dynamic changes in cerebello-thalamo-cortical motor circuitry during progression … · 2010. 3....
TRANSCRIPT
DC
SMa
Cb
Nc
Hd
S
Attcs(msivcof(gueflmtwCCbpcstgt
*UD�EAceeiPcSsV
Neuroscience 166 (2010) 712–719
0d
YNAMIC CHANGES IN CEREBELLO-THALAMO-CORTICAL MOTOR
IRCUITRY DURING PROGRESSION OF PARKINSON’S DISEASEgpL
Kl
Peotmtutfa1asmitg
maapirn
tnc(2(ewCah(2cet
. SEN,a,b A. KAWAGUCHI,b Y. TRUONG,b
. M. LEWISc,d AND X. HUANGa,c,d*
Department of Neurology, School of Medicine, University of Northarolina, Chapel Hill, NC 27599, USA
Department of Biostatistics, School of Public Health, University oforth Carolina, Chapel Hill, NC 27599, USA
Department of Neurology, Pennsylvania State University, Milton S.ershey Medical Center, Hershey, PA 17033, USA
Department of Pharmacology, Pennsylvania State University, Milton. Hershey Medical Center, Hershey, PA 17033, USA
bstract—Both the basal ganglia and cerebellum are knowno influence cortical motor and motor-associated areas viahe thalamus. Whereas striato-thalamo-cortical (STC) motorircuit dysfunction has been implicated clearly in Parkin-on’s disease (PD), the role of the cerebello-thalamo-corticalCTC) motor circuit has not been well defined. Functional
agnetic resonance imaging (fMRI) is a convenient tool fortudying the role of the CTC in vivo in PD patients, but largenter-individual differences in fMRI activation patterns requireery large numbers of subjects in order to interpret data fromross-sectional, case control studies. To understand the rolef the CTC during PD progression, we obtained longitudinalMRI 2 years apart from 5 PD (57�8 yr) and five Controls57�9 yr) performing either externally- (EG) or internally-uided (IG) sequential finger movements. All PD subjects hadnilateral motor symptoms at baseline, but developed bilat-ral symptoms at follow-up. Within-group analyses were per-ormed by comparing fMRI activation patterns between base-ine and follow-up scans. Between-group comparisons were
ade by contrasting fMRI activation patterns generated byhe more-affected and less-affected hands of PD subjectsith the mean of the dominant and non-dominant hands ofontrols. Compared to baseline, Controls showed changes inTC circuits, but PD subjects had increased recruitment ofoth cortical motor-associated and cerebellar areas. Com-ared to Controls, PD subjects demonstrated augmented re-ruitment of CTC circuits over time that was statisticallyignificant when the IG task was performed by the hand thatransitioned from non-symptomatic to symptomatic. This lon-itudinal fMRI study demonstrates increased recruitment ofhe CTC motor circuit concomitant with PD progression, sug-
Correspondence to: X. Huang, Department of Neurology, Penn Stateniversity – Milton S. Hershey Medical Center, H037, 500 Universityrive, Hershey, PA 17033-0850, USA. Tel: �1-717-531-1803; fax:1-717-531-0963.-mail address: [email protected] (X. Huang).bbreviations: BG, basal ganglia (Putamen and Globus Pallidus); CB,erebellum; CHTC, cerebellar hemisphere thalamic cortical; CTC, cer-bello-thalamo-cortical; CVTC, cerebellar vermis thalamic cortical; EG,xternally guided; fMRI, functional magnetic resonance imaging; IG,
nternally guided; MANOVA, multivariate analysis of variance; PD,arkinson’s disease; PMC, primary motor cortex; PreMC, pre-motorortex; ROI, region of interest; SFM, sequential finger movement;MA, supplementary motor area; SMC, somatosensory cortex; SPGR,
upoiled gradient recall; STC, striato-thalamo-cortical; TH, thalamus;m, vermis/paravermis (including bilateral dentate nuclei).
306-4522/10 $ - see front matter © 2010 IBRO. Published by Elsevier Ltd. All rightoi:10.1016/j.neuroscience.2009.12.036
712
esting a role of the CTC circuit in accommodation to, orathophysiology of, PD. © 2010 IBRO. Published by Elseviertd. All rights reserved.
ey words: Parkinson’s disease, longitudinal, FMRI, cerebel-um, neurocircuits, motor control.
arkinson’s disease (PD) is a progressive, neurodegen-rative disorder characterized by asymmetrical onsetf motor symptoms such as bradykinesia, rigidity, andremor. The principal pathology in PD is the loss of dopa-ine neurons in the substantia nigra pars compacta, but
he exact pathophysiology of all PD motor symptoms isnclear. Much of the current understanding of motor con-rol is guided by an elegant model of basal ganglia (BG)unction first developed nearly two decades ago (Albin etl., 1989; Alexander et al., 1990, 1986; DeLong et al.,984). In this classic model, the BG was suggested toffect motor control by modulating cortical function throughtriato-thalamo-cortical motor circuits (STC). Although thisodel provides an adequate explanation for bradykinesia,
t does not address many other aspects of PD motor symp-oms including compensation that occurs during PD pro-ression.
The cerebellum also is an important component inotor control, and is known to influence cerebral corticalctivity via cerebello-thalamo-cortical (CTC) circuits (Afifind Bergman, 1998). These CTC circuits have been im-licated in somatosensory integration (Manzoni, 2007) and
nformation updating (Bonnefoi-Kyriacou et al., 1998). Theole of the cerebellum in PD pathophysiology, however, isot well understood.
The motor deficits of PD are primarily related to voli-ional initiation of movement that has been termed inter-ally-guided (IG). The internal motor deficits in PD subjectsan be overcome by external visual or auditory cuesChuma et al., 2006; Jahanshahi et al., 1995; Nowak et al.,006). Recent functional magnetic resonance imagingfMRI) studies suggested a role for the cerebellar circuit inxternally-guided (EG) tasks (Debaere et al., 2003; Tani-aki et al., 2003, 2006). It has been postulated that theTC circuit may provide a potential compensatory mech-nism in PD to overcome the deficits in the STC circuit, anypothesis supported by several recent fMRI studiesCerasa et al., 2006; Lewis et al., 2007; Taniwaki et al.,006). These latter studies, however, were done usingross-sectional designs. The large inter-individual differ-nces in fMRI activation patterns make it difficult to studyemporal changes in neurocircuitry during PD progression
sing a cross-sectional, case control design.s reserved.
iBscbaactwcsidasp
S
Fmrc1tEc1atbsmntoawafbWp
F
IEhllrmmtfsvei
le1.
Dem
ogra
phic
sof
Par
kins
on’s
dise
ase
subj
ects
inth
isst
udy
bjec
t#
Age
atba
selin
e(y
)H
ande
dnes
sP
rese
ntin
gsy
mpt
oms
Sca
nin
terv
al(m
o)U
PD
RS
-III
scor
esin
“off”
stag
eA
ntip
arki
nson
drug
sat
base
linea
Ant
ipar
kins
ondr
ugs
atfo
llow
-up
Bas
elin
eF
ollo
w-u
p
52R
ight
Dec
reas
edar
msw
ing
and
left
hand
dext
erity
222
3P
ram
ipex
ole
1.5
tidC
arbi
dopa
/Lev
odop
a/E
ntac
apon
e50
tidP
ram
ipex
ole
1tid
Sel
egili
ne5
bid
58R
ight
Ten
sene
ssan
dde
crea
sed
left
hand
dext
erity
213
3C
arbi
dopa
/Lev
odop
a25
/100
tidC
arbi
dopa
/Lev
odop
a25
/100
tid;
37.5
/150
qam
50R
ight
Pos
tura
ltre
mor
ofrig
htha
nd27
44.
5P
ram
ipex
ole
0.5
tidP
ram
ipex
ole
0.5
tidS
eleg
iline
5bi
d55
Rig
htS
tiffn
ess
and
decr
ease
drig
htha
ndde
xter
ity20
67.
5P
ram
ipex
ole
1.5
tidP
ram
ipex
ole
1.5
tidS
eleg
iline
5bi
d71
Rig
htR
ight
hand
rest
ing
trem
or24
710
.5C
arbi
dopa
/Lev
odop
a/E
ntac
apon
e50
tidP
ram
ipex
ole
0.37
5tid
Pra
mip
exol
e1.
5tid
Sel
egili
ne5
bid
oses
are
inm
gex
cept
asno
ted;
qam
�ev
ery
mor
ning
;bi
d�tw
ice
ada
y;tid
�th
ree
times
ada
y.
S. Sen et al. / Neuroscience 166 (2010) 712–719 713
Two recent positron emission tomography (PET) stud-es reported longitudinal changes in PD subjects in cortical,G, and cerebellar structures. Huang et al. (2007) mea-ured dopamine transporter binding and cortical/subcorti-al glucose metabolism changes in PD subjects over time,ut their imaging protocol did not involve a motor task. Innother study, Carbon et al. (2007) assessed both brainctivation and motor performance during an externally-ued motor paradigm in PD and control subjects. Althoughhe changes in brain activation in PD subjects over timeere compared, there were no comparisons made tohanges in Controls (Carbon et al., 2007). The currenttudy investigates the changes in the recruitment patternsn STC and CTC motor circuits in both EG and IG tasksuring the progression of PD as compared to changes oversimilar time period in age-matched Controls. Our data
how clearly that function of the CTC circuit changes withrogression of PD.
EXPERIMENTAL PROCEDURES
ubjects
ive newly diagnosed PD subjects (57�8 yrs) with unilateralotor symptoms and five healthy control subjects (57�9 yrs) were
ecruited for this study. All subjects were strongly right-handed asonfirmed by the Edinburgh Handedness Inventory (Oldfield,971). Three of five PD subjects started with right-hand symp-oms, whereas the other two started with left-hand symptoms.ach PD subject was diagnosed by a movement disorders spe-ialist (XH) based on previously published criteria (Gibb and Lees,988). All subjects were scanned twice approximately 2 yearspart (21.6�4.5 months for PD, and 20.0�2.4 months for Con-rols). All PD subjects had unilateral motor symptoms at baseline,ut developed bilateral symptoms at follow-up. Prior to each fMRIcan, PD subjects were asked to withhold their antiparkinsonedication for approximately 12 h, and all subjects were askedot to imbibe any caffeinated or alcoholic beverages for 24 h prioro the scan (Table 1). Controls were found free of any signs ofther neurological and psychological deficits, and were not takingny drugs with psychoactive properties. Subjects in both groupsere negative for hypothyroidism, vitamin B12 or folate deficiency,nd were free of kidney or liver disease. The study protocolollowed the Helsinki principles, and was reviewed and approvedy the University of North Carolina Institutional Review Board.ritten informed consent was obtained from all subjects who
articipated in the study.
MRI data acquisition
mages were acquired on a 3.0 Tesla Siemens scanner (Siemens,rlangen, Germany) with a birdcage-type standard quadratureead coil and an advanced nuclear magnetic resonance echop-
anar system. The head was positioned along the canthomeataline and foam padding was used to limit head motion. High-esolution T1 weighted anatomical images (3D SPGR, TR�7.7s, TE�14 ms, flip angle�25°, voxel dimensions 1.0�1.0�1.0m3, 176�256 voxels, 160 slices) were acquired for co-registra-
ion and normalization of functional images. A total of 49 co-planarunctional images were acquired using a gradient echoplanarequence (TR�3000 ms, TE�30 ms, flip angle�80°, NEX�1,oxel dimensions 3.0�3.0�3.0 mm3, imaging matrix 64�64 vox-ls). Two radio frequency excitations were performed prior to
mage acquisition to achieve steady-state transverse relaxation.
TabSu
1 2 3 4 5 aD
F
Tsatoetlp(w(tsr
F
Tcrtk
D
csmpptr
at
h(v2ciaetmsssbdsraRwb
sRce(tsiP
Fab d, the re
T
P
SS
S. Sen et al. / Neuroscience 166 (2010) 712–719714
MRI activation paradigm
his study used a modified activation paradigm based on previoustudies in control and PD subjects (Lewis et al., 2007; Sabatini etl., 2000). Briefly, the paradigm consisted of sequential finger-
apping movements (SFM) at 0.5 tap/second using either the rightr the left hand. The sequences were presented with instructionsither to follow the hands on the screen (EG task) or to continue
he finger-tapping sequence (IG task). Each SFM block was 60 song, and each block was preceded and followed by a 30 s rest (R)eriod. Each run consisted of four blocks of rest, EG, and IG taskstotal 10 min, Fig. 1). The finger-tapping sequences of each blockere alternated to prevent memorization from previous blocks
see Table 2). In order to obtain adequate performance on theasks, all subjects practiced the task for about 20 min prior to thecanning session, demonstrating greater than 90% accuracy. Twouns of fMRI data from each subject were included in the analysis.
MRI image pre-processing
he fMRI data were preprocessed using SPM5 software (Well-ome Trust Center for Neuroimaging, London, UK) for spatialealignment and motion correction. The spatial smoothing of func-ional time series was performed with a Gaussian smoothingernel with full width at half maximum (FWHM) � 6�6�6 mm3.
ata analysis
First level statistical analysis. Spatial transformation into aommon coordinate space as traditionally done for most fMRItudies deemphasizes inherent anatomical variability of the hu-an brain that may be amplified by aging and neurodegenerativerocesses. In addition, covarying regions are often of great im-ortance, but are not included in the group methods described inraditional SPM analyses. To circumvent these problems, we car-ied out first level statistical analyses in native space, and gener-
ig. 1. Functional MRI activation paradigm. Subjects performed fingelternating between their two hands. Each run consisted of four blocklocks. For interpretation of the references to color in this figure legen
able 2. Finger sequences used in the fMRI paradigm
aradigm Description of sequences
low sequence number 1 Thumb to digit 2 ¡3¡4¡5
low sequence number 2 Thumb to digit 3 ¡5¡2¡4¡ opented individual T-maps comparing the activation patterns of eachask to rest (i.e., EG vs. Rest and IG vs. Rest).
The anatomical regions of interest (ROIs) were defined onigh resolution T1 images using automatic segmentation softwareAutoSeg, NeuroImage Research and Analysis Laboratories, Uni-ersity of North Carolina at Chapel Hill, NC, USA; Gouttard et al.,007; Joshi et al., 2004) for each subject. This permits statisticalomparisons of similar brain areas across subjects without warp-ng (“normalizing”) the brain. The percent of voxels activated with
t-value �1.96 (corresponding to a P�0.05) were calculated forach ROI in the bilateral STC [Putamen/globus pallidus (BG),halamus (Th), supplementary motor area (SMA), and primaryotor cortex (PMC)] and CTC circuits. For the purpose of this
tudy the CTC circuit was divided into CHTC [cerebellar hemi-phere (CB), Th, lateral premotor cortex (PreMC), and somato-ensory cortex (SMC)] and CVTC (vermis/paravermis includingilateral dentate nuclei (Vm), TH, PreMC and SMC) circuits. Weivided the cerebellum into three segments of two lateral hemi-pheres and one midline vermis/paravermis because theseegions have been suggested to subserve different functions (Afifind Bergman, 1998). A statistical method that compared multipleOIs together, namely multivariate analysis of variance (MANOVA),as employed to compare the neurocircuitry changes within andetween groups as described below.
Within group comparisons between follow-up and baselinecans. Collective values of the percentage of voxels activated inOIs that constitute each of the bilateral STC, CHTC, and CVTCircuits were treated as multivariate dependent variables. Forach group (PD or Control), a randomized block design procedurea special case of two-way MANOVA) was employed to comparehe fMRI activation differences between baseline and follow-upcans. This is a multivariate version of the paired t-test for univar-
ate response settings. The analysis was carried out using SASROC GLM, option MANOVA with two factors: SUBJECT (1
sequences in the following order: Rest, EG-guided, and IG-guided,asks and four blocks of IG-tasks lasting 60 s per block with 30-s restader is referred to the Web version of this article.
and close fists twice ¡ Repeat ¡ Return to beginning of sequence
r-tappings of EG-t
¡ open
and close fists twice ¡ Repeat ¡ Return to beginning of sequence
tfntfb
oaaCcatvvw
u(s
Wf
AsjMtR
ndlho(d
CrengECa
Bl
WnsitwtP
F(mtwnmE
BmSof(cStacm
PceiCs
TC
F
IIICCC
taPMta
b
Pc
Sd
Pe
Tf
a
S. Sen et al. / Neuroscience 166 (2010) 712–719 715
hrough 5 treated as blocks) and TIME (0 for baseline and 1 forollow-up). The blocking effect (SUBJECT) is introduced to elimi-ate the subject-to-subject variations, and our objective was to
est the TIME effect (i.e., the difference between baseline andollow-up scans). Analyses were conducted independently foroth EG and IG tasks within each group.
Between group comparison of PD and Controls. Since twof the five PD subjects had left-sided symptoms, both the more-ffected and less-affected sides of PD subjects were compared ton averaged activity of the dominant and non-dominant hands ofontrols. ROIs that constitute the bilateral STC, CHTC, and CVTCircuits were treated as co-multivariate response variables. In thisnalysis, the change in percent activation over time in the ROIs ofhese circuits in each subject was the multivariate dependentariable, whereas the independent variable was PD status (1�PDs. 0�Controls). Task (EG or IG) and PD status (PD or Controls)ere “dummy-coded” as either 1 or 0.
All comparisons were conducted by utilizing multiple MANOVAssing the PROC GLM command with option MANOVA in SASSystem 9.1, SAS Inc., Cary, NC). The significance of a giventructure in a circuit was probed by a simple t-test.
RESULTS
ithin group comparison between baseline andollow-up
s described in the Methods, the ROI approach permitstatistical comparison of similar brain areas across sub-ects without warping (“normalizing”) the brain. In addition,
ANOVA allows comparison of collective values from mul-iple ROIs between groups. The results of this combinedOI and MANOVA approach are described below.
PD group. PD subjects did not show statistically sig-ificant changes in recruitment of STC and CTC circuitsuring either the EG or IG task over time when using their
ess-affected hand. When using their more-affected hand,owever, PD subjects had increased recruitment over timef the contralateral STC circuit during the IG taskP�0.045, Table 3, Fig. 2A) and the ipsilateral CHTC circuituring the EG task (P�0.001, Table 3, Fig. 2B).
Control group. When using their dominant hand,ontrol subjects did not show any significant changes in
ecruitment patterns of the STC and CTC circuits duringither the EG or IG task over time. When using theiron-dominant hand, however, Control subjects hadreater activation of the ipsilateral CHTC circuit during theG task (P�0.016, Table 4, Fig. 3B) and contralateral
VTC circuit during both the EG and IG tasks (P�0.036nd P�0.042, respectively; Table 4, Fig. 3B).
etween groups neurocircuitry analysis: comparingongitudinal changes between PD and Controls
hen using the less-affected hand that transitioned fromon-symptomatic to symptomatic during the study, PDubjects showed significantly increased recruitment of thepsilateral CVTC circuit over time during IG tasks comparedo Controls (P�0.043, Table 5, Fig. 4). In addition, thereas a trend toward significant increased recruitment of
he ipsilateral C TC circuit over time during IG tasks in
HD subjects compared to Controls (P�0.083, Table 5, t
ig. 4). There were no statistically significant differencesor trends towards significance) in the recruitment ofotor circuits between PD and Controls during the EG
ask (Table 5). When using the more-affected hand thatas already symptomatic at baseline, PD subjects didot demonstrate significant differences in the recruit-ent of motor circuits from that of Controls during eitherG or IG tasks (Table 5).
DISCUSSION
oth the BG and cerebellum are known to influence corticalotor and associated areas via the thalamus, yet only theTC circuit has been implicated clearly in the pathophysiol-gy of PD. Whereas previous cross-sectional studies using
MRI have demonstrated increased activity in cerebellumCerasa et al., 2006; Eckert et al., 2006) and numerousortical areas (Eckert et al., 2006; Haslinger et al., 2001;abatini et al., 2000) in PD, we believe this is the first longi-
udinal fMRI study to assess the temporal changes in STCnd CTC circuits during PD progression. The results from theurrent study are consistent with the hypothesis that CTCotor circuits are involved in PD progression.
As noted earlier, longitudinal PET imaging studies ofD subjects found increased activation in cortical, BG, anderebellar structures over time (Carbon et al., 2007; Huangt al., 2007). The current fMRI data show a similarly
ncreased activation over time in the STC, as well asTC, circuits in PD subjects. Collectively, these studiesuggest that there are decreases in the efficiency (i.e.,
able 3. MANOVA results comparing longitudinal changes in STC andTC motor circuits between baseline and follow-up in PD subjects
ollow-up vs. Baseline
PD(less-affected) PD(more-affected)
EG IG EG IG
psilateral STCa 0.851 0.935 0.738 0.496psilateral CHTCb 0.930 0.307 0.001* 0.501psilateral CVTCc 0.692 0.341 0.094 0.646ontralateral STCd 0.731 0.130 0.863 0.045*ontralateral CHTCe 0.433 0.596 0.369 0.795ontralateral CVTCf 0.249 0.389 0.168 0.865
The percent of voxels activated with a t value�1.96 (correspondingo a P�0.05) was calculated for each ROI in the STC (BG, Th, SMA,nd PMC), CHTC (CB, Th, PreMC, and SMC), and CVTC (Vm, Th,reMC and SMC) circuits. Network analyses were performed usingANOVA with the changes in percent activation of individual ROIs as
he dependent variables. * Represents significant difference.Ipsilateral STC was defined as ipsilateral BG, Th, SMA, and PMC.Ipsilateral CHTC was defined as contralateral CB and ipsilateral Th,reMC, and SMC.Ipsilateral CVTC was defined as Vm and ipsilateral Th, PreMC, andMC.Contralateral STC was defined as contralateral BG, Th, SMA, andMC.Contralateral CHTC was defined as ipsilateral CB and contralateralh, PreMC, and SMC.Contralateral CVTC was defined as Vm and contralateral Th, PreMC,nd SMC.
he need for more recruitment) of structures in STC and
Cd
s3ocaaccco
rave
eMbac
Inm
TitfCsfa
Ffcf . Circuits
T
F
IIICCC
Cia
b
c
d
e
f
S. Sen et al. / Neuroscience 166 (2010) 712–719716
TC circuits used to perform the same motor tasksuring PD progression.
Interestingly, control subjects also showed statisticallyignificant changes in the CTC circuit (see Table 4, and Fig.), emphasizing the importance of using Controls for this typef study. The biological relevance of time-dependent CTChanges in Controls is unclear, but may reflect an age-relatedlteration of this circuit as previously reported (Alexander etl., 2008; Onozuka et al., 2003). Nevertheless, PD subjectslearly demonstrated augmented recruitment of CTC circuitsompared to those of normal aging Controls (Table 5), impli-ating the CTC circuit in accommodation to, or progressionf, PD above and beyond normal aging.
Earlier, we had proposed a model that integrated theole of the CTC circuit into PD pathophysiology (Lewis etl., 2007) that may assimilate some of the seemingly di-ergent basic and clinical results in PD that cannot bexplained only by dysfunction of the STC circuit (Buhmann
ig. 2. Comparison of percent activation changes in (A) STC and (B) Cor each circuit were deduced from MANOVA comparison between bomposing each circuit (see Experimental Procedures). Bars in the figor PD subjects for simple visualization (not for MANOVA calculation)
able 4. MANOVA results comparing longitudinal changes in STC an
ollow-up vs. Baseline
Controls(Non-dominant hand
EG
psilateral STCa 0.734psilateral CHTCb 0.016*psilateral CVTCc 0.212ontralateral STCd 0.811ontralateral CHTCe 0.126ontralateral CVTCf 0.036*
The percent of voxels activated with a t value�1.96 (corresponding t
HTC (CB, Th, PreMC, and SMC), and CVTC (Vm, Th, PreMC and SMCn percent activation of individual ROIs as the dependent variables. *Ipsilateral STC was defined as ipsilateral BG, Th, SMA, and PMC.Ipsilateral CHTC was defined as contralateral CB and ipsilateral Th,Ipsilateral CVTC was defined as Vm and ipsilateral Th, PreMC, andContralateral STC was defined as contralateral BG, Th, SMA, and PContralateral C TC was defined as ipsilateral CB and contralateral
HContralateral CVTC was defined as Vm and contralateral Th, PreMC, and SM
t al., 2003; Cerasa et al., 2006; Haslinger et al., 2001;attay et al., 2002; Sabatini et al., 2000). As we discusselow, we now suggest a modification of this model toccount for the dynamical changes in both STC and CTCircuits during PD progression (Fig. 5).
ncreased CTC recruitment occurs during theon-symptomatic to symptomatic transition in PDotor progression
he increased CTC recruitment in PD subjects was signif-cantly different from Controls only when the hand thatransitioned from non-symptomatic to symptomatic per-ormed the tasks. This result suggests that changes in theTC motor circuit play a role in the emergence of PD motorymptoms. The lack of significant changes in CTC circuitsor the hand that was symptomatic at baseline may reflectn already established compensatory pattern of CTC ac-
ts over time within PD subjects using the more affected hand. P-valuesnd follow-up of the collective percent activation changes within ROIsresent mean�SEM of percent activation changes within a given ROIare defined as listed in the legend of Table 2.
otor circuits between baseline and follow-up in Controls
Controls(Dominant hand)
IG EG IG
0.517 0.529 0.2690.575 0.913 0.7570.438 0.744 0.7490.420 0.251 0.3300.212 0.139 0.5520.042* 0.115 0.457
05) was calculated for each ROI in the STC (BG, Th, SMA, and PMC),. Network analyses were performed using MANOVA with the changes
nts significant differences.
and SMC.
C, and SMC.
TC circuiaseline aures rep
d CTC m
)
o a P�0.) circuits
Represe
PreMC,SMC.MC.
Th, PreM
C.
ts
Ic
TttIc
tasadtcF
cssdrascst2s
oCpvnswpfta
Im
T
Ffcf . Circuits
TC
IIICCC
taPMoaa
b
Pc
Sd
e
Tf
a
S. Sen et al. / Neuroscience 166 (2010) 712–719 717
ivity at baseline without an additional increase in compen-atory activity over time.
ncreased CTC recruitment may representompensation for inadequate STC function
he significant differences in augmented CTC activity be-ween PD and Controls over time were seen only duringhe IG, but not EG, task (Fig. 4). This is not surprising sinceG tasks are thought primarily to be encoded in the STCircuit, whereas EG tasks are thought to be processed
ig. 3. Comparison of percent activation changes in (A) STC and (B)or each circuit were deduced from MANOVA comparisons between bomposing each circuit (see Experimental Procedures). Bars in the figor PD subjects for simple visualization (not for MANOVA calculation)
able 5. MANOVA results comparing longitudinal changes in STC andTC motor circuits between PD and Control subjects
PD(less-affected) vs.Control
PD(more-affected)
vs. Control
EG IG EG IG
psilateral STCa 0.879 0.703 0.477 0.744psilateral CHTCb 0.364 0.083 0.699 0.678psilateral CVTCc 0.419 0.043* 0.710 0.767ontralateral STCd 0.529 0.731 0.613 0.279ontralateral CHTCe 0.654 0.406 0.611 0.603ontralateral CVTCf 0.455 0.277 0.647 0.645
The percent of voxels activated with a t value�1.96 (correspondingo a P�0.05) were calculated for each ROI in the STC (BG, Th, SMA,nd PMC), CHTC (CB, Th, PreMC, and SMC), and CVTC (Vm, Th,reMC and SMC) circuits. Network analyses were performed usingANOVA with the changes in percent activation of individual ROIsver time as the dependent variables and PD status (PD vs. Control)s the independent variable. * Represents significant difference.Ipsilateral STC was defined as ipsilateral BG, Th, SMA, and PMC.Ipsilateral CHTC was defined as contralateral CB and ipsilateral Th,reMC, and SMC.Ipsilateral CVTC was defined as Vm and ipsilateral Th, PreMC, andMC.Contralateral STC was defined as contralateral BG, Th, SMA, and PMC.Contralateral CHTC was defined as ipsilateral CB and contralateralh, PreMC, and SMC.Contralateral C TC was defined as Vm and contralateral Th, PreMC,
tV
nd SMC.
hrough the CTC circuit (see Fig. 5, upper panel) (Lewis etl., 2007; Taniwaki et al., 2006). Thus, the current datauggest that the emergence of symptoms on the less-ffected side during PD progression may be related to aecompensation of the initially functional STC circuit. Thishen leads to the compensatory recruitment of the CTCircuit that permits proper execution of IG motor tasks (seeig. 5 lower panel).
The current study, however, did not demonstrate signifi-ant differences in the recruitment of STC circuits in PDubjects compared to Controls over time. Although this mighteem unexpected at first, it very well may reflect a residualrug effect. Antiparkinson medication has been shown toeduce activity in the STC circuit in PD subjects (Asanuma etl., 2006). The fMRI scans for PD subjects in the currenttudy were obtained when subjects were “off” all PD medi-ations for more than 12 h. Although this “off” stage for PDubjects has been used frequently (Langston et al., 1992),here well may be residual drug effects on STC circuits (Fahn,005). Indeed all of our subjects, except for one, were onome form of dopamine agonist (Table 2).
At this point it worth mentioning that the main limitationf this study is the variability in the data between PD andontrol groups. In comparing PD vs. Controls, our two-sam-le test used the Hotelling’s T2-statistic that includes theariability of the pooled samples, which is larger than theon-pooled sample variance of each group alone. A loss ofignificance can result from small increases in variability, ande believe this affected our analyses, particularly when com-aring the significant changes observed with the more-af-
ected hand in PD subjects to changes in Controls performinghe same task. Future studies with increased sample sizesre needed to test our hypothesis further.
mplications for a striato/cerebello-thalamocorticalodel in PD
he hypothesis that the STC and CTC circuits are func-
uits over time within Controls using the non-dominant hand. P-valuesnd follow-up of the collective percent activation changes within ROIsresent mean�SEM of percent activation changes within a given ROIare defined as listed in the legend of Table 3.
CTC circaseline aures rep
ionally related circuits that are influenced by PD (summa-
rcitcec
pctpCht
Fwh Is compoa MANOV
FtIt
S. Sen et al. / Neuroscience 166 (2010) 712–719718
ized in Fig. 5) offers a more comprehensive view of motorontrol that leads to a variety of testable hypotheses. First,n the normal condition, EG tasks are primarily processedhrough the CTC circuit, with the recruitment of the STCircuit (Fig. 5, Panel A); whereas IG tasks are primarilyncoded in the STC circuit, with recruitment of the CTCircuit (Fig. 5, Panel B). Second, in PD EG tasks also are
ig. 4. Comparison of percent activation changes in CTC circuits overere deduced from MANOVA comparison between PD subjects usingands) of the collective values of percent activation changes within ROctivation changes within a given ROI for simple visualization (not for
ig. 5. Revised striato/cerebello-thalamocortical model for Parkinson’she CTC circuit with recruitment of the STC circuit, whereas (B) IG tas
n PD subjects, (C) EG tasks also are processed through the CTC circuit (its prhe STC circuit (its primary processing center) leading to compensatory recruitrocessed primarily via the CTC circuit throughout theourse of the disease (Fig. 5, Panel C). Finally, in PD IGasks inadequately activate the STC circuit (its primaryrocessing center) leading to compensatory recruitment ofTC circuits (Fig. 5, Panel D). Future studies that test suchypotheses and validate or modify this striato/cerebello-halamocortical model of PD should assist in understand-
een PD and Controls during EG and IG tasks. P-values for each circuit-affected hand and Controls (average of dominant and non-dominantsing each circuit. Bars in the figures represent mean�SEM of percentA calculation). Circuits are defined as listed in the legend of Table 4.
In the normal condition, (A) EG tasks are primarily processed throughimarily encoded in the STC circuit with recruitment of the CTC circuit.
time betwthe less
disease.ks are pr
imary processing center), whereas (D) IG tasks inadequately activatement of the CTC circuit.
ioc
Apfsaw(A
A
A
A
A
A
A
B
B
C
C
C
D
D
E
F
G
G
H
H
J
J
L
L
M
M
N
O
O
S
T
T
S. Sen et al. / Neuroscience 166 (2010) 712–719 719
ng many unexplained aspects of PD motor symptomatol-gy, as well as shedding light on the function of theseircuits in the normal brain.
cknowledgments—We want to thank all subjects who partici-ated in our studies. We also want to thank Dr. Richard Mailman
or his insightful comments and Dr. Michael Flynn for his mentor-hip of SS. Additionally, we acknowledge the assistance of Rox-nne Poole and the UNC Magnetic Resonance Imaging Center. Thisork was supported in part by NIH grants AG21491 and NS060722
XH), institutional Ruth L. Kirschstein National Research Serviceward T32 ES7018 (SS) and RR00046 (UNC - GCRC).
REFERENCES
fifi AK, Bergman RA (1998) Functional neuroanatomy: text and altas.New York: McGraw-Hill.
lbin RL, Young AB, Penney JB (1989) The functional anatomy ofbasal ganglia disorders. Trends Neurosci 12:366–375.
lexander GE, Chen K, Aschenbrenner M, Merkley TL, Santerre-Lemmon LE, Shamy JL, Skaggs WE, Buonocore MH, Rapp PR,Barnes CA (2008) Age-related regional network of magnetic res-onance imaging gray matter in the rhesus macaque. J Neurosci28:2710–2718.
lexander GE, Crutcher MD, DeLong MR (1990) Basal ganglia-thalamocortical circuits: parallel substrates for motor, oculomotor,“prefrontal” and “limbic” functions. Prog Brain Res 85:119–146.
lexander GE, DeLong MR, Strick PL (1986) Parallel organization offunctionally segregated circuits linking basal ganglia and cortex.Annu Rev Neurosci 9:357–381.
sanuma K, Tang C, Ma Y, Dhawan V, Mattis P, Edwards C, KaplittMG, Feigin A, Eidelberg D (2006) Network modulation in thetreatment of Parkinson’s disease. Brain 129:2667–2678.
onnefoi-Kyriacou B, Legallet E, Lee RG, Trouche E (1998) Spatio-temporaland kinematic analysis of pointing movements performed by cerebellarpatients with limb ataxia. Exp Brain Res 119:460–466.
uhmann C, Glauche V, Sturenburg HJ, Oechsner M, Weiller C,Buchel C (2003) Pharmacologically modulated fMRI—cortical re-sponsiveness to levodopa in drug-naive hemiparkinsonian pa-tients. Brain 126:451–461.
arbon M, Felice GM, Dhawan V, Eidelberg D (2007) Correlates ofmovement initiation and velocity in Parkinson’s disease: a longitu-dinal PET study. Neuroimage 34:361–370.
erasa A, Hagberg GE, Peppe A, Bianciardi M, Gioia MC, Costa A,Castriota-Scanderbeg A, Caltagirone C, Sabatini U (2006) Func-tional changes in the activity of cerebellum and frontostriatal re-gions during externally and internally timed movement in Parkin-son’s disease. Brain Res Bull 71:259–269.
huma T, Faruque RM, Ikoma K, Mano Y (2006) Motor learning ofhands with auditory cue in patients with Parkinson’s disease.J Neural Transm 113:175–185.
ebaere F, Wenderoth N, Sunaert S, Van Hecke P, Swinnen SP(2003) Internal vs external generation of movements: differentialneural pathways involved in bimanual coordination performed inthe presence or absence of augmented visual feedback. Neuroim-age 19:764–776.
eLong MR, Alexander GE, Georgopoulos AP, Crutcher MD, MitchellSJ, Richardson RT (1984) Role of basal ganglia in limb move-ments. Hum Neurobiol 2:235–244.
ckert T, Peschel T, Heinze HJ, Rotte M (2006) Increased pre-SMAactivation in early PD patients during simple self-initiated hand
movements. J Neurol 253:199–207.ahn S (2005) Does levodopa slow or hasten the rate of progressionof Parkinson’s disease? J Neurol 252 (Suppl 4):IV37–IV42.
ibb WR, Lees AJ (1988) The relevance of the Lewy body to thepathogenesis of idiopathic Parkinson’s disease. J Neurol Neuro-surg Psychiatry 51:745–752.
outtard S, Styner M, Joshi S, Smith RG, Hazlett HC, Gerig G (2007)Subcortical structure segmentation using probabilistic atlas priors.Medical Image Computing and Computer Assisted InterventionWorkshop, 37–46. Ref Type: Conference proceeding.
aslinger B, Erhard P, Kampfe N, Boecker H, Rummeny E, SchwaigerM, Conrad B, Ceballos-Baumann AO (2001) Event-related func-tional magnetic resonance imaging in Parkinson’s disease beforeand after levodopa. Brain 124:558–570.
uang C, Tang C, Feigin A, Lesser M, Ma Y, Pourfar M, Dhawan V,Eidelberg D (2007) Changes in network activity with the progres-sion of Parkinson’s disease. Brain 130:1834–1846.
ahanshahi M, Jenkins IH, Brown RG, Marsden CD, Passingham RE,Brooks DJ (1995) Self-initiated versus externally triggered move-ments. I. An investigation using measurement of regional cerebralblood flow with PET and movement-related potentials in normaland Parkinson’s disease subjects. Brain 118 (Pt 4):913–933.
oshi S, Davis B, Jomier M, Gerig G (2004) Unbiased diffeomorphicatlas construction for computational anatomy. Neuroimage 23(Suppl 1):S151–S160.
angston JW, Widner H, Goetz CG, Brooks D, Fahn S, Freeman T,Watts R (1992) Core assessment program for intracerebral trans-plantations (CAPIT). Mov Disord 7:2–13.
ewis MM, Slagle CG, Smith AB, Truong Y, Bai P, McKeown MJ,Mailman RB, Belger A, Huang X (2007) Task specific influences ofParkinson’s disease on the striato-thalamo-cortical and cerebello-thalamo-cortical motor circuitries. Neuroscience 147:224–235.
anzoni D (2007) The cerebellum and sensorimotor coupling: lookingat the problem from the perspective of vestibular reflexes. Cere-bellum 6:24–37.
attay VS, Tessitore A, Callicott JH, Bertolino A, Goldberg TE, ChaseTN, Hyde TM, Weinberger DR (2002) Dopaminergic modulation ofcortical function in patients with Parkinson’s disease. Ann Neurol51:156–164.
owak DA, Tisch S, Hariz M, Limousin P, Topka H, Rothwell JC (2006)Sensory timing cues improve akinesia of grasping movements inParkinson’s disease: a comparison to the effects of subthalamicnucleus stimulation. Mov Disord 21:166–172.
ldfield RC (1971) The assessment and analysis of handedness: theEdinburgh inventory. Neuropsychologia 9:97–113.
nozuka M, Fujita M, Watanabe K, Hirano Y, Niwa M, Nishiyama K,Saito S (2003) Age-related changes in brain regional activity duringchewing: a functional magnetic resonance imaging study. J DentRes 82:657–660.
abatini U, Boulanouar K, Fabre N, Martin F, Carel C, Colonnese C,Bozzao L, Berry I, Montastruc JL, Chollet F, Rascol O (2000)Cortical motor reorganization in akinetic patients with Parkinson’sdisease: a functional MRI study. Brain 123 (Pt 2):394–403.
aniwaki T, Okayama A, Yoshiura T, Nakamura Y, Goto Y, Kira J, TobimatsuS (2003) Reappraisal of the motor role of basal ganglia: a functionalmagnetic resonance image study. J Neurosci 23:3432–3438.
aniwaki T, Okayama A, Yoshiura T, Togao O, Nakamura Y, Yamasaki T,Ogata K, Shigeto H, Ohyagi Y, Kira JI, Tobimatsu S (2006) Functionalnetwork of the basal ganglia and cerebellar motor loops in vivo: differentactivation patterns between self-initiated and externally triggered move-
ments. Neuroimage 31:745–753.(Accepted 14 December 2009)(Available online 24 December 2009)