Relevant conflicts of interest/financial disclosures: Nothing to report.
Primary motor cortex long-term plasticity in multiple system atrophy
Article first published online: 2 OCT 2013
Copyright © 2013 Movement Disorder Society
Volume 29, Issue 1, pages 97–104, January 2014
How to Cite
Suppa, A., Marsili, L., Di Stasio, F., Latorre, A., Parvez, AK., Colosimo, C. and Berardelli, A. (2014), Primary motor cortex long-term plasticity in multiple system atrophy. Mov. Disord., 29: 97–104. doi: 10.1002/mds.25668
Full financial disclosures may be found in the online version of this article.
- Issue published online: 23 JAN 2014
- Article first published online: 2 OCT 2013
- Manuscript Accepted: 12 AUG 2013
- Manuscript Revised: 2 AUG 2013
- Manuscript Received: 19 APR 2013
- multiple system atrophy;
- motor cortex;
- motor control
In humans, intermittent and continuous theta-burst stimulation (iTBS and cTBS) elicit long-term changes in motor-evoked potentials (MEPs) reflecting long-term potentiation (LTP)- and depression (LTD)-like plasticity in the primary motor cortex (M1). In this study, we used TBS to investigate M1 plasticity in patients with MSA. We also assessed whether responses to TBS reflect M1 excitability as tested by short-interval intracortical inhibition (SICI), intracortical facilitation (ICF), short-interval intracortical facilitation (SICF), and the input/output curves. We studied 20 patients with MSA and 20 healthy subjects (HS). Patients were clinically evaluated with the Unified Multiple System Atrophy Rating Scale. The left M1 was conditioned with TBS. Twenty MEPs were recorded from the right first dorsal interosseous muscle before TBS and 5, 15, and 30 minutes thereafter. In a subgroup of 10 patients, we also tested MEPs elicited by SICI, ICF, SICF, and input/output curves, before TBS. Between-group analysis of variance showed that at all time points after iTBS MEPs increased, whereas after cTBS they decreased only in HS. In both subgroups tested, patients with predominant parkinsonian and cerebellar features, iTBS and cTBS left MEPs unchanged. MSA patients had reduced SICI, but normal ICF, SICF, and input/output curves. No correlation was found between patients' clinical features and responses to TBS and M1 excitability variables. These findings suggest impaired M1 plasticity in MSA. © 2013 International Parkinson and Movement Disorder Society
MSA manifests with parkinsonian, cerebellar, autonomic, and pyramidal signs. Current classifications distinguish parkinsonian (MSA-P) and cerebellar subtypes (MSA-C). The pathophysiology of motor impairment in MSA remains largely unclear.[1-3]
Previous studies in MSA investigated primary motor cortex (M1) excitability with transcranial magnetic stimulation (TMS) and reported abnormal short-interval intracortical inhibition (SICI), suggesting reduced M1 inhibition.[4-8] Besides M1 excitability measures, it is important to investigate M1 excitability changes over time after applying plasticity-inducing protocols. Plasticity arising from long-term potentiation (LTP) or depression (LTD) operates in M1 as a physiological process required for motor execution and learning.[9-11] Hence, in MSA, altered M1 plasticity might lead to impaired motor execution and learning, thus contributing to the pathophysiology of motor symptoms. In humans, a TMS technique for investigating LTP/LTD-like plasticity entails examining long-term changes in motor-evoked potentials (MEPs; after effects) after theta-burst stimulation (TBS). After intermittent TBS (iTBS), MEPs increase, whereas after continuous TBS (cTBS), they decrease, reflecting LTP/LTD-like plasticity in M1.[12, 13]
No studies have investigated M1 plasticity in MSA or compared TBS responses in MSA-P and MSA-C. Finally, none have verified whether in MSA-P and MSA-C possible changes in LTP/LTD-like plasticity reflect altered M1 excitability as tested by SICI, intracortical facilitation (ICF), short-interval intracortical facilitation (SICF), and input/output curves.[14, 15] Given that TBS elicits different abnormal responses in Parkinson's disease (PD) and progressive supranuclear palsy (PSP),[16-20] information on M1 plasticity in MSA might help to understand pathophysiological differences in PD and atypical parkinsonian syndromes.
We examined M1 LTP/LTD-like plasticity with TBS in MSA-P and MSA-C. We also investigated whether in MSA TBS responses reflect changes in M1 excitability as tested by SICI, ICF, SICF, and input/output curves.
Patients and Methods
We recruited 20 MSA patients (10 men; mean age ± standard deviation [SD]: 62 ± 9.8 years; range, 44–77) and 20 age-matched healthy subjects (HS; 10 men; mean age ± SD: 58.6 ± 11.5 years; range, 36–81). All participants were right-handed. Patients' clinical features are summarized in Table 1. Probable MSA was diagnosed using the 2010 Consensus Criteria. Thirteen patients had the MSA-P variant (6 men; mean age ± SD: 61 ± 11 years; range, 44–75) and 7 had MSA-C (3 men; mean age ± SD: 63.6 ± 8.06 years; range, 54–77). Patients were recruited from the movement disorders clinic at the Department of Neurology and Psychiatry, Sapienza University of Rome (Rome, Italy). Twelve patients continued their usual dopaminergic treatment and 8 took no drugs during the study (Table 1). Motor signs were scored using the Unified Multiple System Atrophy Rating Scale (UMSARS) motor section and the H & Y scale. Cognitive function was evaluated using the Mini–Mental State Evaluation and Frontal Assessment Battery (FAB). Depression was assessed using the Hamilton Depression Rating Scale (HAM-D). Subjects gave their informed consent, and the study was approved by the institutional review board and conformed with the Declaration of Helsinki.
|Case||MSA Subtype||Age(Years)||Gender||Disease Duration(Years)||UMSARS||MMSE||FAB||HAM-D||H & YStage||LEDD (mg)||Experiment|
Subjects were asked to relax and keep their eyes open. Single-pulse TMS was delivered through a monophasic Magstim 200 stimulator (The Magstim Company Ltd, Whitland, UK) connected to a figure-of-eight coil placed over the left M1 in the optimal position for eliciting MEPs in the right first dorsal interosseous (FDI) muscle. Resting motor threshold (RMT) was calculated according to standardized methods. TMS consisted of 20 single pulses delivered at the intensity able to evoke 1-mV MEPs. The same intensity was used for testing MEPs throughout the experiment.
TBS was delivered through a high-frequency biphasic magnetic stimulator (Magstim SuperRapid; Magstim) connected to a figure-of-eight coil placed over the left M1. Active motor threshold (AMT) was calculated according to standardized methods. iTBS sessions consisted in bursts of three pulses at 50 Hz, repeated at 200-ms intervals, delivered in short trains lasting 2 seconds, with an 8-second pause between consecutive trains (600 pulses). cTBS sessions consisted in bursts repeated in a continuous train lasting 40 seconds (600 pulses). Stimulation intensity for TBS was set at 80% of AMT.
To test SICI, ICF, and SICF, we used two monophasic magnetic stimulators, connected by a Y cable to a figure-of-eight coil, placed over the left M1 for evoking MEPs in the contralateral FDI muscle. According to standardized techniques,[14, 15] we elicited SICI with paired pulses at 3 ms and ICF at 10-ms interstimulus intervals. We randomly delivered test pulses at the intensity able to elicit 1-mV MEPs and conditioning pulses at three intensities (70%, 80%, and 90% of AMT; 20 paired pulses for each intensity, 60 paired pulses in total). To elicit SICF, we delivered 20 paired pulses with test pulses at RMT intensity and conditioning pulses at the intensity able to elicit 1-mV MEPs. The 3-ms interstimulus interval for SICF was used to test the possible interference between SICF and SICI as previously reported in HS and in PD.[26, 27] We compared MEPs elicited with the SICI, ICF, and SICF protocols with those elicited by single pulses at the intensity able to elicit 1-mV MEPs.
The input/output curves were measured according to standardized techniques. We randomly delivered 20 single pulses at 100%, 110%, 120%, 130%, 140%, and 150% of RMT.
Recording Techniques and Measurements
Electromyographic activity from the right FDI muscle was recorded using a pair of silver chloride surface electrodes. Electromyographic raw signals were amplified by a Digitimer D360 amplifier (Digitimer Ltd, Welwyn Garden City, UK), digitized at 5 kHz (CED 1401 laboratory interface; Cambridge Electronic Design, Cambridge, UK), and stored on a laboratory computer for off-line analysis with a dedicated software (Signal software; Cambridge Electronic Design). We rejected trials with electromyographic activity greater than 50 µV in a 500-ms time window preceding MEPs. We measured and averaged MEP latency and amplitude.
Experiment 1: TBS-Induced Changes in MEPs in HS and MSA Patients
We used a conditioning-test design. Conditioning stimulation consisted of iTBS or cTBS, given in separate sessions, with an intersession interval of at least 7 days. Twenty test MEPs were recorded before (T0) and 5 (T1), 15 (T2), and 30 minutes (T3) after TBS. The whole group of patients and HS participated (Table 1).
Experiment 2: Intracortical Excitability and Responses to TBS in HS and MSA Patients
Experiment 2 was designed to clarify whether, in MSA, changes in MEPs after TBS reflect SICI, ICF, SICF, and input/output curves, before conditioning. We tested SICI, ICF, SICF, and input/output curves in 10 patients (5 MSA-P: 1 male; mean age ± SD: 59.4 ± 9.4 years; 5 MSA-C: 3 male; mean age ± SD: 65.4 ± 7.7 years) and 10 HS (6 male; mean age ± SD: 62.3 ± 6.5 years) before iTBS or cTBS given in separate sessions, with an intersession interval of at least 7 days (Table 1).
An unpaired Student's t test was used to compare RMTs and AMTs, MEP latency, and stimulus intensity used for evoking MEPs and for conditioning TBS in HS and patients. To test TBS-induced changes in MEPs in HS and patients, we used separate between-group analysis of variance (ANOVA) with “group” and “time” as main factors of analysis. Between-group ANOVA was also used to compare MEPs, before and after TBS, in MSA-P and MSA-C and in patients with or without dopaminergic therapy. Separate repeated-measures ANOVAs were used to compare SICI, ICF, SICF, and input/output curves before iTBS and cTBS in HS and patients. Finally, we used separate between-group ANOVAs with factors group and “condition” to test SICI, ICF, SICF, and input/output curves in HS and patients and to compare these variables in MSA-P and MSA-C and in patients with or without dopaminergic therapy. Tukey's honestly significant difference test was used for post-hoc analysis. Mann-Whitney's U test was used to compare patients' clinical features (UMSARS, H & Y, disease duration, levodopa [L-dopa] equivalent daily dose [LEDD]), in MSA-P and MSA-C, and in patients with or without dopaminergic therapy. Spearman's rank-correlation test was used to assess correlation between patients' clinical features and TBS-induced changes in MEPs, SICI, ICF, SICF, and the input/output curves. P values <0.05 were considered significant. All values are expressed as mean ± standard error (SE).
Experiment 1: TBS-Induced Changes in MEPs in HS and MSA Patients
The unpaired t test showed comparable RMTs and AMTs, MEP latency, intensity for eliciting MEPs at T0, and conditioning TBS in HS and patients (P > 0.05 for all comparisons; Table 2).
|Subjects||RMT||AMT||1 mV||TBS%||RMT||AMT||1 mV||TBS%|
|MSA||36.8 ± 6.5||46.6 ± 8.1||48.7 ± 10||37.2 ± 6.0||37.6 ± 6.5||46.8 ± 7.6||49.1 ± 9.8||37.7 ± 5.9|
|HS||40.7 ± 3.8||48.5 ± 11.3||48.1 ± 6.4||35.2 ± 8.4||39.5 ± 5.9||45.1 ± 3.8||49.1 ± 7.5||35.7 ± 2.9|
TBS-induced changes in MEPs differed significantly in patients and HS, as shown by a significant interaction between factors group and time in the iTBS (F3.114 = 10.1; P < 0.01) and cTBS (F3.114 = 6.74; P < 0.01) sessions. Despite similar MEPs at T0 (P > 0.05 for all comparisons), after conditioning iTBS and cTBS, MEPs differed in patients and HS at T1, T2, and T3 (P < 0.05 for all comparisons). In HS, the effect of factor time was significant after iTBS (F3.57 = 12.9; P < 0.01) and cTBS (F3.57 = 17.58; P < 0.01); MEPs increased after iTBS and decreased after cTBS at T1, T2, and T3 (P < 0.05 for all comparisons). By contrast, in MSA, the factor, time, had a nonsignificant effect in iTBS (F3.57 = 3.57; P = 0.38) and cTBS (F3.57 = 0.48; P = 0.7) sessions (Fig. 1A,B).
When we compared the response to TBS in MSA-P and MSA-C, between-group ANOVA showed that the factor, group, had a nonsignificant effect in the iTBS (F1.18 = 0.02; P = 0.89) and cTBS (F1.18 = 0.002; P = 0.96) sessions. In MSA-P and MSA-C, iTBS and cTBS left MEPs unchanged at all time points (Fig. 1C,D).
Finally, when we tested the effect of dopaminergic therapy in MSA-P and MSA-C, between-group ANOVA again showed a nonsignificant effect of factors group and time (data not shown).
Experiment 2: Intracortical Excitability and Responses to TBS in HS and MSA Patients
Repeated-measures ANOVAs showed the nonsignificant effect of the factor, “experimental session,” demonstrating that SICI and ICF, at all stimulation intensities, SICF, and input/output curves were similar before iTBS and cTBS in HS and patients. Therefore, we averaged SICI, ICF, SICF, and input/output curves recorded before iTBS and cTBS in HS and patients for the subsequent analyses.
Between-group ANOVA comparing SICI in patients and HS showed a significant interaction between factors group and condition (F3.54 = 10.9; P < 0.01). Despite similar MEPs in the two groups (P = 0.13), SICI-induced MEPs differed in HS and patients at all stimulation intensities (P < 0.05 for all comparisons). In healthy subjects, the SICI protocol reduced MEPs, as shown by the significant effect of the factor, condition (F3.27 = 17.35; P < 0.01), and did so at all stimulation intensities (P < 0.01 for all comparisons), whereas, in patients, it left MEPs unchanged (F3.27 = 0.37; P = 0.78) (Figure 2A). When we compared SICI in MSA-P and MSA-C, and in patients with or without dopaminergic therapy, between-group ANOVA showed a nonsignificant effect of the factor, group, demonstrating comparable SICI under all experimental conditions (Fig. 2B).
When we compared ICF in patients and healthy subjects, between-group ANOVA showed a nonsignificant effect of the factor, group (F1.18 = 2.27; P = 0.15). MEPs elicited by test pulses and ICF protocol at all stimulation intensities were comparable in HS and patients (P > 0.05 for all comparisons) (Figure 2A). ICF was also similar in MSA-P and MSA-C as well as in patients with or without dopaminergic therapy (Figure 2B).
Between-group ANOVA comparing SICF in patients and HS showed a nonsignificant effect of the factor, group (F1.18 = 0.37; P = 0.55) (Figure 2A). SICF was comparable also in MSA-P and MSA-C as well as in patients with or without dopaminergic therapy (Figure 2B).
When we compared the input/output curves in HS and patients, between-group ANOVA showed comparable changes in MEPs, as demonstrated by the nonsignificant effect of the factor, group (F1.18 = 1.81; P = 0.2). As the TMS intensity increased, MEPs also increased similarly in HS and patients, as shown by the significant factor, condition, in both groups (F4.36 = 42; P < 0.01; and F4.36 = 39.05; P < 0.01; Fig. 3A). When we compared the input/output curves in MSA-P and MSA-C (Fig. 3B), and in patients with or without dopaminergic therapy, the effect of the factor, group, was nonsignificant.
Finally, between-group ANOVA comparing MEPs after TBS in the 10 HS and 10 patients showed a significant interaction between factors group and time in the iTBS (F3.54 = 9.06; P < 0.01) and cTBS (F3.54 = 10.17; P < 0.01) sessions. Despite similar MEPs at T0 (P > 0.05 for all comparisons), after conditioning, MEPs differed in patients and HS at T1, T2, and T3 (P < 0.05 for all comparisons).
Mann-Whitney's U test showed comparable UMSARS, H & Y, disease duration, and LEDDs in MSA-P and MSA-C as well as in patients with or without dopaminergic therapy (P > 0.05 for all comparisons).
Spearman's test found no correlation between patients' clinical and neurophysiological features.
In this study designed to examine M1 plasticity with TBS in MSA, we found that iTBS increased and cTBS decreased MEPs in HS, whereas in MSA, iTBS and cTBS left MEPs unchanged. When we compared the TBS-induced after effects in MSA-P and MSA-C, we found no difference in the two subgroups. These findings support impaired M1 plasticity in MSA. Finally, in MSA-P and MSA-C, responses to TBS did not reflect M1 excitability as tested by SICI, ICF, SICF, and input/output curves.
We excluded several confoundings possibly affecting responses to TBS. All experimental sessions took place at comparable daytime and at least 1 week elapsed between sessions excluding cumulative effects resulting from repeated TBS. Motor thresholds and intensities for eliciting MEPs were similar in HS and patients. Because we checked patients' electromyographic activity throughout the experiments and none of the recordings showed muscle activity immediately before, during, or after TBS, we can also exclude the possibility that in MSA the altered response to TBS reflected interference between TBS and muscle activity.[29, 30] Given that TMS elicited similar latency MEPs in HS and MSA, and none of the patients had pyramidal signs, we exclude possible confounding from direct corticospinal system impairment.
In MSA, we found reduced SICI and normal ICF, SICF, and input/output curves,[4-8] in treated and untreated patients. These findings might help in differentiating pathophysiological mechanisms underlying MSA and PD, because in PD reduced SICI also reflects increased SICF, and both variables are sensible to L-dopa. In MSA, M1 excitability changes might have influenced the subsequent response to TBS through gating/antigating processes as previously reported in HS. Because the excitability variables did not correlate with the response to TBS, we consider unlikely the occurrence of gating/antigating processes. However, given the relatively small cohort of patients involved in experiment 2, we cannot fully exclude correlation in MSA between excitability changes and response to TBS.
Several mechanisms might explain how TBS elicits altered MEP responses in MSA. TBS resembles experimental protocols in animals mimicking neuronal firing patterns and able to induce LTP or LTD.[12, 13] Hence, the iTBS-induced after effects are thought to reflect homotopic LTP-like plasticity, whereas the cTBS-induced after effects depend on homotopic LTD-like plasticity in M1 interneurons.[12, 13] Accordingly, the reduced iTBS-induced changes in MEPs suggests altered LTP-like plasticity in M1, and the reduced responses to cTBS suggest altered M1 LTD-like plasticity in MSA. Reduced LTP/LTD-like plasticity in MSA might reflect abnormal motor inputs from basal ganglia to cortical motor areas. In MSA, altered M1 plasticity might also reflect intrinsic M1 abnormalities. Histopathological studies demonstrated loss of small- to medium-sized pyramidal neurons in cortical layers V/VI in M1, premotor cortex, and supplementary motor area.[1, 32] As in HS, M1 plasticity also reflects functional connectivity between primary and nonprimary motor areas,[33, 34] and the abnormal M1 plasticity in MSA might also arise from altered connectivity between nonprimary motor areas and M1, in line with neuroimaging studies.
The reduced LTP/LTD-like plasticity agrees with the previously reported reduced responses to short 5-Hz rTMS trains. In HS, during 5-Hz rTMS, MEPs elicited by each stimulus progressively increase reflecting short-term plasticity. Although under specific experimental conditions short-term plasticity interacts with mechanisms underlying TBS-induced after effects, the mechanisms activated by 5-Hz rTMS differ from those responsible for TBS-induced after effects. TBS implies rhythmic gamma-bursts at intensities below the threshold for activating corticospinal neurons. We conclude that short- and long-term plasticity are both abnormal in MSA.
The reduced TBS-induced after effects in MSA resemble those previously reported in de novo and chronically treated PD patients.[17-20] By contrast, the reduced response to TBS in MSA differs from that observed in PSP in whom iTBS elicits exaggerated responses, whereas the cTBS-induced after effects shift from inhibition to facilitation. Neuropathological studies in PD and MSA have detected alpha-synuclein (α-SYN) deposition in several cortical and subcortical brain regions, including M1.1,32 Experimental studies suggest that α-SYN plays a crucial role in neurotransmission and synaptic plasticity. Differently, the pathophysiological hallmark in PSP is tau protein deposition forming neurofibrillary tangles reflecting neurodegenerative pathways different from those responsible for α-SYN deposition. Compared with PD and MSA, in PSP, exaggerated TBS responses might reflect a more prominent cortical degeneration, including loss of M1 inhibitory interneurons.[16, 40, 41]
TBS disclosed no M1 plasticity differences in MSA-P and MSA-C nor did clinical variables correlate with neurophysiological measures. Compared with MSA-P, MSA-C implies disrupted cerebellar-to-thalamus connectivity, possibly eliciting additive effects on M1 plasticity.[42, 43] However, the similar responses to TBS in MSA-P and MSA-C suggest that reduced plasticity reflects common pathophysiological M1 abnormalities, probably secondary to altered thalamocortical inputs.
Although in patients assuming dopaminergic therapy we did not test TBS-induced responses after a drug withdrawal, the observation that TBS-induced plasticity was similar in patients receiving or not receiving dopaminergic medications suggests weak influence of L-dopa on patients' neurophysiological and clinical features.[2, 3]
Experimental studies in nonhuman primates have demonstrated that M1 encodes important movement parameters, including velocity, amplitude, and direction of upper-limb muscle movements. In addition, LTP/LTD-like plasticity in M1 is known to play a role in motor execution and learning.[9, 31] Therefore, we suggest that in MSA, reduced LTP/LTD-like plasticity in M1 interneurons might degrade physiological processes responsible for movement-variable processing and lead to altered motor execution and motor learning. In line with this hypothesis, reduced M1 plasticity might contribute to pathophysiology of bradykinesia and other parkinsonian symptoms in MSA.
In conclusion, in MSA, we found reduced M1 LTP/LTD-like plasticity reflecting abnormalities in the corticobasal ganglia-thalamo-cortical circuit. Our findings might help to clarify the different pathophysiological mechanisms involved in atypical parkinsonian syndromes.
(1) Research Project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript Preparation: A. Writing of the First Draft; B. Review and Critique.
A.S.: 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B
L.M.: 3A, 3B
F.D.S.: 3A, 3B
C.C.: 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B
A.B.: 1A, 1B, 1C, 2A, 2B, 2C, 3A, 3B
Nothing to report.
- 25Hamilton Depression Scale. Berlin: Beltz; 1976., .