Speeded right-to-left information transfer: the result of speeded transmission in right-hemisphere axons?

Neuroscience Letters 380 (2005) 88 92 Speeded right-to-left information transfer: the result of speeded transmission in right-hemisphere axons? Kylie J. Barnett, Michael C. Corballis Research Centre for Cognitive Neuroscience, University of Auckland, Private Bag 92019, Auckland, New Zealand Received 5 October 2004; received in revised form 3 January 2005; accepted 9 January 2005 Abstract Both reaction time (RT) and evoked potential (EP) studies have shown that interhemispheric transfer is faster from the right to the left hemisphere than vice versa. This has been explained either in terms of an asymmetry of callosal fibres or as a result of hemispheric specialization. Here we suggest that it may be due to greater activity resulting from a greater number of fast-conducting, myelinated fibres in the right hemisphere than in the left. Interhemispheric transfer times (IHTTs) were measured in 13 males by comparing latencies and amplitudes of N160 EPs ipsilateral and contralateral to checkerboard stimuli presented to the left or right visual field. IHTT estimates were obtained from three homologous electrode pairs. The shorter IHTT from right-to-left was associated with a concomitant increase in N160 negativity in the right hemisphere. There was no evidence from RTs to stimuli in each visual field to suggest that the right hemisphere was dominant for this task, suggesting that the faster speed of transfer from the right-to-left hemisphere may depend on faster axonal conduction in the right hemisphere relative to the left. 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: EEG; N160; Interhemispheric transfer time (IHTT); Right hemisphere; Asymmetry It has been claimed that interhemispheric transfer is faster from the right to the left (R-to-L transfer) hemisphere than from the left to the right hemisphere (L-to-R transfer). The initial evidence for this came from studies of simple reaction time (RT) in a procedure first devised by Poffenberger [19], in which participants respond with either the left or the right hand to simple visual stimuli in either the left visual field (LVF) or the right visual field (RVF). Mean RT to the uncrossed combinations (i.e., RVF/right hand and LVF/left hand) is subtracted from mean RT to the crossed combination (i.e. LVF/right hand and RVF/left hand). Since the crossed combination implies interhemispheric transfer while the uncrossed combination does not, the crossed uncrossed difference (CUD) provides an estimate of interhemispheric transfer time (IHTT), which is typically in the range 2 6 ms [1]. However, a meta-analysis revealed an asymmetry, such Corresponding author. Present address: Department of Psychology, Department of Genetics, Institute of Neuroscience (TCIN), Rm # 346, Trinity College Dublin, Dublin 2, Ireland. Tel.: +353 1 6088401; fax: +353 6172006. E-mail address: kylie.barnett@tcd.ie (K.J. Barnett). that the difference between the RVF/right hand and LVF/right hand combinations was less than that between LVF/left hand and RVF/left hand combinations, suggesting faster R-to-L transfer than L-to-R transfer [13]. This asymmetry is not always observed, however, and is potentially confounded by overall differences between the hands or between visual fields [5]. A more direct measure of IHTT, and of transfer asymmetry, can be obtained from visual evoked potentials (EPs) by subtracting the latency of the N160 in the hemisphere contralateral to stimulation (direct pathway) from that in the hemisphere ipsilateral to stimulation (callosal pathway) [3,22,17]. IHTT measures from EPs depend on the callosal site, and are typically shorter, and closer to estimates based on RT studies, when derived from EPs at central rather than occipital regions [22,17,20]. More importantly, EP studies confirm that R-to-L transfer is faster than L-to-R transfer [3,2,11]. Visual EPs are absent from the hemisphere ipsilateral to stimulation in commissurotomy and agenesis patients, showing that the corpus callosum determines EP measures of IHTT [12,21]. 0304-3940/$ see front matter 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2005.01.025

K.J. Barnett, M.C. Corballis / Neuroscience Letters 380 (2005) 88 92 89 Two explanations for the asymmetry of transfer (faster R-to-L) have been suggested. Marzi et al. [13] argue that it is best accounted for by an asymmetry in the callosal connections, with a greater number of neurons projecting from the right hemisphere to the left than vice versa (p. 1175). Saron and Davidson [22] similarly suggest an asymmetry in the density of callosal connections in visual areas. This is indirectly supported by the fact that the right occipital cortex is smaller than the left [9]; and one might expect more axons to project from the smaller right cortex than from the larger left cortex. However, there is no direct anatomical evidence for an asymmetry in the number of callosal axons. The second explanation appeals to hemispheric specialization. Nowicka et al. [17] estimated IHTT using stimuli specific to each hemisphere (i.e. words, left hemisphere; spatial gratings, right hemisphere), and found that IHTT was faster when information was transferred from the non-dominant to the dominant hemisphere. In contrast, Braun [1] argued that the hemisphere dominant for a task assists the less able hemisphere. In an EP study, however, Brown and Jeeves [2] found faster R-to-L IHTT in a letter-matching task that had a RVF/left-hemisphere advantage, which makes it unlikely that the dominant verbal hemisphere is aiding the non-dominant hemisphere. Finally, a meta-analysis of EP studies by Brown et al. [3] provides no evidence for the claim that IHTT asymmetry is dependent on hemispheric dominance. We suggest here a third possible explanation, which is that a relative abundance of fast-conducting myelinated axons in the right hemisphere results in both increased activation in and faster transfer from the right-to-left hemisphere. There is at least indirect support for this idea. Miller [14] has proposed that axonal conduction is faster in the right than in the left hemisphere due to a higher concentration of fast-conducting myelinated neurons in the right hemisphere. In a review of 15 studies, he found that 14 showed the right hemisphere to be larger than the left, which might be attributable to a larger number of myelinated neurons in the right hemisphere. The difference was slight perhaps only 1% [10] and was most pronounced in the frontal lobe [6,25]. Relative to the left hemisphere, the right hemisphere has a higher ratio of whiteto-grey matter, especially in frontal and pre-central regions [6], and a lower CT density [4], again suggesting a higher proportion of myelinated neurons. In this present study, we obtained EP estimates of IHTT from three homologous electrode pairs to test the hypothesis that faster R-to-L transfer is indeed related to greater activation in the right hemisphere. Participants were 13 males ranging in age from 21 to 42 years (mean = 30.8, S.D. = 7.4), and with no history of mental illness, neurological disorder, or drug and alcohol abuse. According to the Edinburgh Handedness Inventory [18], three participants were left handed and 10 right handed. In a simple RT task, participants pressed the spacebar in response to presentation of circular checkerboards, with a diameter of 1.4 of visual angle, on a 15 in. SVGA monitor (640 480 pixel resolution). Each checkerboard consisted of alternating black and white squares with an average luminance equal to that of the light grey background, and appeared for 100 ms, with their innermost edge 4 to the left (LVF), right (RVF), or on both (BVF) sides of central fixation. The BVF condition was added for reasons not pertinent to this paper. The experiment was conducted in a quiet, electrically shielded Faraday chamber and participants were monitored via a closed-circuit camera. They sat 57 cm from the screen so that 1 cm corresponded to 1 of visual angle. There were four blocks of 95 trials, two performed with the right hand and two with the left, in a randomly assigned counterbalanced order (i.e. R-L-L-R or L-R-R-L). There was a brief practice with each hand. In each block there were 30 LVF, 30 RVF, 30 BVF and 5 catch trials where no stimulus was presented, randomly ordered. Participants were instructed to press the spacebar as soon as they saw a stimulus appear in any position and to make no response if there was no stimulus (i.e. catch trials). An Anticipation Error message appeared if participants responded prior to stimuli onset. Trials in which an anticipation error or incorrect response occurred were rerun so that there were 380 trials available for analysis. Stimuli were preceded by a central fixation cross that was presented at variable interstimulus intervals (1550, 1750, or 1950 ms). This was done to ensure that participants maintained attention and did not simply respond at fixed intervals. The failure to respond within 3 s was scored as an error. EEG activity was recorded continuously from a 128- channel Ag/AgCl electrode net at a 250-Hz sampling rate. Electrode impedances ranged from 40 to 45 k. EEG data were acquired using a common vertex (Cz) reference and later re-referenced to nazion off-line. Recordings contaminated by eye movements were rejected according to a criterion of 70 V activity recorded in eye channels. Data were segmented and averaged for each unilateral condition (i.e. LVF, RVF), and catch trial conditions were discarded. Mean RTs for each combination of hand and visual field are shown in Table 1. ANOVA showed no significant effects of field (P =.306) or hand (P =.518), and the CUD, as reflected in the interaction between hand and visual field was also insignificant (P =.534); in fact the CUD of 1.7 ms was negative. Hence, the RT analysis was uninformative as to either IHTT or the asymmetry, probably because much larger numbers of trials are needed to yield stable estimates [7]. EP latencies were recorded from electrodes placed according to the standard 10 20 system. The N160 was recorded Table 1 Mean and S.D. RTs (in ms) for stimuli presented to each visual field for each response hand Visual field Response hand Right Left Mean S.D. Mean S.D. LVF 317.4 58.1 313.9 50.2 RVF 316.5 62.0 309.7 52.1

90 K.J. Barnett, M.C. Corballis / Neuroscience Letters 380 (2005) 88 92 Fig. 1. Location of central, parietal and occipital electrode pairs used to measure IHTT. from homologous electrode pairs at central, parietal and occipital sites (C3, C4; P3, P4; O1, O2). Electrode recording sites are shown in Fig. 1. The N160 was defined as the greatest negative amplitude wave occurring between 160 and 200 ms after the onset of the stimulus [17,16]. For each electrode pair IHTTs were calculated by subtracting peak latencies recorded in the hemisphere contralateral to stimulation from those in the hemisphere ipsilateral to stimulation. That is, LVF/right-hemisphere latency minus LVF/left-hemisphere latency provided the measure of R-to-L transfer, while RVF/left-hemisphere latency minus RVF/right-hemisphere latency provided the measure of L-to-R transfer. These latency differences were subjected to ANOVA with direction (R-to-L, L-to-R) and site (central, parietal, occipital) as within-subjects factors. There was a significant main effect of direction [F(1, 12) = 19.085, P =.001], but no main effect of site (P =.670) and no interaction between direction and site [P =.106]. IHTTs were faster from the R-to-L hemisphere (M = 6.97, S.D. = 2.84) than from the L-to-R hemisphere (M = 12.62, S.D. = 3.99). Although the effect was not significant, IHTTs were slightly faster at central sites. The mean directional asymmetry of N160 transfer speeds in both directions at each site is shown in Fig. 2. Fig. 3 plots the average N160 EP components following LVF and RVF stimulation at each site. As a global estimate of hemispheric activation, averaged mean amplitudes ( V) were exported from the time window 160 to 200 ms post-stimulus onset for all central, parietal and occipital electrodes in each visual field (LVF, RVF), for each participant. The amplitude at each electrode in the LVF/righthemisphere condition was compared with that at the homologous electrode in the RVF/left-hemisphere condition (N =38 electrode pairs), using t tests. Chi-square was used to filter out effects due to chance. With 38 comparisons and Chi-square (38 0.05) we can still expect that, by chance, on average 1.9 electrodes will produce a t value that is larger than the critical value. The two hemispheres were characterised by significant differences in activation at 11 electrodes pairs, which is significantly more than the 1.9 expected by chance (χ 2 = 45.87, Fig. 2. Mean directional asymmetry of N160 transfer speed in each direction at each site. Error bars show 1 S.E. of the mean. P <.01). In all 11 pairs the effect was due to increased negativity in the right hemisphere relative to the left. Fig. 4 maps the location of those electrodes characterized by increased activation. It should be noted that both the behavioural and electrophysiological results were unchanged when the three left-handers were eliminated from analyses. Overall, the results confirm that IHHT is faster from R-to-L than from L-to-R. The sampling rate used in the present study (one sample every 4 ms) was lower than that used in other studies (one sample every 2 ms) [3,2]. However, the overall faster IHTT R-to-L obtained here (6.9 ms) falls within the range of 2 16 ms that has been reported using meta-analysis of studies finding speeded R-to-L transfer at parietal and occipital sites. Similarly, faster IHTT R-to-L (range 4 10 ms) at central sites has also been reported using meta-analysis [3]. Differences in transfer times between sites, supports the claim that there are different rates of transfer across different callosal regions [15]. Likewise recent fmri evidence suggests that the behavioural CUD is associated with multiple transfers in parallel including prefrontal decision-making and premotor response regions [8]. Further, as predicted, faster R-to-L transfer was associated with greater activation in the right than in the left hemisphere. The asymmetry is not readily attributable to a righthemispheric dominance, since there was no difference in RT to LVF and RVF stimuli if anything, RT was slightly faster to RVF than to LVF stimuli, favouring the left hemisphere. The greater activation in the right hemisphere might therefore be attributable to intrinsic properties of that hemisphere. Following Miller [14], we suggest that the presence of more rapidly conducting cortico-cortical myelinated axons in the

K.J. Barnett, M.C. Corballis / Neuroscience Letters 380 (2005) 88 92 91 Fig. 3. Grand average N160 ERPs at in the left hemisphere and right hemisphere after LVF and RVF stimulation at each site. right hemisphere might increase the likelihood of neural summation in that hemisphere, so increasing the speed of transfer from the R-to-L hemisphere relative to that from L-to-R. Specifically the evidence finds a larger EP amplitude in the right hemisphere that suggests that in this hemisphere there are relatively rapidly conducting axons, little temporal dispersion, and greater post-synaptic summation. The finding that the EP is larger in the right than the left hemisphere is thus compatible with more rapid conduction in that hemisphere. The faster transmission from R-to-L compared to L-to-R provides direct evidence that cortico-cortical axons originating in the right hemisphere conduct more rapidly than those originating in the left hemisphere. If conduction velocity is slower in the left hemisphere than the right, then the arrival of signals starting at the same time will be more widely dispersed in the left hemisphere. Less signal dispersion in the right hemisphere, means that if post-synaptic neural summation occurs when signals arrive within the time span of one excitatory post-synaptic potential that there will be more chance of neural summation and therefore greater EP amplitude in that hemisphere. The present findings of enhanced EP amplitude in the right hemisphere support these claims. As further support for this hypothesis that there are a greater number of fast-conducting neurons in the right hemisphere, there is evidence for greater coherence of resting EEG in the right hemisphere than in the left [23,24]. Thatcher et al. [23] suggest that there are two determinants of coherence, one based on local connections occurring primarily in grey matter and the other on distant connections that span several centimetres and rely on white matter fibres. The increase in

92 K.J. Barnett, M.C. Corballis / Neuroscience Letters 380 (2005) 88 92 Fig. 4. T-map, location of electrodes with increased N160 negativity. Map shows the location of central, parietal and occipital electrodes with enhanced N160 activation when comparing direct stimulation conditions (i.e. LVF/RH and RVF/LH). right hemisphere white matter [6] suggests that an increase in right hemisphere coherence enables the processing of information over widely distributed regions due to the presence of more rapidly conducting long distance fibres. Further replication of the present findings using a higher sampling rate and coherence analysis would be informative. These findings provide functional electrophysiological evidence regarding the underlying mechanisms that give rise to asymmetrical differences in interhemispheric transfer. Whatever the explanation of our results, they highlight the importance of taking amplitude differences into account when examining differences in IHTT. References [1] C.M.J. Braun, Estimation of interhemispheric dynamics from simple unimanual reaction time to extrafoveal stimuli, Neuropsychol. Rev. 3 (1992) 321 365. [2] W.S. Brown, M.A. Jeeves, Bilateral visual field processing and evoked potential interhemispheric transmission time, Neuropsychologia 31 (1993) 1267 1281. [3] W.S. Brown, E.B. Larson, M.A. Jeeves, Directional asymmetries in interhemispheric transmission time: evidence from visual evoked potentials, Neuropsychologia 32 (1994) 439 448. [4] J.A. Coffman, S. Bloch, Interhemispheric differences in regional density of the normal brain, J. Psychiatr. Res. 18 (1984) 269 275. [5] M.C. Corballis, Hemispheric interactions in simple reaction time, Neuropsychologia 40 (2002) 423 434. [6] R.C. Gur, I.K. Packer, J.P. Hungerbubler, M. Reivich, W.D. Obrist, W.S. Amarnek, H.A. Sackim, Differences in the distribution of grey and white matter in human cerebral hemispheres, Science 207 (1980) 1226 1228. [7] M. Iacoboni, E. Zaidel, Crossed uncrossed difference in simple reaction times to lateralized flashes: between- and within-subjects variability, Neuropsychologia 38 (2000) 535 541. [8] M. Iacoboni, E. Zaidel, Interhemispheric visuo-motor integration in humans: the role of the superior-parietal cortex, Neuropsychologia 42 (2004) 419 425. [9] A. Kertesz, S.E. Black, M. Polk, J. Howell, Cerebral asymmetries on magnetic resonance imaging, Cortex 22 (1986) 117 127. [10] A. Kertesz, M. Polk, S.E. Black, J. Howell, Sex, handedness, and the morphometry of cerebral asymmetry on magnetic resonance imaging, Brain Res. 530 (1990) 40 48. [11] E.B. Larson, W.S. Brown, Bilateral field interactions, hemispheric specialization and evoked potential interhemispheric transmission time, Neuropsychologia 35 (1997) 573 581. [12] G.R. Mangun, S.J. Luck, M.S. Gazzaniga, S.A. Hillyard, Electrophysiological measures of interhemispheric transfer of visual information: studies of split-brain patients, Soc. Neurosci. 17 (1991) 340. [13] C.A. Marzi, P. Bisiacchi, R. Nicoletti, Is interhemispheric transfer of visuomotor information asymmetric? Evidence from a meta-analysis, Neuropsychologia 29 (1991) 1163 1177. [14] R. Miller, Axonal Conduction Time and Human Cerebral Laterality: A Psychobiological Theory, Harwood Academic, Amsterdam, 1996. [15] A.D. Milner, C.R. Lines, Interhemispheric pathways in simple reaction time to lateralized light flash, Neuropsychologia 20 (1982) 171 179. [16] A. Nowicka, E. Fersten, Sex-related differences in interhemispheric transfer time in the human brain, Neuroreport 12 (2001) 4171 4175. [17] A. Nowicka, A. Grabowska, E. Fersten, Interhemispheric transmission of information and functional asymmetry of the human brain, Neuropsychologia 34 (1996) 147 151. [18] R.C. Oldfield, The assessment and analysis of handedness: the Edinburgh inventory, Neuropsychologia 9 (1971) 97 113. [19] A.T. Poffenberger, Reaction time to retinal stimulation with special reference to the time lost in conduction through nervous centres, Arch. Psychol. 23 (1912) 1 73. [20] M.D. Rugg, C.R. Lines, A.D. Milner, Visual evoked potentials to lateralized visual stimuli and the measurement of interhemispheric transmission, Neuropsychologia 22 (1984) 215 225. [21] M.D. Rugg, A.D. Milner, C.R. Lines, Visual evoked potentials to lateralized stimuli in two cases of callosal agenesis, J. Neurol. Neurosurg. Psychiatr. 48 (1985) 367 373. [22] C.D. Saron, R.J. Davidson, Visual evoked potential measures of interhemispheric transfer time in humans, Behav. Neurosci. 103 (1989) 1115 1138. [23] R.W. Thatcher, P.J. Krause, M. Hrybyk, Cortico-cortical associations and EEG coherence: a two-compartment model, Clin. Electroencephalogr. 64 (1986) 123 143. [24] D.M. Tucker, D.L. Roth, T.B. Blair, Functional coherence among cortical regions: topography of EEG coherence, Clin. Electroencephalogr. 63 (1985) 242 250. [25] D.R. Weinberger, D.L. Luchin, J. Morihisa, R.J. Wyatt, Asymmetricl volumes of the right and left frontal and occipital regions of the human brain, Ann. Neurol. 11 (1982) 97 100.