Motor programming of finger sequences of different complexity

Abstract

In a response precuing task, we used behavioral and electrophysiological measures – Contingent Negative Variation (CNV) and the readiness potential (RP) – to investigate the programming of three-element response sequences of different complexity. Precuing effects on foreperiod CNV and RT indicated the use of advance information about response hand and response sequence. Crucially, with advance information about both hand and sequence, heterogeneous response sequences (e.g., 1 → 2 → 2) elicited larger foreperiod CNV activity over medial motor areas than homogeneous response sequences (e.g., 1 → 2 → 3), whereas CNV activity over lateral motor areas was not influenced by sequence complexity. It was only before response execution that lateral but not medial RP activity was stronger for heterogeneous than homogeneous response sequences. Both behavioral and electrophysiological measures indicated finger-order dependent influences on the duration of on-line response programming during response execution.

Introduction

It is generally assumed that a central motor program, which consists of pre-structured motor commands, controls the production of sequential actions (e.g., Keele, 1968; for a review, see Rhodes et al., 2004). This view has received support in behavioral and electrophysiological studies examining the production of response sequences of different length (e.g., Hackley and Miller, 1995, Henry and Rogers, 1960, Klapp, 1995, Klapp, 2003, Magnuson et al., 2004, Rosenbaum et al., 1984, Schröter and Leuthold, 2008, Sternberg et al., 1978, Verwey, 2003). However, little is known about the (brain) mechanisms underlying both the planning and execution of response sequences that differ in structural complexity, that is, response sequences for which only the relationship among response elements but not the number of response elements varies.

The main aim of the present study was to investigate motor programming of three-element response sequences that differ in structural complexity and finger order by analyzing event-related brain potentials (ERPs). Previous ERP studies examining response sequence effects have commonly analyzed the readiness potential (RP) (Kornhuber and Deecke, 1965; for a review, see Shibasaki and Hallett, 2006). Studies concerned with sequence programming revealed larger RPs before sequential, multi-element than single-element responses (e.g., Benecke et al., 1985, Kristeva, 1984, Simonetta et al., 1991). To our knowledge, the ERP study of Prescott (1986) is the only one that examined the production of unilateral response sequences of different structural complexity. Specifically, in a forewarned RT (S1–S2) paradigm, participants were asked to produce in separate blocks of trials either homogeneous (1 → 2 → 3) or heterogeneous (1 → 3 → 2) response sequences consisting of index (1), middle (2) and ring finger (3) key presses. Most relevant for present purposes, Prescott found a larger response-locked RP preceding heterogeneous than homogeneous finger sequences. However, an analysis of the Contingent Negative Variation (CNV; Walter et al., 1964), which reflects preparatory motor activity (e.g., Brunia, 2003, Rohrbaugh and Gaillard, 1983), but also non-motoric anticipatory processes (e.g., Falkenstein et al., 2003, Van Boxtel and Böcker, 2004), was unaffected by structural complexity. The CNV and RP findings of Prescott seem to suggest that structural complexity influences mainly late execution-related motor processes in the brain. However, possible influences of structural complexity on preparatory brain processes are still relatively unknown, because functional neuroimaging and transcranial magnetic stimulation (TMS) studies examined mostly sequence execution (e.g., Chen et al., 1997, Gerloff et al., 1997, Gerloff et al., 1998, Haaland et al., 2004, Van Oostende et al., 1997; but see Elsinger et al., 2006).

Another open issue concerns the brain areas contributing to the execution of finger response sequences of different structural complexity. Thus, Kitamura et al. (1993) suggested that both the supplementary motor area (SMA) and bilateral sensorimotor areas contribute to sequence programming, based on their finding of a larger late RP over both midline and bilateral precentral electrodes before sequential as compared to simultaneous index and middle finger movements. In accord with Kitamura et al.’s proposal, neuroimaging, neurophysiological, and repetitive TMS studies demonstrated influences of structural complexity on both SMA and primary motor cortex (M1) (e.g., Colebatch et al., 1991, Gerloff et al., 1997, Gerloff et al., 1998, Lu and Ashe, 2005, Van Oostende et al., 1997, Wexler et al., 1997).¹ Yet, findings from Arunachalam et al. (2005) indicate that the control of finger sequences by SMA and M1 might depend on finger order. Thus, sequential taps with two fingers were found to be faster in the little-to-index finger direction (4 → 1) than the reverse direction (1 → 4), for both adjacent and non-adjacent fingers. Arunachalam and colleagues speculated that this tapping effect reflects the temporal gradient of finger programming in grasping objects. Moreover, when applying TMS to contralateral M1 after the first tap of a two-finger tapping sequence, disruption of the next tap was larger in the faster (2 → 1) than the slower tapping direction (1 → 2). The authors concluded that tapping in the faster direction is controlled by M1 and in the slower direction by SMA.