Transcript Normal and pathological oscillatory communication in the brain

Normal and pathological
oscillatory communication in the brain
Jaeseung Jeong, Ph.D
Department of Bio and Brain Engineering
KAIST
Complex spatiotemporal dynamics in the Brain
Brain oscillations
• The brain is constantly active, even during deep sleep. In the cerebral
cortex, this spontaneous activity (activity that is not obviously driven
by a sensory input or a motor command) often occurs as periodic,
rhythmic discharges.
Main issues on Oscillations
• Recent improvements in the methods of studying longrange communication have allowed us to address several
important questions:
• What are the common mechanisms that govern local and
long-range communication and how do they relate to the
structure of the brain?
• How does oscillatory synchronization subserve neural
communication?
• What are the consequences of abnormal synchronization?
threshold -> Spike
j
i
Spike reception: EPSP,
summation of EPSPs
ui

Threshold Spike emission
(Action potential)
Spike reception: EPSP
Brain is an information processor
LFP oscillations might be due to regularities in the firing of
only a subset of neurons in a local area, as LFPs represent a
spatial average.
• As no behaviourally relevant task is performed
independently by a single neuron, communication is of the
utmost importance and, ultimately, optimal computational
performance relies on optimal communication.
• Here we use a broad definition of neural communication,
in which a neural element (a single neuron or a population
of neurons) conveys certain aspects of its functional state
to another neural element.
• Neural communication depends on the anatomical
components that connect individual neurons (structure)
and the process of transmitting information (function).
Both aspects affect the overall performance of the system.
EEG recordings
• Structurally, the most striking neuroanatomical feature of
the brain is the abundant connectivity between neurons,
which reflects the importance of neural communication.
• Functionally, oscillations are a prominent feature of
neuronal activity and the synchronization of oscillations —
which reflects the temporally precise interaction of neural
activities — is a likely mechanism for neural
communication.
Neuronal Oscillations
• infra-slow: 0.02-0.1 Hz,
• slow: 0.1-15 Hz (during slow-wave sleep or anesthesia)
– Slow oscillation (0.2-1 Hz),
– Delta (1-4 Hz),
– Spindle (7-15Hz),
– Theta (generated in the limbic system)
• fast: 20-60 Hz,
• ultra-fast: 100-600 Hz.
Fourier transformation of the EEG
Thalamocortical oscillations
• Oscillatory activity is an emerging property of the brain network,
especially the thalamocortical system.
The various oscillatory rhythms generated in the thalamocortical
system are mediated by two types of mechanisms:
• Intrinsic mechanisms, which depend on the interplay between
specific intrinsic currents.
• Extrinsic or network mechanisms, which require the interaction of
excitatory and inhibitory neurons within a population.
• Intrinsic and network mechanisms can work alone (e.g., thalamic
delta oscillations depend on the intrinsic properties of thalamic relay
cells, cortical slow oscillation depends on network properties) or in
combination (e.g., spindles depend on the interaction between
thalamic relay and reticular neurons as well as on their intrinsic
properties).
Infra-slow oscillation
• This type of oscillatory activity has a period within the range of tens
of seconds to a minute.
• Very little is known about the underlying mechanisms of these
oscillations but at least some of the factors responsible for their
generation could depend on non-neuronal dynamics. Infra-slow
activities likely have a cortical origin given that they can be
recorded from small regions of neocortex devoid of their inputs by
means of a surgical undercut.
• What is the functional role?
Indirect evidence suggests that infra-slow oscillations synchronize
faster activities, modulate cortical excitability, and contribute to the
aggravation of epileptic activity during sleep.
Slow Oscillations
• Another form of rhythmic cortical activity that recently has been
investigated extensively in vivo is the so-called "SLOW
OSCILLATION."
• The slow oscillation is characterized by periods of sustained
depolarization interweaved with periods of hyperpolarization and
silence at a rate of between once every 10 seconds to
approximately once per every 2 seconds.
• The depolarized state is associated with low frequency neuronal
firing and is termed the UP state, while the hyperpolarized state is
referred to as the DOWN state.
• The frequent transitions between the UP and DOWN state can
make the membrane potential of the cortical neuron appear as a
single channel recording, even though the slow oscillation is
generated by the interaction of thousands of cells!
Slow oscillations
• During slow-wave sleep and some types of anesthesia the dominant
activity pattern is slow oscillation, with frequency 0.3 - 1 Hz.
• The following observations point to an intracortical origin for this
rhythm:
a. It survives extensive thalamic lesions in vivo and exists in
cortical in vitro preparations.
• It is absent in the thalamus of decorticated cats.
• The slow oscillation is generated in the Cerebral Cortex, since
removal of the thalamus does not block the slow oscillation and
isolation of slabs of cortex retain this activity.
• McCormick et al. demonstrated the slow oscillation in cortical
slices maintained in vitro.
Beta oscillations
• Beta oscillation is the oscillation with frequency range above 12
Hz.
• Low amplitude beta with multiple and varying frequencies is often
associated with active, busy or anxious thinking and active
concentration.
• Rhythmic beta with a dominant set of frequencies is associated
with various pathologies and drug effects, especially
benzodiazepines.
Fast oscillations
• Fast oscillations could be divided on fast (beta and gamma) and
ultra-fast (>100 Hz, ripples).
• Fast oscillations have been implicated in cognitive processes. They
also form a prominent part of sleep EEG signals, when intracortical
electrodes are used. They also occur in association with seizures.
• Ultrafast oscillations could be found during sharp waves and sleep.
The ripples are prominent in the onset of seizures.
Two major categories of fast oscillations
• (Homogeneous) Those which occur in a homogeneous collection of
neurons, all of the same type, for example CA1 pyramidal neurons.
• (Heterogeneous) Those which require synaptic interactions between
two or more populations of neurons, for example CA1 pyramidal
neurons together with fast-spiking CA1 interneurons.
Gamma oscillations
• Gamma oscillation is the rhythmic activity with the frequency
range approximately 26–100 Hz.
• Gamma rhythms may be involved in higher mental activity,
including perception, problem solving, fear, and consciousness.
Persistent gamma
in hippocampus and entorhinal cortex
• This type of gamma is notable for the sparse firing of pyramidal cell
somata, despite the prominent presence of IPSPs in pyramidal cells,
and the dependence of the oscillation on synaptic inhibition.
• Additionally, phasic EPSPs and gap junctions are also required; the
experimental evidence suggests that axonal electrical coupling
between pyramidal cells is an absolute requirement, while
interneuronal electrical coupling plays a modulatory role.
• Persistent gamma is usually induced by bath application of an
appropriate drug, such as kainate or carbachol, and can last for hours.
Persistent gamma in Hippocampus
• Power spectra of the field oscillation reveal
both a gamma peak, and also a faster, but
non-harmonic, peak at >70 Hz.
• Remarkably, the pyramidal cell axonal
plexus, in the hippocampal CA1 region,
when it is surgically separated from
pyramidal cell somata, generates a
continuous very fast oscillation, rather than
gamma.
• It is thought that the mechanism for this
continuous very fast oscillation is similar
to the mechanism for the first type of
"homogeneous" fast oscillation described
above.
• It is apparent that very fast oscillations
(>70 Hz) and persistent gamma oscillations
are intimately related.
Functional role of fast oscillations for
perception and behavior
• Cognitive functions like perception, attention, memory or language
are based on highly parallel and distributed information processing
by the brain.
• One of the major unresolved questions is how information can be
integrated and how coherent representational states can be
established in the distributed neuronal systems subserving these
functions. It has been suggested that this so-called 'binding problem'
may be solved in the temporal domain.
• The hypothesis is that synchronization of neuronal discharges can
serve for the integration of distributed neurons into cell assemblies
and that this process may underlie the selection of perceptually and
behaviourally relevant information.
Functional role of fast oscillations for
perception and behavior
• Moreover, it has been suggested that fast oscillations at
frequencies in the so-called gamma range (> 30 Hz) may help to
entrain spatially separate neurons into synchrony and thus may
indirectly promote the dynamic binding of neuronal populations.
• In accordance with these predictions, states characterized by
synchronized gamma activity have been shown to be associated
with functions like processing of coherent stimuli, perceptual
discrimination, focused attention, short-term memory, or
sensorimotor integration.
• Typically, the observed magnitude of gamma activity is positively
correlated with increased 'processing load' and thus with the level
of vigilance and attention, as well as with the difficulty or
integrative nature of the processing.
The antennal lobe receives input from olfactory receptor neurons; it then
transforms and reformats this input for transfer through projection neurons
to the mushroom body, which is responsible for memory encoding and
retrieval, and to the lateral horn. In turn, inhibitory interneurons in the lateral
horn project to the mushroom body.
Odours elicit global oscillatory activity of 20–30 Hz in the antennal lobe
network (which is composed of local and projection neurons), and this, in
turn, is reflected in the local field potentials (LFPs) of the antennal lobe,
mushroom body and lateral horn. The action potentials of projection
neurons are PHASE-LOCKED to the LFP oscillations in a neuron-, odourand time-specific manner, such that the antennal lobe output is an
evolving 20–30 Hz sequence of synchronized projection neuron spikes.
Inhibitory interneurons provide an important mechanism for
synchronization of the LFP oscillations of remote neural
populations.
According to the network model of Bibbig and co-workers,
synchronization can be mediated by a pair of action
potentials per oscillation cycle (long-interval doublets):
The first spike of the doublet originates from excitatory
input from local principal cells; the second is caused by
excitation from distant principal cells. The timing between
the doublets in a given cycle provides the feedback
required to synchronize the oscillations of distant networks.
Functional consequences of oscillatory driving input to the motoneurons
that relate to breathing have also been shown in rats in vitro.
First, similar to the effect of correlated presynaptic inputs on other
neurons, the timing of action potentials in motor neurons is crucially
affected by oscillatory modulations of input. Motor neuron spike trains are
much less variable and more consistent during oscillatory input.
Second, increased synchronization in the oscillatory input increases the
gain — that is, an increase in the number of action potentials that are
elicited by a given input58 augments the force output in a computer
simulation of a motor neuron pool.
Third, synchronization increases the robustness of the input–output
relationship for motor neurons against changes in neurotransmitter level.
Finally, recent studies indicate that oscillatory communication subserves
gating of information processing and modulates the effects of spike trains,
and so shows a strong dependence on the behavioural state11,
• Three lines of evidence provide experimental support for
extending the concept of gamma-band synchronization as
a mechanism for binding:
• First, there is new experimental evidence that beta
oscillations are important in long-range synchronization,
which is further supported by computational modelling.
• Second, the importance of desynchronization, or, more
specifically, the spatiotemporal balance of synchronization
and desynchronization, is increasingly being recognized.
• Third, there is new evidence that specific synchronization
patterns are directly related to behaviour.
The synchronization index (SI), which quantifies phase synchronization, shown for
five successive stimuli. The x-axis specifies time after presentation of the first target.
Each point represents the mean SI in a 60-ms-window centred at 260 ms after the
respective stimulus. Values at 260 ms quantify the network synchronization to the
first target (T1), whereas values at 114 ms represent the network synchronization
that corresponds to the distractor preceding the first target. The blue line represents
the response seen when both targets were correctly reported and the red line the
response seen when the second target was missed.
The spatial distribution of coherence with the left primary motor cortex (M1) as a
reference region. Only areas with p<0.05 (corrected, one-sample t-test) are
shown. Note that the left thalamus and right cerebellum are projected to the left
surface for easier visualization. The dominant coupling direction (mean
directionality index) is indicated by arrows.
Line thickness indicates the degree of coherence (left) and Granger
causality (right) found between recording sites (labelled 1–6) when a
monkey performs a contraction task. Granger causality quantifies the
directionality of information flow. Only significant interactions are shown.
As, arcuate sulcus; Cs, central sulcus; IPs, intraparietal sulcus; Ls, lateral
sulcus; STs, superior temporal sulcus.
The striatum receives most of the input from the cerebral cortex; in this sense, it is
the doorway to the basal ganglia. The GPi and SNr are the output nuclei of the
basal ganglia and send the main inhibitory output from the basal ganglia back to
the thalamus. The striatum sends its output to the GPi/SNr through a direct
dopamine D1-receptor-mediated pathway and through an indirect dopamine D2receptor-mediated pathway involving the GPe and STN. The direct pathway is
thought to facilitate movement, whereas the indirect pathway is thought to
suppress movement. There is a delicate balance between these two pathways that
is partly maintained by dopamine release from the SNc to the striatum. Dopamine
release inhibits the indirect pathway by stimulating dopamine D2 receptors and
excites the direct pathway by stimulating the dopamine D1 receptors.
Maps of spatially normalized cerebro-muscular and cerebro-cerebral
coherence averaged across 4 patients with right-sided rest tremor. Cerebromuscular coherence at double tremor frequency occurs in the contralateral
primary motor cortex (b). Cerebro-cerebral coherence was computed with
the reference region in the primary motor cortex and averaged for all
patients. Areas of consistent coherence are the lateral (a) and medial (c)
premotor areas, secondary somatosensory cortex (a), posterior parietal
cortex (a) as well as the thalamus/basal ganglia (d) contralateral to the
tremor hand and the cerebellum (e) ipsilateral to the tremor hand. Note that
because of the large distance to the magnetoencephalogram (MEG) sensors,
localization in both subcortical areas is not as precise as at the cortical level.