Effect of Serotonin on Paired Associative Stimulation-Induced Plasticity in the Human Motor Cortex

Abstract

INTRODUCTION

Serotonin (5-HT) is a widely distributed neurotransmitter in the brains of animals and humans, affecting various physiological functions such as learning, memory formation, pain perception, mood and the sleep–wakefulness cycle (Bert et al, 2008; Geyer, 1996; Hasbroucq et al, 1997; Jacobs and Fornal, 1997; Meneses, 1999). One important foundation for these effects might be its impact on neuroplasticity. Activation of serotoninergic subreceptors is shown to affect long-term potentiation (LTP) or long-term depression (LTD) in animal slice preparations, depending on subreceptor type, location and frequency of application (Huang and Kandel, 2007; Kojic et al, 1997; Mori et al, 2001).

In the clinical domain, studies have suggested that depression might be a result of altered brain plasticity (Christoffel et al, 2011; Garcia, 2002; Henn and Vollmayr, 2004; Popoli et al, 2002), on which serotonin has a major impact. Distress has been proposed as one of the main factors preceding depression (Caspi et al, 2003), and in animals it inhibits LTP and facilitates LTD induction (Foy et al, 1987; Rocher et al, 2004; Shors et al, 1989; Xu et al, 1997). In accordance, LTD is facilitated in animal models of depression, which was prevented by chronic application of the selective serotonin reuptake inhibitor (SSRI) fluvoxamine (Holderbach et al, 2007). Besides depression, several studies have demonstrated that SSRIs improve motor functions in stroke patients (Chollet et al, 2011; Dam et al, 1996; Pariente et al, 2001) and in healthy individuals (Loubinoux et al, 1999; Loubinoux et al, 2002a; Loubinoux et al, 2002b; Loubinoux et al, 2005). Again, the physiological basis for this effect might be the impact of serotonin on plasticity.

Recently it was shown that motor cortex plasticity in healthy humans induced by transcranial direct current stimulation (tDCS) was affected by single-dose SSRI. Citalopram enhanced facilitatory plasticity induced by anodal tDCS and converted cathodal tDCS-induced inhibitory plasticity into facilitation (Nitsche et al, 2009). tDCS and paired associative stimulation (PAS) are non-invasive brain stimulation techniques inducing changes in cortical excitability that outlast the stimulation duration (Nitsche et al, 2003b; Nitsche and Paulus, 2000, 2001; Stefan et al, 2000; Wolters et al, 2003). These alterations in cortical excitability are NMDA- and Ca²⁺-dependent (Nitsche et al, 2003a; Stefan et al, 2002; Wolters et al, 2003). tDCS induces non-focal plasticity, affecting relatively non-selectively neuronal populations beneath the large stimulation electrodes via subthreshold resting membrane potential modulation (Nitsche et al, 2008; Nitsche et al, 2007; Purpura and McMurtry, 1965). PAS induces focal and synapse-specific plasticity of the respective target neurons. In PAS, a repetitive electric pulse to a peripheral nerve at an intensity that activates somatosensory afferents is combined with transcranial magnetic stimulation (TMS) over the corresponding area of the primary motor cortex. Depending on the interstimulus interval (ISI), synchronous or asynchronous activation of the target group of neurons, which are motor cortex neurons connected with the respective somatosensory afferents, is accomplished, resulting in excitatory or inhibitory after-effects (Stefan et al, 2000). This mechanism of plasticity induction resembles some characteristics of spike timing-dependent plasticity (STDP), which is closely linked to learning and memory processes.

Interestingly, other neuromodulators have discernable effects of tDCS- and PAS-induced plasticity. Specifically, dopamine, acetylcholine, and nicotine have a focusing, or signal-enhancing effect on facilitatory plasticity (Kuo et al, 2007; Kuo et al, 2008; Monte-Silva et al, 2010; Thirugnanasambandam et al, 2012). These substances abolish tDCS-induced non-focal, but enhance PAS-generated focal facilitatory plasticity. This effect might explain the cognition- and behavior-enhancing impact of these substances.

After having explored the impact of serotonin on tDCS-induced plasticity, we were now interested to explore how this modulator affects PAS-generated neuroplastic cortical excitability alterations. We hypothesize that citalopram enhances PAS-induced focal excitatory plasticity and abolishes focal inhibitory plasticity or convert it into excitation, as it was shown for tDCS (Nitsche et al, 2009).

MATERIALS AND METHODS

Experimental Procedures

Each subject participated in four experimental sessions (PAS25 with citalopram or placebo, PAS10 with citalopram or placebo), which were carried out in randomized order and separated by minimum 1 week. A unique sequence of experimental sessions was randomly generated for each subject individually, which did not match any previously generated one for other subjects. The volunteers were seated in a comfortable chair with head and arm rests. First, the hotspot (the position of coil that produced the largest MEPs of the right ADM) was identified by TMS. Then the stimulation intensity was adjusted to elicit single-pulse MEPs with peak-to-peak amplitudes of on average 1 mV and 25 MEPs were recorded for the first baseline determination. After baseline recording, citalopram or placebo medication was administered. At 2 h after intake of medication, a second baseline was recorded to monitor for a possible influence of the drug on cortical excitability (baseline 2), and TMS intensity was adjusted, if necessary (baseline 3). After that procedure, PAS25 or PAS10 was administered and 25 MEPs were recorded immediately after stimulation and at time points of 5, 10, 15, 20, 25, 30, 60, 90 and 120 min after the stimulation PAS. Further TMS measurements were conducted in the evening of the same day (SE), next morning, at ∼0900 hours (NM), next noon, at ∼1200 hours (NN), and next evening, at ∼1800 hours (NE) (Figure 1). To keep the EMG electrodes and TMS coil at the same place for later measurements, their positions were marked with a waterproof pen. Subjects were blinded for both, stimulation and medication conditions; the experimenter was blinded only for the medication condition.

Open in a separate window

Figure 1

Course of the study. In the beginning of each session, before administration of citalopram or placebo medication, 25 baseline single-pulse MEPs were recorded at an intensity to elicit MEPs with peak-to-peak amplitudes of on average 1 mV. After 2 h, the second baseline was recorded to explore the effect of medication on cortical plasticity, and adjusted, if necessary. After obtaining the second (or third) baseline, PAS was administered and 25 MEPs were recorded immediately after stimulation and at time points of 5, 10, 15, 20, 25, 30, 60, 90, and 120 min after plasticity induction. Further transcranial magnetic stimulation (TMS) measurements were conducted in the evening of the same day (SE), next morning, at ∼0900 hours (NM), next noon, at ∼1200 hours (NN), and next evening, at ∼1800 hours (NE).

RESULTS

All subjects tolerated the procedure well. None of them reported any side effect of either citalopram or the stimulation.

The average baseline MEP values did not significantly differ between groups as revealed by Student's t-tests (paired samples, two-tailed, p>0.05). Citalopram alone did not have any impact on cortical excitability, as revealed by Student's t-tests between first and second baseline values (paired samples, two-tailed, p>0.05; Table 1).

The ANOVA revealed significant main effects of medication (F(1)=5.345; p=0.039), stimulation (F(1)=39.497; p<0.001), stimulation × time (F(14)=15.593; p<0.001) and medication × time (F(14)=2.456; p=0.004) interactions (for details, see Table 2). The main effect of medication is caused by similarly directed effects of citalopram on MEP amplitudes for PAS10, and PAS25. As compared with placebo medication, citalopram enhanced motor cortical excitability. The main effect of stimulation is due to relatively larger MEP amplitudes in the PAS25 condition, as compared with PAS10, irrespective of medication condition or time point. The interaction of stimulation x time refers to different time courses of MEP alterations generated by PAS10, and PAS25. MEP reductions induced by PAS10 lasted longer than those accomplished by PAS25, and were antagonistically directed for up to 90 min after stimulation, but not with regard to the later time points. Finally, the interaction of medication and time course is caused by the MEP-enhancing effect of citalopram on MEP amplitudes, as compared with placebo medication, during the first 30 (PAS25) or 90 (PAS10) min after PAS, but not for later time points.

Post-hoc Student's t-tests show that in the placebo medication conditions, MEPs were significantly enhanced for 30 min after PAS-25 stimulation and diminished for 90 min after PAS10 stimulation. Citalopram abolished PAS10-induced LTD-like plasticity and enhanced PAS25-induced LTP-like plasticity, as compared with the respective placebo medication conditions. Student's t-tests show significant differences between drug and placebo conditions at all time points between 0 and 25 min after PAS10 administration and only at the single time point of 30 min after PAS25 (Figure 2).

Open in a separate window

Figure 2

Impact of citalopram on paired associative stimulation (PAS)-induced neuroplasticity. Shown are baseline-normalized MEP amplitudes after plasticity induction by PAS25 (a) and PAS10 (b) under placebo or citalopram medication conditions up to the evening of the post-stimulation day. (a) In the placebo medication condition, PAS25 induced a significant excitability elevation up to 60 min after stimulation, which was enhanced, but not prolonged, by citalopram. (b). In the placebo medication condition, cortical excitability was significantly reduced after PAS10, this effect was abolished by citalopram. Error bars indicate SEM. Filled symbols indicate significant differences of post-stimulation MEP amplitudes from respective baseline values; asterisks indicate significant differences between the drug and placebo medication conditions at the same time points (Student's t-test, two-tailed, paired samples, p<0.05).

For the effects of citalopram on PAS-induced plasticity with regard to the grand average calculated for the first 30 min after PAS, citalopram had a significant effect on focal excitability-diminishing plasticity, as revealed by respective post-hoc Student's t-tests (Student's t-test, paired samples, two-tailed, p=0.009), whereas only a non-significant tendency towards excitability enhancement after PAS25 stimulation was detected (Student's t-test, paired samples, two-tailed, p=0.126) (Figure 3).

Open in a separate window

Figure 3

Citalopram enhances PAS25-induced excitatory plasticity and abolishes PAS10-induced inhibitory plasticity. Each column represents the mean of baseline-normalized MEP±SEM amplitudes until 30 min after stimulation. Asterisks indicate significant differences between drug and placebo conditions (Student's t-test, two-tailed, paired samples, p<0.05).

DISCUSSION

The results of this study show that serotonin has specific effects on PAS-induced motor cortex plasticity in healthy humans. It abolishes focal LTD-like and trendwise enhances focal LTP-like plasticity induced by PAS10 and PAS25, respectively.

These results go in line with previous studies (Nitsche et al, 2009; Normann et al, 2007). Chronic application of SSRI enhanced facilitatory plasticity and resulted in a shift of inhibitory plasticity of early visual-evoked potentials (VEPs) toward excitation (Normann et al, 2007) or restored LTP induction and suppressed LTD facilitation in distressed animals (Von Frijtag et al, 2001). For the human motor cortex, another study has demonstrated that a single dose of the SSRI citalopram results in enhancement and prolongation of anodal tDCS-induced LTP-like facilitation and conversion of cathodal tDCS-induced LTD-like plasticity into facilitation (Nitsche et al, 2009). In further accordance, animal studies have shown that serotonin can enhance LTP (Huang and Kandel, 2007; Kojic et al, 1997; Machacek et al, 2001; Mori et al, 2001; Ohashi et al, 2002; Park et al, 2012), block the stress-caused inhibition of LTP (Ryan et al, 2008) or block LTD (Normann et al, 2007). Activation of 5-HT receptors was furthermore shown to reverse LTD induction or convert it into LTP (Costa et al, 2012; Kemp and Manahan-Vaughan, 2005).

However, activation of serotoninergic receptors also had opposing results on plasticity in other studies. Some studies have shown negative or no effect of serotonin on LTP induction (Edagawa et al, 1998; Huang and Kandel, 2007; Kojima et al, 2003; Normann et al, 2007; Sanberg et al, 2006), which can be explained by activation of different serotoninergic receptor subtypes, stage of brain development or dosage and frequency of application of 5-HT agonists or antagonists (Mori et al, 2001; Park et al, 2012; Staubli and Otaky, 1994). To explore the reasons of such opposing results, future studies should address specific serotoninergic receptor subtypes, using different 5-HT-receptor agonist or antagonist drugs. It might also make sense to explore different dosages of serotoninergic receptor agonists, as it has been demonstrated that other neuromodulators, such as dopamine, have dose-dependent effects on focal and non-focal plasticity in humans (Monte-Silva et al, 2009; Monte-Silva et al, 2010).

FUNDING AND DISCLOSURE

M-FK, and GB received no financial support or compensation from any individual or corporate entity over the past 3 years for research or professional service and there are no personal financial holdings that could be perceived as constituting a potential conflict of interest. WP is member of Advisory Boards of GSK, UCB, Desitin. MAN is member of Advisory Boards of UCB, Eisai, GSK, and Neuroelectronics.

Acknowledgments

References