Investigating the effects of dopamine on short‐ and long‐latency afferent inhibition

Short‐ and long‐latency afferent inhibition (SAI and LAI respectively) are phenomenon whereby the motor evoked potential induced by transcranial magnetic stimulation (TMS) is inhibited by a sensory afferent volley consequent to nerve stimulation. It remains unclear whether dopamine participates in the genesis or modulation of SAI and LAI. The present study aimed to determine if SAI and LAI are modulated by levodopa (l‐DOPA). In this placebo‐controlled, double‐anonymized study Apo‐Levocarb (100 mg l‐DOPA in combination with 25 mg carbidopa) and a placebo were administered to 32 adult males (mean age 24 ± 3 years) in two separate sessions. SAI and LAI were evoked by stimulating the median nerve and delivering single‐pulse TMS over the motor hotspot corresponding to the first dorsal interosseous muscle of the right hand. SAI and LAI were quantified before and 1 h following ingestion of drug or placebo corresponding to the peak plasma concentration of Apo‐Levocarb. The results indicate that Apo‐Levocarb increases SAI and does not significantly alter LAI. These findings support literature demonstrating increased SAI following exogenous dopamine administration in neurodegenerative disorders.


Introduction
Transcranial magnetic stimulation (TMS) is used to evoke neural circuits resulting in short-latency afferent inhibition (SAI) and long-latency afferent inhibition (LAI) (Chen et al., 1999;Tokimura et al., 2000).SAI and LAI are evoked by delivering a TMS pulse preceded by peripheral nerve stimulation which leads to reduced corticospinal output (Chen et al., 1999;Tokimura et al., 2000).SAI and LAI are evoked when the interstimulus interval between the TMS pulse and nerve stimulation is 20-25 ms (Tokimura et al., 2000) and 200-1000 ms (Chen et al., 1999), respectively.
Pharmacological studies have advanced the understanding of the mechanisms that mediate SAI and LAI.SAI is reduced by lorazepam, an agonist of the GABA A receptor, but is not modulated by baclofen, an agonist of the GABA B receptor (Turco, El-Sayes, Locke et al., 2018).SAI is also modulated by acetylcholine such that it is reduced in the presence of scopolamine, an anti-muscarinic which blocks acetylcholine receptors (Di Lazzaro, Oliviero, Profice et al., 2000) and is reduced or absent in individuals living with Alzheimer's disease (associated with lower acetylcholine).The neural mechanism of LAI is less well understood.LAI is reduced by lorazepam indicating it is modulated by GABA A receptor activity (Turco, El-Sayes, Locke et al., 2018).Both SAI and LAI are reduced in Parkinson's disease (PD) in the presence of dopamine (DA) replacement pharmacotherapy (Dubbioso et al., 2019;Rochester et al., 2014;Sailer et al., 2003) and this may suggest that DA participates in the genesis or modulation of afferent inhibition (Martin-Rodriguez & Mir, 2020;Sailer et al., 2003).
Research has focused on the interplay between DA and acetylcholine in the context of Alzheimer's disease.For example, in Alzheimer's disease, SAI increases by ∼15% following 125 mg levodopa (l-DOPA) (Martorana et al., 2009;Nardone et al., 2014) and by ∼35% following delivery of rotigotine, a DA D2 agonist (Martorana et al., 2013).These effects are suggested to be mediated by dopaminergic modulation of cholinergic release.Further, DA modulates GABA A receptor activity.Specifically, in the striatum, GABA A receptor activity is increased by DA while extrasynaptic GABA A receptors are inhibited by DA (Hoerbelt et al., 2015).Importantly, the mechanism by which l-DOPA effects GABA A receptor activity remains unknown.
To date, there is no placebo-controlled study investigating the effects of DA on SAI and LAI in a population of young adults.In this placebo-controlled, double-anonymized study, we ask whether exogenous DA alters the magnitude of SAI and LAI.Our novel findings indicate that SAI is increased in the presence of l-DOPA while LAI is unaffected.

Ethical approval
The present study was approved by the Hamilton Integrated Research Ethics Board (#7716).The research conformed to the standards set by the Declaration of Helsinki, except for registration in a database.After explanation of the study protocol, and the usual action and side-effects of Apo-Levocarb, all participants provided their written informed consent prior to participation.

Participants
Thirty-two healthy, right-handed males (24 ± 3 years) participated in two sessions at least 1 week apart (18 ± 11 days).All participants were screened for contraindications to TMS and l-DOPA, and right-handedness was confirmed by a modified handedness questionnaire (Oldfield, 1971).To ensure no potential interaction with l-DOPA, participants were excluded if they were taking any prescribed or non-prescribed medications during their participation in the study.This included acetaminophen, ibuprofen, antibiotics, anti-histamines, alcohol, nicotine, cannabis or street drugs.Further, exclusions also involved questions specifically related to TMS and included the presence of head injury or head surgery, metal implants, history of neurological disease or psychiatric illness, blood relatives with a history of seizures, headaches or migraines, and sleep deprivation.All pharmaceuticals and the randomization schedule were prepared by McMaster University Medical Centre pharmacy.All experiments and analyses were performed by experimenters who were blinded to the treatment allocation (S.F., F.A., C.D., K.R.).

Electromyography
Surface EMG electrodes (9 mm Ag-AgCl) were placed on the skin overlying the first dorsal interosseous (FDI) muscle of the right hand.A ground electrode was placed at the styloid process of the wrist.EMG recordings were amplified 1000× (Model 2024F; Intronix Technologies Corporation, Bolton, ON, Canada) and band-pass filtered between 20 Hz and 2.5 kHz.Data were digitized at 5 kHz using an analog-to-digital converter (Power 1401; Cambridge Electronics Design, Cambridge, UK) and analysed using commercial software (Signal, version 6.02; Cambridge Electronics Design).

Peripheral nerve stimulation
Sensory nerve action potentials (SNAPs) were recorded from the median nerve (MN) at the elbow.SNAPs were used to determine the size of the afferent volley evoked by peripheral nerve stimulation during the collection of SAI and LAI (Bailey et al., 2016).SNAPs were elicited using a bar electrode placed over the MN at the wrist with the cathode proximal using a constant current stimulator (Digitimer DS7AH, Welwyn Garden City, UK; square wave pulses of 0.5 ms, 2 Hz).To determine SNAP max , the starting nerve stimulation intensity was set to 5 mA.Twenty stimuli per block were delivered, and the peak-to-peak amplitude of the averaged SNAP was quantified.This procedure continued in stepwise increments of 2 mA until SNAP max was achieved.SNAP max was defined as the intensity (in mA) at which the SNAP ceased to increase greater than 10% for three consecutive blocks (Bailey et al., 2016).During the subsequent collection of SAI and LAI, peripheral nerve stimulation was set to the stimulator intensity required to evoke an averaged peak-to-peak amplitude of 50% of SNAP max (50%SNAP max ).

Electroencephalography
Somatosensory-evoked potentials (SEPs) were recorded using EEG from an electrode positioned at C3 and referenced to Fz (international 10-20 system) and a ground electrode was placed over the clavicle.A bar electrode with the anode (cathode proximal) was used to stimulate the median nerve of the wrist.In total, 500 stimuli (20 μs square wave pulses; 3 Hz) were delivered at 50%SNAP max of the median nerve (Digitimer DS7AH).Signals were averaged over the 500 epochs to identify the latency of the N20 component of the SEP.The N20 was used to determine the interstimulus interval between the nerve stimulation and TMS for subsequent acquisition of SAI.Using the N20 latency, N20 + 2 ms and N20 + 4 ms were used to elicit SAI (Turco, El-Sayes, Locke et al., 2018;Turco, El-Sayes, Savoie et al., 2018).

Transcranial magnetic stimulation
TMS was performed with a 50 mm figure-of-eight branding coil connected to a Magstim 200 2 stimulator J Physiol 602.10 (Magstim, Whitland, UK).The coil was position over the left motor cortex at the location that elicited a motor-evoked potential (MEP) in the right FDI muscle (i.e.motor hotspot).The coil was oriented at a 45 degree angle from the sagittal plane to induce a posterior-anterior current in the cortex.The location and orientation of the coil were registered digitally using Brainsight Neuronavigation (Rogue Research, Montreal, QC, Canada).Resting motor threshold (RMT) was taken as a measure of baseline cortical excitability (Siebner & Rothwell, 2003) and was determined using TMS_MTAT_2.0 freeware (http://clinicalresearcher.org/software.htm).The stimulus intensity was set to 37% of the maximum stimulator output (MSO) and 20 TMS pulses were delivered over the FDI hotspot to accurately determine the RMT (Ah Sen et al., 2017;Ramdeo et al., 2023;Rehsi et al., 2023;Turco, El-Sayes, Locke et al., 2018).

Experimental design
Within a session, participants were either administered Apo-Levocarb (100 mg l-DOPA, a dopamine precursor plus 25 mg carbidopa) or Placebo.The order of interventions delivered were pseudo-randomized across all participants.Dependent measures were acquired prior to drug ingestion (T0), and 1 h after drug ingestion (T1) based on the peak plasma concentration of Apo-Levocarb (Dingemanse et al., 1995;Djaldetti et al., 2003;Martorana et al., 2009).During the 1 h period between T0 and T1, participants remained in the lab and were not permitted to eat or sleep during this time (Fig. 1).They were permitted to drink water, read and use the restroom if needed.

Statistical methods
EMG trials were discarded if the root mean squared amplitude of the EMG signal, within 100 ms before the TMS artifact, exceeded 5 μV.Further, all trials were additionally visually inspected for the presence of muscle contraction in the 100 ms window.Participants were removed from subsequent analyses if the number of trials with movement artifacts exceeded 20% of the total trials.All initial data analyses were performed by experimenters blinded to the allocation of drug (placebo or l-DOPA) and unblinding occurred after all individual data were analysed.SAI and LAI were calculated as the ratio of the mean conditioned MEP (CS-TS) divided by the mean unconditioned MEP (TS) (SAI or LAI = CS − TS/TS × 100%).Normality for all variables was assessed using a Shapiro-Wilks test.If data were not normally distributed, a square root transformation was applied.Data were subsequently ranked if transformation was not effective.Paired t tests were used to assess changes in RMT before and after drug administration.To confirm the presence of inhibition at T0 (i.e.significant inhibition) for each measure (SAI N20+2 , SAI N20+4 and LAI) six Bonferroni-corrected t tests compared the CS-TS vs. TS.A two-way ANOVA was performed on the unconditioned MEP data using within-subjects factors of TIME (two levels: T0, T1) and DRUG CONDITION (two levels: l-DOPA, Placebo).
For SAI, a two-way ANOVA was performed separately for each intervention using within-subjects factors of TIME (two levels: T0, T1) and ISI (two levels: SAI N20+2 , SAI N20+4 ).Importantly, the focus of this research is whether the intervention has an effect over time as performed elsewhere (Di Lazzaro, Oliviero, Saturno et al., 2005;Di Lazzaro, Pilato, Dileone et al., 2005;Turco, El-Sayes, Locke et al., 2018).Next, we determined which ISI (SAI N20+2 , SAI N20+4 ) resulted in the greatest depth of inhibition at T0 for each of the l-DOPA and Placebo visits.Subsequently, the ISI which produced  Average data from 31 participants.%MSO (maximum stimulator output), SAI and LAI are reported in mV.N20 represents the time from nerve stimulation to the onset of the first negative cortical potential corresponding to the N20.Greatest inhibition represents the data from the ISI (N20 + 2 ms or N20 + 4 ms) which produced the greatest inhibition in SAI at T0 in each of the visits.
the greatest inhibition at T0 was used as the ISI for that visit.Using these data, a two-way ANOVA using within-subjects factors of TIME (two levels: T0, T1) and DRUG CONDITION (two levels: l-DOPA, Placebo) was performed.Post hoc analyses were performed using Fisher's least-significant-difference test.
For LAI data, comparisons between T0 and T1 were performed separately for l-DOPA and Placebo intervention through two paired t tests.Next, a two-way ANOVA using within-subjects factors of TIME (two levels: T0, T1) and DRUG CONDITION (two levels: l-DOPA, Placebo) was performed.Last, a paired-samples t test was used to compare l-DOPA and Placebo at T0.
Reliability statistics were performed on the placebo data to assess the measurement error that occurred from repeated testing of an individual.Violation of normality requires a transformation on the dataset, but data were not transformed for this reliability analysis as it would cause a change in the ratio scale.Absolute reliability was determined for the placebo intervention for each dependent measure using the standard error of measurement (SEMeas) values (SEMeas = MeanSquared error ).Furthermore, the SEMeas was used to determine the smallest detectable change (SDC individual ) (SDC individual = SEMeas × 1.96 × √ 2) and SDC group (SDC group = SDCindiv √ n ) where n is the sample size.

Results
The experimental procedures were well tolerated by all participants with no adverse events.One participant was removed from subsequent analyses as the number of trials with movement artifacts exceeded 20% of the total trials obtained.Therefore, all analyses were conducted on a sample size of 31 participants.For the 31 participants, 2.9% of all SAI and 1.9% of all LAI trials were removed due to excess background muscle activity.Of the 31 participants included in the analysis, 17 received l-DOPA as the first intervention while 14 received Placebo first.
Table 1 shows the group-averaged means and standard deviations of all measures in the study.RMT did not differ between days at T0 (paired-samples t test, T 30 = −1.94,P = 0.062) and was not modified by l-DOPA (paired-samples t test, T 30 = −0.18,P = 0.859) or Placebo (paired-samples t test, T 30 = 0.28, P = 0.781).This is consistent with previous findings that showed no effect of l-DOPA on RMT (Priori et al., 1994;Ziemann et al., 1996aZiemann et al., , 1997)).These findings are consistent with the idea that RMT is dependent on changes in neuronal membrane excitability (Chen et al., 1997;Ziemann et al., 1997).Significant inhibition (CS-TS vs. TS) was present at T0 for SAI N20+2 , SAI N20+4 and LAI, as revealed by the main effect of ISI (all P < 0.001).For the unconditioned MEP, a main effect of TIME (F 1,30 = 9.117, P = 0.005) between T0 and T1 was observed using the unconditioned MEP data.
Group-level reliability expressed as %SEMeas was >10% for both interventions demonstrating a large amount of measurement error (Table 3).For SAI, the SDC individual indicates that a minimum change of 51% is needed to be considered a physiological change at an individual level.Our estimations of SDC group based on our sample size is 9.3% which is lower than the 13% increase in SAI observed in the l-DOPA group.For the Placebo group, although not statistically significant, the 4% decrease in SAI falls within measurement error.

LAI
The results of the analysis of LAI are presented in Table 2.For LAI in the l-DOPA intervention there was no significant difference between T0 and T1 (paired-samples t test, T 30 = 1.27,P = 0.213).For LAI in the placebo intervention there was no significant difference between T0 and T1 (paired-samples t test, T 30 = −0.56,P = 0.578).The two-way ANOVA revealed no effect of TIME (F 1,30 = 0.47, P = 0.497) or DRUG CONDITION (F 1,30 < 0.001, P = 0.994) (Fig. 4).Further, there was no difference between l-DOPA and Placebo at T0 (paired-samples t test, T 30 = −0.57,P = 0.572).Table 3 displays the reliability statistics for LAI.%SEMeas shows a large amount of measurement error (%SEMeas > 10%).A minimum change of 12.5% is required to be considered real physiological change at the group level.In the current data, a 14% and 6% change were observed in LAI following l-DOPA and Placebo, respectively.

Discussion
The present study examined the pharmacological influence of l-DOPA on SAI and LAI.SAI was increased following administration of l-DOPA while LAI was not significantly altered.We discuss these findings and their putative neural mechanisms below.
In the present study, we observed an increase of 13% in SAI following l-DOPA.The results we obtained in this study are similar to those reported in Alzheimer's disease where SAI is increased by ∼15-20% following 125 mg l-DOPA (Martini et al., 2023;Martorana et al., 2009;Nardone et al., 2014) and by ∼35% following rotigotine, a D2 agonist (Martorana et al., 2013).Absolute reliability  statistics indicated an SDC group of 9%.Therefore, the change in SAI following l-DOPA was 13% indicating a real change of ∼4%.As such, the change observed in SAI following l-DOPA is not due to measurement error and reflects a real physiological change.SAI can be reliably measured at the group level to track change following an intervention.l-DOPA may increase SAI via increases in GABAergic inhibition.In animal models, iontophoretic injection of DA inhibits action potentials in monkey neocortex (Krnjević & Phillis, 1963) and suppresses spontaneous activity in cortical cells through increased GABA transmission in monkeys (Rolls et al., 1984) and rats (Pirot et al., 1992).In rats, the application of D1 and D2 receptor antagonists increases SEP amplitudes, suggesting DA reduces excitability (Hosp, Hertler et al., 2011).In mice, DA increases GABA activity of cortical neurons, such that administration of quinpirole, a D2 agonist, increases the GABAergic activity from parvalbumin-expressing interneurons to pyramidal neurons in motor cortex (Cousineau et al., 2020).Indeed, DA synapses appear to be on the spines and the distal dendrites of pyramidal neo-cortical neurons, and these synapses suggest DA has an important role in modulating the excitation to these cells (Smiley et al., 1992).D2 receptors are present on pyramidal and GABAergic interneurons in the primary motor cortex (Lidow et al., 1989).Thus, work in animal models demonstrates that DA increases GABAergic activity.
Our previous study revealed decreases in SAI following ingestion of lorazepam, a GABA A receptor agonist that binds to the α subunits (Di Lazzaro, Oliviero, Saturno et al., 2005;Turco, El-Sayes, Locke et al., 2018).In the case of lorazepam, effects are directly related to binding at GABA A receptors, which for SAI leads to a reduction in inhibition (Turco, El-Sayes, Locke et al., 2018).However, it is unclear which α subunits are targeted by lorazepam (Di Lazzaro, Oliviero, Saturno et al., 2005).In the striatum, DA has differential effects on GABA activity; DA increases GABA receptor activity at GABA A receptors that have the α subunit and DA inhibits tonic GABA A currents at extrasynaptic GABA A receptors (Hoerbelt et al., 2015).Further, the mechanisms by which l-DOPA increases GABAergic transmission in cortex are unresolved, and it is unclear which GABA A subunits may be involved in this process.Exploring the interaction between SICI and SAI as performed elsewhere (Alle et al., 2009), in the presence versus absence of l-DOPA, may be one approach to investigate the role of GABA in the genesis or modulation of SAI.SAI was unchanged in studies that tested elderly controls following administration of l-DOPA (Martini et al., 2023;Martorana et al., 2009;Nardone et al., 2014) or a D2 agonist (Martorana et al., 2013).One possibility for the discrepant effects of l-DOPA between our work and others may relate to the depth of SAI such that elderly controls demonstrate ∼40-55% inhibition (Martorana et al., 2009(Martorana et al., , 2013) ) while the participants tested herein demonstrate ∼35-45% inhibition, consistent with SAI obtained in young adults (Bailey et al., 2016;Harasym et al., 2020;Toepp et al., 2019;Tsang et al., 2014Tsang et al., , 2015;;Turco et al., 2017;Turco et al., 2019;Turco, El-Sayes, Locke et al., 2018).Alternatively, the discrepant findings in healthy controls may relate to the density of D2 receptors which decreases by ∼25-50% with age (Morgan, 1987) thereby limiting the impact of the exogenous l-DOPA in older adults tested elsewhere (Martini et al., 2023;Martorana et al., 2009;Nardone et al., 2014).This is the first study to examine the effects of l-DOPA on SAI and LAI in healthy young adults.Alternatively, previous studies included small sample sizes (8-15 controls) and may not have been adequately powered to detect change in healthy controls, contributing to the discrepant findings between studies (Martini et al., 2023;Martorana et al., 2009Martorana et al., , 2013;;Nardone et al., 2014).The present study is the largest study examining the effect of l-DOPA on afferent inhibition and was adequately powered based on the effect size observed for l-DOPA on SAI.We have previously demonstrated that adequate sample sizes are required for reliable group-level changes to be observed in SAI and LAI (Turco et al., 2019).Therefore, the lack of consensus on the effect of DA in healthy controls may relate to ageing and/or the sample size tested.
The present study provides further support that SAI and LAI are differentially affected by pharmacological agents.Specifically, SAI is increased by DA while LAI is not.SAI and LAI are both modulated by lorazepam, a GABA A agonist, such that SAI and LAI are reduced following administration of lorazepam (Di Lazzaro, Oliviero, Saturno et al., 2005;Turco, El-Sayes, Locke et al., 2018).SAI is modulated by cholinergic activity (Di Lazzaro, Oliviero, Meglio et al., 2000;Di Lazzaro, Oliviero, Saturno et al., 2005), but the effects of acetylcholine on LAI are unknown.The data from this study support the notion that SAI and LAI arise from distinct neural mechanisms.

Limitations
The effects of l-DOPA on SAI and LAI may depend on the dosage administered.This study investigated a single dosage of Apo-Levocarb of 125 mg chosen to replicate previous studies (Martini et al., 2023;Martorana et al., 2009;Nardone et al., 2014).However, it is possible that the dosage was suboptimal to alter LAI and that greater or lesser amounts would yield a different outcome.Thirugnanasambandam et al. (2011) demonstrated a complex relationship between l-DOPA dosage and effects on paired associative stimulation (PAS) whereby 25 mg abolished PAS, 100 mg prolonged PAS and 200 mg reversed excitability from PAS.Therefore, future studies may examine the relationship between l-DOPA dosage and the effects on afferent inhibition.A further limitation is that the present study included only males to avoid variability of l-DOPA effects related to female hormones.DA levels vary with progesterone levels (Hidalgo-Lopez & Pletzer, 2017) and D2 receptor binding is greater in females (Kaasinen et al., 2001).Future studies should examine l-DOPA effects on SAI/LAI in females.In line with previous research (Martini et al., 2023;Martorana et al., 2009), we did not include a behavioural assessment to examine the cognitive state of the participant following drug administration.Questionnaires or visual analogue scales performed by the participants would have been useful in confirming that individuals were indeed blinded to the identity of the medication ingested.

Conclusion
This study investigated the effects of l-DOPA on afferent inhibition and revealed that SAI is increased in the presence of l-DOPA while LAI is not significantly altered.These findings advance our understanding of the pharmacological basis of afferent inhibition.

Figure 1 .
Figure 1.Study design Timeline for experimental sessions with two time points that correspond to baseline (T0) and the peak plasma concentration of L-DOPA (T1).

Figure 2 .
Figure 2. Effects of L-DOPA and Placebo on SAI (n = 31) SAI [mean (denoted by red +) ± min and max] expressed as a ratio of the conditioned MEP (CS-TS) to the unconditioned MEP (TS) before (T0) and 1 h after L-DOPA or Placebo ingestion (T1).Data represent the ISI (SAI N20+2 , SAI N20+4 ) which produced the greatest inhibition at T0. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 3 .
Figure 3. Individual SAI data averaged across ISI SAI in individual participants before (T0) and after (T1) L-DOPA (above) and Placebo (below) (n = 31).Participants are displayed on the x-axis.Blue indicates participants who show increased inhibition and orange indicates participants who do not show increased inhibition relative to T0.For L-DOPA and Placebo, 20 and 15 individuals show greater inhibition compared to T0, respectively.[Colour figure can be viewed at wileyonlinelibrary.com]

Figure 4 .
Figure 4. Effects of L-DOPA and Placebo on LAI (n = 31) LAI [mean (denoted by a red +) ± min and max] expressed as a ratio of the conditioned MEP (CS-TS) to the unconditioned MEP (TS) before (T0) and 1 h (T1) after ingestion of L-DOPA or Placebo.[Colour figure can be viewed at wileyonlinelibrary.com]