functional specialization within macaque dorsolateral prefrontal cortex for the maintenance of task...

10
Functional Specialization within Macaque Dorsolateral Prefrontal Cortex for the Maintenance of Task Rules and Cognitive Control Sabeeha Hussein 1,2 , Kevin Johnston 1,2 , Brandon Belbeck 1,2 , Stephen G. Lomber 1,2,3 , and Stefan Everling 1,2,3 Abstract The abilities of switching between and maintaining task rules are fundamental aspects of goal-oriented behavior. The PFC is thought to implement the cognitive processes underling such rule-based behavior, but the specific contributions of the several cytoarchitecturally distinct subfields of PFC remain poorly understood. Here, we used bilateral cryogenic deactiva- tion to investigate the relative contributions of two regions of the dorsolateral PFC (dlPFC)the inferior dlPFC (idlPFC) area, consisting of the cortex lining the caudal principal sulcus, and the dorsally adjacent superior dlPFC (sdlPFC)to different aspects of rule-based behavior. Macaque monkeys performed two variants of a task that required them to alternate unpredic- tably between eye movements toward (prosaccade) or away from (antisaccade) a visual stimulus. In one version of the task, the current rule was overtly cued. In the second, the task rule was uncued, and successful performance required the animals to detect rule changes on the basis of reward outcome and subsequently maintain the current task rule within working memory. Deactivation of the idlPFC impaired the monkeysʼ abil- ity to perform pro- and antisaccades in the uncued task only. In contrast, deactivation of the sdlPFC had no effect on perfor- mance in either task. Combined deactivation of idlPFC and sdlPFC impaired performance on antisaccade, but not prosac- cade, trials in both task variants. These results suggest that the idlPFC is required for mnemonic processes involved in main- tenance of task rules, whereas both idlPFC and sdlPFC together are necessary for the deployment of the cognitive control re- quired to perform antisaccades. Together, these data support the concept of a functional specialization of subregions within the dlPFC for rule-guided behavior. INTRODUCTION Goal-oriented behavior requires the ability to apply rules that specify the most appropriate response to a stimulus within a given context. Rules may be externally cued with fixed, relatively simple stimulusresponse mappings, such as go when a traffic light is green and stop when it is red. In other instances, the appropriate rule is not directly specified by environmental stimuli, but must be selected, maintained, or switched on the basis of be- havioral outcomes. Such uncued rule-related processes are commonly assessed using the WCST (Milner, 1963), which requires subjects to acquire a task rule by trial and error and either maintain or switch rules depending on stayor switchfeedback. Single neuron recordings in the macaque monkey have found neural correlates of both cued and uncued rule representations in PFC (Tsujimoto, Genovesio, & Wise, 2012; Johnston, Levin, Koval, & Everling, 2007; Johnston & Everling, 2006b; Everling & Desouza, 2005; Genovesio, Brasted, Mitz, & Wise, 2005; Wallis & Miller, 2003; Wallis, Anderson, & Miller, 2001; Asaad, Rainer, & Miller, 2000; White & Wise, 1999). In general, these studies suggest little differentiation of such representations across the several subregions comprising this area in primates. In contrast, lesion studies of macaque PFC have provided compelling evidence for regional specialization of cued and uncued rule representations. Ventral prefrontal and orbitofrontal areas are critical for the application of cued rules (Bussey, Wise, & Murray, 2001; Murray, Bussey, & Wise, 2000; Passingham, Toni, & Rushworth, 2000), whereas dorsolateral PFC (dlPFC) lesions either do not or only mildly impair this ability (Gaffan & Harrison, 1989; Petrides, 1982). Recently, Buckley and colleagues trained monkeys on a simplified analog of the WCST that required switching between shape and color matching rules during selection of visual stimuli and investigated the effects of le- sions of several prefrontal subregions on task performance (Buckley et al., 2009). Lesions of inferior dlPFC (idlPFC) in the area of the principal sulcus were found to selectively increase error rates after imposed delay periods between trials. Accordingly, the authors proposed that this area is functionally specialized for the selection and active mainte- nance of uncued rules in working memory. Lesions of cor- tex dorsal to the principal sulcus, which they labeled 1 University of Western Ontario, 2 Brain and Mind Institute, London, Ontario, Canada, 3 Robarts Research Institute, London, Ontario, Canada © 2014 Massachusetts Institute of Technology Journal of Cognitive Neuroscience 26:9, pp. 19181927 doi:10.1162/jocn_a_00608

Upload: stefan

Post on 09-Feb-2017

214 views

Category:

Documents


2 download

TRANSCRIPT

Functional Specialization within Macaque DorsolateralPrefrontal Cortex for the Maintenance of Task

Rules and Cognitive Control

Sabeeha Hussein1,2, Kevin Johnston1,2, Brandon Belbeck1,2,Stephen G. Lomber1,2,3, and Stefan Everling1,2,3

Abstract

■ The abilities of switching between and maintaining taskrules are fundamental aspects of goal-oriented behavior. ThePFC is thought to implement the cognitive processes underlingsuch rule-based behavior, but the specific contributions of theseveral cytoarchitecturally distinct subfields of PFC remainpoorly understood. Here, we used bilateral cryogenic deactiva-tion to investigate the relative contributions of two regions ofthe dorsolateral PFC (dlPFC)—the inferior dlPFC (idlPFC) area,consisting of the cortex lining the caudal principal sulcus, andthe dorsally adjacent superior dlPFC (sdlPFC)—to differentaspects of rule-based behavior. Macaque monkeys performedtwo variants of a task that required them to alternate unpredic-tably between eye movements toward (prosaccade) or awayfrom (antisaccade) a visual stimulus. In one version of the task,the current rule was overtly cued. In the second, the task rule

was uncued, and successful performance required the animalsto detect rule changes on the basis of reward outcome andsubsequently maintain the current task rule within workingmemory. Deactivation of the idlPFC impaired the monkeysʼ abil-ity to perform pro- and antisaccades in the uncued task only.In contrast, deactivation of the sdlPFC had no effect on perfor-mance in either task. Combined deactivation of idlPFC andsdlPFC impaired performance on antisaccade, but not prosac-cade, trials in both task variants. These results suggest that theidlPFC is required for mnemonic processes involved in main-tenance of task rules, whereas both idlPFC and sdlPFC togetherare necessary for the deployment of the cognitive control re-quired to perform antisaccades. Together, these data supportthe concept of a functional specialization of subregions withinthe dlPFC for rule-guided behavior. ■

INTRODUCTION

Goal-oriented behavior requires the ability to apply rulesthat specify the most appropriate response to a stimuluswithin a given context. Rules may be externally cued withfixed, relatively simple stimulus–response mappings,such as go when a traffic light is green and stop whenit is red. In other instances, the appropriate rule is notdirectly specified by environmental stimuli, but must beselected, maintained, or switched on the basis of be-havioral outcomes. Such uncued rule-related processesare commonly assessed using the WCST (Milner, 1963),which requires subjects to acquire a task rule by trialand error and either maintain or switch rules dependingon “stay” or “switch” feedback.

Single neuron recordings in the macaque monkey havefound neural correlates of both cued and uncued rulerepresentations in PFC (Tsujimoto, Genovesio, & Wise,2012; Johnston, Levin, Koval, & Everling, 2007; Johnston& Everling, 2006b; Everling & Desouza, 2005; Genovesio,Brasted, Mitz, & Wise, 2005; Wallis & Miller, 2003; Wallis,

Anderson, & Miller, 2001; Asaad, Rainer, & Miller, 2000;White & Wise, 1999). In general, these studies suggestlittle differentiation of such representations across theseveral subregions comprising this area in primates. Incontrast, lesion studies of macaque PFC have providedcompelling evidence for regional specialization of cuedand uncued rule representations. Ventral prefrontal andorbitofrontal areas are critical for the application of cuedrules (Bussey, Wise, & Murray, 2001; Murray, Bussey, &Wise, 2000; Passingham, Toni, & Rushworth, 2000),whereas dorsolateral PFC (dlPFC) lesions either do not oronly mildly impair this ability (Gaffan & Harrison, 1989;Petrides, 1982). Recently, Buckley and colleagues trainedmonkeys on a simplified analog of the WCST that requiredswitching between shape and color matching rules duringselection of visual stimuli and investigated the effects of le-sions of several prefrontal subregions on task performance(Buckley et al., 2009). Lesions of inferior dlPFC (idlPFC) inthe area of the principal sulcus were found to selectivelyincrease error rates after imposed delay periods betweentrials. Accordingly, the authors proposed that this area isfunctionally specialized for the selection and active mainte-nance of uncued rules in working memory. Lesions of cor-tex dorsal to the principal sulcus, which they labeled

1University of Western Ontario, 2Brain and Mind Institute,London, Ontario, Canada, 3Robarts Research Institute, London,Ontario, Canada

© 2014 Massachusetts Institute of Technology Journal of Cognitive Neuroscience 26:9, pp. 1918–1927doi:10.1162/jocn_a_00608

superior dlPFC (sdlPFC), had no effects. Such regionalspecialization of function is consonant with the well-established anatomical segregation of PFC in primates,based on cytoarchitecture and connectivity (Petrides &Pandya, 1999; Preuss & Goldman-Rakic, 1991; Barbas &Pandya, 1989).An important aspect of rule-based behavior not cap-

tured by the WCST is the relative automaticity of certainstimulus–response pairings that is characteristic of morenaturalistic situations. Some rules may be overlearnedand more automatic, whereas others require more con-trol. Both Stroop (Stroop, 1935) and stimulus–responsecompatibility tasks (Hommel & Prinz, 1997) requiresubjects to override automatic processes in favor of con-trolled responses. In these tasks, incongruent stimulus–response processes are characterized by increases inRT and performance decrements relative to congruentprocesses. Functional imaging studies in humans haveshown that the dlPFC is activated during delay periods inthe Stroop task when subjects have to prepare to namethe print color of a color word (Donohue, Wendelken, &Bunge, 2008; MacDonald, Cohen, Stenger, & Carter,2000) and in the antisaccade task (Munoz & Everling,2004) when they prepare to look away from a sub-sequently presented visual stimulus (Brown, Vilis, &Everling, 2007; Desouza, Menon, & Everling, 2003). Onthe basis of these findings, it has been proposed thatthe ventrolateral PFC is involved in rule maintenancefor stimulus–response associations, whereas the dlPFC isengaged in rule maintenance when subjects must main-tain the goal of overriding a strong response tendency(Bunge, 2004).Taken together, evidence from lesion studies in mon-

keys and fMRI investigations in human participants sug-gests a parcellation of rule-related processes in PFC. Therelative contribution of dlPFC subregions to rule-basedbehavior under conditions requiring maintenance andswitching of automatic and controlled processes, however,remains poorly understood. To address this, we reversiblyand separably deactivated parts of the idlPFC and sdlPFCwhile monkeys performed tasks requiring the implemen-tation of cued or uncued rules. We used bilaterally im-planted cryoloops (Lomber, Payne, & Horel, 1999) toindependently deactivate the idlPFC (cortex in the caudalprincipal sulcus) or the dlPFC (cortex dorsal to the princi-pal sulcus; Figure 1A). The area of cortex deactivated bythese loops was consistent with, but of lesser area, thanthe idlPFC and sdlPFC lesions in the study of Buckleyet al. (2009). Animals performed a task that required themto switch between blocks of prosaccades (requiring asaccade toward a flashed peripheral stimulus) and anti-saccades, which required the suppression of a saccadetoward the stimulus in favor of one toward the oppositedirection (Everling & Desouza, 2005; Figure 1B). In a cuedcondition, the color of the fixation point instructed themonkeys which task to perform on each trial. In an uncuedcondition, the animals had to acquire the current rule

based on reward feedback delivered following each trial.After 15–25 correct trials, the rule was changed withoutwarning and the animals had to switch to the other rule.This task thus required the animals to switch between andmaintain within working memory task rules specifyingrelatively automatic and more controlled responses underexternally cued and uncued conditions.

METHODS

Surgeries

Data were collected from two male rhesus monkeys(Macaca mulatta) weighing 10 and 12 kg. In a thirdrhesus monkey, one of the idlPFC loops became pluggedimmediately after the cooling loop surgery, and no datacould be obtained from this animal for this study. All pro-cedures were carried out in accordance with the guide-lines of the Canadian Council of Animal Care Policy onthe Use of Laboratory Animals and a protocol approvedby the Animal Use Subcommittee of the University ofWestern Ontario Council on Animal Care. In the firstaseptic surgery, a plastic head restraint was implanted(for details, see Johnston & Everling, 2006a). Animalsreceived analgesics and antibiotics postoperatively andwere closely monitored by a university veterinarian. Fol-lowing recovery, monkeys were trained to perform thecued pro-/antisaccade and uncued pro-/antisaccade tasks.After the animals reached a criterion of ∼75% correct per-formance in the uncued task, they underwent a secondsurgery in which stainless steel cryoloops (6 mm ×3 mm) were implanted bilaterally into the posterior por-tion of the principal sulcus (idlPFC loop, covering partsof area 46 and 9/46 and anterior parts of area 8A) andbilaterally on the cortex that lies immediately dorsal tothe principal sulcus (sdlPFC loop, covering parts of area9/46d and area 9; Figure 1A). On the basis of a spread ofdeactivation of 2 mm around the loops (Lomber et al.,1999), we estimated that the idlPFC loop deactivated por-tions of areas 46, 9/46d and v, and 8, whereas the sdlPFCloop deactivated portions of areas 46, 9, 9/46d, and 8(Figure 1B). Cryoloops were constructed from 23-gaugehypodermic stainless steel tubing. The procedures for themanufacturing, surgery, and use of cryoloops have beendescribed in detail (Lomber et al., 1999).

Task

Figure 1C illustrates the pro-/antisaccade switch task.Each trial began with the presentation of a fixation spotat the center of a CRT monitor screen. The animals wererequired to fixate it within a 0.5° × 0.5° window for arandom period of 1100–1400 msec. A visual stimulus(0.2°) was then presented pseudorandomly with equalprobability 8° to the left or right of the fixation spot. Toobtain a liquid reward, monkeys were required to gen-erate a saccade within 500 msec, either to the stimulus

Hussein et al. 1919

location on prosaccade trials or away from the stimulus toits mirror location on antisaccade trials (within a 5° × 5°window). On successful trials, a liquid reward was given200 msec after saccade end. Task switches occurredon a random trial once animals had performed between15 and 25 correct responses. In the cued task, a greenfixation spot instructed prosaccades and a red fixa-tion spot was the instruction to perform antisaccadesfor Monkey G. These instructions were reversed for

Monkey B. In the uncued task, the fixation spot wasalways white, and the monkeys had to acquire the taskrule based on reward delivery (“stay”) or omission(“switch”). Horizontal and vertical eye positions wererecorded at 500 Hz using an Eyelink II system (SRResearch, Kanata, Canada).

Prefrontal Deactivations

We collected data from a total of 103 sessions in whicheither (1) no area was deactivated, (2) the idlPFC wasbilaterally deactivated, (3) the sdlPFC was bilaterally de-activated, or (4) both areas were bilaterally deactivated.Cued and uncued versions of the switch task were runon separate days. During cooling sessions, methanol atroom temperature was pumped through teflon tubingpassing through a methanol ice bath, which was re-duced to subzero temperatures by the addition of dryice. Chilled methanol was then pumped through a cryo-loop and returned to the same reservoir from whichit came. Cryoloop temperature was monitored by anattached microthermocouple. At the beginning of eachcooling session, either two or four pumps (one for eachcyroloop) were turned on. It took on average of 85 secto lower the temperature of the loops to 3°C. After 3–5 min, the experimental task began. We maintained cryo-loop temperature in the range of 1–5°C by adjusting theflow rate of the pump. This range was chosen to inactivateas large an area of cortical tissue as possible (the extentof inactivated tissue is limited to a range of ∼2 mm whencryoloop temperature is reduced to 1–3°C; Lomber et al.,1999) while avoiding potentially harmful subzero tem-peratures at the cortical surface. This ensured that thecortical tissue adjacent to the cryoloop reached tempera-tures below the 20°C threshold for neuronal deactivation(Adey, 1974). Even after months or years of daily coolingdeactivations, there is no alteration to either the structureor function of underlying cortex (Yang, Kennedy, Lomber,Schmidt, & Rothman, 2006). Each experimental sessionlasted between 60 and 70 min to ensure consistent moti-vation of the animals and task performance. We did notobserve any indications of discomfort from the animalsduring any of the cooling sessions. On average, monkeysperformed 50 task switches during each session. Postses-sion, monkeys received liquid until satiation, after whichthey were returned to their home cages. Daily recordsof the weight and health status of the monkeys werekept, and additional fruit was provided.

Data Analysis

Data analysis was performed using custom-designed soft-ware running in Matlab (Mathworks, Natick, MA). Saccadeonset was defined as the time at which eye velocity ex-ceeded 30°/sec and saccade end as the time at which radialvelocity fell below 30°/sec. Trials with broken or incorrectfixation were excluded from further analyses as were trials

Figure 1. Placement of cryoloops and experimental paradigm.(A) Photo of right PFC of Monkey B during cryoloop surgery. Onecryoloop was implanted in the principal sulcus (idlPFC), and theother one was placed over the cortex dorsal to the principal sulcus(sdlPFC). (B) Depiction of estimated area of deactivated cortex,shown in relation to PFC cytoarchitecture as proposed by Petridesand Pandya (1999). AS, arcuate sulcus; PS, principal sulcus.(C) Schematic diagram of the switch task. Monkeys had to alternatebetween blocks of prosaccades (saccade toward a visual stimulus)and blocks of antisaccades (saccade away from the stimulus). Visualstimuli were presented randomly to the left or right side. In thecued condition, the color of the central fixation point instructedthe animals which task to perform. In the uncued condition,monkeys had to acquire and maintain the current task rule byreward feedback that was given at the end of the trial.

1920 Journal of Cognitive Neuroscience Volume 26, Number 9

with RTs <80 msec (anticipations) or >1000 msec (no re-sponse trials). For each experimental session, the averagetime course of pro- and antisaccade performance beforeand after a task switch was calculated, and the time courseacross sessions was computed by averaging the individualsessions from both subjects. Because there was no effectof the task switch on RTs of pro- and antisaccades, withthe exception of the first trial after a switch, which wasassociated with longer RTs, the mean RT of all correctpro- and antisaccades was computed for each sessionand then averaged across sessions.

RESULTS

Effects of DLPFC Deactivation on UncuedTask Performance

Figure 2A shows the monkeysʼ performance on the trialimmediately preceding and on the 15 trials following aswitch from anti- to prosaccades in the uncued condi-tion. The black line depicts mean percentage of correcttrials on control days when no cooling was performed(n = 17). In these sessions, monkeys performed anti-saccades at ∼75% before the task switch. On the first trialpostswitch, performance dropped to ∼25%, a findingwhich was expected as the animals received no warningof the impending switch. Performance then recoveredquickly, reaching ∼50% on the second postswitch trialand ∼90% after six trials. This pattern of behavior is con-sistent with previous studies using this task ( Johnstonet al., 2007; Everling & Desouza, 2005). When the idlPFCwas cooled bilaterally (red line, n = 17), prosaccade per-formance improved only to ∼70%. When we bilaterallydeactivated the sdlPFC (blue line, n = 17) or the idlPFCtogether with the sdlPFC (purple line, n = 17), pro-saccade recovery after the task switch showed a similartime course to that on control days. Figure 2B showsthe animalsʼ performance for task switches from pro- toantisaccades. On control days (black lines), the monkeysʼperformance dropped from ∼85% on the last trial of theprosaccade block to 15% on the switch trial and thenreached ∼75% correct performance after five trials. Fol-lowing deactivation of the idlPFC, monkeys retained theability to switch to the antisaccade task, but their perfor-mance reached a plateau at ∼60%. Cooling of the sdlPFCproduced no effects, whereas combined cooling of theidlPFC and sdlPFC showed a similar deficit to thatobserved when cooling the idlPFC alone.To statistically evaluate the effects of cooling on task

performance, we computed the mean error rates forTrials 6–15 following a task switch. As shown in Figure 2(A, B), performance was fairly stable during this period.Figure 2C displays error rates for the control condi-tion (black), idlPFC cooling (red), sdlPFC cooling (blue),and combined cooling of the idlPFC and sdlPFC (purple)on prosaccade blocks. A one-way ANOVA showed asignificant effect of cooling on error rates for prosaccade

blocks, F(3) = 13.99, p < .0001. Planned post hoc com-parisons (two sample t tests) between the control con-dition and the deactivations demonstrated that errorrates were significantly greater during cooling of idlPFC( p < .00001). Error rates during cooling of the sdlPFCand, surprisingly, combined cooling of the idlPFC andsdlPFC did not differ from control ( p = .83 and p =.32, respectively). Significant effects of cooling on errorrates were also observed on antisaccade blocks (Fig-ure 2D), F(3) = 13.99, p < .000001 (one-way ANOVA).Error rates were elevated significantly during deactivationof the idlPFC ( p < .0001) and combined deactivation ofthe idlPFC and sdlPFC ( p < .0001). These effects weresimilar for Monkey B and Monkey G (Figure 3).

We also tested whether deactivation impaired the abilityto switch between tasks by further investigating the recov-ery of performance observed following a task switch. Asan index of this, we computed the percentage of correctlyperformed trials following an error on the first postswitchtrial. This analysis revealed that cooling had no effects onswitching (anti- to prosaccades, F(3) = 2.48, p = .07; pro-to antisaccades, F(3) = 1.72, p = .17). These resultsdemonstrate that cooling of the idlPFC impaired the perfor-mance of both pro- and antisaccade task blocks whereascooling of the sdlPFC did not increase error rates. Combinedcooling of these two areas impaired performance on anti-saccade blocks but had no effect on prosaccade blocks.

In addition to effects on performance, we tested effectsof cooling on the saccadic RTs of correct pro- and anti-saccade trials. We found no effects of deactivation on pro-saccade trials (Figure 2E), F(3) = 2.01, p = .12 (one-wayANOVA), but a significant effect on antisaccades (Fig-ure 2F), F(3) = 3.8, p = .014. A post hoc test (two samplet test) showed that combined cooling of the idlPFC andsdlPFC increased antisaccade RTs ( p = .002). The effectsof cooling were very similar for both subjects (Figure 4).

Effects of dlPFC Deactivation on CuedTask Performance

We next investigated the monkeysʼ performance whena task instruction was provided throughout each trial bythe color of the central fixation spot. This version ofthe task retained the requirement of switching betweenpro- and antisaccade task blocks and suppression of pro-saccades on antisaccade trials but eliminated the require-ment to acquire the task rule by reward-based feedbackand maintain it in working memory. Figure 2 (G and H)shows percentage of correct responses on prosaccadeand antisaccade blocks, respectively. No effects of cool-ing were observed on prosaccade trials (Figure 2G, I),F(3) = 1.66, p = .20 (one-way ANOVA). On antisaccadetrials, cooling had a small but significant effect on errorrates (Figure 2H, J), F(3) = 3.3, p = .03. Error rates weresignificantly increased during combined cooling of theidlPFC and sdlPFC ( p = .02, t test). The effects of coolingon saccadic RTs were similar to those observed in the

Hussein et al. 1921

Figure 2. Effects of cooling of dorsolateral prefrontal regions on performance and RTs. (A) Performance following task switches from anti-to prosaccades in the uncued condition. The proportion of correct responses is plotted for the trial immediately before the task switchto the 15 trials after a task switch. Data are averaged across sessions from both monkeys. (B) Same as (A) but for task switches fro pro- toantisaccades. (C) Percentage of errors (i.e., responses following the wrong task rule) during the performance of Trials 6–15 following atask switch from anti- to prosaccades in the uncued condition. (D) Same as (C) but for task switches from pro- to antisaccades. (E) RTs ofall correct prosaccades in the uncued condition. (F) Same as (E) but for correct antisaccades. (G–L) Same as (A–F) but for the cued condition.*p < .05.

1922 Journal of Cognitive Neuroscience Volume 26, Number 9

uncued task, with no effect on prosaccade, F(3) = 2.07,p = .13 (one-way ANOVA), and a significant increase onantisaccade trials, F(3) = 3.97, p < .05. Post hoc compar-isons (two-sample t tests) revealed a significant effectof cooling the idlPFC ( p = .01) and combined coolingof the idlPFC and sdlPFC ( p < .005).

DISCUSSION

Our results suggest a functional specialization of sub-regions within the dlPFC for rule-guided behavior. We

found that deactivation of the cortex lining the caudalprincipal sulcus impaired monkeysʼ performance on bothpro- and antisaccade trials in an uncued task variant,which required them to maintain task rules in workingmemory. Deactivation of the same region when the taskrules were fully cued, a task version absent mnemonicdemands, led to only minor performance impairmentsin the antisaccade condition. In contrast, deactivationof the sdlPFC had no effect on error rates in the pro-or antisaccade conditions in either the cued or uncuedversion of the task. Simultaneous deactivation of both

Figure 3. Effects of coolingon performance in Monkey Band Monkey G. (A) Percentageof errors (i.e., responsesfollowing the wrong task rule)during the performance ofTrials 6–15 following a taskswitch in Monkey B. (B) Sameas (A) for Monkey G. *p < .05.

Hussein et al. 1923

areas produced a selective impairment of antisaccadeperformance, which was considerably greater when thetask rule was uncued. In addition to effects on error rates,dlPFC deactivation caused a selective increase in RTs ofantisaccades with the strongest effects elicited by com-bined cooling of both subregions. These effects weresimilar for the cued and uncued conditions, suggestingthat, unlike the selective cooling-induced impairment ofmnemonic processes we observed in the uncued versionof our task, other processes underlying antisaccade gen-eration such as vector inversion and motor preparationare generally impaired by deactivation of dlPFC.

Although cooling of the principal sulcus had strongeffects on task maintenance, we found no evidence ofan effect of deactivation of either dlPFC subregion onthe performance recovery observed following a changeto a new task rule. This finding is inconsistent with arole of the dlPFC in task switching. Consonant with this,functional imaging studies in humans have reportedprefrontal activations during rule maintenance (Crone,Wendelken, Donohue, & Bunge, 2006; Sakai & Passingham,2003; Passingham et al., 2000) but pre-SMA or anterior cin-gulate activations during task set reconfigurations (Croneet al., 2006). Moreover, our own previous work has

Figure 4. RT of Monkey Band Monkey G. (A) RTs ofcorrect saccades in Monkey B.(B) Same as (A) for Monkey G.*p < .05.

1924 Journal of Cognitive Neuroscience Volume 26, Number 9

demonstrated that task rule selectivity of single neuronsfor uncued pro- and antisaccades was low immediately aftera task switch in the DLPFC (caudal areas 9/46d and 46) buthigh in the ACC in monkeys (Johnston et al., 2007), sug-gesting that medial, but not lateral, PFC areas are stronglyinvolved in task switching.Human imaging studies have shown that the ventral PFC

is activated for visual conditional tasks (Toni, Rushworth, &Passingham, 2001; Passingham et al., 2000), whereas moredorsal prefrontal regions are active for tasks in which thesubjects must represent actions to themselves (Deiberet al., 1991; Frith, Friston, Liddle, & Frackowiak, 1991).Although a number of single neuron recording studies innonhuman primates have demonstrated rule-related ac-tivity in the dlPFC (Tsujimoto et al., 2012; Johnston et al.,2007; Johnston & Everling, 2006b; Everling & Desouza,2005; Genovesio et al., 2005; Wallis & Miller, 2003; Walliset al., 2001; Asaad et al., 2000; White & Wise, 1999), thelesion study of Buckley et al. (2009) was the first to dem-onstrate directly that the cortex in the principal sulcus isessential for the maintenance of task rules in uncued tasks.They proposed that this dlPFC subregion is critical forthe selection and maintenance of uncued rules in workingmemory. Our finding that bilateral cooling of the caudalprincipal sulcus selectively impaired performance of bothpro- and antisaccade blocks in the uncued condition isconsistent with this hypothesis. Although it remains un-known whether larger deficits would have been observedhad we deactivated the entire principal sulcus, our re-sults demonstrate that deactivation of the caudal prin-cipal sulcus by itself leads to significant impairments inthe maintenance of task rules in the absence of ongoinginstruction.We observed no increase in error rates during caudal

principal sulcus deactivation in prosaccade blocks in thecued condition. This finding is consistent with severallesion studies in the macaque, which have shown thatventral and orbital lesions impair learning and retentionof stimulus–response associations (Bussey et al., 2001;Murray et al., 2000; Passingham et al., 2000; Hoshi, Shima,& Tanji, 1998; see, however, Wang, Zhang, & Li, 2000),whereas dorsolateral ablations cause no or only mild defi-cits (Gaffan & Harrison, 1989; Petrides, 1982). However,we observed small increases in antisaccade error rate inthe fully cued condition during idlPFC cooling. This find-ing is consistent with a recent study of our group (Koval,Lomber, & Everling, 2011) in which we observed increasedRTs and increased error rates during bilateral deactivationof the caudal principal sulcus in a task in which monkeysperformed cued, randomly interleaved, pro- and antisac-cade trials. In that study, we also recorded the activity ofsaccade-related neurons in a downstream oculomotorstructure, the superior colliculus, and found that deactiva-tion caused a suite of changes in their preparatory, visual,and motor responses consistent with these behavioralchanges. These cooling-induced effects on the activityof superior colliculus neurons provide a plausible mecha-

nism for the increase in error rates and RTs on antisac-cade trials in the cued condition we observed here.

In contrast to the impairments following deactivation ofthe idlPFC, cooling of the sdlPFC did not impair the main-tenance of task rules for pro- or antisaccades. Althoughthis finding is consistent with the study by Buckley et al.(2009), who did not observe any effects following abla-tions of the cortex immediately dorsal to the principalsulcus to the midline, it is surprising given that many sin-gle neuron recording studies have found rule-related orstrategy-related activity in this region (i.e., Genovesioet al., 2005; Wallis et al., 2001). In fact, we have foundmanyneurons with task rule selectivity in this region while mon-keys performed the uncued pro- and antisaccade tasks( Johnston, De Souza, & Everling, 2009; Johnston et al.,2007; Everling & Desouza, 2005). A potential explanationfor this seeming inconsistency has been provided byPassinghamet al. (2000), whohave theorized that the activityof a given neuron within a frontal region may be derivedfrom its connections with other frontal regions and that onlylesion studies can determine whether a particular regionis essential for a given behavior. Following this argument,our data speak against a role of the sdlPFC in rule main-tenance and in the suppression of automatic responses.

Although deactivation of the sdlPFC had no effect ontask rule maintenance in the uncued condition, combineddeactivation of this region with the idlPFC impaired theperformance of anti- but not prosaccade blocks. Thus,deactivation of the sdlPFC reversed the impairments onprosaccade blocks seen after deactivation of the idlPFCalone. Although the behavioral deficits following combineddeactivation of both dlPFC regions are consistent withthe hypothesis that the dlPFC is engaged in rule mainte-nance when subjects must maintain the goal of overridinga strong response tendency (Bunge, 2004), it suggests thattwo subprocesses, endemic to two anatomically distinctdlPFC subregions, contribute to this function. On thebasis of our results, we hypothesize that the idlPFC playsa general role in maintaining uncued task rules whereasthe sdlPFC is specifically involved in the deployment ofcognitive resources in contexts requiring more control.According to this hypothesis, deactivation of the idlPFCimpairs the maintenance of both automatic (i.e., prosac-cades) and less practiced tasks requiring more control(i.e., antisaccades), resulting in the increase in errors ob-served in both tasks. Deactivation of the sdlPFC by itselfmay not impair maintenance of the antisaccade task rulebecause this function could be carried out within the in-tact idlPFC. Combined deactivation of both dlPFC regionswould impair antisaccades, but not prosaccades, becauseit impaired the maintenance of the task rules and shiftedthe behavior toward the more automatic prosaccade taskrule. Further studies combining cryogenic deactivationand single neuron recordings of prefrontal subregionspromise to further refine this model and broaden ourunderstanding of the functional relationships within thismultifarious region of the brain.

Hussein et al. 1925

Acknowledgments

This study was supported by the Canadian Institutes of HealthResearch.

Reprint requests should be sent to Dr. Stefan Everling, RobartsResearch Institute, 1151 Richmond Street, London, OntarioN6A 5B7, Canada, or via e-mail: [email protected].

REFERENCES

Adey, W. R. (1974). Biophysical and metabolic bases ofcooling effects on cortical membrane potentials in thecat. Experimental Neurology, 42, 113–140.

Asaad, W. F., Rainer, G., & Miller, E. K. (2000). Task-specificneural activity in the primate prefrontal cortex. Journalof Neurophysiology, 84, 451–459.

Barbas, H., & Pandya, D. N. (1989). Architecture and intrinsicconnections of the prefrontal cortex in the rhesus monkey.Journal of Comparative Neurology, 286, 353–375.

Brown, M. R., Vilis, T., & Everling, S. (2007). Frontoparietalactivation with preparation for antisaccades. Journal ofNeurophysiology, 98, 1751–1762.

Buckley, M. J., Mansouri, F. A., Hoda, H., Mahboubi, M.,Browning, P. G., Kwok, S. C., et al. (2009). Dissociablecomponents of rule-guided behavior depend on distinctmedial and prefrontal regions. Science, 325, 52–58.

Bunge, S. A. (2004). How we use rules to select actions:A review of evidence from cognitive neuroscience.Cognitive, Affective & Behavioral Neuroscience, 4,564–579.

Bussey, T. J., Wise, S. P., & Murray, E. A. (2001). The roleof ventral and orbital prefrontal cortex in conditionalvisuomotor learning and strategy use in rhesus monkeys(Macaca mulatta). Behavioral Neuroscience, 115, 971–982.

Crone, E. A., Wendelken, C., Donohue, S. E., & Bunge, S. A.(2006). Neural evidence for dissociable components oftask-switching. Cerebral Cortex, 16, 475–486.

Deiber, M. P., Passingham, R. E., Colebatch, J. G., Friston,K. J., Nixon, P. D., & Frackowiak, R. S. (1991). Corticalareas and the selection of movement: A study withpositron emission tomography. Experimental BrainResearch, 84, 393–402.

Desouza, J. F., Menon, R. S., & Everling, S. (2003). Preparatoryset associated with prosaccades and antisaccades inhumans investigated with event-related fMRI. Journalof Neurophysiology, 89, 1016–1023.

Donohue, S. E., Wendelken, C., & Bunge, S. A. (2008).Neural correlates of preparation for action selection as afunction of specific task demands. Journal of CognitiveNeuroscience, 20, 694–706.

Everling, S., & Desouza, J. F. (2005). Rule-dependent activityfor prosaccades and antisaccades in the primate prefrontalcortex. Journal of Cognitive Neuroscience, 17, 1483–1496.

Frith, C. D., Friston, K., Liddle, P. F., & Frackowiak, R. S.(1991). Willed action and the prefrontal cortex in man:A study with PET. Proceedings. Biological Sciences, 244,241–246.

Gaffan, D., & Harrison, S. (1989). A comparison of the effectsof fornix transection and sulcus principalis ablation uponspatial learning by monkeys. Behavioural Brain Research,31, 207–220.

Genovesio, A., Brasted, P. J., Mitz, A. R., & Wise, S. P. (2005).Prefrontal cortex activity related to abstract responsestrategies. Neuron, 47, 307–320.

Hommel, B., & Prinz, W. (Eds.) (1997). Theoretical issuesin stimulus–response compatibility. Amsterdam: Elsevier.

Hoshi, E., Shima, K., & Tanji, J. (1998). Task-dependentselectivity of movement-related neuronal activity in theprimate prefrontal cortex. Journal of Neurophysiology,80, 3392–3397.

Johnston, K., De Souza, J. F. X., & Everling, S. (2009).Monkey prefrontal cortical pyramidal and putativeinterneurons exhibit differential patterns of activitybetween pro- and antisaccade tasks. Journal ofNeuroscience, 29, 5516–5524.

Johnston, K., & Everling, S. (2006a). Monkey dorsolateralprefrontal cortex sends task-selective signals directly tothe superior colliculus. Journal of Neuroscience, 26,12471–12478.

Johnston, K., & Everling, S. (2006b). Neural activity in monkeyprefrontal cortex is modulated by task context and behavioralinstruction during delayed-match-to-sample and conditionalprosaccade-antisaccade tasks. Journal of CognitiveNeuroscience, 18, 749–765.

Johnston, K., Levin, H. M., Koval, M. J., & Everling, S. (2007).Top–down control-signal dynamics in anterior cingulateand prefrontal cortex neurons following task switching.Neuron, 53, 453–462.

Koval, M. J., Lomber, S. G., & Everling, S. (2011). Prefrontalcortex deactivation in macaques alters activity in thesuperior colliculus and impairs voluntary control ofsaccades. Journal of Neuroscience, 31, 8659–8668.

Lomber, S. G., Payne, B. R., & Horel, J. A. (1999). Thecryoloop: An adaptable reversible cooling deactivationmethod for behavioral or electrophysiological assessmentof neural function. Journal of Neuroscience Methods,86, 179–194.

MacDonald, A. W., III, Cohen, J. D., Stenger, V. A., & Carter,C. S. (2000). Dissociating the role of the dorsolateralprefrontal and anterior cingulate cortex in cognitivecontrol. Science, 288, 1835–1838.

Milner, B. (1963). Effects of different brain lesions on cardsorting. Archives of Neurology, 9, 90.

Munoz, D. P., & Everling, S. (2004). Look away: The antisaccadetask and the voluntary control of eye movement. NatureReviews Neuroscience, 5, 218–228.

Murray, E. A., Bussey, T. J., & Wise, S. P. (2000). Role ofprefrontal cortex in a network for arbitrary visuomotormapping. Experimental Brain Research, 133, 114–129.

Passingham, R. E., Toni, I., & Rushworth, M. F. (2000).Specialisation within the prefrontal cortex: The ventralprefrontal cortex and associative learning. ExperimentalBrain Research, 133, 103–113.

Petrides, M. (1982). Motor conditional associative-learningafter selective prefrontal lesions in the monkey.Behavioural Brain Research, 5, 407–413.

Petrides, M., & Pandya, D. N. (1999). Dorsolateral prefrontalcortex: Comparative cytoarchitectonic analysis in thehuman and the macaque brain and corticocorticalconnection patterns. European Journal of Neuroscience,11, 1011–1036.

Preuss, T. M., & Goldman-Rakic, P. S. (1991). Myelo- andcytoarchitecture of the granular frontal cortex andsurrounding regions in the strepsirhine primateGalago and the anthropoid primate Macaca. Journalof Comparative Neurology, 310, 429–474.

Sakai, K., & Passingham, R. E. (2003). Prefrontal interactionsreflect future task operations. Nature Neuroscience, 6,75–81.

Stroop, J. R. (1935). Studies of interference in serial verbalreactions. Journal of Experimental Psychology, 18,643–662.

Toni, I., Rushworth, M. F., & Passingham, R. E. (2001).Neural correlates of visuomotor associations. Spatial

1926 Journal of Cognitive Neuroscience Volume 26, Number 9

rules compared with arbitrary rules. ExperimentalBrain Research, 141, 359–369.

Tsujimoto, S., Genovesio, A., & Wise, S. P. (2012). Comparisonof strategy signals in the dorsolateral and orbital prefrontalcortex. Journal of Neuroscience, 31, 4583–4592.

Wallis, J. D., Anderson, K. C., & Miller, E. K. (2001). Singleneurons in prefrontal cortex encode abstract rules. Nature,411, 953–956.

Wallis, J. D., & Miller, E. K. (2003). From rule to response:Neuronal processes in the premotor and prefrontal cortex.Journal of Neurophysiology, 90, 1790–1806.

Wang, M., Zhang, H., & Li, B. M. (2000). Deficit in conditionalvisuomotor learning by local infusion of bicuculline intothe ventral prefrontal cortex in monkeys. EuropeanJournal of Neuroscience, 12, 3787–3796.

White, I. M., & Wise, S. P. (1999). Rule-dependent neuronalactivity in the prefrontal cortex. Experimental BrainResearch, 126, 315–335.

Yang, X. F., Kennedy, B. R., Lomber, S. G., Schmidt, R. E.,& Rothman, S. M. (2006). Cooling produces minimalneuropathology in neocortex and hippocampus.Neurobiology of Disease, 23, 637–643.

Hussein et al. 1927