The role of the right medial temporal lobe in the control of memory-guided saccades

The role of the hippocampal formation in the control of memory-guided saccades is unclear. We tested two types of memory-guided saccades with short memorization delay in three patients with a lesion affecting the right medial temporal lobe and involving the hippocampal formation. In single memory-guided saccades, testing spatial working memory, the gain of the patient group did not differ from that of an age-matched control group. In contrast, in sequences of memory guided saccades, testing chronological working memory, there was a marked and significant increase in the percentage of erroneous sequences in patients, compared to controls. These results suggest an important role of the hippocampal formation in the memorization of the chronological order of saccade sequences. In contrast, this structure does not appear to be crucial for spatial working memory, used in single memory-guided saccades.     

Information processing in long delay memory-guided saccades: further insights from TMS

The performance of memory-guided saccades with two different delays (3 s and 30 s of memorisation) was studied in eight subjects. Single pulse transcranial magnetic stimulation (TMS) was applied simultaneously over the left and right dorsolateral prefrontal cortex (DLPFC) 1 s after target presentation. In both delays, stimulation significantly increased the percentage of error in amplitude of memory-guided saccades. Furthermore, the interfering effect of TMS was significantly higher in the short delay compared to that of the long delay paradigm. The results are discussed in the context of a mixed model of spatial working memory control including two components: First, serial information processing with a predominant role of the DLPFC during the early period of memorisation and, second, parallel information processing, which is independent from the DLPFC, operating during longer delays.     

Influence of transcranial magnetic stimulation on the execution of memorised sequences of saccades in man

Memorised sequences of saccades are cortically controlled by the supplementary motor area (SMA), as shown in animal experiments and in humans with isolated SMA lesions. We applied transcranial magnetic stimulation (TMS) in eight healthy subjects executing memorised sequences of saccades. Sequences of three targets were presented. Then, upon a go-signal, the subjects had to execute the appropriate sequences. Ten to fifteen sequences were performed in each experiment, and the number of errors were counted. The number of errors increased significantly if TMS was given 80 ms before or 60 ms after the go-signal, with the stimulation coil overlying the SMA. There was no significant increase in errors if different stimulation intervals were chosen (160ms and 120ms before the go-signal; 100 ms, 140 ms or 240 ms after the go-signal), if the coil was positioned inappropriately (e.g. over the occipital cortex), or if the stimulator output was too low. We conclude that TMS can interfere specifically with the function of the SMA during a critical time interval close to the go-signal.     

Effects of repetitive transcranial magnetic stimulation over dorsolateral prefrontal and posterior parietal cortex on memory-guided saccades

We investigated the role of the dorsolateral prefrontal cortex (DLPFC) and the posterior parietal cortex (PPC) in a visuospatial delayed-response task in humans. Repetitive transcranial magnetic stimulation (20 Hz, 0.5 s) was used to interfere temporarily with cortical activity in the DLPFC and PPC during the delay period. Omnidirectional memory-guided saccades with a 3-s delay were used as a quantifiable motor response to a visuospatial cue. The question addressed was whether repetitive transcranial magnetic stimulation (rTMS) over the DLPFC or PPC during the sensory of memory phase affects accuracy of memory-guided saccades. Stimulation over the primary motor cortex served as control. Stimulation over the DLPFC significantly impaired accuracy of memory-guided saccades in amplitude and direction. Stimulation over the PPC impaired accuracy of memory-guided saccades only when applied within the sensory phase (50 ms after cue offset), but not during the memory phase (500 ms after cue offset). These results provide further evidence for a parieto-frontal network controlling performance of visuospatial delayed-response tasks in humans. It can be concluded that within this network the DLPFC is mainly concerned with the mnemonic respresentation and the PPC with the sensory representation of spatially defined perceptual information. Received: 22 April 1996/Accepted: 16 June 1997     

Impairment of extraretinal eye position signals after central thalamic lesions in humans

Accuracy of four different types of memory-guided saccades was studied in two patients with a small central thalamic lesion, probably involving the region of the internal medullary lamina (IML), and in a control group. In the first paradigm, the eyes and head remained immobile between the time of the presentation of the visual target to be remembered and the memory-guided saccade. In the other three paradigms, the eyes were displaced during the same period (before the memory-guided saccade) by either visually-guided saccades, a smooth pursuit eye movement or a body movement (with vestibulo-ocular reflex suppression). Therefore, in these three paradigms, the initial eye displacement required the use of extraretinal eye position to produce accurate memory-guided saccades. Compared with the control group, the two patients had normal accuracy in the first memory-guided saccade paradigm, in which there was no initial eye displacement, but markedly impaired saccade accuracy in the other three paradigms. These results suggest that the cortical areas triggering saccades did not receive correct extraretinal eye position signals. They are consistent with an impairment of the efference copy, which could be distributed to the cortical ocular motor areas by the IML.     

Disconjugate memory-guided saccades to disparate targets: temporal aspects

Memory-guided saccades to disparate targets (i.e., more eccentric for one eye) flashed 1 s earlier become disconjugate (i.e., of different amplitude for the two eyes) after only about 30 trials. After about 225 trials the disconjugacy persists even when the target to remember is no longer disparate. This suggests fast learning based on short-term memorization of disparity. Learning, however, fails to occur if during the training the memory delay for each trial is increased to 2 s. The purpose of the present study was to test the importance of the frequency of stimulus presentation and thereby the rate of saccades. The same memory-guided saccade paradigm was used as in the prior study and a short training period of 225 trials was applied. For each training trial, the memory delay was again 2 s, but the time allocated for fixation of the central dot and the time allocated for fixation of the remembered target in the dark was reduced to increase the frequency of saccades made. Saccades became rapidly disconjugate and their disconjugacy was retained in a subsequent neutral condition using non-disparate targets. These findings indicate that stimulus frequency and thereby saccade frequency is important for disconjugate oculomotor learning based on disparity memorization. Nevertheless, additional experiments using longer memory delays of 3 s or 4 s show a definite failure of memorization and disconjugate learning.     

Contributions from the left PMd and the SMA during sequence retrieval as determined by depth of training

The dorsal premotor cortex (PMd) and the supplementary motor area (SMA) are critical for the acquisition and expression of sequential behavior, but little is known regarding how these regions are recruited when we must simultaneously acquire multiple sequences under different amounts of training. We hypothesized that these regions contribute to the retrieval of sequences at different familiarity levels, with the left PMd supporting sequences of moderate familiarity and the SMA supporting sequences of greater familiarity. Double-pulse transcranial magnetic stimulation (TMS) was applied during the retrieval of six sequences previously learned under three different amounts of exposure during 30 days of training using a discrete sequence production task. TMS led to a significant interaction of sequence error between depth of training and stimulation location. Stimulation of the left PMd increased error during moderate sequence retrieval, whereas stimulation of the SMA increased error during the retrieval of both moderately and extensively trained sequences. The lack of a double dissociation fails to support a direct correspondence between brain region and putative behavioral learning stage. Instead, the interaction suggests that SMA and PMd support the expression of sequences over different, albeit overlapping, time scales. Separate analysis of sequence initiation time did not demonstrate any significant difference between moderately and extensively trained sequences. Instead, stimulation to either region quickened sequence initiation for these sequences, but not for those sequences with poor retrieval performance. This supports the general role of these premotor regions in the maintenance of specific sequence knowledge prior to movement onset.     

Motor execution is necessary to memorize disparity

Binocular saccades in response to briefly flashed, memorized disparate targets (different for the two eyes) become disconjugate following repeated trials. After 15 min of such training, the disconjugacy persists, even when the target to memorize is no longer disparate. This study examines the hypothesis that disparity memorization has a motor basis. We report here three experiments in which subjects were trained for 15-min periods. In experiment 1, subjects made no saccade after target presentation (static training); in experiment 2 subjects intended to make a saccade, but they actually made a saccade in only 10% or 20% of the trials; in experiment 3 subjects made anti-saccades. For all three experiments, the flashed target was disparate and the memory delay for each trial was 1 s. To examine the effects of learning for all three experiments, before and after training, we recorded memory-guided saccades to non-disparate targets (monocular viewing). Experiments 1 and 2 produced inconsistent (before/after training) changes in the disconjugacy of saccades. Thus, the disparity of potential saccade targets had no lasting effect on the disconjugacy of saccades if a saccade was not made. In contrast, the anti-saccades in experiment 3 developed a disconjugacy opposite to the disparity of the remembered target. These findings indicate that the execution of the saccade is necessary to memorize disparity of the target.     

Context-dependent effects of substantia nigra stimulation on eye movements

In a series of now classic experiments, an output structure of the basal ganglia (BG)--the substantia nigra pars reticulata (SNr)--was shown to be involved in the generation of saccades made in particular behavioral contexts, such as when memory was required for guidance. Recent electrophysiological experiments, however, call this original hypothesis into question. Here we test the hypothesis that the SNr is involved preferentially in nonvisually guided saccades using electrical stimulation. Monkeys performed visually guided and memory-guided saccades to locations throughout the visual field. On 50% of the trials, electrical stimulation of the SNr occurred. Stimulation of the SNr altered the direction, amplitude, latency, and probability of saccades. Visually guided saccades tended to be rotated toward the field contralateral to the side of stimulation, whereas memory-guided saccades tended to be rotated toward the hemifield ipsilateral to the side of stimulation. Overall, the changes in saccade vector direction were larger for memory-guided than for visually guided saccades. Both memory- and visually guided saccades were hypometric during stimulation trials, but the stimulation preferentially affected the length of memory-guided saccades. Electrical stimulation of the SNr produced decreases in visually guided saccades bilaterally. In contrast, memory-guided saccades often had increases in saccade latency bilaterally. Finally, we found approximately 10% reduction in the probability of memory-guided saccades bilaterally. Visually guided saccade probability was unaltered. Taken together the results are consistent with the hypothesis that SNr primarily influences nonvisually guided saccades. The pattern of stimulation effects suggests that SNr influence is widespread, altering the pattern of activity bilaterally across the superior colliculus map of saccades.     

Memory-guided saccades: what is memorized?

Summary In order to find out whether extraretinal (oculomotor, internal) input suffices to provide the oculomotor system with the information necessary for saccadic control, two subjects were asked to make memoryguided saccades in complete darkness, after three different location acquisition conditions. These conditions were visually-guided saccades (SA), providing retinal (external) and extraretinal input, visual peripheral target presentation during central target fixation (FI) (external input only), and smooth pursuit (PU) (internal input only). Either 2 or 12 s (delay) after locating the target, the subjects had to make a memory-guided saccade toward it in complete darkness. The results show that whereas these memory-guided saccades were quite accurate for trials with preceding external input, this was not the case with acquisition through internal input alone. Moreover, the accuracy of memory-guided saccades decreased when the delay increased from 2 to 12s for both conditions with retinal input, whereas the accuracy increased for the one condition without retinal input, i.e., the smooth pursuit location acquisition. Furthermore, when both retinal and oculomotor inputs were provided, better accuracy of the memory-guided saccades was observed than with single input.     

Cortical control of memory-guided saccades in man

Summary Memory-guided saccades were electro-oculographically recorded in 30 patients with limited unilateral cerebral infarction, documented by computerized tomographic scan and/or magnetic resonance imaging. The lesions affected either (1) the posterior parietal cortex (PPC), (2) the dorsolateral frontal cortex (DLFC), involving the frontal eye field (FEF) and/or the prefrontal cortex (PFC) (area 46 of Brodmann), or (3) the supplementary motor area in the dorsomedial frontal cortex (DMFC). Patients were divided into 6 groups according to the location (PPC, DLFC, DMFC) and side of the lesions. Both latency and accuracy (expressed as a percentage of error in amplitude) of memory-guided saccades were compared in each group of patients to values obtained from 20 age-matched normal subjects. Latency was significantly increased, for both directions of saccades in the two DLFC groups and in the right PPC group, and for leftward saccades in the left PPC group. The percentage of error in amplitude was also significantly increased for both directions of saccades in the right PPC group and the left DLFC group, and for leftward saccades in the right DLFC group. Results were near the normal values in patients with lesions affecting the DMFC. Thus, both the PPC (essentially on the right side) and the DLFC appear to play a role in the control of memory-guided saccades. It is suggested that the cortical pathway involved in these saccades includes the PPC, the PFC and the FEF, successively. The PPC could have a dual role: visuospatial integration, and early selection and preparation of certain collicular cells by pre-excitation. Both functions could be ensured by two different types of cells, corresponding, in the monkey, to area 7a and to the lateral intraparietal area, respectively. The DLFC could also have a dual role: memorization of visuospatial information by the PFC, and triggering of memory-guided saccades by the FEF.     

The time course of online trajectory corrections in memory-guided saccades

Recent investigations have revealed the kinematics of horizontal saccades are less variable near the end of the trajectory than during the course of execution. Converging evidence indicates that oculomotor networks use online sensorimotor feedback to correct for initial trajectory errors. It is also known that oculomotor networks express saccadic corrections with decreased efficiency when responses are made toward memorized locations. The present research investigated whether repetitive motor timekeeping influences online feedback-based corrections in predictive saccades. Predictive saccades are a subclass of memory-guided saccades and are observed when one makes series of timed saccades. We hypothesized that cueing predictive saccades in a sequence would facilitate the expression of trajectory corrections. Seven participants produced a number of single unpaced, visually guided saccades, and also sequences of timed predictive saccades. Kinematic and trajectory variability were used to measure the expression of online saccadic corrections at a number of time indices in saccade trajectories. In particular, we estimated the minimum time required to implement feedback-based corrections, which was consistently 37 ms. Our observations demonstrate that motor commands in predictive memory-guided saccades can be parameterized by spatial working memory and retain the accuracy of online trajectory corrections typically associated with visually guided behavior. In contrast, untimed memory-guided saccades exhibited diminished kinematic evidence for online corrections. We conclude that motor timekeeping and sequencing contributed to efficient saccadic corrections. These results contribute to an evolving view of the interactions between motor planning and spatial working memory, as they relate to oculomotor control.     

Influence of the supplementary motor area on primary motor cortex excitability during movements triggered by neutral or emotionally unpleasant visual cues

The stronger anatomo-functional connections of the supplementary motor area (SMA), as compared with premotor area (PM), with regions of the limbic system, suggest that SMA could play a role in the control of movements triggered by visual stimuli with emotional content. We addressed this issue by analysing the modifications of the excitability of the primary motor area (M1) in a group of seven healthy subjects, studied with transcranial magnetic stimulation (TMS), after conditioning TMS of SMA, during emotional and non-emotional visually cued movements. Conditioning TMS of the PM or of contralateral primary motor cortex (cM1) were tested as control conditions. Single-pulse TMS over the left M1 was randomly intermingled with paired TMS, in which a conditioning stimulation of the left SMA, left PM or right M1 preceded test stimulation over the left M1. The subjects carried out movements in response to computerised visual cues (neutral pictures and pictures with negative emotional content). The amplitudes of motor-evoked potentials (MEPs) recorded from the right first dorsal interosseous muscle after paired TMS were measured and compared with those obtained after single-pulse TMS of the left M1 under the various experimental conditions. Conditioning TMS of the SMA in the paired-pulse paradigm selectively enhanced MEP amplitudes in the visual-emotional triggered movement condition, compared with single-pulse TMS of M1 alone or with paired TMS during presentation of neutral visual cues. On the other hand, conditioning TMS of the PM or cM1 did not differentially influence MEP amplitudes under visual-emotional triggered movement conditions. This pattern of effects was related to the intensity of the conditioning TMS over the SMA, being most evident with intensities ranging from 110% to 80% of motor threshold. These results suggest that the SMA in humans could interface the limbic and the motor systems in the transformation of emotional experiences into motor actions. Electronic Publication     

The timing and intensity of transcranial magnetic stimulation,and the scalp site stimulated,as variables influencing motor sequence performance in healthy subjects

Objective: The increasing therapeutic use of transcranial magnetic stimulation (TMS) in disorders of cortical excitability raises the need for reliable stimulus variables. Stimulation of cortical motor areas influences motor programming and execution. We investigated the effects of TMS delivered over various cortical motor areas during the reaction time (RT) on the execution of sequential rapid arm movements in healthy subjects. Methods: Subjects performed a five-submovement (S1–S5) motor sequence mainly involving upper limb proximal muscles. RT and movement time (MT) were measured. We delivered late (close to movement onset) and early (close to the go signal) TMS over the primary motor area (M1-FDI hot-spot for the first dorsal interosseus, M1-D hot-spot for the deltoid muscle), the premotor (PM) area, and the supplementary motor area (SMA), using subthreshold and suprathreshold intensity, single and triple pulses. Results: The motor sequence showed a characteristic pattern of submovement duration, S2–S3–S4 being faster than S1 and S5. Late TMS prolonged RT only when high-intensity pulses were delivered over M1-FDI. Single- and triple-pulse TMS over M1-D or M1-FDI significantly prolonged MT with a dose-related effect. Suprathreshold triple-pulse TMS over the PM—but not over the SMA—also lengthened the MT but did not change RT. Early triple-pulse TMS reduced the RT independently from the stimulus intensity and scalp site. SMA and PM—but not M1-D—stimulation also reduced the MT. Single-pulse TMS over the SMA, despite being delivered through a double-cone coil, did not change RT or MT. Conclusions: TMS-induced changes in the kinematics of a sequential arm movement depend closely on the timing of TMS interference, the scalp site stimulated, and the intensity (and number) of stimuli delivered. Late TMS interference inhibits, whereas early interference facilitates, motor performance. The cortical motor region most sensitive to TMS-induced inhibition is that below the scalp site for M1-FDI. In contrast, TMS-induced facilitation has no strict topographic organization. Particularly for MT (although inhibitory and facilitatory effects both depend on stimulation at high intensities) intensity is less crucial than timing of interference and scalp site.     

Comparison of memory- and visually guided saccades using event-related fMRI

Previous functional imaging studies have shown an increased hemodynamic signal in several cortical areas when subjects perform memory-guided saccades than that when they perform visually guided saccades using blocked trial designs. It is unknown, however, whether this difference results from sensory processes associated with stimulus presentation, from processes occurring during the delay period before saccade generation, or from an increased motor signal for memory-guided saccades. We conducted fMRI using an event-related paradigm that separated stimulus-related, delay-related, and saccade-related activity. Subjects initially fixated a central cross, whose color indicated whether the trial was a memory- or a visually guided trial. A peripheral stimulus was then flashed at one of 4 possible locations. On memory-guided trials, subjects had to remember this location for the subsequent saccade, whereas the stimulus was a distractor on visually guided trials. Fixation cross disappearance after a delay period was the signal either to generate a memory-guided saccade or to look at a visual stimulus that was flashed on visually guided trials. We found slightly greater stimulus-related activation for visually guided trials in 3 right prefrontal regions and right rostral intraparietal sulcus (IPS). Memory-guided trials evoked greater delay-related activity in right posterior inferior frontal gyrus, right medial frontal eye field, bilateral supplementary eye field, right rostral IPS, and right ventral IPS but not in middle frontal gyrus. Right precentral gyrus and right rostral IPS exhibited greater saccade-related activation on memory-guided trials. We conclude that activation differences revealed by previous blocked experiments have different sources in different areas and that cortical saccade regions exhibit delay-related activation differences.     

Individual variation in the location of the parietal eye fields: a TMS study

Transcranial magnetic stimulation (TMS) is a popular technique that can be used to investigate the functional role of specific cortical areas with reference to a particular behavioural task. Single-cell recording studies performed in non-human primates have demonstrated that a region of the parietal lobe known as the lateral intraparietal area is specialized in the planning and control of saccadic eye movements. The homologue of this area in humans is termed the parietal eye fields (PEF) and its role in relation to saccades has previously been examined using TMS. In this paper individual variability in the functional effect of parietal TMS on the latency, amplitude and angular direction of visually-guided saccades has been assessed. By examining individual variability in the spatial distribution of scalp-based localization and brain surface anatomy and stereotaxic localizations of the PEF it was shown that the distances between the sites determined by these three methods were not negligible, which raises problems regarding the most reliable anatomical localization technique to use. An assessment of the effect of TMS on saccade metrics (latency, amplitude error and angular error) at a grid of locations over parietal cortex demonstrated a large amount of intra-individual variability in the site where TMS had most affected saccades, leading to the conclusion that there is individual variability in the functional effects of parietal TMS on saccade planning and execution. This study confirms the idea that it may be problematic to use a fixed scalp location for every participant in a study. It may in fact be more appropriate to determine TMS sites functionally on an individual basis if possible. This finding may guide further studies using TMS and saccade planning in order to optimize their capability to investigate this area and to draw meaningful biological conclusions.     

Burst discharges of fastigial neurons in macaque monkeys are driven by vision- and memory-guided saccades but not by spontaneous saccades.

Discharges from 61 saccadic burst neurons in the fastigial oculomotor region were recorded for two trained macaque monkeys during vision-guided or memory-guided saccades or spontaneous saccades in the dark. Although these neurons exhibited vigorous, burst discharges during both vision-guided and memory-guided saccades, only weak bursts were observed during spontaneous saccades in the dark. Especially in 10 of the 61 neurons, saccadic burst discharge was almost completely absent during spontaneous saccades in the dark. These findings suggest that the cerebellum plays an important role in the control of vision-guided saccades as well as memory-guided saccades, but not of spontaneous saccades in the dark.     

Effects of anterior cingulate cortex lesions on ocular saccades in humans

Cerebral blood flow studies in humans suggest that the anterior cingulate cortex (ACC) could be involved in eye movement control. In two patients with a small infarction affecting the posterior part of this area (on the right side) and in ten control subjects, we studied several paradigms of saccadic eye movements: gap task, overlap task, antisaccades (using either a 5° or 25° lateral target), memory-guided saccades with a short (1 s) or long (7 s) delay, and sequences of memory-guided saccades. Compared with controls, patients had normal latency in the gap task but increased latency in the other tasks. The gain of memory-guided saccades was markedly decreased, bilaterally, whatever the duration of the delay. Patients made more errors than controls in the antisaccade task when the 5° lateral target was used, and a higher percentage of chronological errors in the sequences of saccades. These results show that the posterior part of the right ACC plays an important role in eye movement control and suggest that this area could correspond to a “cingulate eye field” (CEF). The role of this hypothetical CEF could be an early activation exerted on the frontal ocular motor areas involved in intentional saccades and also a direct action on brainstem ocular premotor structures. Received: 8 November 1996 / Accepted: 14 October 1997     

Transcranial magnetic stimulation disrupts eye-hand interactions in the posterior parietal cortex

Recent neurophysiological studies have started to shed some light on the cortical areas that contribute to eye-hand coordination. In the present study we investigated the role of the posterior parietal cortex (PPC) in this process in normal, healthy subjects. This was accomplished by delivering single pulses of transcranial magnetic stimulation (TMS) over the PPC to transiently disrupt the putative contribution of this area to the processing of information related to eye-hand coordination. Subjects made open-loop pointing movements accompanied by saccades of the same required amplitude or by saccades that were substantially larger. Without TMS the hand movement amplitude was influenced by the amplitude of the corresponding saccade; hand movements accompanied by larger saccades were larger than those accompanied by smaller saccades. When TMS was applied over the left PPC just prior to the onset of the saccade, a marked reduction in the saccadic influence on manual motor output was observed. TMS delivered at earlier or later periods during the response had no effect. Taken together, these data suggest that the PPC integrates signals related to saccade amplitude with limb movement information just prior to the onset of the saccade.     

Transient disruption of M1 during response planning impairs subsequent offline consolidation

Transcranial magnetic stimulation (TMS) was used to probe the involvement of the left primary motor cortex (M1) in the consolidation of a sequencing skill. In particular we asked: (1) if M1 is involved in consolidation of planning processes prior to response execution (2) whether movement preparation and movement execution can undergo consolidation independently and (3) whether sequence consolidation can occur in a stimulus specific manner. TMS was applied to left M1 while subjects prepared left hand sequential finger responses for three different movement sequences, presented in an interleaved fashion. Subjects also trained on three control sequences, where no TMS was applied. Disruption of subsequent consolidation was observed, but only for sequences where subjects had been exposed to TMS during training. Further, reduced consolidation was only observed for movement preparation, not movement execution. We conclude that left M1 is causally involved in the consolidation of effective response planning for left hand movements prior to response execution, and mediates consolidation in a sequence specific manner. These results provide important new insights into the role of M1 in sequential memory consolidation and sequence response planning.