Alterations in the functional connectivity of a verbal working memory-related brain network in patients with left temporal lobe epilepsy

G Model

ARTICLE IN PRESS

NSL 31392 1–6

Neuroscience Letters xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet

Research article

1

Alterations in the functional connectivity of a verbal working memory-related brain network in patients with left temporal lobe epilepsy

2

3

4

5 6 7 8 9

Q2

Wenli Huang a , Donghong Huang a , Zirong Chen a , Wei Ye b , Zongxia Lv a , Limei Diao c , Jinou Zheng a,∗ a

Department of Neurology, the First Affiliated Hospital, Guangxi Medical University, Nanning 530021, China Department of Radiology, the First Affiliated Hospital, Guangxi Medical University, Nanning 530021, China c Department of Neurology, the First Affiliated Hospital, Guangxi University of Chinese Medicine, Nanning 530023, China b

10

11 12 13 14 15 16 17

h i g h l i g h t s • • • • •

We extracted the VWM network for controls with n-back task from fMRI. We evaluated the RSFC of VWM network in left TLE patients and controls. Left TLE presented a decreased FC in bilateral MFG, IFG, IPL at resting state. The alterations in FC may reflect the impairment of VWM-related network in left TLE. ACCmeanRT (2-back) were not correlated with the FC in VWM-related regions.

18

19 31

a r t i c l e

i n f o

a b s t r a c t

20 21 22 23 24 25

Article history: Received 11 June 2014 Received in revised form 21 April 2015 Accepted 12 June 2015 Available online xxx

26 27 28 29 30

Keywords: Temporal lobe epilepsy Resting-state functional connectivity Verbal working memory

The aim of this study was to investigate the alterations in a verbal working memory (VWM)-related network in left temporal lobe epilepsy (lTLE) at rest. We evaluated 14 patients with lTLE and 14 control subjects by resting-state functional connectivity (RSFC). The region of interest was defined by the voxel with the highest Z-score during a VWM task according to functional magnetic resonance imaging in 16 healthy volunteers. Our study revealed that the network of RSFC was similar to the task-induced network in the healthy volunteers. Moreover, the patients with lTLE exhibited significantly decreased RSFC in the bilateral middle frontal gyrus, the inferior frontal gyrus and the inferior parietal lobule at rest compared to the control subjects. We found no significant correlation between the mean reaction time of the accurate responses in a 2-back task and the mean z-values within the regions that exhibited significant differences in RSFC at the individual level. The alterations in FCs of VWM-related network in lTLE suggested that epileptiform discharges can damage the brain regions, both local focus and remote areas and that the alterations were not associated with VWM performance. © 2015 Published by Elsevier Ireland Ltd.

1. Introduction

Abbreviations: TLE, temporal lobe epilepsy; HS, hippocampal sclerosis; VWM, verbal working memory; RSFC, resting-state functional connectivity; fMRI, functional magnetic resonance imaging; ROI, region of interest; MFG, middle frontal gyrus; IPL, inferior parietal lobule; SPL, superior parietal lobule; IFG, inferior frontal gyrus; PCG, precentral gyrus; PC, precuneus; CPL, cerebellar posterior lobe; SMA, supplementary motor area; SFG, superior frontal gyrus; MTG, middle temporal gyrus; ITG, inferior temporal gyrus; PFC, prefrontal cortex. ∗ Corresponding author. Fax: +86 771 5352627. E-mail address: [email protected] (J. Zheng).

Temporal lobe epilepsy (TLE) is among the most common types of refractory focal epilepsy and is characterized by seizures originating in the temporal lobe structure. TLE has multiple causes, among which hippocampal sclerosis (HS) is particularly notable [1]. Studies have shown that patients with TLE suffer from different degrees of cognitive dysfunction in the domains of intelligence, executive function, alertness and memory [2,3]. Working memory (WM) is a crucial system that provides temporary storage and manipulation of information [4] and is also considered to be the foundation of complex cognitive functions, such as learning,

http://dx.doi.org/10.1016/j.neulet.2015.06.031 0304-3940/© 2015 Published by Elsevier Ireland Ltd.

Please cite this article in press as: W. Huang, et al., Alterations in the functional connectivity of a verbal working memory-related brain network in patients with left temporal lobe epilepsy, Neurosci. Lett. (2015), http://dx.doi.org/10.1016/j.neulet.2015.06.031

32

33 34 35 36 37 38 39 40 41 42

G Model NSL 31392 1–6

W. Huang et al. / Neuroscience Letters xxx (2015) xxx–xxx

2 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87

language comprehension and reasoning [4]. Based on different storage materials, WM is divided into verbal, spatial and object memories [5]. Verbal WM (VWM) depends on the phonological loop and the central executive system to process verbal information, including words, letters and other materials that are primarily linguistically coded [5]. This type of information processing plays a crucial role in daily life. Recently, blood oxygen level-dependent functional magnetic resonance imaging (BOLD fMRI) has been used to observe the neural activity of the entire brain cortex and changes in this activity directly and noninvasively due to this method’s high spatial and temporal resolutions and good repeatability. Resting-state functional connectivity (RSFC) is a type of fMRI analysis that is conducted to observe brain networks and is based on temporal correlations between BOLD intensities in remote brain regions. This measure can determine spontaneous low-frequency neural activity in a brain network. Since Biswal et al. [6] demonstrated the reliability of functional connectivity (FC) by examining the motor system at rest in 1995, numerous similar studies have been performed [7,8] Studies have reported the FCs of TLE at rest [8] and in a WM task [9,10]. Moreover, researchers have studied the changes in the brain networks with the RSFC and a WM task in healthy controls [11]. To date, the alteration in FC of the VWM-related network in TLE at resting state is poorly understood. Numerous studies have demonstrated that impairments of WM exist in patients with TLE. Patients with unilateral TLE are impaired on both verbal and visuospatial WM tasks irrespective of the affected hemisphere [12]. Left TLE (lTLE) patients commit more intrusion errors than right TLE patients on verbal tasks, whereas right TLE patients exhibit an additional impairment for visuospatial WM when compared with lTLE patients [12]. Regarding cerebral functional lateralization, the brain network for VWM exhibit a left hemisphere dominance, whereas the network for spatial WM exhibits a lateralization to the right hemisphere[13]. These findings demonstrate that lTLE patients exhibit strong effects on VWM. However, the association between VWM performance and alteration in FC of the VWM-related network has not been evaluated in lTLE patients. In this study, we investigated the VWM-related network in groups of lTLE patients and healthy controls by applying RSFC with a seed region of interest (ROI), which was defined during a VWM task according to task-related fMRI. We tested the hypothesis that lTLE leads to an alteration in the FC of the VWM-related network during resting state, and that this alteration is associated with VWM performance.

88

2. Materials and methods

89

2.1. Subjects

90 91 92 93 94 95 96 97 98 99 100 101 102 103 104

ARTICLE IN PRESS

Sixteen healthy volunteers (right-handed, nine males and seven females, 26.94 ± 1.64 years old, averaged educational time 14.37 ± 1.75 years) were investigated to examine the VWM network as identified by activation during a verbal n-back task. The subjects of RSFC included 14 patients with lTLE (right-handed, six males and eight females, 25.21 ± 8.03 years old, averaged educational time 12.57 ± 2.79 years) and 14 healthy control subjects (right-handed, seven males and seven females, 27.71 ± 2.58 years old, averaged educational time 14.21 ± 1.93 years). These patients were recruited from the Outpatient Neurologic Clinic of the First Affiliated Hospital of Guangxi Medical University (China). The inclusion criteria for the patients were based on the epilepsy classification of the International League Against Epilepsy as follows: (1) a clinical onset of symptoms that indicated that the location of the epileptogenic zone was in the left temporal lobe; (2) MRI

manifestation of left temporal lobe lesions or left HS; and (3) interictal or ictal electroencephalogram suggesting epileptic discharges in the left temporal lobe. The exclusion criteria were as follows: (1) age <16 years or age >60 years; (2) patients who take drugs that could impair cognition, such as cannabis users and others; and (3) history of mental illness or systemic disease. The subjects were informed in detail about the study and provided written consent to participate. This study was approved by the medical ethics committee of the First Affiliated Hospital of Guangxi Medical University. 2.2. Experimental paradigm Sixteen healthy volunteers performed an n-back task and simultaneously underwent neuroimaging scanning. This task was practiced twice prior to the fMRI scanning to ensure that the subjects completely understood the task. Typically, researchers use words or letters as stimuli in VWM tasks. However, in our study, single-digit numbers (1–9) were used. The processing of digital information can be divided into two types; i.e., verbal and nonverbal. A portion of digital processing belongs to the verbal type, including reading, listening and writing. The task with single-digit numbers as stimuli was designed in a previous study [14]. In this study, an n-back task was used to investigate the VWM network. The task was designed with E-prime 2.0 (Psychology Software Tools, Inc., n.d; Schneider, Eschmann, & Zuccolutto, 2002). The stimuli were presented at the centre of the screen as target cues using the SAMRTEC SA-9800 system. The task included two conditions; i.e., a 0-back and a 2-back condition. Before each condition began, a 2000-ms instruction indicating the condition to the participant was provided. Twenty random case stimuli were presented in each block, which included five targets. Each stimulus lasted for 1000 ms with an interval of 1000 ms. In the 0-back condition, the participants were required to identify the target “5”. In the 2-back condition, the subjects were required to confirm whether each presented number was the same as that presented two stimuli previously. The 2-back condition was alternated with the 0-back control condition three times with a blank delay interval of 18 s. The subjects were instructed to identify the target by pressing the response button with their right index finger as quickly as possible. The n-back experiment lasted for 6 mins. 2.3. Neuropsychological test of VWM To evaluate VWM ability, 14 lTLE patients and 14 healthy controls were required to perform a neuropsychological test of VWM that was identical to that used in the experimental paradigm. The mean reaction times of the accurate responses for the 0-back and the 2-back (ACCmeanRT (0-back) and ACCmeanRT (2-back)) were calculated (Table 1), and a two-way ANOVA was conducted (P < 0.05) with the factors of subjects (controls vs. lTLE; factor A) and conditions (0-back vs.2-back; factor B). 2.4. Data acquisition The fMRI scanning was performed using an Achieva 3T MRI scanner (Philips, Netherlands). We initially scanned 16 healthy volunteers while they performed the n-back task. The SAMRTEC SA-9800 system (Sinorad, China) was used to generate the task stimuli. The subjects’ heads were stabilized with foam pads to minimize head movement. The following parameters were used for axial T1 anatomical imaging: spin-echo sequence (T1 weighted); repetition time = 60 ms; echo time = 16 ms; slice thickness = 5 mm; gap = 1 mm; and field of view = 220 mm × 220 mm. The following parameters were used for axial functional imaging: gradient echo–echo planar imaging sequence (EPI sequence; T2); repetition time = 2000 ms; echo time = 30 ms; slice thickness = 5 mm;

Please cite this article in press as: W. Huang, et al., Alterations in the functional connectivity of a verbal working memory-related brain network in patients with left temporal lobe epilepsy, Neurosci. Lett. (2015), http://dx.doi.org/10.1016/j.neulet.2015.06.031

105 106 107 108 109 110 111 112 113

114

115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142

143

144 145 146 147 148 149 150 151

152

153 154 155 156 157 158 159 160 161 162 163 164

G Model

ARTICLE IN PRESS

NSL 31392 1–6

W. Huang et al. / Neuroscience Letters xxx (2015) xxx–xxx

3

Table 1

Q4 The detailed information about WM performance in both groups. Control group

ACCmeanRTa (0-back) (ms)

ACCmeanRT (2-back)(ms)

lTLE group

ACCmeanRT (0-back) (ms)

ACCmeanRT (2-back)(ms)

01 02 03 04 05 06 07 08 09 10 11 12 13 14

368.20 369.13 416.40 392.73 343.87 324.20 358.60 380.00 381.47 356.40 369.87 361.20 344.47 430.53

428.15 588.60 593.47 451.33 345.07 379.57 344.73 520.75 401.47 436.07 465.86 393.67 398.87 670.21

01 02 03 04 05 06 07 08 09 10 11 12 13 14

433.80 421.30 412.40 432.60 354.33 378.80 316.73 401.50 367.50 401.20 392.33 582.53 379.93 384.33

523.09 530.53 590.60 525.70 405.20 595.70 414.80 512.50 425.60 568.17 525.60 674.67 539.16 556.69

a

Mean reaction times of the accurate responses.

170

gap = 1 mm; field of view = 220 mm × 220 mm; flip angle = 90◦ ; and 31 slices. The task-related fMRI scanning lasted for 6 min, and 180 volumes were obtained. Fourteen patients and 14 controls were instructed to relax their minds with eyes closed and to remain motionless during the rs-fMRI scan. The rs-fMRI parameters were identical to those used in the task-related fMRI.

171

2.5. Data preprocessing

165 166 167 168 169

172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200

201 202 203 204 205 206 207 208 209

2.5.1. N-back fMRI data analysis The neuroimaging data from the task-related fMRI were analyzed with statistical parametric mapping (SPM8; www.fil.ion.ucl. ac.uk). The first five images were discarded to ensure the equilibration of the magnetization. The images were preprocessed separately for each subject. This preprocessing involved slice timing correction, realignment (criteria for data inclusion: maximum head movement <1 mm and maximum rotation <1◦ ) and spatial normalization to an EPI template in Montreal Neurological Institute (MNI) space in SPM8. The data were spatially smoothed with an 8-mm full-width at half-maximum Gaussian smoothing kernel (FWHM). After preprocessing, the images were subject to the first-level analysis according to the following parameters: interscan interval = 2; microtime resolution = 31; microtime onset = 16; serial correlations = AR1; condition 1 = 0-back (onsets: 14, 71, 129; durations: 23); condition 2 = 2-back (onsets: 42, 99, 157; durations: 24); and high-pass filter = 128 Hz. Multiple regressors, including the six head motion parameters, the global mean signal, the white matter signal and the cerebrospinal fluid signal, were used to regress out nuisance covariates. Contrasts were defined to identify the task-positive areas by comparing the 2-back against and 0-back tasks. Group analysis was performed using one-sample t-test in the second-level analysis with a random-effects model. The contrast images from the first-level analysis were selected. The Group statistics was estimated after a file named “SPM.mat” was generated. The results were illustrated with REST V1.8 (http://restingfmri.sourceforge.net, by Song Xiaowei et al; P < 0.05, false discovery rate (FDR) corrected). We selected the voxel with the highest Z-score as the seed ROI for the RSFC analysis. 2.5.2. RSFC analysis The resting state data were processed using DPARSF V2 (http:// resting-fmri.sourceforge.net). The data preprocessing included five discarded images, slice timing correction, head movement correction, spatial normalization, smoothing (FWHM = 4 mm), detrending and temporal bandpass filtering (0.01 Hz to 0.08 Hz). The software automatically regressed out the nuisance covariates. FC was measured based on the seed ROI, which was defined as a 6-mm sphere around the peak voxel. The average time courses were extracted

from the ROI, and the time courses were calculated using Pearson correlation coefficients for each voxel across the whole brain to produce RSFC maps. A fisher’s r-to-z transformation was applied to improve the normality of the correlation coefficients. RSFC z maps were obtained for each subject. The data were statistically analyzed using REST V1.8. Individual RSFC z maps entered into onesample t-test (P < 0.001, AlphaSim corrected, cluster size > 6 voxels). For group analysis, a two-sample t-test was adopted to compare the differences in the VWM-related network (P < 0.05, AlphaSim corrected, cluster size > 85 voxels).

2.6. Correlation analysis To investigate the relationships between the VWM-related brain areas and the VWM function (ACCmeanRT (2-back)), the voxels that exhibited significant differences in FC between the groups were extracted as masks for ROIs. The mean z-values in the ROIs were calculated, and the correlations with the ACCmeanRT (2-back) were determined by Pearson correlation analysis (P < 0.05).

210 211 212 213 214 215 216 217 218 219

220

221 222 223 224 225 226

3. Results

227

3.1. Results of the neuropsychological test

228

The statistical analyses revealed that Ftotal = 13.25, Ptotal = 0.00; FA = 7.41, PA = 0.01; FB = 31.42, PB = 0.00 and FA*B = 0.93, PA*B = 0.34. As expected, significant differences were observed between the lTLE patients and controls in both ACCmeanRT (0-back) (lTLE vs. controls: 404.23 ± 60.20 ms vs. 371.22 ± 28.25 ms) and ACCmeanRT (2-back) (lTLE vs. controls: 527.72 ± 74.02 ms vs. 458.42 ± 99.39 ms).

3.2. Brain activity In the 16 healthy volunteers, the VWM task induced activations in the bilateral middle frontal gyrus (MFG), inferior parietal lobule (IPL), superior parietal lobule (SPL), left inferior frontal gyrus (lIFG), precentral gyrus (PCG), precuneus (PC), cuneus, right cerebellar posterior lobe (rCPL) and supplementary motor area (SMA). The VWM-correlated activated voxels of the left hemisphere were more significant than those of the right hemisphere in terms of spatial distribution and intensity (Fig. 1A), particularly in the lIFG (peak MNI coordinates: x = −51, y = 13, z = 30, Z-score = 5.49, voxels = 556). We selected this region as the seed ROI for the RSFC analysis.

Please cite this article in press as: W. Huang, et al., Alterations in the functional connectivity of a verbal working memory-related brain network in patients with left temporal lobe epilepsy, Neurosci. Lett. (2015), http://dx.doi.org/10.1016/j.neulet.2015.06.031

229 230 231 232 233 234 235

236

237 238 239 240 241 242 243 244 245 246

G Model NSL 31392 1–6

ARTICLE IN PRESS W. Huang et al. / Neuroscience Letters xxx (2015) xxx–xxx

4

Fig. 1. Activation and FC maps. (A) Positive brain activation during a VWM task in a group of 16 healthy volunteers (P < 0.05, FDR corrected). (B,C) Brain areas that were positively correlated with the lIFG in terms of RSFC in the control (B) (P < 0.001, AlphaSim corrected) and lTLE groups (C). (D) Group differences in RSFC (P < 0.05, AlphaSim corrected). The Zs in the lower portion of the figure indicate the Z axis in MNI coordinates. The color bar on the right indicates the Z-scores; higher Z-scores indicate more significant effects.

247

248 249 250 251 252 253 254 255 256 257

258

259 260 261 262 263 264 265 266 267

268

269 270 271 272 273

3.3. RSFC results A one-sample t-test of the RSFC analysis revealed that the regions of the related network comprised a left-lateralized brain network that was similar to the task-related network in both groups. Additionally, the superior frontal gyrus (SFG), middle temporal gyrus (MTG) and inferior temporal gyrus (ITG) were involved in the RSFC (Table 2, Fig. 1B, C). Two-sample t-test revealed that FCs in the bilateral IFG, MFG and IPG in the lTLE group were significantly decreased compared to those of the control group (Table 2, Fig. D). No significant increases in FC were found. 3.4. Results of correlation analysis In both groups, we didn’t observe significant correlations between the ACCmeanRT (2-back) and the mean z-values within the regions that exhibited significant differences in FC at the individual level (lTLE group: left MFG (lMFG): R = −0.04, P = 0.90; lIFG: R = −0.24, P = 0.40; left IPL (lIPL): R = 0.36, P = 0.21; right MFG (rMFG): R = −0.02, P = 0.94; right IFG (rIFG): R = −0.21, P = 0.48; right IPL (rIPL): R = 0.04, P = 0.89; control group: lMFG: R = 0.22, P = 0.45; lIFG: R = −0.19, P = 0.53; lIPL: R = 0.54, P = 0.05; rMFG: R = 0.06, P = 0.83; rIFG: R = 0.13, P = 0.66; rIPL: R = 0.04, P = 0.88). 4. Discussion A previous PET study of normal subjects revealed that VWM brain network is left hemisphere-dominant, particularly in the left frontal lobe (including Broca’s area, the premotor cortex area and the SMA) and the left parietal lobe [13]. The results from the 16 healthy volunteers in this study also corroborate the notion that

the VWM network is lateralized to the left hemisphere. The lIFG exhibits the most significant activation followed by the left parietal lobe. Previous studies have shown that the left posterior parietal cortex mediates verbal information storage [15] and that the left frontal areas are the major regions of the rehearsal of verbal information [13]. In contrast, several studies have indicated that verbal information storage might be subserved by a complex prefrontoparietal network that is not solely localized in the parietal area [16]. Ventrolateral prefrontal cortex (PFC) activation decreases and dorsolateral PFC activation increases with increasing memory load[17]. These results suggest that the ventrolateral PFC might mediate WM storage and that the dorsolateral PFC might mediate WM strategy and facilitate the storage of WM information[17]. Moreover, the activations of the left PCG, PC, cuneus, rCPL and SMA fully indicated that VWM is comprised of a complex of advanced cognitive functions, and requires the involvement and coordination of multiple brain regions. We investigated the VWM-related network in groups of lTLE patients and healthy controls by applying RSFC with lIFG, which exhibited the most significant activation. The same method was used by Waites et al.[7]. Our study found that the RSFC network was similar to the task-related network, but this result was not entirely consistent [18]. The temporal lobe was involved in the RSFC network, but not in the task-related network. Smith et al. [19] identified fairly similar areas when comparing the visual, default mode, sensorimotor and executive control networks during activation and rest. However, some scholars believe that the RSFC method is inferior to task-related fMRI [18]. In this study, the VWM-related brain network was impaired in the lTLE patients as determined by RSFC. This result is consistent with those of previous studies in which the FCs related to WM in patients with TLE were examined [9,10]. For example, Stretton et al.

Please cite this article in press as: W. Huang, et al., Alterations in the functional connectivity of a verbal working memory-related brain network in patients with left temporal lobe epilepsy, Neurosci. Lett. (2015), http://dx.doi.org/10.1016/j.neulet.2015.06.031

274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305

G Model

ARTICLE IN PRESS

NSL 31392 1–6

W. Huang et al. / Neuroscience Letters xxx (2015) xxx–xxx

5

Table 2 Brain areas positively correlated with the lIFG and group differences in RSFC. Group

1556 544 916 292 183 29 60 60

lTLE groupd

lIFG/lMFG/lPCG rIFG/rMFG lIPL/lSPL/lPC/cuneus rIPL lITG rITG SMA rCPL

−51 48 −42 33 −66 57 9 21

12 33 −45 −45 −51 −45 24 −81

30 18 48 36 −15 −21 48 −42

7.48 5.14 5.33 3.67 4.31 3.71 4.89 4.68

1169 377 602 15 56 13 141 81

Control > lTLEe

lMFG rMFG lIFG rIFG lIPL rIPL

− 27 48 −36 54 − 39 48

3 45 30 9 −54 −48

45 15 18 21 33 60

3.76 2.99 3.49 3.89 3.73 3.63

162 93 109 86 101 98

h i l m n o p q r s t u

312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330

Voxels

7.21 5.66 5.61 4.72 4.84 3.94 4.35 4.36

f

311

Z-score

30 21 36 42 −21 −15 51 −51

g

310

z

15 12 −48 −54 −51 −60 12 −75

e

309

y

− 51 48 −39 33 −57 63 6 30

d

308

x lIFGf /lMFGg /lSFGh /lPCGi rIFGj /rMFGk lIPLl /lSPLm /lPCn /cuneus rIPLo /rSPLp lITGq /lMTGr rITGs SMAt rCPLu

c

307

MNIb Coordinates

Control groupc

b

306

Brain region

Montreal Neurological Institute; P < 0.001, AlphaSim corrected; P < 0.001, AlphaSim corrected; P < 0.05, AlphaSim corrected; Left inferior frontal gyrus; Left middle frontal gyrus; Left superior frontal gyrus; Left precentral gyrus; jright inferior frontal gyrus; kright middle frontal gyrus; Left inferior parietal lobule; Left superior parietal lobule; Left precuneus; Right inferior parietal lobule; Right superior parietal lobule; Left inferior temporal gyrus; Left middle temporal gyrus; Right inferior temporal gyrus; Supplementary motor area; Right cerebellar posterior lobe

[10] compared the FCs of the visuospatial WM network extracted from fMRI during WM task. The findings of their authors indicated no significant decreases in the FCs of either HS groups compared to the controls, but an increase in FC was observed [10]. These authors [10] further suggested that chronic uncontrolled seizures might possibly induce white matter degradation and disrupt the FCs of regions both local to and remote from the focus. This previous study also provided further evidence that the integrities of grey matter and white matter are impaired in the areas that are functionally relevant to WM in HS [20]. Moreover, researchers have also revealed that HS can cripple backward connections from the left medial temporal lobe to the left ITG but reinforce bidirectional connections between the IFG and the medial temporal lobe in the contralesional hemisphere [9]. The result might not be a compensation for damage, but it reflected an engagement of the remaining components of a damaged network subserving VWM[9]. Our results provide additional evidence supporting the perspective that the studies above have confirmed that epileptiform discharges can damage the FCs of brain regions. The FCs of the bilateral frontoparietal networks reduced in patients with lTLE in the present study. Waites et al. [7] argued that the changes in the FCs in language areas of patients with lTLE at rest might be related to four possible mechanisms. These mechanisms include the influence of epileptic discharges, altered regional inhibition or excitation, the influence of antiepileptic medication and

structural modifications of the language network [7]. TLE seizures can spread to other regions via the fasciculus. Anatomically, the uncinate fasciculus connects the anterior temporal lobe with the inferior frontal areas, which are damaged in patients with lTLE [21]. The reduced FC of the IFG might be related to damage to the uncinate fasciculus. The superior longitudinal fasciculus and the cingulum also connect to the frontoparietal network. We also presume that these two fasciculi might contribute to the decrease in FC. VWM depends on the phonological loop and the central executive system to process verbal information[5]. Attentional control is part of the central executive system [4] and is responsible for the coordination of the implementation of attentional resources and strategies to ensure the efficacy of the phonological loop. Studies have demonstrated that the rIFG plays important roles in inhibition and attentional control [22]. As a consequence, decreased FC in the rIFG may be associated with impairments of attention [23] and executive function [24] in TLE patients. The PFC is known to mediate executive control[25]. Nevertheless, other studies have shown that the SPL is required for the executive rearrangement of information in WM[26]. In the current study, the FC of the SPL did not decrease, but the FCs of MFG, IFG and IPL were decreased in the lTLE patients. These may suggested that the attentional control of lTLE patients was damaged. As is known to all, the prefrontoparietal network is involved in processing the verbal information[16]. So the reduced

Please cite this article in press as: W. Huang, et al., Alterations in the functional connectivity of a verbal working memory-related brain network in patients with left temporal lobe epilepsy, Neurosci. Lett. (2015), http://dx.doi.org/10.1016/j.neulet.2015.06.031

331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355

G Model NSL 31392 1–6

ARTICLE IN PRESS W. Huang et al. / Neuroscience Letters xxx (2015) xxx–xxx

6

377

FCs of the left hemisphere area in the lTLE patients, indicating that the phonological loop in patients of lTLE was impaired. As expected, significant differences were observed between the lTLE patients and the controls in both ACCmeanRT (0-back) and ACCmeanRT (2-back). These findings suggested that lTLE patients were not only compromised in terms of the working memory system, but also in terms of phonological short-term memory storage. But, we failed to find any significant correlations between the ACCmeanRT (2-back) and the mean z-values within the regions that exhibited significant differences in FC at the individual level in both groups respectively. As we know, VWM is a complex cognitive function that requires the involvement and coordination of multiple brain regions. In this study, only six areas were involved in the correlation but the remaining components were excluded, such as rCPL, SMA and the others. The result may imply that the function of a brain region cannot change the performance of VWM. In addition, lTLE patients’ neuropsychological profile was globally lower than the one of healthy controls. This may explain why we failed to find any significant correlation between the performance and the resting state data. The reduction that we found in the VWMrelated network may just be a consequence of a more widespread reduction of FC in ITLE patients.

378

5. Conclusion

356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376

384

In conclusion, this study demonstrated the lateralization of the VWM network to the left hemisphere. Additionally, alterations in the FC of the VWM-related network in patients with lTLE during resting state suggested that epileptiform discharges can damage the brain regions, both local focus and remote areas and that these alterations were not associated with VWM performance.

385

6. Disclosure

379 380 381 382 383

386

387

None of the authors have anything to disclose. Acknowledgment

390

The National Natural Science Foundation of China (Project number: 81360202) funded this study. We thank Fang Lu, No.1 People’s Hospital of Nanning, Nanning, China, for her assistance.

391

References

Q3 388 389

392 393 394 395 396 397 398 399

[1] H.-G. Wieser, ILAE commission report. Mesial temporal lobe epilepsy with hippocampal sclerosis, Epilepsia 45 (2004) 695–714. [2] B. Hermann, M. Seidenberg, E.J. Lee, F. Chan, P. Rutecki, Cognitive phenotypes in temporal lobe epilepsy, J. Int. Neuropsychol. Soc. 13 (2007) 12–20. [3] J.O. Zheng, B.L. Qin, C. Dang, W. Ye, Z.R. Chen, L. Yu, Alertness network in patients with temporal lobe epilepsy: A fMRI study, Epilepsy Res. 100 (2012) 67–73. [4] A. Baddeley, Working memory: looking back and looking forward, Nat. Rev. Neurosci. 4 (2003) 829–839.

[5] T.D. Wager, E.E. Smith, Neuroimaging studies of working memory, cognitive, affective, & behavioral, Neuroscience 3 (2003) 255–274. [6] B. Biswal, F. Zerrin Yetkin, V.M. Haughton, J.S. Hyde, Functional connectivity in the motor cortex of resting human brain using echo-planar mri, Magn. Reson. Med. 34 (1995) 537–541. [7] A.B. Waites, R.S. Briellmann, M.M. Saling, D.F. Abbott, G.D. Jackson, Functional connectivity networks are disrupted in left temporal lobe epilepsy, Ann Neurol. 59 (2006) 335–343. [8] Z. Haneef, A. Lenartowicz, H.J. Yeh, H.S. Levin, J. Engel, J.M. Stern, Functional connectivity of hippocampal networks in temporal lobe epilepsy, Epilepsia 55 (2014) 137–145. [9] P. Campo, M.I. Garrido, R.J. Moran, F. Maestú, I. García-Morales, A. Gil-Nagel, F. Del Pozo, R.J. Dolan, K.J. Friston, Remote effects of hippocampal sclerosis on effective connectivity during working memory encoding: a case of connectional diaschisis? Cereb. Cortex 22 (2012) 1225–1236. [10] J. Stretton, G. Winston, M. Sidhu, S. Bonelli, M. Centeno, C. Vollmar, R. Cleary, E. Williams, M. Koepp, P. Thompson, Disrupted segregation of working memory networks in temporal lobe epilepsy, NeuroImage: Clin. (2013). [11] E.M. Gordon, A.L. Breeden, S.E. Bean, C.J. Vaidya, Working memory-related changes in functional connectivity persist beyond task disengagement, Hum. Brain Mapp. (2012). [12] D.D. Wagner, V. Sziklas, K.E. Garver, M. Jones-Gotman, Material-specific lateralization of working memory in the medial temporal lobe, Neuropsychologia 47 (2009) 112–122. [13] E.E. Smith, J. Jonides, R.A. Koeppe, Dissociating verbal and spatial working memory using PET, Cereb. Cortex 6 (1996) 11–20. [14] S. Crottaz-Herbette, R. Anagnoson, V. Menon, Modality effects in verbal working memory: differential prefrontal and parietal responses to auditory and visual stimuli, Neuroimage 21 (2004) 340–351. [15] A.M. Owen, K.M. Mcmillan, A.R. LairdE, Bullmore, N-back working memory paradigm: a meta-analysis of normative functional neuroimaging studies, Hum. Brain Mapp. 25 (2005) 46–59. [16] O. Gruber, Effects of domain-specific interference on brain activation associated with verbal working memory task performance, Cereb. Cortex 11 (2001) 1047–1055. [17] B. Rypma, J.S. Berger, M. D’esposito, The influence of working-memory demand and subject performance on prefrontal cortical activity, J. Cogn. Neurosci. 14 (2002) 721–731. [18] R.L. Buckner, F.M. Krienen, B.T. Yeo, Opportunities and limitations of intrinsic functional connectivity MRI, Nat. Neurosci. 16 (2013) 832–837. [19] S.M. Smith, P.T. Fox, K.L. Miller, D.C. Glahn, P.M. Fox, C.E. Mackay, N. Filippini, K.E. Watkins, R. Toro, A.R. Laird, Correspondence of the brain’s functional architecture during activation and rest, Proceedings of the National Academy of Sciences 106 (2009) 13040–13045. [20] G.P. Winston, J. Stretton, M.K. Sidhu, M.R. Symms, P.J. Thompson, J.S. Duncan, Structural correlates of impaired working memory in hippocampal sclerosis, Epilepsia (2013). [21] C.H. Kim, C.K. Chung, B.-B. Koo, J.-M. Lee, J.S. Kim, S.K. Lee, Changes in language pathways in patients with temporal lobe epilepsy: diffusion tensor imaging analysis of the uncinate and arcuate fasciculi, World Neurosurgery 75 (2011) 509–516. [22] A. Hampshire, S.R. Chamberlain, M.M. Monti, J. Duncan, A.M. Owen, The role of the right inferior frontal gyrus: inhibition and attentional control, Neuroimage 50 (2010) 1313–1319. [23] Z. Zhang, G. Lu, Y. Zhong, Q. Tan, Z. Yang, W. Liao, Z. Chen, J. Shi, Y. Liu, Impaired attention network in temporal lobe epilepsy: a resting FMRI study, Neurosci Lett. 458 (2009) 97–101. [24] J. Stretton, P. Thompson, Frontal lobe function in temporal lobe epilepsy, Epilepsy Res. 98 (2012) 1–13. [25] S. Funahashi, Neuronal mechanisms of executive control by the prefrontal cortex, Neurosci. Res. 39 (2001) 147–165. [26] M. Koenigs, A.K. Barbey, B.R. Postle, J. Grafman, Superior parietal cortex is critical for the manipulation of information in working memory, J. Neurosci. 29 (2009) 14980–14986.

Please cite this article in press as: W. Huang, et al., Alterations in the functional connectivity of a verbal working memory-related brain network in patients with left temporal lobe epilepsy, Neurosci. Lett. (2015), http://dx.doi.org/10.1016/j.neulet.2015.06.031

400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463