Springer-Verlag 1998
Exp Brain Res (1998) 123:159±163
B. Gaymard ´ C.J. Ploner ´ S. Rivaud A.I. Vermersch ´ C. Pierrot-Deseilligny
Cortical control of saccades
Abstract Saccadic eye movements are controlled by a cortical network composed of several oculomotor areas that are now accurately localized. Clinical and experimental studies have enabled us to understand their specific roles better. These areas are: (1) the parietal eye field (PEF) located in the intraparietal sulcus involved in visuospatial integration and in reflexive saccade triggering; (2) the frontal eye field (FEF), located in the precentral gyrus, involved in the preparation and the triggering of purposive saccades; and (3) the supplementary eye field (SEF) on the medial wall of the frontal lobe, probably involved in the temporal control of sequences of visually guided saccades and in eye-hand coordination. A putative cingulate eye field (CEF), located in the anterior cingulate cortex, would be involved in motivational modulation of voluntary saccades. Besides these motor areas, the dorsolateral prefrontal cortex (dlPFC) in the midfrontal gyrus is involved in reflexive saccade inhibition and visual shortterm memory. Key words Eye movements ´ Saccade ´ Cerebral cortex ´ Human
Introduction The cortical network involved in saccade control is now well characterized in humans thanks to studies of patients with discrete cortical lesions and functional imaging studies in normal subjects (see for a review PierrotDeseilligny et al. 1995). This network is composed of areas directly triggering saccades and areas concerned with
)
B. Gaymard ( ) ´ S. Rivaud ´ A.I. Vermersch C. Pierrot-Deseilligny INSERM 289, Hôpital de la Salptrire, 47 Bd de l'Hôpital, F-75651 Paris, France e-mail:
[email protected], Fax: 33-1-44-24-36-58 C.J. Ploner Klinik für Neurologie, CharitØ, Humboldt-Universität Berlin, Berlin, Germany
cognitive aspects of saccade control. Areas triggering saccades are: (i) the parietal eye field (PEF), located in the intraparietal sulcus (Müri et al. 1996a), (ii) the frontal eye field (FEF), located in the precentral gyrus (Paus 1996) and (iii) the supplementary eye field (SEF) on the upper part of the medial wall of the frontal lobe (Petit et al. 1996). Areas involved in cognitive aspects of saccade control are: (i) the dorsolateral prefrontal cortex (dlPFC), largely congruent with Brodman's area 46 (B46) (spatial short-term memory, inhibition of reflexive saccades) and (ii) the anterior cingulate gyrus (motivational modulation of voluntary saccades). The latter has been identified quite recently by both clinical and functional imaging and studies (Paus et al. 1993; Petit et al. 1993, 1996; Gaymard et al. 1998). Depending on the behavioral context in which the saccade is performed, the brain recruits different cortical areas within this network. In this review we will concentrate on two simple saccade tasks performed by patients with lesions in one of these cortical areas or with subcortical lesions affecting their efferent pathways. Experimental data from studies with nonhuman primates are briefly discussed when relevant to our results.
Ocular motor paradigms (Fig. 1) The reflexive visually guided saccade task (VGST). In this task, the subject is requested to follow precisely a visual cue that jumps with unpredictable timing and direction to a position 25 to the right or left of a central fixation point. Since the parameters of the task are unpredictable, the corresponding saccades are called reflexive saccades. The memory-guided saccade task (MGST). While the subject fixates a central fixation point in otherwise complete darkness, a visual cue is briefly presented laterally (10 , 15 , 20 , 25 or 30 ). After a delay, during which fixation is maintained, the central fixation point is switched off and the subject is requested to move his or her eyes as
160
Posterior limb of the internal capsule
Fig. 1A±C Saccade paradigms. A Unpredictable visually guided saccade task. B Predictable visually guided saccade task. C Memory-guided saccade task. CT Control target (with the same position as the flash), F flash, L left, R right, T target
precisely as possible to where the cue had been presented (memory-guided saccade). After the saccade the cue is reilluminated, permitting corrective saccades if necessary. In this task, saccades are largely volitional, requiring a short internal representation of the cues' spatial coordinates. In the following, we will only consider accuracy of saccades, as measured by saccade gain (ratio of saccade amplitude to target eccentricity).
Lesions of the parietal lobe PEF After a PEF lesion, all contralateral visually guided saccades (i.e., saccades directed away from the lesion side) become inaccurate with gain decreasing significantly. The same pattern of impairment is seen in memory-guided saccades (Pierrot-Deseilligny et al. 1991b). These results are consistent with experimental data since, in the monkey, area LIP, the equivalent to the PEF in humans, contains both neurons that discharge shortly before saccade triggering and neurons that are active during the delay period of the MGST (Gnadt and Anderson 1988). However, it is questionable whether parietal ocular motor areas are directly involved in memory processes that exceed some hundred milliseconds (Chafee and GoldmanRakic 1994; Müri et al. 1996b; Brandt et al. 1997). The PEF has two main outputs: towards the frontal lobe (Cavada and Goldman-Rakic 1989) and a direct one to the superior colliculus (the parieto-tectal pathway) (Lynch et al. 1985). Since the latter travels through the posterior limb of the internal capsule, lesions in this area selectively disrupt the PEF efferent pathway without affecting cortical processing in the parietal lobe.
In patients with a lesion of the posterior limb of the internal capsule saccades performed in the VGST show similar impairments as after a PEF lesion, i.e., a markedly decreased gain contralateral to the lesion side (Gaymard et al. 1998). However, if the subject is initially informed that the target will constantly appear at the same position (e.g., 25 right) and the target position is thus predictable, this asymmetry in gain disappears and saccades contralateral to the lesion side become accurate. Likewise, memory-guided saccades remain unimpaired (Gaymard et al. 1997). Two conclusions may be drawn from these results. First, the direct parieto-tectal pathway appears to be involved in conditions in which the subject can not predict the target position (i.e., novel targets). As soon as an internal representation of the target is available to the brain, because the target location is either actively memorized or predictable, this pathway is no longer relevant for behavior. Second, since memory-guided saccades are impaired after parietal cortical lesions but unaffected after lesions of the parieto-tectal pathway, cortical processing in the parietal lobe is most probably sent to the frontal lobe rather than directly to the superior colliculus.
Lesions of the frontal lobe dlPFC Accuracy of visually guided saccades is not affected by a dlPFC lesion (Pierrot-Deseilligny et al. 1991a). However, deficits become apparent if the instructions of the task are changed, and the subject is requested to make a saccade in the opposite direction to the target instead of following it. This paradigm, called the antisaccade task, requires inhibition of a reflexive visually guided saccade and triggering of a voluntary nonvisually guided saccade. In this task, patients with a dlPFC lesion show a markedly increased percentage of errors, i.e., erroneous reflexive saccades to the cue, providing evidence for a deficit in reflexive saccade inhibition (Pierrot-Deseilligny et al. 1991a). In the MGST, patients with a dlPFC lesion show a markedly decreased accuracy of saccades (Pierrot-Deseilligny et al. 1991b). Normal visually guided saccades in these patients indicate that this impairment is neither sensory nor motor, but a true memory deficit. Monkeys with a dlPFC lesion have shown the same pattern of deficits, with inaccuracy of memory-guided saccades increasing with the length of the delay (Funahashi et al. 1993). Single-unit recordings from the monkey dlPFC have revealed neurons with spatially selective delay period activity, i.e., activity that maintains visual inputs from a particular portion of the visual field (Funahashi et al. 1989). These neurons could form a neural substrate for visuo-spatial short-term memory in the primate dlPFC. These mnemonic functions of dlPFC neurons seem not to be restricted to eye movements, since it has been shown that the stored spatial information is also relevant to memory-guided limb movements (Fuster 1989).
161
FEF The location of the FEF, well documented in the monkey, has only recently been precisely identified in humans. It was previously thought to be located in area B8, but recent imaging studies have shown that it belongs in fact to area B6, in the precentral gyrus (Paus 1996). On a vertical axis, its location corresponds to the primary hand motor area. After a FEF lesion, visually guided saccades are only slightly affected, with a discrete reduction of contralateral gain. In several patients with small, recent lesions restricted to the FEF and sparing the nearby dlPFC, the percentage of errors in the antisaccade task was normal (Rivaud et al. 1994; Gaymard et al. unpublished data). However, normal antisaccades (i.e., voluntary saccades made away from the target) had an increased latency. Memory-guided saccades show a marked contralateral inaccuracy, with an increasing impairment for longer delays, thus suggesting mnemonic functions for the FEF as well. Interestingly, in contrast to dlPFC lesions, the memory deficit seems to be restricted to eye movements with memory-guided limb movements remaining unaffected. These results are consistent with recent neurophysiological data. The activity of presaccadic movement neurons in the FEF appears to be strongly context-dependent, being absent before spontaneous saccades and maximum before volitional saccades (Bruce and Goldberg 1985). Furthermore, the FEF contains neurons with spatially selective delay period activity, and chemical inactivation of the FEF by local injection of scopolamine affects mostly memory-guided saccades, with a more marked impairment for longer delays (Dias and Bruce 1994). The FEF has access to the brainstem ocular motor structures by at least three main pathways: a direct pathway towards the pons, and two pathways relaying in the superior colliculus ± a direct one, and an indirect one relaying in the basal ganglia, namely the caudate nucleus and the substantia nigra pars reticulata (Leichnetz and Goldberg 1988). Body of the caudate nucleus In a patient with a lesion involving the body of the left caudate nucleus, visually guided saccades were slightly impaired contralaterally whereas a profound reduction of saccade gain was observed in the MGST (Vermersch et al. unpublished data). This deficit was predominantly contralateral, and increased when the delay was lengthened from 1 to 7 s. Memory-guided pointing, performed in otherwise identical conditions to the MGST, was unimpaired. Antisaccades were normal. A neural correlate for these deficits has been identified in the body of the monkey caudate nucleus, where neurons discharge selectively before memory-guided saccades (Hikosaka et al. 1989). Taken together, the abovementioned findings favor a model in which the FEF, the caudate nucleus and probably the substantia nigra pars reticulata participate in a cor-
tico-subcortical memorization network specialized for eye movements. However, the delay-dependence of the deficits in the MGST observed after a dlPFC lesion indicates that this oculomotor network is not sufficient to maintain accurate spatial information over several seconds and may thus perhaps be subordinate to a less specialized representation in the dlPFC. Anterior limb of the internal capsule In patients with a lesion in the anterior limb of the internal capsule, i.e., with a lesion involving the FEF efferent pathways towards the brainstem (superior colliculus and/or pons), all contralateral reflexive and volitional saccades show a decreased gain (Gaymard et al. 1997). The inaccuracy observed in memory-guided saccades obviously results from a pure motor impairment, since the deficit is not influenced by the length of the memorization delay. The fact that deficits can also be observed in reflexive saccades suggests that, besides the parieto-tectal pathway, a parallel frontofugal pathway may also be involved in saccades to targets of unpredictable location.
Medial wall of the frontal lobe SEF Patients with a lesion in the SEF have normal visually guided and memory-guided saccades, but are unable to perform memory-guided sequences of visually guided saccades (Gaymard et al. 1990), especially if the lesion is on the left side (Gaymard et al. 1993; Heide et al. 1995). Recent functional imaging studies have confirmed these results (Heide et al. 1997). In the monkey, the SEF contains neurons that are specifically activated during coordinated eye-hand movements (Mushiake et al. 1996; Chou and Schiller 1997). Furthermore, experimental (Olson et al. 1996) and clinical (Pierrot-Deseilligny et al. 1993; Isral et al. 1995) data suggest that this area is able to control saccades performed in object-centered or exocentric frames of reference. Thus, this area seems to control motor programs of eye and limb movements, in both temporal (chronology of each elementary movement) and spatial (common frame of reference for each elementary movement) domains. The putative cingulate eye field A recent study of two patients with a lesion restricted to the anterior cingulate cortex suggests that this area could contain a cingulate eye field (Gaymard et al. 1998). These patients showed saccade deficits similar to those observed after frontal lobe lesions (FEF, SEF and dlPFC), affecting all types of volitional saccades, i.e., memory-guided saccades, antisaccades and memorized sequences of visually guided saccades. The common lesioned zone of these two
162 Fig. 2 Pathways involved in visually guided saccade generation. The direct parieto-tectal pathway (1), traveling through the posterior limb of the internal capsule, is mainly involved when target location is unpredictable. The involvement of the fronto-fugal pathway (2), traveling through the anterior limb of the internal capsule, increases as the degree of predictability of the task increases. ant Anterior limb, FEF frontal eye field, PEF parietal eye field, post posterior limb, Sup C superior colliculus, VC visual cortex
Fig. 3 Pathways involved in memory-guided generation. Visual information is integrated in the parietal lobe into a spatial map, and stored in the dorsolateral prefrontal cortex. In the context of a saccadic response, an oculomotor signal is stored in a cortico-subcortical oculomotor loop, then sent, when the saccade is required, through a fronto-fugal pathway traveling through the anterior limb of the internal capsule, towards the superior colliculus and the brainstem premotor structures. CN Caudate nucleus, dlPFC dorsolateral prefrontal cortex, FEF frontal eye field, PEF parietal eye field, SNpr substantia nigra pars reticulata, Th thalamus, VC visual cortex
patients was slightly anterior to the vertical anterior commissure line (Vac line of Talairach), an area that corresponds to the locus of activation found in several imaging studies performed on normal humans during ocular motor tasks (Paus et al. 1993; Anderson et al. 1994; O'Sullivan et al. 1995; Petit et al. 1996). Recent electrophysiological studies have identified new premotor areas in the depth of the cingulate sulcus, known as the cingulate motor areas (for a review see Picard and Strick 1995). These areas contain neurons that are activated before limb or face movements and project either to the primary motor cortex or directly to the spinal cord. Whether the anterior cingulate contains analogous ocular premotor areas, as suggested by our results, remains to be substantiated by singleunit studies.
Conclusion We propose two simplified diagrams to illustrate the areas and pathways involved in the two paradigms presented in this review. For reflexive visually guided saccades (Fig. 2), i.e., saccades made towards novel targets, the parietal lobe integrates visuospatial information to generate a motor signal that is sent to the superior colliculus by a direct parieto-tectal pathway traveling through the posterior limb of the internal capsule. A parallel involvement of the frontal lobe and its efferent pathways may depend on the degree of predictability of the visual targets. Hence, when the subject is able to build an internal representation of the location of the target (either from memory or prediction), the direct parieto-tectal pathway
163
is no longer involved. Although visuospatial integration is still performed in the parietal lobe, the signal is sent to the frontal lobe, essentially to the FEF. If the response is delayed (Fig. 3), the signal is maintained in a corticosubcortical network involving the dlPFC, FEF and the basal ganglia (caudate nucleus, substantia nigra pars reticulata and thalamus). When a saccade is required, a motor signal is generated in the FEF and sent through the efferent pathways traveling in the anterior limb of the internal capsule towards the premotor structures of the brainstem, either directly or after a relay in the superior colliculus.
References Anderson TJ, Jenkins IH, Brooks DLJ, Hawken MB, Frackowiak RSJ, Kennard C (1994) Cortical control of saccades and fixation in man: a PET study. Brain 17:1073±1084 Brandt SA, Ploner CJ, Meyer BU, Leistner S, Villringer A (1998) Effects of repetitive transcranial magnetic stimulation over dorsolateral and posterior cortex on memory-guided saccades. Exp Brain Res 118:197±204 Bruce CJ, Goldberg ME (1985) Primate frontal eye fields. I. Single neurons discharging before saccades. J Neurophysiol 53:603± 635 Cavada C, Goldman-Rakic PS (1989) Posterior parietal cortex in rhesus monkey. II. Evidence for segregated corticocortical networks linking sensory and limbic areas within the frontal lobe. J Comp Neurol 287:422±445 Chafee M, Goldman-Rakic PS (1994) Prefrontal cooling dissociated memory- and sensory-guided oculomotor delayed response functions. Soc Neurosci Abstr 20:808 Chou I, Schiller PH (1997) Single cell responses in dorsomedial frontal cortex (DMFC) to eye and arm movements. Soc Neurosci Abstr 23:1552 Dias EC, Bruce CJ (1994) Physiological correlate of fixation disengagement in the primate's frontal eye field. J Neurophysiol 72: 2532±2537 Funahashi S, Bruce CJ, Goldman-Rakic PS (1989) Mnemonic coding of visual space in the monkey's dorsolateral prefrontal cortex. J Neurophysiol 61:331±349 Funahashi S, Bruce CJ, Goldman-Rakic PS (1993) Dorsolateral prefrontal lesions and oculomotor delayed-response performance: evidence for mnemonic scotomas. J Neurosci 13:1479±1497 Fuster JM (1989) The prefrontal cortex. 2nd edn. Raven Press, New York Gaymard B, Pierrot-Deseilligny C, Rivaud S (1990) Impairments of sequences of memory-guided saccades after supplementary motor area lesions. Ann Neurol 28:622±626 Gaymard B, Rivaud S, Pierrot-Deseilligny C (1993) Role of the left and right supplementary motor areas in memory-saccade sequences. Ann Neurol 34:404±406 Gaymard B, Ploner CJ, Rivaud S, Pierrot-Deseilligny C (1997) Eye movement impairments after lesions of the anterior vs posterior limb of the internal capsule: selective involvement of frontal vs parietal oculomotor efferences. Soc Neurosci Abst 23:475 Gaymard B, Rivaud S, Cassarini JF, Dubard T, Rancurel G, Agid Y, Pierrot-Deseilligny C (1998) Effects of anterior cingulate cortex lesions on ocular saccades in humans. Exp Brain Res 120:173±183 Gnadt JW, Anderson RA (1988) Memory related motor planning activity in posterior parietal cortex of macaque. Exp Brain Res 70:216±220
Heide W, Blankenburg M, Zimmermann E, Kömpf D (1995) Cortical control of double-step saccades: implications for spatial orientation. Ann Neurol 38:739±748 Heide W, Binkofski F, Posse S, Seitz RJ, Kömpf D, Freund HJ (1997) Functional anatomy of memory-guided saccade sequences in human cerebral cortex. Soc Neurosci Abst 23:2222 Hikosaka O, Sakamoto M, Usui S (1989) Functional properties of monkey caudate neurons. I. Activities related to saccadic eye movements. J Neurophysiol 61:780±798 Isral I, Rivaud S, Gaymard B, Berthoz A, Pierrot-Deseilligny C (1995) Cortical control of vestibular-guided saccades in man. Brain 118:1169±1183 Leichnetz GR, Goldberg ME (1988) Higher centers concerned with eye movement and visual attention: cerebral cortex and thalamus. In: Büttner-Ennever JA (ed) Neuroanatomy of the oculomotor system. Elsevier, Amsterdam, pp 365±429 Lynch JC, Graybiel AM, Lobeck LJ (1985) The differential projection of two cytoarchitectural subregions of the inferior parietal lobule of macaque upon the deep layers of the superior colliculus. J Comp Neurol 235:242±254 Müri RM, Iba-Zizen MT, Derosier C, Cabanis EA, PierrotDeseilligny C (1996a) Location of the human posterior eye field with functional magnetic resonance imaging. J Neurol Neurosurg Psychiatry 60:445±448 Müri RM, Vermersch AI, Rivaud S, Gaymard B, Pierrot-Deseilligny C (1996b) Effects of single-pulse transcranial magnetic stimulation over the prefrontal and posterior parietal cortices during memory-guided saccades in humans. J Neurophysiol 76:2102± 2106 Mushiake H, Fujii N, Tanji J (1996) Visually guided saccade versus eye-hand reach: contrasting neuronal activity in the cortical supplementary and frontal eye fields. J Neurophysiol 75:2187±2191 Olson CR, Gettner SN (1995) Object-centered direction selectivity in the macaque supplementary eye-field. Science 269:985±988 O'Sullivan EP, Jenkins IH, Henderson L, Kennard C, Brooks DJ (1995) The functional anatomy of remembered saccades: a PET study. Neuroreport 6:2141±2144 Paus T (1996) Location and function of the human frontal eye field: a selective review. Neuropsychologia 734:475±483 Paus T, Petrides M, Evans AC, Meyer E (1993) Role of the human anterior cingulate cortex in the control of oculomotor, manual and speech responses: a positron emission tomography study. J Neurophysiol 70:453±469 Petit L, Orssaud C, Tzourio N, Salamon G, Mazoyer B, Berthoz A (1993) PET study of voluntary saccadic eye movement in humans: basal ganglia±thalamocortical system and cingulate cortex involvement. J Neurophysiol 69:1009±1017 Petit L, Orssaud C, Tzourio N, Crivello F, Berthoz A, Mazoyer B (1996) Functional anatomy of a prelearned sequence of horizontal saccades in humans. J Neurosci 16:3714±3726 Picard N, Strick PL (1996) Medial wall motor areas: a review of their location and functional activation. Cereb Cortex 6:342±353 Pierrot-Deseilligny C, Rivaud S, Gaymard B (1991a) Cortical control of reflexive visually guided saccades in man. Brain 114: 1473±1485 Pierrot-Deseilligny C, Rivaud S, Gaymard B, Agid Y (1991b) Cortical control of memory-guided saccades in man. Exp Brain Res 83:607±617 Pierrot-Deseilligny C, Isral I, Berthoz A, Rivaud S, Gaymard B (1993) Role of the different frontal lobe areas in the control of the horizontal component of memory-guided saccades in man. Exp Brain Res 95:166±171 Pierrot-Deseilligny C, Rivaud S, Gaymard B, Müri R, Vermersch AI (1995) Cortical control of saccades. Ann Neurol 37:557±567 Rivaud S, Müri RM, Gaymard B, Vermersch AI, Pierrot-Deseilligny C (1994) Eye movement disorders after frontal eye field lesions in humans. Exp Brain Res 102:110±120
