular recording from a mouse slice in vitro containing intact connections between VPM and NRT. The recorded cell produced spontaneous spindle oscillations that recurred every 5-10 secs. Note that there are two gaps in the movie. Media:Vibrissalthalamicmodes-spin1.swf
Thalamocortical cells provide the output to neocortex via thalamocortical synapses
Thalamocortical cells have two intrinsic firing modes (see Figure 3): tonic and burst firing modes. Bursts are due to the Ca2+ conductance of T type channels, which gives rise to the low-threshold calcium current (IT) (Steriade et al., 1997). Because of its voltage-dependence, this conductance in thalamocortical cells is controlled by inhibitory inputs from NRT cells. In particular, when an NRT neuron fires a burst of action potentials, it activates both GABAA and GABAB receptors in thalamocortical cells, which hyperpolarizes these neurons and deinactivates low-threshold T-type Ca2+ channels, enabling thalamocortical neurons to produce regenerative calcium spikes that can trigger bursts of action potentials. In response to NRT inhibition, thalamocortical cells fire with a delay on a postinhibitory rebound, and the stronger the inhibition the stronger the rebound excitation. Even without NRT inhibition, if thalamocortical cells are sufficiently hyperpolarized, resulting in the deinactivation of the low-threshold calcium current, excitatory inputs (corticothalamic or lemniscal) can directly drive regenerative calcium spikes that can trigger bursts of action potentials. Moreover, when thalamocortical cells are more depolarized, resulting in the inactivation of the low-threshold calcium current, thalamocortical cells enter the tonic firing mode and produce single action potentials, instead of bursts, in response to excitatory inputs.
The output of the thalamocortical cells is transmitted by thalamocortical synapses that reach the barrel cortex and terminate in layer 4. As thalamocortical fibers ascend to layer 4 they leave fiber collaterals in NRT and in layer 6. In layer 4, thalamocortical fibers produce thalamocortical synapses that release glutamate and trigger EPSPs on cortical neurons.
What determines a thalamocortical mode?
A thalamocortical mode is a particular arrangement of the properties of the thalamocortical network components that gives rise to a distinct input-output function. The properties that are most commonly affected to determine a thalamocortical mode include the membrane potential (Vm), intrinsic firing mode, intrinsic excitability and the strength of synaptic inputs. Neuromodulators act directly on thalamocortical cells and on afferent synapses within the thalamus to change these properties. Neuromodulators may also indirectly influence thalamocortical cells by affecting the activity of the main excitatory and inhibitory inputs, within the trigeminal complex, layer 6 and NRT, respectively. Activity in these afferents changes their strength via short-term synaptic plasticity but may also affect the Vm, firing mode, and intrinsic excitability of thalamocortical cells. Thus, thalamocortical modes are set in a complex way, through direct and indirect effects of neuromodulators affecting several variables. The main variables determining a thalamocortical mode are:
- Vm of thalamocortical cells: This critical variable is highly dynamic because it is affected by most, if not all, neurotransmitters and neuromodulators present in the thalamocortical network.
- Intrinsic excitability: Neurons express a number of voltage-dependent conductances that endow them with different response properties. For example, thalamocortical cells are characterized by strong hyperpolarization-activated cation currents (IH) and low-threshold calcium currents (IT). These currents are not only affected by Vm but can be directly affected by many neuromodulators. Excitatory and inhibitory inputs can affect the intrinsic excitability of thalamocortical cells by changing the Vm and engaging voltage-dependent currents. In addition, when the resting Vm of thalamocortical cells is at the reversal of incoming synaptic inputs, the increased conductance produced by the synaptic inputs can affect the integrative properties of the cell, without changing the Vm, by shunting the membrane (e.g. shunting inhibition).
- Thalamocortical firing mode and rate: Thalamocortical cells are characterized by two distinct firing modes: bursting and tonic. These firing modes are set primarily by the Vm of thalamocortical cells (see Figure 3). Thus, factors that influence Vm also determine firing mode. During bursting, thalamocortical cells produce a cluster of action potentials (usually between 3-6 action potentials) at very high frequencies (> 100 Hz) riding on the low-threshold calcium spike. However, thalamocortical cells are limited by how fast they can produce bursts because of the dependence of bursts on the low-threshold calcium current, which must be deinactivated by hyperpolarization. Thus, cells can usually burst at <15 Hz. In contrast, during tonic firing, cells can produce action potentials at much higher constant firing rates.
- Activity and strength of excitatory and inhibitory afferents: Thalamocortical cells in VPM receive excitatory (glutamatergic) and inhibitory (GABAergic) afferent inputs from three main sources: lemniscal synapses from trigeminal nucleus (mostly Pr5), corticothalamic synapses from layer 6 of barrel cortex and inhibitory synapses from NRT. Activity in the afferent cells drives their synapses and can change the Vm of thalamocortical cells, which can result in changes in intrinsic excitability and intrinsic firing mode. In addition, some of these afferents can activate metabotropic receptors leading to a modulator action on thalamocortical cells. For example, glutamate released from corticothalamic synapses can activate mGLUR, and GABA released from inhibitory synapses can activate GABAB receptors. The frequency and pattern of activity in the synaptic afferents also sets the strength of these synapses by affecting short-term synaptic plasticity and temporal integration. For example, high-frequency activity in corticothalamic synapses will increase release probability at these synapses and enhance the strength of this pathway. In contrast, activity in Pr5 cells will depress lemniscal synapses and decrease the strength of this pathway. Moreover, neuromodulators released in the thalamus can directly affect the efficacy of excitatory and inhibitory neurotransmission rather selectively. For example, acetylcholine and norepinephrine depresses corticothalamic but not lemniscal synaptic strength (Castro-Alamancos and Calcagnotto, 2001; Castro-Alamancos, 2002a).
The main thalamocortical modes
Slow oscillation mode during slow-wave sleep and anesthesia
- When does it happen: The slow oscillation or quiescent mode is considered here as a broad baseline state during which active sensory processing per se does not occur because animals are either sleeping, inattentive/drowsy or anesthetized. While there may well be different modes within these states, we encompass them here within a single mode for simplicity. In this mode, slow synchronous oscillations are common, particularly when animals are sleeping in non-REM sleep. In addition, this mode can be induced by surgical anesthesia, which is when most electrophysiological studies take place. During non-REM sleep, slow wave oscillations occur often but are most prevalent in the deeper stage(s) (typically referred as stage 3 or 3/4). In less deep stages of sleep, slow oscillations can occur interposed with other rhythms, such as spindle oscillations. Similar to non-REM sleep stages, there are also stages of anesthesia. During the surgical anesthesia stage, the slow oscillation mode is evident but can vary significantly depending on the level or plane of surgical anesthesia. For example, the frequency of slow oscillations can vary quite significantly within this mode. Moreover, some anesthetics will tend to produce more rhythmic activity of a certain frequency range than others. Thus, the depth of anesthesia and the specific effects of the anesthetic used are critical at setting the particular characteristics of this mode.
- Spontaneous activity: During slow-wave sleep and surgical anesthesia thalamocortical cells fire at low frequencies producing either bursts or single spikes. This slow activity in thalamocortical cells can be driven by ongoing slow oscillations generated intrinsically in the neocortex (also known as Up and Down states)(Steriade et al., 1997; Rigas and Castro-Alamancos, 2007). But slow activity can also be driven by intrinsic currents in thalamocortical cells and persist in the absence of corticothalamic activity (Hughes et al., 2002; Rigas and Castro-Alamancos, 2007). Thus, full expression of slow oscillations in thalamocortical cells appears to require both thalamic and cortical oscillators (Crunelli and Hughes, 2010). During the slow oscillation mode, thalamocortical cells are usually fairly hyperpolarized close to the reversal potential of K+ (Down), and they may transition for short periods of time to more depolarized states due to synaptic bombardment, usually produced by spontaneous corticothalamic and NRT activity (Up) (Steriade et al., 1997). In this situation, nRt cells can burst and drive strong IPSPs in thalamocortical cells. The hyperpolarization caused by the IPSPs deinactivates IT and activates IH in thalamocortical cells. This sets up thalamocortical cells so that at the outset of the IPSP a rebound depolarization occurs caused by activation of IT. The rebound triggers a burst of action potentials in thalamocortical cells that feedback to NRT and cortex. Such an interplay between NRT and thalamocortical cells repeated in a sequence at 5-12 Hz is responsible for the generation of spindle oscillations that recur every few seconds (McCormick and Bal, 1997). Spindles are waxing and waning rhythms with dominant frequencies of 7–14 Hz, grouped in sequences that last 1–3 sec and recur periodically at 0.1-0.2 Hz (see movie above). Spindle oscillations are common during the slow oscillation mode and are prominent at sleep onset, during loss of awareness, and are prevalent during barbiturate anesthesia, which enhances inhibitory efficacy. Apart from the occasional spindle oscillations, thalamocortical activity in VPM during this state is of low frequency (<1 Hz) (Castro-Alamancos, 2002b; Aguilar and Castro-Alamancos, 2005; Hirata et al., 2006).
- Frequency-dependent sensory responses (rapid sensory adaptation) : Sensory responses driven by whisker stimulation during the slow oscillation mode are of high probability as long as the stimulus is delivered at low frequencies. As soon as the vibrissa stimulus augments in frequency, the thalamocortical response is strongly depressed (Castro-Alamancos, 2002b). Thus, thalamocortical neurons follow high-frequency whisker stimulation with great difficulty in the slow oscillation mode; thalamocortical sensory responses are low-pass filtered. Intracellular recordings in urethane anesthetized rats during the slow oscillation mode show that whisker stimulation evokes EPSP–IPSP sequences in thalamocortical neurons (see movie below), and both the EPSPs and IPSPs depress with repetitive whisker stimulation at frequencies above 2 Hz (Castro-Alamancos, 2002b). The underlying cause of this low-pass filter is the frequency-dependent depression of lemniscal synapses (Castro-Alamancos, 2002a).Thus, sensory inputs at frequencies above 2 Hz reduce the efficacy of lemniscal synapses, which drives the lemniscal EPSP away from the discharge threshold of the cell resulting in a low probability of firing for thalamocortical cells. However, as described below, a major impact of activated states is to change this low-pass filtering. In addition, feedback inhibition from NRT, driven by sensory inputs, also contributes to the low-pass filtering of sensory inputs during the slow oscillation mode (Castro-Alamancos, 2002b). This effect is most notable at the beginning of a high-frequency sensory stimulus train, when IPSPs are more robust and produce a stronger hyperpolarization. However, less effect of feedback inhibition is observed for sensory responses occurring at the end of a long high-frequency train (Castro-Alamancos, 2002b; Hirata et al., 2009). Those responses are mostly depressed by lemniscal synaptic depression with little contribution of synaptic inhibition from NRT.
- Spatial dependent sensory responses (receptive fields and selectivity) : Excitatory receptive fields of VPM cells consist of an excitatory center, the principal whisker (PW), and an excitatory surround, the adjacent whiskers (AWs). For low-frequency sensory inputs, during the slow oscillation mode, the response to the PW (receptive field excitatory center) is much stronger and faster than the response to AWs (receptive field excitatory surround) (Aguilar and Castro-Alamancos, 2005). As mentioned above, for high-frequency sensory inputs, both PW and AW responses are depressed because of the low-pass filtering at the lemniscal pathway. Simultaneous stimulation of the PW and several AWs (i.e. multiwhisker stimulation) produces a response in thalamocortical cells that matches the PW, as if the AWs had not been stimulated (Aguilar and Castro-Alamancos, 2005; Hirata et al., 2006). Interestingly, simultaneous multiwhisker responses are distinguishable from PW responses in the next stage of processing, the barrel cortex (Hirata and Castro-Alamancos, 2008). The size of the receptive field measured as the number of whiskers that evoke a response depends on the level of anesthesia within this mode (Friedberg et al., 1999; Aguilar and Castro-Alamancos, 2005).
- Corticothalamic feedback: The amplitude of corticothalamic EPSPs is relatively small during low frequency corticothalamic activity because corticothalamic synapses have a low release probability and occur at distal portions of the dendritic tree. However, during high frequency corticothalamic activity (>5 Hz) the probability of release at these synapses sharply increases due to synaptic facilitation producing large amplitude EPSPs that can be as powerful than those produced by lemniscal sensory afferents. Thus, the corticothalamic pathway is an activity dependent driver of thalamocortical activity, which can be demonstrated by using electrical stimulation to stimulate corticothalamic fibers (Castro-Alamancos, 2004a). However, it is not clear when corticothalamic cells discharge at high-frequencies to engage the activity dependent facilitation of corticothalamic synapses.
In vivo intracellular recording of a VPM cell during the slow oscillation mode
Spontaneous and whisker-evoked intracellular activity of a VPM thalamocortical cell in a urethane-anesthetized rat. The 20-sec of continuous recording shown in the panel below displays spontaneous spikes and large amplitude lemniscal EPSPs, and two instances of whisker stimulation that evoke the lemniscal EPSPs and feedback IPSPs from NRT. The insets that appear during whisker stimulation show a close-up of responses evoked by the first and tenth stimulus in a 10 Hz train. During the recording the thalamus is in the slow oscillation mode and the VPM cell only evokes spikes (sometimes) to the first stimulus in the whisker train. Note also the large amplitude IPSP to the first stimulus in a train and the depression of both the IPSP and EPSP at 10 Hz. (See (Castro-Alamancos, 2002b) for details) Media:Vibrissalthalamicmodes-VPMvivo4.swf
Activated modes during arousal and BRF stimulation
- When does it happen: The activated mode is typical when animals are awake during arousal, and it is most robust when animals are in a state of vigilance during attentive processing, such as during performance in a behavioral task (Castro-Alamancos, 2004b). A somewhat similar activated mode to that observed during waking occurs when animals enter REM or paradoxical sleep. The activated mode can be induced in anesthetized animals that are in a slow oscillation mode by electrically stimulating the brainstem reticular formation (BRF), and this is a useful method to determine the impact of the activated mode on sensory thalamocortical responses because it allows to compare slow oscillation and activated sensory responses in the same neurons (Moruzzi and Magoun, 1949; Castro-Alamancos, 2002b; Castro-Alamancos and Oldford, 2002; Aguilar and Castro-Alamancos, 2005).
- Spontaneous activity: Spontaneous thalamocortical activity can vary from nil to high frequency tonic firing. During BRF stimulation, thalamocortical activity consists of high frequency tonic firing, which outlasts the stimulation by several seconds (Castro-Alamancos, 2002b; Castro-Alamancos and Oldford, 2002). Bursts in this state are uncommon although they may occur as thalamocortical cells transition from the slow oscillation mode to the activated mode.
- Frequency-dependent sensory responses (rapid sensory adaptation) : In contrast to the slow oscillation mode, during the activated mode (see Figure 4), low-frequency sensory responses are a bit stronger (increase slightly in probability) and become faster (evoked spikes display shorter latencies) (Castro-Alamancos, 2002b; Aguilar and Castro-Alamancos, 2005). But the most robust change occurs at the level of high-frequency sensory processing. Thus, during the activated mode thalamocortical cells robustly enhance their responses to high-frequency sensory signals, virtually eliminating the low-pass filtering typical of the slow oscillation mode (Castro-Alamancos, 2002b). These effects are similar to those produced by cholinergic activation of the thalamus, as discussed below.
- Spatial dependent sensory responses (receptive fields and selectivity) : Excitatory receptive fields of VPM cells consist of an excitatory center, the PW, and an excitatory surround, the AWs. Activation produced by BRF stimulation in anesthetized animals enlarges the excitatory surround of VPM cells (Aguilar and Castro-Alamancos, 2005). Thus, for low-frequency sensory inputs, during quiescent states, the response to the PW (receptive field excitatory center) is much stronger than the response to most AWs (receptive field excitatory surround), but during activation, there is an enhancement of the response to AWs, which can reach response levels similar to the PW. For high-frequency sensory inputs, during quiescent states, both PW and AW responses are depressed because of the low-pass filtering at the lemniscal pathway. However, during the activated mode, there is a significant increase in both PW and AW responses so that they become similar, but PW responses are generally stronger than AW responses at high frequencies (Aguilar and Castro-Alamancos, 2005).
- Corticothalamic feedback: During the activated mode, produced by BRF stimulation in anesthetized animals, corticothalamic responses are further high-pass filtered (i.e. only high-frequency corticothalamic activity is allowed) (Castro-Alamancos and Calcagnotto, 2001). This effect is mimicked by specific neuromodulators (i.e. norepinephrine), as described below.
Activated modes produced by specific neuromodulators
Application of specific neuromodulators into the thalamus leads to characteristic activated modes. Neuromodulators have highly selective effects that set different modes of thalamocortical and corticothalamic information processing. Natural behavioral states are likely set by a combination of neuromodulators acting in synergy.
- When does it happen: Cholinergic neurons in the LDT/PPT complex discharge vigorously during paradoxical sleep and also during wakefulness (el Mansari et al., 1989), and the levels of acetylcholine increase in the thalamus during those states (Williams et al., 1994). Thus, the cholinergic mode is expected to occur during both REM/paradoxical sleep and during states of vigilance in awake animals.
- Spontaneous activity: Cholinergic activation leads to a sharp increase of spontaneous thalamocortical tonic firing in VPM (see Figure 5), which reduces signal to noise ratios (Aguilar and Castro-Alamancos, 2005; Hirata et al., 2006). Typically, cells increase their spontaneous firing by more than 10-fold compared to the slow oscillation mode. The effect of cholinergic activation on spontaneous firing is explained by both a direct depolarization of VPM cells and a suppression of NRT cell firing. The depolarizing effect of acetylcholine on thalamocortical cells is mediated by muscarinic receptors, which block a resting K+ conductance, and the hyperpolarizing effect of acetylcholine on NRT cells is produced by activation of a K+ conductance (McCormick, 1992).
- Frequency-dependent sensory responses (rapid sensory adaptation): During the cholinergic activated mode, low-frequency sensory responses are a bit stronger (increase slightly in probability) and become faster (evoked spikes display shorter latencies) compared to the slow oscillation mode but signal to noise ratios are sharply reduced (Castro-Alamancos, 2002b; Aguilar and Castro-Alamancos, 2005; Hirata et al., 2006). Another robust change occurs at the level of high-frequency sensory processing. Thus, during cholinergic activation thalamocortical cells robustly enhance their responses to high-frequency sensory signals, virtually eliminating the low-pass filtering of sensory signals in the sensory thalamus (Castro-Alamancos, 2002b). The postsynaptic depolarization of thalamocortical neurons produced by cholinergic activation is sufficient to eliminate the effect of lemniscal synaptic depression on the relay of high frequency inputs (Castro-Alamancos, 2002b, a). Cholinergic activation also reduces the effects of inhibition from NRT by hyperpolarizing NRT cells and suppressing IPSPs in thalamocortical cells (Castro-Alamancos, 2002b, a).
- Spatial dependent sensory responses (receptive fields and selectivity) : Cholinergic activation enlarges the excitatory surround of VPM cells just like BRF stimulation does. Thus, for low-frequency sensory inputs during the slow oscillation mode, the response to the PW (receptive field excitatory center) is stronger than the response to AWs (receptive field excitatory surround), but during activation, there is an enhancement of the response to AWs, which can reach response levels similar to the PW. For high-frequency sensory inputs during the slow oscillation mode, both PW and AW responses are depressed because of the low-pass filtering at the lemniscal pathway. However, during cholinergic activation, there is a significant increase in both PW and AW responses so that they become similar, but PW responses are generally stronger than AW responses at high frequencies.
- Corticothalamic feedback: Corticothalamic EPSPs are suppressed by acetylcholine, an effect that is independent of the postsynaptic actions of acetylcholine (Castro-Alamancos and Calcagnotto, 2001). However, cholinergic activation augments low-frequency corticothalamic responses, which reduces the amount of facilitation in corticothalamic responses, making thalamocortical cells responsive to a wide frequency band of cortical signals (Hirata et al., 2006). Hence, during cholinergic activation, the selectivity of VPM cells for high-frequency corticothalamic signals (high-pass filtering) is lost. This may cause a major problem for thalamocortical sensory processing, because it allows low-frequency cortical signals to become effective driver of thalamocortical cells. Such an effect seems undesirable during sensory processing, because thalamocortical cells may not be able to distinguish sensory and cortical inputs. One possibility is that the enhanced responsiveness to low-frequency cortical signals during cholinergic activation is related to sensory experiences that are driven by internal, top-down, representations during paradoxical sleep (when cholinergic activation is strong). During paradoxical sleep, cortical cells may be strong drivers of thalamocortical neurons, which could serve to feed top-down representations to upper layers of primary sensory cortex via the thalamus, perhaps related to sensory experiences during this phase of sleep.
In conclusion, cholinergic activation has the following effects: increases the spontaneous tonic firing of thalamocortical cells; strongly eliminates the low-pass filtering of sensory inputs, allowing high-frequency sensory relays; broadens thalamocortical receptive fields; allows low-frequency corticothalamic signals, which are usually blocked by the high-pass filtering at this connection; reduces signal-to-noise ratios. Possibly, the larger receptive fields, lower signal-to-noise ratios, and broad-frequency spectrum corticothalamic responses characteristic of cholinergic activation may be related to activated modes during paradoxical sleep and non-attentive wakefulness.
- When does it happen: Noradrenergic neurons discharge robustly during high levels of vigilance and attention, reduce their firing during slow-wave sleep, and stop firing during paradoxical sleep (Hobson et al., 1975; Foote et al., 1980; Aston-Jones and Bloom, 1981). Thus, the noradrenergic mode is expected to occur during states of vigilance in attentive animals.
- Spontaneous activity: Noradrenergic activation leads to a reduction of thalamocortical cell firing so that they have basically nil spontaneous firing (see Figure 5). The effect of noradrenergic activation on spontaneous thalamocortical firing is completely mediated by the NRT because during thalamic disinhibition (block of GABA receptors) norepinephrine no longer suppresses thalamocortical cells (Hirata et al., 2006). In fact, during disinhibition, thalamocortical cells in VPM are excited by norepinephrine. Thus, noradrenergic activation strongly excites NRT cells, which inhibit thalamocortical cells in VPM. The depolarizing effect of norepinephrine on NRT cells is mediated by α-adrenergic receptors, and attributable to a decrease of a resting K+ conductance (McCormick, 1992).
- Frequency-dependent sensory responses (rapid sensory adaptation): The effects of norepinephrine on sensory responses are similar to those produced by cholinergic activation but without the increase in spontaneous firing (Hirata et al., 2006). For sensory signals, noradrenergic activation sets sensory processing to a focused and noise-free excitatory receptive field, which contrasts with the broad and noisy excitatory receptive field characteristic of cholinergic activation. Norepinephrine also facilitates the high-frequency responses to whisker stimulation, albeit less effectively than cholinergic activation.
- Spatial dependent sensory responses (receptive fields and selectivity) : Noradrenergic activation enhances AW responses but only for one whisker and for low-frequency responses. Whereas cholinergic activation enhances high-frequency responses for several AWs, norepinephrine only enhances high-frequency responses for the PW. This indicates that high-frequency sensory inputs are highly focused to the center of the receptive field during noradrenergic activation. Consequently, VPM receptive fields are more focused during noradrenergic activation than during cholinergic activation.
- Corticothalamic feedback: Corticothalamic EPSPs are suppressed by norepinephrine, an effect that is independent of the postsynaptic actions of norepinephrine (Castro-Alamancos and Calcagnotto, 2001). Moreover, noradrenergic activation further high-pass filters corticothalamic responses (Hirata et al., 2006). The high-pass filtering ensures that thalamocortical cells are not driven by cortical signals unless those signals arrive at high frequencies. This effect is similar to that observed after BRF stimulation in anesthetized animals. Thus, for corticothalamic signals, noradrenergic activation sets corticothalamic processing to a noise-free high-frequency signal detection mode.
In conclusion, noradrenergic activation may provide a dynamic mechanism to (1) focus thalamocortical receptive fields, (2) high-pass filter corticothalamic signals, and (3) enhance signal-to-noise ratios. Possibly, the more focused receptive fields and higher signal-to-noise ratios during noradrenergic activation reflect a more appropriate, information processing mode for spatial discrimination of sensory inputs.
Active whisking mode
- When does it happen: Rodents use their vibrissae to navigate the environment by performing fast rhythmic vibrissa movements. During active exploration, whisking consists in ellipsoid movements (which are characterized by vibrissa protractions) through the air and over objects at between 4 and 15 Hz.
- Spontaneous activity: During active whisking in air, thalamocortical cell activity in VPM increases compared to non-whisking (Fanselow and Nicolelis, 1999; Lee et al., 2008). Spontaneous activity also increases for most cells in VPM during artificial whisking in air, which is induced by electrical stimulation of motor nerves in a pattern resembling active whisking (Yu et al., 2006).
- Frequency-dependent sensory responses (rapid sensory adaptation): Sensory responses evoked by stimuli delivered during active whisking are usually suppressed compared to during non-whisking. For example, whisker follicle or infraorbital nerve stimulation evokes a smaller field potential and/or fewer spikes in VPM during active whisking periods than during non-whisking (Fanselow and Nicolelis, 1999; Lee et al., 2008). Also, paired-pulse ratios (amplitude of the response to the second stimulus divided by the amplitude of the first) are significantly smaller during non-whisking, indicating stronger paired pulse suppression. Thus, just like during activated modes (Castro-Alamancos, 2002b), thalamocortical cells appear to follow high-frequency stimuli much better during active whisking. During artificial whisking, most VPM cells enhance their response when the whiskers contact an object compared to the response during whisking in air, while other cells suppress their responses (Yu et al., 2006); cells in the ventrolateral portion of VPM appear to convey a pure touch signal because they mostly fire when a whisker contacts an object but not during whisking in air.
- Spatial dependent sensory responses (receptive fields and selectivity): Because whiskers are moving during active whisking it is difficult to map receptive fields during this mode.
- Corticothalamic feedback: Corticothalamic responses have not been monitored during active whisking.
Pathological modes during epilepsy
- When does it happen: Epilepsy has many different causes and there are a number of rodent genetic models that produce spontaneous seizures involving the thalamocortical network. A well known example is the Genetic Absence Epilepsy Rat from Strasbourg (GAERS). Apart from genetic models, the simplest way to generate seizures in the brain is to impair the control that GABA-mediated inhibition has on excitation. This can be accomplished by blocking GABA receptors (disinhibition) using specific antagonists. Disinhibition may occur naturally in the brain due to a variety of mechanisms including withdrawal of inhibitory synapses or death of inhibitory cells caused by various insults, developmental disorders and/or activity-dependent mechanisms.
- Spontaneous activity: In the vibrissa system, block of thalamic GABAA receptors in vivo leads to ~3 Hz activity in thalamocortical cells that is translated into ~3 Hz spike-wave discharges in the neocortex, and these discharges are abolished by subsequent block of thalamic GABAB receptors (Castro-Alamancos, 1999). Work in vitro has shown that when thalamic GABAA receptors are blocked, GABAB-mediated responses are observed in thalamocortical cells due to longer and higher frequency bursts in NRT neurons caused by a reduction of intra-NRT inhibition (Huntsman et al., 1999). The longer time constants of GABAB-mediated hyperpolarization drive the slower ~3 Hz activity, which is then logically abolished by blocking GABAB receptors. This ~3 Hz activity resembles the activity observed in children during absence seizures, and has been proposed as a laboratory model of this disorder (for a recent review see (Beenhakker and Huguenard, 2009).
- Frequency-dependent sensory responses (rapid sensory adaptation): Block of thalamic GABA receptors has robust consequences on thalamocortical sensory responses (Hirata et al., 2009). During high-frequency (10 Hz) whisker stimulation, thalamic disinhibition enhances short-latency multiwhisker (PW and AWs) and PW responses but only of “transition stimuli”, which are those stimuli in between the first stimulus and the last of a 10 Hz train (see (Hirata et al., 2009)). Thalamic disinhibition also enhances long-latency multiwhisker and PW responses evoked by all stimuli in a train regardless of their frequency and position within a train.
- Spatial dependent sensory responses (receptive fields and selectivity) : Thalamic disinhibition slightly enhances the short-latency response of the strongest whisker in the surround during low-frequency stimulation. In addition, thalamic disinhibition enhances the long-latency response of most of the whiskers in the surround during low-frequency stimulation.
- Corticothalamic feedback: During thalamic disinhibition, there are two major effects on corticothalamic responses. First, low-frequency responses are strongly enhanced. Thus responses to all 10 stimuli in a train at 2 and 5 Hz are significantly enhanced by thalamic disinhibition. Second, there are complex effects of thalamic disinhibition on frequency-dependent facilitation evoked by corticothalamic stimulation. Steady-state facilitated responses (i.e., last 5 stimuli in a 10 stimulus train), evoked at 5 and 10 Hz, are further enhanced by disinhibition. However, the last five stimuli in 20 and 40 Hz trains do not reach a steady facilitated state; instead these responses depress after reaching peak facilitation. This depression phenomenon appears to be related to the ability of high-frequency corticothalamic stimulation (facilitation) to trigger epileptic-like discharges (leading to post-discharge depression). These discharges are not evoked during thalamic disinhibition when high-frequency whisker stimulation is used. Thus, it appears that during thalamic disinhibition thalamocortical cells are sensitive to high-frequency corticothalamic activity, which can trigger epileptic-like seizure activity.
Thalamocortical modes set neocortex modes
The considerable differences in spontaneous firing of thalamocortical cells during different thalamic modes may lead to different modes in the barrel neocortex. For example, thalamic noradrenergic and cholinergic activation produce two very distinct modes of thalamocortical firing, and it is possible that this has consequences on cortical activity. Indeed, recent work has shown that the distinct thalamic modes set by these thalamic neuromodulators produces different cortical modes (Hirata and Castro-Alamancos, 2010). Thalamic cholinergic activation of the thalamus makes thalamocortical cells very responsive to whisker stimuli but also increases their spontaneous tonic firing and this leads to cortical activation in the barrel cortex. In contrast, thalamic noradrenergic activation also makes thalamocortical cells very responsive to sensory stimuli but abolishes their spontaneous firing, and this leads to cortical deactivation or slow oscillations in the barrel cortex. Thus, cholinergic and noradrenergic thalamic activation lead to two well differentiated thalamocortical modes that directly set two distinct cortical modes. The cholinergic thalamocortical mode has abundant presynaptic thalamocortical activity (relay cell noise) but little postsynaptic activity (cortical cell spontaneous activity or noise), while the noradrenergic mode has nil presynaptic activity (no thalamocortical relay cell noise) but plenty of postsynaptic activity (cortical noise).
- Aguilar, J R and Castro-Alamancos, M A (2005). Spatiotemporal gating of sensory inputs in thalamus during quiescent and activated states. The Journal of Neuroscience 25: 10990-11002.
- Aston-Jones, G and Bloom, F E (1981). Norepinephrine-containing locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. The Journal of Neuroscience 1: 887-900.
- Beenhakker, M P and Huguenard, J R (2009). Neurons that fire together also conspire together: is normal sleep circuitry hijacked to generate epilepsy? Neuron 62: 612-632.
- Bourassa, J; Pinault, D and Deschenes, M (1995). Corticothalamic projections from the cortical barrel field to the somatosensory thalamus in rats: A single-fibre study using biocytin as an anterograde tracer. European Journal of Neuroscience 7: 19-30.
- Brown, R E; Stevens, D R and Haas, H L (2001). The physiology of brain histamine. Progress in Neurobiology 63: 637-672.
- Castro-Alamancos, M A (1999). Neocortical synchronized oscillations induced by thalamic disinhibition in vivo. The Journal of Neuroscience 19: RC27.
- Castro-Alamancos, M A (2002a). Properties of primary sensory (lemniscal) synapses in the ventrobasal thalamus and the relay of high-frequency sensory inputs. Journal of Neurophysiology 87: 946-953.
- Castro-Alamancos, M A (2002b). Different temporal processing of sensory inputs in the rat thalamus during quiescent and information processing states in vivo. Journal of Physiology 539: 567-578.
- Castro-Alamancos, M A (2002c). Role of thalamocortical sensory suppression during arousal: Focusing sensory inputs in neocortex. The Journal of Neuroscience 22: 9651-9655.
- Castro-Alamancos, M A (2004a). Dynamics of sensory thalamocortical synaptic networks during information processing states. Progress in Neurobiology 74: 213-247.
- Castro-Alamancos, M A (2004b). Absence of rapid sensory adaptation in neocortex during information processing states. Neuron 41: 455-464.
- Castro-Alamancos, M A and Calcagnotto, M E (1999). Presynaptic long-term potentiation in corticothalamic synapses. The Journal of Neuroscience 19: 9090-9097.
- Castro-Alamancos, M A and Calcagnotto, M E (2001). High-pass filtering of corticothalamic activity by neuromodulators released in the thalamus during arousal: in vitro and in vivo. Journal of Neurophysiology 85: 1489-1497.
- Castro-Alamancos, M A and Oldford, E (2002). Cortical sensory suppression during arousal is due to the activity- dependent depression of thalamocortical synapses. Journal of Physiology 541: 319-331.
- Crick, F (1984). Function of the thalamic reticular complex: The searchlight hypothesis. Proceedings of the National Academy of Sciences of the United States of America 81: 4586-4590.
- Crunelli, V and Hughes, S W (2010). The slow (<1 Hz) rhythm of non-REM sleep: a dialogue between three cardinal oscillators. Nature Neuroscience 13: 9-17.
- el Mansari, M; Sakai, K and Jouvet, M (1989). Unitary characteristics of presumptive cholinergic tegmental neurons during the sleep-waking cycle in freely moving cats. Experimental Brain Research 76: 519-529.
- Fanselow, E E and Nicolelis, M A (1999). Behavioral modulation of tactile responses in the rat somatosensory system. The Journal of Neuroscience 19: 7603-7616.
- Foote, S L; Aston-Jones, G and Bloom, F E (1980). Impulse activity of locus coeruleus neurons in awake rats and monkeys is a function of sensory stimulation and arousal. Proceedings of the National Academy of Sciences of the United States of America 77: 3033-3037.
- Friedberg, M H; Lee, S M and Ebner, F F (1999). Modulation of receptive field properties of thalamic somatosensory neurons by the depth of anesthesia. Journal of Neurophysiology 81: 2243-2252.
- Golshani, P; Liu, X B and Jones, E G (2001). Differences in quantal amplitude reflect GluR4- subunit number at corticothalamic synapses on two populations of thalamic neurons. Proceedings of the National Academy of Sciences of the United States of America 98: 4172-4177.
- Hartings, J A; Temereanca, S and Simons, D J (2003). State-dependent processing of sensory stimuli by thalamic reticular neurons. The Journal of Neuroscience 23: 5264-5271.
- Hirata, A and Castro-Alamancos, M A (2008). Cortical transformation of wide-field (multiwhisker) sensory responses. Journal of Neurophysiology 100: 358-370.
- Hirata, A and Castro-Alamancos, M A (2010). Neocortex network activation and deactivation states controlled by the thalamus. Journal of Neurophysiology 103: 1147-1157.
- Hirata, A; Aguilar, J and Castro-Alamancos, M A (2006). Noradrenergic activation amplifies bottom-up and top-down signal-to-noise ratios in sensory thalamus. The Journal of Neuroscience 26: 4426-4436.
- Hirata, A; Aguilar, J and Castro-Alamancos, M A (2009). Influence of subcortical inhibition on barrel cortex receptive fields. Journal of Neurophysiology 102: 437-450.
- Hobson, J A; McCarley, R W and Wyzinski, P W (1975). Sleep cycle oscillation: Reciprocal discharge by two brainstem neuronal groups. Science 189: 55-58.
- Hughes, S W; Cope, D W; Blethyn, K L and Crunelli, V (2002). Cellular mechanisms of the slow (<1 Hz) oscillation in thalamocortical neurons in vitro. Neuron 33: 947-958.
- Huntsman, M M; Porcello, D M; Homanics, G E; DeLorey, T M and Huguenard, J R (1999). Reciprocal inhibitory connections and network synchrony in the mammalian thalamus. Science 283: 541-543.
- Kim, U and McCormick, D A (1998). The functional influence of burst and tonic firing mode on synaptic interactions in the thalamus. The Journal of Neuroscience 18: 9500-9516.
- Landisman, C E et al. (2002). Electrical synapses in the thalamic reticular nucleus. The Journal of Neuroscience 22: 1002-1009.
- Lee, S; Carvell, G E and Simons, D J (2008). Motor modulation of afferent somatosensory circuits. Nature Neuroscience 11: 1430-1438.
- McCormick, D A (1992). Neurotransmitter actions in the thalamus and cerebral cortex and their role in neuromodulation of thalamocortical activity. Progress in Neurobiology 39: 337-388.
- McCormick, D A and von Krosigk, M (1992). Corticothalamic activation modulates thalamic firing through glutamate "metabotropic" receptors. Proceedings of the National Academy of Sciences of the United States of America 89: 2774-2778.
- McCormick, D A and Bal, T (1997). Sleep and arousal: Thalamocortical mechanisms. Annual Review of Neuroscience 20: 185-215.
- McGinty, D J and Harper, R M (1976). Dorsal raphe neurons: Depression of firing during sleep in cats. Brain Research 101: 569-575.
- Moruzzi, G and Magoun, H W (1949). Brain stem reticular formation and activation of the EEG. Electroencephalography and Clinical Neurophysiology 1: 455-473.
- Rigas, P and Castro-Alamancos, M A (2007). Thalamocortical up states: Differential effects of intrinsic and extrinsic cortical inputs on persistent activity. The Journal of Neuroscience 27: 4261-4272.
- Shu, Y and McCormick, D A (2002). Inhibitory interactions between ferret thalamic reticular neurons. Journal of Neurophysiology 87: 2571-2576.
- Sohal, V S and Huguenard, J R (2003). Inhibitory interconnections control burst pattern and emergent network synchrony in reticular thalamus. The Journal of Neuroscience 23: 8978-8988.
- Steriade, M; Jones, E G and McCormick, D A (1997). Thalamus. New York: Elsevier.
- Trulson, M E and Jacobs, B L (1979). Raphe unit activity in freely moving cats: Correlation with level of behavioral arousal. Brain Research 163: 135-150.
- Williams, J A; Comisarow, J; Day, J; Fibiger, H C and Reiner, P B (1994). State-dependent release of acetylcholine in rat thalamus measured by in vivo microdialysis. The Journal of Neuroscience 14: 5236-5242.
- Yu, C; Derdikman, D; Haidarliu, S and Ahissar, E (2006). Parallel thalamic pathways for whisking and touch signals in the rat. PLoS Biology 4: e124.
- Alonso, J-M and Chen, Y (2009). Receptive field. Scholarpedia 4(1): 5393. http://www.scholarpedia.org/article/Receptive_field.
- Bouret, S and Sara, S J (2010). Locus coeruleus. Scholarpedia 5(3): 2845. http://www.scholarpedia.org/article/Locus_coeruleus.
- Braitenberg, V (2007). Brain. Scholarpedia 2(11): 2918. http://www.scholarpedia.org/article/Brain.
- Freund, T and Kali, S (2008). Interneurons. Scholarpedia 3(9): 4720. http://www.scholarpedia.org/article/Interneurons.
- Hille, B (2008). Ion channels. Scholarpedia 3(10): 6051. http://www.scholarpedia.org/article/Ion_channels.
- Izhikevich, E M (2006). Bursting. Scholarpedia 1(3): 1300. http://www.scholarpedia.org/article/Bursting.
- Johnson, D H (2006). Signal-to-noise ratio. Scholarpedia 1(12): 2088. http://www.scholarpedia.org/article/Signal-to-noise_ratio.
- Jonas, P and Buzsaki, G (2007). Neural inhibition. Scholarpedia 2(9): 3286. http://www.scholarpedia.org/article/Neural_inhibition.
- Llinas, R (2008). Neuron. Scholarpedia 3(8): 1490. http://www.scholarpedia.org/article/Neuron.
- Moehlis, J; Josic, K and Shea-Brown, E T (2006). Periodic orbit. Scholarpedia 1(7): 1358. http://www.scholarpedia.org/article/Periodic_orbit.
- Pikovsky, A and Rosenblum, M (2007). Synchronization. Scholarpedia 2(12): 1459. http://www.scholarpedia.org/article/Synchronization.
- Rayner, K and Castelhano, M (2007). Eye movements. Scholarpedia 2(10): 3649. http://www.scholarpedia.org/article/Eye_movements.
- Roberts, E (2007). Gamma-aminobutyric acid. Scholarpedia 2(10): 3356. http://www.scholarpedia.org/article/Gamma-aminobutyric_acid.
- Saper, C (2009). Hypothalamus. Scholarpedia 4(1): 2791. http://www.scholarpedia.org/article/Hypothalamus.
- Sherman, S M (2006). Thalamus. Scholarpedia 1(9): 1583. http://www.scholarpedia.org/article/Thalamus.
- Ward, L M (2008). Attention. Scholarpedia 3(10): 1538. http://www.scholarpedia.org/article/Attention.
- Wilson, C (2008). Up and down states. Scholarpedia 3(6): 1410. http://www.scholarpedia.org/article/Up_and_down_states.