Neuroscience Letters 402 (2006) 216–221
Isoflurane disrupts anterio-posterior phase synchronization
of flash-induced field potentials in the rat
Olga A. Imasa, Kristina M. Ropellab, James D. Wooda, Anthony G. Hudetza,∗
aDepartment of Anesthesiology, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI, United States
bDepartment of Biomedical Engineering, Marquette University, Milwaukee, WI, United States
Received 25 January 2006; received in revised form 16 March 2006; accepted 6 April 2006
Consciousness presumes a set of integrated functions such as sensory processing, attention, and interpretation, and may depend upon both
local and long-range phase synchronization of neuronal activity in cerebral cortex. Here we investigated whether volatile anesthetic isoflurane
at concentrations that produce loss of consciousness (LOC) disrupts long-range anterio-posterior and local anterior synchronization of neuronal
activity in the rat. In six rats, deep electrodes were chronically implanted in the primary visual cortex (V1) and in two areas of the motor cortex
(M1 and M2) for recording of intracortical event-related potentials (ERP). Thirty discrete flashes were presented at random interstimulus intervals
of 15–45s, and ERPs were recorded at stepwise increasing isoflurane concentrations of 0–1.1%. Neuronal synchronization was estimated using
wavelet coherence computed from the ERP data band-pass filtered at 5–50Hz. We found that (1) in the waking state, long-range anterio-posterior
coherence in 5–25Hz and 25–50Hz frequency bands was significantly higher than local anterior coherence; (2) anterio-posterior coherence in
both 5–25Hz and 26–50Hz bands was significantly reduced by isoflurane in a concentration-dependent manner; (3) local anterior coherence was
not affected by isoflurane at any of the concentrations studied. These findings suggest that a disruption of long-range anterio-posterior rather than
local anterior synchronization of neuronal activity precedes the anesthetic-induced loss of consciousness.
© 2006 Elsevier Ireland Ltd. All rights reserved.
Keywords: Synchronization; Consciousness; Anesthesia; Event-related potential; EEG; Gamma
Identification of neuronal targets and mechanisms of hypnotic
thesia research. While significant advances have been made in
understanding of the anesthetic pharmacology and anesthetic-
receptor interaction, a mechanistic model of anesthetic ablation
what phenomenal consciousness is, how it may arise from neu-
rophysiologic events, and how to objectively assess its presence
or absence, has hampered the elucidation of the mechanisms of
Neurophysiologic evidence suggests that consciousness is
an emergent phenomenon of complex functional interactions
of various systems of the brain, including primary sensory and
selective attention [19,27] and working memory . Long-
∗Corresponding author. Tel.: +1 414 456 5622; fax: +1 414 456 6507.
E-mail address: email@example.com (A.G. Hudetz).
has been hypothesized as a potential mechanism of the inte-
gration or “binding” of sensory information necessary for the
emergence of conscious sensory awareness and its respective
behavioral manifestations [4,6,8,25]. A support for this hypoth-
esis comes from various studies in humans and animals that
have shown that synchronization of EEG, local field potentials,
and unit activity among distributed cortical regions, especially
between anterior and posterior cortices [24,29], mediates per-
ception [6,22], attention [2,32], and memory .
neuronal activity among anterior and posterior cortical regions
. A support for this hypothesis comes from a study by John
et al. , who showed in a large patient population that rest-
after the loss of consciousness (LOC) achieved with general
anesthetics. Since the experimental protocol in this study did
not involve an incremental increase in the anesthetic concentra-
tion, this study could not easily isolate the anesthetic-induced
changes in neuronal synchronization precisely at the point of
0304-3940/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved.
O.A. Imas et al. / Neuroscience Letters 402 (2006) 216–221
LOC. Furthermore, the effect of anesthetics on cortical coher-
or animals has not been systematically investigated. Recording
of local field potentials would offer an advantage over scalp
EEG recording because it allows for better spatial localization
of electrophysiologic activity.
In this study, we examined the concentration-dependent
posterior flash-induced phase synchronization of local field
synchronization in the rat’s anterior cortex. We applied wavelet
coherence analysis to assess changes in neuronal synchroniza-
tion as a function of time and frequency. Coherence is a linear
and thus, can be used to assess phase synchronization of neu-
ronal activity. As we were interested in the neural correlate of
waking state and at sedative-hypnotic concentrations of isoflu-
The experimental procedures and protocols were reviewed
and approved by the Institutional Animal Care and Use Com-
mittee (Medical College of Wisconsin, Milwaukee, WI). All
procedures conformed to the Guiding Principles in the Care
MD) and were in accordance with the Guide for the Care and
of animals used and their possible discomfort.
In six adult male Sptrague-Dawley rats, an array of 16
polyimide-insulated tungsten microwire electrodes (Tucker-
land Hills, CA) were stereotaxically implanted for recording of
intracortical field potentials. Using implantation methods previ-
ously described [10,12], the electrode array was positioned in
cortices. The coaxial electrode was implanted in the primary
visual cortex (V1) (7mm posterior, 2–3mm lateral, 2–2.2mm
ible with a 16-channel head stage (Neuralynx, Model HS16,
Tucson, AZ). The coaxial semi-micro electrode was connected
through implanted gold pins, and signals were led to the ampli-
fier through shielded cables. The coaxial electrodes were used
in our previous studies [10–13], and proved effective in record-
ing robust event-related potentials (ERP). We chose to use the
microwire electrode array in the anterior cortical regions, M1
and M2, to examine local neuronal synchronization with multi-
channel recordings. The active electrode tips for visual and
motor cortices were at the same vertical depth. A stainless steel
machine screw in the opposite hemisphere from the array was
for 7–10 days.
On the day of the experiment, the rat was placed in a cylin-
drical plastic restrainer inside a plexiglass anesthesia box under
1.5% isoflurane anesthesia, allowing limited movement of its
head and limbs. The animal was breathing spontaneously, and
its body temperature was controlled at 37◦C. Once all connec-
tions were in place, the anesthetic was turned off, and 1h of
equilibration period was allowed for the animal to regain con-
sciousness and to accommodate to its physical environment.
Following 1h of equilibration, ERPs elicited by binocular
flash stimulation were recorded. Thirty discrete flashes were
proof box. Subsequently, isoflurane concentration was raised
from 0 to 1.1% in increments of 0.1–0.2%, and was allowed to
reach steady state for 20min at each anesthetic level. The anes-
thetic concentration was monitored using a gas analyzer (POET
at each increased concentration.
The signals from all 17 electrodes were amplified simultane-
ously at a gain of 10,000, analog bandpass-filtered at 1–100Hz,
and digitally sampled at 500Hz (WINDAQ Data Acquisition
Software, DATAQ Instruments, Akron, OH).
The qualitative assessment of the microwire recordings from
the anterior cortex revealed that all recordings contained well-
identifiable ERPs that were similar among all 16 channels. For
most spatially separated electrodes (anterior and posterior ends
of the array corresponding to M1 and M2) were selected for
Single-trial 1-s-long ERPs were extracted from V1, M1,
and M2 field-potential records. To remove the low-frequency
afterdischarge component of the ERP characteristic of the rat’s
flash-induced response, the single-trial ERPs were band-pass
filtered to 5–50Hz using a bi-directional Butterworth digital fil-
ter (N=2). The flash-induced afterdischarge activity likely does
not play a role in the mechanism of anesthetic-induced LOC,
as its suppression has been previously demonstrated at very
low anesthetic concentrations with preserved consciousness
The synchronization of neuronal activity between M1 and
V1, M2 and V1, and within the motor cortex (between M1
ence allows an estimation of coherence as a function of time
and frequency, and has an advantage over FFT-based methods,
as it offers sufficient balance between temporal and spectral
resolutions for a subset of short-time, relatively narrow-band,
and fast-changing signals, such as ERP . The mathematical
equation describing wavelet coherence˜Γ2
signals x(n) and y(n) for given frequency scale and translation
factors, s and ?n, is shown below.
xy(s,?n) between two
In Eq. (1),˜SW
auto- and cross-wavelet spectra of x(n) and y(n). Wavelet coher-
ence was computed from single-trial ERPs using Morlet’s
xy(s,?n) are the
O.A. Imas et al. / Neuroscience Letters 402 (2006) 216–221
wavelets with central frequency of 5–50Hz in increments of
1Hz , and then averaged across multiple trials.
As an estimate of band coherence at intermediate (5–25Hz)
and high (26–50Hz) frequencies, the coherence data from each
animal were averaged within two spectral windows of 5–25Hz
and 25–50Hz. Subsequently, to examine changes in the flash-
ulus value by the temporal mean of prestimulus band coherence
of the last 200ms before flash. Finally, the peak BCR was found
in the 0–200ms period after stimulus, since the strongest coher-
ence was localized to that time interval in all experiments.
To test for a significant effect of anesthetic concentration on
BCR, a Repeated Measures ANOVA with planned comparisons
was carried out with the peak BCR as a response variable, the
variable. To test for a significant difference in long-range and
local peak BCR estimates, a paired t-test was performed. In all
tests, p≤0.05 was accepted for statistical significance.
Fig. 1 shows an example from one experiment of the average
ERP from V1 and M1 in the waking state and at various isoflu-
rane concentrations. In V1, isoflurane augmented the amplitude
of the early (0–100ms) ERP component and, in this example,
also suppressed the amplitude of the late (200–300ms) compo-
nent already at 0.5%. In M1, isoflurane reduced the amplitude
of the entire ERP in a concentration-dependent manner. In all
other frontal locations, the effect of isoflurane on the ERP was
similar to that in M1.
The effects of isoflurane on M1-V1 and M2-V1 coherence
were similar in all experiments. For that reason, the correspond-
ing BCR and peak BCR data were pooled and averaged across
all experiments. These averages will be referred to as M1/M2-
V1 BCR and M1/M2-V1 peak BCR to represent long-range
anterio-posterior coherence. Fig. 2 shows group-average effects
Fig. 1. Event-related potentials in primary visual (V1) and primary motor (M1)
cortices in the waking state and at increased isoflurane concentrations from the
activity averaged across 30 trials. In V1, the anesthetic agent augmented the
amplitude of the early (0–100ms) ERP component and, in this example, also
suppressed the amplitude of the late (200–300ms) component already at 0.5%.
In M1, isoflurane reduced the amplitude of the entire ERP in a concentration-
dependent manner. All consecutive analyses were performed on the single-trial
of isoflurane on M1–M2 and M1/M2-V1 BCR in intermedi-
ate (5–25Hz) and high (26–50Hz) frequency bands. Note that
the standard error associated with BCR was due to animal-to-
animal variability in both the time of occurrence and amplitude
of the BCR peak. An insert in each of the four panels of Fig. 2
shows the standard errors associated with the time of occur-
rence of the BCR peak. The standard errors associated with the
peak BCR amplitude are included in Fig. 3. Fig. 2 shows that
Fig. 2. Concentration-dependent effect of isoflurane on band coherence ratio (BCR) between primary visual and motor cortices (M1/M2-V1), and within motor
cortex (M1–M2) in intermediate (5–25Hz) and high (26–50Hz) frequency bands. The standard error associated with BCR was due to animal-to-animal variability in
both the time of occurrence and amplitude of the BCR peak. An insert in each of the four panels shows the standard errors associated with the time of occurrence of
the BCR peak as a function of isoflurane concentration. Isoflurane had no effect on M1–M2 BCR either in the intermediate or in the high frequency band. However,
isoflurane produced a concentration-dependent reduction in M1/M2-V1 BCR in both bands. Note that in the waking state and at sedative concentrations (0.3–0.5%)
M1/M2-V1 BCR was localized to the 0–200ms time interval and was considerably higher than M1–M2 BCR in both frequency bands.
O.A. Imas et al. / Neuroscience Letters 402 (2006) 216–221
Fig. 3. Concentration-dependent effect of isoflurane on peak band coherence ratio (peak BCR) between primary visual and motor cortices (M1/M2-V1), and within
motor cortex (M1–M2) in intermediate (5–25Hz) and high (26–50Hz) frequency bands. The peak BCR represents the maximum band coherence ratio between 0 and
200ms after flash. Bars represent averages from all animals, and the corresponding standard error bars represent indices of variability in the peak BCR amplitude.
Significant differences from waking control and between waking controls are indicated by * and †, respectively at (p≤0.05) level. Isoflurane had no effect on
M1–M2 peak BCR either in the intermediate or in the high frequency band. However, isoflurane significantly reduced M1/M2-V1 peak BCR in both bands in a
concentration-dependent manner. Note also that in the waking state, M1/M2-V1 peak BCR was significantly higher than M1–M2 peak BCR in both bands.
bands. Note that in the waking state and at sedative concentra-
time interval and was considerably higher than M1–M2 BCR in
both frequency bands.
Fig. 3 shows group-average effects of isoflurane on M1/M2-
V1 and M1–M2 peak BCR in intermediate and high frequency
bands. Consistent with Fig. 2, isoflurane had no effect on
M1–M2 peak BCR either in the intermediate or in the high fre-
quency band, whereas it produced a concentration-dependent
reduction in M1/M2-V1 peak BCR in both bands. Note also
that in the waking state, M1/M2-V1 peak BCR was signifi-
the M1/M2-V1 peak BCR was 3.8 and 3.2 times greater than
M1–M2 peak BCR in intermediate and high frequency bands,
respectively. Some difference in the concentration dependence
of BCR (Fig. 2) and peak BCR (Fig. 3) is explained by the fact
that the precise temporal location of the peak BCR between 0
and 200ms varied from animal to animal.
In summary, we found that long-range anterio-posterior
coherence in both, intermediate (5–25Hz) and high (26–50Hz)
frequency bands was significantly stronger than local ante-
rior coherence in the waking state. The volatile anesthetic
isoflurane significantly reduced anterio-posterior coherence in
a concentration-dependent manner, but had no effect on local
The role of long-range cortical interactions in conscious-
ness has been suggested by a number of studies in humans and
animals [21,26,31]. Recently, it has been proposed that visual
consciousness itself may depend upon a sparse network of neu-
rons involved in recurrent interactions between anterior and
posterior cortical regions [5,18]. Our goal was to find the neural
correlate of anesthetic-induced unconsciousness presumably as
an abrupt change in neural activity pattern. Instead of an abrupt
ual, concentration-dependent reduction in long-range anterio-
ning from wakefulness to LOC. This graded suppression of
coherence is consistent with the known effect of general anes-
thetics on various indices of EEG and evoked potential activity
that are routinely used for anesthetic depth monitoring to assess
our current findings one cannot ascertain if a critical value of
coherence would predict LOC. Future studies on coherence
among cortical and subcortical regions previously implicated
in conscious perception may help in determining the neural cor-
relate of LOC.
Also of interest is our finding that long-range coherence
greatly exceeded local coherence in the waking and sedated
states. Furthermore, local coherence was unaffected by isoflu-
rane at any of the concentrations studied. Taken together, these
findings suggest that local interactions do not play a significant
role in the mechanism of anesthetic-induced LOC.
The anesthetic-induced suppression of anterio-posterior
coherence has been previously found by John et al.  in all
in contrast to coherence in lower frequency bands, coherence in
gamma (20–60Hz) band only was restored to the pre-LOC lev-
time required for emergence from than for induction of anesthe-
thus could not ascertain if anterio-posterior cortical interactions
O.A. Imas et al. / Neuroscience Letters 402 (2006) 216–221
at any frequency are restored upon return of consciousness in
the rat. The study of long-range coherence during emergence as
thus represents another potential extension of this project.
The results of this work could be compared to our previously
published study  that used transfer entropy as an alterna-
tive measure of multiregional functional interactions. While the
goal of both studies was to assess anesthetic-induced changes
in cortico-cortical functional interactions, two distinct method-
ologies, transfer entropy and wavelet coherence, allowed us to
evaluate different aspects of these interactions. Namely, trans-
fer entropy is a probabilistic measure of directional information
transfer between two regions that is based on the ability to
edge of another signal, but this method provides no information
about temporal phase relation between two signals. Conversely,
wavelet coherence is a measure of constancy of phase relation-
ship between two signals over time, but it does not provide any
ilar results provides support for the contention that the mecha-
cortical functional interactions.
How anesthetic agents may suppress long-range synchro-
nization of neuronal activity is unclear. Volatile anesthetics
including isoflurane are known to potentiate neurotransmission
at GABAAreceptors involved in maintaining local synchrony
within a monosynaptic neuronal network [3,9]. However, this
mechanism may not apply to long-range cortical interactions,
as the latter may largely depend on polysynaptic circuits. Long-
range pathways may be more susceptible to anesthetic suppres-
sion because multiple potential sites are available for anesthetic
action, and the superposition of anesthetic effects at these sites
may occur. The results of neuronal network modeling also sug-
gest that the anesthetic-induced suppression of a single unit
along a long-range polysynaptic pathway may lead to a con-
duction failure of the entire pathway , which is consistent
with our result. Another interesting potential mechanism under-
lying the anesthetic-induced suppression of long-range cortical
interactions may be a disturbance of the coincidence detection
midal cells .
In conclusion, our findings suggest that isoflurane at increas-
ing concentrations that lead to LOC preferentially impairs
anterio-posterior rather than local anterior cortical phase syn-
chronization of local field potentials, consistent with the pro-
posed role of anterio-posterior interactions in conscious percep-
for the design and construction of electronic equipment, and
Samhita S. Rhodes, Ph.D. for the implementation of the peak-
 E.D. Bigler, Neurophysiology, neuropharmacology and behavioral rela-
tionships of visual system evoked after-discharges: a review, Biobehav.
Rev. 1 (1977) 95–111.
 J.J. Bouyer, M.F. Montaron, A. Rougeul, P. Buser, [Parietal electrocor-
tical rhythms in the cat: their relation to a behavior of focused attention
and possible mesencephalic control through a dopaminergic pathway],
C. R. Seances Acad. Sci. D 291 (1980) 779–783.
 G. Buzsaki, Hippocampal GABAergic interneurons: a physiological per-
spective, Neurochem. Res. 26 (2001) 899–905.
 F. Crick, C. Koch, Towards a neurobiological theory of consciousness,
Sem. Neurosci. 2 (1990) 263–275.
 F. Crick, C. Koch, A framework for consciousness, Nat. Neurosci. 6
 S.M. Doesburg, K. Kitajo, L.M. Ward, Increased gamma-band synchrony
precedes switching of conscious perceptual objects in binocular rivalry,
NeuroReport 16 (2005) 1139–1142.
 R.K. Ellerkmann, V.M. Liermann, T.M. Alves, I. Wenningmann, S.
Kreuer, W. Wilhelm, H. Roepcke, A. Hoeft, J. Bruhn, Spectral entropy
and bispectral index as measures of the electroencephalographic effects
of sevoflurane, Anesthesiology 101 (2004) 1275–1282.
 A.K. Engel, W. Singer, Temporal binding and the neural correlates of
sensory awareness, Trends Cogn. Sci. 5 (2001) 16–25.
 A. Hasenstaub, U. Shu, B. Haider, U. Kraushaar, D.A. McCormick,
Inhibitory postsynaptic potentials carry synchronized frequency infor-
mationin active corticalnetworks,
 O.A. Imas, K.M. Ropella, B.D. Ward, J.D. Wood, A.G. Hudetz, Volatile
anesthetics enhance flash-induced gamma oscillations in rat visual cor-
tex, Anesthesiology 102 (2005) 937–947.
 O.A. Imas, K.M. Ropella, B.D. Ward, J.D. Wood, A.G. Hudetz,
Volatile anesthetics disrupt frontal-posterior recurrent information trans-
fer at gamma frequencies in rat, Neurosci. Lett. 387 (2005) 145–
 O.A. Imas, K.M. Ropella, J.D. Wood, A.G. Hudetz, Halothane augments
event-related gamma oscillations in rat visual cortex, Neuroscience 123
 O.A. Imas, J.D. Wood, A.G. Hudetz, Isoflurane disrupts frontoparietal
coherence during visual flash stimulation in rat, Society for Neuroscience
35th Annual Meeting, Washington, D.C., 2005.
 I.A. Iselin-Chaves, H.E. El Moalem, T.J. Gan, B. Ginsberg, P.S. Glass,
Changes in the auditory evoked potentials and the bispectral index fol-
lowing propofol or propofol and alfentanil, Anesthesiology 92 (2000)
 E.R. John, The anesthetic cascade: a theory of how anesthesia suppresses
consciousness, Anesthesiology 102 (2005) 447–471.
 E.R. John, L.S. Prichep, W. Kox, P. Valdes-Sosa, J. Bosch-Bayard,
E. Aubert, M. Tom, F. diMichele, L.D. Gugino, Invariant reversible
QEEG effects of anesthetics, Conscious. Cogn. 10 (2001) 165–
 Y. Kaminoh, H. Kamaya, C. Tashiro, I. Ueda, Multi-Unit and Multi-
Path system of the neural network can explain the steep dose-response
of MAC, J. Anesth. 18 (2004) 94–99.
 V.A. Lamme, Recurrent corticocortical interactions in neural disease,
Arch. Neurol. 60 (2003) 178–184.
 M.M. Muller, T. Gruber, A. Keil, Modulation of induced gamma band
activity in the human EEG by attention and visual information process-
ing, Int. J. Psychophysiol. 38 (2000) 283–299.
 P.S. Myles, K. Leslie, J. McNeil, A. Forbes, M.T. Chan, Bispectral
index monitoring to prevent awareness during anaesthesia: the B-aware
randomised controlled trial, Lancet 363 (2004) 1757–1763.
 A. Pascual-Leone, V. Walsh, Fast backprojections from the motion to the
primary visual area necessary for visual awareness, Science 292 (2001)
 E. Rodriguez, N. George, J.P. Lachaux, J. Martinerie, B. Renault, F.J.
Varela, Perception’s shadow: long-distance synchronization of human
brain activity, Nature 397 (1999) 430–433.
O.A. Imas et al. / Neuroscience Letters 402 (2006) 216–221 Download full-text
 K.M. Ropella, O.A. Imas, Coherence, Encyclopedia of Biomedical Engi-
neering, vol. 2, Wiley Publishers, Inc., 2005, pp. 914–925.
 J. Sarnthein, H. Petsche, P. Rappelsberger, G.L. Shaw, A. von Stein,
Synchronization between prefrontal and posterior association cortex dur-
ing human working memory, Proc. Natl. Acad. Sci. U.S.A. 95 (1998)
 W. Singer, Neuronal synchrony: a versatile code for the definition of
relations? Neuron 24 (1999) 49–65, 111-25.
 R. Srinivasan, D.P. Russell, G.M. Edelman, G. Tononi, Increased syn-
chronization of neuromagnetic responses during conscious perception,
J. Neurosci. 19 (1999) 5435–5448.
 P.N. Steinmetz, A. Roy, P.J. Fitzgerald, S.S. Hsiao, K.O. Johnson, E.
Niebur, Attention modulates synchronized neuronal firing in primate
somatosensory cortex, Nature 404 (2000) 187–190.
 M.M. Struys, E.W. Jensen, W. Smith, N.T. Smith, I. Rampil, F.J.
Dumortier, C. Mestach, E.P. Mortier, Performance of the ARX-derived
auditory evoked potential index as an indicator of anesthetic depth: a
comparison with bispectral index and hemodynamic measures during
propofol administration, Anesthesiology 96 (2002) 803–816.
 C. Summerfield, J.A. Mangels, Coherent theta-band EEG activity pre-
dicts item-context binding during encoding, NeuroImage 24 (2005)
 C. Tallon-Baudry, O. Bertrand, C. Delpuech, J. Pernier, Stimulus speci-
ficity of phase-locked and non-phase-locked 40Hz visual responses in
human, J. Neurosci. 16 (1996) 4240–4249.
 M. Thiebaut de Schotten, M. Urbanski, H. Duffau, E. Volle, R. Levy, B.
Dubois, P. Bartolomeo, Direct evidence for a parietal-frontal pathway
subserving spatial awareness in humans, Science 309 (2005) 2226–2228.
 H. Tiitinen, J. Sinkkonen, K. Reinikainen, K. Alho, J. Lavikainen,
R. Naatanen, Selective attention enhances the auditory 40-Hz transient
response in humans, Nature 364 (1993) 59–60.
 S. Weiss, P. Rappelsberger, Long-range EEG synchronization during
word encoding correlates with successful memory performance, Brain
Res. Cogn. Brain Res. 9 (2000) 299–312.
 P.F. White, J. Tang, H. Ma, R.H. Wender, A. Sloninsky, R. Kariger, Is
the patient state analyzer with the PSArray2 a cost-effective alternative
to the bispectral index monitor during the perioperative period? Anesth.
Analg. 99 (2004) 1423–1435.