. The deeper layers of the midbrain superior colliculus (SC) contain a topographic motor map in which a localized population of cells is recruited for each saccade, but how the brain stem decodes the dynamic SC output is unclear. Here we analyze saccade-related responses in the monkey SC to test a new dynamic ensemble-coding model, which proposes that each spike from each saccade-related SC neuron adds a fixed, site-specific contribution to the intended eye movement command. As predicted by this simple theory, we found that the cumulative number of spikes in the cell bursts is tightly related to the displacement of the eye along the ideal straight trajectory, both for normal saccades and for strongly curved, blink-perturbed saccades toward a single visual target. This dynamic relation depends systematically on the metrics of the saccade displacement vector, and can be fully predicted from a quantitative description of the cell's classical movement field. Furthermore, we show that a linear feedback model of the brain stem, which is driven by dynamic linear vector summation of measured SC firing patterns, produces realistic two-dimensional (2D) saccade trajectories and kinematics. We conclude that the SC may act as a nonlinear, vectorial saccade generator that programs an optimal straight eye-movement trajectory. I N T R O D U C T I O NThe midbrain superior colliculus (SC) is a sensorimotor interface that is critically involved in the control of rapid gaze shifts. An important problem in understanding its role in gaze control is how the spatial distribution of movement-related activity in its motor map is ultimately transformed into the temporal code carried by motor neurons (Sparks and HartwichYoung 1989). In this study, we analyzed saccade-related responses in the monkey SC to test a new theoretical framework for the involvement of the SC in the generation of saccades in two dimensions (2D). First, we present a novel analysis of SC spike trains that provides evidence for dynamic vector summation of movement contributions provided by each spike of each cell in the active population. We then analyze the spatialtemporal distribution of SC activity. The results are used to test the predictions and emerging properties of our new ensemblecoding theory, which assumes dynamic, linear decoding of the SC population activity by the brain stem saccade generator. Finally, we propose and test a new quantitative description of dynamic SC movement fields that is implied by our theory. In what follows, we first explain why a new approach is called for by highlighting the main findings that have led to several controversies. These controversies include 1) static versus dynamic involvement of the SC, 2) vector summation versus vector averaging of the population activity, and 3) feedforward versus feedback involvement of the SC. Earlier theories and controversies

The coordination between eye and head movements during a rapid orienting gaze shift has been investigated mainly when subjects made horizontal movements towards visual targets with the eyes starting at the centre of the orbit. Under these conditions, it is difficult to identify the signals driving the two motor sys tems, because their initial motor errors are identical and equal to the coordinates o f the sensory stimulus (i.e. reti nal error). In this paper, we investigate head-free gaze saccades of human subjects towards visual as w ell as au ditory stimuli presented in the two-dimensional frontal plane, under both aligned and unaligned initial fixation conditions. Although the basic patterns for eye and head movements were qualitatively comparable for both stim ulus modalities, systematic differences were also ob tained under aligned conditions, suggesting a task-de pendent movement strategy. Auditory-evoked gaze shifts were endowed with smaller eye-head latency differences, consistently larger head movements and smaller concom itant ocular saccades than visually triggered movements. By testing gaze control for eccentric initial eye positions, we found that the head displacement vector was best re lated to the initial head motor-error (target-re-head), rather than to the initial gaze error (target-re-eye), re gardless of target modality. These findings suggest an in dependent control of the eye and head motor systems by commands in different frames of reference. However, we also observed a systematic influence o f the oculomotor response on the properties of the evoked head move ments, indicating a subtle coupling between the two sys tems. The results are discussed in view of current eyehead coordination models.

A major challenge in computational neurobiology is to understand how populations of noisy, broadly-tuned neurons produce accurate goal-directed actions such as saccades. Saccades are high-velocity eye movements that have stereotyped, nonlinear kinematics; their duration increases with amplitude, while peak eye-velocity saturates for large saccades. Recent theories suggest that these characteristics reflect a deliberate strategy that optimizes a speed-accuracy tradeoff in the presence of signal-dependent noise in the neural control signals. Here we argue that the midbrain superior colliculus (SC), a key sensorimotor interface that contains a topographically-organized map of saccade vectors, is in an ideal position to implement such an optimization principle. Most models attribute the nonlinear saccade kinematics to saturation in the brainstem pulse generator downstream from the SC. However, there is little data to support this assumption. We now present new neurophysiological evidence for an alternative scheme, which proposes that these properties reside in the spatial-temporal dynamics of SC activity. As predicted by this scheme, we found a remarkably systematic organization in the burst properties of saccade-related neurons along the rostral-to-caudal (i.e., amplitude-coding) dimension of the SC motor map: peak firing-rates systematically decrease for cells encoding larger saccades, while burst durations and skewness increase, suggesting that this spatial gradient underlies the increase in duration and skewness of the eye velocity profiles with amplitude. We also show that all neurons in the recruited population synchronize their burst profiles, indicating that the burst-timing of each cell is determined by the planned saccade vector in which it participates, rather than by its anatomical location. Together with the observation that saccade-related SC cells indeed show signal-dependent noise, this precisely tuned organization of SC burst activity strongly supports the notion of an optimal motor-control principle embedded in the SC motor map as it fully accounts for the straight trajectories and kinematic nonlinearity of saccades.

While firing rate is well established as a relevant parameter for encoding information exchanged between neurons, the significance of other parameters is more conjectural. Here, we show that regularity of neuronal spike activities affects sensorimotor processing in tottering mutants, which suffer from a mutation in P/Q-type voltage-gated calcium channels. While the modulation amplitude of the simple spike firing rate of their floccular Purkinje cells during optokinetic stimulation is indistinguishable from that of wild-types, the regularity of their firing is markedly disrupted. The gain and phase values of tottering's compensatory eye movements are indistinguishable from those of flocculectomized wild-types or from totterings with the flocculus treated with P/Q-type calcium channel blockers. Moreover, normal eye movements can be evoked in tottering when the flocculus is electrically stimulated with regular spike trains mimicking the firing pattern of normal simple spikes. This study demonstrates the importance of regularity of firing in Purkinje cells for neuronal information processing.

Sound localization in humans relies on binaural differences (azimuth cues) and monaural spectral shape information (elevation cues) and is therefore the result of a neural computational process. Despite the fact that these acoustic cues are referenced with respect to the head, accurate eye movements can be generated to sounds in complete darkness. This ability necessitates the use of eye position information. So far, however, sound localization has been investigated mainly with a fixed head position, usually straight ahead. Yet the auditory system may rely on head motor information to maintain a stable and spatially accurate representation of acoustic targets in the presence of head movements. We therefore studied the influence of changes in eye-head position on auditory-guided orienting behavior of human subjects. In the first experiment, we used a visual-auditory double-step paradigm. Subjects made saccadic gaze shifts in total darkness toward brief broadband sounds presented before an intervening eye-head movement that was evoked by an earlier visual target. The data show that the preceding displacements of both eye and head are fully accounted for, resulting in spatially accurate responses. This suggests that auditory target information may be transformed into a spatial (or body-centered) frame of reference. To further investigate this possibility, we exploited the unique property of the auditory system that sound elevation is extracted independently from pinna-related spectral cues. In the absence of such cues, accurate elevation detection is not possible, even when head movements are made. This is shown in a second experiment where pure tones were localized at a fixed elevation that depended on the tone frequency rather than on the actual target elevation, both under head-fixed and -free conditions. To test, in a third experiment, whether the perceived elevation of tones relies on a head- or space-fixed target representation, eye movements were elicited toward pure tones while subjects kept their head in different vertical positions. It appeared that each tone was localized at a fixed, frequency-dependent elevation in space that shifted to a limited extent with changes in head elevation. Hence information about head position is used under static conditions too. Interestingly, the influence of head position also depended on the tone frequency. Thus tone-evoked ocular saccades typically showed a partial compensation for changes in static head position, whereas noise-evoked eye-head saccades fully compensated for intervening changes in eye-head position. We propose that the auditory localization system combines the acoustic input with head-position information to encode targets in a spatial (or body-centered) frame of reference. In this way, accurate orienting responses may be programmed despite intervening eye-head movements. A conceptual model, based on the tonotopic organization of the auditory system, is presented that may account for our findings.

Recently, we proposed an ensemble-coding scheme of the midbrain superior colliculus (SC) in which, during a saccade, each spike emitted by each recruited SC neuron contributes a fixed minivector to the gaze-control motor output. The size and direction of this ‘spike vector’ depend exclusively on a cell’s location within the SC motor map (Goossens and Van Opstal, in J Neurophysiol 95: 2326–2341, 2006). According to this simple scheme, the planned saccade trajectory results from instantaneous linear summation of all spike vectors across the motor map. In our simulations with this model, the brainstem saccade generator was simplified by a linear feedback system, rendering the total model (which has only three free parameters) essentially linear. Interestingly, when this scheme was applied to actually recorded spike trains from 139 saccade-related SC neurons, measured during thousands of eye movements to single visual targets, straight saccades resulted with the correct velocity profiles and nonlinear kinematic relations (‘main sequence properties– and ‘component stretching’) Hence, we concluded that the kinematic nonlinearity of saccades resides in the spatial-temporal distribution of SC activity, rather than in the brainstem burst generator. The latter is generally assumed in models of the saccadic system. Here we analyze how this behaviour might emerge from this simple scheme. In addition, we will show new experimental evidence in support of the proposed mechanism.

Trigeminal reflex blinks evoked near the onset of a saccade cause profound spatial-temporal perturbations of the saccade that are typically compensated in mid-flight. This paper investigates the influence of reflex blinks on the discharge properties of saccade-related burst neurons (SRBNs) in intermediate and deep layers of the monkey superior colliculus (SC). Twenty-nine SRBNs, recorded in three monkeys, were tested in the blink-perturbation paradigm. We report that the air puff stimuli, used to elicit blinks, resulted in a short-latency ( approximately 10 ms) transient suppression of saccade-related SRBN activity. Shortly after this suppression (within 10-30 ms), all neurons resumed their activity, and their burst discharge then continued until the perturbed saccade ended near the extinguished target. This was found regardless whether the compensatory movement was into the cell's movement field or not. In the limited number of trials where no compensation occurred, the neurons typically stopped firing well before the end of the eye movement. Several aspects of the saccade-related activity could be further quantified for 25 SRBNs. It appeared that 1) the increase in duration of the high-frequency burst was well correlated with the (two- to threefold) increase in duration of the perturbed movement. 2) The number of spikes in the burst for control and perturbed saccades was quite similar. On average, the number of spikes increased only 14%, whereas the mean firing rate in the burst decreased by 52%. 3) An identical number of spikes were obtained between control and perturbed responses when burst and postsaccadic activity were both included in the spike count. 4) The decrease of the mean firing rate in the burst was well correlated with the decrease in the velocity of perturbed saccades. 5) Monotonic relations between instantaneous firing rate and dynamic motor error were obtained for control responses but not for perturbed responses. And 6) the high-frequency burst of SRBNs with short-lead and long-lead presaccadic activity (also referred to as burst and buildup neurons, respectively) showed very similar features. Our findings show that blinking interacts with the saccade premotor system already at the level of the SC. The data also indicate that a neural mechanism, rather than passive elastic restoring forces within the oculomotor plant, underlies the compensation for blink-related perturbations. We propose that these interactions occur downstream from the motor SC and that the latter may encode the desired displacement vector of the eyes by sending an approximately fixed number of spikes to the brainstem saccadic burst generator.

A longstanding but still controversial hypothesis is that longterm depression (LTD) of parallel fiber-Purkinje cell synapses in the cerebellum embodies part of the neuronal information storage required for associative motor learning. Transgenic mice in which LTD is blocked by Purkinje cell-specific inhibition of protein kinase C (PKC) (L7-PKCI mutants) do indeed show impaired adaptation of their vestibulo-ocular reflex, whereas the dynamics of their eye movement performance are unaffected. However, because L7-PKCI mutants have a persistent multiple climbing fiber innervation at least until 35 d of age and because the baseline discharge of the Purkinje cells in the L7-PKCI mutants is unknown, factors other than a blockage of LTD induction itself may underlie their impaired motor learning. We therefore investigated the spontaneous discharge of Purkinje cells in alert adult L7-PKCI mice as well as their multiple climbing fiber innervation beyond the age of 3 months. We found that the simple spike and complex spike-firing properties (such as mean firing rate, interspike interval, and spike count variability), oscillations, and climbing fiber pause in the L7-PKCI mutants were indistinguishable from those in their wild-type littermates. In addition, we found that multiple climbing fiber innervation does not occur in cerebellar slices obtained from 3-to 6-month-old mutants. These data indicate (1) that neither PKC inhibition nor the subsequent blockage of LTD induction disturbs the spontaneous discharge of Purkinje cells in alert mice, (2) that Purkinje cell-specific inhibition of PKC detains rather than prevents the developmental conversion from multiple to mono-innervation of Purkinje cells by climbing fibers, and (3) that as a consequence the impaired motor learning as observed in older adult L7-PKCI mutants cannot be attributable either to a disturbance in the baseline simple spike and complex spike activities of their Purkinje cells or to a persistent multiple climbing fiber innervation. We conclude that cerebellar LTD is probably one of the major mechanisms underlying motor learning, but that deficits in LTD induction and motor learning as observed in the L7-PKCI mutants may only be reflected in differences of the Purkinje cell signals during and/or directly after training. Key words: heterosynaptic plasticity; motor learning; multiple climbing fiber innervation; cerebellar and vestibular nuclei; genetic manipulation; phosphorylationA challenge faced by neuroscience is to understand how model systems of information storage in the brain, such as long-term potentiation and long-term depression (LTD), function in neural circuits that control behavioral learning. Cerebellar LTD is an attenuation of the granule cell axon-Purkinje cell synapse that occurs after conjunctive stimulation of the granule cell axons and climbing fiber inputs (Ito et al., 1982;Linden and Connor, 1995). It has been suggested that LTD underlies several forms of motor learning, including adaptation of the vestibulo-ocular reflex (VOR) and eye bl...