Dopaminergic modulation of cortical function in patients with Parkinson's disease.код для вставкиСкачать
Dopaminergic Modulation of Cortical Function in Patients with Parkinson’s Disease Venkata S. Mattay, MD,1 Alessandro Tessitore, MD,1 Joseph H. Callicott, MD,1 Alessandro Bertolino, MD,1 Terry E. Goldberg, PhD,1 Thomas N. Chase, MD,2 Thomas M. Hyde, MD,1 and Daniel R. Weinberger, MD1 Patients with idiopathic Parkinson’s disease suffer not only from classic motor symptoms, but from deficits in cognitive function, primarily those subserved by the prefrontal cortex as well. The aim of the current study was to investigate the modulatory effects of dopaminergic therapy on neural systems subserving working memory and motor function in patients with Parkinson’s disease. Ten patients with stage I and II Parkinson’s disease were studied with functional magnetic resonance imaging, during a relatively hypodopaminergic state (ie, 12 hours after a last dose of dopamimetic treatment), and again during a dopamine-replete state. Functional magnetic resonance imaging was performed under three conditions: a working memory task, a cued sensorimotor task and rest. Consistent with prior data, the cortical motor regions activated during the motor task showed greater activation during the dopamine-replete state; however, the cortical regions subserving working memory displayed greater activation during the hypodopaminergic state. Interestingly, the increase in cortical activation during the working memory task in the hypodopaminergic state positively correlated with errors in task performance, and the increased activation in the cortical motor regions during the dopamine-replete state was positively correlated with improvement in motor function. These results support evidence from basic research that dopamine modulates cortical networks subserving working memory and motor function via two distinct mechanisms: nigrostriatal projections facilitate motor function indirectly via thalamic projections to motor cortices, whereas the mesocortical dopaminergic system facilitates working memory function via direct inputs to prefrontal cortex. The results are also consistent with evidence that the hypodopaminergic state is associated with decreased efficiency of prefrontal cortical information processing and that dopaminergic therapy improves the physiological efficiency of this region. Ann Neurol 2002;51:156 –164 DOI 10.1002/ana.10078 Although patients with idiopathic Parkinson’s disease (PD) usually do not present with cognitive symptoms, there is growing evidence that the motor symptoms characteristic of this disease are also accompanied by a pattern of progressive neuropsychological impairment.1 These deficits are similar to those seen in patients with frontal lobe damage and include problems with verbal learning,2 problem solving,3 self-ordered pointing,4 temporal sequencing,5 delayed response,6 conditional associative learning,2,7 implicit learning,7 and forward planning.8 Because most patients with PD do not suffer from primary frontal lobe pathology, these cognitive deficits must be a direct or indirect consequence of degeneration of ascending catecholaminergic projections from the brainstem. It is generally believed that degeneration of the sub- stantia nigra pars compacta (SNpc) with subsequent depletion of dopamine in the putamen and resulting disruption of basal ganglia–thalamocortical loops results in the classical motor signs and symptoms of PD. Akinesia is thought to result from functional deafferentation of the supplementary motor area (SMA),9 –12 whereas bradykinesia and rigidity appear to result from functional deafferentation of the motor cortex.13 By contrast, the neurobiological basis of the frontal cognitive impairment has not been clearly established. Two mechanisms have been proposed,14 including either an alteration in outflow of the caudate nuclei to frontal cortex via the thalamus15 or diminished dopamine activity in the frontal lobes themselves consequent to degeneration of the frontal projections of the VTA and other nigral cell groups.16 –21 The aim of the present From the 1Clinical Brain Disorders Branch, National Institute of Mental Health, 2Experimental Therapeutics Branch, National Institute of Neurological Diseases and Stroke, National Institutes of Health, Bethesda, MD. Published online Jan 31, 2002 Received Aug 6, 2001, and in revised form Sep 24. Accepted for publication Sep 26, 2001. 156 © 2002 Wiley-Liss, Inc. Address correspondence to Dr Mattay, Clinical Brain Disorders Branch, National Institute of Mental Health, Bldg 10, Center Drive, Room 4S-235, National Institutes of Health, Bethesda, MD 20982-1379. E-mail: firstname.lastname@example.org study was to delineate these mechanisms using blood oxygen level-dependent (BOLD) functional magnetic resonance imaging (fMRI), a method sensitive to changes in the oxygen content in the blood, while patients performed a prefrontal cortex (PFC)-dependent task, the N-Back working memory task (WMT), in a relatively dopamine-depleted (ie, “drug-off”) state and relatively dopamine-replete (ie, “drug-on”) state. To specifically contrast the effects of dopamine on WM circuitry with those on the motor circuitry, we studied patients while they performed both a WMT and a visually paced motor task. If the prefrontal cognitive deficits are primarily attributable to compromised outflow from the caudate nuclei to the frontal cortex via the thalamus,15 one might see more PFC activation during the “drug-on” state than during the relatively “drug-off” state— similar to the activation pattern changes seen in the cortical motor areas after dopaminergic therapy.10 –12 Theoretically, dopamine diminishes inhibitory basal ganglia output to the thalamus and, by relative disinhibition, leads to increased activity of thalamocortical projection neurons.22 In contrast, if the prefrontal cognitive deficits are secondary to a direct cortical dopamine deficiency caused by reduced input through the mesocortical dopamine pathway,17–21, 23–24 the physiological response might be qualitatively different. In this case, the hypodopaminergic state would be associated with decreased cortical physiological efficiency in information processing; that is, a larger neuronal pool may be required to carry out the task in the dopaminedepleted state than during the dopamine-replete state. This latter supposition is consistent with evidence from animal and neuroimaging studies that suggests that monoamines including dopamine optimize the signalto-noise response of pyramidal neurons in both a taskspecific and region-specific manner to improve cortical efficiency.25–29 Greater activation under these circumstances may reflect a breakdown in local circuit inhibition,30 –32 as well as the recruitment of additional cortical resources to meet cognitive demands and maintain proficiency. Patients and Methods Subjects Ten patients with idiopathic PD participated in the study. Patients were recruited from the Parkinson’s clinic of the Experimental Therapeutics Branch of the National Institute of Neurological Disorders and Stroke, and from the surrounding community. fMRI was conducted under a protocol approved by the Institutional Review Board of the National Institute of Mental Health, Intramural Research Program. All subjects gave written informed consent before participation in this protocol. Participants in the study included 2 patients with Hoehn and Yahr stage I and 8 patients with stage II (mean age ⫹SE ⫽ 55 ⫾ 2.6 years) under treatment with Sinemet (L-Dopa ⫹ carbidopa), alone or in combination with dopamine agonists. Before being scanned, all patients underwent a thorough neurological examination. To preclude effects of order and of novelty in terms of the MRI environment and the task paradigms, all patients underwent a structural MRI scan and were made to practice the tasks until they reached ceiling in their performance on a separate day before the fMRI sessions. Each patient was studied twice, once during the “drugoff” state (ie, 12 hours after their last dopaminergic drug dose the night before) and for a second time during the “drug-on” state (ie, 1 to 2 hours after the first dose of the day). Dopaminergic drug doses were adjusted to provide an optimal therapeutic response. A brief neurological examination was carried out before each session to rate the severity of patients’ motor functions (Unified Parkinson’s Disease Rating Scale [UPDRS], motor rating items 19 to 31).33 Patients were asked to refrain from nicotine and caffeine for at least 4 hours and from over-the-counter medications for 24 hours before the MRI studies. Data Acquisition BOLD fMRI data were collected on a standard 1.5T Signa scanner (General Electric, Milwaukee, WI), using gradientecho whole-brain fast spiral acquisition to acquire 36 interleaved slices (3.75mm isotropic, TE ⫽ 24ms, TR ⫽ 56ms, flip angle ⫽ 85 degrees, FOV ⫽ 24cm, matrix ⫽ 64 ⫻ 64).34 Working Memory and Motor Tasks BOLD fMRI was conducted while subjects performed a variation of the n-back WMT as previously described.35,36 N-Back refers to the number of previous stimuli that the subject had to recall. The stimuli consisted of numbers (1– 4) shown in random order and displayed at the points of a diamond-shaped box. During each treatment condition, two task combinations were administered: (1) eight cycles of the 2-Back (WMT) alternating with the 0-Back (sensorimotor task), and (2) eight cycles of the 0-Back (which is essentially a visually paced motor task), alternating with a rest condition. Each task combination was obtained in 4 minutes and 16 seconds, 128 whole-brain images with 16 images in each cycle (8 during the 2-Back and 8 during the 0-Back, or 8 during the 0-Back and 8 during rest). Owing to time constraints, only 8 of the 10 patients were able to complete the “0-Back versus rest” paradigm. Further, performance data from 1 of the remaining 8 patients demonstrated multiple involuntary button presses (probably because of excessive hand tremor) during the “drug-off” state. Therefore, the data from this patient were not included in the final group analysis for the “0-Back versus rest” comparison. Image Processing and Data Analysis After reconstruction of the individual time volumes, each three-dimensional brain volume was registered to the first in the time series, using a tricubic-spline interpolation.37 Data sets were then chosen for their high quality (scan stability), as demonstrated by small motion correction (⬍2 voxels) and matched voxel variance across the two sessions.35,38 The in- Mattay et al: Dopaminergic Modulation of Cortical Function 157 Table 1. Clinical Ratings, Task Accuracy, and Reaction Time Measurements Drug State Drug-off Drug-on a UPDRS (Motor Subscale) Mean ⫾ SEa 2-Back (% Correct) Mean ⫾ SE 10 Pt 2-Back (RT ms) Mean ⫾ SE 10 Pt 0-Back (% correct) Mean ⫾ SE 7 Pt 0-Back (RT ms) Mean ⫾ SE 7 Pt 8.8 ⫾ 2.6 5.0 ⫾ 1.8 80.2 ⫾ 4.6 86.8 ⫾ 4.2 710 ⫾ 73 747 ⫾ 93 91.5 ⫾ 5.9 93.9 ⫾ 1.5 846 ⫾ 158 783 ⫾ 116 p ⬍ 0.003. UPDRS ⫽ Unified Parkinson’s Disease Rating Scale; SE ⫽ standard error; RT ⫽ reaction time. dividual whole-brain data were then spatially normalized to a common stereotactic space (Montreal Neurological Institute [MNI] template) via Automated Image Registration 3.08.39,40 Voxel-wise signal intensities were ratio normalized to the whole brain mean and detrended in a linear fashion with the baseline at each voxel set to 100.35,38 The data were then smoothed with a Gaussian filter (8 ⫻ 8 ⫻ 8mm) to control further for inter-individual variance in sulcal and gyral anatomy. Image analysis software and general linear model statistics (SPM 99) were used to create group maps for the “2-Back versus 0-Back” comparison (WM challenge) and for the “0-Back versus rest” comparison (motor function) during both the “drug-off” and “drug-on” states. The height threshold value was set to p ⬍ 0.05 (corrected for multiple comparisons) with an extent threshold of three voxels. We then performed a “drug ⫻ task” interaction analysis to assess the effect of dopaminergic therapy on task-related activation. To determine the relationship between drug-related changes in BOLD response and change in motor function or WM performance, we performed voxel-wise correlation analyses. To accomplish these correlation analyses, we first carried out a “drug ⫻ task” analysis for each subject separately for each task condition (“2-Back vs 0-Back” and “0-Back vs rest”) and created contrast images. We then performed a second-level correlation analysis on these contrast images, using the simple regression model in SPM99 with two measures of improvement in motor function (ie, percentage change in reaction time (RT) and percentage change in UPDRS scores relative to the “drug-off” state) as covariates of interest for the motor function analysis, and percent difference in performance scores of the 2-Back task between the two drug conditions as a covariate of interest for the WM data analysis. Behavioral Data Analysis We performed a Student’s t-test on the clinical ratings (UPDRS–motor subscale scores), task performance, and RT measures to look for significant differences across drug conditions. Task performance was measured as the percentage of correct responses. Results Clinical Ratings, Task Accuracy, and Reaction Time Measurements All patients showed a significant improvement in motor symptoms during the “drug-on” state when com- 158 Annals of Neurology Vol 51 No 2 February 2002 pared with the “drug-off” state (Table 1). No significant difference was found in performance of the 0-Back task between the two drug states. Although the mean accuracy score on the 2-Back task tended to be lower during the “drug-off” state, the difference did not reach statistical significance; 6 patients performed worse during the “drug-off” state, 2 patients performed equally well during both drug conditions, and 2 patients had a marginally lower score during the “drugon” state. There were no statistically significant differences in RT for both task paradigms across the two drug states; however, the mean RT for the 0-Back task during the “drug-on” state was lower than during the “drug-off” state; 5 of the 7 subjects showed a decrease in RT on the 0-Back task after dopaminergic therapy. During the two-back state, the mean RT was marginally and nonsignificantly longer during the “drug-on” state relative to the “drug-off” state, because of paradoxically longer RTs in 4 subjects during the “drug-on” state. While the basis of this is unclear, 2 of these 4 subjects also had a lower accuracy score during the “drug-on” state, and 1 subject had similar accuracy scores during both states. Functional Magnetic Resonance Imaging A main task effect during both “drug-off” and “drug-on” sessions mapped to very similar locales. Consistent with earlier studies,35,41 the spatial distribution of the task-related responses included PFC (BA 9-10/44 – 46), pericingulate cortex (including the medial frontal gyrus), anterior cingulate (BA 24, 32), and parietal cortex (BA 7, 39 – 40) bilaterally. However, the spatial extent of activation in these cortical regions was much greater during the “drug-off” state than during the “drug-on” state. This is further corroborated by the off ⬎ on “drug ⫻ task” interaction analysis (see Fig 1 and Table 2, part a), which shows the cortical areas that are significantly more active during the “drug-off” state relative to the “drugon” state. There were also significant correlations between the relatively greater degree of activation in these cortical regions and the degree of decrement in perforWM CHALLENGE. mapped to very similar cortical locales. The spatial distribution of the task effects included supplementary motor area (SMA; BA 6), lateral premotor cortex (LPM; BA 6; left ⬎ right), sensorimotor cortex (SMC; BA 4, 3,2 and 1; left ⬎ right) , parietal cortex (PAR; BA 39, 40) and cerebellum (CER; right ⬎ left) bilaterally. However, consistent with evidence from earlier neuroimaging literature10 –12 and contrary to the effect observed during the 2-Back task, the spatial extent of activation in these cortical regions was much greater during the “drug-on” state than during the “drug-off” state. This is further corroborated by the on ⬎ off “drug ⫻ task” interaction analysis (see Fig 2a and Table 3, part a), which shows the cortical areas that are significantly more active during the “drug-on” state relative to the “drug-off” state. This increase in activation after dopaminergic therapy correlated significantly with improvement in RT (see Table 2, part a, and Fig 2b). Moreover, the increase in activation after dopaminergic therapy in the right primary motor cortex, right PAR, SMA, and left LPM covaried with improvement in motor function (as determined by a change in UPDRS motor subscale scores) (see Table 3, part c). This is the opposite direction of the relationship of activation and behavior seen in PFC during the WMT. No areas exhibited a significant off ⬎ on “drug ⫻ task” interaction. Fig 1. Brain regions with greater activation during working memory in the hypodopaminergic state. (A) Representative slices from a group activation map showing brain regions with a significant off ⬎ on “drug ⫻ task” interaction during working memory challenge. Slices are arranged in radiological orientation (ie, right side of brain to viewer’s left and vice versa). Patients in the “drug-off” state show more activation of the network subserving working memory. See Table 2, part A, for coordinates of voxels with significant interaction. (B) Group map of cortical regions that show significant correlations between drug-related differences in BOLD response and a measure of difficulty in performing the working memory task during the “drug-off” state. See Table 2, part B, for coordinates of significant voxels and statistics. mance during the “drug-off” state relative to the “drugon” state, ie, a decrease in accuracy correlated with an increase in cortical activation (see Fig 1b, see Table 2, part b). In other words, the hypodopaminergic state was associated with decreased efficiency in cortical processing during WM. There were no areas showing a significant on ⬎ off “drug ⫻ task” interaction. Image analysis limited to the 4 atypical subjects with prolonged RT in the “drug-on” state also showed the same pattern as the whole group. MOTOR FUNCTION (“0-BACK VS REST”). As with WM, there was a main effect of task on both sessions that Discussion We found that during a WM challenge in patients with uncomplicated PD, the extent of cortical activation (particularly in the PFC, PAR and cingulate regions) was greater in the “drug-off” state relative to the “drug-on” state. Although there was no significant difference in task performance across the two drug conditions, correlation analysis revealed that the worse the patients performed, the more cortical tissue they activated (see Fig 1b). This finding of increased activity in the cortical areas subserving WM during the “drug-off” state even in the context of slightly poorer cognitive function, is in contrast to the changes we observed in the activity of cortical areas subserving motor function. Consistent with earlier reports,10 –12 we noted greater activation in the supplementary motor, primary motor, and parietal cortices in the “drug-on state” (see Fig 2a and Table 3). This increase in activity in cortical motor regions was associated with improvement in motor symptoms as determined by the motor subscale of UPDRS and a decrease in RT during the motor task (see Fig 2b). These activation changes in the cortical motor regions are in agreement with the view that the motor signs and symptoms in patients with PD result from functional deafferentation of excitatory thalamic inputs to the cortical motor regions probably due to disruption of neural processing through basal ganglia– thalamocortical loops.22 By contrast, the changes we Mattay et al: Dopaminergic Modulation of Cortical Function 159 Table 2. Effects on Regions During Working Memory Challenge Talairach Coordinates (x,y,z) Region (Brodmann’s Area) a: Regions that showed a significant off ⬎ on drug ⫻ task interaction Left dorsal PFC (BA 10) (⫺34 51 5) Left dorsal PFC (BA 45/46) (⫺40 36 5) Right dorsal PFC (BA 9) (33 25 34) Right PFC (BA 8) (33 11 43) Anterior cingulate (BA 32/8) (0 29 40) Anterior cingulate (BA 32/6) (3 15 46) Right insular cortex (39 ⫺15 10) Left insular cortex (⫺39 ⫺8 6) Right PFC (BA 8) (19 25 46) Left PFC (BA 8) (⫺23 15 49) Left parietal cortex (BA 40) (⫺50 ⫺21 23) Right parietal cortex (BA 40) (23 ⫺41 27) Right precuneus (BA 7) (7 ⫺47 53) Maximum z value Cluster Level Corrected p 3.09 2.88 4.24 3.13 2.92 2.56 3.00 2.19 2.56 2.15 2.85 2.43 2.38 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.04 0.008 0.001 0.000 0.000 0.003 b: Regions that showed a significant correlation between drug-induced increase in BOLD response and decrement in performance Right PFC (BA 10) (23 47 5) 2.26 0.05 Right PFC (BA 46) (23 44 11) 2.33 0.05 Left PFC (BA 10) (⫺34 47 5) 2.29 0.03 Left parietal cortex (BA 39) (⫺46 ⫺51 19) 4.29 0.000 Left parietal cortex (BA 39/40) (⫺34 ⫺51 33) 2.6 0.000 Right precuneus (BA 19) (26 ⫺80 18) 3.57 0.000 Left superior temporal gyrus (BA 22) (⫺50 ⫺47 14) 3.92 0.000 PFC ⫽ prefrontal cortex; BA ⫽ Brodmann’s area. observed in the cortical regions subserving WM, namely the more focused activation in the PFC after dopaminergic therapy, are more consistent with our alternate hypothesis, namely, that relatively under active dopaminergic terminals in PFC would impact directly on its information processing, in terms of decreased efficiency or automaticity of this processing (ie, greater extent of cortical activation may be required for the same or poorer performance). Although this interpretation is speculative, the qualitatively differential physiological response to dopaminergic therapy observed during the WM and motor tasks raises the possibility that a mechanism other than the nigrostriatal dopaminergic system may mediate the changes in the WM network. We discuss this interpretation within the context of the mesocortical dopaminergic system and prefrontal function, as well as the degeneration of this system in PD. Mesocortical Dopaminergic System and Prefrontal Function Anatomical studies indicate that the frontal cortex receives dopaminergic fibers42 from the mesencephalon and that dopamine receptors, particularly D1, mediate dopamine effects in PFC related to WM.43 Neurophysiological and neuropharmacological data confirm the role of these receptors in PFC function, especially 160 Annals of Neurology Vol 51 No 2 February 2002 WM.44 – 46 Dopamine afferents to pyramidal neurons synapse on soma, but also on dendritic spines in close proximity to glutamate inputs from other cortical neurons. Dopamine inputs to the dendritic shafts of local circuit neurons also appear to be in close proximity to glutamate presynaptic terminals.47 On the basis of these observations, it has been suggested that dopamine gates the excitatory impact of associative cortical information mediated by intracortically projecting glutamate neurons and by locally recurrent collaterals of pyramidal neurons.48 Evidence from electrophysiological studies indicates that stimulation of the ventral tegmental area in animals with an intact mesocortical system causes partial depolarization and diminished spontaneous firing of action potentials of prefrontal pyramidal neurons, which appear to be primed for associative inputs.49,50 In animals, dopamine is released in PFC during WM,51 and depleting PFC of dopamine or blocking D1 receptors diminishes performance on these tasks.52,53 Recent studies of neuronal interactions in tissue slice preparations indicate that D1 receptors modulate glutamate-induced excitability of intrinsic PFC neurons.29 Consistent with this anatomical and electrophysiological information, experiments in behaving monkeys have demonstrated that dopamine modulates the firing responses of pyramidal neurons and improves the prefrontal physiological signal to stages of the disease the brain compensates for these physiological alterations by recruiting extraneous neuronal assemblies. Advancing disease may decrease the efficacy of these compensatory mechanisms as patients become cognitively symptomatic. Furthermore, the patterns we observed may be different in patients with more advanced disease, as they may have more extensive degeneration of dopaminergic neurons. Fig 2. Brain regions with greater activation during the motor task in the dopamine replete state. (A) Representative slices from a group activation map showing brain regions with a significant on ⬎ off “drug ⫻ task” interaction during the motor task (“0-Back vs rest”). Consistent with evidence from earlier neuroimaging data and contrary to our observations during working memory challenge (Fig 1a), there was more activation of the cortical motor regions during the “drug-on” state relative to the “drug-off” state. See Table 3, part A, for coordinates of voxels with significant interaction. (B) Group map of cortical regions that show significant correlations between drug-induced change in BOLD response during the motor task with a measure of improvement in reaction time. See Table 3, part B, for coordinates of significant voxels. noise during executive cognition and WMT, respectively.45,54 This signal-to-noise-enhancing effect of monoamines on cortical neuronal activity has also been supported by neuroimaging studies in humans that used monoaminergic drugs.27,36, 55–58 Our finding of more focal activation after dopaminergic therapy is therefore in conceptual agreement with the evidence from the above studies and supports the view that dopamine “focuses” prefrontal neuronal assemblies to respond more precisely and efficiently to specific stimuli. Consistent with this hypothesis, Cools and colleagues recently demonstrated working memory task-related rCBF decreases in the right DLPFC of patients with PD following dopaminergic therapy when compared to the hypodopaminergic state.59 In the hypodopaminergic state of PD, this focusing function is presumably altered and physiological STN ratio is diminished. Our data support this conclusion. It is plausible that to maintain cognitive proficiency in early Degeneration of the Mesocortical Dopaminergic System in PD Whereas the hallmark of PD is dopaminergic neuronal loss in SNpc leading to striatal denervation,60 there is also evidence of considerable damage to the mesocorticolimbic dopaminergic system, which originates largely in the VTA and medial SNpc.23,24,61,62 Consistent with evidence of mesocortical dopaminergic deafferentation from postmortem neurochemical studies, investigators have used positron emission tomographic imaging techniques to demonstrate neocortical monoamine terminal loss in the frontal cortex of patients with PD17 and a reduction in presynaptic dopaminergic function, not only in the nigrostriatal system, but also in the mesocortical system in patients with early PD.19 –21 A positive correlation between F-18 fluorodopa binding in the frontal cortex and performance on prefrontal tasks has also been reported.19 In another study, based on the absence of a significant correlation between C-11 Nomifensine (a radioligand of dopamine and norepinephrine presynaptic reuptake sites) binding in the caudate and WM performance, Marie and colleagues18 suggested that the prefrontal dopaminergic system plays a more prominent role in WM, rather than the striatal dopaminergic system. In essence, these imaging data combined with the previously mentioned evidence of cortical signal-to-noise-enhancing effects of dopamine support our hypothesis that direct cortical involvement of the mesocortical dopaminergic system may be responsible for decreased efficiency of PFC function in patients with early PD. Although cognitive performance may benefit from dopaminergic therapies aimed at improving motor symptoms, higher doses required in later stages of the disease have been reported to exacerbate cognitive deficits in some patients, possibly owing to excessive dopamine receptor stimulation in striatum or PFC30,31,46,63– 65 or to extension of the degenerative process to additional neuronal populations. To our knowledge, this is the first study in which the physiological effects of dopaminergic modulation on both PFC function and motor function in patients with PD were assayed simultaneously. Although our results and conclusions are not based on direct in vivo measurement of extrastriatal dopaminergic cortical activity, they are well supported by, and consistent with, findings in experimental animal paradigms. These re- Mattay et al: Dopaminergic Modulation of Cortical Function 161 Table 3. Effects on Regions during Motor Task Talairach Coordinates (x,y,z) Region (Brodmann’s Area) a: Regions that showed a significant on ⬎ off drug ⫻ task interaction Right primary motor cortex (BA 4) (33 ⫺18 55) Left primary motor cortex (BA 4) (⫺43 ⫺18 38) Right premotor area (BA 6) (30 ⫺4 45) Supplementary motor area (BA 6) (⫺10 ⫺22 68) Right parietal cortex (BA 7) (19 ⫺62 56) Left parietal cortex (BA 40) (⫺13 ⫺70 52) Left parietal cortex (BA 40) (⫺46 ⫺37 27) Maximum z value Cluster Level Corrected p 2.67 2.14 2.56 3.08 3.53 2.72 3.21 0.03 0.04 0.02 0.04 0.000 0.000 0.000 b: Regions that showed a significant correlation between drug-induced increase in BOLD response and improvement in reaction time Left primary motor cortex (BA 4) (⫺34 ⫺12 44) 2.31 0.03 Left premotor area (BA 6) (⫺10 ⫺18 68) 2.78 0.01 Supplementary motor area (BA 6) (⫺4 ⫺22 64) 1.90 0.01 Left parietal cortex (BA 7) (⫺27 ⫺51 53) 2.51 0.000 Left parietal cortex (BA 40) (⫺34 ⫺44 46) 1.94 0.000 c: Regions that showed a significant correlation between drug-induced increase in BOLD response during the motor task and improvement in UPDRS scores Right primary motor cortex (BA 4) (26 ⫺14 58) 2.75 0.03 Left premotor area (BA 6) (⫺17 ⫺12 61) 2.44 0.03 Supplementary motor area (BA 6) (⫺10 ⫺8 65) 2.51 0.04 Right parietal cortex (BA 7) (⫺33 ⫺44 50) 2.75 0.003 BA ⫽ Brodmann’s area. sults support the notion that dopamine modulates cortical circuitry invoked by the WMT differently than it does the cortical motor system. Consistent with the model reported by Alexander and colleagues,22 it appears that dopamine modulates motor function through the nigrostriatal–thalamocortical loop, whereas its effect on WM function appears to be primarily through the mesocortical dopaminergic system. Our results also show that even early in the disease, when there are no gross deficits in cognition at the behavioral level, the brain regions subserving WM function are performing additional work to accomplish the task; that is, they are inefficient. The authors thank Dr Joseph A. Frank (Laboratory of Diagnostic Radiology Research, NIH), and Dr Jeff Duyn (Laboratory of Functional and Molecular Imaging, NINDS, NIH), for providing fMRI resources and expertise; Dr Richard Coppola (Clinical Brain Disorders Branch, NIMH, NIH), for task development; Dr Saumitra Das, and Dr Sam Lee (Clinical Brain Disorders Branch, NIMH, NIH), for research assistance; Dr Andreas Meyer-Lindenberg, Dr Phillip Kohn, Dr Ahmad Hariri, and Dr Francesco Fera (Clinical Brain Disorders Branch, NIMH, NIH), for advice on fMRI data analysis; and Marge Gillespie, RN (Experimental Therapeutics Branch, NINDS, NIH), for her help with patient recruitment. 162 Annals of Neurology Vol 51 No 2 February 2002 References 1. Goldman W, Baty JD, Buckles VD, et al. Cognitive and motor functioning in Parkinson’s disease. Arch Neurol 1998;55: 674 – 680. 2. Taylor A, Saint-Cyr J, Lang AE. Memory and learning in early Parkinson’s disease: evidence for a frontal lobe syndrome. Brain Cogn 1990;2:211–238. 3. Elsinger P, Grattan L. Frontal lobe and frontal-striatal substrates for different forms of human cognitive flexibility. Neuropsychologia 1993;31:17–28. 4. Pillon B, Dubois B, Cusimano G, et al. Does cognitive impairment in Parkinson’s disease result from non-dopaminergic lesions? J Neurol 1989;52:201–206. 5. Sagar H, Sullivan E, Gabrieli J. Temporal ordering and short-term memory deficits in Parkinson’s disease. Brain 1988;111:525–539. 6. Postle B, Jonides J, Smith E. Spatial, but not object delayed response is impaired in early Parkinson’s disease. Neuropsychology 1997;11:171–179. 7. Jackson G, Jackson S, Harrison J. Serial reaction time learning and Parkinson’s disease: evidence for a procedural learning deficit. Neuropsychologia 1995;33:577–593. 8. Ogden J, Growden J, Corkin S. Deficits on visuospatial tests involving forward planning in high-functioning parkinsonians. Neuropsychiatry Neuropsychol Behav Neurol 1990;3:125–139. 9. DeLong M. Primate models of movement disorder of basal ganglia origin. Trends Neurosci 1990:281–285. 10. Jenkins I, Fernandez W, Playford ED, et al. Impaired activation of the supplementary motor area in Parkinson’s disease is reversed when akinesia is treated with apomorphine. Ann Neurol 1992;32:749 –757. 11. Rascol O, Sabatini U, Chollet F, et al. Normal activation of the supplementary motor area in patients with Parkinson’s disease undergoing long-term treatment with levodopa. J Neurol Neurosurg Psychiatry 1994;57:567–571. 12. Sabatini U, Boulanouar K, Fabre N, et al. Cortical motor reorganization in akinetic patients with Parkinson’s disease: a functional MRI study. Brain 2000;123:394 – 403. 13. Hoover J, Strick P. Multiple output channels in the basal ganglia. Science 1993;259:819 – 821. 14. Kulisevsky J. Role of dopamine in learning and memory. Drugs Aging 2000;16:365–379. 15. Owen A, Doyon J, Dagher A, et al. Abnormal basal ganglia outflow in Parkinson’s disease identified with PET. Brain 1998; 121:949 –965. 16. Leenders K, Brown R, Salmon P, et al. The relationship between “frontal” function in patients with Parkinson’s disease and brain dopaminergic activity as measured by PET. Neurology 1990;40(suppl 1):168. 17. Marie R, Barre L, Rioux P, et al. PET imaging of neocortical monoaminergic terminals in Parkinson’s disease. J Neural Transm 1995;9:55–71. 18. Marie R, Barre L, Dupuy B, et al. Relationships between striatal dopamine denervation and frontal executive tests in Parkinson’s disease. Neurosci Lett 1999;260:77– 80. 19. Rinne J, Portin R, Ruottinen H, et al. Cognitive impairment and the brain dopaminergic system in Parkinson disease. Arch Neurol 2000;57:470 – 475. 20. Ouchi Y, Kanno T, Okada H, et al. Presynaptic and postsynaptic dopaminergic binding densities in the nigrostriatal and mesocortical systems in early Parkinson’s disease: a doubletracer positron emission tomography study. Ann Neurol 1999; 46:723–731. 21. Ouchi Y, Yoshikawa E, Okada H, et al. Alterations in binding site density of dopamine transporter in the striatum, orbitofrontal cortex, and amygdala in early Parkinson’s disease: compartmental analysis for ␤-CFT binding with positron emission tomography. Ann Neurol 1999;45:601– 610. 22. Alexander G, DeLong M, Strick P. Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annu Rev Neurosci 1986;9:357–381. 23. Javoy-Agid F, Agid Y. Is the mesocortical dopaminergic system involved in Parkinson’s disease? Neurology 1980;30:1326 – 1330. 24. Scatton B, Javoy-Agid, Rouquier L, et al. Reduction of cortical dopamine, noradrenaline, serotonin and their metabolites in Parkinson’s disease. Brain Res 1983;275:321–328. 25. Foote SL, Freedman R, Oliver PA. Effects of putative neurotransmitters on neuronal activity in monkey auditory cortex. Brain Res 1975;86:229 –242. 26. Segal M, Bloom FE. The action of norepinephrine in the rat hippocampus. III. Hippocampal cellular responses to locus coeruleus stimulation in the awake rat. Brain Res 1976;107: 499 –511. 27. Mattay V, Berman KF, Ostrem JL, et al. Dextroamphetamine enhances “neural network-specific” physiological signals: a positron-emission tomography rCBF study. J Neurosci 1996; 16:4816 – 4822. 28. Goldman-Rakic P. The cortical dopamine system: role in memory and cognition. Adv Pharmacol 1998;42:707–711. 29. Seamans J, Durstewitz D, Christie BR, et al. Dopamine D1/D5 receptor modulation of excitatory synaptic inputs to layer V prefrontal cortex neurons. Proc Natl Acad Sci USA 2001;98: 301–306. 30. Williams G, Goldman-Rakic P. Modulation of memory fields by dopamine D1 receptors in prefrontal cortex. Nature 1995; 376:572–575. 31. Williams G, Goldman-Rakic P. Blockade of dopamine D1 receptors enhances memory fields of prefrontal neurons in primate cerebral cortex. Nature 1995;376:572–575. 32. Yang C, Seamans J, Gorelova N. Developing a neuronal model for the pathophysiology of schizophrenia based on the nature of electrophysiological actions of the dopamine in the prefrontal cortex. Neuropsychopharmacology 1999;21:161–194. 33. Fahn S, Elton R, and members of the UPDRS development committee. Unified Parkinson’s disease rating scale. In: Fahn S, Marsden CD, Goldstein M, Calne DB, eds. Recent developments in Parkinson’s disease. Macmillan: New York, 1987. 34. Yang Y, et al. Fast 3D functional magnetic resonance imaging at 1.5T with spiral acquisition. Magn Reson Med 1996;36: 620 – 626. 35. Callicott JH, Mattay VS, Bertolino A, et al. Physiological characteristics of capacity constraints in working memory as revealed by functional MRI. Cereb Cortex 1999;9:20 –26. 36. Mattay V, Callicott JH, Bertolino A, et al. Effects of dextroamphetamine on cognitive performance and cortical activation. NeuroImage 2000;12:268 –275. 37. Ostuni J, Santha AK, Mattay VS, et al. Analysis of interpolation effects in the reslicing of functional MR images. J Comput Assist Tomogr 1997;21:803– 810. 38. Mattay VS, Frank JA, Santha AK, et al. Whole-brain functional mapping with isotropic MR imaging. Radiology 1996;201: 399 – 404. 39. Woods RP, Grafton ST, Watson JD, et al. Automated image registration. II. Intersubject validation of linear and nonlinear models. J Comput Assist Tomogr 1998;22:153–165. 40. Woods RP, Grafton ST, Holmes CJ, et al. Automated image registration. I. General methods and intrasubject, intramodality validation. J Comput Assist Tomogr 1998;22:139 –152. 41. Callicott J, Bertolino A, Mattay VS, et al. Physiological dysfunction of the dorsolateral prefrontal cortex in schizophrenia revisited. Cereb Cortex 2000;10:1078 –1092. 42. Gaspar P. Topography and collateralization of the dopaminergic projections to motor and lateral prefrontal cortex in owl monkeys. J Comp Neurol 1992;325:1–21. 43. Goldman-Rakic P, Muly E, Williams G. D1 receptors in prefrontal cells and circuits. Brain Res Rev 2000;31:295–301. 44. Sawaguchi T, Goldman-Rakic PS. D1 Dopamine receptors in pre-frontal cortex: involvement in working memory. Science 1991;251:947–950. 45. Sawaguchi T. The role of D1-dopamine receptors in working memory-guided movements mediated by frontal cortical areas. Parkinsonism Relat Disord 2001;7:9 –19. 46. Murphy BL, Arnsten AF, Goldman-Rakic PS, Roth RH. Increased dopamine turnover in the prefrontal cortex impairs spatial working memory performance in rats and monkeys. Proc Natl Acad Sci USA 1996;93:1325–1329. 47. Goldman-Rakic P, Leranth C, Williams SM, et al. Dopamine synaptic complex with pyramidal neurons in primate cerebral cortex. Proc Natl Acad Sci USA 1989;86:9015–9019. 48. Sesack S, Snyder C, Lewis D. Axon terminals immunolabeled for dopamine or tyrosine hydroxylase synapse on GABAimmunoreactive dendrites in rat and monkey cortex. J Comp Neurol 1995;363:264 –280. 49. O’Donnell P, Lewis BL, Lerman D, et al. Effects of neonatal hippocampal lesions in prefrontal cortical pyramidal cell responses to VTA stimulation. Soc Neurosci Abs 1999;25:1569. 50. Lewis B, O’Donnell P. Ventral tegmental area afferents to the prefrontal cortex maintain membrane potential “up” states in pyramidal neurons via D1 dopamine receptors. Cereb Cortex 2000;10:1168 –1175. 51. Watanabe M, Kodama T, Hikosaka K. Increase of extracellular dopamine in primate prefrontal cortex during a working memory task. J Neurophysiol 1997;78:2795–2798. Mattay et al: Dopaminergic Modulation of Cortical Function 163 52. Sawaguchi T. Catecholaminergic effects on neuronal activity related to a delayed response task in monkey prefrontal cortex. J Neurophysiol 1990;63:1385–1400. 53. Sawaguchi T. Effects of dopamine antagonists on neuronal activity related to a delayed response task in monkey prefrontal cortex. J Neurophysiol 1990;63:1401–1412. 54. Sawaguchi T. Catecholamine sensitivities of neurons related to a visual reaction time task in the monkey prefrontal cortex. J Neurophysiol 1987;58:1100 –1122. 55. Daniel DG, Berman KF, Weinberger DR. The effect of apomorphine on regional cerebral blood flow in schizophrenia. J Neuropsychiatry Clin Neurosci 1989;1:377–384. 56. Friston KJ, Grasby PM, Bench CJ, et al. Measuring the neuromodulatory effects of drugs in man with positron emission tomography. Neurosci Lett 1992;141:106 –110. 57. Dolan RJ, Fletcher P, Frith CD, et al. Dopaminergic modulation of impaired cognitive activation in the anterior cingulate cortex in schizophrenia. Nature 1995;378:180 –182. 58. Daniel DG, Weinberger DR, Jones DW, et al. The effect of amphetamine on regional cerebral blood flow during cognitive activation in schizophrenia. J Neurosci 1991;11:1907–1917. 164 Annals of Neurology Vol 51 No 2 February 2002 59. Cools R, Stefanova E, Barker RA, et al. Dopaminergic modulation of high-level cognition in Parkinson’s disease: The role of prefrontal cortex revealed by PET. In press. 60. Bernheimer H, Birkmayer W, Hornykiewicz O. Brain dopamine and the syndromes of Parkinson and Huntington. J Neurol Sci 1973;20:415– 455. 61. Jellinger K. Post mortem studies in Parkinson’s disease—is it possible to detect brain areas for specific symptoms. J Neural Transm 1999;56(suppl):1–29. 62. Gaspar P, Duyckaerts C, Alvarez C, et al. Alterations of dopaminergic and noradrenergic innervations in motor cortex in Parkinson’s disease. Ann Neurol 1991;30(3):365–374. 63. Verma A, Moghaddam B. NMDA receptor antagonists impair prefrontal cortex function as assessed via apatial delayed alteration performance in rats: modulation by dopamine. J Neurosci 1996;16:373–379. 64. Arnsten A. Catecholamine modulation of prefrontal cortical cognitive function. Trends Cogn Sci 1998;2:436 – 447. 65. Arnsten AF. Catecholamine regulation of the prefrontal cortex. J Psychopharmacol 1997;11:151–162.