Histamine (H3) receptors modulate the excitatory

Brain Research 1064 (2005) 1 – 9 www.elsevier.com/locate/brainres

Research Report

Histamine (H3) receptors modulate the excitatory amino acid receptor response of the vestibular afferents Hortencia Cha´vez a,b,*, Rosario Vega a, Enrique Soto a a

Instituto de Fisiologıá-BUAP, Universidad Autońoma de Puebla, Apartado Postal 406, Puebla, Pue. cp 72000, Me´xico b Facultad de Estomatologıá, Universidad Autońoma de Puebla, Me´xico Accepted 15 October 2005 Available online 28 November 2005

Abstract Although the effectiveness of histamine-related drugs in the treatment of peripheral and central vestibular disorders may be explained by their action on the vestibular nuclei, it has also been shown that antivertigo effects can take place at the peripheral level. In this work, we examined the actions of H3 histaminergic agonists and antagonists on the afferent neuron electrical discharge in the isolated inner ear of the axolotl. Our results indicate that H3 antagonists such as thioperamide, clobenpropit, and betahistine (BH) decreased the electrical discharge of afferent neurons by interfering with the postsynaptic response to excitatory amino acid agonists. These results lend further support to the idea that the antivertigo action of histamine-related drugs may be caused, at least in part, by a decrease in the sensory input from the vestibular endorgans. The present data show that the inhibitory action of the afferent neurons discharge previously described for BH is not restricted to this molecule but is also shared by other H3 antagonists. D 2005 Elsevier B.V. All rights reserved. Theme: Neurotransmitters, modulators, transporters, and receptors Topic: Other neurotransmitters Keywords: Betahistine; Thioperamide; Clobenpropit; Hair cell; R-a-methylhistamine; AMPA; Kainic; Quisqualic

1. Introduction In the vestibular endorgans, hair cells make synaptic contact with the afferent terminals. Efferent neurons originating from the brainstem exert control influences from the central nervous system to the hair cell and to the afferent endings, controlling the afferent flow of information. Efferent fibers synapse on type-II hair cells and on the calyx afferent endings of type-I hair cells. The efferent synapse has been shown to be mediated by acetylcholine (ACh) in both cases [19,20]. The calcitonin-gene-related peptide (CGRP), substance P, and opioid peptides have been shown to be released from efferent neurons [1,37].

* Corresponding author. Instituto de Fisiologıá-BUAP, Universidad Autońoma de Puebla, Apartado Postal 406, Puebla, Pue. cp 72000, Me´xico. E-mail address: [email protected] (H. Cha´vez). 0006-8993/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2005.10.027

Hair cells make synaptic contact with afferent endings transmitting information about displacement of their hair bundle secondarily to a mechanical stimuli. The synapse between hair cells and primary afferent endings has been shown to be mediated by excitatory amino acids (EAA) [2,39,40]. Various other substances including opioid peptides [44], nitric oxide [16], ATP [3,36], and histamine [7,21,23,41] have been shown to modulate the hair-cellafferent synaptic transmission, acting both at the pre- and the postsynaptic terminals. Histamine plays an important role as a neurotransmitter and neuromodulator in the central nervous system (CNS). Histamine receptors are classified into four subtypes: H1, H2, H3, and H4 receptors [22]. Betahistine (BH; N-a-methyl-2pyridylethylamine), a histamine-like substance, has been used in the treatment of vertigo, motion sickness, and various vestibular disorders of central and peripheral origin [18,24,25,31,34]. BH activity may be explained by its direct

2

H. Cha´vez et al. / Brain Research 1064 (2005) 1 – 9

interaction with histamine receptors on which BH has a complex function, as a partial agonist of H1 receptors and as an antagonist of H3 receptors [43]. Although the action of histamine-related drugs in the treatment of vestibular disorders may be explained by their influence on the histamine receptors in vestibular nuclei [42,46], it has been suggested that antivertigo effects of BH can also take place at the peripheral level by improving the microcirculation of the labyrinth [15,27,29] or by modulating the afferent neuron discharge of the semicircular canal ampullar receptors [7]. Histamine and other imidazole-containing substances were found to increase the ampullar nerve firing rate, whereas both H1 and H2 histamine antagonists effectively inhibited ampullar nerve activity [23]. Inhibition of histidine decarboxylase reduced ampullar nerve firing in a dose-dependent manner. These observations indicate a physiological role for histamine in the vestibular endorgans [23]. Previous reports have shown that BH produces an inhibitory action on the afferent discharge of the semicircular canals of the frog and of the axolotl [7,9,10]. Evidence indicates that BH significantly reduces the response of the primary afferent neurons of the vestibular system to EAA agonists, thus significantly decreasing the synaptic input from hair cells [41]. Our aim was to contribute to defining the mechanism of action of histaminergic drugs at the endorgans of the vestibular system.

The isolated vestibule was transferred to a recording chamber and continuously perfused with Ringer solution of the following composition (in mM): KCl 2.5, NaCl 111, CaCl2 1.8, MgCl2 1, glucose 10, HEPES 5, pH 7.4. Multiunit extracellular recordings were obtained from the lateral semicircular canal nerve using a suction electrode (Fig. 1). Electrical activity amplified by means of a conventional AC amplifier (P-15, Grass Inst., Boston, MA) was filtered at cutoff frequencies of 100 and 3000 Hz and monitored by using an oscilloscope (Tektronix, Beaverton, OR). The signal was also led to a magnetic tape recorder and to a window discriminator (Model 121, WPI, Sarasota, FL), the output of which was connected to a microcomputer for on-line analysis of the discharge rate [38]. In some of the experiments, the preparation was mechanically stimulated. For this purpose, the recording chamber, manipulators, and amplifier were mounted on a rotating table driven by a DC motor (Aerotech, Pittsburgh, PA). The output of a function generator (model 8904A, Hewlett-Packard, Palo Alto, CA) was fed to the motor controller to produce sinusoidal accelerations (0.2 Hz, 72- s 1) to the rotating table [40]. Typically, the preparation was stimulated during 30 s periods under control conditions and after drug administration. Mechanical stimulation of the semicircular canals allowed us to compare the response to exogenously applied substances, such as EAA agonists, with the response of the system to natural mechanical stimulation.

2. Material and methods Animal care and procedures were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and with the Health General Law Regulations for Research in Health edited by the Secretary of Health in Mexico (Reglamento de la Ley General de Salud en Materia de Investigacioń para la Salud of the Secretarıá de Salud). All experiments and procedures were reviewed and approved by the appropriate committee of the Institution. The number of animals used was kept to the minimum necessary for a meaningful interpretation of the data. Experiments were made in wild larval axolotl (Ambystoma tigrinum, 30 to 60 g body weight). A total of 146 experimental preparations were used for the whole experimental series. Animals were anesthetized by immersion in 3-aminobenzoic acid ethyl ester (0.1% in water) and subsequently decapitated. The otic capsule was opened ventrally and the inner ear structures identified as previously reported [40]. The nerve fibers originating from the anterior and of the lateral canals were identified. Fibers originating from the lateral canal were dissected and cut before their confluence with the utricular and anterior canal fibers. Then, the nerve fibers were dissected from the covering endothelium. The cartilaginous otic capsule, including the whole vestibular system, was cut and isolated from the skull.

Fig. 1. Recording of the electrical activity of the semicircular canal nerve in the isolated inner ear preparation. The otic capsule and the labyrinth were isolated from the skull and mounted in a recording chamber the lateral canal lying in the horizontal plane. The lateral canal nerve was sucked into a glass electrode connected to an AC amplifier, an oscilloscope, and a window discriminator. Spike activity of the semicircular canal nerve was analyzed online. In some experiments, the whole preparation including the amplifier and manipulators was mounted on a rotating table and stimulated with sinusoidal accelerations by means of a DC servocontrolled motor. LC, lateral canal; AC, anterior canal; PC, posterior canal; S, saccule; U, utricle; L, lagena.


Drugs were applied by bath perfusion or by microperfusion as indicated in each case. For bath application, the desired drug concentration was added to the perfused Ringer solution, and the recording chamber bath was completely replaced using a rapid exchange suction –perfusion system. For microperfusion, 20 Al of the drug was pressure ejected over 2 s from a micropipette (100 Am tip diameter, flow rate 10 Al/s) positioned near (less than 500 Am) the origin of the afferent endings. Concentrations given herein are those originally in the pipette. Because the bath volume was 2 ml, the drug concentration decays exponentially to about 1% of its original value in a few milliseconds [47]. The microperfusion experiments allowed us to produce a concentration peak of the drug in the vicinity of the tip of the perfusion pipette. In contrast, bath application produced a slowly rising, sustained change in the concentration of the drug. The reason for using microperfusion is that this allows us to study EAA receptors avoiding desensitization phenomena that would otherwise be produced by bath application. No drug was tested unless the resting discharge was stable. When more than one drug was used, sufficient time (at least 10 min) was allowed to wash off the previously used drug, and no drug was used until the resting-discharge level returned to the control level. Kainic acid (KA), quisqualic acid (QA), thioperamide, clobenpropit, and Ra-methyl histamine were obtained from Sigma Chemical Co. (St. Louis, MO), a-amino-3-hydroxy-5-methyl-isoxale4-propionic acid (AMPA) was from Tocris Cookson (Bristol, UK), betahistine was from Solvay Pharmaceuticals (Solvay, France), and aminoethylpyridine (M1) was from Formenti (Formenti, Milan, Italy).

3

3. Results Recordings of the electrical discharge of the semicircular canal neurons were reliably obtained from the isolated inner ear of the axolotl. The H3 receptor antagonists inhibited the afferent discharge and the response of afferent endings to EAA agonists. 3.1. Betahistine and M1 concentration response curve The actions of BH and of its metabolite M1 on the electrical discharge of vestibular afferents were tested in concentrations ranging from 10 AM to 10 mM. As reported [7,41], BH inhibits the electrical activity of the vestibular afferent neurons, its effect being of higher potency on the basal resting discharge than on the mechanically evoked response, with an IC50 of 250 AM for the resting and 800 AM for the mechanically evoked discharge (n = 32; Fig. 2). Perfusion of the BH metabolite M1 [8,9] also produces a significant inhibition of the semicircular canal afferent neuron discharge. M1 was tested in concentrations ranging from 100 nM to 10 mM (n = 23). Although M1 action began at lower concentrations, M1 potency was similar to that of BH in inhibiting afferent discharge, with an IC50 for the resting discharge of 400 AM and for the mechanically evoked discharge 630 AM (Fig. 2). 3.2. Betahistine and excitatory amino acid receptor interactions From previous experiments, we have hypothesized that histamine receptor activation, particularly H3Ireceptors,

Fig. 2. BH and M1 actions on the electrical discharge of vestibular afferent neurons. In panels (A and B), frequency versus time plots showing the effects of M1 on the resting and mechanically evoked activity of the semicircular canal afferent neurons. Plots shows the resting activity and the response to mechanical stimulation under control conditions and after 5- and 10-min bath perfusion of 1 AM and 10 mM M1. In panel (C) is the dose response curve for BH and M1. Points represent the mean T standard error of at least 4 experiments. Data were fitted with a dose response function. Data are the mean inhibitory concentrations.

4


Fig. 3. BH and AMPA interactions. Panels (A and B) are the plots of the discharge frequency versus time of the response of the semicircular canal afferent neurons to microperfusion (20 Al) of 10 AM AMPA under control conditions (A) and after 10-min bath perfusion of 1 mM BH (B). In panel (C) is the bar graph showing AMPA excitatory action under control conditions and after 10-min bath perfusion of 1 mM BH. Bath application of BH reduced the response to AMPA by 37 T 1% (n = 3). ips: impulses per second.

interacts with the postsynaptic response of the EAA receptors in the afferent endings [41]. To test for this possibility, the interactions of BH and the EAA agonists AMPA, QA, and KA were studied. The bath perfusion of 1 mM BH for 10 min followed by the microperfusion of 10 AM AMPA produced a 37 T 1% (n = 3) reduction of the AMPA peak response (Fig. 3). Bath perfusion of 1 mM BH also reduced the response to 10 AM KA by 45 T 10% (n = 4) and 1 AM QA by 40 T 3% (n = 3). By bathing the preparation with high Mg2+ (10 mM), low 2+ Ca (0.01 mM) saline solutions, both the resting discharge and the mechanically evoked response of the afferent fibers were completely abolished. This procedure produced a physiological uncoupling of the hair cells from the afferent neurons. This effect is caused by an inhibition of the

presynaptic Ca2+ current in hair cells and the blocking of the subsequent neurotransmitter release. In this experimental condition, the application of AMPA or QA produced a strong excitatory response from the afferent neurons. Because neurotransmitter release is blocked, this is a pure postsynaptic effect. Bath perfusion of 10 AM BH reduced the response to 1 AM QA and 1 AM AMPA (Figs. 4A and B). Bath perfusion of BH in the range of 0.1 AM to 1 mM antagonized the response produced by AMPA application with an IC50 of 0.5 AM (n = 15, Fig. 4C). The inhibitory action of BH on the responses caused by AMPA and QA was dose-dependent, showing a statistically significant inhibitory effect (paired Student’s t test, P < 0.05) at concentrations of 10 AM for QA (n = 3) and of 1 AM for AMPA (n = 3).

Fig. 4. BH and AMPA interaction. In panels (A and B) are plots of the discharge frequency versus time of the semicircular canal afferent neurons. Preparation was bathed with a high Mg2+ (10 mM) low Ca2+ (0.01 mM) saline solution, which determines that there was no resting discharge from the afferent fibers. The microperfusion of 1 AM AMPA (black) or 1 AM QA produced a significant excitatory action that lasted for about 3 min. Bath application of 10 AM BH for 10 min produced a significant reduction in the response of afferent neurons to AMPA and to QA (gray). In panel (C) is the concentration response of BH inhibition of AMPA responses in the afferent neurons. Inhibitory action of BH on AMPA effects shows an IC50 of 0.5 AM. Each point represents the mean T SE of at least 3 experiments. ips: impulses per second.


3.3. Histaminergic (H3) actions Thioperamide and clobenpropit are two well-known H3 antagonists. To test whether the inhibitory action of BH on the EAA responses was caused by its antagonistic action on H3 receptors or mediated through some other mechanisms, the interaction of thioperamide and clobenpropit with EAA receptor agonists was also studied. Bath perfusion of thioperamide (0.01 to 100 AM, n = 33) produced a concentration-dependent inhibitory action on the resting discharge of the semicircular canal afferent neurons with an IC50 of about 100 AM and nonsignificant action on the mechanically evoked response (Figs. 5A and C). Inhibitory action of thioperamide on the resting discharge becomes significant for concentrations > 10 AM. Bath perfusion of clobenpropit for 10 min at different concentrations (0.01 and 10 AM, n = 44) produced significant effects on the resting discharge of the afferent neurons at concentrations above 0.01 AM with an IC50 of about 26 AM. For the mechanical response, inhibitory effect was significant for concentrations above 1 AM with an IC50 of about 100 AM (Figs. 5B and D). For EAA agonists, both thioperamide (n = 5) and clobenpropit (n = 8) at concentrations