www.elsevier.com/locate/braindev

Brain & Development 31 (2009) 562–567

Original article

Maturation of inhibitory and excitatory motor cortexpathways in children

Michael Walther a,1, Steffen Berweck b,2, Joachim Schessl d, Michaela Linder-Lucht a,1,Urban M. Fietzek b,2, Franz X. Glocker c,3, Florian Heinen b,2, Volker Mall a,*

a Division of Neuropediatrics and Muscular Disorders, Department of Pediatrics and Adolescent Medicine,

University of Freiburg, Mathildenstrasse 1, 79106 Freiburg, Germanyb Paediatric Neurology and Developmental Medicine, University of Munich, Lindwurmstr. 4, 80337 Munich, Germany

c Dept. of Neurology, University of Freiburg, Breisacher Strasse 64, 79106 Freiburg, Germanyd Friedrich-Baur-Institute, Clinic of Neurology, Ludwig-Maximilians-University, Ziemssenstr. 1, 80336 Munich, Germany

Received 23 December 2008; received in revised form 3 February 2009; accepted 16 February 2009

Abstract

Objective: To study intracortical inhibition and facilitation with paired-pulse transcranial magnetic stimulation in children, ado-lescents and adults. Methods: Paired-pulse transcranial magnetic stimulation (interstimulus intervals (ISI): 1, 3, 5, 10 and 20 ms) wasapplied over the primary motor cortex (M1) in 30 healthy subjects (range 6–30 years, median age 15 years and 8 months, SD 7,9)divided in three groups: adults (P 18 years), adolescents (> 10 and < 18 years) and children (6 10 years). Results: We observedsignificantly less intracortical inhibition (SICI) in children’s M1 compared to that of adults. Adolescents showed significantly lessSICI at the 5 ms interval than did adults. No significant differences were apparent in intracortical facilitation (ICF). Conclusion:We postulate that, as in adults, the maturing M1 possesses horizontal glutamatergic cross-links that represent the neuronal substrateof excitatory intracortical pathways. GABAergic interneurons, the neuronal substrate of inhibitory intracortical pathways, maturebetween childhood and adulthood. Reduced GABAergic inhibition may facilitate neuronal plasticity and motor learning in children.� 2009 Elsevier B.V. All rights reserved.

Keywords: Children; Neuronal plasticity; Cortical excitability; Intracortical inhibition; Intracortical facilitation; Motor cortex maturation; Trans-cranial magnetic stimulation

1. Introduction

Intracortical inhibition and facilitation are two inter-acting phenomena representing a homeostatic regulatory

0387-7604/$ - see front matter � 2009 Elsevier B.V. All rights reserved.

doi:10.1016/j.braindev.2009.02.007

Abbreviations: AMPA, a-amino-3-hydroxyl-5-methyl-4-isoxazole-propionamma aminobutyric acid; ICF, intracortical facilitation; ISI, interstimulus intprimary motor cortex; MEP, motor-evoked potential; MT, motor thresholderror of mean; SICI, short interval intracortical inhibition; TMS, transcranial

* Corresponding author. Tel.: +49 (761) 270 4310; fax: +49 (761) 270 434

1 Tel.: +49 761 270 4315.2 Tel.: +49 89 5160 7851.3 Tel.: +49 761 270 5001.

concept in synaptic plasticity. Animal experiments andanimal motor cortex slice preparation studies revealedexcitatory horizontal pathways in the cortical layer II/III that span the M1 and are blocked by inhibitory

te; cMEP, conditioned MEP; EMG, electromyography; GABA, ga-erval; LTD, long-term depression; LTP, long-term potentiation; M1,; NMDA, N-methyl-D-aspartic acid; PP, paired-pulse; SEM, standardmagnetic stimulation; TS, test stimulus; ucMEP, unconditioned MEP.

4.

M. Walther et al. / Brain & Development 31 (2009) 562–567 563

mechanisms [1–3]. The motor cortexs moment-to-moment plasticity is induced by a change in the intracor-tical inhibition level. It regulates the strength of excitatorypathways, thus mediating the intensity of intracorticalfacilitation [4]. In cat M1 slice preparation studies, thelevel of GABAergic intracortical inhibition correlatesinversely to activity-dependent and long-term plasticityeffects [5]. The GABergic system A’s maturation processduring brain development has been well documented inmature mice M1 in in-vitro rodent experiments with cor-tical cultures [6–8]. During nascent synaptogenesis,GABA was initially revealed as an excitatory neurotrans-mitter, shifting to inhibitory characteristics during thedevelopment of the K+/Cl� Transporter [7]. GABAsinhibitory intensity is thus developmentally determined.This, and an additional increase in glutamatergic trans-mission, lead to a matured, narrowed GABA/glutamateratio [8].

Transcranial magnetic stimulation (TMS) hasenabled us to better understand the cortical circuits inhuman adult M1, namely, that they are responsible foractivity-dependent neuronal plasticity. Kujirai et al.demonstrated intracortical inhibition (SICI) at the inter-stimulus intervals (ISI) of 1–6 ms in their TMS-investi-gation using a paired-pulse paradigm, while ISI of 10and 20 ms induce intracortical facilitation (ICF) [9].Further studies demonstrate this inhibition as GABAer-gic, showing the reinforcement of glutamate-mediatedexcitatory pathways after blocking the GABA uptake[10–13].

These results provide evidence of an activity-depen-dent dynamic interaction in adult human M1 similarto the moment-to-moment synaptic plasticity in animalexperiments. This interactions function seems to behavein an almost linear and reciprocal manner [11,13] – thusa high GABAergic inhibition level leads to less, and alow GABAergic level to greater synaptic plasticity. Itremains unclear whether GABAergic intracortical inhi-bition involves a maturation process in human M1 aswell.

This study evaluates the hypothesis of a maturationprocess in intracortical inhibition and facilitation evalu-ated using the short interval paired-pulse paradigm inchildren, adolescents and adults.

2. Methods

2.1. Participants

Our study was approved by the Ethics Committee ofthe University of Freiburg, Germany (Number 181/07).All participants gave informed consent; parents pro-vided informed consent for participants under 18 years.

Thirty healthy subjects were included (19 male, 11female), 10 children (6 10 years, n = 10, mean age = 8 -years and 3 months, range 6–10 years), 10 adolescents

(>10 and < 18 years, n = 10, mean age = 12 years and10 months, range 11–17 years) and 10 adults (P18 years, n = 10, mean age = 25 years and 8 months,range 21–30 years).

The lower age limit of 10 years in the intermediate‘‘adolescents” group was set following a study con-ducted in our department that revealed relatively greatindividual variation in how the brain matures at thatage [14]. Corticospinal pathway values above that agehave also been shown to be adult-like [15].

2.2. Transcranial magnetic stimulation

Subjects sat in upright position on a chair. A 70 mmfigure-eight-shaped stimulation coil was centered tan-gentially on the scalp over the contralateral motor cor-tex, corresponding to the EMG-recorded firstinterosseus dorsalis muscle (MDI) and adjusted to theposition whence the maximum MEP amplitude wasobtained. The coil’s handle was pointed in a posteriordirection. Orientation was achieved using the preauricu-lar and parasagittal line, and the coil position was keptconstant throughout the experiment. MEPs wererecorded from the first interosseous dorsalis muscleusing the belly-tendon recording technique with surfaceAg-AgCl electrodes (diameter 8 mm). Responses wererecorded using a Multi-Liner electromyograph (JaegerToennies, Hochberg, Germany). The corner frequenciesof the bandpass-filter were set at 2 Hz and 10 kHz,respectively. Sampling rate was 5 kHz. Two monophasicmagnetic stimulators (Magstim 200, Magstim Com-pany, Whitland, UK) were connected with a bistimdevice and discharged into the figure-eight-shaped coil.All TMS procedures, including the evaluation of themotor threshold (MT), were performed using the bistimdevice; MT determination therefore included the changein stimulus intensity by the bistim set-up.

2.3. Motor threshold (MT)

Relaxed MT was determined by raising the stimulusintensity in 1% step-wise increments according to inter-national standards (minimal EMG-response (P 50 lV)in at least 5 of 10 successive stimulations) [16].

2.4. SICI/ ICF

Stimuli were applied while the subjects were relaxing.The intensity of the conditioning stimulus was 20%below the motor threshold. At this low intensity, theTMS-impulse does not produce significant corticospinalactivation [17]. The test stimulus (TS) magnitude was setto produce an MEP at a range of 200–600 lV whengiven alone. An interstimulus interval (ISI) of 1, 3,5 ms (revealing SICI and) 10 and 20 ms (revealingICF) between conditioning and test stimulus was used

564 M. Walther et al. / Brain & Development 31 (2009) 562–567

throughout the experiments. Conditioned and uncondi-tioned stimuli were applied in a randomized order. 10stimuli were applied for every ISI 10 stimuli appliedand documented for off-line analysis. We analyzed theEMG baseline (pre-stimulus period, 100 ms in each sin-gle-trial) by square root means of single-trial EMG datain order to document relaxation in children similar tothat in to adults.

2.5. Statistic analysis

The SICI/ICF ratio was calculated as the quotient ofconditioned motor-evoked potential (cMEP) andunconditioned motor-evoked potential (ucMEP). Statis-tic analysis was performed for the group-mean of everysingle SICI/ICF-ratio. Due to abnormally-distributeddata, we used nonparametric tests. In a first step, weemployed a one-way analysis of variance by ranks(Kruskal–Wallis test) to test the influence of age onSICI/ICF ratios. In case of significance, a nonparamet-ric test for two unrelated samples (Mann–Whitney test)was used to test differences between two age groups.

3. Results

3.1. Single pulse analysis

Relaxation control based on analysis of the squareroot mean of single-trial rectified EMG data revealedno significant difference between children and adults(p = 0.326). We obtained reproducible MEPs from all

Table 1The table depicts the mean, range and standard error of the mean (SEM) of thof age groups was tested by means of the Kruskal–Wallis test for each ISI. DWhitney test. The footnotes illustrate significant results from the statistical a

ISI (ms) MEP in mV (child)

1* Mean 0.7361a

Range 0.1–1.25SEM 0.37

3* Mean 0.9721b

Range 0.13–1,74SEM 0.47

5** Mean 0.9581b

Range 0.32–1.89SEM 0.48

10 Mean 0.983Range 0.34–2.06SEM 0.54

20 Mean 2.071Range 1.13–3.24SEM 0.74

Kruskal–Wallis Test: *p 6 0.01; **p 6 0.001.Mann-Whitney-Test:ap 6 0.05bp 6 0.0011(child against adult)2(child against adolescent)3(adolescents against adult)

participants. The resting MT mean fluctuated between72.30% of maximum stimulator output (MSO) (range40–95%) in children, and 53.9% of MSO (range 45–65%) in adults (p = 0.07). The mean TS magnitude toproduce an MEP between 200 and 600 lV rangedbetween 85% of MSO (range 55–100%) in children,77% in adolescents (range 45–90%), and 59% of MSO(range 40–70%) in adults.

3.2. Cortical excitability

The Kruskal–Wallis test depicts a significant influ-ence of age on the inhibitory ISIs (1, 3 and 5 ms), butno significant influence on the facilitating ISIs (10 and20 ms). The Mann–Whitney-Test shows significantlylower intracortical inhibition for the ISIs 1, 3 and 5 msbetween children and adults, whereas we noted no sig-nificant differences, except for the 5 ms ISI, betweenadults and adolescents. Among the adolescents, the5 ms ISI likewise showed lower intracortical inhibitioncompared to the 1 ms and 3 ms ISI. There were no sig-nificant differences between the inhibitory ISIs in chil-dren and adolescents (Table 1, Figs. 1 and 2).

4. Discussion

Our study revealed a significant influence of age onintracortical inhibition, but no influence on intracorticalfacilitation. We observed significantly lower SICI inchildren compared to adults in all the inhibitory ISItested (1, 3 and 5 ms). We found significantly lower SICI

e SICI/ICF-ratio for each ISI in the different age groups. The influenceifferences between two age groups were tested by means of the Mann-nalyses.

MEP in mV (adolescents) MEP in mV (adults)

0.685 0.2241a

0.16–5.56 0.06–0.570.78 0.140.939 0.2451b

0.14–3.59 0.07–0.571.27 0.151.4413b 0.3271b,3b

0.53–4.04 0.04–0.571.11 0.172.168 2.1340.57–7.38 0.57–6.262.07 1.653.080 2.1391.58–4.92 0.73–5.211.15 1.42

Fig. 1. Boxes in Fig. 1 and 2 show minimum, first quartile, median,third quartile and maximum. If a significant influence was proven inthe Kruskal–Wallis test, differences between age groups were evaluatedby the Mann–Whitney test. Significant differences compared to adults(p < 0.05) are marked (�). (a) Trace I and II show sample EMGdischarges of an adult participant. Lower MEP amplitude after paired-pulse stimulation at the 3 ms ISI (II) compared to EMG-amplitudeafter single-pulse TMS stimulation (I) reveals intracortical inhibition.Traces III and IV show sample EMG discharges of a child. Stimulationin paired-pulse TMS mode at 3 ms ISI (IV) depicts slightly higherEMG-amplitude than in single-pulse stimulation, showing that facil-itation in this ICI is even possible in the maturing M1 (III). (b) TheSICI ratios of the short ISIs (1, 3, 5 ms) are depicted in boxplots for thethree groups. Compared to adults. Significant lower inhibition inchildren was evident in all ISIs. Compared with adults and children,the group of adolescents revealed no significant differences in the groupof the inhibitory ISIs, except for the 5 ms ISI, which showed asignificantly higher ICI in adolescents than in adults.

Fig. 2. (a) Sample EMG Traces of an adult (I and II) and child (IIIand IV) revealed higher EMG-amplitude (II and IV) in the paired-pulse TMS mode (ISI 20 ms) than in single-pulse TMS mode (I andIII). (b) ICF ratios of the longer ISIs (10, 20 ms) are depicted inboxplots for the three groups. No significant influence of age groups onthe facilitating ISIs was apparent.

M. Walther et al. / Brain & Development 31 (2009) 562–567 565

in adolescents than in adults in the 5 ms ISI. No othersignificant age-effect in intracortical facilitation wasapparent. This confirms our hypothesis that GABAergicintracortical inhibition experiences a maturationprocess.

The application of TMS paired-pulse technique inchildren requires thorough evaluation of methodicalaspects. SICI and ICF correlate closely with the ucMEPamplitude. It is well known that MEP amplitudes aresmaller in children than adults, even when using thesame stimulus output. This is due to their higher motorthreshold [14,18]. We therefore standardized the MEPamplitude at a range of 200–600 lV. Furthermore, SICIand ICF must be evaluated under relaxed conditions, asmuscle facilitation is known to eradicate intracortical

566 M. Walther et al. / Brain & Development 31 (2009) 562–567

inhibition at the 2 ms interval [19]. This effect could alsoapply to other ISIs. We analyzed the EMG baseline toverify relaxation in children and adults. The square rootmean of rectified EMG data revealed no significant dif-ference in relaxation between the groups. Taking thisdata into account, we conclude that the differencesfound in SICI are contingent upon maturation and arenot due to methodology.

Since GABA seems to be the responsible procurer ofintracortical inhibition in animal studies [5,3,20], TMSstudies demonstrated GABAergic origin of SICI inhumans as well [10,12] There is little knowledge aboutintracortical inhibition in the maturing human cortex,but some evidence of lower SICI in juvenile M1, as dem-onstrated in the 2 ms ISI [14]. Their finding is consistentwith our results.

There are data from postmortem human slice prepa-ration studies that likewise support the idea of a matu-ration process of the GABAergic system in the humancortex. As in rodents, GABA conduct in the prenatalhuman cortex as excitatory neurotransmitters, however,during the first days after birth, GABA’s functionswitches from an excitatory to an inhibitory role [21].Brooks–Kayal and Pritchett demonstrated a multiplica-tion of GABAA receptors during human brain matura-tion, linked with the diminution of severalbenzodiazepine-sensitive subtypes to just one in adult-hood [22]. Those studies results were analogous toGABAergic maturation in animal studies, and they con-cur with our results as well. A multiplication of GABAA

receptors may explain our in vivo evidence of increasingintracortical inhibition from childhood to adulthood.

If the structural maturation of the GABAergic systemin animals resembles that of humans, we should be jus-tified in presuming similarities in the functional role ofGABAergic intracortical inhibition maturation for syn-aptic plasticity in animals and humans as well. Recentdata from rodents showed a developmental shift in theeffectiveness of an LTP-inducing protocol during corti-cal maturation. In that study, the efficiency of synapticplasticity is in inverse correlation to the developingage. This effect could be eradicated by blocking GAB-Aergic inhibition [23]. Taking their data into account,our results support the idea of a synaptic-plasticity mat-uration process, mediated by GABAergic intracorticalinhibition in humans.

In contrast to SICI, intracortical facilitation is a func-tion of the strength of excitatory neuronal circuits med-iated by glutamatergic synapses [24,25]. As with to theGABAergic system, animal studies support the conceptof an maturation process of the glutamatergic system.The more excitation input the glutamatergic circuitsreceive after birth, the more they begin to replace GABAas crucial excitatory neurotransmitters [21]. Our resultsdepicted a strong ICF in children for both ISI (10 msand 20 ms). This presupposes preexisting glutamatergic

circuits mediated by inhibition-modulated synaptic plas-ticity in that age group.

5. Conclusion

Our results provide evidence that the inhibitoryGABAergic system undergoes a maturation process.Because the GABAergic system is believed to be a keyregulator of synaptic plasticity, our findings seem toindicate an enhanced plasticity capacity in the first dec-ade of primary motor cortex maturation, which wouldfacilitate motor learning during that period. Additionalstudies are necessary to determine whether motor learn-ing is the only aspect possibly influenced by this matura-tion process of inhibition, or if there are other potentialeffects, for example on the recovery from brain insults injuvenile brains. This may have an impact on motorlearning and recovery after brain injury which has tobe assessed in further studies.

Acknowledgement

This paper was presented at the 3rd German-Japa-nese Symposium of Pediatric Neurology on September2008, Munich, Germany.

References

[1] Keller A. Intrinsic connections between representation zones inthe cat motor cortex. Neuroreport 1993;4:515–8.

[2] Aroniadou VA, Keller A. The patterns and synaptic properties ofhorizontal intracortical connections in the rat motor cortex. JNeurophysiol 1993;70(4):1553–69.

[3] Hess G, Jacobs KM, Donoghue JP. N-methyl-D-aspartate recep-tor mediated component of field potentials evoked in horizontalpathways of rat motor cortex. Neuroscience 1994;61:225–35.

[4] Jacobs KM, Donoghue JP. Reshaping the cortical motor map byunmasking latent intracortical connections. Science1991;251(4996):944–7, 22.

[5] Keller A. Intrinsic synaptic organization of the motor cortex.Cereb Cortex 1993;3:430–41.

[6] Swanwick CC, Murthy NR, Mtchedlishvili Z, Sieghart W, KapurJ. Development of gamma-aminobutyric acidergic synapses incultured hippocampal neurons. J Comp Neurol2006;10(495(5)):497–510.

[7] Jensen K, Jensen MS, Bonefeld BE, Lambert JD. Developmentalincrease in asynchronous GABA release in cultured hippocampalneurons. Neuroscience 2000;101(3):581–8.

[8] Klueva J, Meis S, deLima AD, Voigt T, Munsch T. Develop-mental downregulation of GABAergic drive parallels formationof functional synapses in cultured mouse neocortical networks.Dev Neurobiol 2008;68(7):934–49.

[9] Kujirai T, Caramia MD, Rothwell JC, Day BL, Thompson PD,Ferbert A, et al. Corticocortical inhibition in human motorcortex. J Physiol (Lond) 1993;471:501–19.

[10] Werhahn KJ, Kunesch E, Noachtar S, Benecke R, Classen J.Differential effects on motorcortical inhibition induced by block-ade of GABA uptake in humans. J Physiol 1999;517(Pt 2):591–7.

[11] Ziemann U, Rothwell JC, Ridding MC. Interaction betweenintracortical inhibition and facilitation in human motor cortex. JPhysiol (Lond) 1996;496(Pt 3):873–81.

M. Walther et al. / Brain & Development 31 (2009) 562–567 567

[12] Ziemann U, Lonnecker S, Paulus BJ, Steinhoff W. The effect oflorazepam on the motor cortical excitability in man. Exp BrainRes 1996;109:127–35.

[13] Ziemann U. TMS and drugs. Clin Neurophysiol2004;115:1717–29.

[14] Mall V, Berweck S, Fietzek UM, Glocker FX, Oberhuber U,Walther M, Schessl J, et al. Low level of intracortical inhibition inchildren shown by transcranial magnetic stimulation. Neuropedi-atrics 2004;35(2):120–5.

[15] Muller K, Ebner B, Homberg V. Maturation of fastest afferentand efferent central and peripheral pathways: no evidence for aconstancy of central conduction delays. Neurosci Lett1994;166:9–12.

[16] Rossini PM, Barker AT, Berardelli A, Caramia MD, Caruso G,Cracco RQ, et al. Non-invasive electrical and magnetic stimula-tion of the brain, spinal cord and roots: basic principles andprocedures for routine clinical application. Report of an IFCNcommittee. Electroencephalogr clin Neurophysiol 1994;91:79–92.

[17] Di Lazzaro V, Restuccia D, Oliviero A, Profice P, Ferrara L,Insola A, et al. Effects of voluntary contraction on descendingvolleys evoked by transcranial stimulation in conscious humans. JPhysiol (Lond) 1998;508:625–34.

[18] Garvey MA, Ziemann U, Bartko JJ, Denckla MB, Barker CA,Wassermann EM. Cortical correlates of neuromotor developmentin healthy children. Clin Neurophysiol 2003;114:1662–70.

[19] Hanajima R, Furubayashi T, Iwata NK, Shiio Y, Okabe S,Kanazawa I, et al. Further evidence to support different mech-anisms underlying intracortical inhibition of the motor cortex.Exp Brain Res 2003;151:427–34.

[20] Hess G, Aizenman CD, Donoghue JP. Conditions for theinduction of long-term potentiation in layer II/III horizontalconnections of the rat motor cortex. J Neurophysiol1996;75(5)::1765–78.

[21] Ben-Ari Y, Tseeb V, Raggozzino D, Khazipov R, Gaiarsa JL. c-Aminobutyric acid (GABA): a fast excitatory transmitter whichmay regulate the development of hippocampal neurones in earlypostnatal life. Prog Brain Res 1994;102:261–73.

[22] Brooks-Kayal AR, Pritchett DB. Developmental changes inhuman c-aminobutyric acidA receptor subunit composition.Ann Neurol 1993;34:687–93.

[23] Meredith RM, Floyer-Lea AM, Paulsen O. Maturation of long-term potentiation induction rules in rodent hippocampus: role ofGABAergic inhibition. J Neurosci 2003;3(23(35)):11142–6.

[24] Ziemann U, Chen R, Cohen LG, Hallett M. Dextromethorphandecreases the excitability of the human motor cortex. Neurology1998;51:1320–4.

[25] Schwenkreis P, Witscher K, Janssen F, Addo A, Dertwinkel R,Zenz M, et al. Influence of the N-methyl-D-aspartate antagonistmemantine on human motor cortex excitability. Neurosci Lett1999;270:137–40.