Étude des neurosciences ( texte en anglais)

Frontiers in Human Neuroscience www.frontiersin.org November 2010 | Volume 4 | Article 198 | 1

HUMAN NEUROSCIENCE

Original Research Article

published: 22 November 2010

doi: 10.3389/fnhum.2010.00198rons in gamma generation is more ambiguous. While the presence of sufficient glutamatergic activity to recruit IN activity is essential in most models and in vitro preparations, the decay rate of synaptic excitation is typically thought to be less important. In contrast, the origin of low-frequency cortical activity (e.g., 8 Hz – “alpha” band activity) has been tied to intrinsic biophysical properties of cortical pyramidal neurons (Silva et al., 1991; Castro-Alamancos and Connors, 1996; Jones et al., 2000), and to reciprocal thalamocortical loops (Contreras and Steriade, 1995; Hughes and Crunelli, 2005; Jones et al., 2009).

Optogenetic tools provide a novel avenue for investigating the cellular mechanisms controlling neocortical rhythmicity in vivo, by permitting selective drive of distinct neural populations. Using the light-sensitive cation channel Channelrhodopsin-2 (ChR2; Deisseroth et al., 2006) to drive either FS or RS in neocortex, we recently found that selective FS stimulation from 8 to 200 Hz led to increased local field potential (LFP) power above baseline at the driving frequency, specifically in the gamma band (most pronounced at 32–64 Hz) (Cardin et al., 2009). In contrast, selective RS activation showed a peak in enhanced power at 8 Hz, with significantly weaker higher frequency recruitment (Figure 1A).

Introduction

Neocortical oscillations are believed by many to be essential to perception and memory, and several neurologic and psychiatric diseases are characterized by alterations in these patterns (Herrmann and Demiralp, 2005; Jensen et al., 2007; Tallon-Baudry, 2009). However, any given neocortical rhythm may alternatively be an epiphenomenon of neural information processing. Developing a mechanistic, cellular-level understanding of the origin of these rhythms is likely key to identifying what, if any, computational role they have, and to understanding the changes underlying their altered expression in disease.

Until recently it was not possible to selectively record and drive distinct cell populations in vivo, making it difficult to decisively identify the precise contributions of specific excitatory and inhibitory mechanisms to rhythm generation in fully embodied networks. Computational modeling and in vitro experimental work have provided a leading source of insight. Extensive work has supported the critical role of fast-spiking (FS) interneurons (INs) and the time constant of GABAA synaptic inhibition in controlling gamma rhythmicity (e.g., Whittington et al., 1995, 2000; Borgers and Kopell, 2005; Bartos et al., 2007; Mann and Paulsen, 2007). The role of oscillatory activity in pyramidal, regular-spiking (RS) excitatory neuComputational

modeling of distinct neocortical oscillations driven by cell-type selective optogenetic drive: separable resonant circuits controlled by low-threshold spiking and fast-spiking interneurons

Dorea Vierling-Claassen1,2, Jessica A. Cardin3, Christopher I. Moore1 and Stephanie R. Jones2*

1 McGovern Institute of Brain Research, Massachusetts Institute of Technology, Cambridge, MA, USA

2 Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, USA

3 Department of Neurobiology, Yale University School of Medicine, New Haven, CT, USA

Selective optogenetic drive of fast-spiking (FS) interneurons (INs) leads to enhanced local field potential (LFP) power across the traditional “gamma” frequency band (20–80 Hz; Cardin et al., 2009). In contrast, drive to regular-spiking (RS) pyramidal cells enhances power at lower frequencies, with a peak at 8 Hz. The first result is consistent with previous computational studies emphasizing the role of FS and the time constant of GABAA synaptic inhibition in gamma rhythmicity. However, the same theoretical models do not typically predict low-frequency LFP enhancement with RS drive. To develop hypotheses as to how the same network can support these contrasting behaviors, we constructed a biophysically principled network model of primary somatosensory neocortex containing FS, RS, and low-threshold spiking (LTS) INs. Cells were modeled with detailed cell anatomy and physiology, multiple dendritic compartments, and included active somatic and dendritic ionic currents. Consistent with prior studies, the model demonstrated gamma resonance during FS drive, dependent on the time constant of GABAA inhibition induced by synchronous FS activity. Lower-frequency enhancement during RS drive was replicated only on inclusion of an inhibitory LTS population, whose activation was critically dependent on RS synchrony and evoked longer-lasting inhibition. Our results predict that differential recruitment of FS and LTS inhibitory populations is essential to the observed cortical dynamics and may provide a means for amplifying the natural expression of distinct oscillations in normal cortical processing.

Keywords: gamma, GABA, channelrhodopsin, low-threshold spiking, fast-spiking, interneurons, somatosensory cortex

Edited by:

Thilo Womelsdorf, Robarts Research Institute London, Canada

Reviewed by:

Gustavo Deco, Universitat Pompeu Fabra, Spain

Nancy J. Kopell, Boston University, USA

*Correspondence:

Stephanie R. Jones, Athinoula A. Martinos Center for Biomedical Imaging, 149 Thirteenth Street, Suite 2301, Charlestown, MA 02129, USA. e-mail: srjones@nmr.mgh.harvard.edu

Frontiers in Human Neuroscience www.frontiersin.org November 2010 | Volume 4 | Article 198 | 2

Vierling-Claassen et al. Engaging distinct oscillatory neocortical circuits

Figure 1 | Experimental findings of contrasting resonance properties with RS and FS drive. (A) Mean power ratio in each frequency band in response to light activation of FS (filled circles; n = 14 sites in six PV-Cre mice) and RS (open circles; n = 13 sites in five CamKII-Cre mice) cells at those frequencies. (B) Average spike probability per light pulse cycle in light-activated FS and RS cells in the PV-Cre and CamKII-Cre mice, respectively (figures adapted from Cardin et al., 2009).

While these FS effects agree

...