Please cite this article in press as: Polanı´a et al., The Importance of Timing in Segregated Theta Phase-Coupling for Cognitive Performance, Current Biology (2012), doi:10.1016/j.cub.2012.05.021 Current Biology 22, 1–5, July 24, 2012 ª2012 Elsevier Ltd All rights reserved
DOI 10.1016/j.cub.2012.05.021
Report The Importance of Timing in Segregated Theta Phase-Coupling for Cognitive Performance Rafael Polanı´a,1,2,3,* Michael A. Nitsche,1,3 Carolin Korman,1 Giorgi Batsikadze,1 and Walter Paulus1 1Department of Clinical Neurophysiology, Georg-August University of Go¨ttingen, 37075 Go¨ttingen, Germany 2Laboratory for Social and Neural Systems Research, Department of Economics, University of Zurich, 8006 Zurich, Switzerland
Summary Functional cortical circuits for central executive functions have been shown to emerge by theta (w6 Hz) phasecoupling of distant cortical areas [1–3]. It has been repeatedly shown that frontoparietal theta coupling at w0 relative phase is associated with recognition, encoding, short-term retention, and planning [1, 4, 5]; however, a causal link has not been demonstrated so far. Here we used transcranial alternating current stimulation [6–8] simultaneously applied at 6 Hz over left prefrontal and parietal cortices with a relative 0 (‘‘synchronized’’ condition) or 180 (‘‘desynchronized’’ condition) phase difference or a placebo stimulation condition, whereas healthy subjects performed a delayed letter discrimination task. We show that exogenously induced frontoparietal theta synchronization significantly improves visual memory-matching reaction times as compared to placebo stimulation. In contrast, exogenously induced frontoparietal theta desynchronization deteriorates performance. The present findings provide for the first time evidence of causality of theta phase-coupling of distant cortical areas for cognitive performance in healthy humans. Moreover, the results demonstrate the suitability of transcranial alternating current stimulation to artificially induce coupling or decoupling of behaviorally relevant brain rhythms between segregated cortical regions. Results and Discussion Sensory and association areas of the human brain are organized in a distributed manner [9], requiring an efficient communication mechanism to integrate responses across different cortical regions to guide behavior. How the human brain can achieve this relatively fast and efficient integration of information has been the topic of intensive research in the last two decades. A growing amount of studies suggests phase synchronization as a fundamental neural mechanism in cognitive functions requiring large-scale integration of distributed neural activity, supporting both neural communication and plasticity [1, 2, 10]. Phase synchronization in distributed cortical networks during cognitive performance occurs in a wide portion of the spectrum of oscillatory brain activity starting from theta (w4–8 Hz), alpha (w8–12 Hz), beta (w13–30 Hz), and going up to gamma oscillations (>30 Hz), where functional coupling
3These
authors equally contributed to this work *Correspondence:
[email protected]
at—and between—each of these frequency bands appears to coordinate different aspects of behavior [2, 11–13]. In particular, frontoparietal theta coupling at w0 relative phase is associated with a great variety of cognitive processes such as recognition, encoding, short-term retention, and planning [1, 4, 5]. Despite the large amount of empirical data, so far the majority of these studies have provided only correlative evidence for the impact of theta phase synchronization on cognitive performance, whereas its causal role still awaits empirical evidence. In the present study, we first add correlative evidence for the relevance of frontoparietal theta phasecoupling during cognitive performance and, subsequently, we provide its causal evidence. Experiment 1: EEG Experiment In a first set of experiments, we explored the relevance of oscillatory phase synchronization on performance in a delayed letter recognition task by means of electroencephalogram (EEG) analysis in healthy volunteers (Figure 1; see also Supplemental Experimental Procedures available online). Initially, we analyzed the probe period of the task (Figure 1), which involves comparison of a previously displayed and remembered letter with the ‘‘probe’’ cue, a match or nonmatch decision and selection of the appropriate action. Several studies suggest that the integration of sensory information that is subsequently used to guide behavior is carried out through the interconnection of synchronized distributed cortical networks, including the dorsolateral prefrontal cortex (DLPFC) and the posterior parietal cortex (PPC) [12, 14, 15]. Thus, in the present study, we centered the attention on the left DLPFC and PPC. There are three main reasons to focus the analysis on the left hemisphere in the present study: First, the paradigm used here is a working memory and sensorimotor decision-making task with a verbal component, and it has been previously suggested that mainly left DLPFC and PPC are active during similar letter recognition paradigms [16, 17]. Second, we recently showed—using a two letter variant of the same task—that left DLPFC and PPC are directly involved in the temporal storage of memorized letters, where high-gamma oscillations—which were found to represent the temporally memorized information—were modulated by theta activity [17]. Although in that study the contents of working memory were explored during the maintenance periods, we hypothesize that theta-synchronized activity between left PPC and DLPFC should be relevant during cue comparison and the match or nonmatch process. Third, it has been strongly suggested that left-hemisphere frontal and parietal regions are involved in motor selection and preparation [18, 19]. Based on these arguments, we selected EEG electrodes approximately belonging to these cortical regions a priori to investigate phase synchronization during memory matching periods (Supplemental Experimental Procedures; Figure S1). We used the weighted phase lag index (WPLI), a recently introduced debiased index of phase synchronization unaffected by volume conduction and signal amplitude [20] (Supplemental Experimental Procedures; Figure S3). We identified a significant increase of phase synchronization at w4–7 Hz occurring w200–500 ms
Please cite this article in press as: Polanı´a et al., The Importance of Timing in Segregated Theta Phase-Coupling for Cognitive Performance, Current Biology (2012), doi:10.1016/j.cub.2012.05.021 Current Biology Vol 22 No 14 2
presentation of the masks in the delayed letter recognition task. These correlations did not yield significant effects (for all cases r < 0.08, p > 0.45). Hence, the results of this experiment add correlative evidence that timing-dependent phase synchronization of these interregional oscillations is crucial for improved behavior.
Figure 1. Behavioral Task and Stimuli Participants performed a delayed letter discrimination task. Observers attended to three sample letters (‘‘L,’’ ‘‘T,’’ and ‘‘C,’’ subtending 2 of visual angle) that were briefly presented (350 ms) in randomized order. Immediately after each of the letters was presented, a mask stimulus displaying all possible line segments forming the letter stimulus L, T, or C was presented for 1 s to interrupt visual processing of the target shape. After the third mask was presented, a numerical cue (2 ) indicated whether to remember the first, second, or third letter. Probe period: After a 1.5 s delay interval, a test letter was presented and participants indicated as fast as possible whether it matched the numerically cued letter or not (right hand if it matched, left hand otherwise). The probe letter was always one of the initially presented letters. Each experimental session consisted of 90 trials, in which the numerical cue was evenly pseudorandomized. See Supplemental Experimental Procedures section for further details.
after memory probe onset (WPLI z 0.15, normalized WPLI: z = 4.7, confidence interval a(0.01/99%) = (3.1, 6.6); Figures 2B and 2E) (Supplemental Experimental Procedures; Figure S1). Reaction times (RTs) were on average w500 ms after probe onset (506 6 11) and accuracies were near perfect (96% 6 2%). The relative phase difference between both regions was w0 (circular mean = 20.025 6 0.18); Rayleigh test for circular uniformity: p < 0.001, z = 306; Figure 2C). Interestingly, at these frequencies and latencies (w4–7 Hz and w200–500 ms, respectively; Figure 2B), we found a strong positive correlation between the absolute value of the relative phase and reaction times (rPearson’s = 0.26, p < 0.005, Figure 2D; nonparametric statistics supporting this significant correlation can be found in Figure S1), i.e., the closer the phase difference was to 0 the lower the RTs in the matching periods. In order to examine whether this correlation was related to contamination from effects of evoked activity (notice, however, that these analyses were based on the imaginary part of the cross-spectrum), we carried out the following control measurements: First, reaction times and phase difference of theta activity were independently regressed with theta power in both F3 and P3. Additional regressions with the ratio and difference of theta power between F3 and P3 were carried out as well. These regressions did not yield significant effects (for all cases r < 0.11, p > 0.24). Second, it has been shown extensively that the poststimulus presentation time period induces a significant increase of interregional phase synchronization in theta and other frequency bands. Therefore, we investigated whether the correlation between reaction times and phase difference in theta is merely due to the presentation of visual stimuli. The same poststimulus latencies used in the initial analysis (200–400 ms) were used to perform linear regressions between RTs and phase difference, however, after the
Experiment 2: Theta tACS Experiment Based on the results obtained in experiment 1, we hypothesized that an exogenous boost of frontoparietal theta coupling (0 relative phase) should improve reaction times during the memory matching periods, whereas an exogenous induction of ‘‘desynchronization’’ (i.e., 180 relative phase) may deteriorate performance. In order to test this hypothesis, we applied transcranial alternating current stimulation (tACS), which is a noninvasive brain stimulation tool that has been successfully used to entrain oscillatory cortical activity in circumscribed cortical areas [6, 8, 21, 22] (Supplemental Experimental Procedures). We applied tACS at 6 Hz over the left prefrontal and parietal cortices (at F3 and P3, respectively) with a relative 0 (‘‘synchronized’’ condition) or 180 (‘‘desynchronized’’ condition) phase difference or a placebo stimulation condition (Figure 3), while healthy participants (n = 18) performed the same delayed letter discrimination task used in experiment 1 (tACS was administrated during the whole behavioral task, which lasted for w14 6 1.5 min for all subjects and all sessions). All participants received the three stimulation conditions in balanced order (Table S1). RTs during matching memory were submitted to a repeated-measures ANOVA with stimulation (6 Hz_0 , 6 Hz_180 , Sham) and response hand (left, right) as factors (see Supplemental Experimental Procedures for more details). We found a main effect of stimulation (F2,34 = 8.82, p < 0.005; Table S2). Post hoc t tests revealed that RTs in the ‘‘desynchronized’’ condition were significantly larger than in sham [t(17) = 3.41, p < 0.01] and synchronized conditions [t(17) = 3.51, p < 0.01]. Additionally, RTs in the sham condition were significantly slower than in the synchronized condition [t(17) = 1.76, p < 0.05] (Figure 3D). The interaction stimulation 3 response hand was not significant (F2,34 = 1.37, p = 0.28; Table S2), suggesting that tACS-induced modulation in performance is not significantly affected by the response hand used in each trial. Thus, in line with our hypothesis, an exogenous boost of frontoparietal theta coupling improves RTs during the matching periods, whereas an exogenous induction of a 180 relative phase deteriorates reaction performance. Experiment 3: Control Experiment In order to test whether the results obtained in experiment 2 are due to any in or out of phase stimulation or perhaps due to some artifact of the stimulation protocol, we repeated the same experiment (n = 18 new subjects with respect to the previous experiments of this study), however, this time applying a different stimulation frequency: 35 Hz. The main reasons to choose this frequency are as follows: First, phase synchronization and the regression of RTs with absolute phase difference were not significant: r < 0.06 (p > 0.51) and Z < 1 respectively. Second, 35 Hz is out of the range where tACS may induce perception of peripheral flickers at the stimulation intensity used in the present study (see Supplemental Experimental Procedures). In this control experiment, we did not find significant effects for stimulation (F2,34 = 0.86, p = 0.43; Figure 3E) or the interaction stimulation 3 response hand (F2,34 = 1.051, p = 0.36; Table S2). In order to show that
Please cite this article in press as: Polanı´a et al., The Importance of Timing in Segregated Theta Phase-Coupling for Cognitive Performance, Current Biology (2012), doi:10.1016/j.cub.2012.05.021 Coupling/Decoupling Frontoparietal Theta Phase 3
Figure 2. EEG Experiment (A–E) Amplitude of the event related potentials from channels F3 and P3 (A) and their weighted phase lag index (B). Circular histogram representing P3-F3 absolute phase difference (C), linear regression between normalized RTs and the P3-F3 absolute phase difference (D), and normalized WPLI with respect to surrogate data (E) analyzed in the time-frequency interval marked with the white dotted rectangle in (B) are shown. (F) A representative trial band-pass filtered at 6 6 1 Hz is shown to illustrate theta phase synchronization occurring w200 ms after probe onset. Times (x axis) in (A), (B), and (F) are shown with respect to probe onset (see ‘‘probe period’’ description in Figure 1 legend). Error bar in (E) represents bootstrap-t confidence intervals for an alpha (0.01/99%) = (3.1, 6.6). For details on how these figures were generated, please see Supplemental Experimental Procedures section. See also Figure S1.
the effects induced by 6 Hz tACS differed from the effects of 35 Hz stimulation, we performed an additional two-way ANOVA where stimulation frequency (6 Hz and 35 Hz; between subjects) and stimulation synchronization (synchronized, desynchronized, sham; within subjects) were included as factors. We found a significant effect for the interaction stimulation frequency 3 stimulation synchronization (F2,68 = 3.78, p < 0.05).
General Remarks and Implications for Future Work The present findings provide evidence for causality between frontoparietal theta phase-coupling and cognitive performance in healthy humans, however, without ruling out the potential role of segregated phase synchronization at other frequency bands in cognitive processes [2, 12]. Hereby, phase synchronization-dependent effect on performance is in favor for exact timing of these interregional oscillations to improve behavior. Thus, further supporting the idea that network synchronization conveys information widely distributed in the brain to guide behavior [12, 15]. It has been reported that phase synchronization modulates spike timedependent plasticity (STDP) [2], and thus it might be speculated that exogenously driven interregional coupling modulated performance, at least partially, via induction of neuroplastic alterations of functional connectivity. Here, it can be argued that a heightened state of excitability that would modulate performance related to slow-wave oscillations would have to be timed according to the delay of communication between both regions, i.e., a nonzero phase. Therefore, is it required to apply tACS with a nonzero phase timed according the communication delay between two regions to further improve behavior? Our results suggest that this might be not strictly necessary for segregated brain regions such as DLPFC and PPC. In accordance, several studies—empirical as well as computational modeling ones—have reported 0 phase lag for theta phase synchronization between distant brain regions to occur despite of their separation by long axonal conduction delays [23] and have suggested that these
Please cite this article in press as: Polanı´a et al., The Importance of Timing in Segregated Theta Phase-Coupling for Cognitive Performance, Current Biology (2012), doi:10.1016/j.cub.2012.05.021 Current Biology Vol 22 No 14 4
Figure 3. tACS Experiment (A) ‘‘Desynchronized’’ condition (tACS_180 ): Electrodes over F3 and P3 with a 6 Hz frequency stimulation with a 180 relative phase. (B) Sham condition: For sham stimulation sessions, the current was applied for 30 s at the beginning of the stimulation and then turned off (20 s linear down-ramping until 0 mA was reached). (C) ‘‘Synchronized’’ condition (tACS_0 ): Electrodes over F3 and P3 with a 6 Hz frequency stimulation with a 0 relative phase. The return electrode was located over Cz. For more details regarding the tACS parameters, please see Supplemental Experimental Procedures section. (D and E) Reaction times during the memory matching periods are split for each of the stimulation conditions (tACS_180 red, placebo stimulation black, tACS_0 blue) in the 6 Hz experiment (D) and the control experiment (E). Exogenous boost of frontoparietal theta coupling improved reaction times at the memory matching period, whereas exogenous induction of 180 relative phase deteriorates performance. The application of 35 Hz tACS did not result in significant changes. Error bars represent SEM; *p < 0.05, **p < 0.01. See also Figure S2.
synchronizations play a key role in cognitive processes [2, 24]. Alternatively, it might be the case that the observed effects were directly caused by coupling (or decoupling) of frontoparietal theta synchronization without tACS-induced synaptic plasticity. With the current design of the experiment in which
tACS was continuously delivered, a decision for one or the other hypothesis is not possible. A brief stimulation at the time of the probe presentation would be sufficient to observe the same effects, however, under the assumption that amplitude-enhancement trough tACS-induced phase alignment occurs in the order of milliseconds. On the other hand, tACS-induced STDP would predict that the effect outlasts the stimulation period. Here, it might be speculated that the improvement in reaction times in the 6 Hz_0 condition could be due to motor cortex stimulation via the Cz reference electrode. Theoretically, stimulation of motor cortex might increase cortical excitability at the motor cortex level, thereby reducing RTs. We investigated this possibility through measures of motor evoked potentials (MEPs) induced by single pulse transcranial magnetic stimulation (TMS). We did not find a significant change of motor cortex excitability following tACS (details in Figure S2). In an additional experiment, we show that RTs in a simple motor response task are not affected using the 6 Hz_0 protocol (Figure S2). The latter results address two important matters: First, the RT improvement with the 6 Hz_0 protocol in the working memory task cannot be explained by motor cortex stimulation via the Cz reference electrode, and second, frontoparietal theta coupling plays a causal role for more complex cognitive processes than simple stimulus decoding to motor response generation.
Please cite this article in press as: Polanı´a et al., The Importance of Timing in Segregated Theta Phase-Coupling for Cognitive Performance, Current Biology (2012), doi:10.1016/j.cub.2012.05.021 Coupling/Decoupling Frontoparietal Theta Phase 5
In the present study, no EEG measures during tACS were carried out because of technical difficulties to separate brain activity from the continuous alternating electric field induced by tACS. Therefore, we have no direct proof that our tACS protocol increased theta activity under the stimulated areas. However, recent studies in humans suggest that tACS is capable of entraining brain oscillations and modulating brain activity in a frequency- and topographic-specific manner [8, 21, 22]. Moreover, in a recent work, Ozen and colleagues [25] investigated the direct physiological effects of tACS— however, at lower frequencies, e.g., 1.7 Hz—in chronically implanted rats, i.e., while receiving tACS. Interestingly, the investigators found that neocortical neurons oscillate in phase with the oscillatory electric field applied over the scalp, thus providing direct physiological evidence that tACS is capable of exogenously entraining cortical activity at the externally applied frequency. However, these results should be further validated for higher frequencies and in humans. Taken together, the results of the present study motivate us to further seek for the causal relevance of interregional oscillatory cortical activity for cognitive and behavioral processes via noninvasive stimulation in humans more directly than before. Moreover, the possibility that tACS can successfully be used to artificially induce coupling or decoupling of behaviorally relevant brain rhythms between segregated cortical regions might be of potential relevance for the treatment of neurological diseases, such as Alzheimer’s disease, schizophrenia, and autism, where abnormal behavior correlates with ‘‘out-of-phase’’ interregional brain synchronization [26, 27]. On the other hand, exogenously induced desynchronization might also be useful in neurological disorders such as epilepsy, where exaggerated interregional brain coupling is reported [27]. Experimental Procedures The experiments conform to the Declaration of Helsinki, and the experimental protocol was approved by the Ethics Committee of the University of Go¨ttingen. Details regarding the experimental procedures can be found in the Supplemental Information. Supplemental Information Supplemental Information includes three figures, two tables, and Supplemental Experimental Procedures and can be found with this article online at doi:10.1016/j.cub.2012.05.021. Acknowledgments This work was supported by the Rose Foundation (R.P. and W.P.) and by the Bernstein Focus Neurotechnology BFNT Go¨ttingen (W.P.). Received: February 22, 2012 Revised: April 20, 2012 Accepted: May 8, 2012 Published online: June 7, 2012 References 1. Sauseng, P., and Klimesch, W. (2008). What does phase information of oscillatory brain activity tell us about cognitive processes? Neurosci. Biobehav. Rev. 32, 1001–1013. 2. Fell, J., and Axmacher, N. (2011). The role of phase synchronization in memory processes. Nat. Rev. Neurosci. 12, 105–118. 3. Liebe, S., Hoerzer, G.M., Logothetis, N.K., and Rainer, G. (2012). Theta coupling between V4 and prefrontal cortex predicts visual short-term memory performance. Nat. Neurosci. 15, 456–462.
4. Mizuhara, H., and Yamaguchi, Y. (2007). Human cortical circuits for central executive function emerge by theta phase synchronization. Neuroimage 36, 232–244. 5. Perfetti, B., Moisello, C., Landsness, E.C., Kvint, S., Pruski, A., Onofrj, M., Tononi, G., and Ghilardi, M.F. (2011). Temporal evolution of oscillatory activity predicts performance in a choice-reaction time reaching task. J. Neurophysiol. 105, 18–27. 6. Zaehle, T., Rach, S., and Herrmann, C.S. (2010). Transcranial alternating current stimulation enhances individual alpha activity in human EEG. PLoS ONE 5, e13766. 7. Kanai, R., Chaieb, L., Antal, A., Walsh, V., and Paulus, W. (2008). Frequency-dependent electrical stimulation of the visual cortex. Curr. Biol. 18, 1839–1843. 8. Thut, G., Schyns, P.G., and Gross, J. (2011). Entrainment of perceptually relevant brain oscillations by non-invasive rhythmic stimulation of the human brain. Front. Psychol. 2, 170. 9. Buzsa´ki, G. (2006). Rhythms and the Brain (Oxford: Oxford University Press). 10. Rodriguez, E., George, N., Lachaux, J.P., Martinerie, J., Renault, B., and Varela, F.J. (1999). Perception’s shadow: long-distance synchronization of human brain activity. Nature 397, 430–433. 11. Hipp, J.F., Engel, A.K., and Siegel, M. (2011). Oscillatory synchronization in large-scale cortical networks predicts perception. Neuron 69, 387–396. 12. Siegel, M., Donner, T.H., and Engel, A.K. (2012). Spectral fingerprints of large-scale neuronal interactions. Nat. Rev. Neurosci. 13, 121–134. 13. Siegel, M., Engel, A.K., and Donner, T.H. (2011). Cortical network dynamics of perceptual decision-making in the human brain. Front. Hum. Neurosci. 5, 21. 14. Heekeren, H.R., Marrett, S., and Ungerleider, L.G. (2008). The neural systems that mediate human perceptual decision making. Nat. Rev. Neurosci. 9, 467–479. 15. Schall, J.D. (2001). Neural basis of deciding, choosing and acting. Nat. Rev. Neurosci. 2, 33–42. 16. Oztekin, I., McElree, B., Staresina, B.P., and Davachi, L. (2009). Working memory retrieval: contributions of the left prefrontal cortex, the left posterior parietal cortex, and the hippocampus. J. Cogn. Neurosci. 21, 581–593. 17. Polanı´a, R., Paulus, W., and Nitsche, M.A. (2012). Noninvasively decoding the contents of visual working memory in the human prefrontal cortex within high-gamma oscillatory patterns. J. Cogn. Neurosci. 24, 304–314. 18. Schluter, N.D., Krams, M., Rushworth, M.F., and Passingham, R.E. (2001). Cerebral dominance for action in the human brain: the selection of actions. Neuropsychologia 39, 105–113. 19. Rushworth, M.F., Ellison, A., and Walsh, V. (2001). Complementary localization and lateralization of orienting and motor attention. Nat. Neurosci. 4, 656–661. 20. Vinck, M., Oostenveld, R., van Wingerden, M., Battaglia, F., and Pennartz, C.M. (2011). An improved index of phase-synchronization for electrophysiological data in the presence of volume-conduction, noise and sample-size bias. Neuroimage 55, 1548–1565. 21. Feurra, M., Bianco, G., Santarnecchi, E., Del Testa, M., Rossi, A., and Rossi, S. (2011). Frequency-dependent tuning of the human motor system induced by transcranial oscillatory potentials. J. Neurosci. 31, 12165–12170. 22. Pogosyan, A., Gaynor, L.D., Eusebio, A., and Brown, P. (2009). Boosting cortical activity at Beta-band frequencies slows movement in humans. Curr. Biol. 19, 1637–1641. 23. Bibbig, A., Traub, R.D., and Whittington, M.A. (2002). Long-range synchronization of gamma and beta oscillations and the plasticity of excitatory and inhibitory synapses: a network model. J. Neurophysiol. 88, 1634–1654. 24. Seidenbecher, T., Laxmi, T.R., Stork, O., and Pape, H.C. (2003). Amygdalar and hippocampal theta rhythm synchronization during fear memory retrieval. Science 301, 846–850. 25. Ozen, S., Sirota, A., Belluscio, M.A., Anastassiou, C.A., Stark, E., Koch, C., and Buzsa´ki, G. (2010). Transcranial electric stimulation entrains cortical neuronal populations in rats. J. Neurosci. 30, 11476–11485. 26. Schnitzler, A., and Gross, J. (2005). Normal and pathological oscillatory communication in the brain. Nat. Rev. Neurosci. 6, 285–296. 27. Uhlhaas, P.J., and Singer, W. (2006). Neural synchrony in brain disorders: relevance for cognitive dysfunctions and pathophysiology. Neuron 52, 155–168.
