Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 Jan 1.
Published in final edited form as: Brain Lang. 2007 Apr 6;104(1):42–50. doi: 10.1016/j.bandl.2007.02.007

Functional connectivity in an fMRI study of semantic and phonological processes and the effect of L-DOPA

Madalina E Tivarus 1, Ashleigh Hillier 2, Petra Schmalbrock 3, David Q Beversdorf 4
PMCID: PMC2430741  NIHMSID: NIHMS37266  PMID: 17418888

Abstract

We describe an fMRI experiment examining the functional connectivity (FC) between regions of the brain associated with semantic and phonological processing. We wished to explore whether L-Dopa administration affects the interaction between language network components in semantic and phonological categorization tasks, as revealed by FC. We hypothesized that L-Dopa would decrease FC due to restriction of the semantic network. During two test sessions (placebo and L-Dopa) each participant performed two fMRI runs, involving phonological and semantic processing. A number of brain regions commonly activated by the two tasks were chosen: left inferior frontal, left posterior temporal and left fusiform gyri, and left parietal cortex. FC was calculated and further analyzed for effects of either the drug or task. No drug effect was found. A significant main effect for task was found, with a greater average correlation for the phonological task than for the semantic task. These findings suggest that language areas are activated in a more synchronous manner for phonological tasks than for semantic tasks. This may relate to the fact that phonological tasks are mediated to a greater extent within language areas, whereas semantic tasks likely require greater interaction outside of the language areas. Alternatively, this may be due to differences in the attentional requirements of the two tasks.

Keywords: functional MRI, functional connectivity, semantic, phonological, dopamine, L-DOPA

INTRODUCTION

Understanding the effects of pharmacological manipulation on semantic networks may have implications ranging from language perception (Angwin et al., 2004; Kischka et al., 1996) to problem solving (Beversdorf et al., 1999; Beversdorf et al., 2002; Silver et al., 2004). This may also reveal insight into the impairments exhibited by some patient populations (Arnott et al., 2001). In order to understand these complex behaviors, one must first begin to understand how these manipulations affect brain regions associated with language processing.

Studies of patients with brain lesions have shown that the representations of word meaning and form can be differentiated (for a review see Rapp & Caramazza, 1996). However functional deficits are often associated with large lesions and make it difficult to relate specific deficits to specific anatomical regions. Evidence form neuroimaging studies complements the findings from brain damaged patients and makes it possible to identify specific brain areas associated with a specific task (Cabeza & Nyberg, 2000; Crosson et al., 2003; Devlin et al., 2003; Fiez & Petersen, 1998; Martin, 2003; McDermott et al., 2003; Poldrack et al., 1999; Price, 2000; Thompson-Schill et al., 1997). The findings of these studies revealed the existence of a vast network of brain structures that contribute to these language processes, including anterior and posterior, cortical and subcortical brain regions. In examining different aspects of language processing, such as phonological and semantic processing, no task is ‘process-pure’. All phonological tasks can involve some semantic processing and all semantic tasks can include some phonological processing. However, whereas brain activation associated with these two language processes overlaps to a marked degree, some differences in brain activation have been revealed between tasks directed at semantic and phonological processing. The left inferior prefrontal cortex (LIPC) participates in a network that activates during controlled processing of both semantic and non-semantic information (Bookheimer, 2002; Devlin et al., 2003; Gabrieli et al., 1998; Poldrack et al., 1999; Price, 2000; Thompson-Schill et al., 1997; Wagner et al., 2001), whereas posterior regions appear to have a relatively greater involvement in retrieval of stored information such as word meaning (posterior middle temporal gyrus-BA21) (Demb et al., 1995; Gold & Buckner, 2002; McDermott et al., 2003; Poldrak et al., 1999; Schivde & Thompson-Schill, 2004) or sound (parietal cortex-BA7/40)(McDermott et al., 2003; Schivde & Thompson-Schill, 2004). These studies go further and reveal some regional specificity within the prefrontal cortex: anterior LIPC (BA 45/47/10) as involved more in semantic processing (Gold & Buckner, 2002; McDermott et al., 2003; Demb et al., 1995; Poldrak et al., 1999; Schivde & Thompson-Schill, 2004) and posterior LIPC (BA 44/6) in phonological processing (Gold & Buckner, 2002; Poldrack et al., 1999; McDermott et al., 2003; Schivde & Thompson-Schill, 2004). Mc Dermott et al. (2003) have also found that areas of the middle frontal gyrus (BA6, pre-SMA) bilateral fusiform gyrus (BA37) and bilateral cerebellum are similarly activated by both phonological and semantic tasks relative to baseline (fixating on a cross-hair). Subcortical structures have also been found to contribute to both semantic and phonological processing. Crosson et al. (2003) showed that the left dorsal caudate and the ventral anterior thalamus contribute to retrieval of words from pre-existing lexical stores, due to direct connections with the pre-SMA. They have also suggested that right basal ganglia activity contributes by suppressing right frontal activity from interfering with language production.

Although, the areas of the brain involved in semantic and phonological processing have been identified and further specifications of functional distinctions within these areas have been made, the interaction between these areas during language processing has not been extensively reported to our knowledge. Therefore, we intended to contribute by adding information about the functional connectivity in the brain during language processes, specifically semantic and phonological processing.

Complex cognitive functions and behaviors are mediated by interconnected neural networks (Mesulam, 1990). The anatomical connections can be characterized by using diffusion weighted imaging. In contrast, the functional interaction between brain regions activated during performance of a task can be characterized by using fMRI to examine functional connectivity. Functional connectivity has been defined as the “temporal correlation between spatially remote neurophysiological events” (Friston, 1994a). In neuroimaging, functional connectivity offers indirect evidence of communication or collaboration between areas of the brain but offers “no insight into how these correlations are mediated” (Friston, 1994a). Functional connectivity is calculated by extracting fMRI time series from brain regions of interest. Then correlation coefficients (cc) are calculated between the time series of two regions of interest in order to determine to what degree they are functionally connected (Arfanakis et al., 2000). Such methodology has been used to examine functional connectivity within the motor, visual and auditory systems (Arfanakis et al., 2000), and during visual delayed recognition (Rissman et al., 2004). This methodology has also revealed decreased functional connectivity during sentence comprehension and working memory in high functioning autism (Just et al., 2004; Koshino et al., 2005), a clinical population characterized by impairments in utilization of context, resulting in atypical performance consistent with ‘underconnectivity’ on tasks involving the semantic and associative networks (Beversdorf et al., 2000). Therefore, we wished to examine functional connectivity for semantic and phonological processes.

Behavioral research has demonstrated a role of the catecholamine neurotransmitter systems in modulation of cognition, more specifically of language networks. L-Dopa is a converted into both dopamine and norepinephrine. These act to “amplify strong signals and dampen weak ones” (Dehaene et al., 1999), and therefore they increase the signal-to-noise ratio of cortical information processing. This is supported by evidence that dopamine enhances N-methyl-D-aspartate (NMDA) mediated excitatory postsynaptic currents (EPSCs) in layer V of the rat prefrontal cortex, while reducing non-NMDA mediated components of EPSCs, thus selectively enhancing sustained synaptic inputs (Seamans et al., 2001). These findings are proposed as a mechanism for the dopaminergic role in sustaining working memory (Winterer & Weinberger, 2004; Durstewitz & Seamans, 2002). This also suggests that increased levels of catecholamines may increase the focus of activation in semantic networks, and therefore it should decrease the spread of activation. Previous behavioral research demonstrated that administration of L-Dopa resulted in restriction of the semantic network in a lexical priming experiment (Angwin et al., 2004; Kischka et al., 1996). This result along with studies of semantic network spread in neurologically impaired patients such as schizophrenic patients and patients with Parkinson’s disease, suggests that dopamine may play a role in modulation of semantic processing (Spitzer et al., 1993, Spitzer et al., 1994; Watters & Patel, 1999). Therefore, we also wished to explore whether administration of L-Dopa also affects the interaction between language network components in semantic and phonological categorization tasks, as revealed by functional connectivity.

MATERIALS AND METHODS

Research participants

Sixteen participants, 8 male and 8 female, mean age 28.3 years (range 21–49 years), average educational level 15 years, were recruited to participate in these studies. The sample size in this within-subject design would be expected to yield a significant effect of L-DOPA since previous experiments on semantic priming yielded significant results with a similar number of between-subject comparisons (Kischka et al., 1996). They were all native English speakers, right handed, as assessed with the Edinburgh Handedness Inventory (Oldfield, 1971), with no history of psychiatric or neurological problems, or of learning disabilities such as dyslexia. Additional exclusion criteria related to L-Dopa administration were: previous or current use of drugs that stimulate or block dopamine, previous history of unprovoked hallucinations, and previous adverse reaction to drugs that stimulate dopamine. All participants reported normal or corrected-to-normal vision. In order to avoid the hemodynamic effects of these agents on BOLD signal, the participants were asked to abstain form caffeine and nicotine at least 2 hours prior to the study. They were also screened to comply with the MRI safety requirements (no metal implants or prostheses, no metal objects in their bodies, non-claustrophobic). Written consent was obtained from all participants in accordance with the regulations of the Institutional Review Board of The Ohio State University.

Drug administration

Due to the unavoidable inter-subject variability in brain activation detected with fMRI, the detection of drug effects benefits from a within subject (or repeated measures) design in which each participant is scanned following administration of placebo as well as following administration of one active compound (drug). We used for this study a combination of L-Dopa (100mg) and carbidopa (25mg). L-Dopa, a precursor for dopamine, crosses the blood-brain barrier and is administered to increase dopamine levels in the brain. L-Dopa is rapidly absorbed from the small intestine, and reaches peak plasma levels in 1–2 hours after oral dose, with a half-life between one and three hours. The dose of L-DOPA is derived from previous semantic priming experiments (Kischka et al., 1996). Because only 1–3% of administered L-Dopa reaches the brain, to achieve therapeutic levels either a high dose or administration along with dopa decarboxylase inhibitor, is necessary. Carbidopa is such an inhibitor typically administered with L-Dopa to minimize peripheral adverse reactions (e.g. nausea and vomiting). It does not cross the brain-blood barrier and does not prevent central conversion.

Testing started 90 minutes after the oral administration of the pill, either placebo or L-Dopa/carbidopa. Both the participants and the researchers were blind to the pill content. The order of drug administration was counterbalanced. To avoid confounds such as practice and task difficulty, the order of the language tests was also counterbalanced, and the test sessions for each participant were separated by at least a week. Heart rate and blood pressure were measured before administration of the drug, and then immediately before and after the MRI scans.

Materials

Four groups of word lists, two for the semantic processing and two for the phonological processing, were used as stimuli in this experiment. Each group was comprised of four word lists, for a total of 16 lists presented to each participant over the entire experiment. One semantic and one phonological group of word lists was presented at each imaging session. These were selected and modified from the sets used by McDermott et al (2003), which have demonstrated activation patterns on fMRI that can detect some differentiation between the semantic and phonological language processes. The median word length was 5 letters for both semantic and phonological lists, and the median Kucera-Francis written word frequency (Kucera and Francis, 1967) was 32 per million for the semantic lists and 30 per million for the phonological lists, with these characteristics matched between the four different lists. The four list groups were counterbalanced within and between sessions, with each version of the lists being presented equal times in each experimental condition, and no version being repeated for the same participant, allowing within subject comparisons for each condition.

During each test session (one session per drug condition), each participant performed two scanning runs. One of the runs consisted of alternating blocks of semantic task (24 sec) and rest (30 sec), and the other run consisted of alternating blocks of phonological task (24 sec) and rest (30 sec), for a total of 4min and 6 sec for each run, as shown in Fig. 1 and 2.

Fig 1.

Fig 1

Open in a new tab

Design of a semantic run and an example of a semantic task block (sugar, taste, candy are related by meaning to the cue word SWEET, boat is not).

Fig 2.

Fig 2

Open in a new tab

Design of a phonological run and an example of a semantic task block (brand, land, sand rhyme with the cue word STAND, clips does not).

Each word list constituted a task block (24 sec) and consisted of 15 words: for the semantic condition 10 of the words in the list were related by meaning and 5 were unrelated to a cue word, while for the phonological condition 10 of the words rhymed and 5 did not rhyme with the cue word. The design of a block was as follows: a cue word was presented in capital letters for 3 seconds followed by the list of 15 words. Each word in the list was presented for approximately 1100ms, with a 300ms inter-stimulus interval (a blank screen), for a total of for 1.4 seconds for each word, as presented in Fig. 1 and 2.

Participants were instructed to attend to the meaning or sound of the lists presented. They responded by pressing one of the two buttons on an fMRI-compatible response system (Lumina LP 400, Cedrus Corp., San Pedro, CA) as follows: for the semantic processing run, they were asked to respond by pressing the YES button if the word in the list was related by meaning to the cue, and by pressing the NO button if the word was not related to the cue; for the phonological task, participants were asked to respond by pressing YES if the word rhymed with the cue and by pressing NO if the word did not rhyme with the cue. Response times and accuracy were recorded and all stimuli were presented visually one at a time on the center of a screen using SuperLab experiment generator software (Version 2.0, Cedrus Corp., San Pedro, CA). Participants were able to see the screen through a mirror mounted on the head coil. A practice trial was performed inside the scanner, prior to data acquisition, until participants were comfortable with the task and the scanner conditions.

Imaging data acquisition

Images were collected with a 1.5 T General Electric (Milwaukee, WI) Signa scanner with a quadrature head coil. The scanner was also equipped with an fMRI acquisition software, BrainWaveRT (General Electric Milwaukee, WI), which allows for preparation, acquisition and visualization of functional images. Structural T1-weighted images were acquired for anatomic localization and co-registration, using the 3D FAST SPGR pulse sequence (256×128 matrix; 240mm FOV; 64 axial slices; 2.5mm slice thickness). The BOLD contrast functional data were collected using a gradient echo EPI pulse sequence (TR=3s; TE=40ms; α=90; FOV=240mm; matrix 64×64, 28 axial slices for whole brain coverage; 5mm slice thickness). The first four image volumes were acquired to allow stabilization of longitudinal magnetization, and were discarded before data analysis.

Imaging data analysis

Individual and group activation maps

The imaging data were analyzed using FSL (Smith et al, 2004) software. Pre-statistics processing was performed using FSL and consisted of motion correction (Jenkinson et al., 2002), non-brain signal removal (Smith, 2002), Gaussian spatial smoothing (FWHM 5mm), intensity normalization of all volumes by the same factor and high-pass temporal filtering.

In order to identify the areas of the brain activated by different tasks, individual and group analyses were performed with FSL. Statistical analysis was carried out by first using FILM (FMRIB’s Improved Linear Model) to create individual activation maps, and then FLAME (FMRIB’s Local Analysis of Mixed Effects) (Behrens et al, 2003) was used to generate average group maps. Z statistic images were obtained using cluster analysis with clusters determined by Z>3 and a cluster significance threshold of p=0.05 (Friston et al., 1994b; Froman et al., 1995; Worsley et al., 1992). Registration to the Montreal Neurological Institute (MNI) standard brain was also carried out (Jenkinson & Smith, 2001; Jenkinson et al., 2002).

Individual activation maps were obtained for each participant for each of the four scans. Due to excessive motion during the scans, 4 sets of data were discarded, resulting in a total of 60 data sets: semantic processing during placebo treatment (n=15), phonological processing during placebo treatment (n=15), semantic processing during L-Dopa treatment (n=15), and phonological processing during L-Dopa treatment (n=15). For these remaining data sets, average group maps were obtained for each of the four scans performed by participants and corrections for multiple comparisons (t-tests) were performed to determine differences in activation patterns during semantic versus phonological processing as well as to identify any effects the drug may have on these.

ROI design and average signal time series analysis

Functional connectivity was computed for each participant, for each task and drug condition. A number of regions the brain activated by the two tasks, were selected a priori for each participant, based on the findings of McDermott et al (2003), and included the left inferior frontal cortex (BA 44/45/46)- LIFG, left fusiform gyrus (BA37)-LFUS, left parietal cortex (BA7)-LPAR and left middle temporal gyrus (BA 21/22)-LMTG. Spherical ROIs (radius 10mm) with the center in the above mentioned anatomical regions were drawn on each participant’s high resolution images. These were registered (Jenkinson and Smith, 2001; Jenkinson et al., 2002) to the activation images in each task and drug condition so that they overlaid the active voxels in the areas of interest. Average time series for all voxels included in the ROI were then extracted for each subject using FEATQUERY, part of FSL.

Correlations of these time series between pairs of ROIs were computed by calculating the correlation coefficient between the time series plots for each ROI pair for each subject. Fisher’s Z-transformation was applied to the set of correlations and further analyzed for effects of drug, task and ROI in a 2*2*6 (drug*task*ROI pairs) repeated measures ANOVA. Furthermore, for each ROI pair, functional connectivity was compared separately, using a 2*2 ANOVA for L-Dopa vs. placebo and semantic vs. phonological task.

Behavioral data analysis

Mean response times (RT) and error rates were calculated for each participant for each of the four scans. All response errors were removed and response times (RT) below 200ms were considered outliers and excluded from further analysis. This resulted in the exclusion of less than 3% of the data.

A repeated measures 2*2 (task*drug) ANOVA was carried out to detect the effect of either the drug or the task on response times.

RESULTS

Behavioral results

Mean response times (RT) were calculated for each participant for each of the four testing conditions, and are presented in Table 1. A repeated measures 2*2 (task*drug) ANOVA was carried out to detect the effect of either drug or task on response times. A significant main effect was found for task [F(1,15)=64.5, p<0.00005] with a longer response time recorded for the semantic task (mean RT =668 msec, SD=11) as compared to the phonological task (mean RT= 602, SD=12). Neither a main effect for drug, nor interactions effects were found. Furthermore, no significant drug effects were observed on heart rate or blood pressure.

Table 1.

Mean response times (msec) for the two tasks in each treatment condition (SD-standard deviations).

Drug L-Dopa Placebo
Task semantic phonological semantic phonological
RT(SD) 668 (47) 604 (51) 669 (49) 600 (57)
Open in a new tab

Imaging results-activation maps

Group activation maps obtained using FSL’s cluster analysis reveal a pattern of activation similar to that obtained by McDermott et al (2003). The pattern of brain activation associated with the processing of semantic relationship and phonological processing in the two drug conditions is shown in Fig 3. Relative to the baseline (fixating the cross-hair) both semantic and phonological tasks activated a set of common regions of the brain comprised of the left inferior frontal cortex- LIFG (BA44/45) extending into the premotor and motor areas, bilateral middle frontal gyrus-MFG (BA46/9), left posterior middle temporal gyrus-LpMTG (BA 21/22), left fusiform gyrus-LFUS (BA37), bilateral occipital cortex (BA 17/18/19), bilateral premotor cortex.

Fig 3.

Fig 3

Open in a new tab

Average group activation maps during semantic and phonological processing under the two drug treatments.

Differences in brain activation for the two types of processes obtained by performing corrected multiple comparisons, were as follows: semantic processing preferentially activated (p< 0.001) the anterior LIFG (BA47) and posterior dorsal LIFG (BA44), bilateral MFG (BA46/9), left supramarginal gyrus (BA40) as well as LpMTG (BA21/22); phonological processing showed preferential activation (p< 0.001) in the posterior LIFC going into the premotor and motor areas (BA44/6) and the left parietal cortex-LPAR (BA7). These results agree with McDermott et al (2003). No significant differences were found in the average maps, between drug conditions for the two different tasks.

Functional connectivity results

In a repeated measures 2*2*6 (task*drug*ROIpair) ANOVA, a significant main effect for task was found [F(1,14)=6.597, p=0.022], with an average correlation over both drugs and all ROI pairs for the phonological task (mean cc=0.506; SE=0.06) greater than that for the semantic task (mean cc= 0.446; SE=0.05) (Fig. 4).

Fig 4.

Fig 4

Open in a new tab

Mean correlation coefficient for the semantic and phonological tasks (average over all ROI pairs and drug conditions).

There was no significant main effect for drug nor a significant drug*task interaction. There was also a main effect for the ROI pair [F(5,70)=7.319, p<0.0005]. Therefore, further 2*2 (task*drug) repeated measures ANOVAs were performed for each of the 6 ROI pairs. For the LIFG-LMTG, LIFG-LPAR, LMTG-LFUS pairs, the average correlation coefficients for both drugs were significantly greater for the phonological than for the semantic process. These results are presented in Table 2 and the significant findings in Fig. 5 (a-c). Due to the effects of L-DOPA previously reported on semantic priming, comparisons were also performed for each ROI pair for the effect of L-DOPA in order to detect whether the overall drug comparison might fail to detect effects on individual ROI pairs. Since priming involves recognition of words, our particular focus was on comparisons between visual perceptual brain regions and language perceptual brain regions. For the LFUS-LPAR pair a main drug effect was found, with a higher correlation coefficient across tasks for L-Dopa than for placebo. These results are also presented in Table 2.

Table 2.

Results of repeated measures 2*2 ANOVAs for each ROI pair (DA-L-Dopa; Pla-placebo)

ROI pair Main and interaction effects Mean cc (SE)
F P
1 2
LIFG&LMTG Task (1-sem; 2-phono) 0.384 (0.062) 0.487 (0.076) 5.036 0.042
Drug (1-DA; 2-Pla) 0.471 (0.068) 0.400 (0.077) 1.216 0.289
Task*Drug 1.266 0.280
LIFG&LPAR Task (1-sem; 2-phono) 0.522 (0.047) 0.591 (0.055) 5.636 0.032
Drug (1-DA; 2-Pla) 0.563 (0.057) 0.055 (0.060) 0.042 0.840
Task*Drug 0.167 0.689
LIFG&LFUS Task (1-sem; 2-phono) 0.491 (0.058) 0.518 (0.066) 0.505 0.489
Drug (1-DA; 2-Pla) 0.493 (0.067) 0.515 (0.062) 0.161 0.694
Task*Drug 0.052 0.822
LMTG&LPAR Task (1-sem; 2-phono) 0.384 (0.058) 0.413 (0.074) 0.515 0.485
Drug (1-DA; 2-Pla) 0.419 (0.068) 0.378 (0.076) 0.378 0.548
Task*Drug 0.933 0.350
LMTG&LFUS Task (1-sem; 2-phono) 0.368 (0.065) 0.452 (0.067) 7.295 0.017
Drug (1-DA; 2-Pla) 0.449 (0.058) 0.371 (0.081) 1.706 0.213
Task*Drug 0.165 0.691
LFUS&LPAR Task (1-sem; 2-phono) 0.527 (0.058) 0.576 (0.054) 1.361 0.263
Drug (1-DA; 2-Pla) 0.606 (0.055) 0.496 (0.060) 4.889 0.044
Task*Drug 0.302 0.591
Open in a new tab

Fig 5.

Fig 5

Open in a new tab

Mean correlation coefficient for semantic and phonological tasks for a) LIFG-LMTG pair, b) LIFG-LPAR pair, c) LMTG-LFUS pair (average over the two drug conditions).

DISCUSSION

fMRI was used to study functional connectivity associated with semantic and phonological processing and whether this is affected by L-Dopa. Brain activation maps of the semantic and phonological networks were obtained and functional connectivity was calculated as the degree of correlation between the activation time series data of two brain areas commonly activated by the two tasks.

Behavioral data, consisting of response times to the stimuli, were recorded. These results demonstrate that phonological processing was accomplished faster than semantic processing. This could reflect differences in attentional resources required by the two tasks. This could also reflect the level of encoding associated with semantic and phonological tasks: a ‘deeper’ level of encoding, corresponding to the semantic process, took longer to accomplish, while the ‘shallower’ level equivalent to the phonological process allowed faster responses (Craik & Lockhart, 1972; Kapur et al, 1994; Mottron, 2001). Response times were not affected by the treatment condition.

The fMRI results agreed with previous findings, showing the network of brain regions involved in semantic and phonological processes: left inferior frontal cortex- LIFG (BA44/45) extending into the premotor and motor areas, bilateral middle frontal gyrus-MFG (BA46/9), left posterior middle temporal gyrus-LpMTG (BA 21/22), left fusiform gyrus-LFUS (BA37), bilateral occipital cortex (BA 17/18/19), and bilateral premotor cortex.

Left inferior prefrontal cortex (LIPC) participates in a network activated by controlled processing of both semantic and non-semantic information (Bookheimer, 2002; Devlin et al., 2003; Gabrielli et al., 1998; Price, 2000; Poldrack et al., 1999; Thompson-Schill et al, 1997; Wagner et al., 2001). There is a distinction in this region between anterior and posterior areas of this region, which seem to be relatively specialized for either semantic (the anterior part (BA 45/47/10)) or phonological (posterior regions (BA 44/6)) processing. The posterior regions of the brain activated in these tasks seem to have a relatively greater involvement in retrieval of stored information such as word meaning (BA21) (Demb et al., 1995; Gold and Buckner, 2002; McDermott et al., 2003; Poldrack et al., 1999; Shivde & Thompson-Schill, 2004) or sound (BA7/40) (McDermott et al., 2003; Shivde & Thompson-Schill, 2004). Therefore, the findings on our activation maps correlate with the general knowledge about the distribution of the semantic and phonological networks in the brain (Crosson et al., 2003; Devlin et al., 2003; McDermott et al., 2003; Poldrack et al., 1999; Thompson-Schill et al., 1997).

In addition to this, the functional connectivity analysis demonstrated the degree of collaboration between language specific areas when processing semantic and phonological information. Language areas were activated in a more synchronous manner for phonological tasks than for semantic tasks. Since semantic tasks had a greater response time, this apparent increase in functional connectivity during the phonological tasks does not appear to be an epiphenomenon of a greater proportion of time spent processing during a task. This may be due to the higher degree of perceptual coherence of the phonological stimuli. Examination of specific ROI pairs did reveal that this increase in functional connectivity for the phonological task occurred between the fusiform and middle temporal areas, which might support this hypothesis since this may represent the communication between the visual word form area (fusiform) (Beversdorf et al., 1997) and Wernicke’s area (posterior middle temporal cortex). Alternatively, perhaps this is due to phonological decisions being mediated to a greater degree within language areas, and therefore demonstrates greater coherence of activation within language areas, whereas semantic tasks likely require greater interaction between language areas and other regions outside of the language areas. Examination of specific ROI pairs revealed increased phonological functional connectivity for the LIFG-LMTG and LIFG-LPAR pairs. This may represent greater synchronous language-specific information for phonological tasks when Broca’s area (LIFG) utilizes information for language-related decisions primarily from Wernicke’s area (posterior LMTG) and language activated parietal areas (LPAR), as compared to semantic tasks which may require greater input from other areas. However, these hypotheses based on specific ROI pairs must be considered purely speculative at this point and in need of further study due to the presence of multiple comparisons. It is also possible that because the phonological task required less attentional resources than the semantic task, an automatic processing of these stimuli has happened, which may have led to better correlation of the signal from different brain regions than in the case of semantic processing. Similarly, as the phonological task was performed faster, this may have led to greater fluctuation in signal intensity in all brain regions, resulting in an apparent increase in functional connectivity for the phonological task.

There were no significant drug effects on either the activation patterns or the functional connectivity, in contrast to the L-Dopa restriction of the semantic network suggested behaviorally in priming paradigms (Angwin et al., 2004; Kischka et al., 1996). Therefore, our study cannot contribute any experimental evidence to the constructs of signal-to-noise or semantic network effects of the dopaminergic system. However, in the priming experiments, the task required fast recognition of target words after the presentation of prime words (varying in semantic distance and stimulus onset asynchrony). Therefore, our current task significantly differed as it involved categorization of words rather than word recognition, without involvement of a priming effect. This would likely result in a more controlled and less automatic processing during our paradigm. Since this functional connectivity study examined L-Dopa effect on a slower, more controlled task, further work will be needed to examine the effect of this drug on functional connectivity in more rapid tasks, such as semantic priming. The lack of a drug effect may also relate to the differences in deployment of attentional resources between the more automatic priming tasks and our more controlled tasks. Of interest, the one ROI pair that did appear to be affected by L-DOPA was the LFUS-LPAR pair, which may represent drug effects modulating word perception perhaps resulting from interactions with fronto-striatal regions containing a higher density of dopaminergic projections, which might be expected to result in priming effects as well. However, this must be interpreted with caution as no overall drug effect was seen and no other ROI pair revealed a drug effect.

In conclusion, the finding of increased functional connectivity for phonological tasks as compared to semantic tasks raises important questions regarding the nature of functional connectivity. Further research will be necessary examining how this relates to perception and response selection for tasks varying in contextual complexity Further work will also be necessary to understand how the dopaminergic system affects brain activation using a range of other semantic tasks.

Acknowledgments

This research was funded by the Alumni Grants for Graduate Research and Scholarship (AGGRS) from The Ohio State University Graduate School and by Davis Medical Research Grant from the Ohio State University Medical Center. Dr. Beversdorf is also funded by grants from NIDA (R21 DA015734) and NINDS (K23 NS43222).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Contributor Information

Madalina E. Tivarus, Department of Neurology, The Ohio State University and University of Rochester

Ashleigh Hillier, Department of Neurology, The Ohio State University and University of Massachusetts-Lowell.

Petra Schmalbrock, Department of Radiology, The Ohio State University.

David Q. Beversdorf, Department of Neurology, The Ohio State University

References

  1. Angwin AJ, Chenery HJ, Copland DA, Arnott WL, Murdoch BE, Silburn PA. Dopamine and semantic activation: an investigation of masked direct and indirect priming. Journal of International Neuropsychological Society. 2004;10:15–25. doi: 10.1017/S1355617704101033. [DOI] [PubMed] [Google Scholar]
  2. Arfanakis K, Cordes D, Haughton VM, Moritz CH, Quigley MA, Meyerand ME. Combining independent component analysis and correlation analysis to probe interregional connectivity in fMRI task activation datasets. Magnetic Resonance Imaging. 2000;18:921–930. doi: 10.1016/s0730-725x(00)00190-9. [DOI] [PubMed] [Google Scholar]
  3. Arnott AL, Chenery HL, Murdoch BE, Silburn PA. Semantic priming in Parkinson’s disease: evidence for delayed spreading activation. Journal of Clinical and Experimental Neuropsychology. 2001;23:502–519. doi: 10.1076/jcen.23.4.502.1224. [DOI] [PubMed] [Google Scholar]
  4. Behrens TE, Woolrich MW, Smith S. Human Brain Mapping. New York: 2003. Jun, Multi-Subject Null Hypothesis Testing Using a Fully Bayesian Framework: Theory. [Google Scholar]
  5. Beversdorf DQ, Hughes JD, Steinberg BA, Lewis LD, Heilman KM. Noradrenergic modulation of cognitive flexibility in problem solving. Neuroreport. 1999;10:2763–2767. doi: 10.1097/00001756-199909090-00012. [DOI] [PubMed] [Google Scholar]
  6. Beversdorf DQ, Ratcliffe NR, Rhodes CH, Reeves AG. Pure alexia: clinical-pathologic evidence for a lateralized visual language association cortex. Clinical neuropathology. 1997;16(6):328–31. [PubMed] [Google Scholar]
  7. Beversdorf DQ, Smith BW, Crucian GP, Anderson JM, Keillor JM, Barrett AM, Hughes JD, Felopulos GJ, Bauman ML, Nadeau SE, Heilman KM. Increased discrimination of “false memories” in autism spectrum disorder. Proceedings of the National Academy of Science of the USA. 2000;97(15):8734–8737. doi: 10.1073/pnas.97.15.8734. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Beversdorf DQ, White DM, Chever DC, Hughes JD, Bornstein RA. Central beta-adrenergic modulation of cognitive flexibility. Neuroreport. 2002;13:2505–2507. doi: 10.1097/00001756-200212200-00025. [DOI] [PubMed] [Google Scholar]
  9. Bookheimer S. Functional MRI of language: new approaches to understanding the cortical organization of semantic processing. Annual Reviews of Neuroscience. 2002;25:151–188. doi: 10.1146/annurev.neuro.25.112701.142946. [DOI] [PubMed] [Google Scholar]
  10. Cabeza R, Nyberg L. Imaging cognition II: An empirical review of 275 PET and fMRI studies. Journal of Cognitive Neuroscience. 2000;12(1):1–47. doi: 10.1162/08989290051137585. [DOI] [PubMed] [Google Scholar]
  11. Craik FIM, Lockhart RS. Levels of Processing: A Framework for Memory Research. Journal of Verbal Learning and Verbal Behavior. 1972;11:671–684. [Google Scholar]
  12. Crosson B, Benefield H, Cato MA, Sadek JR, Moore AB, et al. Left and right basal ganglia and frontal activity during language generation: contributions to lexical, semantic, and phonological processes. Journal of International. Neuropsychological Society. 2003;9(7):1061–1077. doi: 10.1017/S135561770397010X. [DOI] [PubMed] [Google Scholar]
  13. Dehaene S, Jonides J, Smith EE, Spitzer M. Thinking and Problem Solving. In: Zigmond MJ, et al., editors. Fundamental Neuroscience. 1. San Diego: Academic Press; 1999. pp. 1543–1564. [Google Scholar]
  14. Demb JB, Desmond JE, Wagner AD, et al. Semantic encoding and retrieval in the left inferior prefrontal cortex: a functional MRI study of task difficulty and process specificity. Journal of Neuroscience. 1995;15:5870–5878. doi: 10.1523/JNEUROSCI.15-09-05870.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Devlin JT, Matthews PM, Rushworth MF. Semantic processing in the left inferior prefrontal cortex: a combined functional magnetic resonance imaging and transcranial magnetic stimulation study. Journal of Cognitive Neuroscience. 2003;15(1):71–84. doi: 10.1162/089892903321107837. [DOI] [PubMed] [Google Scholar]
  16. Durstewitz D, Seamans JK. The computational role of dopamine D1 receptors in prefrontal cortex. Neural Networks. 2002;15:561–572. doi: 10.1016/s0893-6080(02)00049-7. [DOI] [PubMed] [Google Scholar]
  17. Gabrieli JD, Poldrack RA, Desmond JE. The role of left prefrontal cortex in language and memory. Proceedings of the National Academy of Science of the USA. 1998;95(3):906–913. doi: 10.1073/pnas.95.3.906. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gold BT, Buckner RL. Common prefrontal regions coactivate with dissociable posterior regions during controlled semantic and phonological tasks. Neuron. 2002;35(4):803–812. doi: 10.1016/s0896-6273(02)00800-0. [DOI] [PubMed] [Google Scholar]
  19. Fiez JA, Petersen SE. Neuroimaging studies of word reading. Proceedings of the National Academy of Science of the USA. 1998;95(3):914–21. doi: 10.1073/pnas.95.3.914. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Friston KJ. Functional and Effective Connectivity in Neuroimaging: A Synthesis. Human Brain Mapping. 1994a;2:56–78. [Google Scholar]
  21. Friston KJ, Worsley KJ, Frakowiak RSJ, Mazziotta JC, Evans AC. Assessing the significance of focal activations using their spatial extent. Human Brain Mapping. 1994b;1:214–220. doi: 10.1002/hbm.460010306. [DOI] [PubMed] [Google Scholar]
  22. Froman SD, Cohen JD, Fitzgerald M, Eddy WF, Mintun MA, Noll DC. Improved assessment of significant activation in functional magnetic resonance imaging (fMRI): Use of a cluster-size threshold. Magnetic Resonance in Medicine. 1995;33:636–647. doi: 10.1002/mrm.1910330508. [DOI] [PubMed] [Google Scholar]
  23. Gold BT, Buckner RL. Common prefrontal regions coactivate with dissociable posterior regions during controlled semantic and phonological tasks. Neuron. 2002;35(4):803–812. doi: 10.1016/s0896-6273(02)00800-0. [DOI] [PubMed] [Google Scholar]
  24. Jenkinson M, Smith S. A global optimization method for robust affine registration of brain images. Medical Image Analysis. 2001;5:143–156. doi: 10.1016/s1361-8415(01)00036-6. [DOI] [PubMed] [Google Scholar]
  25. Jenkinson M, Bannister P, Brady M, Smith S. Improved optimization for the robust and accurate linear registration and motion correction of brain images. NeuroImage. 2002;17:825–841. doi: 10.1016/s1053-8119(02)91132-8. [DOI] [PubMed] [Google Scholar]
  26. Just MA, Cherkassky VL, Keller TA, Minshew NJ. Cortical activation and synchronization during sentence comprehension in high-functioning autism: evidence of underconnectivity. Brain. 2004;127:1811–1821. doi: 10.1093/brain/awh199. [DOI] [PubMed] [Google Scholar]
  27. Kapur S, Craik FIM, Tulving E, Wilson AA, Houle S, Brown GM. Neuroanatomical correlates of encoding in episodic memory: levels of processing effect. Proceedings of the National Academy of Science of the USA. 1994;91(6):2008–2011. doi: 10.1073/pnas.91.6.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kischka U, Kammer T, Maier S, Weisbrod M, Thimm M, Spitzer M. Dopaminergic modulation of semantic network activation. Neuropsychologia. 1996;34:1107–1113. doi: 10.1016/0028-3932(96)00024-3. [DOI] [PubMed] [Google Scholar]
  29. Koshino H, Carpenter PA, Minshew NJ, Cherkassky VL, Keller TA, Just MA. Functional connectivity in an fMRI working memory task in high-functioning autism. NeuroImage. 2005;24:810–821. doi: 10.1016/j.neuroimage.2004.09.028. [DOI] [PubMed] [Google Scholar]
  30. Kucera H, Francis WN. Computational Analysis of Present-Day American English. Providence: Brown University Press; 1967. [Google Scholar]
  31. Martin RC. Language processing: functional organization and neuroanatomical basis. Annual Review of Psychology. 2003;54:55–89. doi: 10.1146/annurev.psych.54.101601.145201. [DOI] [PubMed] [Google Scholar]
  32. McDermott KB, Petersen SE, Watson JM, Ojemann JG. A procedure for identifying regions preferentially activated by attention to semantic and phonological relations using functional magnetic resonance imaging. Neuropsychologia. 2003;41(3):293–303. doi: 10.1016/s0028-3932(02)00162-8. [DOI] [PubMed] [Google Scholar]
  33. Mesulam MM. Large Scale Neurocognitive Networks and Distributed Processing for Attention, Language and Memory. Annals of Neurology. 1990;28(5):597–613. doi: 10.1002/ana.410280502. [DOI] [PubMed] [Google Scholar]
  34. Mottron L, Morasse K, Belleville S. A study of memory functioning in individuals with autism. The Journal of Child Psychology and Psychiatry. 2001;42(2):253–60. [PubMed] [Google Scholar]
  35. Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia. 1971;9(1):97–113. doi: 10.1016/0028-3932(71)90067-4. [DOI] [PubMed] [Google Scholar]
  36. Poldrack RA, Wagner AD, Prull MW, Desmond JE, Glover GH, Gabrielli JD. Functional specialization for semantic and phonological processing in the left inferior prefrontal cortex. NeuroImage. 1999;10:15–35. doi: 10.1006/nimg.1999.0441. [DOI] [PubMed] [Google Scholar]
  37. Price CJ. The anatomy of language: contributions from functional neuroimaging. Journal of Anatomy. 2000;197:335–359. doi: 10.1046/j.1469-7580.2000.19730335.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Rapp BC, Caramazza A. Disorders of Lexical Processing and the Lexicon. In: Gazzaniga MS, Bizzi E, editors. The Cognitive Neurosciences. Cambridge, Massachusetts: MIT Press; 1995. p. 901. [Google Scholar]
  39. Rissman J, Gazzaley A, D’Esposito M. Measuring functional connectivity during distinct stages of a cognitive task. NeuroImage. 2004;23:752–763. doi: 10.1016/j.neuroimage.2004.06.035. [DOI] [PubMed] [Google Scholar]
  40. Rugg MD, Wilding EL. Retrieval processing and episodic memory. Trends in Cognitive Sciences. 2000;4:108–115. doi: 10.1016/s1364-6613(00)01445-5. [DOI] [PubMed] [Google Scholar]
  41. Seamans JK, Durstewitz D, Christie BR, Stevens CF, Sejnowski TJ. Dopamine D1/D5 receptor modulation of excitatory synaptic inputs to layer V prefrontal cortex neurons. Proceedings of the National Academy of Sciences of the United States of America. 2001;98:301–306. doi: 10.1073/pnas.011518798. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Shivde G, Thompson-Schill SL. Dissociating semantic and phonological maintenance using fMRI. Cognitive, Affective, and Behavioral Neuroscience. 2004;4(1):10–19. doi: 10.3758/cabn.4.1.10. [DOI] [PubMed] [Google Scholar]
  43. Silver JA, Hughes JD, Bornstein RA, Beversdorf DQ. Effect of anxiolytics on cognitive flexibility in problem solving. Cognitive and Behavioral Neurology. 2004;17:93–97. doi: 10.1097/01.wnn.0000119240.65522.d9. [DOI] [PubMed] [Google Scholar]
  44. Smith SM. Fast robust automated brain extraction. Human Brain Mapping. 2002;17:143–155. doi: 10.1002/hbm.10062. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, et al. Advances in functional and structural MR image analysis and implementation as FSL. NeuroImage. 2004;23(S1):208–219. doi: 10.1016/j.neuroimage.2004.07.051. [DOI] [PubMed] [Google Scholar]
  46. Spitzer M, Braun U, Maier S, Hermle L, Maher BA. Indirect semantic priming in schizophrenic patients. Schizophrenia research. 1993;11:71–80. doi: 10.1016/0920-9964(93)90040-p. [DOI] [PubMed] [Google Scholar]
  47. Spitzer M, Weisker I, Winter M, Maier S, Hermle L, Maher BA. Semantic and phonological priming in schizophrenia. Journal of Abnormal Psychology. 1994;103:485–94. doi: 10.1037//0021-843x.103.3.485. [DOI] [PubMed] [Google Scholar]
  48. Thompson-Schill SL, D’Esposito M, Aguirre GK, Farah MJ. Role of left inferior prefrontal cortex in retrieval of semantic knowledge: a reevaluation. Proceedings of the National Academy of Science of the USA. 1997;94(26):14792–14797. doi: 10.1073/pnas.94.26.14792. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Wagner AD, Pare-Blagoev EJ, Clark J, Poldrack RA. Recovering meaning: left prefrontal cortex guides controlled semantic retrieval. Neuron. 2001;31:329–338. doi: 10.1016/s0896-6273(01)00359-2. [DOI] [PubMed] [Google Scholar]
  50. Watters PA, Patel M. Semantic processing deficits in patients with Parkinson’s disease: degraded representation or defective retrieval? Journal of Psychiatry and Neuroscience. 1999;24(4):322–332. [PMC free article] [PubMed] [Google Scholar]
  51. Winterer G, Weinberger DR. Genes, dopamine and cortical signal-to-noise ratio in schizophrenia. Trends in Neurosciences. 2004;27:683–690. doi: 10.1016/j.tins.2004.08.002. [DOI] [PubMed] [Google Scholar]
  52. Worsley KJ, Evans AC, Marrett S, Neelin P. A three-dimensional statistical analysis for CBF activation studies in human brain. Journal of the Cerebral Blood Flow and Metabolism. 1992;12:900–918. doi: 10.1038/jcbfm.1992.127. [DOI] [PubMed] [Google Scholar]

RESOURCES