Volume 173, Issue 14 p. 2263-2277
RESEARCH PAPER
Free Access

Functional modulation of glycine receptors by the alkaloid gelsemine

Cesar O Lara

Cesar O Lara

Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile

These authors contributed equally to this workSearch for more papers by this author
Pablo Murath

Pablo Murath

Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile

These authors contributed equally to this workSearch for more papers by this author
Braulio Muñoz

Braulio Muñoz

Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile

Search for more papers by this author
Ana M Marileo

Ana M Marileo

Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile

Search for more papers by this author
Loreto San Martín

Loreto San Martín

Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile

Search for more papers by this author
Victoria P San Martín

Victoria P San Martín

Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile

Search for more papers by this author
Carlos F Burgos

Carlos F Burgos

Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile

Search for more papers by this author
Trinidad A Mariqueo

Trinidad A Mariqueo

Department of Pharmacology, School of Medicine, University of Talca, Chile

Search for more papers by this author
Luis G Aguayo

Luis G Aguayo

Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile

Search for more papers by this author
Jorge Fuentealba

Jorge Fuentealba

Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile

Search for more papers by this author
Patricio Godoy

Patricio Godoy

IfADo-Leibniz Research Centre for Working Environment and Human Factors at the Technical University Dortmund, Dortmund, Germany

Search for more papers by this author
Leonardo Guzman

Leonardo Guzman

Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile

Search for more papers by this author
Gonzalo E Yévenes

Corresponding Author

Gonzalo E Yévenes

Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile

Correspondence Dr Gonzalo E. Yévenes, Department of Physiology, Faculty of Biological Sciences, University of Concepcion, Chile.

E-mail: [email protected]

Search for more papers by this author
First published: 29 April 2016
Citations: 36

Abstract

Background and Purpose

Gelsemine is one of the principal alkaloids produced by the Gelsemium genus of plants belonging to the Loganiaceae family. The extracts of these plants have been used for many years, for a variety of medicinal purposes. Coincidentally, recent studies have shown that gelsemine exerts anxiolytic and analgesic effects on behavioural models. Several lines of evidence have suggested that these beneficial actions were dependent on glycine receptors, which are inhibitory neurotransmitter-gated ion channels of the CNS. However, it is currently unknown whether gelsemine can directly modulate the function of glycine receptors.

Experimental Approach

We examined the functional effects of gelsemine on glycine receptors expressed in transfected HEK293 cells and in cultured spinal neurons by electrophysiological techniques.

Key Results

Gelsemine directly modulated recombinant and native glycine receptors and exerted conformation-specific and subunit-selective effects. Gelsemine modulation was voltage-independent and was associated with differential changes in the apparent affinity for glycine and in the open probability of the ion channel. In addition, the alkaloid preferentially targeted glycine receptors in spinal neurons and showed only minor effects on GABAA and AMPA receptors. Furthermore, gelsemine significantly diminished the frequency of glycinergic and glutamatergic synaptic events without altering the amplitude.

Conclusions and Implications

Our results provide a pharmacological basis to explain, at least in part, the glycine receptor-dependent, beneficial and toxic effects of gelsemine in animals and humans. In addition, the pharmacological profile of gelsemine may open new approaches to the development of subunit-selective modulators of glycine receptors.

Abbreviations

  • AP5
  • D(-)-2-amino-5-phosphovaleric acid
  • CNQX
  • 6-cyano-7-nitroquinoxaline-2,3-dione
  • mIPSCs
  • miniature inhibitory postsynaptic currents
  • TTX
  • tetrodotoxin
  • Tables of Links

    TARGETS
    Ligand-gated ion channelsa
    AMPA receptors
    GABAA receptors
    α1 glycine receptors
    α2 glycine receptors
    α3 glycine receptors
    β1 glycine receptors
    Transportersb
    KCC2 (K-Cl co-transporter), SLC12A5
    LIGANDS
    3α,5α-tetrahydroprogesterone, allopregnanolone
    AMPA
    AP5
    Bicuculline
    CNQX
    GABA
    TTX
    • These Tables list key protein targets and ligands in this article which are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Southan et al., 2016) and are permanently archived in the Concise Guide to PHARMACOLOGY 2015/16 (a,bAlexander et al., 2015a,b).

    Introduction

    Strychnine-sensitive glycine receptors are anion-selective neurotransmitter-gated inhibitory ion channels. They belong to the Cys-loop receptor superfamily, together with the inhibitory GABAA receptor and the excitatory nicotinic ACh and 5-HT3 receptors. Glycine receptors are pentameric complexes composed of α and β subunits, which can form receptors composed exclusively by α subunits (i.e. homomeric) or by α and β subunits (i.e. heteromeric). Each subunit shares the basic structure of the Cys-loop receptor subunits, with an amino-terminal extracellular domain, four transmembrane domains (TM) and a large intracellular loop between TM3 and TM4 (Lynch, 2004; Zeilhofer et al., 2012a). To date, molecular cloning studies have identified four highly conserved α subunits (α1–4) and only one β subunit. Molecular, electrophysiological and immunocytochemical studies have determined that these isoforms present differences in their expression patterns and their biophysical properties, which are possibly linked to their proposed roles in physiology and disease (Legendre, 2001; Lynch, 2009).

    Glycine receptors mediate fast synaptic inhibition in the spinal cord, brain stem and in some selected CNS areas (Zeilhofer et al., 2012a,b). Additional evidence has shown that these receptors can also modulate neuronal excitability through tonic inhibition (Mitchell et al., 2007; Xu and Gong, 2010; Salling and Harrison, 2014; Liu et al., 2015) and presynaptic modulation (Turecek and Trussell, 2001; Jeong et al., 2003; Kunz et al., 2012; Choi et al., 2013). Synaptic glycine receptors interact with the scaffolding protein gephyrin through molecular interactions of the β subunits with the E domain of gephyrin, allowing the formation and stabilization of inhibitory synapses at postsynaptic sites (Tyagarajan and Fritschy, 2014). The current knowledge indicates that synaptic glycine receptors are mainly composed of heteromeric receptors, while extrasynaptic and presynaptic glycine receptors, on the other hand, are likely to exist at least partly as homomeric receptors (Lynch, 2009).

    The relevance of glycinergic inhibition has been described in early studies by applying the glycine receptor antagonist strychnine to rodents (Callister and Graham, 2010). These experiments defined the critical importance of these receptors on muscle tone, motor control and sensory processing. These results were later expanded by genetic studies in mice and humans carrying mutations in glycine receptor genes that generated altered glycinergic inhibition leading to hyperekplexia and startle disease (Harvey et al., 2008). Recent evidence has determined pivotal roles for glycinergic inhibition in inflammation-induced chronic pain hypersensitivity, brainstem respiratory control and temporal lobe epilepsy (Harvey et al., 2004; Eichler et al., 2008; Manzke et al., 2010).

    Despite the importance of the glycinergic inhibition in physiological processes and in disease states, only few agonists, antagonists and allosteric modulators for glycine receptors have been characterized (Yévenes and Zeilhofer, 2011a). In addition, with the exception of the classical alkaloid strychnine, most of the known molecules targeting glycine receptors are also modulators of other receptors. Agonists and positive allosteric modulators targeting specific subtypes of glycine receptors may lead to novel therapeutic strategies against a variety of diseases involving altered synaptic glycinergic inhibition, such as hyperekplexia and chronic pain (Xiong et al., 2012, 2014). In this context, recent studies have shown that gelsemine, a natural alkaloid from the Gelsemium genus plants of the Loganiaceae family, binds to spinal glycine receptors and exerts analgesic and anxiolytic actions in behavioural models (Liu et al., 2013; Meyer et al., 2013; Zhang et al., 2013). However, the functional effects of gelsemine on glycine receptors are still largely unknown. In this study, we characterized the effects and mechanisms underlying the modulation of different glycine receptor subtypes by gelsemine. By analyzing recombinant glycine receptors through electrophysiology, we determined that gelsemine exerts subunit-specific effects on different glycine receptor subtypes. We furthermore determined that gelsemine modulates spinal glycinergic synapses. Thus, our results define the actions of the alkaloid gelsemine on glycine receptors, contributing to our current understanding of its pharmacological actions in vivo.

    Methods

    Cell culture and transfection

    HEK293 cells (CRL-1573; American Type Culture Collection, Manassas, VA, USA) were cultured using standard methods. The cells were transfected using XfectTM Transfection Reagent (Clontech, San Francisco, CA, USA) using 0.5 μg of cDNA plasmids encoding glycine receptor α subunits and 0.5 μg of eGFP. To study heteromeric glycine receptors, we transfected 0.4 μg of α subunits/eGFP plasmids plus 2.0 μg of β subunit cDNA to avoid the formation of homomeric glycine receptors (Yévenes et al., 2010). All recordings were made 18–36 h after transfection.

    Animals and cultures of spinal cord neurons

    All animal care and experimental protocols of this study complied with the ethical protocols established by the National Institutes of Health (NIH, USA) and were supervised and approved by the Bioethical Committee of the University of Concepción. A total of 20 animals were used in this study. The animal studies are reported as recommended by the ARRIVE guidelines (Kilkenny et al., 2010; McGrath & Lilley, 2015).

    Cultured spinal cord neurons were prepared as described (Tapia and Aguayo, 1998; Yévenes et al., 2003). A pregnant mouse (C57BL/6J) was placed in a closed bucket with isoflurane before cervical dislocation. The whole spinal cord from five–six mouse embryos (E13–14 days) was plated at 300 000 cells/mL onto 18 mm glass coverslips coated with poly-L-lysine (70–150 kDa; Trevigen, Gaithersburg, MD, USA). The feeding medium consisted of 90% minimal essential medium (GIBCO, Rockville, MD, USA), 5% heat-inactivated horse serum (Hyclone, Logan, UT USA), 5% FBS (GIBCO, Rockville, MD, USA), and a mixture of nutrient supplements. The electrophysiological experiments were performed after 14–17 days of culture.

    Electrophysiology

    Glycine-evoked currents were recorded from transfected HEK293 cells and from cultured spinal neurons in the whole-cell voltage-clamp configuration at room temperature (20–24°C) at a holding potential of −60 mV (Yévenes et al., 2010, Yévenes and Zeilhofer, 2011b). Patch electrodes (3–4 mΩ) were pulled from borosilicate glass and were filled with (in mM): 120 CsCl, 8 EGTA, 10 HEPES (pH 7.4), 4 MgCl2, 0.5 GTP and 2 ATP. The external solution contained (in mM) 140 NaCl, 5.4 KCl, 2.0 CaCl2, 1.0 MgCl2, 10 HEPES (pH 7.4) and 10 glucose. Whole-cell recordings were performed with an Axoclamp 200B amplifier (Molecular Devices, Sunnyvale, CA, USA) and acquired using Clampex 10.1 or Axopatch 10.0 software. Data analysis was performed off-line using Clampfit 10.1 (Axon Instruments, Sunnyvale, CA, USA) and MiniAnalysis 6.0.3 (Synaptosoft, CA, USA). Exogenous glycine-evoked currents were obtained using a manually applied pulse (3–4 s) of the agonist and an outlet tube (200 μm ID) of a custom-designed gravity-fed microperfusion system positioned 50–100 μm from the recorded cell. A stepper motor-driven rapid solution exchanger system was used for the desensitization experiments (Warner, USA). The percentage of desensitizing current and the decay time constant of glycine receptors in the absence or in the presence of gelsemine were obtained from whole-cell current traces of 5 s duration. The rise and decay time was fitted to a single exponential function and was calculated off-line using Clampfit 10.1 or MiniAnalysis 6.0.3. The EC10 values for the recombinant and neuronal receptors were obtained experimentally after the successive application of increasing concentrations of glycine (3–3000 μM), GABA (0.01–100 μM) or AMPA (0.01–100 μM). The concentration–response curve parameters (EC50 and Hill coefficients, nH) were obtained from the fitted curve of normalized concentration–response data points to the equation Iagonist = Imax (agonist)nH / ((agonist)nH + (EC50)nH). The mean maximal current (Imax) indicated corresponds to the average maximal current elicited by saturating concentrations of the agonist.

    Gelsemine stocks were prepared in high purity distilled water and subsequently diluted into the recording solution on the day of the experiment. In most of the experiments, gelsemine was co-applied with the corresponding agonist, without any pre-application. Glycinergic miniature inhibitory postsynaptic currents (mIPSCs) in spinal neurons were isolated using bicuculline (5 μM), CNQX (4 μM), D(-)-2-amino-5-phosphovaleric acid (AP5, 50 μM) and tetrodotoxin (TTX, 0.3 μM). The remaining mIPSCs were blocked by strychnine (1 μM). Glutamatergic synaptic events were recorded at a holding potential of −70 mV using a low-chloride based intracellular solution composed of (in mM): 120 K-gluconate, 10 KCl, 8 EGTA, 10 HEPES (pH 7.4), 2 MgCl2, 0.5 GTP and 2 ATP. The glutamatergic miniature excitatory postsynaptic currents (mEPSCs) were recorded in the presence of extracellular TTX (0.3 μM). Under these conditions, all the remaining synaptic events were blocked by a cocktail of CNQX (4 μM) and AP5 (50 μM), confirming their glutamatergic nature and the absence of chloride currents.

    The methodologies employed for the single channel recordings have been previously published (Yévenes et al., 2008, 2010; Marabelli et al., 2013). The patch pipettes had tip resistances of 10–20 mΩ and were manually fire polished in a microforge (Narishige, Japan). The outside-out recordings (i.e. for α1 and α2 glycine receptors) were obtained at −60 mV, whereas the cell-attached recordings (i.e. α3 glycine receptors were obtained at +60 mV. The data were filtered (1 kHz low-pass 8-pole Butterworth) and acquired at 5–20 kHz using an Axopatch 200B amplifier and a 1322A Digidata (Axon Instruments, Union City, CA, USA). Data were acquired using pClamp software and analysed off-line with Clampfit 10.1 (Axon Instruments, Union City, CA). Solutions were applied to cells using a stepper motor-driven rapid solution exchanger (Fast-Step, Warner Instrument Corp.).

    Data analysis

    The data and statistical analysis comply with the recommendations on experimental design and analysis in pharmacology (Curtis et al., 2015). All values were expressed as mean ± SEM of normalized agonist-activated currents. Statistical comparisons were performed using Student t-tests or the Kolmogorov–Smirnov (K–S) test. P < 0.05 was considered statistically significant. Multiple comparisons were analysed with ANOVA followed by a Bonferroni post hoc test. For statistical analyses, at least six cells were analysed per condition. All the statistical analyses and plots were performed with MicroCal Origin 6.0/8.0 (Northampton, MA, USA) and MiniAnalysis 6.0.3. software.

    Materials

    Gelsemine hydrochloride was obtained from Extrasynthese (Genay Cedex, Lyon Nord, France). TTX citrate was purchased from Alomone Labs (Jerusalem, Israel). CNQX, D-AP5, strychnine and bicuculline were from Tocris (Bristol, UK). HEPES was from US Biological (Salem, MA, USA). All other reagents were from Sigma-Aldrich (St. Louis, MO,USA).

    Results

    We first examined the sensitivity of the three most abundant glycine receptor α subunits (α1, α2 and α3) to different concentrations of gelsemine. For all these subunits, we performed experiments on receptors composed of only α subunits (i.e. homomeric) and also on receptors made up of α and β subunits (i.e. heteromeric). Application of gelsemine alone to cells expressing homomeric glycine receptors did not elicit any detectable change in the holding currents (Supporting Information Fig. S1), suggesting the absence of agonistic activity. Using a sub-saturating concentration of glycine (EC10) to activate its receptors, we found that low micromolar concentrations of gelsemine potentiated currents through homomeric α1 glycine receptors expressed in HEK293 cells (Figure 1A). The modulation elicited by gelsemine on α1 glycine receptors displayed a bell-shaped profile, showing a concentration-dependent potentiation in the range between 0.1 to 50 μM and displaying inhibition at concentrations of gelsemine above 100 μM. Interestingly, the expression of the β subunit switched the potentiation elicited by gelsemine to a concentration-dependent inhibition (Figure 1A). For example, 10 μM gelsemine induced a potentiation of glycine-activated currents through α1 glycine receptors, whereas currents through α1β glycine receptors were not significantly modulated. On the other hand, 50 μM gelsemine still potentiated α1 glycine receptors but inhibited α1β glycine receptors. We then analysed whether the effects of gelsemine on homomeric glycine receptors composed of α2 and α3 subunits were equivalent to those observed in α1 glycine receptors. Interestingly, gelsemine showed a consistent concentration-dependent inhibition indicating a subunit-specific effect on homomeric glycine receptors (Figure 1B–D). In contrast with the modulation of glycine receptors composed of α1 subunits, the presence of the β subunit did not significantly influence the inhibition of α2 and α3 glycine receptors by the alkaloid (Figure 1A–C). Further analyses indicated that the gelsemine-induced inhibition of homomeric and heteromeric α2 and α3 glycine receptors displayed similar pharmacological profiles (Figure 1, Table 1). On the other hand, the gelsemine-elicited potentiation of homomeric α1 glycine receptors appeared to be highly specific and strongly influenced by the expression of the β subunit.

    Details are in the caption following the image
    Modulation of recombinant glycine receptors (GlyR) by gelsemine. (A). Whole-cell current traces evoked by 15–20 μM glycine before and after the application of gelsemine (10 μM or 50 μM) to HEK293 cells expressing α1GlyRs or α1βGlyRs. The graph summarizes the effect of different gelsemine concentrations (0.01–300 μM) on the glycine-activated currents. (B.) Current traces evoked by 40–60 μM glycine before and after the application of gelsemine (50 μM) to HEK293 cells expressing α2GlyRs or α2βGlyRs. The plot summarizes the effects of different gelsemine concentrations (0.1–300 μM) on the glycine-activated currents. (C.) Current traces evoked by 50–70 μM glycine before and after the application of gelsemine (50 μM) to HEK293 cells expressing α3GlyRs or α3β GlyRs. The plot summarizes the effects of different gelsemine concentrations (0.1–300 μM) on the glycine-activated currents. (D). Summary of the gelsemine effects (25 μM) on homomeric and heteromeric GlyRs. *, P < 0.05; significant difference between α1GlyRs and the other receptors studied; ANOVA followed by Bonferroni post hoc test, F(5,55) = 12.37. The values of the gelsemine-induced inhibition of homomeric and heteromeric GlyRs were not significantly different. Data are means ± SEM. of cells expressing α1(n = 12), α1β(n = 13), α2(n = 8), α2β(n = 10), α3(n = 12) and α3β (n = 8).
    Table 1. Effects of gelsemine on homomeric and heteromeric glycine receptors (GlyR) expressed in HEK293 cells
    GlyR EC50 (μM) IC50 (μM) nH Maximal modulation (%) Maximal modulation concentration (μM) n
    α1 0.59 ± 0.02 4.68 ± 0.3 54.8 ± 13.5 10 12
    α1β N.C. N.C. −79.2 ± 3.8 300 13
    α2 31.1 ± 3.8 1.17 ± 0.2 −94.5 ± 4.5 300 8
    α2β 42.0 ± 3.7 1.58 ± 0.2 −87.0 ± 1.8 300 10
    α3 36.7 ± 3.1 1.70 ± 0.2 −90.3 ± 3.2 300 12
    α3β 49.2 ± 6.8 0.94 ± 0.1 −85.0 ± 9.8 300 8
    • Values are indicated as mean ± SEM. from the indicated number of cells. The EC50 value of potentiation for α1glycine receptors was calculated from the concentration–response curve data points between 0.01 and 50 μM. The statistical analyses did not show significant differences between the values of gelsemine-induced inhibition of homomeric and heteromeric receptors composed of the α2 and α3 subunits. N.C. Not calculated.

    We next aimed at determining the mechanisms underlying the modulation of glycine receptors by gelsemine. We first analysed the effects of the alkaloid on the glycine sensitivity of homomeric glycine receptors. Analysis of concentration–response curves revealed that gelsemine (50 μM) significantly decreased the apparent affinity for glycine in α2 and α3 glycine receptors without changing the maximal current amplitudes (Figure 2, Table 2). On the other hand, 10 μM of the alkaloid shifted the glycine sensitivity curve of α1 glycine receptors towards the left and did not affect the maximal current amplitudes (Figure 2, Table 2). Thus, these results suggest that gelsemine modulates glycine receptors by modifying the apparent affinity for glycine. However, although the alkaloid did not affect the maximal current amplitude elicited by high concentrations of the agonist, it may still have affected the desensitization rates of the ion channel. Further analyses of the glycine-activated currents stimulated by saturating agonist concentrations indicated that gelsemine did not modify the fraction of desensitized current or the decay time constant of homomeric α1 glycine receptors (Figure 2D–E, Table 2). Similarly, the alkaloid did not alter the desensitization profile of α2 and α3 glycine receptors (Figure 2D–E, Table 2), suggesting that the alkaloid has negligible effects on glycine receptors fully activated by the agonist. Overall, these data indicate that the differential glycine receptor modulation elicited by gelsemine can be explained at least partly by differential changes in the apparent affinity for glycine, rather than by altering the desensitization rates.

    Details are in the caption following the image
    Effects of gelsemine on the agonist sensitivity and on the desensitization rates of different glycine receptor (GlyR) subunits. A–C. Concentration–response curves to glycine in the absence and the presence of gelsemine from cells expressing α1 (panel A, n = 12), α2 (panel B, n = 10) or α3 (panel C, n = 13) GlyRs. (D). Examples of current traces from GlyRs using a saturating concentration of glycine (1000–3000 μM) in the absence or presence of gelsemine (10 μM for α1GlyRs and 50 μM for α2/α3GlyRs). (E). The graphs summarize the effects of gelsemine on the percentage of desensitized current and on the decay time constant of the glycine-evoked currents. The alkaloid did not significantly influence the desensitization rates. Data are means ± SEM. of cells expressing α1(n = 12), α2(n = 10) and α3(n = 13).
    Table 2. Effects of gelsemine on the glycine sensitivity and desensitization profile of homomeric glycine receptors (GlyRs) expressed in HEK293 cells
    GlyR EC50 (μM) nH Imax (pA) Decay time (s) Desensitized current (%)
    α1 79.1 ± 2.8 3.51 ± 0.41 1773 ± 243 3.97 ± 0.95 32.1 ± 4.05
    +G 37.5 ± 1.0* 3.36 ± 0.26 1602 ± 299 3.63 ± 0.90 36.3 ± 5.99
    α2 96.5 ± 4.3 1.68 ± 0.13 1587 ± 293 2.24 ± 0.68 61.2 ± 7.40
    +G 172.5 ± 6.3* 1.67 ± 0.08 1864 ± 686 2.54 ± 0.43 66.5 ± 7.00
    α3 253.8 ± 10.35 2.05 ± 0.14 1995 ± 458 3.60 ± 0.72 71.4 ± 6.50
    +G 407.3 ± 14.0* 1.89 ± 0.10 1850 ± 420 3.38 ± 0.68 66.6 ± 8.31
    • Values of the concentration–response curve parameters (EC50 and Hill coefficients, nH) were obtained from the curve fits of normalized concentration–response data points from the indicated number of cells. Desensitization parameters were calculated from glycine-activated current traces obtained using 1000–3000 μM of agonist. The symbol +G denotes the presence of 10 μM (α1glycine receptors) or 50 μM of gelsemine (α2 and α3glycine receptors). The values of each receptor subtype were compared in the absence or presence of the alkaloid. Maximal current (Imax), decay time and percentage of desensitized current are presented as mean ± SEM. from 10–13 cells.
    • * P < 0.05.

    We then studied the influence of the membrane potential on gelsemine effects through whole-cell recordings performed at potentials ranging from −90 to +60 mV. From these recordings, we constructed current–voltage (I–V) relationships of glycine-evoked currents in the absence or in the presence of gelsemine (Figure 3A–C). The results showed that the relationship of the glycine-evoked current amplitude with the membrane voltage and the reversal potential was not altered extensively by gelsemine. In addition, the data showed that neither the potentiation of α1 glycine receptors nor the inhibition of α2 or α3 glycine receptors by the alkaloid was influenced significantly by the membrane potential (Figure 3D–F). These results suggest that the gelsemine-induced potentiation and inhibition of glycine-evoked chloride currents is likely to involve the direct modulation of the ion channel function by the alkaloid and that the inhibition is not related to alterations on the chloride flux due to pore blockade. To further explore the mechanisms involved in the effects of the alkaloid on glycine receptors, we next performed single channel recordings of α1 and α2 glycine receptors in the outside-out configuration (Yévenes et al., 2010) and of α3 glycine receptors in the cell-attached configuration (Marabelli et al., 2013). We first studied the mechanisms underlying the potentiation of homomeric α1 glycine receptors by gelsemine. Application of 10 μM of gelsemine to membrane patches expressing α1 glycine receptors enhanced the normalized open probability (nPo) and the frequency of ion channel openings by (Figure 4A–C). On the other hand, studies conducted in cells expressing homomeric α2 and α3 glycine receptors showed that 50 μM gelsemine decreased the open probability and the frequency of ion channel openings (Figure 4E–L). In all the receptors examined, the alkaloid did not elicit any significant change in their main conductances (Figure 4C-G-K). Overall, these results are in agreement with our data obtained using whole-cell currents and are consistent with a mechanism of ion channel modulation, resulting from the fast direct binding of gelsemine to the glycine receptor structure.

    Details are in the caption following the image
    The positive and negative effects of gelsemine on the glycine-evoked currents in different glycine receptor (GlyR) subunits are voltage-independent. A–C. Voltage–current relationships (I–V plots) for α1 (panel A), α2 (panel B) or α3 (panel C) GlyRs in the absence and presence of gelsemine. The recordings were performed using membrane potentials ranging from −90 to +60 mV with steps of 30 mV. The test concentration of the alkaloid was 10 μM for α1GlyRs and 50 μM for α2/α3GlyRs. (D–F). The plots summarize the mean percentage of potentiation (α1GlyR, panel D) or the mean percentage of inhibition (α2/α3GlyRs, panels E, F) elicited by gelsemine on glycine-evoked currents recorded at different membrane potentials. Both the positive and negative effects of the alkaloid were not significantly influenced by the membrane potential (ANOVA followed by Bonferroni post hoc test). Data are means ± SEM. of seven cells for each glycine receptor subunit.
    Details are in the caption following the image
    Gelsemine effects on the single-channel activity of glycine receptor (GlyR) α1, α2 and α3 subunits. (A). Single-channel recordings in the outside-out configuration from cells expressing α1GlyRs before and during the application of 10 μM of gelsemine. (B, C). The graphs show that gelsemine significantly enhanced the open probability of α1GlyRs, but did not modify the main conductance. * P < 0.05, paired Student t-test. (D). The bar graph summarizes the percentage change elicited by 10 μM of gelsemine on the single-channel activity of α1GlyRs. Differences were significant. * P < 0.05; paired Student t-test. (E). Recordings from cells expressing α2GlyRs before and during the application of 50 μM of gelsemine. (F, G). The plots summarize the effects of gelsemine on the open probability and the main conductance of α2GlyRs. * P < 0.05, paired Student t-test. (H). The bar graph summarizes the percentage change elicited by 50 μM of gelsemine on the single-channel activity of α2GlyRs. * P < 0.05; paired Student t-test. (I). Single-channel recordings in the cell-attached configuration from cells expressing α3GlyRs before and in the presence of 50 μM of gelsemine. (J, K). Gelsemine significantly decreased the open probability of α3GlyRs, but did not modify the main conductance. * P < 0.05; paired Student t-test. (L). The bar graph summarizes the percentage change elicited by 50 μM of gelsemine on the single-channel activity of α3GlyRs. * P < 0.05; paired Student t-test. Data were obtained from eight patches in each condition. γ, main conductance; f, frequency; NPo, open probability.

    To assess the modulation of native glycine receptors by gelsemine, we carried out electrophysiological experiments on cultures of spinal neurons from mice (Yévenes et al., 2003; Mariqueo et al., 2014). These neurons were taken primarily from embryos at E13/14 and were subsequently cultured for 14–17 days. At the point of our electrophysiological experiments, these cultures comprise a heterogeneous population of neurons, containing mostly glial cells and interneurons. During early periods of the in vitro development of these cultures, the neurons preferentially express α2 glycine receptors (i.e. until 8–10 days in vitro), and the glycine-activated currents depolarize the cell membrane (Tapia and Aguayo, 1998). In more intermediate and mature stages (i.e. >14 days in vitro), the activation of glycine receptors produced hyperpolarization of the membrane potential and inhibition of neuronal firing (Tapia and Aguayo, 1998). These cultured spinal neurons preferentially express synaptic heteromeric α1β glycine receptors possibly mixed with minor amounts of homomeric α1 or α2 glycine receptors (Van Zundert et al., 2004; Mariqueo et al., 2014). In line with previous evidence, the glycine sensitivity of these neurons displayed an EC50 value of 29.1 ± 1.1 μM (Figure 5A), suggesting a predominant expression of both homo and heteromeric α1-containing glycine receptors (Aguayo et al., 2004; Mariqueo et al., 2014). Using a sub-saturating concentration of glycine (15 μM), subsequent experiments showed a concentration-dependent inhibition of the glycine-evoked currents elicited by gelsemine (0.1–200 μM, Figure 5B–C). The calculated IC50 value of 42.4 ± 4.4 μM for inhibition resembles the sensitivity of recombinant α1β glycine receptors and α2-containing glycine receptors to the alkaloid (Figure 1). Further recordings using saturating concentrations of glycine (i.e. 1000 μM) revealed that gelsemine did not significantly affect the amplitude or the desensitization profile of the glycine-evoked current (Figure 5D–E). These results thus demonstrate that low micromolar concentrations of gelsemine negatively modulate native glycine receptors expressed in mouse spinal cord neurons.

    Details are in the caption following the image
    Effects of gelsemine on spinal glycine, GABAA and AMPA receptors. (A). Glycine concentration–response curve obtained from cultured spinal cord neurons (n = 7). (B). Glycine-activated currents from a spinal neuron in the absence and presence of increasing concentrations of gelsemine. (C). The graph describes the concentration-dependent inhibition of the neuronal glycine-evoked current by gelsemine (0.1–300 μM, n = 8). (D). Current traces from spinal glycine receptors (GlyRs) activated by 1000 μM of glycine in the absence or the presence of 50 μM of gelsemine. (E). The graphs summarize the effects of gelsemine on the fraction of desensitized current and on the decay time constant of the glycine-evoked currents. Differences were not significant. (F). Concentration–response curves to AMPA (n = 8) or GABA (n = 6) from spinal cord neurons. (G). Examples of AMPA or GABA-activated currents from spinal neurons in the absence and presence of 50 μM of gelsemine. (H). Sensitivity of spinal GABAA and AMPA receptors to gelsemine. Because the two data sets did not fit properly to the Hill equation, the IC50 values were not calculated (n = 8 for both receptors). (I). Summary of the gelsemine effects (50 μM) on spinal glycine, GABAA and AMPA receptors. The effects of gelsemine on GlyRs were significant. * P < 0.05; significant effect of gelsemine;. The gelsemine-induced effects on glycine, GABAA and AMPA receptors were also significantly different. + P < 0.05; ANOVA followed by Bonferroni post hoc test; F (5,48) = 12.00. Data are means ± SEM. of glycine activated-currents (vehicle (n = 5) and gelsemine (n = 8)), GABA-activated currents (vehicle (n = 8) and gelsemine (n = 11)) and AMPA-activated currents (vehicle (n = 8) and gelsemine (n = 13)) obtained from spinal neurons.

    We next aimed to determine whether the alkaloid could modify the activity of other ligand-gated ion channels involved in phasic synaptic transmission in the spinal cord. Therefore, we analysed the alkaloid effects on the agonist-evoked currents through inhibitory GABAA and excitatory glutamate AMPA receptors. Both AMPA and GABA-evoked currents in these neurons showed concentration-dependent responses and displayed EC50 values of 1.89 ± 0.25 μM and 3.18 ± 0.7 μM, respectively (Figure 5F). Using sub-saturating concentrations of these agonists (≈EC10, 0.25 μM of AMPA and 2 μM of GABA), we found that gelsemine did not have significant effects on GABAA and AMPA receptors at concentrations below 50 μM (Figure 5G–H). However, concentrations above 100 μM of the alkaloid partially inhibited GABA-activated currents (Figure 5H). Conversely, the AMPA-evoked currents were not affected by gelsemine (Figure 5H). Additional analyses comparing the sensitivity of spinal AMPA, GABAA and glycine receptors to 50 μM of gelsemine or to vehicle revealed that only spinal glycine receptors were significantly inhibited (Figure 5I). These analyses also showed that the extent of glycine receptor inhibition was statistically different from the effects of gelsemine on GABAA and AMPA receptors (Figure 5I). These results suggest that low micromolar concentrations (i.e. ≤50 μM) of gelsemine may preferentially target glycine receptors rather than GABAA or AMPA receptors in spinal cord neurons.

    All the results described above suggest that gelsemine can modulate glycine-evoked currents in recombinant and neuronal glycine receptors. To assess the potential effects of the alkaloid in a synaptic context, we next investigated whether gelsemine could modulate the glycinergic synaptic activity of this spinal cord neuronal preparation. Therefore, we recorded glycinergic mIPSCs from cultured spinal cord neurons at −60 mV (Figure 6). The glycinergic mIPSCs were pharmacologically isolated by the application of the voltage-gated sodium channel blocker TTX, the GABAA receptor antagonist bicuculline, the AMPA receptor antagonist CNQX and the NMDA receptor antagonist AP5 to the extracellular solution. The remaining synaptic currents were completely blocked by the glycine receptor inhibitor strychnine, confirming their glycinergic nature. Application of 10 μM of gelsemine to these neurons generated a trend towards lower frequency of glycinergic mIPSCs (Figure 6C). Using 50 μM of gelsemine, all the neurons displayed a significantly diminished frequency of glycinergic mIPSCs and an altered cumulative probability of the inter-event intervals of the glycinergic mIPSCs (Figure 6A–D). Under these conditions, only 4/10 of the neurons tested displayed synaptic events in the presence of the alkaloid. Interestingly, the average amplitude and the cumulative probability of these remaining miniature postsynaptic events were not significantly affected by gelsemine (Figure 6E–F). Analyses of the glycinergic mIPSC kinetics revealed that gelsemine did not modify the rise time (Figure 6G–H). Gelsemine (50 μM) induced a shorter decay time, by nearly ≈25%, which, however, was not statistically significant ( P = 0.16, ANOVA followed by Bonferroni post hoc test).

    Details are in the caption following the image
    Gelsemine modulation of glycinergic neurotransmission. (A). Examples of current traces showing the glycinergic synaptic activity before and during the application of 50 μM gelsemine from a single spinal neuron. (B). The scatter graph depicts the effect of 50 μM of gelsemine on the frequency of synaptic events in 10 spinal neurons. The alkaloid completely abolished the presence of glycinergic mIPSC (i.e. frequency value equals zero) in 6/10 of the neurons examined. (C). The graph summarizes the effects of 10 and 50 μM of the alkaloid on the average frequency of glycinergic mIPSCs. The average frequency of the glycinergic synaptic events was significantly diminished only by the application of 50 μM of gelsemine. * P < 0.05, significant difference; ANOVA followed by Bonferroni post hoc test; F (2,29) = 5.27. Control (n = 17), 10 μM (n = 8) and 50 μM (n = 10) of gelsemine. (D). Cumulative probability of the inter-event intervals of glycinergic mIPSCs in the absence or the presence of 50 μM of gelsemine. The distribution was significantly altered by the alkaloid. * P < 0.05; Kolmogorov–Smirnoff test. (E). The bar plot summarizes the effects of 10 and 50 μM on the amplitude of glycinergic mIPSCs. Differences were not significant. (F). Cumulative probability of the amplitudes of glycinergic mIPSCs in the absence or the presence of 50 μM of gelsemine. Differences were not significant. (G). Representative average current traces before and during the application of 50 μM of gelsemine. (H). The bar graphs summarize the effects of gelsemine on the mIPSC kinetics. Differences were not significant; ANOVA followed by Bonferroni post hoc test.

    In order to evaluate whether these synaptic effects are exclusive for glycine receptors, we next evaluated the sensitivity of the glutamatergic synaptic transmission to 10 and 50 μM of gelsemine (Figure 7). To this end, we recorded miniature glutamatergic excitatory postsynaptic currents (mEPSCs) from cultured spinal cord neurons at −70 mV using a low-chloride intracellular solution in the presence of extracellular TTX. Our results showed that the application of gelsemine did not modify the average amplitude of the glutamatergic events (Figure 7B). The cumulative probability of the mEPSCs amplitudes was also not altered (Figure 7C). Interestingly, all the neurons recorded displayed a significantly diminished frequency of synaptic glutamatergic events and an altered cumulative probability of the inter-event intervals of the glutamatergic mEPSCs in the presence of 50 μM of gelsemine (Figure 7D–F). Altogether, these data suggest that gelsemine can exert inhibitory effects on both glycinergic and glutamatergic neurotransmissions likely through presynaptic mechanisms.

    Details are in the caption following the image
    Modulation of glutamatergic neurotransmission by gelsemine. (A). Current traces showing the glutamatergic synaptic activity before and during the application of 50 μM gelsemine. (B). The bar plot summarizes the effects of gelsemine (10–50 μM) on the amplitude of glutamatergic mEPSCs. Differences were not significant. (C). Cumulative probability of the amplitudes of glutamatergic mEPSCs in the absence or the presence of 50 μM of gelsemine. Differences were not significant. (D). The scatter graph depicts the effect of 50 μM of gelsemine on the frequency of glutamatergic events on eight spinal neurons. (E). The graph summarizes the effects of 10 and 50 μM of the alkaloid on the average frequency of glutamatergic mEPSCs. The frequency of the synaptic events was significantly diminished by the application of 50 μM of gelsemine. * P < 0.05; ANOVA followed by Bonferroni post hoc test; (F (2,22) = 4.89. Control (n = 8), 10 μM (n = 5) and 50 μM (n = 8) of gelsemine. (F). Cumulative probability of the inter-event intervals of glutamatergic synaptic events in the absence or the presence of 50 μM of gelsemine. The distribution was significantly modified by the alkaloid. * P < 0.05; Kolmogorov–Smirnoff test.

    Discussion

    A limited number of molecules have been characterized as exogenous modulators of the functions of glycine receptors (Yévenes and Zeilhofer, 2011a). Here, we described the modulation exerted by the alkaloid gelsemine on recombinant and native glycine receptors. Our results indicate that gelsemine is a direct modulator of these receptors, displaying subunit-specificity and conformation-selective effects. The alkaloid exerted these actions by changing the apparent affinity and the open probability of the ion channel in a voltage-independent fashion. Results from spinal cord neurons showed that native glycine receptors were more sensitive to gelsemine modulation than GABAA and AMPA receptors. Furthermore, the alkaloid diminished the frequency of glycinergic and glutamatergic synaptic events without altering their amplitude.

    Our results demonstrated that gelsemine exerts subunit-specific and configuration-selective actions on glycine receptors. Previous reports have characterized subunit-specific actions of a variety of modulators on these receptors, such as several cannabinoid ligands and tropeines (Supplisson and Chesnoy-Marchais, 2000; Yang et al., 2008, Yévenes and Zeilhofer, 2011b). Electrophysiological studies, for example, have determined that the putative endocannabinoid N-arachydonoyl-glycine potentiated homomeric glycine receptors composed of the α1 subunit, but inhibited the function of glycine receptors composed of the α2 and α3 subunits (Yang et al., 2008, Yévenes and Zeilhofer, 2011b). In addition, modulatory actions displaying bell-shape profiles for Zn2+ ions and tropeines have been previously described (Bloomenthal et al., 1994; Chesnoy-Marchais, 1996; Laube et al., 2000). The influence of the β glycine receptor subunit on many of these pharmacological profiles however has received less attention. The evidence has shown that β subunits modified either the potency or the efficacy of several allosteric modulators (Miller et al., 2005; Liu et al., 2010) or did not influence the modulation (Pistis et al., 1997; Shan et al., 2001; Miller et al., 2005; Hejazi et al., 2006; Yang et al., 2008). Nevertheless, Supplisson and Chesnoy-Marchais (2000) have determined that β subunit expression switched the modulation by the tropeine tropisetron of α2 glycine receptors from inhibition to potentiation. Our results showed that homomeric glycine receptors composed of the α2 or α3 subunits were inhibited by the alkaloid gelsemine, whereas α1 glycine receptors showed a potentiation in the chloride current at concentrations below 100 μM. The potentiation, however, was switched to inhibition by the expression of the β subunits, indicating that homomeric α1 glycine receptors have a particular molecular configuration that is pivotal for the enhancement of the receptor function. Therefore, from our results, it is possible to suggest that at least a part of the selective potentiation of homomeric α1 glycine receptors by gelsemine may lie on unique elements present at α1–α1 interfaces. Because the β subunits can also shape glycine binding sites and negatively affect the gelsemine potentiation of α1 glycine receptors, it is possible that just the presence of α1–β interfaces in the receptor complex directly contributes to switch off the potentiation of the glycine-evoked currents. Thus, residues present in the β subunit would favour the gelsemine-induced inhibition of α1-containing glycine receptors. A more definitive explanation of the residues involved in the subunit-specificity of gelsemine is limited by the lack of knowledge regarding the gelsemine binding sites within glycine receptors. Nevertheless, due to their structural similitude with strychnine, it is likely that the gelsemine binding sites are in close association with the glycine and strychnine binding sites. This notion is supported by recent studies performed in spinal cord membranes, which showed that gelsemine displaced the 3H-strychnine binding (Zhang et al., 2013). Considering this evidence and the recently reported crystal structures of homomeric glycine receptors (Du et al., 2015; Huang et al., 2015), we can speculate that gelsemine binds to the glycine receptorin a manner similar to strychnine, likely in the same pocket or around its vicinity. However, from these crystallographic studies and the primary sequences of the different glycine receptor subunits, it can be inferred that the critical residues for strychnine binding in these structures are conserved between α1, α2 and α3 subunits. Assuming that the binding of gelsemine occurs on the same site, these observations thus suggest that the subunit-specific effects of gelsemine are possibly related with residues outside the strychnine binding site. These set of unrecognized residues may locally influence the orthosteric site or may generate differences in the gating events after gelsemine binding. The identification of these residues in future studies may contribute to elucidate the mechanistic framework behind gelsemine selectivity on different glycine receptors and possibly will foster the development of new glycine receptor modulators.

    By studying cultured spinal cord neurons, our results demonstrated that gelsemine reduced the frequency of glycinergic and glutamatergic synaptic events. Because gelsemine did not affect the amplitude of the miniature synaptic events, it is possibly that the gelsemine effects involve the modulation of a presynaptic target. Our results analyzing glycine-evoked currents of recombinant and native glycine receptors strongly suggest that one of the possible gelsemine-sensitive elements at the presynaptic site is the glycine receptor. On the other hand, the results obtained by analyzing the AMPA and GABA-evoked currents in the presence of gelsemine partly discard a main role of presynaptic AMPA or GABAA receptors on the gelsemine modulation of glycinergic and glutamatergic activities. The existence of presynaptic glycine receptors has been well demonstrated by several groups. Studies in several CNS regions have shown that the activation of presynaptic glycine receptors can enhance the frequency of synaptic events elicited by glutamate, GABA and glycine synaptic release in a neuron-specific manner (Turecek and Trussell, 2001; Jeong et al., 2003; Kunz et al., 2012; Choi et al., 2013). The increase on the neurotransmitter release is mainly due to a glycine-evoked activation of chloride channels, which depolarizes the axon terminal (Turecek and Trussell, 2001). Thus, it seems likely that a selective modulation of presynaptic glycine receptors may regulate the frequency of glycinergic or glutamatergic events in other systems. Indeed, a recent study coincidentally showed that an exogenous cannabinoid ligand was able to potentiate presynaptic homomeric glycine receptors, enhancing the frequency of glycinergic synaptic events in a mouse model of hyperkeplexia (Xiong et al., 2014). In addition, others have suggested that ethanol increased the frequency of glycinergic mIPSCs through presynaptic mechanisms in spinal cord neurons (Mariqueo et al., 2014). Although these evidences and our observations certainly suggest that the modulation of presynaptic glycine receptors may underlie at least a part of the gelsemine effects at the synaptic level, the molecular identity of such presynaptic receptors in our neuronal preparation is currently unknown. In addition, besides the possible effects on presynaptic glycine receptors, our results cannot rule out the possibility that gelsemine may bind and regulate other unrecognized protein targets, such as presynaptic GPCRs (Pernia-Andrade et al., 2009; Zeilhofer et al., 2012b). Thus, our observations at this point should be interpreted with care and more experiments are needed to determine the precise mechanisms underlying the modulation of glycinergic and glutamatergic synaptic transmission by gelsemine.

    Whether the electrophysiological evidences described above can directly explain the actions of gelsemine in animals and humans is also a complex matter. The extracts of Gelsemium genus plants of the Loganiaceae family have been used for many years for a range of medicinal purposes (Dutt et al., 2010; Jin et al., 2014). However, several reports have also shown that these preparations are toxic to animals and humans (Rujjanawate et al., 2003; Dutt et al., 2010; Jin et al., 2014). Our results from the present study may help to understand the toxic actions of gelsemine in a more direct way. Gelsemine intoxication is characterized by the presence of dyspnoea, convulsions, tremors, respiratory arrest and death. Therefore, the effects of high gelsemine concentrations (>100 μM) described in our experiments correlate well with a global depression of glycinergic activity and with the toxic effects of the alkaloid in animals and humans (Rujjanawate et al., 2003; Jin et al., 2014).

    On the other hand, the beneficial actions of gelsemine are more complex to explain and are currently under investigation. Our study nevertheless still can shed some light on the current understanding of these actions. The beneficial actions of gelsemine include anti-proliferative activity (Zhao et al., 2010), as well as anxiolytic (Liu et al., 2013; Meyer et al., 2013) and analgesic actions (Zhang et al., 2013). The anxiolytic actions of gelsemine are difficult to explain because glycine receptors have been not traditionally linked with anxiety disorders. However, experiments performed in spinal cord slices have shown that gelsemine can stimulate the endogenous production of neurosteroids, such as 3α,5α-tetrahydroprogesterone (3α-5α-THP; also known as allopregnanolone), through a mechanism dependent on glycine receptor activation (Venard et al., 2008). Coincidentally, recent experiments in mice have found that low doses of strychnine largely attenuated the gelsemine-induced anxiolytic activity (Liu et al., 2013). These results thus suggest that the glycine receptor activity can regulate the production of endogenous neurosteroids that subsequently enhance the function of GABAA receptors. Based on our results, and because the anxiolytic effects of gelsemine were blocked by strychnine, we can hypothesize that the potentiation of α1 glycine receptors is the most plausible mechanism by which glycine receptors can modulate the production of neurosteroids. However, the cellular pathways involved in the neurosteroid synthesis in the CNS are complex, and they can be synthesized either by neurons or by glial cells (Belelli and Lambert, 2005). In addition, neurosteroid levels are dynamic and are influenced by several environmental factors, such as aging, pregnancy, stress or abuse drug consumption. Furthermore, the function and the expression pattern of homomeric α1 glycine receptors are not well described (Lynch, 2009), and the modulation of different GABAA subtypes by gelsemine has not yet been addressed. Despite these concerns, it is possible to suggest that at least some neurosteroid-producing cells are regulated by activity at glycine receptors. The interaction of gelsemine with such glycine receptors may directly modify the membrane potential of these cells, contributing to the stimulation of neurosteroid synthesis and the potentiation of the GABAA receptor function.

    The analgesic effects of gelsemine reported by Zhang and co-workers (Zhang et al., 2013) and the results of the present work showing a negative modulation of α3 glycine receptors by gelsemine are also difficult to reconcile in a straightforward way. There is good evidence that the prostaglandin EP2 receptor-mediated inhibition of synaptic glycine receptors containing the α3 subunit, in dorsal horn neurons has a critical role in the development and maintenance of inflammatory pain (Harvey et al., 2004). Positive allosteric modulators targeting α3 glycine receptors thus may restore the loss of inhibitory control in these neurons producing beneficial analgesic effects in chronic pain states of inflammatory origin (Xiong et al., 2012; Zeilhofer et al., 2012b). On the other hand, the critical role of α3 glycine receptors for the antinociceptive effects of gelsemine is based to date on genetic ablation experiments showing that the gelsemine-induced analgesic effects in a rat model of neuropathic pain were diminished after the i.t. injection of siRNA for α3 glycine receptors (Zhang et al., 2013). However, it should be noted that experiments using α3 glycine receptor knock-out mice have demonstrated that the EP2 receptor-mediated inhibition of α3 glycine receptors exerts a critical role in inflammatory pain (Harvey et al., 2004) but appears to be not relevant for the development of neuropathic pain hypersensitivity (Hösl et al., 2006). Taken together, these data demonstrate that the role of α3 glycine receptors in neuropathic pain is still not clear. Thus, the effects of α3 glycine receptor modulators in the context of neuropathic pain are too complex to interpret simply. In addition, the neurophysiological mechanisms underlying different forms of pain are diverse. For example, neuropathic pain occurs through an altered chloride homeostasis of dorsal horn neurons mediated by down-regulation of the KCC2 transporter, which can turn the GABA and glycine activity from inhibitory to excitatory (see Zeilhofer et al., 2012a). Furthermore, recent data have demonstrated the existence of LTP of glycinergic synapses on dorsal horn GABAergic neurons, which may contribute to the loss of inhibitory control at the dorsal horn and the development of chronic pain (Chirila et al., 2014). Taking in consideration these results, the inhibition of the α3 glycine receptor activity induced by gelsemine may control at least part of the altered glycinergic tone in neuropathic pain, contributing to the analgesic effects of the alkaloid reported in behavioural assays. All these points support the notion that additional electrophysiological and behavioural experiments (possibly on genetically modified mice) are necessary to clarify the role of different glycine receptor subunits on the mechanisms underlying the anxiolytic and analgesic effects of gelsemine in animals and humans.

    From the results reported here, we can conclude that the alkaloid gelsemine is a direct modulator of recombinant and native glycine receptors, which can exert conformation-specific and subunit-selective effects. In addition, the alkaloid negatively modulates the frequency of glycinergic and glutamatergic synaptic events. Hence, these data provide a pharmacological basis to explain, at least in part, the glycine receptor-dependent beneficial and toxic effects of gelsemine on animals and humans. In addition, the pharmacological profile of gelsemine and the molecular sites underlying the subunit-specific modulation may open new approaches to the development of subunit-selective glycine receptor modulators.

    Acknowledgements

    The authors thank Lauren Aguayo, Ixia Cid and Daniela Gonzalez for their outstanding technical assistance. This work was supported by FONDECYT 1140515 and VRID-Universidad de Concepcion 213.033.106-1.0 (to G.E.Y). Cesar O. Lara is supported by the Graduate School of the University of Concepcion (Master Program in Human Physiology).

      Author contributions

      C.O.L., P.M., B.M., A.M.M., L.S.M., V.S.M., C.F.B., T.A.M. and G.E.Y. performed experiments and analysed the data. G.E.Y., L.G.A., J.F., P.G. and L.G. designed the research. C.O.L., C.F.B., B.M. and G.E.Y. wrote the paper. All the authors read and commented on the paper.

      Conflict of interest

      The authors declare no conflicts of interest.

      Declaration of transparency and scientific rigour

      This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research recommended by funding agencies, publishers and other organizations engaged with supporting research.