2 the New Neuropsychology of Sleep: Implications for Psychoanalysis

7/25/2019 2 the New Neuropsychology of Sleep: Implications for Psychoanalysis

1/28

This article was downloaded by: [Gazi University]On: 17 August 2014, At: 22:15Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: MortimerHouse, 37-41 Mortimer Street, London W1T 3JH, UK

Neuropsychoanalysis: An Interdisciplinary Journal

for Psychoanalysis and the NeurosciencesPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/rnpa20

The New Neuropsychology of Sleep: Implications for

PsychoanalysisJ. Allan Hobson

a

aLaboratory of Neurophysiology, Massachusetts Mental Health Center, 74 Fenwood Road,

Boston, MA 02115

Published online: 09 Jan 2014.

To cite this article:J. Allan Hobson (1999) The New Neuropsychology of Sleep: Implications for Psychoanalysis,Neuropsychoanalysis: An Interdisciplinary Journal for Psychoanalysis and the Neurosciences, 1:2, 157-183, DOI:

10.1080/15294145.1999.10773258

To link to this article: http://dx.doi.org/10.1080/15294145.1999.10773258

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose ofthe Content. Any opinions and views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be reliedupon and should be independently verified with primary sources of information. Taylor and Francis shallnot be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and otherliabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
http://dx.doi.org/10.1080/15294145.1999.10773258http://www.tandfonline.com/action/showCitFormats?doi=10.1080/15294145.1999.10773258http://www.tandfonline.com/page/terms-and-conditionshttp://www.tandfonline.com/page/terms-and-conditionshttp://dx.doi.org/10.1080/15294145.1999.10773258http://www.tandfonline.com/action/showCitFormats?doi=10.1080/15294145.1999.10773258http://www.tandfonline.com/loi/rnpa20


2/28

57

The New Neuropsychology of Sleep: Implications

for Psychoanalysis

J Allan Hobson

Abstract:

n

his 1895 Project for a Scientific Psychology,

Sigmund Freud clearly stated his goal: to integrate the workings

of the mind with the workings of the brain. But in his day, too

little neurobiology was known to make his goal attainable and he

tried instead to understand such fascinating normal phenomena

as dreaming in exclusively psychodynamic terms. A century later,

Freud s brilliant but entirely speculative dream theory

is

in need

of radical revision,

not complete overhaul, because dreams,

as well as other unusual states of consciousness, can finally be

approached from the solidfoundation ofmodem neuroscience. In

other words, the goals of Freud s Project are at last within our

grasp. Ironically, an obstacle to progress

is

the tenacious adher-

ence oforthodox psychoanalysis to a theory that even by the stan-

dards of its originator,

is

now clearly outm04ed.

n

this paper, recent positron emission tomography (PET)

imaging and brain lesion studies

in

humans are integrated with

new basic research findings at the cellular level in animals to

explain how the formal cognitive features

of

dreaming may be the

combinedproduct

of

a shift in the neuromodulatory balance

of

the

brain and a related redistribution of regional blood

flow

The

human P T data indicate a preferential activation in rapid eye

movement (REM) sleep of the pontomesencephalic brainstem and

of

limbic and paralimbic cortical structures involved in the emo-

tional and mnemonic aspects of cognition. The pontine brainstem

mechanisms controlling the neuromodulatory balance of the brain

in rats and cats include noradrenergic and serotonergic influences

which enhance waking and impede REM via anticholinergicmech-

anisms, and cholinergic mechanisms which are essential to R

sleep and only come into ful l play when the serotonergic and

noradrenergic systems are inhibited.

n

REM, the brain thus be-

comes activated but processes its internally generated data

in

a

manner quite different from that

of

waking.

This paper combines with permission) the content

of

a recently pub

lished review by the author and his colleagues Robert Stickgold and Ed

ward Pace-Schott

(NeuroReport,

9:R9-R14, 1998) with an essay prepared

for a presentation to the Neuro-Psychoanalysis group

of

the

New

York

Psychoanalytic Institute on November 7, 1998. The author s research is

supported by NIMH grants MH13923 and MH48832, a NASA grant, and

by the Mind-Body Network of the John

D

and Catherine

T

MacArthur

Foundation.

J Allan Hobson is Professor of Psychiatry, Harvard Medical School;

Laboratory

of

Neurophysiology, Massachusetts Mental Health Center,

Boston.

The New Neuropsychology

n

Psychoanalysis

The field of sleep and dream research has recently

been invigorated by convergent new data from two

complementary neuropsychological sources Maquet

et aI., 1996; Braun et aI., 1997; Nofzinger, Mintun,

Wiseman, Kupfer, and Moore, 1997; Solms, 1997). In

the brain imaging and brain lesion studies to be re

viewed in detail below, the evidence for a strong brain

stem role in human REM sleep dream generation

complements the cellular and molecular level data in

animal studies and reveals an unexpectedly prominent

role

of

the limbic system in the selection and elabora

tion

of

dream plots. The emerging picture

of

dreaming

as

the synthesis

of

emotional and sensorimotor data

generated by the distinctive mechanisms

of

brain acti

vation in REM sleep will be

of

interest to all who

share Sigmund Freud s early vision of a psychology

founded on the solid base

of

neuroscience Freud,

1895) even as it forces revision

of

his highly specula

tive dream theory Freud, 1900).

Insofar as psychoanalysis remains committed to

Freud s view

of

dreaming, or to any revisions of that

view that retain the notion

of

disguise-censorship as

the mechanism of dream bizarreness, the new results

do not provide the faintest modicum of support. Also

without support is the corollary assumption that

dreaming affords privileged access to unconscious mo

tives via the technique

of

free association to bizarre

dream material.

If, instead, the modern psychoanalyst takes the

more open view, that afforded by the Activa

tion-Synthesis Hypothesis, that dreaming is a physio

logical projection test revealing, rather than

concealing, emotionally salient concerns, the prospect

is quite bright. The question is, can psychoanalysis

afford to admit that Freud was wrong about the dream

theory? Even if some would agree that it cannot any


3/28

158

longer afford not to do so, the consequences

such

an avowal are profound and far-reaching. Because the

dream theory is so foundational, its renunciation

forces a major reformulation upon the whole field.

One face-saving approach is to revisit the 1895

Project for a Scientific Psychology, which Freud

abandoned in favor the 1900 dream theory, and to

proclaim that the goals he set himself in that work are

the very same goals we now see to be within reach.

The intervening century could thus be viewed

as

an

unfortunate interlude fraught with the perpetuation

many regrettable errors but with such important suc

cesses

as

the work

John Bowlby on attachment and

separation. With that idea in mind let us now review

the recent data, and in its light examine the current

status Freud s dream theory.

rain ased Differences between Waking and

Dreaming

The demonstration that the human brain activation

pattern REM sleep is distinctly different from that

waking has an important bearing upon our concep

tion of how conscious states are generated by the

brain. It supports the hypothesis that quite different

mechanisms underlie waking and dreaming conscious

ness and that those differentiated mechanisms are

causally determinant the differences in our subjec

tive experiences

the two states. For some time, it has

been the cognitive similarities between waking and

dreaming that have been emphasized by dream psy

chologists (Antrobus, 1983; Foulkes, 1990, 1993,

1997; Moffitt, 1995). These similarities have been as

cribed to brain activation processes that were thought

to be identical in the two states and this inference

identity was supported by evidence from neurophysi

ology shared electrical and ionic mechanisms for

cortical EEG activation seen in both REM sleep and

waking (Llinas and Pare, 1991). Besides being unable

to account for the robust differences between wake

and dream consciousness (Kahn, Pace-Schott, and

Hobson, 1997), this inference is now clearly in need

amendment on physiological

g r o u n ~

Until now, the best available candidate mecha

nism for the differentiation dreaming from waking

cognition has been the drastic reduction in the release

the neuromodulators norepinephrine and serotonin

which drop from their steady high levels in waking to

almost zero in the REM sleep cats and rats (Hobson,

1988, 1990, 1992, 1994, 1997a; Mamelak and Hob

son, 1989; Hobson and Stickgold, 1994, 1995a). We

J.

Allan Hobson

suggest that the newly described differences in re

gional activation found in humans may result from the

same neuromodulatory differentiation found in animal

studies (see Hobson and Steriade, 1986; Steriade and

McCarley, 1990, for reviews) and predict that it is just

a matter of time before more sophisticated imaging

confirms these chemical differences in the human

brain too. Indeed, a REM-related decline in central

nervous system (eNS) serotonin has recently been

demonstrated in humans using depth electrodes and

microdialysis (Wilson et aI

1997).

rain Mind States and the Studyof Consciousness

One of the strongest supports for the scientifically hy

pothesized unity

brain and mind comes from the

changes in conscious experience that

we

all experi

ence when we doze off, fall deeply asleep, and, later,

dream. The initial loss of contact with the outside

world at sleep onset with its flurry

fleeting hypna

gogic images, the deeply unconscious oblivion

sleep early in the night, and the gripping hallucinoid

scenarios of late-night dreams, all have such strong

and meaningful underpinnings in brain physiology as

to make all but certain the idea that our conscious

experience

is the brain-mind s awareness

its own

physiological states (Hobson, 1994, 1997).

But whether or not they are accepted

as

firm

proof of brain-mind identity, these simultaneous sub

jective and objective events encourage the concept

a unified system which we call the brain-mind (Kahn

et aI 1997). And they further encourage a detailed

accounting in the separable analytic domains the

neurophysiology and psychology the events that

change, or remain the same, as the brain changes state.

It is within this paradigm

simultaneous conscious

state and brain-state change that I now review and

integrate data from three sources:

1

The formal and quantitative characteristics con

sciousness in waking, sleeping, and dreaming;

2

The cellular and molecular level brain events that

have been measured in awake, NREM, and REM

sleeping animals; and

3

The neuropsychological analysis

the effect

brain lesions and regional blood flow changes upon

the conscious states humans.

REM Sleep Dreaming Defined

Formal Features of

R

Sleep Dreams

REM sleep dreams have several distinctive formal fea

tures which the underlying brain state must somehow


4/28

h New Neuropsychology of Sleep

determine. They include: sensorimotor hallucinations;

bizarre imagery; the delusional belief that one is

awake; diminished self-reflective awareness; orienta

tional instability; narrative structure; intensification of

emotion; instinctual behaviors; attenuated volition;

and very poor memory. Table 1 summarizes these fea-

TABLE 1

The Formal Features of REM Sleep Dreaming

Hallucinations Especially visual and motoric, but occasionally in

any and all sensory modalities Snyder, 1970; McCarley and

Hoffman, 1981; Hobson, 1988; Zadra, Nielsen, and Don

deri, 1998).

Bizareness Incongruity imagery is

strange, unusual, or impossi

ble); Discontinuity imagery and plot can change, appear, or

disappear rapidly); Uncertainty persons, places, and events

are often bizarrely uncertain by waking standards) McCarley

and Hoffman, 1981; Porte and Hobson, 1986; Hobson et aI

1987; Hobson, 1988; Mamelak and Hobson, 1989; Williams,

Merritt, Rittenhouse, and Hobson, 1992; Reinsel, Antrobus,

and Wollman, 1992; Hobson and Stickgold, 1994, 1995b; Re

vonsuo and Salmivalli, 1995).

Delusion We are consistently duped into believing that we are

awake unless we cultivate lucidity) Purcell, Mullington,

Moffitt, Hoffman, and Pigea, 1986; LaBerge, 1990, 1992; Bar

rett, 1992; Kahan and LaBerge, 1994; Hobson, 1997b).

Self-reflection absent or greatly reduced relative to waking Recht

schaffen, 1978; Barrett, 1992; Bradley, Hollifield, and

Foulkes, 1992).

Lack of orientational stability Persons, times, and places are fused,

plastic, incongruous, and discontinuous McCarley and Hob

son, 1981; Hobson et aI., 1987; Hobson, 1988; Williams et

aI 1992; Rittenhouse et al., 1994; Stickgold, Rittenhouse, and

Hobson, 1994; Revonsuo and Salmivalli, 1995).

Narrative story lines Explain and integrate all the dream elements

in a confabulatory manner Foulkes, 1985; Cipolli and Poli,

1992; Hobson, 1988; Hunt, 1991; Montangero, 1991; Bla

grove, 1992).

Emotions increased Intensified and predominated by fear-anxiety

Nielsen, Deslauriers, and Baylor, 1991; Merritt, Stickgold,

Pace-Schott, Williams, and Hobson, 1994; Domhoff, 1996).

Instinctual programs especially fight-flight) often incorporated

Hobson and McCarley, 1977; Hobson, 1988; Jouvet, 1999).

Volitional control greatly attenuated Hartmann, 1966; Purcell et

aI 1986).

Memory deficits across dream-wake, wake-dream and

dream-dream transitions Cipolli, Baroncini, Fagioli, Fumai,

and Salzaruo, 1987; Fagioli, Cipolli, and Tuozzi, 1989;

Goodenough, 1991; Roussy et aI 1996; Pace-Schott,

Stickgold, and Hobson, 1997a,b; Roussy, Gonthier, Raymond,

Mercier, and DeKoninck, 1997).

159

tures and documents their identification and quantifi

cation. Our discussion is based upon our

psychophysiological theory

of

brain-mind isomor

phism.

We

assume that any enhancement or impair

ment)

of

any psychological function e.g., dreaming)

will be mirrored by enhancement or impairment) of

its physiological substrate s function e.g., REM

sleep). We have emphasized these formal aspects of

dreaming because they are noted in all REM sleep

dreams regardless of their specific narrative content.

We expect that REM sleep neurobiology will be able

to explain more about such features than it now can

about specific dream content.

ream Motor Hallucinosis Fictive Movement

The nature of the motor hallucinations of dreams de

serves special comment because it suggests that brain

mechanisms subserving active motor behaviors are

brought into play during REM. Thus, even office

bound intellectuals never dream of what they do every

day: sitting at their desks reading, writing, or analyzing

data Hartmann, 1996). Instead they ski, swim, fly or

play tennis in their dreams whether or not they have

recently done any of these things in their waking lives.

In contrast to the deficits in memory functions dis

cussed below, REM sleep dream consciousness rou

tinely has

more

motor hallucinatory content than

NREM consciousness and perhaps even more than

most waking fantasy. In this case we must look for

what has been

added

to brain function in REM. We

might expect to find an enhancement of those physio

logical processes which subserve internal visuomotor

activation. We would then predict selective activation

of

the visual system, the basal ganglia, the motor cor

tex, or subcortical motor pattern generators. The neu

rophysiological studies and PET data from humans

confirm these predictions.

ream Emotion

Emotion is a subjective experience that is intensified

in dreams. To account for the documented prominence

of

anxiety-fear, elation, and anger in dreams Nielsen

et aI., 1991; Merritt et aI., 1994; DoIhhoff, 1996) we

would not be surprised to find selective activation of

the limbic brain and this is the prediction most strongly

supported by the new neuroimaging evidence Maquet

et aI., 1996; Braun et aI., 1997; Nofzinger et aI., 1997).

That dream emotion is usually consistent with the


5/28

6

dream narrative Foulkes, Sullivan, Kerr, and Brown,

1988 and bizarre incongruities between emotion and

narrative are rarer than incongruities among other

dream elements Merritt et al., 1994 , can be explained

by viewing dream emotion as a primary shaper

of

plots

rather than as a reaction to them Seligman and Yellen,

1987 . Thus in a classic anxiety dream, the plot may

shift from feeling lost, to not having proper creden

tials, adequate equipment, or suitable clothing, to miss

ing a train. These plots all satisfy the driving

emotion nxiety while

being only very loosely as

sociated with one another.

Dream ognition

The distinctively discontinuous and incongruous na

ture of dream cognition can be measured

as

a construct

termed bizarreness Hobson, Hoffman, Helfand, and

Kostner, 1987; Hobson, 1988 . Bizarreness in turn re

flects the hyperassociative quality of REM sleep

dream consciousness. The instability of time, place,

and, most strikingly, person is a qualitatively unique

feature of REM sleep dreams. A dream character may

thus have the name of one

of

our friends but the wrong

face, hairstyle, or clothing. Other dream characters are

true chimeras having some

of

the features of one indi

vidual and some

of

another. Even the sexual identity

of

dream characters is fluid, and this ambiguity can

be anatomically explicit, not just psychological.

Dream mnesia and Related ognitive Deficits

The loss

of

memory in REM sleep makes dreaming

consciousness much more difficult to recall than wak

ing consciousness. This phenomenological deficit logi

cally implies a physiological deficit: some functional

process, present and responsible for memory in wak

ing, is absent or at least greatly diminished in REM

sleep.

In our attempt to explain dream amnesia, we look

with interest at such functional deficits as the loss

of

noradrenergic and serotonergic modulation in REM

sleep. This is because these very neuromodulators

have been shown, in many human and animal studies,

to be critical to learning and memory and to such

memory-enhancing cognitive functions as perception

and attention via their direct CNS effects

as

well as

their indirect peripheral mechanisms Kandel and

Schwartz, 1982; Quatermain, 1983; Frith, Dowdy,

Ferrier, and Crow, 1985; Clark, Geffen, and Geffen,

Allan Hobson

1987, 1989; Kandel, 1989; McGaugh, 1990, 1995;

Coull, Middleton, Sahakian, and Robbins, 1992;

Witte, Gordon-Lickey, and Marrocco, 1992; Robbins

and Everitt, 1994; Cahill, Prins, Weber, and

McGaugh, 1994; Abel et aI 1995 . This REM-related

aminergic demodulation is best viewed

as

a subtrac

tion of noradrenaline and serotonin from the varied

neuromodulatory mixture facilitating waking cogni

tion, a mixture which,

of

course, includes acetylcho

line e.g., Hasselmo and Bower, 1993 which remains

abundant during REM.

The loss of orientational stability which is at the

cognitive root of dream bizarreness and the loss of

self-reflective awareness which is the basis of the de

lusion that we are awake in our dreams are two re

lated deficits which could be caused by the aminergic

demodulation of the brain in REM sleep. But we have

wondered, is there more to it than that? Could the

frontal lobes be selectively inactivated during REM

sleep? At least two of the new PET studies reviewed

in a later section now suggest that this is so Maquet

et aI 1996; Braun et aI., 1997 .

REM Sleep Neurophysiology

In

1953, Eugene Aserinsky and Nathaniel Kleitman,

working in Chicago, discovered that the brian-mind

exhibited periodic self-activation during sleep. At reg

ular 90- to 100-minute intervals they observed the

spontaneous emergence of electroencephalographic

EEG desynchronization, accompanied by clusters of

rapid saccadic eye movements or REMs together

with acute accelerations of heart and respiration rates.

When subjects were awakened and asked to report

their antecedent mental activity, REM sleep was asso

ciated with longer, more vivid, more motorically ani

mated, and more bizarre accounts than NREM

Foulkes, 1962,1982,1985,1993; Rechtschaffen, Ver

done, and Wheaton, 1963; Goodenough, Lewis, Sha

piro, Jaret, and Sleser, 1965; Ogilvie, Hunt, Sawicki,

and Samahalski, 1982; Foulkes and Schmidt, 1983;

Antrobus, 1983; Hobson, 1990; Cavallero, Cicogna,

Natale, Occhionero, and Zito, 1992; Waterman, Elton,

and Kenemans, 1993; Stickgold, Pace-Schott, and

Hobson, 1994; Antrobus, Kondo, and Reinsel, 1995;

Casagrande, Violani, Lucidi, Buttinelli, and Bertini,

1996; Porte and Hobson, 1986; Kahn et aI 1997 .

Thus, while some dreaming can occur in other stages

of sleep Foulkes, 1962; Cavallero et aI 1992; for

reviews see Foulkes, 1967, 1985; Kahn et aI 1997;

Nielson, 1999 , it is REM neurophysiology which


6/28

The

New Neuropsychology of Sleep

161

Structural

Model

Cholinergic R Sleep Generation

REM

- NE 5HT

NREM

ake

The reciprocal interaction hypothesis (McCarley

and Hobson, 1975) provided a formal model for the

aminergic-cholinergic interplay at the synaptic level

and a mathematical model

of

the dynamics

of

the neu

robiological control system (Figure

1 .

In this section

we review subsequent

work

that has led to the alter

ation and elaboration (Figure 2)

of

the model.

C Activation Level A

1

..Ir 4l

B Dynamic

Model

t> RIM..:.Q

REM On . /

-----

Although there is abundant evidence for a pontine per

ibrachial cholinergic mechanism

of

REM generation

centered in the pedunculopontine (PPT) and laterodor

sal tegmental (LDT) nuclei (for recent reviews see

Hobson, 1992; Hobson, Datta, Calvo, and Quattrochi,

Figure

1.

The Original Reciprocal Interaction Model

of

physiological

mechanisms determining alterations in activation level. (A)

Structural

model

of

Reciprocal Interaction

REM-on cells

of

the pontine reticular

formation are cholinoceptively excited and/or cholinergically excitatory

(ACH

at their synaptic endings. Pontine REM-off cells are noradrener

gically (NE)

or

serotonergically (5HT) inhibitory

-

at their synapses. (B)

Dynamic Model

During waking the pontine aminergic system is tonically

activated and inhibits the pontine cholinergic system. During NREM sleep

aminergic inhibition gradually wanes and cholinergic excitation recipro

cal ly waxes. At REM sleep onset aminergic inhibit ion is shut

off

and

cholinergic excitation reaches its high point. (C)

ctivation level

As a

consequence

of

the interplay

of

the neuronal sys tems shown in A and B,

the net activation level

of

the brain is at equally high levels in waking and

REM sleep and a t abou t hal f this peak level in NREM sleep (Hobson,

1992).

The Reciprocal Interaction Hypothesis

The discovery

of

the ubiquity

of

REM sleep in mam

mals provided the brain side of the brain-mind state

question with an animal model (Dement and Wolpert,

1958; Jouvet and Michel, 1959; Jouvet, 1962; Snyder,

1966; Dallaire, Toutain, and Ruckebusch, 1974).

While animal studies showed that potent and wide

spread activation

of

the brain did occur in REM sleep,

it soon became clear that Moruzzi and Magoun s

(1949) concept

of

a brainstem reticular activating sys

tem required extension and modification to account for

the differences between the behavioral and subjective

concomitants

of

waking and those of REM sleep

(Hobson and Brazier, 1981).

A conceptual breakthrough was made possible by

the discovery of the chemically specific neuromodula

tory subsystems

of

the brainstem (Dahlstrom and

Fuxe, 1964; for reviews see Foote, Bloom, and Aston

Jones, 1983; Hobson and Steriade, 1986; Jacobs and

Azmita, 1992) and

of

their differential activity in wak

ing (noradrenergic and serotonergic systems on,

cholinergic system damped) and REM sleep

(noradrenergic and serotonergic systems off, choliner

gic system undamped) (Trulson and Jacobs, 1970; Chu

and Bloom, 1973, 1974; Hobson, McCarley, and Wyz

inski, 1975; McCarley and Hobson, 1975; McGinty

and Harper, 1976; Cespuglio, Faridi, Gomez, and

Jouvet, 1981; Aston-Jones and Bloom, 1981; Lydic,

McCarley, and Hobson, 1983, 1987; Jacobs, 1986;

Rasmussen, Morilak, and Jacobs, 1986; Reiner, 1986;

Lydic, Baghdoyan, and Lorinc, 1987; Steriade and

McCarley, 1990).

The resulting model

of

reciprocal interaction

(McCarley and Hobson, 1975) provided a theoretical

framework for experimental interventions at the cellu

lar and molecular level that has vindicated the notion

that waking and dreaming are at opposite ends

of

an

aminergic-cholinergic neuromodulatory continuum,

with NREM sleep holding an intermediate position

(Figure 1). This spectrum

of

brain activity across the

states

of

waking, NREM, and REM must be the neuro

biological substrate of the conscious experience asso

ciated with these states. We now devote our careful

attention to a review

of

the cellular and molecular

level details, in the context

of

the reciprocal interac

tion concept, to provide a basis for our discussion of

the new human imagery data in a later section.

most strongly supports dream psychology (Kahn et aI.,

1997). For this reason, we restrict our integrative ef

forts to the neuropsychology

of

REM sleep dreaming.


7/28

162

J. Allan Hobson

GABAergic nuclei

e.g.

substantia

nigra

pars

reticulat8

REM

Off REM On

B

Ach

+0:

1 i

5-HT 5.

:

----..., --- ----...-------,

Figure

2.

Synaptic Modifications

of

the Original Reciprocal Interaction

Model Based Upon Recent Findings. Reported data from animal (cat and

rodent) are shown as solid lines, some

of

the recently proposed putative dy

namic relationships are shown as dotted lines, and

references are indicated

by numbers see key below .

The

exponential magnification

of

cholinergic

output predictedby theoriginalmodel (Figure 1)

can

alsooccur in this model

with mutuallyexcitatory cholinergic-non-cholinergic interactions (7) taking

the place of the previously postulated, mutually excitatory cholinergic-cho

linergic interactions. Therefore the basic shape

of

reciprocal interaction s

dynamic model (illustrated in Figure

IB)

and its reluctant alternation

of

be

havioral state (illustrated in Figure 1C) would also result from this revised

model. The additional synaptic details

can

be superimposed

on

this revised

reciprocal interaction model without altering the basic effects

of

aminergic

and cholinergic influences on the REM sleep cycle.

For

example: i. Excit

atory cholinergic-non-cholinergic interactions utilizing ACh and the excit

atory amino acid transmitters enhance firing

of

REM

-on cel ls (6, 7) whi le

inhibitory noradrenergic (4), serotonergic (3) and autoreceptor cholinergic

(1) interactions suppress REM-on cells.

Cholinergic effects upon aminer

gic neurons are bothexcitatory (2), as hypothesized in the originalreciprocal

interactionmodel andmay also operate via presynaptic influences

on

norad

renergic-serotonergic as well as serotonergic-serotonergic circuits (8).

Inhibitory cholinergic autoreceptors (1)could contributeto the inhibition of

LOT

and

PPT

cholinergic neurons whichis also causedby noradrenergic (4)

and serotonergic (3) inputs.

iv.

GABAergic influences(9, 10) as well asother

neurotransmitters suchas adenosine and nitric oxide (see text) maycontrib

ute to the modulation

of

these interactions.

Abbreviations:

open circles, ex

citatory postsynaptic potentials; closed circles, inhibitory postsynaptic

potentials; mPRF, medial pontine reticular formation; PPT, pedunculopon

tine tegmental nucleus; LOT, laterodorsal tegmental nucleus; LCa, peri-lo

cus coeruleus alpha; 5HT, serotonin; NE, norepinephrine; ACh,

acetylcholine; GL, glutamate; AS, aspartate; GABA, gamma-aminobutyric

acid.

References:

1. Baghdoyan, Fleegal, Lydic, 1997; EI Manseri, Sa

kai,

Jouvet, 1990; Kodama

Honda, 1996;Leonard

Llinas, 1990, 1994;

Luebke, McCarley, Greene, 1993; Roth, Fleegal, Lydic, Bagndoyan,

1996; Sakai

Koyama, 1996; Sakai, EIManseri,

Jouvet, 1990. 2. Egan

North, 1985, 1986.3. Horner,Sanford,Annis, Pack, Morrison, 1997;Leo

nard Llinas, 1994; Luebke, Green, Semba, Kamodi, McCarley, Reiner,

1992; Thakkar, Strecker,

McCarley, 1997. 4. Sakai,

Koyama, 1996.

5. Portas, Thakkar, Rannie,

McCarley, 1996. 6. Sakai

Koyama, 1996;

Sakai Onoe, 1997; Vanni-Mercier, Sakai, Lin, 1989; Yamamoto, Ma

melak, Quattrochi,

Hobson, 1990a, b. 7. Greene

McCarley, 1990; Leo

nard Llinas, 1994; Sakai Koyama, 1996. 8. Li, Greene, Rannie,

McCarley, 1997. 9. Nitz Siegel, 1997; Datta, 1997b; Datta, Curro-Dossi,

Pare, Oakson, Steriade, 1991. 10. Porkka-Heiskanen, Strecker, Stenberg,

Bjorkum, McCarley, 1997a.

GA A

locus

Ach

t

coeruleus

2.

Experimental REM Sleep Induction and Suppression

Microinjection of cholinergic agonist or cholinesterase

inhibitors into many areas of the paramedian pontine

reticular formation induces REM sleep (Baghdoyan,

1993; Datta, 1995, 1997a; McCarley, Greene, Rannie,

and Portas, 1995; McCarley et aI., 1997), not all pon

tine PPT and LDT neurons are cholinergic (Steriade

et aI., 1988; Leonard and Llinas, 1990, 1994; Kang

and Kitai, 1990; Kamondi, Williams, Hutcheson, and

Reiner, 1992; Sakai and Koyama, 1996) and cortical

acetylcholine release may be as high during wake

fulness as during sleep (Jasper and Tessier, 1971; Mar

rosu et aI., 1995; Jimenez-Capdeville and Dykes,

1996).

The original claim, that themedial pontine reticu

lar formation (mPRF) was cholinergic, was clearly in

error. While many

of

the mPRF cells are excited by

acetylcholine as originally hypothesized, their own ex

citatory neurotransmitter now appears to be glutamate,

not acetylcholine. For this and other reasons to be

discussed below, reciprocal interaction (McCarley and

Hobson, 1975) and reciprocal inhibition (Sakai, 1988)

models for control of the REM sleep cycle by brain

stem cholinergic and aminergic neurons have recently

been questioned (Leonard and Llinas, 1994). Specifi

cally, the self-stimulatory role of acetylcholine on

pontine PGO-bursting neurons has not been confirmed

in in vitro slice preparations (Leonard and Llinas,

1995). For example, ACh has been shown to hyperpo

larize cell membranes in slice preparations

of

the ro

dent parabrachial nucleus (Egan and North, 1986)

LDT (Luebke, McCarley, and Greene, 1993; Leonard

and Llinas, 1994) and PPT (Leonard and Llinas,

1994). Similarly, LDT and PPT neurons with burst

discharge properties most like those hypothesized to

occur in PGO-burst neurons ( type I neurons) may

not be cholinergic (Leonard and Llinas, 1990).

Much evidence remains, however, that the recip

rocal interaction model accurately describes essential

elements of REM sleep cycle control even though

some

of

its detailed synaptic assumptions need correc

tion, as shown in Figure 2. Numerous findings confirm

the hypothesis that cholinergic mechanisms are essen

tial to the generation

of

REM sleep and its physiologi

cal signs (for recent reviews see Hobson, Lydic, and

Baghdoyan, 1986; Hobson and Steriade, 1986; Sakai,

1988; Steriade andMcCarley, 1990; Jones, 1991; Hob

on, 1992b; Hobson et aI., 1993; Datta, 1995, 1997).

A selection of many recent examples follows.

mesopontine

tegmentum

1

7. GL.

;1 _ 7. Ach

cholinergic non-cholinergic

PPT

and LOT mPRF and Lea

- Ach +OAch - NE

~ 1 . ~

6. 4.

5-HT

3.

raphe

NE+Q 0

B +Ach

B

GABA

i

14


8/28

The New Neuropsychology Sleep

Rodrigo-Angulo, McCarley, and Hobson, 1987; Bagh

doyan, Lydic, Callaway, and Hobson,

1989;

Vanni

Mercier, Sakai, and Lynn, 1989; Velazquez-Moctez

uma, Gillin, and Shiromani, 1989; Yamamoto, Ma

melak, Quattrochi, and Hobson, 1990a,b; Hobson et

aI 1993 . In addition to these short-term REM induc

tion sites, carbachol injection into a pontine site in the

caudal peribrachial area has been shown to induce

long-term over seven days REM enhancement

Calvo, Datta, Quattrochi, and Hobson, 1992; Datta,

Calvo, Quattrochi, and Hobson, 1992; Datta, Quat

trochi, and Hobson, 1993 . In rats, it has been difficult

to enhance REM sleep with carbachol Deurveiller,

Hans, and Hennevin, 1997 , but rat strains which are

genetically supersensitive to ACh show enhanced

REM sleep Benca et. aI., 1996 . In addition to the

well-known suppression

REM by muscarinic antag

onists Hobson et aI., 1986 , the new presynaptic anti

cholinergic agents have also been shown

to

block

REM Salin-Pascual and Jimenez-Anguiano, 1995;

Capece, Efange, and Lydic, 1997 .

holinergic Neurons and

R

Sleep

Cholinergic type II and III PPT and LDT neurons

have firing properties which make them well suited

for the tonic maintenance

REM Leonard and Lli

nas, 1990 , and both PPT and LDT neurons show spe

cifically c-fos and foslike immunoreactivity Fos-LI

following carbachol-induced REM sleep Shiromani,

Malik, Winston, and McCarley, 1995; Shiromani,

Winston, and McCarley, 1996 suggesting that they

participate in the genesis that state. Low amplitude

electrical stimulation the LDT enhances subsequent

REM sleep Thakkar, Portas, and McCarley, 1996

while electrical stimulation of the cholinergic LDT

evokes excitatory postsynaptic potentials in pontine

reticular formation neurons and these EPSPs can be

blocked by scopolamine Imon, Ito, Dauphin, and

McCarley, 1996 . The excitatory amino acid, gluta

mate, when microinjected into the cholinergic PPT,

increases REM sleep in a dose-dependent manner

Datta, 1997b; Datta and Siwek, 1997 .

cetylcholine Release and R sleep

Microdialysis studies show enhanced release

en

dogenous acetylcholine in the medial pontine reticular

formation during both natural Kodama, Takahashi,

andHonda, 1990 and carbachol-induced Lydic, Bagh-

163

doyan, and Lorine, 1991 REM sleep. Thalamic ACh

concentration mesopontine origin is higher in both

wake and REM than in NREM Williams, Comisarow,

Day, Fibiger, and Reiner, 1994 , and a REM-specific

increase

ACh in the lateral geniculate body has

been observed Kodama and Honda, 1996 . Both mus

carinic and nicotinic receptors participate in the depo

larization thalamic nuclei by the cholinergic

brainstem Curro-Dossi, Pare, and Steriade, 1991 .

holinergic Mediation of PGO Waves

PGO input to the LGB is cholinergic Steriade, Pare,

Parent, and Smith, 1988 , and can be antidromically

traced to pontine PGO-burst neurons Sakai and

Jouvet, 1980 . Stimulation

mesopontine neurons in

duces depolarization cortically projecting thalamic

neurons Curro-Dossi et

aI

1991 . Neurotoxic lesions

pontomesencephalic cholinergic neurons reduce the

rate PGO spiking Webster and Jones, 1988 and

PGO wavescan be blocked by cholinergic antagonists

Hu, Bouhassira, Steriade, and Deschenes, 1988 . It

may not be an exaggeration to state that the evidence

for cholinergic REM-sleep generation is now

so

over

whelming and so widely accepted that this tenet the

reciprocal interaction model is an established prin

ciple.

minergic Inhibition of the holinergic R

Generator

At the heart the reciprocal interaction concept is

the idea that cholinergic REM sleep generation can

only occur when the noradrenergic and serotonergic

mediators waking release their inhibitory constraint

the cholinergic generator. The evidence for inhibi

tory serotonergic and noradrenergic influences on cho

linergic neurons and REM sleep is now also quite

strong.

Decreased Serotonin Release in Natural

R

Sleep

In the cat, extracellular levels

serotonin are higher

in waking than in NREM and higher in NREM than

REM in the dorsal raphe Portas and McCarley, 1994

and the medial pontine reticular formation Iwakiri,

Matsuyama, and Mori, 1993 . This same state-depen

dent pattern is observed in the hypothalamus the rat

Auerbach, Minznberg, and Wilkinson, 1989; Imeri,


9/28

164

DeSimoni, Giglio, Clavenna, and Mancia, 1994).

Moreover, reduced extracellular serotonin concentra

tion in REM sleep has recently been demonstrated in

the human amygdala, hippocampus, orbitofrontal cor

tex and cingulate cortex Wilson et aI., 1997). Since

most of these structures show selective activation in

PET images of REM sleep, it can be inferred that the

human limbic system is turned on but demodulated

during dreaming.

Serotonergic Suppression of holinergic Systems and

R Sleep

Serotonergic neurons from the dorsal raphe have been

shown to synapse o n LDT and PPT neurons Honda

and Semba, 1994). Serotonin has been shown to hy

perpolarize rat cholinergic LDT cells in vitro Luebke

et aI., 1992; Leonard and Llinas, 1994) and to reduce

the proportion

of

REM sleep in vivo Horner, Sanford,

Annis, Pack, and Morrison, 1997). Serotonin has been

shown to counteract the REM-like carbachol-induced

atonia

of

hypoglossal motoneurons Kubin, Reignier,

et aI., 1994; Kubin, Tojima, Reignier, Pack, and Da

vies, 1996; Okabe and Kubin, 1997).

Suppression of

R

by Serotonin gonists

Microinjection

of

the serotonin agonist 8-0H-DPAT

into the peribrachial region impedes REM initiation

in cats Sanford et aI., 1994) and systemic injection

of

8-0H-DPAT into serotonin-depleted rats also sup

presses REM Monti et aI., 1994). Simultaneous unit

recording has shown that microinjection of 8 0H-

DPAT selectively suppressed the firing

of

REM-on

but not REM-and wake-on cells

of

the cholinergic

LDT and PPT Thakkar et aI., 1997). In vivo microdia

lysis of serotonin agonists into the dorsal raphe nu

cleus DRN) decreased DRN levels

of

serotonin

presumably via serotonin autoreceptors on DRN

cells) which in turn increased REM sleep percent Por

tas, Thakkar, Rainnie, and McCarley, 1996). Meso

pontine injection of a serotonin agonist depressed ACh

release in the lateral geniculate body Kodama and

Honda, 1996).

Suppression

of

R by Endogenous Norepinephrine

and its gonists

Locus coeruleus neurons have been shown to become

quiescent during REM in the monkey Rajkowski, Si-

J. Allan Hobson

lakov, Ivanova, and Aston-Jones, 1997) as well as in

the cat and rat Hobson and Steriade, 1986). Electrical

stimulation of the pons in the vicinity of the noradren

ergic) locus coeruleus reduced REM sleep in rats

Singh and Mallick, 1996). The alpha-2 noradrenergic

agonist clonidine suppresses REM in human subjects

Nicholson and Pascoe, 1991; Gentili etaI.,

1996) and

the cat Tononi, Pompeiano, and Cirelli, 1991) while

the noradrenergic antagonist idazoxan increases REM

when injected into the pontine reticular formation

of

cats Bier and McCarley, 1994). That the REM-sup

pressive effects

of

serotonin and norepinephrine are

additive is indicated by the suppression of REM sleep

in humans by acute dosage

of

antidepressant drugs

which inhibit the reuptake

of

serotonin, norepineph

rine, or both Vogel, 1975; Nicholson, Belyavin, and

Pascoe, 1989; Vogel, Buffenstein, Minter, and Hen

nessy, 1990).

Like cholinergic enhancement, aminergic sup

pression of REM sleep is now an established principle.

The 5-HT

1A

serotonin receptor may be of the greatest

importance in the inhibition

of

cholinergic firing in

the cat PPT Sanford et aI., 1994) and LDT Sanford

et aI., 1997) while the alpha-l receptor may be the

most important site for adrenergic REM suppression

Ross, Gresch, Ball, Sanford, and Morrison, 1995).

Modification

of

the Reciprocal Interaction Model

Modifications of simple reciprocal inhibition or inter

action models, which are consonant with recent find

ings, have been proposed for the brainstem control

of

REM sleep. All such modifications retain one or both

of the major tenets

of

the reciprocal interaction model:

cholinergic facilitation and adrenergic inhibition

of

REM.

Leonard and Llinas 1994) suggest in regard

to

the McCarley and Hobson 1975) model that indi

rect feedback excitation via cholinergic inhibition of

an inhibitory input or cholinergic excitation

of

an ex

citatory input or some combination

of

the two could

replace direct feedback excitation in their

model p.

327). A similar mutually excitatory or mutually inhibi

tory interaction between REM-on cholinergic and

REM-on noncholinergic mesopontine neurons has

also been proposed Sakai and Koyama, 1996). Such

a mechanism is depicted in Figure 2.

From recent in vitro studies in the rat, the follow

ing elaboration of reciprocal interaction has been pro

posed by

Li

et

ai.

in the McCarley laboratory Li et

aI., 1997). During waking, presynaptic nicotinic facili-


10/28

The

New Neuropsychology

of

Sleep

tation

of

excitatory locus coeruleus noradrenergic in

puts to the dorsal raphe enhances serotonergic firing.

During REM, when the locus coeruleus is silent, this

same presynaptic nicotinic input may facilitate sero

tonergic self-inhibition by the raphe neurons them

selves. In vivo microdialysis studies

of

GABA in the

cat further postulate selective suppression of norad

renergic locus coeruleus neurons by GABAergic inhi

bition during REM as

proposed by Nitz and Siegel

1997 .

It is important to realize that many of the studies

questioning reciprocal interaction Egan and North,

1986; Leonard and Llinas, 1990, 1994; Luebke et aI.,

1992 have been carried out on in vitro rodent models.

Exploring the relationship of these findings

to

the in

vivo mechanisms generating REM sleep signs in the

cat is only in its early stages Hobson et aI 1993;

Datta, 1995; Sakai and Koyama, 1996 . It seems possi

ble, for example, that the hyperpolarization by ACh

of cholinergic cells cited in these studies might be

explained by the presence

of

ACh autoreceptors which

contribute to homeostatic control of cholinergic activ

ity Leonard and Llinas, 1990, 1994; EI Manseri et

aI 1990; Sakai et aI 1990; Sakai and Koyama, 1996;

Kodama and Honda, 1996; Roth et

aI

1996; Bagh

doyan et aI., 1997 . In contrast to the hyperpolariza

tion

of

some mesopontine cholinergic neurons

by

cho

linergic agonists, in vitro studies have shown the

majority of medial pontine reticular formation

mPRF neurons to be depolarized by carbachol

Greene and McCarley, 1990 . This suggests that the

exponential self-stimulatory activation which can be

triggered by cholinergic stimulation in diverse meso

and medial pontine sites Hobson and Steriade, 1986;

Hobson et aI., 1986, 1993; Steriade and McCarley,

1990 may involve excitatory neurons which are non

cholinergic. Such cholinergic self-regulation com

bined with cholinergic-noncholinergic mutual

excitation is schematized in Figure

2

We conclude that the two central ideas

of

the

model are strongly supported by subsequent research:

1 noradrenergic and serotonergic influences enhance

waking and impede REM via anticholinergic mecha

nisms; 2 cholinergic mechanisms are essential to

REM sleep and come into full play only when the

serotonergic and noradrenergic systems are inhibited.

By restricting our discussion to cholinergic and

aminergic mechanisms, we do not exclude the contri

butions to the modulation

of

behavioral state of other

neuromodulatory systems such

as

GABAergic systems

Nitz and Siegel, 1997 ; nitroxergic systems Leonard

and Lydic, 1997 ; glutamatergic systems Qnoe and

165

Sakai, 1995 ; glycinergic systems Chase, Soja, and

Morales, 1989 ; histaminergic systems Saper, Sherin,

and Elmquist, 1997 ; adenosinergic systems McCar

ley et aI., 1997 ; or the neuropeptides Bourgin et aI.,

1997 . Nor do we exclude the contributions

of

numer

ous nonpontine structures such as the basal forebrain

Szymusiak, 1995 ; hypothalamus Saper et aI 1997 ;

the amygdala Sanford, Ross, Tejani-Butt, and Mor

rison, 1995 ; thalamic nuclei Mancia and Marini,

1997 ; central gray area Sastre, Buda, Kitahama, and

Jouvet, 1996 ; or the medulla Chase and Morales,

1990 . These other systems are reviewed elsewhere

Kahn et aI., 1997; Hobson, Pace-Schott, and

Stickgold, 2000 . Rather, we emphasize here those

aminergic and cholinergic mechanisms associated

with the executive control of REM sleep in reciprocal

interaction/inhibition models McCarley and Hobson,

1975; Sakai, 1988; Steriade and McCarley, 1990 .

While the studies we have reviewed here are nec

essarily restricted to data obtained in subhuman

models

of

REM sleep, an abundant psychopharmaco

logical literature provides indirect evidence that the

same mechanisms operate at the cellular and molecu

lar level in the human brain Sitaram, Wyatt, Dawson,

and Gillin, 1976; Sitaram, Moore, and Gillin, 1978a,b;

Hobson and Steriade, 1986; Nicholson et aI 1989;

Vogel et aI., 1990; Gillin, Sutton, and Ruiz, 1991;

Perry and Perry, 1995 . We now turn our attention to

new, more direct evidence supporting the assumptions

of cross-species homology.

um n Neuropsychology

Until recently, the experimental study of human REM

sleep dreaming has been limited on the physiological

side by the poor resolving power

of

the EEG. Even

expensive and cumbersome evoked-potential and

computer-averaging approaches have not helped to an

alyze and compare REM-sleep physiology with that

of waking in an effective way. This limitation has

probably reinforced the erroneous idea that the brain

activation picture

of

REM sleep and waking are identi

calor

at least, very similar.

Fortunately, technological advances in the field

of human brain imaging have now made it possible to

describe a highly selective regional activation pattern

of

the brain in REM sleep. At the same time, experi

ments

of

nature, in the form

of

strokes, have allowed

the locale

of

brain lesions to be correlated with deficits

or accentuations of dream experience in patients Dor

icchi and Violani, 1993; Solms, 1997 . The remark-


11/28

66

J. Allan Hobson

KEY: 1Increase; 1Decrease; - No Change.

P T maging Studies

of

R Sleep reaming

ably complementary results

of

these two approaches

are summarized in Table

2

TABLE 2

Imaging

of

Brain Activation in REM

nd

the Effects of Brain

Lesions on Dreaming

to be important for spatial imagery construction, an

important aspect

of

dream cognition. As Maquet (Ma

quet and Franck, 1997) emphasizes, those cortical ar

eas activated in REM are rich in afferentation from

the amygdala (anterior cingulate, right parietal opercu

lum) while those areas with sparse amygdalar affer

entation (prefrontal cortex, parietal cortex, and

precuneus) were deactivated in REM. Maquet et al.

interpreted their data in terms of the selective pro

cessing, in REM, of emotionally influenced memories

(see also Braun et aI., 1997; Maquet and Franck,

1997).

In another H

2

5

0 PET study, Braun et

aI

(1997)

replicated the Maquet group s findings

of

a consistent

REM-related brainstem, limbic, and paralimbic activa

tion. When REM-sleep brain activity was compared

to brain activity in delta NREM, with presleep waking

and with postsleep waking, Braun et

aI

showed rela

tive activation

of

the pons, the midbrain, the anterior

hypothalamus, the hippocampus, the caudate, and the

medial prefrontal, caudal orbital, anterior cingulate,

parahippocampal, and inferior temporal cortices in

REM sleep as compared to each of the above three

conditions (Table 3). Based on these observations, the

Braun group offered the following speculations which

are relevant

to

the neurology of dreaming.

The ascending reticular activation of REM

sleep may proceed relatively more via a ventral cholin

ergic route from the brainstem through the basal fore

brain rather than via the dorsal route through the

thalamus which is preferred in waking.

2

The activation of the cerebellar vermis in

REM sleep may reflect an input from the brainstem

vestibular nuclei and thus constitute a source of neu

ronal activation causing fictive movement in dreams

(Leslie and Ogilvie, 1996; Hobson, Stickgold, Pace

Schott, and Leslie, 1997).

3

The strong REM sleep-related activation of

the

basal ganglia

suggests that these subcortical struc

tures may play an important role in ascending thala

mocortical activation. The mediating network links

brainstemto the basal ganglia via the intralaminar thal

amic nuclei and proceeds to the cortext via the ventral

anterior and ventromedial thalamic nuclei. Because

this network contains multiple regulatory back projec

tions to the pedunculopontine tegmentum, a possible

role for the basal ganglia in the rostral transmission

of

PGO waves is suggested. The basal ganglia may

initiate motor activity and be related to the ubiquity

of hallucinated motion in dreams (Hobson, 1988;

Porte and Hobson, 1996).

(right)

PET Studies of Lesions Studies of

Activation in REM Effects on

Dreaming

REGION

Pontine Tegmentum

Limbic Structures

Striate Cortex

Extrastriate Cortex

Parietal Operculum

Dorsolateral Prefrontal

Cortex

Mediobasal FrontalCortex

Two very recent and entirely dependent PET studies

confirm the importance

of

the pontine brainstem in

the REM sleep activation of the human brain (Braun

et aI., 1997; Maquet et aI., 1996). This is an important

advance because it validates, for the first time, the

experimental animal data on the critical and specific

role of the pontine brainstem in REM-sleep genera

tion. At the same time these new studies also provide

important new data for our understanding

of

dream

synthesis by the forebrain. Instead

of

the global, re

gionally nonspecific picture of forebrain activation

that had been suggested by EEG studies, all of these

new imaging studies indicate a preferential activation

of limbic and paralimbic regions of the forebrain in

REM sleep compared to waking or to NREM sleep

(Braun et aI., 1997, Maquet et aI., 1996; Nofzinger et

aI., 1997). One important implication

of these discov

eries is that dream emotion may be a primary shaper

of dream plots rather than playing the secondary role

in dream plot instigation that was previously hypothe

sized (Hobson and McCarley, 1977).

Maquet et

aI

(1996) used an H

2

5

0 positron

source to study REM-sleep activation in their subjects

who were subsequently awakened for the solicitation

of dream reports. In addition to the pontine tegmen

tum, significant activation was seen in both amygdalae

and the anterior cingulate cortex (Table 3). Signifi

cantly, despite the general deactivation in much of the

parietal cortex, Maquet et

aI

reported activation of

the right parietal

operculum

brain region thought


12/28

The New Neuropsychology of Sleep

167

TABLE 3

Subcortical and Cortical Regional

Brain

Activation

and

Deactivation Revealed by Recent PET Studies

Comparing

REM Sleep

with Waking

and

with NREM Sleep

Comparison

REM vs. all other stages REM vs. waking

REM vs. pre- ( post-*) REM vs. NREM 3 4

sleep waking

Study

Maquet et aI., 1996 Nofzinger et aI., 1997 Braun et aI., 1997 Braun et aI., 1997

Technique H

2

0

18PDG

H

2

0

H

2

0

Subcortical Areas

Brainstem

Pontine tegmentum

increase

increase (R

increase

Midbrain

increase*

increase

Dorsal mesencephalon

increase

Diencephalon

Thalamus

increase: L

increase

Hypothalamus

increase: R, Lat

increase: A-POA

increase: A-POA

Limbic system

Left amygdala

increase

increase

Right amygdala

increase

Septal nuclei

increase

Hippocampus

increase*

increase

Basal ganglia striatum

Caudate

increase: A, I, L increase*

increase

Putamen

increase

Ventral striatum (n. ac-

increase

increase

cumbens, sub.innominata)

Cerebellum

increase

increase

(vermis)*

(vermis)

Cortical Areas

Frontal

Dorsolateral Prefrontal

decrease:

increase

decrease: 46*

L: 10, 46 47

R: 8, 9, 10, 11, 46

Opercular

decrease: 45*

Paraolfactory

increase

Lateral orbital

increase: 11, 12

decrease: 11 *

Caudal orbital

increase

increase

Gyrus rectus

increase

Parietal

Brodmann area 40

increase: R A 40

decrease: 40*

(supramarginal gyrus)

decrease: L 40

Angular gyrus

decrease: 39*

Precuneus

decrease

Temporal

Middle

increase R

Posterior superior

increase: 22

Inferior fusiform

increase: 37, 19 increase: 37, 19

(postsleep only)

Occipital

Postrolandic sensory

increase

Limbic ssociated

Medial (prelimbic) prefrontal

increase: R 32

increase: 10

increase: 10

Anterior cingulate

increase: 24

increase: 24

increase: 32*

increase: 32

Posterior cingulate

decrease: 31

dec: R sm. areas

decrease*

Infralimbic

increase: 25

Insula

increase: L

decrease-P

increase: A I

Parahippocampal

increase

increase: 37*

increase: 37

Entorhinal

increase

inc. (in fusiform)

Temporal pole

increase: 38

Abbr: L-Ieft h ~ m i s p h r R-right hemisphere; A-anterior; P-posterior; C-caudal; M-medial; Lat.-lateral; I-inferior; S-superior; A-POA anterior preoptic area; all numerals

Brodmann s area

*Change relative to both pre- and postsleep waking (Braun et

aI. , 1997


13/28

68

4

The REM-associated activation

of

unimodal

associative visual (Brodmann areas 19 and 37) and

auditory (Brodmann area 22) cortex contrasts with the

maintained (NREM and REM) sleep-related deactiva

tion

of

heteromodal association areas in the frontal and

parietal cortices. Interestingly, the inferior temporal

cortex (Brodmann areas 19 and 37) contains the fusi

form gyrus, a structure known to

be

involved in human

face recognition (McCarthy, Puce, Gore, and Truett,

1997) another common,

if

often bizarrely uncertain,

dream feature.

5

The REM-associated increase in activation

of

the

limbic associated medial prefrontal area

contrasts

with the prominent decrease in the executive portions

of

the frontal cortex (dorsolateral and orbital prefron

tal cortices). This medial area, which has the most

abundant limbic connections in the prefrontal cortex,

has been associated with arousal and attention. Disrup

tion of this area has been shown to cause confabula

tory syndromes formally similar to dreaming (Braun et

aI., 1997). Interestingly, lesions

of

the anterior limbic

cortex, especially the neighboring anterior cingulate,

often result in a distinctive syndrome in which dream

ing increases in vivacity and reality and dreaming be

come confused (Solms, 1997). From these findings as

well as primary visual cortex deactivat ion in REM,

the Braun group has recently suggested that REM con

stitutes, in the cortex, a unique condition

of

internal

information processing (between extrastriate and lim

bic cortices) functionally isolated from input (via stri

ate cortex) or output (via frontal cortex) to the external

world (Braun et aI., 1998).

Confirming the widespread limbic activation of

the human brain in REM, Nofzinger et

aI

(1997) de

scribed increased glucose utilization in the lateral hy

pothalamic area and the amygdaloid complex using an

I8F-fluoro-deoxyglucose (FDG) PET technique (Table

3). Nofzinger et

aI

note that,

The

largest area

of

activat ion is a bilateral confluent paramedian zone

which extends from the septal area into ventral stria

tum, infralimbic, prelimbic, orbitofrontal and anterior

cingulate

cortex

(p. 192). The authors suggest that

an important function

of

REM sleep is the integration

of

neocortical function with basal forebrain hypothala

mic motivational and reward mechanisms.

An equally interesting H

Q

PET finding, rele

vant to the cognitive deficits in self-reflective aware

ness, orientation, and memory during dreaming was

significant

deactivation

in REM,

of

a vast area

of

dorsolateral prefrontal cortex (Maquet et aI., 1996;

Braun et al., 1997). Using SPECT, a similar decrease

in cerebral blood flow to frontal areas during REM

J. Allan

Hobson

had earlier been noted (Madsen et aI., 1991). This dor

solateral prefrontal deactivation during REM, how

ever, was not replicated by the Nofzinger et

aI

FDG

study and this discrepancy remains to be clarified. U

s-

ing the finer time-resolution offered by functional MRI

(fMRI) imaging (Huang-Hellinger et aI., 1995; Ives et

aI., 1997; Sutton et aI., 1997) this area of research can

be expected to provide more detail in the near future.

The fact that considerable portions

of

executive

and association cortex are far less active in REM than

in waking led Braun et

aI

(1997) to speculate that

R M

sleep may constitute a state

of

generalized

brain activity with the specific exclusion

of

executive

systems which normally participate in the highest or

der analysis and integration

of

neural information

(p. 1190). In terms

of

cortical-subcortical networks,

Braun et aI suggest further that

the

limbic loop

connecting ventral striatum, anterior thalamus and

paralimbic cortices, appear to be activated during

REM sleep However, the prefrontal or associa

tion loop, connecting the caudate, dorsomedial thala

mus and prefrontal cortices

appears to be activated

only in a partial

or

fragmentary

way

(p. 1191). Fig

ure 3 integrates findings from these first three PET

studies comparing REM sleep to other states.

Loss Dreaming fter Cerebral Lesions

An entirely complementary set

of

findings and conclu

sions has been reached following a neuropsychological

survey of 332 clinical cases with cerebral lesions

(Solms, 1997). The 112 patients who reported a

global cessation

of

dreaming had damage either in

the parietal convexity or suffered disconnections

of

the mediobasal frontal cortex from the brainstem and

diencephalic limbic regions. Solms, who was appar

ently unaware

of

the recent PET studies, cited the

much earlier single subject report

of

a PET glucose

activation of limbic and prelimbic structures in REM

(Heiss, Pawlik, Herholz, Wagner, and Wienhard,

1985). With respect to the visual imagery aspect, a

decrease in the vivacity

of

dreaming was reported

by two patients with damage to the seat

of

normal

vision in the medial-occipital-temporal cortex (espe

cially areas V

3

,

V

3a

,

and V

4

but not VI, V

s

or V

6

Solms (1997) also-reports that his patients with pon

tine lesions continued to dream and concludes that the

pons is not necessary for human dreaming. Based

upon the difficulty

of

suppressing REM by experimen

tal lesions

of

the pons in animals, we suggest an alter

native explanation. It seems to us that any lesion


14/28

Th e

New Neuropsychology

of

Sleep

Dorsolateral

Prefrontal

ortex

ctivated in R

Deactivated

in R

Figure 3. Convergent Findings

on

Relative Regional Brain Activation and

Deactivation in REM Compared to Waking. A schematic sagittal view

of

the human brain showing those areas of relative activation and deactivation

in REM sleep compared to waking and/or NREM sleep which were re

ported in

two or more

of the three PET studies published to date (Maquet

et aI., 1996; Braun et aI., 1997; Nofzinger et aI., 1997). Only those areas

which could be easily matched between two or more studies are schemati

cally illustrated here and a realistic morphology of the depicted areas is

not implied. Note that considerably more extensive areas of activation are

reported in the individual studies and these more detailed findings are

given in Table 3. The depicted areas in this figure are thus most realistically

viewed as representative portions of large eN S areas subserving similar

functions (e.g., limbic-related cortex, ascending activation pathways and

multimodal association cortex).

capable of destroying the pontine REM sleep genera

tor mechanism would have to be

so

extensive as to

eliminate consciousness altogether.

motionally Salient emory Processing

Concerning the functional significance of the imaging

results, all three

of

the image study authors assign

REM sleep a role in the processing of emotion in mem

ory systems (Maquet et aI., 1996; Braun et aI. 1997;

Nofzinger et aI., 1997; Maquet and Franck, 1997). Ad

ditionally, both the Maquet and Braun groups suggest

the possible origin

of

dream emotionality in REM

associated limbic activation and dream-associated ex

ecutive deficiencies in REM-associated frontal deacti

vation (Braun et aI., 1997; Maquet and Franck, 1997).

Additional findings support this proposed cortico-lim

bic interaction. First,

as

shown in Table 3 the cingu

late cortex has consistently shown increased activation

in REM in other PET studies (Buchsbaumet

aI.

1989).

Second, FDG PET activation of anterior medial struc

tures, including the anterior cingulate and medial fron

tal cortex, was found to correlate with REM density

in the REM period during which FDG uptake occurred

169

(Hong, Gillin, Dow, Wu, and Buchsbaum, 1995). Al

though these authors interpret this medial activity cor

relation with REM density

as

resulting from the

activity

of

midline attentional systems in response to

cortically generated dream imagery (Hong et aI.

1995), it is equally possible that activation

of

these

structures reflects the limbic and paralimbic activity in

REM suggested by the Nofzinger, Maquet, and Braun

studies (Maquet et aI., 1996; Braun et aI., 1997; Nof

zinger et aI. 1997). Finally, a recent study

of

human

limbic structures with depth electrodes has shown that

a distinctive rhythmic delta-frequency EEG pattern

occurs only during REM sleep (Mann, Simmons, Wil

son, Engel, and Bragin, 1997).

The regional activation during REM may reflect

a specific activation

of

subcortical, and cortical, limbic

structures for the adaptive processing of emotional and

motivational learning (Maquet et aI. 1996; Nofzinger

et aI. 1997). Such processing may, in turn, account

for the emotionality and psychological salience of

REM sleep dreams (Hobson, 1988; Braun et aI. 1997).

Some support for this comes from a PET (glucose)

study showing correlation between content-analyzed

dream anxiety and medial frontal activation

(Gottschalk et

aI.

1991).

reud sDream Theory Revised in the Light of

Modern Sleep

an d

Dream Research

There are five basic tenets to Freud s disguise--eensor

ship dream theory which can now be contrasted with

aspects of the activation-synthesis hypothesis.

The instigation

dreaming was, for Freud,

caused by the upsurge

of

unconscious wishes follow

ing suspension of their wake-state repression. We

would now say that dreaming is caused by brain acti

vation during sleep. In NREM sleep, residual brain

activation is at a low level hence dreaming is less

intense and less sustained, whereas in REM the brain

activation is

as

vigorous as that in waking but that

activation is both biochemically and regionally differ

entiated from waking.

The bizarre character dreaming was, for

Freud, determined by the disguise and censorship

of

the unconscious wishes. In order to protect conscious

ness from disruption (which would otherwise lead to

awakening), such defensive processes as displacement,

condensation, and symbolization were called into

play. Even the visual hallucinosis of dreaming was

viewed by Freud as a regression to the sensory side


15/28

7

that served to neutralize the impact o the unconscious

wishes on consciousness.

We

would now say that dreaming is bizarre be

cause o the distinctive neurophysiology o REM sleep

with its shift from top-down cortical control in waking

to bottom-up phasic autoactivation epitomized by the

PGO process. The distinctive regional activation pat

tern

o

the forebrain (with limbic, paralimbic, and pa

rietal cortical regions taking precedence over the

dorsolateral prefrontal cortex), and the global shift in

the biochemical mode

o

information processing

(caused by the noradrenergic and serotonergic demod

ulation) also contribute to dream bizarreness.

The visual nature dreams is for activation-syn

thesis, not defensive as Freud asserted, but rather the

direct result o both bottom-up activation processes

(which also convey information about brainstem/eye

movement commands). The degree to which primary

vs. secondary unimodal and multimodal cortices medi

ate dream vision and the relative influence

o

subcorti

cal--cortical vs. corticocortical activation processes

remains to be clarified. But the answers to these fasci

nating questions have no direct bearing on the dis

guise-censorship postulate which is the heart o the

Freudian dream theory.

While dreaming, we never sit and watch but are

constantly moving through dream space. Thus dreams

are not properly thought o as simply visual but more

accurately as visuomotor. This feature was not recog

nized by Freud even he did correctly infer the neces

sity

o

inhibiting motor output to prevent the

behavioral enactment o the fictive movement o

dreams. This is a key point for activation-synthesis

(and any modern general theory

o

hallucinosis) be

cause o the apparent link between brainstem motor

pattern generators and cortical sensory processes. A

strong implication is that far from being a feedback

regression to the sensory side, dream hallucinosis re

sults from a feed-forward conveyance

o

data about

motor intentions to the sensory processors in the up

per brain.

The emotional character

dreams

was not easily

dealt with by Freud's theory. He repeatedly side

stepped the obvious inconsistency between the pres

ence o sleep-disruptive anxiety and his mechanistic

postulate o disguise and censorship o the psychonox

ious dream content. The presence o dream anxiety

also inveighed against his guardian o sleep func

tional hypothesis. In other parts o Freudian theory

anxiety is viewed as a symptom o inadequate compro

mise between id and ego functions and he supposed

it to so operate in dreaming sleep.

J. Allan Hobson

Activation-synthesis has always regarded anxiety

(which is the leading emotion in all dreams and all

dreamers [Nielsen et aI., 1991; Merritt et aI., 1994;

Domhoff, 1996]) as the primary product

o

limbic lobe

activation. This speculative hypothesis has now re

ceived powerful support from recent PET studies

(showing amygdala, parahippocampal cortex, and an

terior cingulate gyrus activation in human REM

sleep).

In this sense anxiety (and the two other prominent

dream emotions, elation and anger) can be seen as

major dream plot organizers and shapers

o

the emo

tional salience o dreams. For the reform-minded psy

choanalyst, this should be heard as brain-music

o

divine inspiration because it suggests that dreaming is

a conscious state in which the impact o emotion upon

cognition is more clearly demonstrated than in wak

ing. A major reason for this shift must be that, together

with the limbic lobe activation, the seat

o the execu

tive control o cognition in the dorsolateral prefrontal

cortex is deactivated in REM

But let

us

make no mistake. This cortex vs. limbic

system construct in no way vindicates disguise-censor

ship nor does it support any interpretive scheme that

carries any trappings o that dogma. On the contrary,

it favors the view that dreams are, in part, the transpar

ent exposure

o

an individual's cognitive associations

to, and means o coping with, anxiety (where anxiety

is viewed

as

an inherent, existential emotion that is

designed to interact with cognition in an adaptive

disruptive manner).

mnesia for Dreams Freud ascribed dream for

getting to repression. Why repression was needed

the dreams had already been bowdlerized by disguise

and censorship was never made clear. Activation-syn

thesis says that both the dream forgetting following

awakening and the defective episodic memory that oc

curs within dreams is a simple, state-dependent amne

sia compounded by the global aminergic demodulation

and the deactivation

o working memory mechanisms

in the dorsolateral prefrontal cortex. Activation-syn

thesis further asserts that the reason that dream recall

is enhanced by spontaneous or stimulated awakenings

from REM sleep is because the aminergic demodula

tion and prefrontal cortical deactivation o REM are

suddenly reversed. As is well known, even these post

arousal recollections are evanescent probably because

it takes several minutes to reinstate waking cognition.

From the Neuron to the Dream: Freud's

Project Realized

Taken together, these new neuroimaging and brain le

sion studies strongly suggest that the forebrain activa-


16/28

The New Neuropsychology o Sleep

tion and synthesis processes underlying dreaming are

very different from those o waking. Not only is REM

sleep chemically biased, but the preferential choliner

gic neuromodulation and aminergic demodulation are

associated with selective activation of the subcortical

and cortical limbic structures which mediate emotion

and with relative inactivation o the frontal cortex

which mediates directed thought . A unifying neuro

biological hypothesis is that the regional blood flow

changes are causally linked to the neuromodulatory

dynamics in the following way. Those areas which are

inactivated in REM are those undergoing aminergic

demodulation but are uncompensated by cholinergic

modulation while the activated areas are those heavily

targeted by cholinergic modulatory neurons.

Whatever the link between the neuromodulatory

and regional blood flow data, these findings greatly

enrich and inform the integrated picture ofREM sleep

dreaming as emotion-driven cognition with deficient

memory, orientation, volition, and analytic thinking.

And now that we know that there is a close fit between

the animal and human data regarding the mechanism

and pattern

o

brain activation in REM sleep, we are

in a much stronger position to strengthen the brain

based theory

o

dreaming first proposed twenty years

ago Hobson and McCarley, 1977 . We will now at

tempt to integrate the newly discovered facts from the

human imaging studies with the cellular and molecular

level findings gleaned from the animal model in order

to answer four questions: 1 What is the origin

o

dreaming? 2 Why are dreams cognitively distinc

tive? 3 Why are dreams forgotten? 4 What is the

function o dreaming?

The Origin reaming

Dreaming is a state o consciousness arising from the

activation o the brain in REM sleep. The brain activa

tion which underlies dreaming is, like that o waking,

a result o the excitation o forebrain circuits by im

pulses arising in the ascending activation systems

o

the brainstem e.g., pontine and midbrain reticular ac

tivating systems and basal forebrain e.g., cholinergic

Nucleus Basalis

o

Meynert . This activation process

prepares the forebrain to process data with associated

cognitive awareness. But REM-sleep brain activation

differs from that o waking in three important ways:

There is selective activation o occipital, parietal

and limbic zones with a selective inactivation

o

frontal regions.

7

2 The mechanism o the brainstem triggering o fore

brain activation involves the spontaneous excitation

o

cholinergic neurons in the pontomesencephalic

LDT and PPT nuclei. This occurs as the inhibitory

restraintupon themdeclines with the near total arrest

o

firing by noradrenergic neurons in the locus coer

ulus and serotonergic neurons in the raphe nuclei.

3

Besides the recruitment o the pontine and mesen

cephalic reticular formation which mediate the

tonic thalamocortical activation the disinhibited

cholinergic system appears to

pl y

role in provid

ing the activated forebrain with phasic activation

signals, the PGO waves, that have two targets o

particular relevance to dream theory:

a The lateral geniculate body and posterolateral

cerebral cortex, the presumed substrates of

visual imagery in dreaming, and

b Limbic and paralimbic structures, the pre

sumed substrates o emotion and emotionally

salient dream memories.

The istinctive Nature ream ognition

The selective activation process described above may

account for such distinctive cognitive features o

dreaming as:

The intense and vivid visual hallucinosis, due to

autoactivation o the visual brain;

2 The intense emotions, especially anxiety, elation,

and anger, due to the autoactivation o the amyg

dala, and more medial limbic structures;

3 The delusional belief that we are awake, the lack

o directed thought, the loss of self-reflective

awareness, and the lack o insight about illogical

and impossible dream experience, due to the com

bined and possibly related effects of aminergic de

modulation and the selective inactivation o the

frontal cortices;

4

The bizarre cognition o dreaming which is charac

terized by incongruities and discontinuities o

dream characters, loci, and actions, due to an orien

tational instability caused by:

a the chaotic nature o the pontine autoactiva

tion process and its sporadic engagement o

association cortices,

b the absence o frontal cortical monitoring,

and

c the memory deficits.


17/28

172

J. Allan Hobson

Figure 4 Physiological Signs and Regional Brain Mechanisms

of

REM

Sleep Dreaming Separated into the Activation (A), Input Source (I), and

Modulation (M) Functional Components of the AIM Model (see Hobson,

1992a; Hobson et aI., 2000). Dynamic changes in A, I and M during REM

sleep dreaming are noted adjacent to each figure. Note that these are highly

schematized depictions which illustrate global processes and do not at

tempt to comprehensively detail all the brain structures and their interac

tions which may be involved in REM sleep dreaming (see text, Table 3

and Hobson et aI., 2000 for additional anatomic details).

ream orgetting

The practically total amnesia that most humans have

for their dream consciousness is most likely a joint

product

of

the aminergic demodulation and the frontal

deactivation of REM sleep. Cellular and molecular

level studies

of

learning and memory all concur in

supporting a role for the aminergic neuromodulators,

especially serotonin, norepinephrine, and dopamine

(Flicker, McCarley, and Hobson, 1981; Kandel and

Schwartz, 1982; Quartermain, 1983; Libet, 1984; Frith

et aI 1985; Montarolo et aI 1986; Kandel, 1989;

Abel et aI 1995). Without their mediation, signals

which arrive at a postsynaptic neuron may have in

stantaneous effects upon its membrane potential but

lack the specific second messenger instruction needed

by intracellular metabolic substrates to store a record

of the membrane events.

At first glance, this failure to record intercellular

transactions would seem to be at odds with the en

hancement of learning hypothesis advanced in the next

section. But

if

we recall that it is consolidation, not

acquisition that is hypothetically enhanced, it may be

quite useful

to

direct the brain-mind to the exclusive

task of processing information already acquired in

waking and

to ignore or

even

disc rd the

informa

tion that it necessarily generates as it self-activates in

the interests of consolidation. On this view, the dream

is the often meaningless sometimes meaningful noise

that is made when the brain enters its active, memory

consolidation mode.

Of course, it is also quite possible, and even prob

able, that a key aspect of memory consolidation in

volves emotional salience. But whether this aspect is

very different from that operating in the waking

st te s those psychologists who regard dreaming as

a privileged communication from the unconscious

mind still

hold rem ins

to be established.

The unction

reaming

ort x

minergically

demodulated

t

Recent memory

tOrientation

en lY input

blocked

Real wortd data unavailable

otor

outputblocked

Real

action

impossible

Pons

minergic neurons off

tNE

t T

Chotinergicneurons

on

t ch

Modulation

Input Source

P O syst m tumed

2 the New Neuropsychology of Sleep: Implications for Psychoanalysis

Documents