2 November 2007. A new study of monkeys suggests that the dorsolateral prefrontal cortex (DLPFC), a brain region that has repeatedly been implicated in the pathophysiology of negative symptoms in schizophrenia, is crucial to detecting and resolving conflicts in information processing during cognitive tasks. However, the study also found that the anterior cingulate cortex (ACC), another brain area frequently associated with deficits in conflict resolution in patients with schizophrenia or brain injury (Kerns et al., 2005; Gehring and Knight, 2000), appears to play no important role in these processes.

In the new research, by Farshad Mansouri and Keiji Tanaka at the Cognitive Brain Mapping Laboratory at Saitama University in Japan, in collaboration with Mark Buckley of Oxford University, monkeys were trained to perform a modified version of the Wisconsin Card Sorting Test (WCST), a common neuropsychological instrument used to detect frontal lobe dysfunction. Schizophrenia patients have been reported to exhibit decreased frontal lobe activity, or “hypofrontality,” in neuroimaging studies, and their poor performance on the WCST has made the test a standard tool in schizophrenia research (e.g., Prentice et al., 2007).

Making matches
In the WCST, a subject is asked to match cards featuring colored symbols in various shapes, colors, and numbers to stimulus cards, on which the symbols may match in color, shape, number, or some combination of the three. The subject is not provided beforehand with a rule for proper matches, but after each match he or she is told whether the choice is correct or incorrect. After several trials, subjects deduce the matching rule, but the person administering the test then repeatedly changes the rule, again without informing the subject. The degree of behavioral flexibility displayed by the subject in resolving conflicts between the rules and reducing matching errors is a sensitive measure of frontal lobe dysfunction.

In the modified WCST used by Mansouri and colleagues, monkeys were presented with three test items, each of which could match or not match a sample for color and/or shape. In a “high-conflict” condition, the sample matched one of the test items in color and another in shape, and did not match a third test item in either category. In a “low-conflict” condition, the sample matched one test item in color and shape and did not match the other test items at all.

Compensating for conflict
After 8 months of training, the monkeys’ brains were lesioned, either along the principal sulcus of the DLPFC or the anterior cingulate sulcus. Both groups performed as well as a group of control monkeys on the test, and all three groups showed slower reaction times in the presence of conflict.

According to a “conflict-monitoring hypothesis” proposed in the literature on cognitive control (Botvinick et al., 2004), the ACC plays an active role in detecting conflict and errors in cognitive tasks, and activates compensatory mechanisms to resolve conflict and improve performance. As reported online in the October 25 Sciencexpress, the Mansouri group tested this hypothesis by comparing the performance of both groups of lesioned monkeys when a low-conflict trial was followed by a high-conflict trial (an “LH pairing”) and when a high-conflict trial was followed by a second high-conflict trial (“HH pairing”). The researchers found that the control and ACC-lesioned groups showed a significant performance improvement in the second trial of HH pairings versus the second trial of LH pairings, suggesting that these monkeys adjusted their behavior in response to the high-conflict condition of the HH pairings’ first trial. But the DLPFC-lesioned group made no such adjustments, and performed equivalently in the LH and HH pairings.

A memory mechanism?
To explore how neural activity in the DLPFC might contribute to conflict-induced behavioral adjustment, the team made a series of single-cell recordings in two monkeys in which the DLPFC was intact while the monkeys performed the modified WCST. Using data only from pairings in which the correct response was made in the second trial of a pairing, the researchers found that in 15 out of 146 cells the conflict condition was strongly correlated with spike frequency: spikes were higher in 11 cells in the LH pairings, and higher in four cells in the HH pairings. Because these spike patterns were consistently observed during an eye-fixation period between each trial in a pairing, the authors propose that they reflect a mnemonic neural representation of the conflict level of a given trial that can be called upon to guide behavior in a subsequent trial.

Mansouri and colleagues conclude that the DLPFC, but not the ACC, is necessary for conflict-induced behavioral adjustment. This finding meshes well with some recent neuropsychological data (Fellows and Farah, 2005), but would seem to stand in contrast to many neuroimaging studies (e.g., MacDonald et al., 2000; Pardo et al., 1990; Peterson et al., 1999) that postulate a major role for the ACC in conflict resolution, and also in the psychopathology of schizophrenia.—Peter Farley.

Reference:
Mansouri FA, Buckley MJ, Tanaka K. Mnemonic function of the lateral prefrontal cortex in conflict-induced behavioral adjustment. Sciencexpress. 2007 Oct 25. Abstract