Mechanisms of Nicotinic Modulation of Glutamatergic Neuroplasticity in Humans

Abstract

The impact of nicotine (NIC) on plasticity is thought to be primarily determined via calcium channel properties of nicotinic receptor subtypes, and glutamatergic plasticity is likewise calcium-dependent. Therefore glutamatergic plasticity is likely modulated by the impact of nicotinic receptor-dependent neuronal calcium influx. We tested this hypothesis for transcranial direct current stimulation (tDCS)-induced long-term potentiation-like plasticity, which is abolished by NIC in nonsmokers. To reduce calcium influx under NIC, we blocked N-methyl-d-aspartate (NMDA) receptors. We applied anodal tDCS combined with 15 mg NIC patches and the NMDA-receptor antagonist dextromethorphan (DMO) in 3 different doses (50, 100, and 150 mg) or placebo medication. Corticospinal excitability was monitored by single-pulse transcranial magnetic stimulation-induced motor-evoked potential amplitudes after plasticity induction. NIC abolished anodal tDCS-induced motor cortex excitability enhancement, which was restituted under medium dosage of DMO. Low-dosage DMO did not affect the impact of NIC on tDCS-induced plasticity and high-dosage DMO abolished plasticity. For DMO alone, the low dosage had no effect, but medium and high dosages abolished tDCS-induced plasticity. These results enhance our knowledge about the proposed calcium-dependent impact of NIC on plasticity in humans and might be relevant for the development of novel nicotinic treatments for cognitive dysfunction.

dextromethorphan, neuroplasticity, nicotine, nicotinic receptors, transcranial direct current stimulation

Introduction

Nicotine (NIC), a tertiary amine compound, is the primary psychoactive agent of tobacco smoke and is responsible for its addictive properties (Heishman et al. 1994; Levin et al. 2006). Beyond its addictive component, NIC has prominent effects on cognition in healthy humans (Grundey et al. 2012a; 2012b). Furthermore, NIC improves learning and attention in patients suffering from Alzheimer's disease, in which the cholinergic system is hypoactive (Wilson et al. 1995; White and Levin 1999).

Nicotinic receptors (nAChRs) form a heterogeneous family of ligand-gated ion channels that are differently expressed in many regions of the central nervous system (CNS) and, furthermore, modulate the effects of a wide diversity of transmitter pathways, including the cholinergic system itself, by both post- and presynaptic mechanisms, and dopamine, serotonin, norepinephrine, glutamate/N-methyl-d-aspartate (NMDA), γ-aminobutyric acid (GABA), opioid, and histaminergic systems (Levin and Simon 1998; Gotti and Clementi 2004), which are implicated in the generation and modulation of plasticity. Plasticity is thought to be the main physiological foundation for learning and memory formation (Rioult-Pedotti et al. 1998, 2000). The molecular basis for the functional heterogeneity of nAChRs is the existence of a gene family encoding at least 9 α “α₂–α₁₀” and 3 β “β₂–β₄” subunits (Gotti and Clementi 2004).

Both α₄β₂ and α₇ nAChRs, which are ligand-gated cation channels, appear to be critical for functional effects of NIC and its impact on cortical excitability and neuroplasticity (Burnashev 1998; Dajas-Bailador and Wonnacott 2004). These receptors, which are widely expressed throughout the CNS, influence synaptic plasticity and cognitive function by regulating calcium permeability as well as neurotransmitter release (Lisman 2001; Gotti and Clementi 2004; Browne et al. 2010; Huang et al. 2010).

In recent years, noninvasive brain stimulation protocols have been developed which enable the generation of long-term potentiation (LTP)- and long-term depression (LTD)-like plasticity in humans (Ziemann and Siebner 2008). Transcranial direct current stimulation (tDCS) induces cortical excitability modifications by long-lasting tonic subthreshold stimulation via relatively large electrodes positioned over cortical areas and thus produces a relatively nonfocal kind of plasticity, whose direction depends on the direction of current flow (Nitsche and Paulus 2000, 2001; Nitsche et al. 2003a). This protocol induces NMDA-receptor and calcium channel-dependent plasticity of the glutamatergic system (Liebetanz et al. 2002; Nitsche et al. 2003b; 2004a). Anodal tDCS over the motor cortex has a subthreshold depolarizing effect on neuronal membranes, thus resulting in enhanced excitability, while cathodal tDCS has antagonistic effects (Nitsche and Paulus 2000). Stimulation for some minutes results in respective neuroplastic excitability alterations, which outlast the stimulation for >1 h (Nitsche and Paulus 2001; Nitsche et al. 2003b).

The impact of the cholinergic system on tDCS-induced plasticity was explored in recent years in the human motor cortex as a model system. Enhancement of cholinergic activity via application of the cholinesterase-inhibitor rivastigmine resulted in an abolishment of LTP-like tDCS-generated plasticity (Kuo et al. 2007). These effects seem to be primarily controlled by nAChRs, since in nonsmokers NIC had similar effects as rivastigmine (Thirugnanasambandam et al. 2011; Grundey et al. 2012b). Moreover, these effects depend on nAChRs with calcium channel properties, since varenicline, a high-affinity partial agonist to α₄β₂ and full agonist to α₇ receptors, had comparable effects on tDCS-induced plasticity (Batsikadze et al. 2014).

These effects of NIC and nicotinic agents on facilitatory plasticity are compatible with a calcium-dependent mechanism. However, at present it is unclear why NIC in nonsmokers abolishes tDCS-induced facilitatory plasticity. Given the calcium hypothesis of plasticity, dependent on the amount of neuronal calcium concentration, LTP or LTD is induced. For LTP induction, higher calcium concentration as for LTD induction is required. Moreover, transition zones between LTP and LTD induction do exist, and too high calcium concentration might prevent LTP by potassium channel activation (Lisman 2001; Misonou et al. 2004). Thus, addition of NIC to anodal tDCS, given that both agents enhance calcium influx, might result in a respective calcium overflow, which then converts plasticity. Indeed such calcium-dependent nonlinear plasticity alterations have been demonstrated for tDCS (Batsikadze et al. 2013; Monte-Silva et al. 2013).

In the present study, we aimed to further elucidate the mechanism by which the activation of nAChRs alters neuroplasticity in humans, especially with regard to calcium-dependent mechanisms, and interaction with the glutamatergic system. We therefore combined anodal LTP-like plasticity-inducing tDCS under NIC with NMDA-receptor block by dextromethorphan (DMO). If NIC abolishes or converts tDCS-induced plasticity via enhanced calcium influx, then reduction of calcium influx via NMDA-receptor block should help to recover plasticity under specific dosages. Specifically we hypothesized that blocking NMDA receptors to different degrees should have nonlinear effects on anodal tDCS-induced plasticity: small and medium dosages of NMDA-receptor block should reestablish facilitatory plasticity under NIC due to a gradual diminution of calcium influx, whereas high-dosage NMDA-receptor block should abolish plasticity. We choose NMDA-receptor block in order to explore the mechanisms of NIC on tDCS-induced plasticity in this study. tDCS is assumed to induce plasticity of the glutamatergic system. In this kind of plasticity, NMDA receptors, which have calcium channel properties, are critically involved (Nitsche et al. 2003b; 2004b). To make NIC a potentially relevant substance to improve cognitive functions, as for Alzheimer's disease, it is important to unravel mechanisms of action, also with regard to worsening effects on physiological parameters. This potentially will make it possible to develop targeted and fine-tuned interventions in future.

Materials and Methods

Subjects

Thirteen healthy human volunteers (5 males/ 8 females) aged 26.4 ± 4.0 years were recruited. All of them were nonsmokers, none of them had smoked tobacco for at least 3 years before the study. All subjects were right-handed according to the Edinburgh handedness inventory (Oldfield 1971). None of them took any medication, had a history of neuropsychiatric or medical disease, present pregnancy, or metallic head implants. All volunteers gave written informed consent and were compensated for participation. The experiments were approved by the Ethics Committee of the University of Göttingen and conformed to the Declaration of Helsinki.

Transcranial Direct Current Stimulation

tDCS was administered by a battery-driven constant current stimulator (neuroConn GmbH, Ilmenau, Germany) through a pair of rubber electrodes covered with saline-soaked sponges (5 × 7 cm). The motor cortex electrode was fixed over the area representing the right abductor digiti minimi muscle (ADM) and the return electrode contra-laterally above the right supraorbital ridge. The connecting cables were positioned at the electrodes in posterior direction. Subjects received 1 mA of excitability-enhancing anodal tDCS for 13 min, which induces motor cortex excitability alterations lasting for ∼60 min after stimulation (Nitsche and Paulus 2001; Nitsche et al. 2003a), combined with all drug conditions or placebo (PLC) medication in different experimental sessions.

Monitoring of Motor Cortex Excitability

Transcranial magnetic stimulation (TMS)-elicited motor-evoked potentials (MEP) were recorded to measure excitability changes of the representional motor cortical area of the right ADM. Single-pulse TMS was conducted by a Magstim 200 magnetic stimulator (Magstim Company, Whitland, Dyfed, UK) at a frequency of 0.25 Hz with a figure of 8-shaped coil (diameter of one winding 70 mm, peak magnetic field, 2.2T). The coil was held tangentially to the scalp at an angle of 45° to the sagittal plane with the coil handle pointing laterally and posterior. The optimal position was defined as the site where stimulation resulted consistently in the largest MEPs. Surface EMG was recorded from the right ADM with Ag–AgCl electrodes in a belly-tendon montage. The signals were amplified and filtered with a time constant of 10 ms and a low-pass filter of 2.5 kHz, then digitized at an analog-to-digital rate of 5 kHz and further relayed into a laboratory computer using the Signal software and CED 1401 hardware (Cambridge Electronic Design). TMS intensity was adjusted to elicit, on average, baseline MEPs of 1 mV peak-to-peak amplitude and kept constant for the post-tDCS measures. MEPs were stored for offline analysis.

Pharmacological Intervention

NIC transdermal patches (Nicorette Depotpflaster, Pfizer, releasing 15 mg NIC over 16 h) or placebo patches were applied to all subjects in combination with DMO or placebo capsules under anodal tDCS. Related dosages of NIC have been shown to affect cognition (Min et al. 2001; Myers et al. 2004; Poltavski and Petros 2005) and are sufficient to influence CNS physiology (Thirugnanasambandam et al. 2011; Grundey et al. 2012a, 2013, 2015). The patch was applied 6 h before the start of the stimulation. This is the approximate time for the plasma level to reach its maximum following application of the patch (Nørregaard et al. 1992). The patch was retained until the end of the last measurements of the experiment on the afternoon of the second day. To counteract possible systemic side effects of NIC, participants were instructed to take 20 mg domperidone, a peripheral-acting dopamine D2-receptor antagonist (Barone 1999) with antiemetic effects, in case of need. Domperidone at 20 mg alone exerts no effects on motor cortical excitability, and is not expected to cross the blood–brain barrier (Barone 1999; Thirugnanasambandam et al. 2011; Grundey et al. 2013). Almost all subjects (11 out of 13) took domperidone in all experimental sessions.

DMO was administered in dosages of 50, 100, or 150 mg in different sessions of the experiments. DMO blocks NMDA receptors (Wong et al. 1988; Tortella et al. 1989; Franklin and Murray 1992; Netzer et al. 1993), which have calcium channel properties and are relevant for glutamatergic plasticity induction (Artola and Singer 1987; Iriki et al. 1989; Kirkwood et al. 1993; Hess et al. 1994, 1996). For this drug, the maximal plasma level is achieved 2 h after oral intake (Silvasti et al. 1987; Schadel et al. 1995), and the respective dosages suffice to elicit prominent effects in the CNS (Ziemann et al. 1998; Liebetanz et al. 2002; Nitsche et al. 2003b; Monte-Silva et al. 2013).

Experimental Procedures

First, subjects were seated comfortably in a chair with head- and armrest and asked to relax completely. EMG electrodes were placed over the right ADM as described above. The exact position of these electrodes was marked with a waterproof pen. Then TMS was applied over the left motor cortex representational area of the right ADM to determine the spot with the consistently highest MEPs in the resting ADM (optimal site), which was then marked with a waterproof pen in order to ensure that the TMS coil was positioned over the same spot during the whole course of the experiment. TMS intensity was adjusted to elicit MEP with peak-to-peak amplitudes of an average of 1 mV (S1 mV). Then, 25 MEPs were recorded for the determination of the first baseline (baseline 1 = bl1). After baseline recording, NIC patch or placebo patch was adhered to the left upper arm and remained there until the end of the last after-measurement at the following afternoon. 4 h after patch application, DMO, at one of the above-mentioned dosages, or placebo medication was administered. 6 h after patch application, again 25 MEPs were recorded and, if needed, TMS intensity was adjusted to 1 mV (baseline 2 = bl2). After that procedure, the anodal tDCS protocol was administered, followed by immediate recording of at least 25 MEPs at time points of 0, 5, 10, 15, 20, 25, 30, 60, 90, and 120 min. Further, after-measurements were conducted in the evening (ev) of the stimulation day, next morning (nmor), and next afternoon (nanoon). Sessions were conducted in randomized order, and an intersession interval of at least 1 week was obligatory to avoid interferences. This minimum 1-week inter-session interval was chosen because the after-effects of the tDCS protocol (with or without NIC) last shorter than 24 h, and the half-lives of DMO (Guenin et al. 2014) and NIC (Feyerabend et al. 1985) (after patch removal) are 2–4 h. Thus 1 week interval should be sufficient to avoid interferences. All participants received anodal tDCS under NIC only, DMO only, complete placebo medication, and combinations of DMO and NIC. This resulted in 8 sessions per subject (Fig. 1). The participants were blinded to treatment conditions. The experimenter was not blinded to medication conditions.

Figure 1.

Course of the study: after adjusting the TMS intensity to elicit MEP amplitudes of 1 mV (S1 mV), 25 MEPs were recorded at this stimulus intensity and the mean MEP amplitude was calculated (bl1). Then a NIC or PLC patch was adhered to the left upper arm. Four hours after patch application, DMO at one of the dosages 0 (PLC medication), 50, 100, 150 mg was administered. Then, 6 h after patch application, again 25 MEPs were recorded at the adjusted baseline stimulus intensity and the mean MEP amplitude was calculated (bl2). Then anodal tDCS was administered, followed by immediate recording of 25 MEPs at the time points 0, 5, 10, 15, 20, 25, 30, 60, 90, 120 min, in the evening (ev) of the stimulation day and in the morning (nmor) and afternoon of the day following the plasticity induction procedure (nanoon).

Analysis and Statistics

We powered the study for a moderate effect size according to our hypothesis that DMO would change the suppressive effect of NIC on tDCS-induced plasticity. Thus, assuming an effect size of 0.4 for the time × NIC × DMO interaction of a repeated-measure analysis of variance (ANOVA) as principal statistical test with a power of 80% and a 2-sided probability of type I error of 5%, a minimum of 12 subjects would be necessary; however to account for any discontinuation, we increased the estimated sample to 10%, resulting in 13 subjects.

The individual means of the 25 MEP amplitudes recorded at baselines 1, 2, and all time points after plasticity induction were calculated. The post-intervention mean MEP amplitudes from each subject were then normalized to the respective individual mean baseline (baseline 2) MEP-amplitude (quotient of post- versus preintervention MEP amplitudes). Then, these normalized MEP amplitudes were pooled together session-wise by calculating the grand average across subjects for each condition and time point. A repeated-measures ANOVA was performed on the above-mentioned data. MEP amplitude was the dependent variable including all time points up to next afternoon after tDCS. Nicotine (NIC/PLC), DMO (50, 100, 150 mg DMO/PLC), and time points were included as within-subjects factors. Mauchly's sphericity test was performed and Greenhouse–Geisser correction applied when necessary.

Conditional on significant results of the ANOVA, exploratory Student's t-tests (paired samples, 2-tailed, P < 0.05, not corrected for multiple comparisons) were performed to compare the MEP amplitudes before and after tDCS within each condition for each time point, as well as for TMS-intensities between different medication conditions within a given time bin. A P-value of <0.05 was considered significant for all statistical analyses. No adjustment for multiple comparisons was performed because of the exploratory character of these analyses (Bender and Lange 2001). All results are given as mean and standard error of mean (SEM).

To compare effects of NIC and different dosages of DMO on plasticity, additionally averaged MEPs for the first 30 min after stimulation were calculated for each subject per experimental session and normalized to baseline 2. Then, these averaged MEP values for each dosage condition were compared with the respective placebo condition by a repeated-measures ANOVA including the within-subject factors NIC and DMO. Since single measures at later time points showed significant deviation from baseline or the PLC condition, we performed another ANOVA for the pooled time bins from 60 to nanoon to rule out systematic effects.

To check for drug-induced alterations (NIC/PLC, DMO/PLC, or NIC/DMO) of MEP amplitude and the percentage of maximum TMS stimulator output (%MSO) to acquire the baseline MEP amplitude of 1 mV, as well as for identity of MEP amplitudes and %MSO between medication conditions, these values were compared before (bl1) and after drug administration (bl2), and within bl1 and bl2 between medication groups. An ANOVA was performed for MEP-amplitude values and %MSO for bl1 and bl2, and medication condition as within subject factors.

Results

Four hours after NIC patch application 3 subjects experienced mild nausea and 2 subjects vomited. Six hours after NIC patch application with oral intake of 150 mg of DMO 5 subjects experienced mild nausea. Eleven subjects took 20 mg of domperidone in all sessions. No side effects were reported under 50 or 100 mg of DMO. The 2 remaining participants tolerated all drugs well.

Comparison of baseline values resulted in a significant effect for the factor time in the ANOVA for TMS intensity [F_1,12 = 5.28, P = 0.04, partial η² = 0.306]. This was caused by a significant difference between baselines 1 and 2 for the NIC/DMO150 condition (Table 1). No other difference between baseline values was identified.

Table 1

MEP amplitudes and stimulation intensity before and after drug administration (bl1 and bl2)

TMS parameter	Medication condition	Baseline 1	Baseline 2
MEP	PLC/PLC	0.91 ± 0.04	0.92 ± 0.04
%MSO		59.69 ± 2.13	61.00 ± 2.29
MEP	PLC/DMO50	0.94 ± 0.05	0.97 ± 0.03
%MSO		63.38 ± 3.17	62.77 ± 3.26
MEP	PLC/DMO100	0.93 ± 0.05	1.04 ± 0.06
%MSO		62.23 ± 2.74	60.46 ± 3.34
MEP	PLC/DMO150	0.96 ± 0.03	1.04 ± 0.03
%MSO		64.15 ± 3.40	63.08 ± 3.29
MEP	NIC/PLC	0.96 ± 0.03	0.96 ± 0.03
%MSO		63.00 ± 3.55	62.92 ± 3.76
MEP	NIC/DMO50	0.88 ± 0.04	0.99 ± 0.03
%MSO		63.15 ± 3.81	61.92 ± 3.28
MEP	NIC/DMO100	0.99 ± 0.03	0.98 ± 0.03
%MSO		63.54 ± 3.41	63.31 ± 3.39
MEP	NIC/DMO150	0.95 ± 0.03	0.95 ± 0.04
%MSO		64.77 ± 2.94	60.69 ± 3.01**

Note: Shown are the mean MEP amplitudes ± SEM and stimulation intensity (percentage of maximum stimulator output, %MSO) mean ± SEM of baselines 1 and 2. The intensity of TMS was adjusted to elicit MEP with a peak-to-peak amplitude of ∼1 mV (baseline 1). A second baseline (baseline 2) was recorded 6 h after patch application to determine the impact of the drugs on cortical excitability.

**P < 0.01 when compared with baseline 1 (Student's t-test, 2-tailed, paired samples).

The principal ANOVA conducted for the main experiment revealed a significant main effect for the factor Time [F_13,143 = 4.511; P < 0.0001, partial η² = 0.291] and a significant 3-way interaction across NIC, DMO doses and Time [F_39,429 = 1.716; P < 0.01, partial η² = 0.135].

Effect of NMDA-Receptor Block on tDCS-Induced Plasticity

As revealed by post hoc Student's t-tests, in the placebo medication condition MEPs were significantly (P < 0.05) enhanced for up to 30 min after anodal tDCS when compared with respective baseline values. Under doses of 50 mg DMO, the obtained MEPs did not differ from those obtained under placebo medication up to 90 min after tDCS. Under 100 mg DMO, the post hoc analysis shows that MEP amplitudes did not differ from baseline values for any time point, but differed significantly from the placebo medication condition immediately after tDCS (P < 0.01), 10 min (P < 0.05), 20 min (P < 0.05), and 30 min after tDCS (P < 0.05). The MEP amplitudes obtained for the highest dose of DMO also did not differ from baseline values for up to the last measure at the day of stimulation, and were significantly altered when compared with the respective placebo medication condition at the 25-min interval (P < 0.05) (Fig. 2).

Figure 2.

NMDA-receptor antagonist DMO effects on tDCS-induced facilitatory plasticity. The graph shows baseline(bl2)-standardized motor-evoked potential (MEP) amplitudes on the Y-axis plotted at different time points following anodal tDCS under PLC medication or doses of 50, 100, and 150 mg DMO up to the afternoon of the post-stimulation day. In the PLC medication conditions, anodal tDCS induced a significant excitability elevation for up to 30 min after stimulation, which was abolished by 100 and 150 mg DMO. Fifty milligrams of DMO intake did not relevantly affect tDCS-induced plasticity. Filled symbols indicate statistically significant deviations of post-stimulation MEP amplitudes from respective baseline values. *Indicate significant differences between the placebo and DMO100 condition and ⁺ indicate significant differences between the placebo and DMO150 condition at the same time points (Student's t-test, 2-tailed, paired samples, P < 0.05). Error bars indicate standard error of mean. Further TMS measurements were conducted in the evening of the stimulation day (ev), next morning (nmor), and next afternoon (nanoon).

The Impact of NIC Under NMDA-Receptor Block on tDCS-Induced Plasticity

Under application of NIC patch without DMO, Student's t-test analysis showed that the tDCS-induced excitability enhancement was abolished. However, when combined with 100 mg DMO, facilitatory plasticity was reestablished. MEPs were significantly enhanced for 25 min after anodal tDCS when compared with the respective baseline values. MEP under this drug combination differed also significantly from the NIC without DMO condition at 10 (P < 0.01), 15 (P < 0.01) and 20 min after tDCS (P < 0.05). MEP under NIC patch with low and high doses of DMO differed from baseline values and the NIC without DMO condition only at 5 min after tDCS (high-dosage DMO) and nmor (low-dosage DMO), respectively (Fig. 3).

Figure 3.

Nicotinergic impact and its interaction with different doses of DMO on tDCS-induced neuroplasticity. The graph shows baseline(bl2)-standardized MEP amplitudes on the Y-axis plotted at different time points following anodal tDCS under NIC patch and DMO doses of 50, 100, and 150 mg or PLC medication up to the afternoon of the post-stimulation day. Under administration of NIC/PLC the tDCS-induced excitability enhancement was abolished. Under NIC/DMO 100 mg, facilitatory plasticity was reestablished and MEPs were significantly enhanced for 25 min after anodal tDCS, when compared with the respective baseline values. The MEPs of NIC/DMO under 50 and 150 mg DMO were not relevantly different from baseline and the NIC/PLC condition. Filled symbols indicate statistically significant deviations of post-stimulation MEP amplitudes from respective baselines.*indicate significant differences between the NIC/PLC and NIC/DMO100 condition, **indicate significant differences between the NIC/PLC and NIC/DMO50 condition and ⁺ indicate significant differences between NIC/PLC and NIC/DMO150 conditions at the same time points (Student's t-test, 2-tailed, paired samples, P < 0.05). Error bars indicate standard error of mean. Further TMS measurements were conducted in the evening of the stimulation day (ev), next morning (nmor), and next afternoon (nanoon).

Comparison of NIC and Different Dosages of DMO for Pooled Time Bins (up to 30 min and from 60 min to Nanoon) After Anodal tDCS

For the grand average calculated for the first 30 min, the ANOVA revealed significant effects for the 2-way interaction between NIC and DMO doses [F_3,36 = 4.222; P = 0.012, partial η² = 0.26].

Student's t-tests showed that a medium dose of DMO had a significant reestablishing effect on anodal tDCS-induced excitability plasticity, which was abolished by NIC alone (P < 0.01). Low and high dosages of DMO did not show this effect. In the placebo patch condition, medium and high doses of DMO had significant abolishing effects on the tDCS-induced excitability enhancement (P < 0.05) (Fig. 4).

Figure 4.

Impact of NIC and different dosages of DMO on tDCS-induced facilitatory plasticity for MEP pooled for 30 min after tDCS. Facilitatory plasticity is selectively reestablished by 100 mg DMO under NIC administration. NIC alone abolishes anodal tDCS-induced plasticity. Medium and High doses of DMO abolished tDCS-induced facilitatory plasticity in the placebo patch condition. Each column represents the mean of baseline-normalized MEP ± SEM amplitudes until 30 min after stimulation. *indicate significant differences between PLC/DMO and PLC/PLC or NIC/DMO and NIC/PLC (Student's t-test, 2-tailed, paired samples, P < 0.05).

For the later time points (60 min to nanoon), the ANOVA revealed no statistically significant differences (Fig. 5).

Figure 5.

Impact of NIC and different dosages of DMO on tDCS-induced facilitatory plasticity for MEP pooled from 60 min to nanoon after tDCS. There were no significant effects for the late measurements.

As shown by Figures 4 and 5, the impact of the interventions on plasticity is largely restricted to the first 30 min after plasticity induction.

Discussion

The results of this study show an interaction between nAChR activation and glutamatergic plasticity, which is suggested to be controlled by alterations of calcium influx. In accordance with previous studies, NIC alone abolished LTP-like plasticity induced by anodal tDCS. This effect might be caused by calcium overflow. Dosage-dependent reestablishment of plasticity under NMDA-receptor block via DMO might be due to a reduction of calcium influx. These results stress the neuromodulatory effect of NIC on glutamatergic plasticity. Taking into account that LTP is a relevant physiological foundation of learning and memory formation, this might also explain partially heterogeneous effects of NIC on cognitive processes.

For the separate NIC and DMO conditions, the results are fairly identical with those of previous studies, which explored the impact of these drugs on tDCS-induced plasticity (Liebetanz et al. 2002; Nitsche et al. 2003b; Kuo et al. 2007; Thirugnanasambandam et al. 2011; Grundey et al. 2013).

For MEP comparisons between drug administrations before tDCS, these were largely unchanged. The only significant effect—shown for TMS intensity after NIC/DMO150 application—might be due to arousal caused by side effects, however, it should not relevantly affect the main results of the study, since MEPs were identical to the other medication conditions after adjustment. Apart from this single deviation, none of the drug combinations alone affected size of MEPs, which is in accordance with previous studies (Ziemann and Siebner 2008; Thirugnanasambandam et al. 2011; Nitsche et al. 2012).

Proposed Mechanisms of Action

Nicotinic acetylcholine receptors are widely expressed throughout the CNS and are known to be involved in various cognitive functions such as attention, learning, memory consolidation, arousal, and sensory perception (Levin et al. 1992; Albuquerque et al. 2009).

In the brain, the predominant subtypes of functional nAChR are the homomeric α₇ and the heteromeric α₄β₂ receptor (Alkondon and Albuquerque 2004; Machaalani et al. 2010). Both receptors increase intracellular calcium levels by serving as pre- and post-synaptic ligand-gated calcium channels (Burnashev 1998; Dajas-Bailador and Wonnacott 2004). Several studies have shown that α₇ nAChRs can modulate the release of various neurotransmitters including glutamate, GABA, dopamine, and noradrenaline and thus have the potential to participate in a range of neurological functions (Alkondon et al. 1997, 1999; Summers et al. 1997; Li et al. 1998; Schilstrom et al. 1998; Maggi et al. 2001; Huang et al. 2010).

The plasticity-abolishing effects of NIC on tDCS-induced facilitatory plasticity in the present study are in line with the results of previous ones (Thirugnanasambandam et al. 2011; Grundey et al. 2013, 2015), in which global nAChR activation resulted in abolishment of this kind of plasticity. The results of a recently conducted study, in which the α₄β₂ and α₇ receptor agonist varenicline had comparable effects, suggest a relevant role of calcium in these effects (Batsikadze et al. 2014).

We hypothesized that addition of NIC to anodal tDCS will increase calcium concentration to a level that overshoots the concentration window for LTP induction (Thirugnanasambandam et al. 2011; Grundey et al. 2012b, 2013). NMDA-receptor activation and nAChR contribute to post-synaptic calcium influx, and thus calcium overflow presumably caused by hyper-activation of nAChR via NIC can be counterbalanced by down-regulating calcium influx via NMDA receptors. This mechanism would not require direct interaction between both receptors. To test this, we blocked calcium influx by a NMDA-receptor blocker which should have similar effects as a calcium channel blocker. Here, a direct effect of NMDA-receptor block on calcium influx via nAChR might be not very probable with regard to the α₇ receptor subtype, given the spatial distance between α₇ and NMDA receptors, as shown in some recently published articles (Shaffer et al. 2007; Zappettini et al. 2014). This situation might be different for interaction between α₄β₂ receptors and NMDA receptors, which seem to interact directly (Levin et al. 2002; Batsikadze et al. 2014). In accordance with our hypothesis, under administration of NIC and 100 mg DMO, a reestablishment of facilitatory plasticity was observed, suggesting that the calcium-decreasing effects of DMO under this dose are sufficient to reduce calcium concentration to a level which induces LTP.

The DMO 50 mg intake under NIC application however did not reestablish tDCS-generated facilitatory plasticity significantly, probably caused by an insufficient reduction of calcium influx by this low dosage, which is in line with its missing effect on tDCS-induced plasticity when given alone. The high-dosage DMO condition (150 mg) was already known to block tDCS-generated plasticity independent from nAChR activation in a couple of studies, including the present one (Liebetanz et al. 2002; Nitsche et al. 2003b; Monte-Silva et al. 2013), and it is likely that this dosage blocks NMDA receptors and prevents calcium influx to a degree which cannot be compensated by NIC-driven calcium influx.

Taken together, these results are suggestive for an impact of NIC on plasticity possibly via modulation of intracellular calcium concentration. Due to the calcium-dependency of LTP and LTD (Lisman 2001; Misonou et al. 2004), this would explain not only plasticity-enhancing, but also—reducing effects of the substance, depending on brain states, including other sources of calcium influx, as introduced here by stimulation-induced plasticity. It should however be stressed that these assumptions are to a certain degree speculative, because we could not monitor neuronal calcium influx directly. The relevance of these mechanisms for plasticity-related cognitive and behavioral processes awaits yet to be explored systematically.

General Remarks

The proposed calcium-dependent mechanisms of the effects of NIC on stimulation-induced plasticity might explain also different effects of this agent on noisy and non-noisy plasticity, as suggested by previous studies of our group, where a signal-to-noise-enhancing or focusing effect of NIC was suggested. In these studies, the impact of NIC on diffuse, not synapse-specific LTP-like plasticity, as accomplished by anodal tDCS, was compared with that of paired associative stimulation (PAS), which induces synapse-specific plasticity of somatosensorymotor cortex connections by combination of a peripheral mixed nerve stimulation with motor cortex TMS (in the classical paradigm one combined stimulus is applied every 20 sec for 30 min). In the case of PAS, simultaneous arrival of the somatosensory afferent and the TMS stimulus at the level of the motor cortex induces LTP-like glutamatergic calcium-dependent plasticity of motor cortex–somatosensory synapses (Stefan et al. 2000; Wolters et al. 2003). In contrast to the LTP-like plasticity-abolishing effects of NIC on tDCS-induced plasticity, PAS effects were strengthened by NIC in foregoing studies in nonsmokers (Kuo et al. 2007; Thirugnanasambandam et al. 2011). One might speculate that the tonic diffuse stimulation accomplished by tDCS is more prone to calcium overflow than the phasic synapse-specific PAS stimulation protocol. Consequently, abolishment of LTP-like plasticity would only be present in the tDCS condition. This hypothetical explanation should be explored more directly in future studies. At present, alternative explanations, such as different effects of NIC on Hebbian-like plasticity, as induced by PAS, and membrane polarization-induced plasticity, as generated by tDCS—although also for induction of plasticity by tDCS suprathreshold activity of neuronal populations is required (Fritsch et al. 2010)—cannot be ruled out.

These physiological mechanisms might also at least partially explain the effects of NIC on cognition. Beyond positive effects on various cognitive processes, also none or negative effects were described (Levin et al. 2002; Kumari et al. 2003; Jacobsen et al. 2005; Swan and Lessov-Schlaggar 2007; Grundey et al. 2015). This at least partially heterogeneous effect of NIC might be explained by presumably nonlinear physiological effects, which were not within the scope of the present project, but should be explored in future studies.

Moreover, the impact of NIC on physiology and cognitive processes is state-dependent. In smokers under NIC withdrawal, LTP-like plasticity could not be induced by tDCS and PAS alone, but was restituted in the case of NIC administration (Grundey et al. 2012a). Likewise, smokers showed reduced performance in a working memory task under NIC withdrawal when compared with nonsmokers, which was however improved selectively in smokers by NIC administration (Grundey et al. 2015). Since nAChRs with calcium channel properties are desensitized in smokers, these results could be explained by reduced calcium influx in smokers under NIC withdrawal, which is reestablished by NIC consumption, and therefore improves calcium-dependent processes (Lester and Dani 1994; Wooltorton et al. 2003; Grundey et al. 2012a). Conversely, in nonsmokers the more “optimal” calcium concentration without NIC will prevent further improvement by this substance. This state-dependent mechanism might also be relevant for the impact of NIC in neuropsychiatric diseases, in which the cholinergic system is hypoactive. Here, activation of nAChRs might be able to restitute respective processes (Wilson et al. 1995; White and Levin 1999).

The results of the present study suggest moreover that substance administration might have to be fine-tuned to result in optimal effects, since both too high and too low calcium concentration will likely compromise functions.

Limiting Conditions

Some limitations of the present study should be taken into account. NIC was administered in a single dosage in this experiment; therefore, it was not possible to determine a dose-dependent effect of NIC on plasticity, which was shown for other neuromodulators, like dopamine (Monte-Silva et al. 2010; Fresnoza et al. 2014). Furthermore, calcium influx was not directly measured in this study, thus respective mechanistic explanations of the results are hypothetical at present. The same holds true for the mechanistic speculations about cognitive effects.

MEP values obtained for the later time points seem to be somewhat smaller than those measured at baseline, although this effect was not significant for respective pooled values. This could imply that excitability at the bl2 measure was enhanced due to medication. However, since DMO and NIC did not modulate single-pulse MEP amplitudes in other studies (Nitsche et al. 2004b; Ziemann and Siebner 2008; Grundey et al. 2013) this seems not very probable. Moreover, since NIC and DMO as applied here have a half-lives of 2 (Feyerabend et al. 1985) and 2–4 h (Guenin et al. 2014), respectively, exceeding the duration of the main effects of the intervention on MEP, this would not relevantly affect the results during the first hour after tDCS, where the main effects were present. Alternatively, an arousal-related effect due to side effects of the medication cannot be ruled out completely.

With regard to methodological limitations, it should be mentioned that this study was single-blinded, and that the TMS coil was hand-held. However, identification of the hotspot of the target muscle allows correction of the coil position in the case of accidental movements, which will result in reduced MEP size, and the advantage of neuronavigated procedures for motor cortex stimulation seems to be gradual (Cincotta et al. 2010). We furthermore compensated for presumed larger variability of MEP amplitudes by obtaining a relatively high number of stimuli per time bin.

Funding

This work was supported by the Deutsche Forschungsgemeinschaft (DFG grant NI 683/4-2 “Impact of the nicotinergic alpha7 receptor in smokers and nonsmokers”) within the DFG priority programme “Nicotine: Molecular and Physiological Effects in Central Nervous System”. Marcelo Di Marcello Valladão Lugon was supported by CAPES, Brazil.

Notes

Conflict of Interest: None declared.

Author notes

†

Ester Miyuki Nakamura-Palacios and Michael A. Nitsche contributed equally to this study.