вход по аккаунту


Cortical control of spatial memory in humans The visuooculomotor model.

код для вставкиСкачать
Cortical Control of Spatial Memory in
Humans: The Visuooculomotor Model
Charles Pierrot-Deseilligny, MD,1 René M. Müri, MD,2 Sophie Rivaud-Pechoux, PhD,1
Bertrand Gaymard, MD, PhD,1 and Christoph J. Ploner, MD3
We review current knowledge of the cortical control of spatial memory, studied using visuooculomotor paradigms.
Spatial memory is an essential cognitive process that can be involved in preparing motor responses. Our knowledge of
spatial memory in humans recently has progressed thanks to the use of ocular saccades as a convenient model of motor
behavior. Accuracy of memory-guided saccades, made to the remembered locations of visual targets, is a reflection of
spatial memory. For the performance of memory-guided saccades with brief delays (up to 15–20 seconds), that is,
involving short-term spatial memory, lesion studies have shown that the posterior parietal cortex, the dorsolateral prefrontal cortex, and the frontal eye field play significant roles. Studies of memory-guided saccades using transcranial
magnetic stimulation have suggested that the right posterior parietal cortex is involved at the initial stage (<300 milliseconds) of visuospatial integration, whereas the dorsolateral prefrontal cortex in both hemispheres controls the following
phase of short-term memorization, the frontal eye field mainly serving to trigger saccades. The new concept of a mediumterm spatial memory has emerged from a behavioral study of memory-guided saccades in normal subjects, showing a
paradoxical spontaneous improvement of spatial memory after delays of approximately 20 seconds. Lesion studies have
shown that the parahippocampal cortex could specifically control this medium-term spatial memory. Last, different
experimental and clinical arguments suggest that, after a few minutes, the hippocampal formation finally takes over the
control of spatial memory for long-term spatial memorization. Therefore, spatial memory involved in the memorization
of visual items could be successively controlled by the dorsolateral prefrontal cortex (short-term spatial memory), the
parahippocampal cortex (medium-term spatial memory), and the hippocampal formation (long-term spatial memory),
depending on specific periods of times. The applicability of this simple visuooculomotor model of spatial memory to
other types of stimuli and general motoricity has yet to be confirmed.
Ann Neurol 2002;52:10 –19
Spatial memory is a cognitive function allowing a subject to memorize the locations of diverse types of sensory stimuli occurring in the environment and to
guide, after their disappearance and variable delays, a
correct motor response of a part of the body or the
whole body toward the remembered locations of these
stimuli. In animals and humans, spatial memory has
been extensively studied for either short delays (of a
few seconds) or long delays (more than a few minutes).
However, this traditional dichotomy between “shortterm” and “long-term” memory appears to be an oversimplification because memory is continuously reorganizing.1,2 In fact, little is known about spatial memory
with intermediate delays, that is, between a few seconds and a few minutes. Moreover, several cortical areas have been identified as being involved in the control of spatial memory, but the time limits of their
control remain relatively unclear.
Four main complementary methods may be used to
study spatial memory or any other cerebral function in
humans: (1) in normal subjects, purely behavioral studies serve to define the normal characteristics of the investigated paradigms; (2) in patients, lesion studies are
necessary to determine whether a given cortical area is
crucial in the control of a specific paradigm, that is, by
showing that after damage to this area significant abnormalities exist in the execution of the paradigm; (3)
in normal subjects, functional imagery, which is the
method with the best spatial resolution, is useful to define the precise locations of the studied areas and also
to show the extension of the cortical network involved
in a paradigm; and (4) in normal subjects, transcranial
magnetic stimulation (TMS), a method with a good
temporal resolution, allows the specific time in the paradigm at which the studied cortical area is crucial for
its control to be determined.
From the 1Service de Neurologie 1 (Assistance Publique-Hôpitaux
de Paris) and Institut National de la Santé et de la Recherche Médicale 289, Hôpital de la Salpêtrière, Paris, France; 2Eye Movement
Research Laboratory and Department of Neurology, Inselspital, Bern,
Switzerland; and 3Klinik für Neurologie, Charité, Berlin, Germany.
Published online Jun 18, 2002, in Wiley InterScience
( DOI: 10.1002/ana.10273
Received Aug 27, 2001, and in revised form Apr 2, 2002. Accepted
for publication Apr 10, 2002.
© 2002 Wiley-Liss, Inc.
Address correspondence to Prof Pierrot-Deseilligny, Clinique Paul
Castaigne, Hôpital de la Salpêtrière, 47 Bd de l’Hôpital, 75651
Paris Cedex 13, France. E-mail:
Several stimuli and motor responses may be used to
study spatial memory, but the simpler the stimuli and
movements, the easier it is to study them and interpret
the results. This is why simple visual stimuli and single
eye movement responses such as ocular saccades frequently have been used as models in animals and humans to study this function. The paradigm generally
used is that of memory-guided saccades (MGSs; Fig
1A): the subject, initially fixating a central point, is instructed to memorize the location of a flashed target
occurring in the peripheral visual field, without looking
at it; then, after a variable delay, the central fixation
point is switched off, which is the “go” signal for the
subject to make a voluntary eye movement (ie, an
MGS) to the remembered location of the flash. The
accuracy of this single eye movement reflects the accuracy of spatial memory, provided that perceptual and
oculomotor deficits have been ruled out. Indeed, when
an ocular motor deficit also exists, which is the case in
some patients, several saccades rather than a single saccade may be made to reach the final eye position,
which then gives a better evaluation of spatial memory
than the amplitude of the initial (first) saccade.
In the last 15 years, this paradigm often has been
used to study short-term spatial memory with delays of
a few seconds, but results with longer delays only recently have been reported. The aim of this article is to
review briefly the various findings that have been obtained with MGSs using the aforementioned four complementary methods, to synthesize our current understanding of the cortical control of spatial memory. The
main concept emerging from the different studies on
this visuooculomotor model is that spatial memory is
controlled by successive cortical areas, depending on
specific periods of time between a few seconds and a
few minutes. Thus, the results relating to short-term,
medium-term, and long-term spatial memory will be
reviewed, successively.
Short-term Spatial Memory
Short-term spatial memory is the working memory
used for current, ongoing behavior.3,4 The delays used
in animal or human experiments to study this memory
system usually comprise between 1 and 6 seconds,
based on the assumption that the upper limit in this
system is greater than such delays.
Cortical Structures Involved
In monkeys, single-neuron recordings have shown that
the dorsolateral prefrontal cortex (DLPFC) is involved
in the control of short-term spatial memory used in
MGS paradigms.4 –7 In humans, lesion studies have
suggested that several cortical areas, including the
DLPFC, are involved in MGSs.8 –11 These results have
been corroborated by functional imagery, showing that
a large network of frontoparietal areas is, in fact, active
during such a paradigm.12–14 However, note that the
MGS paradigm comprises three successive phases, involving different types of physiological mechanisms
(see Fig 1A): (1) a first phase of perception, during
which the visual stimulus (a peripheral flashed target) is
presented, involving both the visual (occipital) and attentional (parietal) areas; (2) a second phase, related to
memorization (during the delay), starting after the visual stimulus presentation and under the control of the
cortical area actually involved in spatial memory; in
fact, the beginning of this second phase corresponds to
visuospatial integration, a posterior parietal process occurring just after the visual stimulus presentation and
allowing the subject to know further the memorized
position of the stimulus in relation not only to the eyes
but probably also to the body; and (3) after the go
signal, the final phase of movement, during which the
MGS is triggered by the frontal and parietal motor areas and accuracy of spatial memory, reflected by that of
the saccade, is measured. Therefore, during the last 10
years, we first studied spatial memory using electrooculographic recordings of MGSs in two main series of
patients with selective damage to several cortical areas.8,10 The DLPFC was studied because in monkeys
this area appears to be essential for the control of shortterm spatial memory.4 –7 In humans, the DLPFC could
be located in area 46 and the adjacent area 9 of Brodmann. 5,8,13,15,16 The posterior parietal cortex (PPC)
was studied because in monkeys this area is involved in
visuospatial integration.17,18 Furthermore, it has been
suggested that the PPC plays a role in short-term spatial memorization itself and also in saccade triggering.11,17 In humans, this area could be located, as in
the monkey, in the intraparietal sulcus, a location that
recently was confirmed using functional imagery.19,20
The frontal eye field (FEF) was studied because this
area is involved in the triggering of all intentional saccades, including MGSs.8 –11,21–23 The human FEF is
located in the precentral gyrus and sulcus, mainly
around the intersection with the superior frontal sulcus.24,25 Last, the supplementary eye field (SEF)26 also
was studied because functional imaging studies have
shown that this area is activated in all saccade paradigms.12,13,25 The SEF is located in humans in the upper part of the paracentral sulcus, just anteriorly to the
supplementary motor area.25,27 The results of lesion
studies showed that the accuracy of MGS was impaired
after lesions affecting either the DLPFC, the PPC, or
the FEF, but not after lesions of the SEF.8 –10,28,29
This underlines that, although functional imagery has
shown that the SEF is activated during MGSs,12–14
this area is not actually crucial to the execution of these
single saccades, a result that also has been confirmed in
monkeys.30 In contrast, lesion studies have shown that
the role of the SEF could be essential for learning and
performing motor programs using either several types
Pierrot-Deseilligny et al: Spatial Memory
Fig 1. Short-term spatial memory. (A) Memory-guided saccade paradigm: the peripheral flashed target (visual stimulus) to memorize
is first presented for 50 milliseconds while the subject is fixating a central fixation point (perception); after a variable delay (memorization), the central fixation point is switched off, which is the go signal for the subject to perform the memory-guided saccade in
complete darkness to the remembered position of the peripheral flashed target (movement); accuracy of this saccade (or of the final
eye position if there are several saccades) is a reflection of spatial memory; the error is the correction made, if necessary, when the
target is presented again. M ⫽ midline; R ⫽ right. (B) Results of a transcranial magnetic stimulation (TMS) study made on the
right hemisphere in eight normal subjects performing memory-guided saccades (with a 2-second delay).31 Results of TMS are compared with those without stimulation; note that (1) after posterior parietal cortex (PPC) stimulation, saccade accuracy was significantly impaired when stimulation was performed 260 milliseconds after the flashed peripheral target (p ⬍ 0.01), (2) after dorsolateral prefrontal cortex (DLPFC) stimulation, saccade accuracy was impaired when stimulation was performed during the
memorization period (ie, beyond 500 milliseconds; p ⬍ 0.001), and (3) after the go signal (2 seconds), accuracy was normal in
both cases.
of body movements comprising saccades or a sequence
of a several successive saccades.10,11,28 Finally, although
lesion studies in humans showed that the DLPFC, the
PPC, and the FEF are indeed essential for the correct
performance of MGSs, such studies are no more able
Annals of Neurology
Vol 52
No 1
July 2002
than functional imaging studies to tell us how the control of these different cortical areas is chronologically
organized and which of them is more particularly involved in short-term spatial memory. Another method
therefore clearly was needed to answer such questions.
Chronology of Cortical Involvement
Initially developed to evoke motor potentials in the
limbs from stimulation over the primary motor cortex,
TMS also has been used in research to study other cortical areas and nonmotor functions. Resulting in an activation, in particular, when using repetitive stimulation, or an inactivation or inhibition, in particular,
when using single-pulse stimulation, of the stimulated
cortical area lasting only a few milliseconds, which is
similar to a brief functional lesion, TMS can be used to
determine at which specific time (or times) a paradigm
is critically controlled by this area (or even the network
directly related to this area). We used single-pulse
TMS, which was applied over the right PPC or the
right DLPFC in normal subjects at different times of
an MGS paradigm.31 Saccade accuracy was impaired
for PPC stimulation performed 260 milliseconds after
the onset of the visual stimulus, but not later in the
paradigm (see Fig 1B); in contrast, saccade accuracy
was impaired for DLPFC stimulation performed during the memorization period (approximately 1 second),
but not earlier in the paradigm. More recently, we
showed that MGS accuracy also was impaired after left
DLPFC stimulation performed during the memorization period but was normal after left PPC stimulation.32 Repetitive TMS over the right or left DLPFC
resulted in analogous results.33 Moreover, TMS studies
have confirmed that the FEF plays a major role in the
triggering of MGSs by modulating saccade latency34
and have shown that the PPC also may be involved at
this triggering stage.31,32 Overall, these different results
suggest that (1) the right PPC is involved in the control of saccade accuracy during the brief initial visuospatial integration stage, immediately after the stimulus
presentation, but not later during the memorization
phase of the paradigm, and is not, therefore, crucial in
the control of short-term spatial memory; this conclusion is now also supported by experimental results35;
(2) the left PPC is either uninvolved or less involved in
the control of saccade accuracy, a result that probably
is related to the well-known dominance in humans of
the right cerebral hemisphere for visuospatial functions;
(3) in contrast, both the right and left DLPFC control
saccade accuracy during the memorization phase of this
paradigm; (4) last, the FEF is mainly involved in MGS
triggering, to which the PPC also may contribute.
Thus, these TMS studies succeeded in determining the
chronology of the cortical control of MGSs, confirming that short-term spatial memory involved in these
saccades is controlled by the DLPFC, after a parietal
visuospatial integration and before a triggering by the
motor areas (ie, mainly the FEF; see Fig 3B). In the
monkey, the DLPFC is reciprocally connected with
both the PPC and the FEF,5,36 with therefore the existence of an anatomical substrate for such a cortical
Note that, according to the results of lesion studies
in humans29,37 or to electrophysiological data in the
monkey,38 the FEF also may contribute to the control
of short-term memorization of visual targets required
for forthcoming ocular saccades. However, the FEF
does not appear to be crucial in the memorization of
visual targets serving to guide other body movements
such as arm movements.37 Overall, these different results suggest that the FEF, in association with the
DLPFC, significantly contributes to the memorization
of visual targets used to guide saccades, whereas other
body movements could mainly depend on the DLPFC
for their memorized guidance. Last, although the subcortical pathways are beyond the scope of this review
dealing with the cortical control of spatial memory, it
should be added that MGSs made with short delays are
also under the control of different parts of the basal
Medium-term Spatial Memory
After various lesion and TMS studies on short-term
spatial memory, further questions that then needed to
be resolved using the same MGS paradigm were to determine (1) how long the transient DLPFC control of
spatial memory lasts, and (2) what cerebral structures
take over the control of memorized information after
the DLPFC.
Evidence for a Medium-term Spatial Memory
Accuracy of MGSs has been shown to decrease during
the first seconds of the memory delay.40 This is mainly
reflected in increased variable targeting errors, that is,
scatter of MGS end points around the to-beremembered target position and probably represents information loss in short-term spatial memory.40,41 In a
recent study, we used this delay dependency of targeting errors to investigate possible temporal limits for
short-term spatial memory.42 Normal subjects performed an MGS task with unpredictable varied memory delays of 0.5 to 30 seconds. Accuracy of MGSs as
a function of delay length was measured, because it was
reasoned that any transition from short-term spatial
memory to a more stable form of memory should lead
to a stabilization of the accuracy of MGSs.42 The results confirmed that variable targeting errors initially
increased linearly with delay length (Fig 2A). Surprisingly, after maximum errors at delays of approximately
20 seconds (⫹5 seconds), this process did not stabilize
but reverted partially, and accuracy of MGSs improved
significantly with longer delays. This time-limited decay of spatial information strongly suggests involvement of a more stable memory system at longer delays
and thus allows for the hypothesis of an upper temporal limit of approximately 20 seconds for short-term
spatial memory in humans. This temporal limit fits remarkably well with the limited temporal stability of
Pierrot-Deseilligny et al: Spatial Memory
Fig 2. Medium-term spatial memory. (A) Delay dependency of variable targeting errors of memory-guided saccades (MGSs).42 Left
panel shows exemplary results from a single subject. Accuracy of MGSs is expressed as gain (gain ⫽ 1: precise saccade; gain ⬍1: hypometria; gain ⱖ1: hypermetria). Note that in this subject, after an initial progressive increase in gain variability up to a maximum
at a 15-second delay, there was a sudden and surprising decrease in gain variability at a 20-second delay. Right figure shows group
results from 16 subjects: note that there was a linear increase in variable targeting errors of MGSs in the group up to a maximum of
an approximately 20-second delay (95% confidence interval, 15–25 seconds); at longer delays, variable errors decreased, that is, improved, significantly. (single asterisk) p ⬍ 0.05; (double asterisk) p ⬍ 0.01 significant difference to 0.5-second delay; (dagger) p ⬍
0.05 significant difference to a 20-second delay. Data are adapted from Ploner and colleagues.42 (B) Inaccuracy of MGSs in patients
with lesions of the right hippocampal formation (HF; ie, hippocampus and entorhinal cortex) in patients with lesions of the right HF
and perirhinal cortex (HF⫹PRC), in patients with lesions of the right HF, perirhinal cortex, and parahippocampal cortex
(HF⫹PRC⫹PHC), and in controls. Note that in patients with involvement of the parahippocampal cortex, there was significant inaccuracy of MGSs at delays longer than 20 seconds contralateral to the lesion side. (asterisks) p ⬍ 0.01. Modified from Ploner and colleagues51 (HF patients) and Ploner and colleagues52 (controls, HF⫹PRC patients, HF⫹PRC⫹PHC patients). (C) Anatomy of the
medial temporal lobe. Three coronal brain sections perpendicular to the Sylvian fissure are arranged from rostral to caudal. The distance from the temporal pole is shown on the top left of each section (in millimeters). Arrows mark the approximate margins of the
cytoarchitectonic subregions of the medial temporal lobe. (asterisk) Collateral sulcus. EC ⫽ entorhinal cortex; GP ⫽ globus pallidus;
H ⫽ hippocampus; LGN ⫽ lateral geniculate nucleus; NA ⫽ amygdala; NC ⫽ caudate nucleus; PHC ⫽ parahippocampal cortex;
PRC ⫽ perirhinal cortex; PU ⫽ putamen; TH ⫽ thalamus; VI ⫽ lateral ventricle; VIII ⫽ third ventricle.
Annals of Neurology
Vol 52
No 1
July 2002
spatially selective neuronal activity in the monkey
DLPFC and the known temporal properties of shortterm plasticity in prefrontal neurons.4,41 Moreover, a
similar relationship between delay length and memory
performance is a well-known phenomenon in animal
models of memory consolidation.43 In these models,
memory formation proceeds successively through temporally distinct phases, with each phase representing a
distinct and independent memory system. The transition between successive memory systems is marked by
brief and transient decreases in memory performance
(“dips”), with subsequent memory improvement.43
Here, the observed improvement of spatial memory at
delays longer than 20 seconds provides evidence for the
existence of a distinct, more stable medium-term spatial memory, becoming behaviorally relevant for such
delays and storing spatial information largely independent of short-term spatial memory.42 This new finding
then led us to determine which cerebral structure could
be involved in such medium-term spatial memory.
Role of the Parahippocampal Cortex
Previous lesion studies in humans have suggested that
the medial temporal lobe (MTL) is the region containing the neuronal substrate of the putative mediumterm memory. Both in the famous case of Patient
H.M. and in a subsequent large series of patients with
unilateral MTL removals, visual and spatial memory
was deficient at delays of 24 seconds but was unimpaired at delays of 16 and 12 seconds, respectively.44,45
However, the MTL is a complex anatomical region,
consisting of the HF (ie, hippocampus and entorhinal
cortex), the perirhinal cortex (PRC), and the parahippocampal cortex (PHC).46 In the human brain, the
PRC and PHC are laterally adjacent to the HF and
line the collateral sulcus, the PRC rostrally, and the
PHC caudally (see Fig 2C). The HF is directly connected with both cortices, but it is mainly the PHC
which is reciprocally connected with neocortical areas
involved in spatial cognition, that is, the PPC and
DLPFC.47,48 A special role of the PHC for spatial cognition is further supported by the fact that patients
with lesions of the PHC are disorientated when navigating in previously unfamiliar environments.49 However, it has proven difficult to define the precise cognitive deficit underlying this syndrome.49 Whereas
lesion studies in monkeys have clearly shown that the
PRC and PHC carry visual memory functions independent of the HF,50 possible spatial memory functions of the PRC and PHC only rarely have been investigated. Thus, in two recent studies, we investigated
the question of possible differential contributions of
MTL subregions to spatial memory in humans with
postsurgical lesions of the right MTL.51,52 Patients
with selective lesions of the HF, the HF plus the PRC,
and the HF plus the PRC and PHC performed the
same MGS task with unpredictable varied memory delays of up to 30 seconds. The results showed that only
patients with PHC involvement showed a significant
inaccuracy of MGSs at delays longer than 20 seconds
contralateral to the lesion side. The other patient
groups showed no deficit51,52 (see Fig 2B). MGSs with
short delays (ie, ⬍20 seconds) were normal, a result
confirming the findings of our previous study of patients with large lesions of the right MTL.53 Although
it is difficult to definitely rule out a cumulative effect
of lesion to MTL subregions, further analysis showed
that rostrocaudal lesion extent and lesion volume per se
did not explain the significant spatial memory deficit
with PHC lesions.51,52 Therefore, these results suggest
that spatial memory functions within the MTL are not
restricted to the HF. Furthermore, they show that the
right PHC contributes to spatial memory at delays beyond DLPFC-based short-term spatial memory. The
PHC, but not the HF or the PRC, therefore is a likely
neuronal substrate for medium-term spatial memory of
single visual items in humans. This cognitive function
is consistent with the anatomical position of the PHC
between the DLPFC and the HF and also may contribute to the aforementioned navigational capabilities.
It remains, however, to be determined whether the left
PHC in humans plays a role analogous to that of the
right PHC.
Long-term Spatial Memory
Two of the other major questions in this research field
are to determine (1) which structure is involved for
longer delays, that is, for long-term spatial memorization, and (2) when the role of this structure becomes
Role of the Hippocampal Formation
There is ample evidence, from both single-neuron recordings and lesion studies in monkeys, that the HF is
involved in spatial memory processes.54 Apart from
neurons encoding space relative to the actual body coordinates, many neurons in the primate hippocampus
appear to represent space largely independent of the
actual body coordinates and also in association to nonspatial information.54 Indeed, the HF is situated, by
virtue of its extensive direct and indirect connections
with areas involved in spatial processing and object recognition,46,48 at the highest level of integration of perceptual information and thus is in a privileged position
to bind together information from different sensory
modalities and spatial coordinate frames into a spatial
memory map of the environment. Therefore, it is
probable that spatial memory deficits with hippocampal lesions depend on the number of associations between to-be-remembered items in a given spatial memory task, that is, factors that are not assessed by
standard MGS paradigms. However, this notion does
Pierrot-Deseilligny et al: Spatial Memory
not exclude the possibility that the HF is also involved
in memory of single spatial items.
memorization. Last, three major questions are now
raised about this model.
Time of Long-term Memorization
Previous studies in patients with autoptically verified
lesions restricted to the HF have shown that memory
for locations of a simple spatial array of objects is impaired with delays of 5 minutes.55,56 Thus, it appears
that at delays of some minutes the hippocampus is actively involved in spatial memory, at least in tasks
where no or only a few associations between spatial
items have to be remembered. This interpretation
would parallel the known delay-dependent memory
deficits for simple object information seen in monkeys
with lesions to the HF, in which memory is normal at
delays of 60 seconds, but impaired at delays of 10 minutes.57,58 Thus, the HF is a likely candidate region for
a long-term spatial memory system, subserving spatial
memory formation beyond PHC-based medium-term
spatial memory.
Selection of Information
First, how is salient visuospatial information selected
and how is such memorized information retrieved? At
the perceptive stage, we recently have shown that selection is influenced by environmental and behavioral
conditions60; this selection is, of course, mainly under
the control of the parietal attentional processes,18 with
a rapid transmission of such selected information in the
DLPFC for short-term memorization.61 Then, the anterior cingulate cortex could play an active role in
memorization. This area is involved in motivation,
which is a cognitive process preparing intentional
movements, including MGSs according to one of our
recent lesion studies.62 Furthermore, the anterior cingulate cortex is anatomically and functionally located
between the memory areas and motor areas, receiving
afferences from the DLPFC and the medial temporal
region and projecting to the primary motor cortex and
the FEF.62,63 Therefore, the ACC could contribute to
mediating toward the motor areas salient memorized
spatial information stored in the different frontal and
temporal areas controlling spatial memory. This represents an interesting research field and currently is under investigation.
Conclusions and Future Developments
The results of these different studies allow us to suggest
that, after an initial visuospatial integration stage under
the control mainly of the right PPC (up to approximately 300 milliseconds), spatial memory for single visual items is controlled, successively, by the DLPFC for
its short-term component (delays of up to approximately 20 seconds), the PHC for its medium-term
component (ie, delays longer than 20 seconds and up
to a few minutes), and, last, the HF for its long-term
component (after a few minutes; Fig 3A). However, it
must be acknowledged that the temporal properties of
this visuooculomotor model also may depend on the
strength of memory encoding.59 It thus is possible that
the abovementioned transitional delays between memory systems may change with stimulus duration, stimulus salience, or the degree of interference with concurrent information. In our experiments, we did not
systematically investigate these possibilities. Moreover,
although there is no doubt about the role of the HF in
the control of long-term spatial memory in general, it
remains to be confirmed by further studies that (1)
long-term spatial memory involved in the MGS paradigms used here is also controlled by this structure, and
(2) the critical transitional time between the mediumterm and long-term memorized components indeed
does occur, at least for single spatial items, after a delay
of approximately a few minutes as already has been
shown for other memory modalities. However, some
aspects of the cortical control of spatial memory defined in this visuooculomotor model are original findings, in particular, the new concept of an intermediary
link between short-term and long-term memorizations
and the involvement of the PHC in such medium-term
Annals of Neurology
Vol 52
No 1
July 2002
Interactions between Areas
Second, in this visuooculomotor model, how is memorized information coordinated between the different
structures successively involved in spatial memory?
Theoretically, there are two main ways in which this
control could be exerted (see Fig 3B): (1) it might be
organized in parallel, with each structure receiving the
starting integrated visuospatial signals from the PPC
and progressively constructing its own memorized information, which could become mature at a specific
critical time; then, this structure could control its stabilized memorized information for a limited moment,
before rapidly losing it at another critical time, whereupon it could be replaced by the structure involved in
the following specific period of time; (2) control also
might be serially organized, with each structure being
completely dependent on memorized information held
and transmitted at a critical time by the preceding
structure, before directly transmitting such information
at another, later critical time to the following structure;
and 3) control might be mixed, combining these two
mechanisms. The spontaneous improvement of certain
memory abilities after a delay of 20 seconds in normal
subjects already suggests that a part of the control is
exerted in parallel (see above). However, other studies
are required so as to determine whether a part of the
control of spatial memory is also serially organized.
Fig 3. Cortical control of spatial memory. (A) Schematic representation of spatial memory accuracy depending on time: during the
first 300 milliseconds, the (posterior) parietal cortex controls visuospatial integration; then, short-term spatial memory (up to 15–20
seconds), medium-term spatial memory (probably up to a few minutes), and long-term spatial memory (after a few minutes) are
controlled, successively, by the (dorsolateral) prefrontal cortex, the parahippocampal cortex, and the hippocampal formation, respectively, with therefore a different structure crucially involved for each of these four periods of time. (B) Hypothetical representation on
the cortex of the areas involved in spatial memory control: after an initial spatial integration of multimodal sensory input (1) in the
PPC, spatial memorized information is transmitted in parallel (solid arrows) and/or serially (dotted arrows) to the DLPFC (shortterm spatial memory), the PHC (medium-term spatial memory), and the HF (long-term spatial memory) before multimodal motor
output (2), with triggering performed in the FEF for a memory-guided saccade or the PMC for other body movements. Note that
the arrows represent either well-known direct tracts or possible (?) direct tracts (see Barbas and Mesulam66 for the projection of the
PHC to the FEF); cs ⫽ central sulcus; DLPFC ⫽ dorsolateral prefrontal cortex; FEF ⫽ frontal eye field; HF ⫽ hippocampal formation (entorhinal cortex ⫹ hippocampus); ips ⫽ intraparietal sulcus; PHC ⫽ parahippocampal cortex; pcs ⫽ precentral sulcus;
PMC ⫽primary motor cortex; PPC ⫽ posterior parietal cortex.
Generalization of the Model
Third, is this simple visuooculomotor model of spatial memory applicable to the memorization of other
types of stimuli (ie, more complex visual, auditory,
and somesthetic stimuli) and to the rest of body motoricity? It is already known that the same regions,
the PPC for spatial integration, and then at least the
DLPFC and HF for spatial memory, are involved for
Pierrot-Deseilligny et al: Spatial Memory
other types of stimuli and in the preparation of other
movements.44,64,65 This means that, when stimuli
other than single visual items occur and when body
movements other than ocular saccades are performed,
different areas or cells are active within the same regions, and finally the primary motor cortex replaces
the FEF for triggering body movements (see Fig 3B).
However, the existence of a similar spatial memory
organization for these other stimuli and movements,
in which the PHC also could play a major role, has
yet to be demonstrated using appropriate paradigms
and methods. Therefore, further studies are required
to determine more precisely the organization of the
cortical control of spatial memory, our knowledge of
which has, however, markedly progressed in recent
years thanks to investigations on this visuooculomotor
1. Squire LR, Kandel ER. Memory. New York: Freeman, 1999.
2. McGaugh JL. Memory—a century of consolidation. Science
2000;287:248 –251.
3. Baddeley A. The fractionation of working memory. Proc Natl
Acad Sci USA 1996;93:13468 –13472.
4. Goldman-Rakic PS. Regional and cellular fractionation of
working memory. Proc Natl Acad Sci USA 1996;93:
5. Goldman-Rakic PS. Circuitry of primate prefrontal cortex and
regulation of behavior by representational memory. In: Mountcastle VB, ed. Handbook of physiology. Section 1. Vol 5. Part
1. Bethesda: American Physiological Society, 1987:373– 417.
6. Funahashi S, Bruce CJ, Godman-Rakic PS. Mnemonic coding
of space in the monkey’s dorsolateral prefrontal cortex. J Neurophysiol 1989;61:331–349.
7. Sawaguchi T, Iba M. Prefrontal cortical representation of
visuospatial working memory in monkeys examined by local
inactivation with muscimol. J Neurophysiol 2001;86:
8. Pierrot-Deseilligny C, Rivaud S, Gaymard B, Agid Y. Cortical
control of memory-guided saccades in man. Exp Brain Res
1991;83:607– 617.
9. Rivaud S, Müri RM, Gaymard B, et al. Eye movement disorders after frontal eye field lesions in humans. Exp Brain Res
1994;102:110 –120.
10. Israël I, Rivaud S, Gaymard B, et al. Cortical control of
vestibular-guided saccades in man. Brain 1995;118:1169 –1183.
11. Pierrot-Deseilligny C, Rivaud S, Gaymard B, et al. Cortical
control of saccades. Ann Neurol 1995;37:557–567.
12. O’Sullivan EP, Jenkins IH, Henderson L, et al. The functional
anatomy of remembered saccades: a PET study. Neuroreport
13. Sweeney JA, Mintun MA, Kwee M, et al. Positron emission
tomography study of voluntary saccadic eye movements and
spatial working memory. J Neurophysiol 1996;75:454 – 468.
14. Heide W, Binkofski F, Seitz RJ, et al. Activation of frontoparietal cortices during memorized triple-step sequences of saccadic eye movements: an fMRI study. Eur J Neurosci 2001;13:
15. Rajkowska G, Goldman-Rakic PS. Cytoarchitectonic definition
of prefrontal areas in the normal human cortex. II. Variability
in locations of area 9 and 46 and relationship to the Talairach
coordinate system. Cereb Cortex 1995;5:323–337.
Annals of Neurology
Vol 52
No 1
July 2002
16. Petrides M, Pandya DN. Dorsolateral prefrontal cortex: comparative cytoarchitectonic analysis in the human and the macaque brain and corticocortical connection patterns. Eur J Neurosci 1999;11:1011–1036.
17. Andersen RA. Visual and eye movement functions of the
posterior parietal cortex. Annu Rev Neurosci 1989;12:
377– 403.
18. Colby CL, Goldberg ME. Space and attention in parietal cortex. Annu Rev Neurosci 1999;22:319 –349.
19. Müri RM, Iba-Zizen MT, Derosier C, et al. Location of the
human posterior eye field with functional magnetic resonance
imaging. J Neurol Neurosurg Psychiatry 1996;6:445– 448.
20. Milea D, Lobel E, Lehericy S, et al. Functional MRI mapping
of parietal and cingular activity during voluntary saccades. Soc
Neurosci Abstr 2001:71.27.
21. Goldberg ME, Bruce CJ. Primate frontal eye fields. III. Maintenance of spatially accurate saccade signal. J Neurophysiol
1990;64:489 –508.
22. Terao Y, Fukuda H, Ugawa Y, et al. Visualization of the information flow through human oculomotor cortical regions by
transcranial magnetic stimulation. J Neurophysiol 1998;80:
936 –946.
23. Tehovnik EJ, Sommer MA, Chou I-H, et al. Eye fields in the
frontal lobes of primates. Brain Res Rev 2000;32:413– 448.
24. Paus T. Location and function of the human frontal eye field:
a selective review. Neuropsychology 1996;734:475– 483.
25. Carter N, Zee DS. The anatomical localization of saccades using functional imaging studies and transcranial magnetic stimulation. Curr Opin Neurol 1997;10:10 –17.
26. Schlag J, Schlag-Rey M. Evidence for a supplementary eye field.
J Neurophysiol 1987;57:179 –200.
27. Grosbras MH, Lobel E, Van de Moortele PF, et al. An anatomical landmark for the supplementary eye fields in human
revealed with functional magnetic resonance imaging. Cereb
Cortex 1999;9:705–711.
28. Gaymard BM, Rivaud S, Pierrot-Deseilligny C. Role of the left
and right supplementary motor areas in memory-saccade sequences. Ann Neurol 1993;3:404 – 406.
29. Ploner CJ, Rivaud-Pechoux S, Gaymard BM, et al. Errors of
memory-guided saccades in humans with lesions of the frontal
eye field and the dorsolateral prefrontal cortex. J Neurophysiol
1999;82:1086 –1090.
30. Sommer MA, Tehovnik EJ. Reversible inactivation of macaque
dorsomedial frontal cortex: effects on saccades and fixations.
Exp Brain Res 1999;124:429 – 446.
31. Müri RM, Vermersch A-I, Rivaud S, et al. Effects of singlepulse transcranial magnetic stimulation over the prefrontal and
posterior parietal cortices during memory-guided saccades in
humans. J Neurophysiol 1996;76:2102–2106.
32. Müri RM, Gaymard B, Rivaud S, et al. Hemispheric asymmetry in cortical control of memory-guided saccades. A transcranial magnetic stimulation study. Neuropsychologia 2000;38:
33. Brandt SA, Ploner CJ, Meyer BU, et al. Effects of repetitive
transcranial magnetic stimulation over dorsolateral prefrontal
and posterior parietal cortex on memory-guided saccades. Exp
Brain Res 1998;118:197–204.
34. Wipfli M, Felblinger J, Mosimann UP, et al. Double-pulse
transcranial magnetic stimulation over the frontal eye field facilitates triggering of memory-guided saccades. Eur J Neurosci
35. Constantinidis C, Steinmetz MA. Neuronal activity in posterior
parietal area 7a during the delay periods of a spatial memory
task. J Neurophysiol 1996;76:1352–1355.
36. Chafee MV, Goldman-Rakic P. Inactivation of parietal and
prefrontal cortex reveals interdependence of neural activity during memory-guided saccades. J Neurophysiol 2000;83:
1550 –1566.
37. Gaymard B, Ploner CJ, Rivaud-Pechoux S, Pierrot-Deseilligny
C. The frontal eye field is involved in spatial short-term memory but not in reflexive saccade inhibition. Exp Brain Res 1999;
129:288 –301.
38. Umeno MM, Goldberg ME. Spatial processing in the monkey
frontal eye field, II: Memory responses. J Neurophysiol 2001;
86:2344 –2352.
39. Hikosaka O, Takikawa Y, Kawagoe R. Role of the basal ganglia
in the control of purposive saccadic eye movements. Physiol
Rev 2000;80:953–978.
40. White JM, Sparks DL, Stanford TR. Saccades to remembered
target locations: an analysis of systematic and variable errors.
Vision Res 1994;34:79 –92.
41. Wang X-J. Synaptic reverberation underlying mnemonic persistent activity. Trends Neurosci 2001;24:455– 463.
42. Ploner CJ, Gaymard B, Rivaud S, et al. Temporal limits of
spatial working memory in humans. Eur J Neurosci 1998;10:
794 –797.
43. De Zazzo J, Tully T. Dissection of memory formation: from
behavioural pharmacology to molecular genetics. Trends Neurosci 1995;18:212–218.
44. Sidman M, Stoddard LT, Mohr JP. Some additional observations of immediate memory in a patient with bilateral hippocampal lesions. Neuropsychologia 1968;6:245–254.
45. Rains GD, Milner B. Right-hippocampal contralateral-hand effect in the recall of spatial location in the tactual modality.
Neuropsychologia 1994;32:1233–1242.
46. Amaral DG, Insausti R. The hippocampal formation. In: The
human nervous system. Paxinos G, ed. San Diego: Academic
Press, 1990:711–755.
47. Goldman-Rakic PS, Selemon LD, Schwartz ML. Dual pathways connecting the dorsolateral prefrontal cortex with the hippocampal formation and parahippocampal cortex in the rhesus
monkey. Neuroscience 1984;12:719 –743.
48. Suzuki WA, Amaral DG. Perirhinal and parahippocampal cortices of the macaque monkey: cortical afferents. J Comp Neurol
49. Aguirre GK, D’Esposito M. Topographical disorientation. A
synthesis and taxonomy. Brain 1999;122:1613–1628.
50. Zola-Morgan S, Squire LR, Amaral DG, Suzuki WA. Lesions
of the perirhinal and parahippocampal cortex that spare the
amygdala produce severe memory impairment. J Neurosci
1989;9:4355– 4370.
51. Ploner CJ, Gaymard BM, Ehrlé N, et al. Spatial memory deficits in patients with lesions affecting the medial temporal neocortex. Ann Neurol 1999;45:312–319.
52. Ploner CJ, Gaymard BM, Rivaud-Péchoux S, et al. Lesions affecting the parahippocampal cortex yield spatial memory deficits in humans. Cereb Cortex 2000;10:1211–1216.
53. Müri RM, Rivaud S, Timsit S, et al. The role of the right
medial temporal lobe in the control of memory-guided saccades. Exp Brain Res 1994;101:165–168.
54. Rolls ET. Spatial view cells and the representation of place in
the primate hippocampus. Hippocampus 1999;9:467– 480.
55. Cave CB, Squire LR. Equivalent impairment of spatial and
nonspatial memory following damage to the human hippocampus. Hippocampus 1991;1:329 –340.
56. Rempel-Clower NL, Zola SM, Squire LR, Amaral DG. Three
cases of enduring memory impairment after bilateral damage
limited to the hippocampal formation. J Neurosci 1996;16:
57. Alvarez P, Zola-Morgan S, Squire LR. Damage limited to the
hippocampal region produces long-lasting memory impairment
in monkeys. J Neurosci 1995;15:3796 –3807.
58. Leonard BW, Amaral DG, Squire LR, Zola-Morgan S. Transient memory impairment in monkeys with bilateral lesions of
the entorhinal cortex. J Neurosci 1995;15:5637–5659.
59. McClelland JL, McNaughton BL, O’Reilly RC. Why there are
complementary learning systems in the hippocampus and
neocortex: insights from the successes and failures of connectionist models of learning and memory. Psychol Rev 102:419 – 457.
60. Ploner CJ, Ostendorf F, Brandt SA, et al. Behavioral relevance
modulates access to spatial working memory in humans. Eur
J Neurosci 2001;13:357–363.
61. Kastner S, Ungerleider LG. Mechanisms of visual attention in
the human cortex. Annu Rev Neurosci 2000;23:315–341.
62. Gaymard B, Rivaud S, Cassarini JF, et al. Effects of anterior
cingulate cortex lesions on ocular saccades in humans. Exp
Brain Res 1998;120:173–183.
63. Morecraft RJ, Van Hoesen GW. Convergence of limbic input
to the cingulate motor cortex in the rhesus monkey. Brain Res
Bull 1998;20:209 –232.
64. Goldman-Rakic PS, Bates JF, Chafee MV. The prefrontal cortex and internally generated motor acts. Curr Opin Neurobiol
1992;2:830 – 835.
65. Culham JC, Kanwisher NG. Neuroimaging of cognitive functions in human parietal cortex. Curr Opin Neurobiol 2001;11:
66. Barbas H, Mesulam MM. Organization of afferent input to
subdivisions of area 8 in the rhesus monkey. J Comp Neurol
1981;200:407– 431.
Pierrot-Deseilligny et al: Spatial Memory
Без категории
Размер файла
202 Кб
mode, visuooculomotor, memory, cortical, human, spatial, control
Пожаловаться на содержимое документа