The ability to wake up every morning and to fall asleep at night is something that most people take for granted. However, damage to a brain region called the central thalamus can cause a range of consciousness-related disorders, including memory problems, excessive sleeping, and even comas. For example, cell death within the central thalamus has been associated with severely disabled patients following traumatic brain injury.

Previous studies have found that electrically stimulating the neurons in the central thalamus can change whether an animal is drowsy or awake and alert. However, it was not clear whether a single group of neurons in the central thalamus was responsible for switching the brain’s state between sleep and wakefulness, or how this would work.

Liu, Lee, Weitz, Fang et al. have now used a technique called optogenetics to stimulate specific neurons in the central thalamus of rats, by using flashes of light. Stimulation was combined with several techniques to monitor the response of other brain regions, including fMRI imaging that shows the activity of the entire brain.

The results showed that rapidly stimulating the neurons in the central thalamus – 40 or 100 times a second – led to widespread brain activity and caused sleeping rats to wake up. In contrast, stimulating the neurons of the central thalamus more slowly – around 10 times a second – suppressed the activity of part of the brain called the sensory cortex and caused rats to enter a seizure-like state of unconsciousness. Further investigation identified a group of inhibitory neurons that the central thalamus interacts with to carry out this suppression.

The results suggest that the central thalamus can either power the brain to an “awake” state or promote a state of unconsciousness, depending on how rapidly its neurons are stimulated. Future work will seek to translate these results to the clinic and investigate how stimulation of the central thalamus can be optimized to reduce cognitive deficits in animal models of traumatic brain injury.

The thalamus plays an important role in coordinating global brain signals responsible for cognition and normal waking behavior (Sherman and Guillery, 1996; Llinás et al., 1998; Mitchell et al., 2014). The central thalamus and intralaminar nuclei, in particular, have been postulated to play a critical and unique function in regulating arousal, attention, and goal-directed behavior (Schiff and Pfaff, 2009; Mair et al., 2011). This idea dates back to the first demonstrations that direct and indirect electrical stimulations of central thalamus control cortical electroencephalography (EEG) dynamics and elicit behavioral transitions between drowsiness/relaxation and wakefulness/attention (Moruzzi and Magoun, 1949; Hunter and Jasper, 1949; Fuster, 1958). Since the initial identification of central thalamus’s causal effect on brain state and behavior, significant support for its role in arousal regulation has come from anatomical and histological studies. Steriade and Glenn (Steriade and Glenn, 1982) identified a monosynaptic pathway from the mesencephalic reticular formation to the central lateral (CL) and paracentral (PC) nuclei of central thalamus that projects to cerebral cortex and striatum. In addition to this input, the central thalamus receives projections from other arousal systems, including norepinephrinergic innervation from locus coeruleus (Vogt et al., 2008) and cholinergic innervation from the upper brainstem and basal forebrain (Heckers et al., 1992). In combination with these inputs, the diffuse projections of central thalamus allow it to influence the overall excitability of cortex during states of attention. For example, virtually all relay cells of the CL nucleus project to both striatum and cerebral cortex (Deschenes et al., 1996).

Studies on the physiological properties of central thalamus also show that it is tightly coupled to arousal regulation. First, variations in the level of activity within the intralaminar nuclei are linked to changes in behavioral alertness (Kinomura et al., 1996; Shirvalkar et al., 2006; Mair and Hembrook, 2008; Schiff et al., 2013; Giber et al., 2015), including transitions during the normal sleep–wake cycle and acute cognitive enhancements such as improved working-memory and sustained attention (Baker et al., 2012). Similarly, lesions of the central thalamus can produce enduring cognitive impairments, including reduced attentional processing and memory (Guberman and Stuss, 1983; Mair et al., 1998; Van Der Werf et al., 1999; Newman and Burk, 2005), hypersomnolence (Bassetti et al., 1996), or even coma (Castaigne et al., 1981; Plum, 1991). Indeed, neuronal loss across central thalamus has been associated with severely disabled and vegetative patients following severe traumatic brain injury (Maxwell et al., 2006). In addition, electrical stimulation of the central thalamus at low frequencies is associated with absence seizures and behavioral arrest in animal models (Hunter and Jasper, 1949) and human subjects (Velasco, 1996). Human imaging studies have also found that anesthesia-induced loss of consciousness is associated with disrupted thalamocortical functional connectivity in regions consistent with the intralaminar nuclei (Akeju, 2014).

According to the mesocircuit hypothesis of forebrain dysfunction, the central thalamus, which has a strong activating role in driving cortical and striatal neurons (Schiff, 2010), is under tonic inhibition by GABAergic pallidal neurons (Grillner et al., 2005). This GABAergic population is itself inhibited by the striatal neurons driven by central thalamus, creating a positive feedback loop. Thus, when the thalamostriatal and thalamocortical projections from central thalamus are partially lost due to brain injury, it causes disinhibition of the pallidum and increased inhibition of the remaining central thalamus neurons, which further reduces cortical activation. While this down-regulation is predicted to have broad modulatory impact on global dynamics, deep brain stimulation (DBS) of central thalamus has been explored as a potential means of reversing its effects and facilitating arousal regulation in the minimally conscious state (Shirvalkar et al., 2006; Schiff et al., 2007; Mair and Hembrook, 2008; Smith et al., 2009; Mair et al., 2011; Schiff, 2012; Fridman et al., 2014; Gummadavelli et al., 2015). Despite success in a single-subject clinical study (Schiff et al., 2007), identification of circuit-level mechanisms that link therapeutic efficacy of central thalamus DBS to specific stimulation parameters remains challenging and at present limits the clinical efficacy of DBS in subjects with traumatic brain injury. Thus, while significant progress has been made in understanding the connections of central thalamus and its behavioral correlates (Van der Werf et al., 2002; Schiff, 2008), relatively little is known about the dynamic function of these circuits, and no clear mechanism exists to explain how – or indeed, if – a single population in central thalamus can act as a switch on the global brain state.

To overcome such obstacles and dissect the dynamic influence of central thalamus on global brain networks, we combined targeted optogenetic control of excitatory relay neurons within central thalamus with whole-brain fMRI readouts, EEG, and single-unit recordings. The combination of optogenetic stimulation with fMRI (ofMRI) has been demonstrated to be an effective method for mapping the functional role of specific genetically- and spatially-defined neuronal populations at various brain regions and different frequencies of stimulation (Lee et al., 2010; Desai et al., 2011; Weitz et al., 2015; Liang et al., 2015; Takata et al., 2015; Duffy et al., 2015; Byers et al., 2015). In the present work, we sought to apply this approach to uncover the downstream effects of distinct firing patterns by central thalamus relay cells at the whole-brain level – a visualization uniquely possible with ofMRI. Prior work beginning with Morison and Dempsey (Morison and Dempsey, 1942), who envisioned “dissecting… the electrical activity of the cortex on the basis of its relations with the thalamus,” has shown that high- and low-frequency electrical stimulation of thalamic nuclei can produce distinct cortical activation patterns. For example, early studies of the intralaminar nuclei per se showed that low-frequency stimulation evokes slow-wave activity and spindle bursts in cortical EEG, which are associated with primary generalized absence seizures, loss of consciousness, and drowsiness (Hunter and Jasper, 1949; Jasper, 1949; Seidenbecher and Pape, 2001). Conversely, high-frequency electrical stimulation has been shown to desynchronize the cortical EEG signal (Moruzzi and Magoun, 1949), which is associated with behavioral arousal. While these studies set early hypotheses on the mechanisms of arousal regulation, the non-selective nature of electrical stimulation has prevented the observed responses from being attributed specifically to relay cells, and not synaptic afferents or fibers of passage mixing together in a bulk activation effect. More importantly, although a picture of whole brain activity can be vaguely inferred from electrophysiology recordings, these techniques cannot provide a direct visualization of activity across individual brain regions over the entire brain. Because ofMRI can provide spatial and temporal information on the whole-brain scale during perturbations of specific neural circuitry, we applied this technique to study the causal role of central thalamus relay neurons in activating forebrain networks.

In addition, following on novel results described below, we were led to examine the interplay between central thalamus and zona incerta (ZI), a subcortical region implicated in the modulation of absence seizures in rats (Shaw et al., 2013). The ZI, a mostly GABAergic region, has been shown to limit the transmission of ascending sensory information via feedforward inhibition of higher order thalamic nuclei (Barthó et al., 2002; Trageser and Keller, 2004; Lavallée et al., 2005; Trageser et al., 2006). Such activity can induce a state of reduced sensory processing, similar to the behavioral quiescence induced by low-frequency central thalamus stimulation. These studies suggest a powerful control exerted by ZI over brain state and higher level processing, much like central thalamus. However, the possible involvement of ZI in central thalamus arousal circuits remains unexplored. Here, we investigated the electrophysiology responses of this region during targeted stimulation of central thalamus at frequencies that either facilitate or suppress attention and arousal. To determine whether ZI plays an active role in these circuits, we also used optogenetics to specifically inhibit neurons in this region during central thalamus stimulation. This experimental paradigm was used to infer ZI’s functional contribution to the central thalamus-driven brain circuit dynamics measured with fMRI and electrophysiology.

Previously proposed mechanisms of arousal regulation have focused on the physiological and anatomical specialization of neurons within central thalamus (Schiff, 2008). Here, using a combination of ofMRI and electrophysiological recordings, we directly visualize whole-brain network activations produced by selective stimulation of central thalamus relay neurons and reveal novel insight into frequency-dependent gating of forebrain arousal.

Since the earliest observations that electrical stimulation of central thalamus exerts frequency-dependent effects on behavior and EEG rhythms (Moruzzi and Magoun, 1949; Hunter and Jasper, 1949), behavioral arousal and cognition have been tightly linked with cortical activation (i.e. low-amplitude, high-frequency oscillations), while behavioral arrest has been linked with cortical deactivation (i.e. high-amplitude, low-frequency oscillations). More recently, it has also been shown that pharmacologically-induced changes in thalamic firing levels can switch cortical dynamics between activation and deactivation (Hirata and Castro-Alamancos, 2010). However, no studies have characterized the specific changes in activity that simultaneously occur across the whole intact brain during these events to explain how the same neuronal population can selectively switch arousal state. Here, we show that distinct firing patterns of excitatory neurons in the central thalamus drive these opposing EEG rhythms, lead to dramatic differences in the spatial extent of forebrain recruitment, and switch the region’s downstream influence on cortex from excitation to inhibition. Notably, high-frequency EEG patterns evoked during 40 and 100 Hz stimulation associate with robust activation of frontal cortex, motor cortex, somatosensory cortex, and striatum – regions that receive widespread glutamatergic projections from intralaminar nuclei (Jones and Leavitt, 1974; Berendse and Groenewegen, 1990; Groenewegen and Berendse, 1994; Smith et al., 2004; Hoover and Vertes, 2007; Hunnicutt et al., 2014). On the other hand, slow-wave oscillations evoked during 10 Hz stimulation are associated with limited forebrain activation and strong inhibition of somatosensory cortex. The frequency-dependent generation of spindle-like oscillations, which are known to underlie brain synchronization at the onset of sleep (Steriade et al., 1993), suggest that these differences may in part be due to the engagement of thalamocortical networks responsible for sleep and loss of perceptual awareness during 10 Hz stimulation. In the context of our findings, it is also noteworthy that for some rat strains with absence seizures, a specific 10 Hz generator in somatosensory cortex has been independently proposed (van Luijtelaar and Sitnikova, 2006).

The broadly contrasting cortical responses between low and high frequencies of central thalamus stimulation imply the frequency-dependent activation of a GABAergic population. In this study, we investigated the behavior of ZI – a region rich in GABAergic neurons that also sends direct inhibitory projections to sensory thalamus and sensory cortex (Lin et al., 1990; Benson et al., 1992; Nicolelis et al., 1995; Kolmac and Mitrofanis, 1999; Barthó et al., 2002). We found that spindle-like oscillations were uniquely evoked when central thalamus was stimulated at 10 Hz and somatosensory cortex was inhibited. Stimulation at this frequency in asleep rats also evoked absence seizure-like freezing and spike-waves in a majority of animals, a cortical pattern known to be modulated by ZI projections (Shaw et al., 2013). Importantly, suppressing the incertal activity evoked by 10 Hz central thalamus stimulation with halorhodopsin reduced the cortical inhibition (Figure 5H), suggesting a key role for ZI in modulating this response. Indeed, it has been previously suggested that rhythmic incertal activity contributes to membrane hyperpolarizations and sustained high-voltage cortical rhythms through GABAergic incertofugal pathways (Shaw et al., 2013). Our data support this hypothesis, and link such a pathway to whole-brain, directly visualized fMRI activity patterns.

Given the presence of GABAergic projections from ZI to central thalamus (Barthó et al., 2002), activity in ZI may also act to limit forebrain activation, as observed with ofMRI during 10 Hz stimulation, through incertal-thalamic feedback. This incerto-thalamic pathway would parallel the previously reported gating of ascending sensory information at the level of thalamus by ZI (Trageser and Keller, 2004; Lavallée et al., 2005). The hypothesized feedforward and feedback inhibition via ZI both suggest a direct projection from central thalamus to ZI, which our fluorescence imaging data support (Figure 4H). However, we note that previous tracing studies failed to identify thalamic input specifically from intralaminar nuclei to ZI (Shammah-Lagnado et al., 1985). In summary, our findings provide the first demonstration that arousal regulation driven by central thalamus has a causal and frequency-dependent influence on ZI, and that suppressing the recruitment of ZI modulates the brain-wide dynamics driven by central thalamus stimulation. Specifically, our results suggest that the frequency-dependent depression of cortical activity is in part mediated by extrinsic inhibitory signals originating from ZI.

An additional mechanism that could contribute to the evoked suppression of cortex is feedforward thalamocortical inhibition – the process by which relay neurons drive inhibitory post-synaptic potentials (IPSPs) in pyramidal cells via fast spiking cortical interneurons (Agmon and Connors, 1991; Porter et al., 2001; Cruikshank et al., 2007). Low-frequency (10 Hz) stimulations of certain thalamic nuclei in vivo yield strong hyperpolarization of cortical neurons putatively via this process (Castro-Alamancos and Connors, 1996a; 1996c). It is also interesting to note that the ‘recruiting response’, characterized by an enhanced cortical response during low frequency electrical stimulation of intralaminar nuclei (Morison and Dempsey, 1942), is hypothesized to originate from this inhibition (Castro-Alamancos and Connors, 1997). Our observation that individual stimuli sometimes trigger spikes in cortex (Figure 3—figure supplement 1) is consistent with the possibility that these phenomena occur during the delivered 10 Hz optogenetic stimuli. However, intracellular and laminar recordings are needed to more conclusively resolve this issue. More recently, a study utilizing optogenetic targeting showed that non-specific ‘matrix’ thalamocortical neurons preferentially drive inhibitory interneurons in cortical layer I (Cruikshank et al., 2012). Moreover, they found that IPSPs generated by stimulation of matrix neurons (which constitute the nuclei targeted here [Jones, 2001]) remain sustained during repeated stimuli compared to those evoked by stimulation of non-matrix neurons. Given the above findings, it is possible that interneuron-mediated thalamocortical inhibition, in addition to the demonstrated role of ZI, may contribute to the observed cortical responses. However, to the best of our knowledge, there have been no in vivo studies demonstrating that cortical interneuron-to-pyramidal cell inhibition is stronger at low frequencies of intralaminar thalamic stimulation compared to high frequencies.

In addition to providing novel insight into the mechanisms of arousal regulation by central thalamus, our study offers important insight into the cellular origins of the fMRI BOLD signal. While there is a growing body of evidence suggesting that negative BOLD signals reflect local decreases in neuronal activity (Shmuel et al., 2006; Pasley et al., 2007; Allen et al., 2007; Devor et al., 2007; Sotero and Trujillo-Barreto, 2007), the nature of this signal remains a subject of debate and holds significant potential for the interpretation of functional imaging studies (Schridde et al., 2008; Ekstrom, 2010). It has been shown that different sensory stimuli can evoke positive and negative BOLD signals in the same cortical area, which are linked to increases and decreases in neural activity, respectively (Shmuel et al., 2006). Building upon these studies, we found that direct stimulation of central thalamus excitatory neurons at different frequencies leads to activation or suppression of neuronal activity in a downstream cortical location, which is coupled with positive and negative BOLD signals, respectively (Figure 3). These findings strongly support the hypothesis that a major component of the negative BOLD signal derives from decreases in neuronal activity and are consistent with previous reports of tight neural-hemodynamic coupling in the somatosensory cortex (Huttunen et al., 2008).

Our results are also consistent with a previous study by Logothetis et al., which showed that low frequencies (<50 Hz) of electrical microstimulation in the thalamic lateral geniculate nucleus evoke negative BOLD responses in the monosynaptically connected V1 cortex, while higher frequencies (100–200 Hz) evoke positive BOLD responses in the same region (Logothetis et al., 2010). We observed similar results in the somatosensory cortex, which is monosynaptically connected to the stimulated intralaminar nuclei (Van der Werf et al., 2002) (Figure 2—figure supplement 1). The study by Logothetis et al. also found that cortical regions which are polysynaptically connected to the lateral geniculate nucleus, such as V2, even exhibit negative BOLD responses at high frequencies of stimulation (>60 Hz). It was proposed that these polysynaptic deactivations result from frequency-dependent disynaptic inhibition, the process by which pyramidal cells in cortex inhibit local and remote pyramidal cells via GABAergic interneurons. Unlike the study by Logothetis et al., we did not observe significant negative BOLD signals in either mono- or polysynaptically connected regions of cortex during high frequencies of stimulation. Furthermore, given the bias of corticocortical disynaptic inhibition toward higher frequencies (Silberberg and Markram, 2007), this microcircuit is unlikely to be driving the observed suppression of cortex during 10 Hz central thalamus stimulation.

In the context of electrical stimulation, our study helps dissociate the confounding effects of (a) delivering stimulation at a certain frequency (which can preferentially recruit certain neuronal elements (McIntyre and Grill, 2002)) and (b) the excited neuronal population firing at a specific frequency. With electrical stimulation, it has been impossible to dissociate these two effects in vivo, since the frequency of stimulation and preferential recruitment of specific neuronal populations could not be decoupled (McIntyre and Grill, 2002). This made it difficult to explain, for example, the relationship between stimulation parameters and the therapeutic efficacy of DBS. Using targeted, temporally precise, optogenetic stimulation in the current study allowed us to selectively excite a single group of neuronal elements and identify their specific role in creating distinct modes of network function. The use of electrical stimulation instead would have prevented us from gaining this unique insight into the specific role of excitatory central thalamus neurons and their spiking frequency.

Finally, in the context of central thalamus DBS, our study offers important insight into the identification of proper stimulation targets and parameters that are needed before the therapeutic application of central thalamus stimulation can reach its full clinical potential. In particular, the images from ofMRI experiments (Figure 2 and Figure 2—figure supplement 3) reveal dramatic differences in global brain dynamics that can result from controlling one parameter of stimulation (i.e. frequency). Furthermore, the widespread activation of cortex and striatum observed at high frequencies of stimulation adds to a growing body of evidence that the central thalamus is a highly appropriate target for the remediation of acquired cognitive disabilities via forebrain recruitment. In a more general context that extends beyond stimulation of central thalamus, the ofMRI techniques we employ here are generalizable and can be universally applied to study the mechanisms underlying DBS for other target regions and disorders. With this knowledge, stimulation paradigms can be optimized to accelerate clinical translation for a wide range of neurological disorders that currently lack such treatment, paving the way for the development of next-generation DBS therapies.

1. 1. Agmon A
   2. Connors BW

   (1991) Thalamocortical responses of mouse somatosensory (barrel) cortex in vitro

   Neuroscience 41:365–379.

   https://doi.org/10.1016/0306-4522(91)90333-J

   • PubMed
   • Google Scholar
2. 1. Akeju O
   2. Loggia ML
   3. Catana C
   4. Pavone KJ
   5. Vazquez R
   6. Rhee J
   7. Contreras Ramirez V
   8. Chonde DB
   9. Izquierdo-Garcia D
   10. Arabasz G
   11. Hsu S
   12. Habeeb K
   13. Hooker JM
   14. Napadow V
   15. Brown EN
   16. Purdon PL

   (2014) Disruption of thalamic functional connectivity is a neural correlate of dexmedetomidine-induced unconsciousness

   eLife 3:e04499.

   https://doi.org/10.7554/eLife.04499

   • PubMed
   • Google Scholar
3. 1. Allen EA
   2. Pasley BN
   3. Duong T
   4. Freeman RD

   (2007) Transcranial magnetic stimulation elicits coupled neural and hemodynamic consequences

   Science 317:1918–1921.

   https://doi.org/10.1126/science.1146426

   • PubMed
   • Google Scholar
4. 1. Aravanis AM
   2. Wang LP
   3. Zhang F
   4. Meltzer LA
   5. Mogri MZ
   6. Schneider MB
   7. Deisseroth K

   (2007) An optical neural interface: in vivo control of rodent motor cortex with integrated fiberoptic and optogenetic technology

   Journal of Neural Engineering 4:S143–156.

   https://doi.org/10.1088/1741-2560/4/3/S02

   • PubMed
   • Google Scholar
5. 1. Baker JL
   2. Wei XF
   3. Ryou JW
   4. Butson CR
   5. Schiff ND
   6. Purpura KP

   (2012)

   Modulation of behavior and global brain oscillations with central thalamic deep brain stimulation in the non-human primate

   Society for Neuroscience Abstract 597.14:.

   • Google Scholar
6. 1. Bandettini PA
   2. Jesmanowicz A
   3. Wong EC
   4. Hyde JS

   (1993) Processing strategies for time-course data sets in functional mri of the human brain

   Magnetic Resonance in Medicine 30:161–173.

   https://doi.org/10.1002/mrm.1910300204

   • PubMed
   • Google Scholar
7. 1. Barthó P
   2. Freund TF
   3. Acsády L

   (2002) Selective GABAergic innervation of thalamic nuclei from zona incerta

   The European Journal of Neuroscience 16:999–1014.

   https://doi.org/10.1046/j.1460-9568.2002.02157.x

   • PubMed
   • Google Scholar
8. 1. Bassetti C
   2. Mathis J
   3. Gugger M
   4. Lovblad KO
   5. Hess CW

   (1996) Hypersomnia following paramedian thalamic stroke: a report of 12 patients

   Annals of Neurology 39:471–480.

   https://doi.org/10.1002/ana.410390409

   • PubMed
   • Google Scholar
9. 1. Benson DL
   2. Isackson PJ
   3. Gall CM
   4. Jones EG

   (1992) Contrasting patterns in the localization of glutamic acid decarboxylase and Ca2+/calmodulin protein kinase gene expression in the rat central nervous system

   Neuroscience 46:825–849.

   https://doi.org/10.1016/0306-4522(92)90188-8

   • PubMed
   • Google Scholar
10. 1. Berendse HW
    2. Groenewegen HJ

    (1990) Organization of the thalamostriatal projections in the rat, with special emphasis on the ventral striatum

    The Journal of Comparative Neurology 299:187–228.

    https://doi.org/10.1002/cne.902990206

    • PubMed
    • Google Scholar
11. 1. Byers B
    2. Lee HJ
    3. Liu J
    4. Weitz AJ
    5. Lin P
    6. Zhang P
    7. Shcheglovitov A
    8. Dolmetsch R
    9. Pera RR
    10. Lee JH

    (2015) Direct in vivo assessment of human stem cell graft-host neural circuits

    NeuroImage 114:328–337.

    https://doi.org/10.1016/j.neuroimage.2015.03.079

    • PubMed
    • Google Scholar
12. 1. Castaigne P
    2. Lhermitte F
    3. Buge A
    4. Escourolle R
    5. Hauw JJ
    6. Lyon-Caen O

    (1981) Paramedian thalamic and midbrain infarct: clinical and neuropathological study

    Annals of Neurology 10:127–148.

    https://doi.org/10.1002/ana.410100204

    • PubMed
    • Google Scholar
13. 1. Castro-Alamancos MA
    2. Connors BW

    (1996a)

    Cellular mechanisms of the augmenting response: short-term plasticity in a thalamocortical pathway

    Journal of Neuroscience 16:7742–7756.

    • PubMed
    • Google Scholar
14. 1. Castro-Alamancos MA
    2. Connors BW

    (1996b)

    Spatiotemporal properties of short-term plasticity sensorimotor thalamocortical pathways of the rat

    Journal of Neuroscience 16:2767–2779.

    • PubMed
    • Google Scholar
15. 1. Castro-Alamancos MA
    2. Connors BW

    (1996c) Short-term plasticity of a thalamocortical pathway dynamically modulated by behavioral state

    Science 272:274–277.

    https://doi.org/10.1126/science.272.5259.274

    • Google Scholar
16. 1. Castro-Alamancos  MA
    2. Connors BW

    (1997) Thalamocortical synapses

    Progress in Neurobiology 51:581–606.

    https://doi.org/10.1016/S0301-0082(97)00002-6

    • PubMed
    • Google Scholar
17. 1. Christie IN
    2. Wells JA
    3. Southern P
    4. Marina N
    5. Kasparov S
    6. Gourine AV
    7. Lythgoe MF

    (2013) FMRI response to blue light delivery in the nave brain: implications for combined optogenetic fMRI studiess

    NeuroImage 66:634–641.

    https://doi.org/10.1016/j.neuroimage.2012.10.074

    • PubMed
    • Google Scholar
18. 1. Contreras D
    2. Destexhe A
    3. Sejnowski TJ
    4. Steriade M

    (1997)

    Spatiotemporal patterns of spindle oscillations in cortex and thalamus

    Journal of Neuroscience 17:1179–1196.

    • PubMed
    • Google Scholar
19. 1. Cruikshank SJ
    2. Ahmed OJ
    3. Stevens TR
    4. Patrick SL
    5. Gonzalez AN
    6. Elmaleh M
    7. Connors BW

    (2012) Thalamic control of layer 1 circuits in prefrontal cortex

    Journal of Neuroscience 32:17813–17823.

    https://doi.org/10.1523/JNEUROSCI.3231-12.2012

    • PubMed
    • Google Scholar
20. 1. Cruikshank SJ
    2. Lewis TJ
    3. Connors BW

    (2007) Synaptic basis for intense thalamocortical activation of feedforward inhibitory cells in neocortex

    Nature Neuroscience 10:462–468.

    https://doi.org/10.1038/nn1861

    • PubMed
    • Google Scholar
21. 1. Dempsey EW
    2. Morison RS

    (1943)

    The electrical activity of a thalamocortical relay system

    Am J Physiol 138:0283–0296.

    • Google Scholar
22. 1. Desai M
    2. Kahn I
    3. Knoblich U
    4. Bernstein J
    5. Atallah H
    6. Yang A
    7. Kopell N
    8. Buckner RL
    9. Graybiel AM
    10. Moore CI
    11. Boyden ES

    (2011) Mapping brain networks in awake mice using combined optical neural control and fMRI

    Journal of Neurophysiology 105:1393–1405.

    https://doi.org/10.1152/jn.00828.2010

    • PubMed
    • Google Scholar
23. 1. Deschênes M
    2. Bourassa J
    3. Parent A

    (1996) Striatal and cortical projections of single neurons from the central lateral thalamic nucleus in the rat

    Neuroscience 72:679–687.

    https://doi.org/10.1016/0306-4522(96)00001-2

    • PubMed
    • Google Scholar
24. 1. Devor A
    2. Tian P
    3. Nishimura N
    4. Teng IC
    5. Hillman EM
    6. Narayanan SN
    7. Ulbert I
    8. Boas DA
    9. Kleinfeld D
    10. Dale AM

    (2007) Suppressed neuronal activity and concurrent arteriolar vasoconstriction may explain negative blood oxygenation level-dependent signal

    Journal of Neuroscience 27:4452–4459.

    https://doi.org/10.1523/JNEUROSCI.0134-07.2007

    • PubMed
    • Google Scholar
25. 1. Duffy BA
    2. Choy M
    3. Chuapoco MR
    4. Madsen M
    5. Lee JH

    (2015) MRI compatible optrodes for simultaneous LFP and optogenetic fMRI investigation of seizure-like afterdischarges

    NeuroImage 123:173–184.

    https://doi.org/10.1016/j.neuroimage.2015.07.038

    • Google Scholar
26. 1. Ekstrom A

    (2010) How and when the fMRI BOLD signal relates to underlying neural activity: the danger in dissociation

    Brain Research Reviews 62:233–244.

    https://doi.org/10.1016/j.brainresrev.2009.12.004

    • PubMed
    • Google Scholar
27. 1. Ellender TJ
    2. Harwood J
    3. Kosillo P
    4. Capogna M
    5. Bolam JP

    (2013) Heterogeneous properties of central lateral and parafascicular thalamic synapses in the striatum

    The Journal of Physiology 591:257–272.

    https://doi.org/10.1113/jphysiol.2012.245233

    • PubMed
    • Google Scholar
28. 1. Engel SA
    2. Glover GH
    3. Wandell BA

    (1997) Retinotopic organization in human visual cortex and the spatial precision of functional MRI

    Cerebral Cortex 7:181–192.

    https://doi.org/10.1093/cercor/7.2.181

    • PubMed
    • Google Scholar
29. 1. Fang Z
    2. Lee JH

    (2013) High-throughput optogenetic functional magnetic resonance imaging with parallel computations

    Journal of Neuroscience Methods 218:184–195.

    https://doi.org/10.1016/j.jneumeth.2013.04.015

    • PubMed
    • Google Scholar
30. 1. Fridman EA
    2. Beattie BJ
    3. Broft A
    4. Laureys S
    5. Schiff ND

    (2014) Regional cerebral metabolic patterns demonstrate the role of anterior forebrain mesocircuit dysfunction in the severely injured brain

    Proceedings of the National Academy of Sciences of the United States of America 111:6473–6478.

    https://doi.org/10.1073/pnas.1320969111

    • Google Scholar
31. 1. Fuster JM

    (1958) Effects of stimulation of brain stem on tachistoscopic perception

    Science 127:150.

    https://doi.org/10.1126/science.127.3290.150

    • PubMed
    • Google Scholar
32. 1. Giber K
    2. Diana MA
    3. Plattner VM
    4. Dugué GP
    5. Bokor H
    6. Rousseau CV
    7. Maglóczky Z
    8. Havas L
    9. Hangya B
    10. Wildner H
    11. Zeilhofer HU
    12. Dieudonné S
    13. Acsády L

    (2015) A subcortical inhibitory signal for behavioral arrest in the thalamus

    Nature Neuroscience 18:562–568.

    https://doi.org/10.1038/nn.3951

    • PubMed
    • Google Scholar
33. 1. Glover GH
    2. Lee AT

    (1995) Motion artifacts in fMRI: comparison of 2DFT with PR and spiral scan methods

    Magnetic Resonance in Medicine 33:624–635.

    https://doi.org/10.1002/mrm.1910330507

    • PubMed
    • Google Scholar
34. 1. Grillner S
    2. Hellgren J
    3. Ménard A
    4. Saitoh K
    5. Wikström MA

    (2005) Mechanisms for selection of basic motor programs--roles for the striatum and pallidum

    Trends in Neurosciences 28:364–370.

    https://doi.org/10.1016/j.tins.2005.05.004

    • PubMed
    • Google Scholar
35. 1. Groenewegen HJ
    2. Berendse HW

    (1994) The specificity of the ‘nonspecific’ midline and intralaminar thalamic nuclei

    Trends in Neurosciences 17:52–57.

    https://doi.org/10.1016/0166-2236(94)90074-4

    • PubMed
    • Google Scholar
36. 1. Guberman A
    2. Stuss D

    (1983) The syndrome of bilateral paramedian thalamic infarction

    Neurology 33:540.

    https://doi.org/10.1212/WNL.33.5.540

    • PubMed
    • Google Scholar
37. 1. Gummadavelli A
    2. Motelow JE
    3. Smith N
    4. Zhan Q
    5. Schiff ND
    6. Blumenfeld H

    (2015) Thalamic stimulation to improve level of consciousness after seizures: evaluation of electrophysiology and behavior

    Epilepsia 56:114–124.

    https://doi.org/10.1111/epi.12872

    • PubMed
    • Google Scholar
38. 1. Heckers S
    2. Geula C
    3. Mesulam MM

    (1992) Cholinergic innervation of the human thalamus: dual origin and differential nuclear distribution

    The Journal of Comparative Neurology 325:68–82.

    https://doi.org/10.1002/cne.903250107

    • PubMed
    • Google Scholar
39. 1. Hirata A
    2. Castro-Alamancos MA

    (2010) Neocortex network activation and deactivation states controlled by the thalamus

    Journal of Neurophysiology 103:1147–1157.

    https://doi.org/10.1152/jn.00955.2009

    • PubMed
    • Google Scholar
40. 1. Hoover WB
    2. Vertes RP

    (2007) Anatomical analysis of afferent projections to the medial prefrontal cortex in the rat

    Brain Structure & Function 212:149–179.

    https://doi.org/10.1007/s00429-007-0150-4

    • PubMed
    • Google Scholar
41. 1. Horner RL
    2. Stephenson R
    3. O'Donnell C

    (2003)

    Instrumentation and methods for chronic studies of sleep and breathing in rodents

    Sleep Related Breathing Disorders: Experimental Models and Therapeutic Potential, (Marcel Dekker, Inc) 171:19–55.

    • Google Scholar
42. 1. Hunnicutt BJ
    2. Long BR
    3. Kusefoglu D
    4. Gertz KJ
    5. Zhong H
    6. Mao T

    (2014) A comprehensive thalamocortical projection map at the mesoscopic level

    Nature Neuroscience 17:1276–1285.

    https://doi.org/10.1038/nn.3780

    • PubMed
    • Google Scholar
43. 1. Hunter J
    2. Jasper HH

    (1949) Effects of thalamic stimulation in unanaesthetised animals; the arrest reaction and petit mal-like seizures, activation patterns and generalized convulsions

    Electroencephalography and Clinical Neurophysiology 1:305–324.

    https://doi.org/10.1016/0013-4694(49)90043-7

    • PubMed
    • Google Scholar
44. 1. Huttunen JK
    2. Gröhn O
    3. Penttonen M

    (2008) Coupling between simultaneously recorded BOLD response and neuronal activity in the rat somatosensory cortex

    NeuroImage 39:775–785.

    https://doi.org/10.1016/j.neuroimage.2007.06.042

    • PubMed
    • Google Scholar
45. 1. Jackson JI
    2. Meyer CH
    3. Nishimura DG
    4. Macovski A

    (1991) Selection of a convolution function for fourier inversion using gridding [computerised tomography application]

    IEEE Transactions on Medical Imaging 10:473–478.

    https://doi.org/10.1109/42.97598

    • PubMed
    • Google Scholar
46. 1. Jasper H

    (1949) Diffuse projection systems: the integrative action of the thalamic reticular system

    Electroencephalography and Clinical Neurophysiology 1:405–420.

    https://doi.org/10.1016/0013-4694(49)90213-8

    • PubMed
    • Google Scholar
47. 1. Jones EG
    2. Leavitt RY

    (1974) Retrograde axonal transport and the demonstration of non-specific projections to the cerebral cortex and striatum from thalamic intralaminar nuclei in the rat, cat and monkey

    The Journal of Comparative Neurology 154:349–377.

    https://doi.org/10.1002/cne.901540402

    • PubMed
    • Google Scholar
48. 1. Jones EG

    (2001) The thalamic matrix and thalamocortical synchrony

    Trends in Neurosciences 24:595–601.

    https://doi.org/10.1016/S0166-2236(00)01922-6

    • PubMed
    • Google Scholar
49. 1. Kim DH
    2. Adalsteinsson E
    3. Spielman DM

    (2003) Simple analytic variable density spiral design

    Magnetic Resonance in Medicine 50:214–219.

    https://doi.org/10.1002/mrm.10493

    • PubMed
    • Google Scholar
50. 1. Kinomura S
    2. Larsson J
    3. Guly s B. z.
    4. Roland PE

    (1996) Activation by attention of the human reticular formation and thalamic intralaminar nuclei

    Science 271:512–515.

    https://doi.org/10.1126/science.271.5248.512

    • PubMed
    • Google Scholar
51. 1. Kolmac C
    2. Mitrofanis J

    (1999) Distribution of various neurochemicals within the zona incerta: an immunocytochemical and histochemical study

    Anatomy and Embryology 199:265–280.

    https://doi.org/10.1007/s004290050227

    • PubMed
    • Google Scholar
52. 1. Lavallée P
    2. Urbain N
    3. Dufresne C
    4. Bokor H
    5. Acsády L
    6. Deschênes M

    (2005) Feedforward inhibitory control of sensory information in higher-order thalamic nuclei

    Journal of Neuroscience 25:7489–7498.

    https://doi.org/10.1523/JNEUROSCI.2301-05.2005

    • PubMed
    • Google Scholar
53. 1. Lee JH
    2. Durand R
    3. Gradinaru V
    4. Zhang F
    5. Goshen I
    6. Kim DS
    7. Fenno LE
    8. Ramakrishnan C
    9. Deisseroth K

    (2010) Global and local fMRI signals driven by neurons defined optogenetically by type and wiring

    Nature 465:788–792.

    https://doi.org/10.1038/nature09108

    • PubMed
    • Google Scholar
54. 1. Liang Z
    2. Watson GD
    3. Alloway KD
    4. Lee G
    5. Neuberger T
    6. Zhang N

    (2015) Mapping the functional network of medial prefrontal cortex by combining optogenetics and fMRI in awake rats

    NeuroImage 117:114–123.

    https://doi.org/10.1016/j.neuroimage.2015.05.036

    • PubMed
    • Google Scholar
55. 1. Lin C
    2. Nicolelis M
    3. Schneider J
    4. Chapin J

    (1990) A major direct GABAergic pathway from zona incerta to neocortex

    Science 248:1553–1556.

    https://doi.org/10.1126/science.2360049

    • PubMed
    • Google Scholar
56. 1. Llinás R
    2. Ribary U
    3. Contreras D
    4. Pedroarena C

    (1998) The neuronal basis for consciousness

    Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 353:1841–1849.

    https://doi.org/10.1098/rstb.1998.0336

    • PubMed
    • Google Scholar
57. 1. Logothetis NK
    2. Augath M
    3. Murayama Y
    4. Rauch A
    5. Sultan F
    6. Goense J
    7. Oeltermann A
    8. Merkle H

    (2010) The effects of electrical microstimulation on cortical signal propagation

    Nature Neuroscience 13:1283–1291.

    https://doi.org/10.1038/nn.2631

    • PubMed
    • Google Scholar
58. 1. Mair RG
    2. Burk JA
    3. Porter MC

    (1998) Lesions of the frontal cortex, hippocampus, and intralaminar thalamic nuclei have distinct effects on remembering in rats

    Behavioral Neuroscience 112:772–792.

    https://doi.org/10.1037/0735-7044.112.4.772

    • PubMed
    • Google Scholar
59. 1. Mair RG
    2. Hembrook JR

    (2008) Memory enhancement with event-related stimulation of the rostral intralaminar thalamic nuclei

    Journal of Neuroscience 28:14293–14300.

    https://doi.org/10.1523/JNEUROSCI.3301-08.2008

    • PubMed
    • Google Scholar
60. 1. Mair RG
    2. Onos KD
    3. Hembrook JR

    (2011) Cognitive activation by central thalamic stimulation: the yerkes-dodson law revisited

    Dose-Response 9:313–331.

    https://doi.org/10.2203/dose-response.10-017.Mair

    • PubMed
    • Google Scholar
61. 1. Mattis J
    2. Brill J
    3. Evans S
    4. Lerner TN
    5. Davidson TJ
    6. Hyun M
    7. Ramakrishnan C
    8. Deisseroth K
    9. Huguenard JR

    (2014) Frequency-dependent, cell type-divergent signaling in the hippocamposeptal projection

    Journal of Neuroscience 34:11769–11780.

    https://doi.org/10.1523/JNEUROSCI.5188-13.2014

    • PubMed
    • Google Scholar
62. 1. Mattis J
    2. Tye KM
    3. Ferenczi EA
    4. Ramakrishnan C
    5. O'Shea DJ
    6. Prakash R
    7. Gunaydin LA
    8. Hyun M
    9. Fenno LE
    10. Gradinaru V
    11. Yizhar O
    12. Deisseroth K

    (2012) Principles for applying optogenetic tools derived from direct comparative analysis of microbial opsins

    Nature Methods 9:159–172.

    https://doi.org/10.1038/nmeth.1808

    • Google Scholar
63. 1. Maxwell WL
    2. MacKinnon MA
    3. Smith DH
    4. McIntosh TK
    5. Graham DI

    (2006) Thalamic nuclei after human blunt head injury

    Journal of Neuropathology and Experimental Neurology 65:478–488.

    https://doi.org/10.1097/01.jnen.0000229241.28619.75

    • PubMed
    • Google Scholar
64. 1. McIntyre CC
    2. Grill WM

    (2002)

    Extracellular stimulation of central neurons: influence of stimulus waveform and frequency on neuronal output

    Journal of Neurophysiology 88:1592–1604.

    • PubMed
    • Google Scholar
65. 1. Mirsky AF
    2. Van Buren JM

    (1965) On the nature of the “absence” in centrencephalic epilepsy: a study of some behavioral, electroencephalographic and autonomic factors

    Electroencephalography and Clinical Neurophysiology 18:334–348.

    https://doi.org/10.1016/0013-4694(65)90053-2

    • PubMed
    • Google Scholar
66. 1. Mitchell AS
    2. Sherman SM
    3. Sommer MA
    4. Mair RG
    5. Vertes RP
    6. Chudasama Y

    (2014) Advances in understanding mechanisms of thalamic relays in cognition and behavior

    Journal of Neuroscience 34:15340–15346.

    https://doi.org/10.1523/JNEUROSCI.3289-14.2014

    • PubMed
    • Google Scholar
67. 1. Morison RS
    2. Dempsey EW

    (1942)

    A study of thalamo-cortical relations

    Am J Physiol 135:0281–0292.

    • Google Scholar
68. 1. Moruzzi G
    2. Magoun HW

    (1949) Brain stem reticular formation and activation of the EEG

    Electroencephalography and Clinical Neurophysiology 1:455–473.

    https://doi.org/10.1016/0013-4694(49)90219-9

    • PubMed
    • Google Scholar
69. 1. Newman LA
    2. Burk JA

    (2005) Effects of excitotoxic thalamic intralaminar nuclei lesions on attention and working memory

    Behavioural Brain Research 162:264–271.

    https://doi.org/10.1016/j.bbr.2005.03.018

    • PubMed
    • Google Scholar
70. 1. Nicolelis MAL
    2. Chapin JK
    3. Lin RCS

    (1992) Somatotopic maps within the zona incerta relay parallel GABAergic somatosensory pathways to the neocortex, superior colliculus, and brainstem

    Brain Research 577:134–141.

    https://doi.org/10.1016/0006-8993(92)90546-L

    • PubMed
    • Google Scholar
71. 1. Nicolelis MAL
    2. Chapin JK
    3. Lin RCS

    (1995) Development of direct GABAergic projections from the zona incerta to the somatosensory cortex of the rat

    Neuroscience 65:609–631.

    https://doi.org/10.1016/0306-4522(94)00493-O

    • PubMed
    • Google Scholar
72. 1. Pasley BN
    2. Inglis BA
    3. Freeman RD

    (2007) Analysis of oxygen metabolism implies a neural origin for the negative BOLD response in human visual cortex

    NeuroImage 36:269–276.

    https://doi.org/10.1016/j.neuroimage.2006.09.015

    • PubMed
    • Google Scholar
73. Book

    1. Paxinos G
    2. Watson C

    (2005)

    The Rat Brain in Stereotaxic Coordinates

    Burlington: Elsevier Academic Press.

    • Google Scholar
74. 1. Pinskiy V
    2. Tolpygo AS
    3. Jones J
    4. Weber K
    5. Franciotti N
    6. Mitra PP

    (2013) A low-cost technique to cryo-protect and freeze rodent brains, precisely aligned to stereotaxic coordinates for whole-brain cryosectioning

    Journal of Neuroscience Methods 218:206–213.

    https://doi.org/10.1016/j.jneumeth.2013.03.004

    • PubMed
    • Google Scholar
75. 1. Plum F

    (1991) Coma and related global disturbances of the human conscious state

    Normal and Altered States of Function 9:359–425.

    https://doi.org/10.1007/978-1-4615-6622-9_9

    • Google Scholar
76. 1. Porter JT
    2. Johnson CK
    3. Agmon A

    (2001)

    Diverse types of interneurons generate thalamus-evoked feedforward inhibition in the mouse barrel cortex

    Journal of Neuroscience 21:2699–2710.

    • PubMed
    • Google Scholar
77. 1. Schiff ND
    2. Giacino JT
    3. Kalmar K
    4. Victor JD
    5. Baker K
    6. Gerber M
    7. Fritz B
    8. Eisenberg B
    9. Biondi T
    10. O'Connor J
    11. Kobylarz EJ
    12. Farris S
    13. Machado A
    14. McCagg C
    15. Plum F
    16. Fins JJ
    17. Rezai AR

    (2007) Behavioural improvements with thalamic stimulation after severe traumatic brain injury

    Nature 448:600–603.

    https://doi.org/10.1038/nature06041

    • PubMed
    • Google Scholar
78. Book

    1. Schiff ND
    2. Pfaff DW

    (2009) Neural Perspectives on Activation and Arousal

    In: Schiff Nicholas D., Pfaff Donald W., editors. Handbook of neuroscience for the behavioral sciences. Hoboken, NJ, USA: John Wiley & Sons, Inc. pp. 454–460.

    https://doi.org/10.1002/9780470478509.neubb001024

    • Google Scholar
79. 1. Schiff ND
    2. Shah SA
    3. Hudson AE
    4. Nauvel T
    5. Kalik SF
    6. Purpura KP

    (2013) Gating of attentional effort through the central thalamus

    Journal of Neurophysiology 109:1152–1163.

    https://doi.org/10.1152/jn.00317.2011

    • PubMed
    • Google Scholar
80. 1. Schiff ND

    (2008) Central thalamic contributions to arousal regulation and neurological disorders of consciousness

    Annals of the New York Academy of Sciences 1129:105–118.

    https://doi.org/10.1196/annals.1417.029

    • PubMed
    • Google Scholar
81. 1. Schiff ND

    (2010) Recovery of consciousness after brain injury: a mesocircuit hypothesis

    Trends in Neurosciences 33:1–9.

    https://doi.org/10.1016/j.tins.2009.11.002

    • PubMed
    • Google Scholar
82. 1. Schiff ND

    (2012) Moving toward a generalizable application of central thalamic deep brain stimulation for support of forebrain arousal regulation in the severely injured brain

    Annals of the New York Academy of Sciences 1265:56–68.

    https://doi.org/10.1111/j.1749-6632.2012.06712.x

    • PubMed
    • Google Scholar
83. 1. Schridde U
    2. Khubchandani M
    3. Motelow JE
    4. Sanganahalli BG
    5. Hyder F
    6. Blumenfeld H

    (2008) Negative BOLD with large increases in neuronal activity

    Cerebral Cortex 18:1814–1827.

    https://doi.org/10.1093/cercor/bhm208

    • PubMed
    • Google Scholar
84. 1. Seidenbecher T
    2. Pape H-C

    (2001) Contribution of intralaminar thalamic nuclei to spike-and-wave-discharges during spontaneous seizures in a genetic rat model of absence epilepsy

    European Journal of Neuroscience 13:1537–1546.

    https://doi.org/10.1046/j.0953-816x.2001.01537.x

    • PubMed
    • Google Scholar
85. 1. Shammah-Lagnado SJ
    2. Negrão N
    3. Ricardo JA

    (1985) Afferent connections of the zona incerta: a horseradish peroxidase study in the rat

    Neuroscience 15:109–134.

    https://doi.org/10.1016/0306-4522(85)90127-7

    • PubMed
    • Google Scholar
86. 1. Shaw FZ
    2. Liao YF
    3. Chen RF
    4. Huang YH
    5. Lin RC

    (2013) The zona incerta modulates spontaneous spike-wave discharges in the rat

    Journal of Neurophysiology 109:2505–2516.

    https://doi.org/10.1152/jn.00750.2011

    • PubMed
    • Google Scholar
87. 1. Sherman SM
    2. Guillery RW

    (1996)

    Functional organization of thalamocortical relays

    Journal of Neurophysiology 76:1367–1395.

    • PubMed
    • Google Scholar
88. 1. Shirvalkar P
    2. Seth M
    3. Schiff ND
    4. Herrera DG

    (2006) Cognitive enhancement with central thalamic electrical stimulation

    Proceedings of the National Academy of Sciences of the United States of America 103:17007–17012.

    https://doi.org/10.1073/pnas.0604811103

    • PubMed
    • Google Scholar
89. 1. Shmuel A
    2. Augath M
    3. Oeltermann A
    4. Logothetis NK

    (2006) Negative functional MRI response correlates with decreases in neuronal activity in monkey visual area V1

    Nature Neuroscience 9:569–577.

    https://doi.org/10.1038/nn1675

    • PubMed
    • Google Scholar
90. 1. Silberberg G
    2. Markram H

    (2007) Disynaptic inhibition between neocortical pyramidal cells mediated by martinotti cells

    Neuron 53:735–746.

    https://doi.org/10.1016/j.neuron.2007.02.012

    • PubMed
    • Google Scholar
91. 1. Smith AC
    2. Shah SA
    3. Hudson AE
    4. Purpura KP
    5. Victor JD
    6. Brown EN
    7. Schiff ND

    (2009) A bayesian statistical analysis of behavioral facilitation associated with deep brain stimulation

    Journal of Neuroscience Methods 183:267–276.

    https://doi.org/10.1016/j.jneumeth.2009.06.028

    • PubMed
    • Google Scholar
92. 1. Smith Y
    2. Raju DV
    3. Pare JF
    4. Sidibe M

    (2004) The thalamostriatal system: a highly specific network of the basal ganglia circuitry

    Trends in Neurosciences 27:520–527.

    https://doi.org/10.1016/j.tins.2004.07.004

    • PubMed
    • Google Scholar
93. 1. Smith Y

    (2008) The thalamus

    Neuroscience in Medicine pp. 419–442.

    https://doi.org/10.1007/978-1-60327-455-5_26

    • Google Scholar
94. 1. Sotero RC
    2. Trujillo-Barreto NJ

    (2007) Modelling the role of excitatory and inhibitory neuronal activity in the generation of the BOLD signal

    NeuroImage 35:149–165.

    https://doi.org/10.1016/j.neuroimage.2006.10.027

    • PubMed
    • Google Scholar
95. 1. Steriade M
    2. Glenn LL

    (1982)

    Neocortical and caudate projections of intralaminar thalamic neurons and their synaptic excitation from midbrain reticular core

    Journal of Neurophysiology 48:352–371.

    • PubMed
    • Google Scholar
96. 1. Steriade M
    2. McCormick D
    3. Sejnowski T

    (1993) Thalamocortical oscillations in the sleeping and aroused brain

    Science 262:679–685.

    https://doi.org/10.1126/science.8235588

    • PubMed
    • Google Scholar
97. 1. Takata N
    2. Yoshida K
    3. Komaki Y
    4. Xu M
    5. Sakai Y
    6. Hikishima K
    7. Mimura M
    8. Okano H
    9. Tanaka KF

    (2015) Optogenetic activation of CA1 pyramidal neurons at the dorsal and ventral hippocampus evokes distinct brain-wide responses revealed by mouse fMRI

    PloS One 10:e0121417.

    https://doi.org/10.1371/journal.pone.0121417

    • PubMed
    • Google Scholar
98. 1. Trageser JC
    2. Burke KA
    3. Masri R
    4. Li Y
    5. Sellers L
    6. Keller A

    (2006) State-dependent gating of sensory inputs by zona incerta

    Journal of Neurophysiology 96:1456–1463.

    https://doi.org/10.1152/jn.00423.2006

    • PubMed
    • Google Scholar
99. 1. Trageser JC
    2. Keller A

    (2004) Reducing the uncertainty: gating of peripheral inputs by zona incerta

    Journal of Neuroscience 24:8911–8915.

    https://doi.org/10.1523/JNEUROSCI.3218-04.2004

    • PubMed
    • Google Scholar
100. 1. Van der Werf YD
     2. Weerts JGE
     3. Jolles J
     4. Witter MP
     5. Lindeboom J
     6. Scheltens P

     (1999) Neuropsychological correlates of a right unilateral lacunar thalamic infarction

     Journal of Neurology, Neurosurgery & Psychiatry 66:36–42.

     https://doi.org/10.1136/jnnp.66.1.36

     • Google Scholar
101. 1. Van der Werf YD
     2. Witter MP
     3. Groenewegen HJ

     (2002) The intralaminar and midline nuclei of the thalamus. anatomical and functional evidence for participation in processes of arousal and awareness

     Brain Research Reviews 39:107–140.

     https://doi.org/10.1016/S0165-0173(02)00181-9

     • PubMed
     • Google Scholar
102. 1. van Luijtelaar G
     2. Sitnikova E

     (2006) Global and focal aspects of absence epilepsy: the contribution of genetic models

     Neuroscience and Biobehavioral Reviews 30:983–1003.

     https://doi.org/10.1016/j.neubiorev.2006.03.002

     • PubMed
     • Google Scholar
103. 1. Velasco M
     2. Velasco F
     3. Velasco AL
     4. Brito F
     5. Jiménez F
     6. Marquez I
     7. Rojas B

     (1997) Electrocortical and behavioral responses produced by acute electrical stimulation of the human centromedian thalamic nucleus

     Electroencephalography and Clinical Neurophysiology 102:461–471.

     https://doi.org/10.1016/S0013-4694(96)95203-0

     • PubMed
     • Google Scholar
104. 1. Vogt BA
     2. Hof PR
     3. Friedman DP
     4. Sikes RW
     5. Vogt LJ

     (2008) Norepinephrinergic afferents and cytology of the macaque monkey midline, mediodorsal, and intralaminar thalamic nuclei

     Brain Structure & Function 212:465–479.

     https://doi.org/10.1007/s00429-008-0178-0

     • PubMed
     • Google Scholar
105. 1. Weitz AJ
     2. Fang Z
     3. Lee HJ
     4. Fisher RS
     5. Smith WC
     6. Choy M
     7. Liu J
     8. Lin P
     9. Rosenberg M
     10. Lee JH

     (2015) Optogenetic fMRI reveals distinct, frequency-dependent networks recruited by dorsal and intermediate hippocampus stimulations

     NeuroImage 107:229–241.

     https://doi.org/10.1016/j.neuroimage.2014.10.039

     • PubMed
     • Google Scholar

[Editors’ note: this article was originally rejected after discussions between the reviewers, but the authors were invited to resubmit after an appeal against the decision.]

Thank you for choosing to send your work entitled "Frequency-selective control of forebrain networks by central thalamus" for consideration at eLife. Your full submission has been evaluated by Timothy Behrens (Senior Editor), Andrew King (Reviewing Editor) and three peer reviewers, and the decision was reached after discussions between the reviewers. Based on our discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife.

Although we were impressed with the interdisciplinary approach used in this study, particularly the combination of optical stimulation and fMRI, as well as its potential therapeutic implications, a number of significant concerns were raised by the reviewers. While some of these can be dealt with by changes to the text, other methodological concerns will likely require further experiments. In particular, the effects of varying the frequency of optical stimulation appear to be confounded by changes in pulse duration and number. Two of the reviewers also felt that the section on the zona incerta was particularly weak.

Reviewer #1:

Liu and colleagues have used a combination of fMRI, optogenetics, electrophysiology and video EEG monitoring in a rodent model to map thalamocortical and forebrain neural circuits. This work has several important contributions. The authors show that mode of stimulation can change response patterns of neurons dramatically in these central circuits. In one case, high-frequency stimulation of excitatory relay neurons in the central thalamus caused a widespread positive BOLD signal in multiple forebrain regions.

In contrast, low frequency (10 Hz) stimulation led to less neural circuit modulation, negative BOLD signal in the somatosensory cortex and loss of consciousness. This low frequency stimulation was also associated with spindle like activity in the zona incerta. The authors postulate that the zona incerta may be playing a roll in the inhibition and spindling through a feedback circuit with the central thalamus.

The authors' work suggests that negative BOLD signals could be associated with large scale inhibition in brain networks.

The work draws a compelling distinction between stimulating at a given frequency and observing population spiking activity at a given frequency after stimulation. The optogenetics disambiguate these two factors by allowing the stimulation of a specific neuronal population.

Finally, the work has enormous translational potential. The authors demonstrate substantial forebrain network activation with stimulation of the central thalamus. This observation clearly speaks to the potential use of these findings to guide developmental work on DBS for therapeutic purposes. In particular, the authors show that changing frequency of stimulation can change the brain response from broad activation of forebrain circuits for high frequency stimulation to decreased arousal and negative BOLD signals for low frequency stimulation. One could imagine using the first stimulation pattern or patterns similar to it help enhance arousal in patients who are minimally conscious or in a vegetative states, whereas the second pattern might prove useful for treating seizures. At any rate, the results offer a principled way to investigate the basic science of DBS and to learn more about the brain's arousal circuits by combining two modalities.

The manuscript is well-written. It is a significant basic neuroscience and translational neuroscience contribution. I have no major concerns.

Reviewer #2:

In this study, Liu et al carry out a multilevel exploration of the role of central thalamus activation in the control of brain states. By using optogenetic activation of the central thalamus in combination with fMRI (and other techniques), they demonstrate how the activation of somas in the central thalamus is enough to induce large changes in cortical BOLD signal, later correlated with single unit firing. The authors demonstrate that while high (40-100 Hz) frequency stimulation of the central thalamus activates different cortical (and subcortical) areas causing arousal, low (10 Hz) frequencies reduce cortical activity. The responses of zona incerta to the same stimulation patterns are compatible with a potential role of this area in the switch between both effects, at least those concerning some areas of the cortex. Even when a definitive causal and mechanistic relationship is not established, the study is relevant for both basic neuroscience – it brings together the single neuron level to the whole brain activation and it gives a feasible systems level interpretation of the phenomena – and also for the applied, clinical implications.

Main concerns:

1) The inclusion of the zona incerta as a potential intermediary of the cortical inhibition at 10 Hz provides a potential mechanism and it is valuable. However the study of this is not well justified in the text. It is not even mentioned in the Introduction and it is not well justified in the Abstract. Why the zona incerta? What inputs does it receive? Is central thalamus projecting to zona incerta? For example in the Introduction all these explanations are given for the pallidum, but then the zona incerta is the one included in the study. In the anatomical study reported in the manuscript, the ZI is never mentioned as a place of distribution of the EYFP axons.

2) In the subsection “Low-frequency central thalamus stimulation drives incertal oscillations” it is said that ZI sends direct gabaergic projections to sensory cortex (somatosensory; Nicolelis, Chaplin and Lin, 1995), and the study is carried out by recording from S1 and Z1. However, this does not explain how the brain state switch would be in prefrontal cortex or other cortical areas. It is possible that the authors have considered this and therefore it should be explicitly discussed in the text.

3) In the study of ZI the 10 Hz thalamic stimulation evokes spike and wave discharges in the EEG but this is not reported in the other recordings in the cortex during 10 Hz stim. 10 Hz CT stimulation is reported silencing the cortex, however spike and wave is a rather visible activity that does not silence the neurons. Shouldn't the spike and wave be seen in the single unit recordings?

4) Subsection “Stimulation of central thalamus relay neurons drives widespread forebrain activation in vivo”, fourth paragraph: stimulating at different frequencies always has the problem of giving the same number of pulses versus giving the stimulation for the same period of time. I understand that here the second option was taken. This means that the number of pulses was (in 6 minutes) 1200 when at 10 Hz, 4800 when at 40 Hz and 12000 at 100 Hz. In order to rule out the effect of the differential number of pulses, and since there are six blocks of repetitions, it would strengthen the conclusions to demonstrate that the differences in cortical activation remain for equivalent number of pulses. Probably the authors already have some data showing this that could be included in a supplemental figure. It is also confusing that Figure 4B shows stimulation periods that are variable and therefore (probably) adjusted to number of pulses. Was a different protocol used in this case?

5) In general there is a lack of description of the control (pre-stimulus) activity. Is the departing point slow wave activity? Is the decrease in firing during 10 Hz stimulation independent from the pre-existing activity patterns?

Reviewer #3:

This study deals with an interesting issue, the role of central thalamic nuclei in controlling cortical activation. However, there are several problems with this paper.

First, the authors seem to have a poor understanding of the effects that thalamic stimulation delivered at different frequencies has on the cortex. This has been studied previously but there is very little reflection of that knowledge here. I can think of a series of studies in the rat that investigated the cortical effects of frequency-dependent electrical stimulation in the VL/VA thalamic region (adjacent to the targeted CL/PC region) in rats. Those studies showed that stimulation in various thalamic nuclei evoke short-latency frequency-depend excitatory cortical responses that can augment (spreading significantly when stimulating in VL/VA) or depress (when stimulating in VPM/VPL). Importantly, at 10 Hz, stimulation always drives strong long-lasting feedforward inhibition, which hyperpolarizes cortical cells immediately after each short-latency excitatory response (e.g. see Fig. 1 in Castro-Alamancos and Connors, 1996b). This inhibition of cortical cells may explain the negative BOLD observed in S1 during 10 Hz stimulation (20 sec); under this stimulation regimen, cortical cells will be hyperpolarized most of the time.

Second, the idea that the suppression of neural activity observed at 10 Hz would be caused by the zona incerta seems unlikely. There are numerous other possibilities. For example, the very robust drive of feedforward inhibition in neocortex at 10 Hz is more parsimonious. In general, this part of the study is very unconvincing.

Third, there are significant methodological issues to consider. Most important is the fact that the authors chose to keep the duty cycle constant across stimulation frequencies. This means that at 10 Hz the light stimulus per cycle lasts 30 ms and at 100 Hz it lasts 3 ms. This is very problematic because the changes that are being attributed to frequency can easily be attributed to pulse duration, which has very significant effects on the responses evoked. Thus, everywhere in the paper where the authors allude to frequency, they are also changing pulse duration, which is a major confound. The only reasonable way to keep the total light duration constant when studying frequency is by increasing the train duration for the lower frequencies and keeping the pulse duration constant.

Fourth, the selectivity of the stimulation, which is based mostly on the placement of the fiber, is a major limitation because the nuclei that are targeted are very small and have complex shapes. It seems inevitable that adjacent structures will also be driven by the light. A Cre-mouse that would limit the expression to the targeted nuclei would be a much better choice in this particular case.

Fifth, regarding the idea that the thalamus serves to drive cortical activation, previous work has shown that neuromodulation of neural activity in sensory thalamus leads to activation or deactivation in the related sensory cortex (Hirata and Castro-Alamancos, 2010). This work is directly related to the study at hand but was not considered here.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for submitting your work entitled "Frequency-selective control of forebrain networks by central thalamus" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by Tim Behrens (Senior Editor) and Andrew King (Reviewing Editor).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Several of their original concerns have been satisfactorily addressed in the revised version and the reviewers have acknowledged the improvements made to the manuscript, particularly through the inclusion of the new pulse width and train duration imaging data. Nevertheless, some substantive issues remain that need to be addressed before a final decision can be made (references cited are given below).

1) One of the reviewers argues that your claim that the ZI is the source of cortical inhibition during 10 Hz stimulation is still not well supported. Whilst agreeing that thalamocortical responses evoked in the sensory cortices by stimulating the sensory thalamus at 10 Hz are depressed both in vivo (Castro-Alamancos and Connors, 1996b) and in slices (Cruikshank et al., 2010), the reviewer has pointed out that thalamocortical responses evoked in neocortex by stimulating other thalamic nuclei (midline, intralaminar, and VA/VL) are different in most respects, both in vivo (Castro-Alamancos and Connors, 1996b) and in slices (Cruikshank et al., 2012). Stimulation of these other thalamic nuclei in vivo at 10 Hz produces facilitating (spindle-like) responses in cortex, not depressing responses. In fact, work in slices also supports this. A direct quote from the in vitro paper (Cruikshank et al., 2012) cited by the authors in the rebuttal emphasizes this, "responses during repetitive stimulation were far more sustained for matrix than for core sensory TC pathways".

Since the 1940s it has been known that stimulating thalamic nuclei at different frequencies evokes three main types of responses in cortex (Dempsey and Morison): primary responses, augmenting responses and recruiting responses. A more recent account of this literature, including more recent studies done in the rat (Dempsey and Morison worked in cats), is presented in Castro-Alamancos and Connors, 1997 (not cited by the authors, but highly relevant). Primary responses occur when specific thalamic nuclei are stimulated, such as sensory thalamus. Primary responses depress with frequency. Augmenting and recruiting responses occur when VL, VA, VM, midline or intralaminar thalamic nuclei are stimulated. These responses are similar in that they produce enhanced cortical responses at stimulation frequencies between 7-14 Hz. They differ in their cortical depth profile, indicating that they target different layers (as expected from the known projections), and in the size of the primary response they evoke. VL stimulation evokes augmenting responses with a small primary response (small compared to sensory pathways) in motor cortical areas, which is surface positive. In contrast, recruiting responses are mostly surface negative, lacking the surface positive primary response of augmenting responses. The present study did not record field potential responses in cortex, so it is difficult to know if the responses they evoke at 10 Hz in cortex are augmenting or recruiting. But a consistent finding from intracellular recordings is that strong feedforward inhibition is evoked in cortex in vivo by stimulating the thalamus at 10 Hz (Castro-Alamancos and Connors, 1996c; Castro-Alamancos and Connors, 1996a). In fact, the area stimulated by the authors is known to evoke recruiting responses, which should yield feedforward inhibitory responses with weaker short-latency excitation (surface positivity) than augmenting responses.

The reviewer concludes that by citing papers based on specific thalamic nuclei, such as core sensory thalamic nuclei (Cruikshank et al., 2010), is not appropriate, because it is well known that the pathways stimulated in this study have different response properties.

2) The reviewer has also criticized your argument that high frequency activity evokes increased levels of interneuron-driven cortical inhibition relative to excitation, whereas low frequency activity does not, on the grounds that the relevance of the cited paper (Galarreta and Hestrin, 1998) to the issue at hand is doubtful because this study only deals with inhibition driven by cortical stimuli, not thalamic stimuli. You need to look at feedforward inhibition driven by thalamic stimulation, preferably in vivo. The other cited slice paper (Cruikshank et al., 2012), looked at feedforward inhibition driven by thalamic inputs but only marginally (only 1 example panel in Figure 6D and no group data). It is hard to derive any strong conclusions other than to say that inhibition is there, albeit depressed compared to the 1st stimulus.

3) The reviewer also points out that white matter stimulation does not mimic thalamic stimulation, let alone central thalamus stimulation, so this should not be used as an argument for the effects of thalamic stimulation on cortical activity.

4) In response to your comment that, in contrast to Fig. 1 in Castro-Alamancos and Connors (1996b), you did not observe spikes on every stimulus presentation at 10 Hz, the reviewer points out that this is based on VL stimulation and recording in the motor cortex, which evokes an augmenting response. If in that figure (or Fig. 3 of the same paper) the surface positive field potential component and associated corresponding excitatory drive (which produces the spikes) are eliminated, what is left is a recruiting response, as defined originally by Dempsey and Morison. Similar to spindles, recruiting responses trigger spikes only on some of the waves at 10 Hz. Thus, 3.6 Hz would be consistent with this. It would be helpful if you were to present PSTHs using each stimulus of the 10 Hz train as the marker (Figure 3D).

5) Your evidence that inhibition of the zona incerta during 10 Hz central thalamus stimulation reverses the cortical inhibition effect is based on AAV delivery of halorhodopsin, which will not discriminate between inhibitory and excitatory neurons. Consequently, light delivery might not only inhibit ZI cells, but other cells too, including adjacent thalamocortical cells. It is therefore possible that the effects observed are due to inhibition of thalamocortical cells. Similarly, could the mCherry fibers observed in cortex actually be thalamocortical fibers? It is also possible that your stimulation could be activating the NRt, which would drive the observed spindles and the recruiting responses in the cortex at 10 Hz, in which case there would be no need for the extrinsic inhibition purportedly provided by ZI. These possibilities should be addressed in the paper. Unless they can be ruled out, conclusions about the role of the zona incerta should be modified accordingly.

6) In discussing alternative hypotheses of the mechanism of cortical inhibition, including disynaptic thalamocortical inhibition (see Discussion, paragraph four) and disynaptic corticocortical inhibition (see Discussion, paragraph six), you should include relevant literature (e.g. rat in vivo thalamic stimulation), instead of marginally related work (slice studies that used white-matter or cortical stimulation, hippocampal studies, etc.). Care should also be taken to cite the literature accurately. For example, you argue that "central thalamus at different frequencies can lead to distinct behavioral responses" and cite Morison and Dempsey (1942), but Dempsey and Morison did not study behavior. Galarreta and Hestrin (1998) did not study thalamocortical responses (please correct this in the subsection “Central thalamus stimulation frequency controls cortical excitation/inhibition balance”).

References cited:

Cruikshank, S.J., H. Urabe, A.V. Nurmikko, and B.W. Connors, Pathway-specific feedforward circuits between thalamus and neocortex revealed by selective optical stimulation of axons. Neuron, 2010. 65(2): p. 230-45.

Galarreta, M. and S. Hestrin, Frequency-dependent synaptic depression and the balance of excitation and inhibition in the neocortex. Nat Neurosci, 1998. 1(7): p. 587-94.

[Editors’ note: the author responses to the first round of peer review follow.]

We were very pleased to see that two of the three reviewers found our manuscript suitable for publication, with such comments as “This work has several important contributions,” “The work has enormous translational potential,” and “The study is relevant for both basic neuroscience… and also for the applied, clinical implications.”

We have fully addressed the few criticisms offered by the reviewers through additional clarification in writing, experiments, and analyses in the attached revised manuscript. Below, we provide a point-by-point response to the comments by Reviewers #2 and #3. In the instances where we omitted citing key literature, we sincerely apologize for the oversight and have addressed the issue carefully.

Given the overwhelmingly positive reviews from two of the three reviewers, and our direct rectification of the two issues noted in our rejection letter (frequency vs. pulse width effects, and interpretation of zona incerta recordings) with additional experiments, we request the opportunity to have this revised manuscript re-evaluated by the reviewers.

Reviewer #2:

1) The inclusion of the zona incerta as a potential intermediary of the cortical inhibition at 10 Hz provides a potential mechanism and it is valuable. However the study of this is not well justified in the text. It is not even mentioned in the Introduction and it is not well justified in the Abstract. Why the zona incerta? What inputs does it receive? Is central thalamus projecting to zona incerta? For example in the Introduction all these explanations are given for the pallidum, but then the zona incerta is the one included in the study. In the anatomical study reported in the manuscript, the ZI is never mentioned as a place of distribution of the EYFP axons.

We thank the reviewer for their critical feedback of our writing and have revised the manuscript accordingly to address the listed concerns. Specifically, in the Introduction, we provide context for our investigation of zona incerta, citing prior demonstrations of its influence on attention and arousal, and explain that its possible involvement in central thalamus arousal circuits has not yet been studied. Furthermore, in the Abstract, we more explicitly justify the investigation as one to explore the mechanism underlying the negative BOLD signal observed in cortex.

While we speculate that a monosynaptic connection may exist between the transfected neurons in central thalamus and zona incerta based on observed distributions of EYFP axons (see subsection “Cortical inhibition driven by central thalamus stimulation depends on evoked incertal activity”), previous tracing studies have failed to identify significant input to zona incerta from intralaminar nuclei. This clarification has been added to the Discussion. We have kept our discussion of thalamo-striatal projections in the Introduction to provide context for the strong BOLD responses evoked in this region.

2) In the subsection “Low-frequency central thalamus stimulation drives incertal oscillations” it is said that ZI sends direct gabaergic projections to sensory cortex (somatosensory; Nicolelis, Chaplin and Lin, 1995), and the study is carried out by recording from S1 and Z1. However, this does not explain how the brain state switch would be in prefrontal cortex or other cortical areas. It is possible that the authors have considered this and therefore it should be explicitly discussed in the text.

To address this concern, we have included several hypotheses in the revised manuscript. For example, during 10 Hz stimulation, cortical deactivation in EEG and the absence of significant forebrain recruitment with fMRI may be controlled by spindle-like oscillations, which were observed in zona incerta, and associated sleep networks (see Discussion). GABAergic projections from zona incerta to central thalamus may also act as a form of negative feedback to suppress the region’s broad thalamocortical projections.

3) In the study of ZI the 10 Hz thalamic stimulation evokes spike and wave discharges in the EEG but this is not reported in the other recordings in the cortex during 10 Hz stim. 10 Hz CT stimulation is reported silencing the cortex, however spike and wave is a rather visible activity that does not silence the neurons. Shouldn't the spike and wave be seen in the single unit recordings?

As correctly summarized by the reviewer, we found that 10 Hz stimulation of central thalamus evokes spike-and-wave discharges in EEG and neuronal silencing in single-unit recordings. Although the reviewer suggests that these phenomena are contradictory, the spatial localization of the two recordings are quite distinct. The decrease in single-unit firing was measured specifically in somatosensory cortex, which was the only region to exhibit a negative BOLD signal (For example, motor and cingulate cortex exhibited positive responses; Figure 2A). In contrast, the EEG electrode was placed on the dura and located above frontal cortex. Therefore, the recorded signal represents the integration of a larger volume that likely includes excited regions (e.g. motor and cingulate cortex). Indeed, the reviewer’s comment highlights the novelty of our results. While our observation of cortical spike-and-wave activity is consistent with previous studies of thalamic stimulation, the whole-brain visualization afforded by fMRI allowed us to predict the presence of neuronal silencing in a highly localized region of cortex, which we later went on to confirm with electrophysiology. This demonstrates that the whole-brain visualization afforded by ofMRI is a key advantage over other methods in deconstructing complex neural circuits.

4) Subsection “Stimulation of central thalamus relay neurons drives widespread forebrain activation in vivo”, Fourth paragraph: stimulating at different frequencies always has the problem of giving the same number of pulses versus giving the stimulation for the same period of time. I understand that here the second option was taken. This means that the number of pulses was (in 6 minutes) 1200 when at 10 Hz, 4800 when at 40 Hz and 12000 at 100 Hz. In order to rule out the effect of the differential number of pulses, and since there are six blocks of repetitions, it would strengthen the conclusions to demonstrate that the differences in cortical activation remain for equivalent number of pulses. Probably the authors already have some data showing this that could be included in a supplemental figure. It is also confusing that Figure 4B shows stimulation periods that are variable and therefore (probably) adjusted to number of pulses. Was a different protocol used in this case?

The reviewer is correct in their observation that stimulation frequency cannot be varied independently of all other parameters, including pulse width, total light delivery, and pulse train duration. In this experiment, we kept the duty cycle constant in order to mitigate the possible confounding factor of variations in total light delivery, which can lead to heating artifacts (Christie et al., 2013). We apologize if this was not clear to the reviewer, and have clarified this decision in the revised manuscript (Results; Figure 4B legend).

As a result of keeping the duty cycle constant, the pulse width associated with each frequency was necessarily different (30 ms for 10 Hz, 7.5 ms for 40 Hz, and 3 ms for 100 Hz). The alternative option – keeping the pulse width fixed and varying the pulse train duration to maintain total light delivery – would lead to uninterpretable results due to differences in neuronal adaptation that occur over the course of a 20 s pulse train compared to a 2 s pulse train (Note that a 10-fold difference would be required for comparing 10 and 100 Hz).

To address this concern as best as possible, we have performed additional fMRI experiments keeping both the pulse width and train duration constant. Our results show that both the increase in forebrain activation with frequency and the transition from negative to positive BOLD signals in cortex is preserved (Figure 2—figure supplement 3). Although these stimulus trains deliver different amounts of total light, the fact that we see these phenomena while varying frequency and keeping either the total amount of light delivery (Figures 2 and 3) or pulse width (Figure 2—figure supplement 3) constant suggests that frequency is the main factor in driving these effects.

5) In general there is a lack of description of the control (pre-stimulus) activity. Is the departing point slow wave activity? Is the decrease in firing during 10 Hz stimulation independent from the pre-existing activity patterns?

To address this concern, we have added a supplementary figure, which shows that the pre-stimulus forebrain EEG activity was consistent across frequencies of stimulation (Figure 6—figure supplement 1). Thus, the differences in evoked behavior and electrographic responses can be attributed to the stimulation paradigm itself.

With regards to the decrease in cortical firing, the magnitude of change during 10 Hz stimulation was well correlated with the pre-stimulus firing rate (n = 46 neurons; R² = 0.78, see Author response image 1). This correlation is consistent with expectations, since cells which have a high baseline firing rate can be suppressed more than those with a low baseline firing rate. The data also show that cortical cells exhibiting a wide range of baseline firing rate were inhibited, emphasizing the robustness of central thalamus-driven cortical inhibition. To generate this plot, the average firing rate over the 20 s period before stimulation was compared with the average firing rate during the 5-10 s interval after stimulation began (that is, when the most neurons were silenced; see Table 1).

Author response image 1

Download asset Open asset

Because the single-unit recordings in cortex were collected without simultaneous forebrain EEG, we cannot compare the magnitude of neuronal silencing to the broader state of cortex prior to stimulation.

Reviewer #3:

This study deals with an interesting issue, the role of central thalamic nuclei in controlling cortical activation. However, there are several problems with this paper. First, the authors seem to have a poor understanding of the effects that thalamic stimulation delivered at different frequencies has on the cortex. This has been studied previously but there is very little reflection of that knowledge here. I can think of a series of studies in the rat that investigated the cortical effects of frequency-dependent electrical stimulation in the VL/VA thalamic region (adjacent to the targeted CL/PC region) in rats. Those studies showed that stimulation in various thalamic nuclei evoke short-latency frequency-depend excitatory cortical responses that can augment (spreading significantly when stimulating in VL/VA) or depress (when stimulating in VPM/VPL). Importantly, at 10 Hz, stimulation always drives strong long-lasting feedforward inhibition, which hyperpolarizes cortical cells immediately after each short-latency excitatory response (e.g. see Fig. 1 in Castro-Alamancos and Connors, 1996b). This inhibition of cortical cells may explain the negative BOLD observed in S1 during 10 Hz stimulation (20 sec); under this stimulation regimen, cortical cells will be hyperpolarized most of the time.

We thank the reviewer for their critical feedback of our writing and data interpretation, and apologize for several key studies being omitted from our Introduction/Discussion that we now include. In particular, we now discuss feedforward inhibition and the possibility that it underlies the observed frequency-dependent silencing of sensory cortex. After careful consideration of existing literature and our own results, we conclude that feedforward thalamocortical inhibition is unlikely to explain the effects reported here. Several lines of evidence point to this conclusion. First, cortical IPSPs driven by feedforward inhibition typically depress after a few repeated stimuli (Cruikshank et al., 2012; Cruikshank et al., 2010). Second, high frequency activity evokes increased levels of interneuron-driven cortical inhibition relative to excitation, whereas low frequency activity does not (Galarreta and Hestrin, 1998). Third, white matter stimulation increases inhibitory synaptic activity in cortex during 40 Hz, but not 10 Hz, stimulation (Contreras and Llinas, 2001). The figure highlighted by the reviewer (Fig. 1 in Castro-Alamancos and Connors, 1996b) is informative in highlighting the frequency-dependent properties of thalamic stimulation, but suggests that 10 Hz stimulation still evokes cortical spikes with every stimulus. We did not observe this type of response, with the average firing rate in cortex being 3.6 ± 3.4 Hz (n = 11 neurons, mean ± std) during the 20 s period of 10 Hz stimulation.

To provide additional support that the zona incerta plays a role in the observed cortical inhibition, we performed a new series of experiments in which zona incerta is inhibited using optogenetics during 10 Hz stimulation of central thalamus (Figure 5). We report that targeted inhibition of the zona incerta during 10 Hz central thalamus stimulation reverses the cortical inhibition effect. That is, suppressing zona incerta activity has a net excitatory effect on cortex. These data suggest that zona incerta plays a causal role in contributing to the cortical inhibition reported.

Second, the idea that the suppression of neural activity observed at 10 Hz would be caused by the zona incerta seems unlikely. There are numerous other possibilities. For example, the very robust drive of feedforward inhibition in neocortex at 10 Hz is more parsimonious. In general, this part of the study is very unconvincing.

As noted above, we have performed a new series of experiments using optical inhibition of zona incerta to demonstrate this region’s causal role in the frequency-dependent inhibition of cortex. In addition, we have significantly revised the manuscript to discuss alternative hypotheses of the cortical inhibition’s mechanism, including disynaptic thalamocortical inhibition (see Discussion, paragraph four) and disynaptic corticocortical inhibition (see Discussion, paragraph six).

Third, there are significant methodological issues to consider. Most important is the fact that the authors chose to keep the duty cycle constant across stimulation frequencies. This means that at 10 Hz the light stimulus per cycle lasts 30 ms and at 100 Hz it lasts 3 ms. This is very problematic because the changes that are being attributed to frequency can easily be attributed to pulse duration, which has very significant effects on the responses evoked. Thus, everywhere in the paper where the authors allude to frequency, they are also changing pulse duration, which is a major confound. The only reasonable way to keep the total light duration constant when studying frequency is by increasing the train duration for the lower frequencies and keeping the pulse duration constant.

As explained in our response to Reviewer #2, we chose to keep the duty cycle constant in these experiments to keep both the total amount of light delivery and pulse train duration constant. Maintaining a constant amount of light delivery is important to mitigate possible confounding factors associated with heating (Christie et al., 2013). Similarly, maintaining constant pulse train duration is necessary to avoid possible differences in neuronal adaptation that would occur when changing the train duration by such a large factor (e.g. an order of magnitude between 10 and 100 Hz).

To provide additional insight regarding this potential confound, we performed additional fMRI experiments keeping both the pulse width and train duration constant. Our results show that the transition from negative to positive BOLD signals in cortex is preserved, as well as the changes in activation volume in cortex and striatum (Figure 2—figure supplement 3). Although these stimulus trains deliver different amounts of total light, the fact that we see these phenomena while varying frequency and keeping either the total amount of light delivery (Figures 2 and 3) or pulse width (Figure 2—figure supplement 3) constant suggests that frequency is the main determinant in driving these effects.

Fourth, the selectivity of the stimulation, which is based mostly on the placement of the fiber, is a major limitation because the nuclei that are targeted are very small and have complex shapes. It seems inevitable that adjacent structures will also be driven by the light. A Cre-mouse that would limit the expression to the targeted nuclei would be a much better choice in this particular case.

We agree with the reviewer that the issue of stimulation selectivity is a very important one, and we paid careful attention to it throughout the experimental design process. In particular, we used a large sample size (n = 10-16 animals) and ensured that the stimulation location was within a very narrow range across animals using high-resolution in vivo imaging and excluding over half the original cohort (n = 47) from analysis. To ensure that stimulation was primarily limited to the targeted nuclei, we also titrated the power level to achieve an activation cone of appropriate scale (Figure 1A). Given these measures, the authors are confident in the claim that the effects reported in this manuscript are primarily driven by the intralaminar nuclei of central thalamus. Because there is no genetic marker that differentiates intralaminar nuclei from the rest of thalamus, development of a Cre-mouse line is currently not possible. Nevertheless, we agree that such an experiment would be ideal.

Fifth, regarding the idea that the thalamus serves to drive cortical activation, previous work has shown that neuromodulation of neural activity in sensory thalamus leads to activation or deactivation in the related sensory cortex (Hirata and Castro-Alamancos, 2010). This work is directly related to the study at hand but was not considered here.

We sincerely apologize for this oversight. We have added this reference to the revised manuscript.

References cited:

Cruikshank SJ, Urabe H, Nurmikko AV, & Connors BW (2010) Pathway-specific

feedforward circuits between thalamus and neocortex revealed by selective

optical stimulation of axons. Neuron 65(2):230-245.

Galarreta M & Hestrin S (1998) Frequency-dependent synaptic depression and

the balance of excitation and inhibition in the neocortex. Nat Neurosci 1(7):587-

594.

Contreras D & Llinas R (2001) Voltage-sensitive dye imaging of neocortical

spatiotemporal dynamics to afferent activation frequency. J Neurosci

21(23):9403-9413.

[Editors’ note: the author responses to the re-review follow.]

Several of their original concerns have been satisfactorily addressed in the revised version and the reviewers have acknowledged the improvements made to the manuscript, particularly through the inclusion of the new pulse width and train duration imaging data. Nevertheless, some substantive issues remain that need to be addressed before a final decision can be made (references cited are given below).

Many of the findings presented in our study are consistent with the anatomical and physiological properties of thalamus and thalamocortical projections that have been documented separately over the last half century. It is notable, however, that our experimental paradigm is unique and fundamentally different compared to those employed in previous studies due to our novel approach of selective excitation of central thalamic relay neurons using optogenetic techniques and the visualization of whole-brain response patterns using fMRI. The several well-characterized thalamocortical properties cited by the reviewer originate from thalamic stimulation in related but distinct experimental preparations (e.g. electrical microstimulation or optogenetic stimulation in reduced preparations), which though clearly directly related, may produce different effects. The precise frequency selectivity of the whole-brain network responses shown here has simply not been demonstrated before (other approaches have been limited to looking at only some pieces of the network at one time). With that said, we agree that many aspects of our findings have interpretations in the context of prior literature. One reviewer relates our findings to the well-known recruiting response, which we agree could possibly co-exist with and/or contribute to the observed responses (see Discussion paragraph five). However, there is a much more clear and compelling comparison to a fourth thalamocortical physiological mode described since the 1940s (in addition to the three already noted by the reviewer) – the production of high voltage spike-wave activity associated with behavioral arrest. This mode is directly supported by our data, as we show that 10 Hz stimulations generate spike-wave EEG and behavioral arrest patterns, and concretely link our stimulation paradigm to the zona incerta (see below). We now more clearly present the evidence to frame and support this interpretation in the revised manuscript.

1) One of the reviewers argues that your claim that the ZI is the source of cortical inhibition during 10 Hz stimulation is still not well supported. Whilst agreeing that thalamocortical responses evoked in the sensory cortices by stimulating the sensory thalamus at 10 Hz are depressed both in vivo (Castro-Alamancos and Connors, 1996b) and in slices (Cruikshank et al., 2010), the reviewer has pointed out that thalamocortical responses evoked in neocortex by stimulating other thalamic nuclei (midline, intralaminar, and VA/VL) are different in most respects, both in vivo (Castro-Alamancos and Connors, 1996b) and in slices (Cruikshank et al., 2012). Stimulation of these other thalamic nuclei in vivo at 10 Hz produces facilitating (spindle-like) responses in cortex, not depressing responses. In fact, work in slices also supports this. A direct quote from the in vitro paper (Cruikshank et al., 2012) cited by the authors in the rebuttal emphasizes this, "responses during repetitive stimulation were far more sustained for matrix than for core sensory TC pathways".

We agree with the reviewer’s claim that thalamocortical inhibition mediated by cortical interneurons cannot be ruled out completely. We would like to thank the reviewer for pointing this out and have revised the manuscript accordingly. First, we changed the wording in the Discussion from “A competing hypothesis to explain the suppression of cortex […]” to “An additional mechanism that could contribute to the evoked suppression of cortex […]”. We believe this highlights the possibility that feedforward thalamocortical inhibition may play a role in the observed response, while also acknowledging the novel, causal role of zona incerta discovered in our experiments. Second, we removed our previous argument that thalamocortical inhibition becomes depressed after repeated stimuli, including the citation that referenced stimulation of the VB complex. Furthermore, we make note of Cruikshank et al.’s recent finding that the matrix neurons of central thalamus can generate relatively sustained IPSPs (see Discussion, paragraph four). We conclude that our results are compatible with feedforward thalamocortical inhibition acting in conjunction with the novel zona incerta-modulated inhibition revealed here.

We also added citation (Castro-Alamancos and Connors, 1996b) listed by the reviewer in reference to short-term plasticity of thalamocortical synapses.

Since the 1940s it has been known that stimulating thalamic nuclei at different frequencies evokes three main types of responses in cortex (Dempsey and Morison): primary responses, augmenting responses and recruiting responses. A more recent account of this literature, including more recent studies done in the rat (Dempsey and Morison worked in cats), is presented in Castro-Alamancos and Connors, 1997 (not cited by the authors, but highly relevant). Primary responses occur when specific thalamic nuclei are stimulated, such as sensory thalamus. Primary responses depress with frequency. Augmenting and recruiting responses occur when VL, VA, VM, midline or intralaminar thalamic nuclei are stimulated. These responses are similar in that they produce enhanced cortical responses at stimulation frequencies between 7-14 Hz. They differ in their cortical depth profile, indicating that they target different layers (as expected from the known projections), and in the size of the primary response they evoke. VL stimulation evokes augmenting responses with a small primary response (small compared to sensory pathways) in motor cortical areas, which is surface positive. In contrast, recruiting responses are mostly surface negative, lacking the surface positive primary response of augmenting responses. The present study did not record field potential responses in cortex, so it is difficult to know if the responses they evoke at 10 Hz in cortex are augmenting or recruiting. But a consistent finding from intracellular recordings is that strong feedforward inhibition is evoked in cortex in vivo by stimulating the thalamus at 10 Hz (Castro-Alamancos and Connors, 1996c; Castro-Alamancos and Connors, 1996a). In fact, the area stimulated by the authors is known to evoke recruiting responses, which should yield feedforward inhibitory responses with weaker short-latency excitation (surface positivity) than augmenting responses. The reviewer concludes that by citing papers based on specific thalamic nuclei, such as core sensory thalamic nuclei (Cruikshank et al., 2010), is not appropriate, because it is well known that the pathways stimulated in this study have different response properties.

We agree with the reviewer that papers based on stimulation of specific thalamic nuclei should be avoided and removed citation (Cruikshank et al., 2010) from the Discussion. We also include citations (Castro-Alamancos and Connors, 1997) and (Castro-Alamancos and Connors, 1996c) listed above, and discuss the recruiting response and its hypothesized mechanism of hyperpolarization driven by thalamic stimulation. However, we wish to add that the list of three cortical responses described by the reviewer (primary, augmenting, and recruiting) omits another key mode of thalamocortical activation – namely, the generation of spike-wave cortical activity and behavioral arrest that is known in human subjects as the absence seizure and is modeled in various rodent strains. Indeed, the reference provided by the reviewer (Castro-Alamancos and Connors, 1997) also ends with a discussion of this general phenomenon. However, it is not presented as an isolated mode of cortical response, likely because it is a network property and is not well isolated in the experimental contexts in which the other three responses are characterized. We would point out that our experimental framework of whole-brain imaging is uniquely suited to characterize the circuit-level differences of stimulation frequency across the entire brain during sustained, steady-state activation (approximating the typical approach used in deep brain stimulation) rather than the methods used to study primary, augmenting, and recruiting responses (which are generated by shorter duration stimulation paradigms and lack specific behavioral correlates).

Here, we demonstrate that 10 Hz optogenetic stimulation during sleep produces slow spike-wave activity and freezing behavior in the majority of animals, consistent with absence seizure-like responses. We are led to zona incerta directly from these data and existing literature on rodent absence seizures that points to the ZI’s role in modulating spike-wave discharges (see Introduction). We also wish to emphasize that we are not the first group to suggest a key role for incertofugal inhibition in spike-wave activity or to probe this possibility experimentally. For example, one study (Shaw et al., 2013) concluded, “Membrane hyperpolarization due to the intensive inhibitory GABAergic actions is beneficial for maintaining brain rhythms […] Delayed rhythmic incertal activity may help to develop membrane hyperpolarization that builds up sustained high-voltage cortical rhythms through GABAergic incertofugal pathways.” This point has been clarified in the revised manuscript (see Discussion).

2) The reviewer has also criticized your argument that high frequency activity evokes increased levels of interneuron-driven cortical inhibition relative to excitation, whereas low frequency activity does not, on the grounds that the relevance of the cited paper (Galarreta and Hestrin, 1998) to the issue at hand is doubtful because this study only deals with inhibition driven by cortical stimuli, not thalamic stimuli. You need to look at feedforward inhibition driven by thalamic stimulation, preferably in vivo. The other cited slice paper (Cruikshank et al., 2012), looked at feedforward inhibition driven by thalamic inputs but only marginally (only 1 example panel in Figure 6D and no group data). It is hard to derive any strong conclusions other than to say that inhibition is there, albeit depressed compared to the 1st stimulus.

We agree with the reviewer and removed this argument (including citation Galarreta and Hestrin, 1998) from the Discussion. Unfortunately, to the best of our knowledge, no study has directly compared the frequency-dependent strength of cortical interneuron-mediated thalamocortical inhibition during intralaminar thalamic stimulation in vivo. Therefore, it is difficult to predict the extent to which this inhibition would vary between our 10, 40, and 100 Hz stimulations. Intracellular recordings would be an appropriate means to further investigate this issue, a discussion point we have added to the revised manuscript. Until these experiments are conducted, our results stand on their own as a novel and exciting demonstration that zona incerta plays a causal role in modulating cortical inhibition induced by central thalamus stimulation. While zona incerta has previously been implicated in modulating spike-wave discharges and associated absence seizures, we were able to link this relationship to specific whole-brain activity patterns and a previously described inhibitory pathway originating from ZI.

3) The reviewer also points out that white matter stimulation does not mimic thalamic stimulation, let alone central thalamus stimulation, so this should not be used as an argument for the effects of thalamic stimulation on cortical activity.

We agree that white matter stimulation does not accurately mimic thalamic stimulation and removed this citation from the Discussion. This reference had been included in the previous submission due to the general lack of other studies comparing cortical inhibition between different frequencies of stimulation. We note that the results described in our manuscript help to fill this void.

4) In response to your comment that, in contrast to Fig. 1 in Castro-Alamancos and Connors, 1996a, you did not observe spikes on every stimulus presentation at 10 Hz, the reviewer points out that this is based on VL stimulation and recording in the motor cortex, which evokes an augmenting response. If in that figure (or Fig. 3 of the same paper) the surface positive field potential component and associated corresponding excitatory drive (which produces the spikes) are eliminated, what is left is a recruiting response, as defined originally by Dempsey and Morison. Similar to spindles, recruiting responses trigger spikes only on some of the waves at 10 Hz. Thus, 3.6 Hz would be consistent with this. It would be helpful if you were to present PSTHs using each stimulus of the 10 Hz train as the marker (Figure 3D).

As requested by the reviewer, we included PSTHs of cortical firing during 10 Hz central thalamus stimulation in the revised manuscript (see Figure 2—figure supplement 4). These figures show that spike events have a non-uniform distribution over time and are more likely to occur after a 6-34 ms latency from the onset of individual light pulses. While such a pattern is expected to occur during disynaptic thalamocortical inhibition (and is therefore compatible with the reviewer’s hypothesis), it should be noted that it does not confirm or implicate the presence of thalamocortical inhibition mediated by cortical interneurons. Instead, this pattern only suggests that the evoked inhibition is not strong enough to completely eliminate the spikes evoked monosynaptically by excitatory thalamocortical afferents. In other words, while the PSTHs are consistent with cortical interneurons driving the observed inhibition (in the sense that they show a stimulus-driven excitatory response), the observed histogram distributions would likely also be present if inhibition was mediated by an external source such as zona incerta.

5) Your evidence that inhibition of the zona incerta during 10 Hz central thalamus stimulation reverses the cortical inhibition effect is based on AAV delivery of halorhodopsin, which will not discriminate between inhibitory and excitatory neurons. Consequently, light delivery might not only inhibit ZI cells, but other cells too, including adjacent thalamocortical cells. It is therefore possible that the effects observed are due to inhibition of thalamocortical cells. Similarly, could the mCherry fibers observed in cortex actually be thalamocortical fibers? It is also possible that your stimulation could be activating the NRt, which would drive the observed spindles and the recruiting responses in the cortex at 10 Hz, in which case there would be no need for the extrinsic inhibition purportedly provided by ZI. These possibilities should be addressed in the paper. Unless they can be ruled out, conclusions about the role of the zona incerta should be modified accordingly.

To address this concern, we added a quantitative characterization on the volume of eNpHR-expressing tissue activated by 593 nm light delivery (see Figure 5—figure supplement 1). These calculations indicate that the cone of eNpHR activation predominately falls within the zona incerta. Furthermore, virtually none of the cone includes the thalamus (which already has a low mCherry expression-level compared to ZI), suggesting that the effects observed are not due to inhibition of thalamocortical cells. Notably, the incomplete coverage of zona incerta by the 593 nm cone of activation may explain why inhibition is not fully reversed across all recorded cells in cortex (Figure 5H).

A similar argument can be made regarding direct excitation of the reticular nucleus suggested by the reviewer. As shown in Figure 1A, the cone of tissue directly excited by 473 nm light is far removed from the reticular nucleus. Thus, it is highly unlikely that such a phenomenon could be driving the observed spindles and inhibition during 10 Hz stimulation.

Regarding the mCherry fibers reported in cortex, it has been known for over two decades that zona incerta projects to sensory cortex (see Lin et al., Science, 1990). Thus, while we agree that the mCherry-expressing axons observed in cortex may not necessarily originate from zona incerta, we also maintain that such specificity is not necessary to support our conclusions on zona incerta-mediated inhibition. As originally stated in the manuscript, these axons are simply consistent with previously reported projections.

6) In discussing alternative hypotheses of the mechanism of cortical inhibition, including disynaptic thalamocortical inhibition (see Discussion, paragraph four) and disynaptic corticocortical inhibition (see Discussion, paragraph six), you should include relevant literature (e.g. rat in vivo thalamic stimulation), instead of marginally related work (slice studies that used white-matter or cortical stimulation, hippocampal studies, etc.). Care should also be taken to cite the literature accurately. For example, you argue that "central thalamus at different frequencies can lead to distinct behavioral responses" and cite Morison and Dempsey (1942), but Dempsey and Morison did not study behavior. Galarreta and Hestrin (1998) did not study thalamocortical responses (please correct this in the subsection “Central thalamus stimulation frequency controls cortical excitation/inhibition balance”).

We thank the reviewer for pointing out these errors. We replaced the Dempsey and Morison reference with two studies demonstrating distinct behavioral responses (e.g. arrest/absence seizures) during low-frequency stimulation of intralaminar nuclei in vivo. We also replaced the citation with one that studied thalamocortical responses (Castro-Alamancos and Connors, 1996b), and added studies utilizing in vivo thalamic stimulation in rat to the Discussion (Castro-Alamancos and Connors, 1996a.; Casto-Alamancos and Connors, 1996c). We appreciate how these changes have helped to improve our manuscript.

Author details

1. Jia Liu

   Department of Neurology and Neurological Sciences, Stanford University, Stanford, United States

   Contribution

   JL, Collected the fMRI data, Acquisition of data

   Contributed equally with

   Hyun Joo Lee, Andrew J Weitz and Zhongnan Fang

   Competing interests

   The authors declare that no competing interests exist.

2. Hyun Joo Lee

   Department of Neurology and Neurological Sciences, Stanford University, Stanford, United States

   Contribution

   HJL, Performed EEG, in vivo single-unit, and field potential electrophysiology recordings

   Contributed equally with

   Jia Liu, Andrew J Weitz and Zhongnan Fang

   Competing interests

   The authors declare that no competing interests exist.

3. Andrew J Weitz

   1. Department of Neurology and Neurological Sciences, Stanford University, Stanford, United States
   2. Department of Bioengineering, Stanford University, Stanford, United States

   Contribution

   AJW, Analyzed the fMRI and electrophysiology data and wrote the paper

   Contributed equally with

   Jia Liu, Hyun Joo Lee and Zhongnan Fang

   Competing interests

   The authors declare that no competing interests exist.

4. Zhongnan Fang

   1. Department of Neurology and Neurological Sciences, Stanford University, Stanford, United States
   2. Department of Electrical Engineering, Stanford University, Stanford, United States

   Contribution

   ZF, Analyzed the data and helped make figures and videos

   Contributed equally with

   Jia Liu, Hyun Joo Lee and Andrew J Weitz

   Competing interests

   The authors declare that no competing interests exist.

5. Peter Lin

   Department of Neurology and Neurological Sciences, Stanford University, Stanford, United States

   Contribution

   PL, Conducted behavioral experiments

   Competing interests

   The authors declare that no competing interests exist.

6. ManKin Choy

   Department of Neurology and Neurological Sciences, Stanford University, Stanford, United States

   Contribution

   MKC, Performed quantitative immunohistochemistry

   Competing interests

   The authors declare that no competing interests exist.

7. Robert Fisher

   Department of Neurology and Neurological Sciences, Stanford University, Stanford, United States

   Contribution

   RF, Scored the video-EEG data

   Competing interests

   The authors declare that no competing interests exist.

8. Vadim Pinskiy

   Cold Spring Harbor Laboratory, Cold Spring Harbor, United States

   Contribution

   VP, Conducted optical brain slice imaging

   Competing interests

   The authors declare that no competing interests exist.

9. Alexander Tolpygo

   Cold Spring Harbor Laboratory, Cold Spring Harbor, United States

   Contribution

   AT, Conducted optical brain slice imaging

   Competing interests

   The authors declare that no competing interests exist.

10. Partha Mitra

    Cold Spring Harbor Laboratory, Cold Spring Harbor, United States

    Contribution

    PM, Advised VP and AT, and helped design the experiments

    Competing interests

    The authors declare that no competing interests exist.

11. Nicholas Schiff

    Department of Neurology, Weill Cornell Medical College, New York, United States

    Contribution

    NS, Helped with the design of the experiments, interpretation of the data, and writing of the paper

    Competing interests

    The authors declare that no competing interests exist.

12. Jin Hyung Lee

    1. Department of Neurology and Neurological Sciences, Stanford University, Stanford, United States
    2. Department of Bioengineering, Stanford University, Stanford, United States
    3. Department of Electrical Engineering, Stanford University, Stanford, United States
    4. Department of Neurosurgery, Stanford University, Stanford, United States

    Contribution

    JHL, Designed and planned all experiments, supervised and organized the study, analyzed and interpreted data, wrote the paper, and outlined the figures and videos

    For correspondence

    ljinhy@stanford.edu

    Competing interests

    The authors declare that no competing interests exist.

• 8,105

  Page views

• 1,584

  Downloads

• 83

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.