level of consciousness of the patient.
Shortly afterward, in 1935, the Belgian neu-
rophysiologist Frederic Bremer
28
(see also
29
) ex-
amined the EEG waveforms in cats into which
he had placed lesions of the brainstem. He
found that after a transection between the me-
dulla and the spinal cord, a preparation that he
called the encephale isole, or isolated brain, ani-
mals showed a desynchronized (low voltage, fast,
i.e., waking) EEG pattern and appeared to be
fully awake. However, when he transected the
neuraxis at the level between the superior and
inferior colliculus, a preparation he called the
cerveau isole, or isolated cerebrum, the EEG
showed a synchronized, or high-voltage, slow-
wave pattern indicative of deep sleep and the
animals were behaviorally unresponsive. Bremer
concluded that the forebrain fell asleep due to
the lack of somatosensory and auditory sen-
sory inputs. He did not address why the ani-
mals failed to respond to visual inputs either
with EEG desynchronization or by making ver-
tical eye movements (as do patients who are
locked in).
This issue was addressed after World War II
by Moruzzi and Magoun,
30
who placed more
selective lesions in the lateral part of the mid-
brain tegmentum in cats, interrupting the as-
cending somatosensory and auditory lemniscal
pathways, but leaving the paramedian reticular
core of the midbrain intact. Such animals were
deaf and did not appear to appreciate somato-
sensory stimuli, but were fully awake, as indi-
cated both by EEG desynchronization and mo-
tor responses to visual stimuli. Conversely, when
they placed lesions in the paramedian reticu-
lar formation of the midbrain, the animals still
showed cortical-evoked responses to somato-
sensory or auditory stimuli, but the background
EEG was synchronized and the animals were
behaviorally unresponsive. Later studies showed
that electrical stimulation of the midbrain re-
ticular core could excite forebrain desynchro-
nization.
31
These observations emphasized the
midbrain reticular core as relaying important
arousing influences to the cerebral cortex, and
this pathway was labeled the ascending reticular
activating system. The origin of the pathway was
not established in this early work.
Subsequent studies, in which transecting
lesions were placed sequentially at different
levels of the brainstem in cats, demonstrated
that transections at the midpontine level or
caudally down to the lower medulla resulted in
animals that acutely spent most of their time in
12
Plum and Posner’s Diagnosis of Stupor and Coma
Box 1–2 The Thalamus, Basal Forebrain, and Generation
of EEG Waves
The origin of the sinusoidal appearance of the waveforms in the EEG remained a
mystery until the 1980s. Although it was understood that the EEG voltages are
due to the summated excitatory postsynaptic potentials in dendrites of cortical
neurons, the reason for the synchronous waves of dendritic potentials remained
elusive. The waves of postsynaptic potentials in the cerebral cortex are now un-
derstood to be due to the intrinsic burst firing of neurons in the thalamus, basal
forebrain, and the cortex itself, which produce waves of postsynaptic potentials in
cortical neurons.
When the membrane potential of burst neurons is close to their firing threshold,
they fire single action potentials that transmit sensory and other information. How-
ever, when burst neurons have been hyperpolarized to membrane potentials far
below their usual threshold for firing sodium action potentials, a low-threshold
calcium channel is deinactivated. When the low-threshold calcium channel is trig-
gered, calcium entry brings the membrane potential to a plateau that is above the
threshold for firing sodium action potentials. As a result, a series of sodium spikes are
fired, until sufficient calcium has entered the cell to activate a calcium-activated
potassium current. This potassium current then brings the cell back to a hyperpo-
larized state, terminating the burst of action potentials. The more deeply the resting
A
B
Waking
Slow-wave sleep
Bursts
Bursts
Ca
2+
Na
+
100 ms
50 mV
Single spikes
Single spikes
Thalamic
firing
intracellular
Thalamic
firing
extracellular
EEG
0.5 s
Figure B1–2. Thalamic relay neurons have transmission and burst modes of firing. (A) During
transmission mode, which operates mainly during wakefulness, individual neurons in the thalamus fire
single spikes in patterns that reflect their incoming afferent inputs. This correlates with a desynchro-
nized electroencephalogram. (B) During burst mode, the thalamic neurons are hyperpolarized by
gamma-aminobutyric acid (GABA)-ergic afferents, deinactivating a low-threshold calcium current with
a long plateau. This brings the cell above the threshold for firing sodium action potentials, which are
fired in a burst, until this is terminated by a calcium-activated potassium current that hyperpolarizes and
silences the cell. These bursts tend to fire rhythmically, in correspondence with high-voltage slow waves
in the EEG, which reflect large volleys of synchronized excitatory inputs reaching cortical dendrites.
(From Saper, C. Brain stem modulation of sensation, movement, and consciousness. Chapter 45 in:
Kandel, ER, Schwartz, JH, Jessel, TM. Principles of Neural Science. 4th ed. McGraw-Hill, New York,
2000, pp. 871–909. By permission of McGraw-Hill.)
(continued)
13
a wakeful state.
32
Thus, the lower brainstem
was thought to play a synchronizing, or sleep-
promoting, role.
33
Transections from the ros-
tral pons forward produced EEG slowing and
behavioral unresponsiveness. Periods of fore-
brain arousal returned after several days if the
animals were kept alive. However, it is clear
that the slab of tissue from the rostral pons
through the caudal midbrain (the mesopon-
tine tegmentum) contains neural structures
that are critically important to forebrain
arousal, at least in the acute setting.
At the time, little was known about the ori-
gins of ascending projections from the meso-
pontine tegmentum to the forebrain, and the
arousal effect was attributed to neurons in the
reticular formation. However, more recent stud-
ies have shown that projections from the meso-
pontine tegmentum to the forebrain arise from
several well-defined populations of neurons.
The major source of mesopontine afferents that
span the entire thalamus is a collection of cho-
linergic neurons that form two large clusters,
the pedunculopontine and laterodorsal tegmen-
tal nuclei.
34
These neurons project through the
paramedian midbrain reticular formation to
the relay nuclei of the thalamus (which innervate
specific cortical regions), as well as the midline
and intralaminar nuclei (which innervate the
entire cortex more diffusely), and the reticu-
lar nucleus. As noted in Box 1–2, the reticular
nucleus plays a critical role in regulating thala-
mocortical transmission by profoundly hyper-
polarizing thalamic relay neurons via GABA
B
receptors.
35
Cholinergic inputs in turn hyper-
polarize the reticular nucleus. Other neurons in
the cholinergic pedunculopontine and later-
odorsal tegmental nuclei send axons into the
lateral hypothalamus, where they may contact
populations of neurons with diffuse cortical pro-
jections (see below). Neurons in the pedun-
culopontine and laterodorsal tegmental nuclei
fire fastest during REM sleep (see Box 1–3) and
wakefulness,
36
two conditions that are charac-
terized by a low-voltage, fast (desynchronized)
EEG. They slow down during non-REM
(NREM) sleep, when the EEG is dominated by
high-voltage slow waves (Figure B1–3A).
membrane potential of the cells is hyperpolarized, the less frequent but longer
the bursts become.
The bursting behavior of neurons in the thalamic relay nuclei, which are a major
source of cortical inputs, is often thought to be a major source of cortical EEG.
The synchrony is credited to the thalamic reticular nucleus, which is a thin sheet
of GABAergic neurons that covers the thalamus like a shroud. Thalamic axons on
their way to the cerebral cortex, and cortical projections to the thalamus, give
off collaterals to the reticular nucleus as they pass through it. Neurons in the re-
ticular nucleus provide GABAergic inputs to the thalamic relay nuclei, which hy-
perpolarizes them and sets them into bursting mode.
However, there is evidence that the synchrony of EEG rhythms across the ce-
rebral cortex is due in large part to corticocortical connections, and that even
isolated slabs of cortex can set up rhythmic slow-wave potentials.
26
Recent evidence
suggests that the basal forebrain may play a critical role in entraining cortical rhyth-
mic activity. Basal forebrain neurons also fire in bursts that are time-locked to
cortical rhythms. In addition, cell-specific lesions of the basal forebrain can elimi-
nate fast cortical rhythms, including those associated with wakefulness and rapid
eye movement (REM) sleep, whereas large cell-specific thalamic lesions have sur-
prisingly little effect on the cortical EEG.
27
Thus, the waveforms of the cortical EEG appear to be due to complex interac-
tions among the burst neurons in the thalamus, cortex, and basal forebrain, all of
which receive substantial inputs from the ascending arousal system.
Box 1–2 The Thalamus, Basal Forebrain, and Generation
of EEG Waves (cont.)
14
Plum and Posner’s Diagnosis of Stupor and Coma
In addition, at the mesopontine level the
brainstem contains at least three different mo-
noamine groups whose axons project through
the hypothalamus to the cerebral cortex.
42
The noradrenergic locus coeruleus projects
through the paramedian midbrain reticular for-
mation and the lateral hypothalamus, inner-
vating the entire cerebral cortex diffusely.
43
Serotoninergic neurons in the dorsal and
median raphe nuclei project through a similar
course.
44
Mixed in with the serotoninergic
neurons are a smaller number of dopaminer-
gic cells, which are an extension of the ventral
tegmental dopamine group along the midline
of the midbrain, into the area under the ce-
rebral aqueduct.
45
These dopaminergic neu-
rons also project through the paramedian
midbrain reticular formation. Some of them
innervate the midline and intralaminar nuclei
of the thalamus, and others pass through the
lateral hypothalamus to the basal forebrain
and prefrontal cortex. Evidence from single-
unit recording studies in behaving animals
indicates that neurons in these monoaminergic
nuclei are most active during wakefulness, slow
down during slow-wave sleep, and stop almost
completely during REM sleep.
46–49
Application of monoaminergic neurotrans-
mitters to cortical neurons produces complex
responses.
35,50–52
In most cases, there is inhi-
bition resulting in a decrease in background
firing, although firing induced by the specific
stimulus to which the neuron is best tuned may
not be reduced to as great a degree as back-
ground firing. In an awake and aroused in-
dividual, this alteration in firing may result in
an improvement in signal-to-noise ratio, which
may be critical in sharpening cortical informa-
tion processing to avoid misperception of stim-
uli, such as occurs during a delirious state.
Although the cholinergic and monoaminergic
neurons in the mesopontine tegmentum have
traditionally been thought to play a major role
in regulating wake-sleep states, lesions of these
cell groups have relatively little effect on wake-
sleep states or cortical EEG.
53
Recent studies
0.2 mV
1 s
A
B
Figure 1–1. Electroencephalogram (EEG) from a cat in which Frederic Bremer transected the cervicomedullary
junction (A), showing a normal, desynchronized waking activity. However, after a transection at the midcollicular level (B),
the EEG consisted of higher voltage slow waves, more typical of sleep or coma. (From Saper, C. Brain stem modulation
of sensation, movement, and consciousness. Chapter 45 in: Kandel, ER, Schwartz, JH, Jessel, TM. Principles of Neural
Science. 4th ed. McGraw-Hill, New York, 2000, pp. 871–909. By permission of McGraw-Hill.)
Pathophysiology of Signs and Symptoms of Coma
15
Box 1–3 Wake-Sleep States
In the early days of EEG recording, it was widely assumed that sleep, like coma,
represented a period of brain inactivity. Hence, it was not surprising when the EEG
appearance of sleep was found to resemble the high-voltage, slow waves that ap-
pear during coma. However, in 1953, Aserinsky and Kleitman
37
reported the curi-
ous observation that, when they recorded the EEG as well as the electromyogram
(EMG) and the electro-oculogram (EOG) overnight, their subjects would period-
ically enter a state of sleep in which their eyes would move and their EEG would
appear to be similar to waking states, yet their eyes were closed and they were
deeply unresponsive to external stimuli.
37,38
This condition of REM sleep has also been called desynchronized sleep (from
the appearance of the EEG) as well as paradoxical sleep. More detailed study of
the course of a night of sleep revealed that the REM and NREM periods tend
to alternate in a rhythmic pattern through the night.
39–41
During active wakefulness, the EEG gives the appearance of small, desynchro-
nized waves and the EMG is active, indicating muscle activity associated with
waking behavior. In quiet wakefulness, the EEG often begins to synchronize, with
8- to 12-Hz alpha waves predominating, particularly posteriorly over the hemi-
sphere. Muscle tone may diminish as well. As sleep begins, the EEG rhythm drops
to the 4- to 7-Hz theta range, muscle tone is further diminished, and slowly roving
eye movements emerge (stage I NREM). The appearance of sleep spindles (waxing
and waning runs of alpha frequency waves) and large waves in the 1- to 3-Hz delta
range, called K complexes, denotes the onset of stage II NREM. The subject may
then pass into the deeper stages of NREM, sometimes called slow-wave sleep, in
which delta waves become a progressively more prominent (stage III) and then
dominant (stage IV) feature. During these periods, eye movements are few and
muscle tone drops to very low levels. This usually takes about 45 to 60 minutes, and
then the subject often will gradually emerge from the first bout of slow-wave sleep
to stage I again.
At this point, the first bout of REM sleep of the night often occurs. The subject
abruptly transitions into a desynchronized, low-voltage EEG, with rapid and
EEG
EMG
EOG
Awake
Sleep stage 1
2
3
4
REM
1 s
50 V
Figure B1–3A. The main features of a polysomnogram showing the eye movements (electro-oculogram
[EOG]), muscle tone (electromyogram [EMG]), and electroencephalogram (EEG) across the different
stages of sleep and wakefulness. During wakefulness, the EEG is desynchronized, the EMG is active, and
there are spontaneous eye movements. During non-rapid eye movement (NREM) sleep, the EEG be-
comes progressively slower, the EMG less active, and eye movements slow down or become slowly roving.
During REM sleep, there is a rapid transition to a desynchronized EEG, and irregular, rapid eye move-
ments, but the EMG becomes minimal, consistent with atonia. (From Rechtschaffen, A, and Siegel, J.
Sleep and dreaming. Chapter 47 in: Kandel, ER, Schwartz, JH, Jessel, TM. Principles of Neural Science.
4th ed. McGraw-Hill, New York, 2000, pp. 936–947. By permission of McGraw-Hill.)
(continued)
16
vigorous eye movements and virtually complete loss of muscle tone, except in the
muscles of respiration. The first bout of REM sleep during the night typically lasts
only 5 to 10 minutes, and then the subject will transition into stage I NREM, and
again begin to descend gradually into deeper stages of NREM sleep.
As the night progresses, the subject typically will spend progressively less time in
the deeper stages of NREM sleep, and more time in REM sleep, so that most of the
REM sleep for the night comes in the last few bouts. Spontaneous awakenings during
the night typically occur from the lighter stages of NREM sleep. Active dreams
Hours of night
1
2
3
4
5
6
7
4
3
2
1
REM
Awake
Old age
1
2
3
4
5
6
7
4
3
2
1
REM
Awake
Sleep stages
Early adulthood
1
2
3
4
5
6
7
4
3
2
1
REM
Awake
Childhood
Figure B1–3B. The stages of sleep through the night in a child, young adult, and older person. There is
usually regular progression from wakefulness through the stages of non-rapid eye movement (NREM)
sleep into its deepest stages, then progression back to light NREM sleep before the first REM episode
of the night. With successive cycles through the night, the amount of deeper NREM sleep becomes
less, and the amount of REM becomes greater. With aging, the amount of deep NREM sleep dimi-
nishes, and sleep fragmentation with more frequent awakenings is seen. (From Rechtschaffen, A, and
Siegel, J. Sleep and dreaming. Chapter 47 in: Kandel, ER, Schwartz, JH, Jessel, TM. Principles of Neural
Science. 4th ed. McGraw-Hill, New York, 2000, pp. 936–947. By permission of McGraw-Hill.)
(continued)
17
by Lu and Saper (unpublished) have focused on
neurons in the mesopontine tegmentum that
provide inputs to the basal forebrain, which is
critical for maintaining a wakeful state. Popula-
tions of neurons in the pre-locus coeruleus area
and medial parabrachial nucleus have intense in-
puts to the basal forebrain. Cell-specific lesions
of these neurons produce profound coma, sug-
gesting that they may be a major source of the
ascending arousal influence.
In addition, along the course of the ascending
cholinergic and monoaminergic axons through
the rostral midbrain reticular formation, there
are many additional neurons that project to the
thalamic relay, midline, and intralaminar nu-
clei.
34
Most of these neurons appear to be gluta-
matergic, and they may amplify the arousal signal
that arises in the mesopontine tegmentum. On
the other hand, they do not appear to be cap-
able of maintaining a waking state in the case of
acute loss of the influence from the mesopontine
neurons.
Along the course of the ascending arousal
systems, as they pass through the hypothala-
mus, are several hypothalamic cell groups that
augment the ascending projection to the basal
forebrain and cerebral cortex. These include his-
taminergic neurons in the tuberomammillary
nucleus as well as several populations of neu-
rons in the lateral hypothalamic area, all of which
project diffusely to the cerebral cortex and
innervate the intralaminar and midline thala-
mus.
54
There is considerable evidence that the
histaminergic input in particular is important
for maintaining a wakeful state. Histamine H
1
blockers impair wakefulness in both animals
and humans,
55
and transgenic mice lacking H
1
receptors have impairment of arousal responses
induced by intraventricular injection of the pep-
tide orexin.
56
Transgenic mice lacking histidine
decarboxylase show a deficit in wakefulness in-
duced by a novel environment, and mice in-
jected with an inhibitor of this key enzyme for
histamine synthesis similarly show less wake-
fulness.
57
Some of the lateral hypothalamic neurons
contain orexin,
58
a peptide that is associated
with arousal, and others contain melanin-
concentrating hormone
59,60
or GABA.
61
Many
neurons in the lateral hypothalamic area, in-
cluding those that contain orexin, fire fastest
during wakefulness and slow down during
both slow-wave and REM sleep.
62,63
Alterna-
tively, the firing of some lateral hypothalamic
neurons, which are likely to contain melanin-
concentrating hormone, increases during REM
sleep.
38,64,65
In addition, the ascending monoaminergic
and hypothalamic projections pass through the
basal forebrain, and along their pathway to the
cerebral cortex, they encounter and are aug-
mented further by additional populations of
occur predominantly during REM sleep, although many subjects report passive
dreams and ideation during NREM sleep as well.
This pattern, which is typical of young adults, changes dramatically across a life-
time. Infants spend much more time asleep, and much more time in the deeper
stages of NREM sleep, than adults. The amount of stages III and IV NREM sleep
Dostları ilə paylaş: |