Conflict processing in the anterior cingulate cortex constrains response priming

Abstract

A prominent function of the anterior cingulate cortex (ACC) is to process conflict between competing response options. In this study, we investigated the role of conflict processing in a response-priming task in which manual responses were either validly or invalidly cued. Examining electrophysiological measurements of oscillatory brain activity on the source level, we found response priming to be related to a beta power decrease in the premotor cortex and conflict processing to be linked to a theta power increase in the ACC. In particular, correlation of oscillatory brain activities in the ACC and the premotor cortex showed that conflict processing reduces response priming by slowing response time in valid trials and lowering response errors in invalid trials. This relationship emerged on a between subjects level as well as within subjects, on a single trial level. These findings suggest that conflict processing in the ACC constrains the automatic priming process.

Introduction

The anterior cingulate cortex (ACC) plays an important role in the adjustment of cognitive control in conflict resolution between relevant and irrelevant information (for a review, see Mansouri et al., 2009). Such cognitive control is needed when goal-irrelevant automatic or prepotent information interferes with goal-relevant information in the selection of the appropriate behavior. For example, the automatic response to extinguish fire is to pour water on the fire source. However, this automatic response is best overridden by throwing table salt, baking soda, or a damp towel on the fire source in the case of an oil fire in the kitchen.

Two current theories of how the ACC processes conflict are the conflict-monitoring theory and the regulatory theory of ACC function. The former theory asserts that the ACC detects and signals conflict to other areas, such as the dorsolateral prefrontal cortex (DLPFC), which then implement cognitive control (Botvinick et al., 2004, Carter and van Veen, 2007). The conflict-monitoring theory assumes a monitoring role for the ACC in a broad range of situations in which conflict might occur from early perceptual to late motor processing. On the other hand, the latter theory asserts that the ACC exerts cognitive control itself and, thus, has a more direct response-regulatory function in resolving conflict (Paus et al., 1998, Posner and DiGirolamo, 1998). The regulatory theory is primarily based on evidence that indicates an important role for the ACC in response selection at the late motor stage. Actually, although the ACC has been shown to signal conflict already from the early perceptual stage of information processing, it is most strongly responsive to conflict at the late motor stage (Milham et al., 2001, van Veen et al., 2001) and post-response conflict due to response errors (Braver et al., 2001, Carter et al., 1998). Consistently, the ACC with its strong connectivity to motor areas has been suggested to be in part a motor structure (Matsumoto et al., 2003, Paus, 2001).

Examining the effects of response conflict and error processing on event-related brain potentials (ERPs), ACC activity has been found to be related to the frontocentral N2 component and error-related negativity (ERN) (Debener et al., 2005, van Veen and Carter, 2002). In addition, induced frontocentral ACC theta oscillations have been shown to indicate error detection (Luu et al., 2004, Marco-Pallarés et al., 2008, Trujillo and Allen, 2007) and to signal conflict of prepotent responses in the classical Stroop task (Hanslmayr et al., 2008) and the Eriksen flanker task (Cavanagh et al., 2009). Moreover, during conflict in prepotent responding, a cross-talk between the ACC and the DLPFC as reflected by long-range theta phase coupling between these brain areas has been found suggesting that when conflict of a prepotent response arises the ACC engages the DLPFC to implement control (Cavanagh et al., 2009, Hanslmayr et al., 2008). In the present response-priming study, we sought to examine the role of theta oscillations in the selection of validly and invalidly primed motor responses.

In response priming, response processing is faster when a target to respond with one hand is shortly preceded by a valid cue that primes the same-hand response and more error-prone when it is preceded by an invalid cue that primes the other-hand response (Sterr and Dean, 2008). Response priming has been studied with neuroimaging (Deiber et al., 1996, Lee et al., 1999) and ERPs (Leuthold and Jentzsch, 2002, Mathews et al., 2006). Consistently, lateralized activity in the premotor cortex has been found to be related to response priming. For example, Dehaene et al. (1998) reported that subliminal primes trigger lateralized readiness potentials (LRP) indicating a preparatory activation of the primed response. In addition to response-related evoked activity, there is good evidence that induced beta oscillations in the motor cortex are linked to movement preparation and movement execution (Kaiser et al., 2001, Pfurtscheller and Neuper, 1997). Consistently, in an inhibition-of-return task, beta power in the motor cortex has been found to be responsive both to cue and target presentation (Pastötter et al., 2008). Thereby, induced beta oscillations and evoked readiness potentials are suggested to arise from different neuronal mechanisms providing different information in movement preparation (Shibasaki and Hallett, 2006). In the present study, we sought to examine the role of beta oscillations in response priming.

In the present EEG experiment, a localization-discrimination task was used in which, half-and-half, manual responses were primed by cues or not. Cuing was exogenous (peripheral cues) and uninformative (50% cue validity). Conflict arose when an actual response (e.g., a left-hand response) had been preceded by an invalid cue (e.g., a cue that primed a right-hand response). We examined induced brain oscillations on the source level and hypothesized that beta power in the premotor cortex is linked to response priming and theta power in the ACC is related to conflict processing. To examine whether conflict processing in the present task affects the priming process, we analyzed the relation of theta activity in the ACC and beta activity in motor sources both between and within subjects.