The impact of basal ganglia lesions on sensorimotor synchronization, spontaneous motor tempo, and the detection of tempo changes

Abstract

The basal ganglia (BG) are part of extensive subcortico-cortical circuits that are involved in a variety of motor and non-motor cognitive functions. Accumulating evidence suggests that one specific function that engages the BG and associated cortico-striato-thalamo-cortical circuitry is temporal processing, i.e., the mechanisms that underlie the encoding, decoding and evaluation of temporal relations or temporal structure. In the current study we investigated the interplay of two processes that require precise representations of temporal structure, namely the perception of an auditory pacing signal and manual motor production by means of finger tapping in a sensorimotor synchronization task. Patients with focal lesions of the BG and healthy control participants were asked to align finger taps to tone sequences that either did or did not contain a tempo acceleration or tempo deceleration at a predefined position, and to continue tapping at the final tempo after the pacing sequence had ceased. Performance in this adaptive synchronization-continuation paradigm differed between the two groups. Selective damage to the BG affected the abilities to detect tempo changes and to perform attention-dependent error correction, particularly in response to tempo decelerations. An additional assessment of preferred spontaneous, i.e., unpaced but regular, production rates yielded more heterogeneous results in the patient group. Together these findings provide evidence for less efficient processing in the perception and the production of temporal structure in patients with focal BG lesions. The results also support the functional role of the BG system in attention-dependent temporal processing.

Introduction

Individuals continuously adjust their behavior to environmental changes. The underlying adaptive process unfolds in time and involves the cyclic processing of motor and sensory information [1]. The question arises whether this cyclic processing is merely intrinsically temporal or to what extent temporal information is actually processed and used as a source of information in itself. Some internal representation of temporal structure is a prerequisite for efficient timing, which in turn bears the potential to optimize adaptive processes. Efficient timing of behavior implies temporally appropriate reactive and proactive actions. The latter depend on anticipation and predictions about the temporal structure of future changes or events as well as the ability to temporally align behavior with these events. Both aspects converge in sensorimotor synchronization (SMS).

SMS is a specific form of adaptive interaction with the environment. It is an extensively studied process that merges motor and non-motor components in a single setting. SMS can be characterized as the temporal coordination of a motor rhythm with an external rhythm (for reviews see [2], [3]). This temporal coordination can be conceived as synchronization. Synchronization denotes the “adjustment of rhythms of oscillating objects due to their weak interaction” [4]. An oscillation is defined by its phase, relative to another oscillation, and period, and provides a means to describe the temporal relation of the events that constitute a rhythm in terms of frequency, i.e., the repetition of similar events in a specific amount of time. Complex rhythmic activity and synchronization between different rhythms constitute central aspects of life. Physiological rhythms interact continuously with each other and the environment in order to mediate between internal and external events [5]. In cognition, this implies adaptation of internally-generated to external rhythms via unidirectional coupling which eventually leads to stimulus-driven synchronization of behavior.

Both SMS and temporal processing have been modeled on the basis of oscillations and the fundamental dissociation of automatic and controlled sub-processes. For example, temporal processing is hypothesized to rest on the coincidental activation of medium spiny neurons in the BG by ensembles of cortical neural oscillations [6], [7]. In SMS, the period of an adaptive oscillatory timekeeper is assumed to reflect the temporal structure of the pacing signal, thereby establishing a reference for the timing of successive motor commands [8]. In this context, the current study investigates a combination of these aspects by providing BG patients and controls with an oscillatory perceptual input whose temporal structure needs to be exploited to generate oscillatory motor behavior. Error correction mechanisms adjust the phase and period of the internal timekeeper oscillation if it is confronted with a perturbation, i.e., a change in the temporal structure of the pacing signal. These error correction mechanisms are dissociable based on their dependence on attention, the intention to adapt, and awareness of a tempo change [9]. Whereas phase correction depends solely on the intention to maintain synchrony and can therefore be considered automatic, period correction depends also on attention and awareness of the tempo change.

A comparable dissociation between automatic and controlled mechanisms has been proposed with respect to temporal processing [10]. More specifically, the cerebellum (CE) performs automatic or pre-attentive, short-range, event-based temporal processing, while in parallel, the BG and cortico-striato-thalamo-cortical circuitry engage in attention-dependent, longer-range, interval-based temporal processing [7], [11], [12]. Attention, or the ability to attentively detect a change in temporal structure may therefore not only decide upon the type of error correction, but also upon the primary temporal processing system. Temporal structure may thereby influence cognitive processes, e.g., as the basis for attentional set shifting and sequence coordination [13]. This view is consistent with Dynamic Attending Theory DAT [14], which proposes that attention can be modeled as a self-sustained oscillation capable of entrainment. Within the framework of DAT, the temporal structure of a stimulus guides the allocation of attentional resources thereby evoking stimulus-driven attending [15]. On this basis, attentive adaptation to tempo change in SMS would depend on the parallel engagement of pre-attentive and attention-dependent temporal processing systems, as well as phase and period correction to adjust internal timekeeper and/or attention oscillations.

The role of the BG in temporal processing and in SMS has been investigated primarily in patients with Parkinson's disease (PD), albeit with mixed results [16], [17], [18]. Studies involving PD patients also suggest difficulties in beat extraction and the comparison of rhythmic sequences [19]. However, PD is a progressive neurodegenerative disease, and besides medication and different PD subgroups [20] some of the heterogeneity of the respective results may be due to the variable extent of cortical damage in this population, which can be minimal or absent in patients with BG lesions [21], [22]. Studies on SMS involving patients with focal lesions of the BG are scarce. Aparicio et al. [23] used a synchronization–continuation paradigm and found no evidence for impaired temporal processing performance in this group. Different tasks, stimulus characteristics, and cognitive sets add further complexity to the identification of specific BG and cerebellar temporal processing functions [24].

Besides attention, temporal range seems to be an important factor, as temporal processing evolves across different timescales that may map onto different physiological and psychological mechanisms [10], [24], [25]. A well-defined boundary between short-range and longer-range temporal processing remains elusive, although values around 500 ms [26] and close to 1000 ms [27] have been suggested. Fraisse [28] emphasizes that synchronization is most stable for tempi between 400 and 800 ms, while the intermediary tempo of 600 ms is most representative for unpaced, spontaneous motor activity. This has been confirmed in more recent studies [29] that also found a correlation between individual spontaneous motor tempo (SMT) and a preferred perceptual tempo [30]. It corresponds to the “indifference interval” or “indifference zone” that is neither systematically overestimated nor underestimated [28], [31]. A tempo of 600 ms is within the range for optimal pulse sensation [32] and tempo sensitivity, for a review see [33], [34]. Although originating from a different perspective, Karmarkar and Buonomano [35] hypothesize that temporal processing between 400 and 800 ms may be accurately performed by mechanisms underlying both time perception and time estimation. Hence, the SMS task in the current study incorporated a base tempo of 600 ms and tempo changes that induced shifts relative to this base tempo. This procedure should perturb the synchronization of internal timekeeper and/or attention oscillations. We expected that damage to the cortico-striato-thalamo-cortical attention-dependent temporal processing system due to BG lesions should lead to difficulties in the evaluation of temporal structure and consequently in maintaining attentive synchrony. These difficulties should compromise the ability to detect and to assess a tempo change which should in turn lead to a lesser degree of attention-dependent period correction during SMS in the patient group, while automatic phase correction should be preserved.

We assessed SMT before and after the main SMS task to explore whether the SMT of patients with BG lesions differs from that of healthy controls and whether it would be influenced by the intervening SMS task. For example, stronger reliance on the unimpaired cerebellar short-range system in the patient group may be reflected by a preference for faster SMT rates. SMT is not constrained by simple biomechanical mechanisms [28] and in the absence of an external pacemaker it has to rely on internally generated temporal structure and simultaneous monitoring of temporal regularity. We hypothesized that faster SMT rates in the patient group could reflect stronger reliance on the unaffected cerebellar short range temporal processing system. BG lesions should compromise both the consistency of internally generated temporal structure and the monitoring process, which in turn should lead to increased tapping variability in the patient group. In general, a better understanding of these fundamental mechanisms and the way they are altered in this specific patient group is not only important for modeling the mechanisms underlying the adaptive interaction with the environment but may also be helpful in optimizing the temporal structure of compensatory strategies used in therapeutic settings.