[1] 14w VAGAL AFFERENT MODULATION OF SPINAL TRIGEMINAL NEURONAL RESPONSES TO DURAL ELECTRICAL STIMULATION IN RATS

[1] 254w Vagus nerve stimulation (VNS) is an approved antiepileptic and antidepressant treatment, which has recently shown promise as a therapy for drug-resistant primary headaches. Specific neurobiological mechanisms underlying its anticephalgic action are not elucidated, partly because of the deficiency of research-related findings. The spinal trigeminal nucleus (STN) plays a prominent role in pathophysiology of headaches by modulating pain transmission from intracranial structures to higher centers of the brain. To determine whether vagal stimulation may affect trigeminovascular nociception, we investigated the effects of VNS on the STN neuronal activity in the animal model of headache. In anesthetized rats the spike activity of the STN neurons with convergent orofacial and meningeal inputs was monitored, and the changes in neuronal responses to electrical stimulation of the dura mater under preconditioning or under continuous electrical stimulation of the left cervical vagus nerve were studied. Preconditioning vagal afferent stimulation (200-ms train of pulses at 30 Hz applied before each dural stimulus) did not produce substantial changes in the STN spike activity. However, continuous VNS with frequency of 10 Hz in 48% of cases significantly suppressed trigeminal neuronal responses to dural electrical stimulation. In line with the decrease in evoked activity, the VNS-induced depression of ongoing neuronal firing was observed. Although the inhibitory effect was prevailing, 29.5% of STN neurons were facilitated by VNS, whereas 22.5% were unresponsive to the stimulation. These results provide an evidence of VNS-induced modulation of trigeminovascular nociception, and therefore contribute to a deeper understanding of neurophysiological mechanisms underlying effects of vagal stimulation in chronic drugresistant headaches.

[1] 180w In the last two decades, the vagus nerve stimulation (VNS) has become the most widely used non-pharmacological treatment for refractory epilepsy and depression (Kosel and Schlaepfer, 2002;Albert et al., 2009;Bajbouj et al., 2010;Ruffoli et al., 2011). Furthermore, VNS is currently considered by clinicians as a valid therapy for the prevention of drug-resistant primary headaches (Lenaerts et al., 2008;Broggi et al., 2009Broggi et al., , 2010)). It has been reported that at least half of patients receiving VNS had reductions in frequency and intensity of their chronic cluster, migraine or daily headaches which were not only associated with epilepsy or depression (Sadler et al., 2002;Hord et al., 2003;Cecchini et al., 2009), but also referred to as primary disorders (Mauskop, 2005). Furthermore, a case has been described in which mechanical (prolonged finger pressure) stimulation of the cervical portion of the vagal nerve trunk was able to abort migraine attacks (Di Stani et al., 2007). Additionally, the use of VNS as an adjunctive therapy allowed decreasing the numbers and dosages of prophylactic anticephalgic drugs received by the patients (Hord et al., 2003;Mauskop, 2005).

[2] 139w The current explanations of the VNS-induced relief of primary headaches remain highly speculative and are mostly based on evidences of general antinociceptive action of VNS. The analgesic effect of the vagal afferent stimulation has been reported in numerous behavioral pain tests in laboratory animals. For instance, the VNS-induced inhibition of the nociceptive digastric reflex induced by intense tooth-pulp stimulation (Maixner et al., 1991;Bossut et al., 1992), reduction of the cumulative duration of rubbing and scratching the injection site in the orofacial formalin test (Bohotin et al., 2003b), and latency increase of the tail-flick or the hind paw withdrawal response to noxious heat (Ren et al., 1989;Aicher et al., 1991;Thurston and Randich, 1991;Bohotin et al., 2003a) were reported. In epileptic patients treated with VNS, an increase in mechanical pain threshold was noted (Kirchner et al., 2000(Kirchner et al., , 2006)).

[3] 124w The exact mechanisms by which stimulation of vagal afferents may reduce pain are not clear. Several neurophysiological studies reported that VNS had predominantly inhibitory effect on activity of lumbar and thoracic spinal neurons (Thies and Foreman, 1983;Hobbs et al., 1989;Ren et al., 1991;Evans et al., 1994). In addition, the vagal afferent stimulation has been found to suppress the spike frequency and c-fos expression occurring in trigeminal and trigeminothalamic neurons in response to noxious orofacial stimulation (Bossut and Maixner, 1996;Bohotin et al., 2003b). Furthermore, the VNS-induced inhibition of activity of tooth pulp-responsive units in the trigeminal nuclei oralis and caudalis (Takeda et al., 1998;Tanimoto et al., 2002), as well as in the ventral posteromedial nucleus of the thalamus (Nishikawa et al., 1999) has been reported.

[4] 96w These findings indicate that VNS may inhibit nociceptive neurotransmission both at the spinal and at the supraspinal level. However, the neurobiological mechanisms underlying the anticephalgic action of VNS remain unclear. Currently it is hypothesized that VNS, through widespread ascending pathways of the nucleus of the solitary tract, could alter activity of pain-modulating brain structures responsible for headache (Hord et al., 2003;Mauskop, 2005;Lenaerts et al., 2008;Broggi et al., 2009Broggi et al., , 2010)). Direct experimental evidence for this concept is lacking, mainly due to the deficiency of animal model studies pursuing understanding of VNS action in headache.

[5] 90w Several lines of evidence suggest a prominent role of the spinal trigeminal nucleus (STN) in pathophysiology of headaches. The second-order nociceptive neurons in the STN are shown to be intimately involved in pain transmission from intracranial structures to higher centers of the brain (Goadsby, 2005;Goadsby et al., 2009;Messlinger, 2009). To date, no studies have addressed a question whether vagal afferent stimulation modifies the activity of the STN neurons associated with meningeal nociception. Such data could provide an important contribution to understanding of neurophysiological mechanisms underlying VNS action in drug-resistant headaches.

[6] 71w Therefore, in the present work we investigated the effects of VNS on the STN neuronal activity in the animal model of headache. In order to determine whether vagal stimulation may affect trigeminovascular nociception, we monitored the spike activity of the STN neurons with convergent orofacial and meningeal inputs and studied the changes in their responses to electrical stimulation of the dura mater under electrical stimulation of the left cervical vagus nerve.

[1] 84w Extracellular recordings were made from 52 neurons within the caudal part of the STN (Fig. 1A). The recorded cells were located in the region of the nucleus defined by a rostrocaudal direction from 0.5 to 1.5 mm caudal to the obex and mediolaterally from 2.0 to 2.5 mm left to the middle line at the depth of 0.3-1.3 mm from the dorsal surface of the spinal cord. Nineteen (36.5%) neurons were located in lamina III, and 33 cells (63.5%) were located in laminae IV-V.

[2] 70w All the neurons received convergent afferent inputs from the ipsilateral dura mater and facial skin; their cutaneous mechanoreceptive fields were located in the periorbital area (N = 23), on the vibrissa pad (N = 17) and on the upper lip (N = 12; Fig. 1B). All 52 spinal trigeminal neurons responded to both noxious (pinch) and nonnoxious (touch) mechanical stimulation and were therefore classified as wide-dynamic range neurons (Fig. 1C).

[3] 94w Recorded units showed a wide range (1-23 spikes/s) of frequencies of initial ongoing activity (Fig. 1D). The mean rate of ongoing firing was 8.2 ± 1.3 spikes/s (N = 52). All tested neurons showed an excitatory response to electrical stimulation of the dura mater with the mean latency of 11.9 ± 0.4 ms (N = 52), corresponding to activation of Ad-fibers (Fig. 1E). The number of spikes in a single response varied between 3 and 6 per one trial. Initially, the mean rate of evoked firing was 4.4 ± 0.4 spikes/stimulus (N = 52).

[4] 171w Continuous stimulation of the vagus nerve in 21 (48%) of 44 tested trigeminal neurons caused a pronounced (more than 20%) decrease in their electrically evoked activity (Figs. 2A, 4A). According to the Friedman test, the VNS-induced changes in responses of these cells to dural stimulation were highly significant (P = 0.002, Fr = 12.1). Under vagal stimulation, the mean rate of evoked firing decreased to 62 ± 4% of the baseline level in this group (2.2 ± 0.3 spikes/stimulus versus 3.7 ± 0.4 spikes/stimulus in baseline, N = 21, P < 0.0001, Wilcoxon signed rank test; Fig. 4A). The mean latency of the responses did not change significantly (12.4 ± 0.9 ms under the VNS versus 11.6 ± 0.6 ms initially, N = 21, P = 0.46, Wilcoxon signed rank test). Five minutes after the end of the vagal stimulation, electrically induced neuronal activity raised up to 89 ± 14% (N = 21) and was not significantly different from the initial level (P = 0.23, Wilcoxon signed rank test; Fig. 4A).

[5] 174w In 17 (81%) of 21 trigeminal neurons with a decrease in responses to electrical stimulation of the dura mater under continuous VNS, the vagal-induced suppression of ongoing activity was also observed (Fig. 3A). Out of the remaining 4 neurons, the ongoing activity was facilitated by VNS in 2 neurons and in 2 it did not change. In total, the ongoing firing in this neuronal group was substantially suppressed by vagal stimulation (P = 0.0012, Fr = 13.4, Friedman test). Under the VNS, the mean discharge rate significantly fell to 59 ± 13% of the initial level (5.1 ± 1.2 versus 10.0 ± 1.9 spikes/s initially, N = 21, P = 0.002, Wilcoxon signed rank test; Fig. 4A). In contrast to the transient modulation of neuronal responses, the VNS-induced suppression of ongoing activity in these cells was long-lasting. Five minutes after the end of the vagal stimulation, the discharge rate was still 51 ± 17% (N = 21) and remained significantly lower than the initial level (P = 0.013, Wilcoxon signed rank test; Fig. 4A).

[6] 157w In the other 13 (29.5%) out of 44 tested trigeminal cells, the continuous VNS caused substantial facilitation of responses to dural electrical stimulation (Fig. 2B, 4B). The excitatory changes of evoked activity in this neuronal group reached the highest level of significance (P = 0.0002, Fr = 16.9, Friedman test). Under the VNS, the mean rate of dural stimulation-induced firing increased up to 153 ± 7% of the baseline level (7.6 ± 1.2 versus 4.9 ± 0.8 spikes/stimulus in baseline, N = 13, P = 0.0002, Wilcoxon signed rank test; Fig. 4B). The mean latency of the responses did not change significantly (12.7 ± 0.7 ms under VNS versus 12.5 ± 0.7 ms initially, N = 13, P = 0.57, Wilcoxon signed rank test). Five minutes after the VNS, the evoked neuronal firing declined to 99 ± 8% (N = 13) and was comparable to the initial level (P = 0.77, Wilcoxon signed rank test; Fig. 4B).

[7] 142w Eleven (85%) neurons of the latter group, under continuous VNS also showed a facilitation of ongoing activity (Fig. 3B). In the remaining 2 units, the ongoing activity was either inhibited or unaffected by the vagal stimulation. In total, the ongoing firing in this neuronal group was substantially enhanced (P = 0.011, Fr = 8.9, Friedman test). During the vagal stimulation, the mean discharge rate increased up to 264 ± 49% of the initial level (18.8 ± 2.5 versus 7.2 ± 1.9 spikes/s initially, N = 13, P = 0.0078, Wilcoxon signed rank test; Fig. 4B). However, the VNS-induced enhancement of neuronal activity was short-lasting. Five minutes after the end of the vagal stimulation, the ongoing firing fell to 102 ± 33% (N = 13) and was not significantly different from the baseline level (P = 0.81, Wilcoxon signed rank test; Fig. 4B).

[8] 113w Ten (22.5%) out of 44 tested trigeminal neurons were considered unresponsive to the continuous vagal stimulation. The changes in their reactions to the electrical stimulation of the dura mater under the VNS did not exceed 12%. In total, the initial evoked firing in this group (4.4 ± 0.3 spikes/stimulus, N = 10) was not significantly altered by the vagal stimulation (4.3 ± 0.4 spikes/stimulus, N = 10, P = 0.81, Fr = 0.67, Friedman test; Fig. 4C). Similarly, the VNS had no effect on the ongoing activity of these cells (9.5 ± 2.3 versus 9.8 ± 2.9 spikes/s in baseline, N = 10, P = 0.14, Fr = 4.0, Friedman test; Fig. 4C).

[9] 79w It should be noted that the three groups of trigeminal neurons-the ones inhibited by the vagal stimulation, the activated and the unresponsive ones-did not show differences either in their baseline electrically evoked activity (P = 0.20, KW = 3.21, Kruskal-Wallis test) or in the initial rate of ongoing firing (P = 0.06, KW = 5.62, Kruskal-Wallis test). Moreover, cells belonging to the inhibited, activated or unresponsive group did not demonstrate any specific for the group location within the STN.

[1] 99w The presented study demonstrates for the first time that continuous stimulation of the left cervical vagus nerve produces predominantly inhibitory effect on responses of the convergent spinal trigeminal neurons to electrical stimulation of the dura mater. Simultaneously with the decrease in evoked activity, the majority of tested trigeminal units demonstrated VNS-induced depression of ongoing firing. The observed effects depended on mode, duration and frequency of vagal stimulation: preconditioning VNS with a frequency of 30 Hz for 200 ms was less effective in modulating spinal trigeminal neuronal activity than continuous stimulation of the same intensity with frequency of 10 Hz.

[2] 180w The electrophysiological model of craniovascular nociception used in this study is based on modern knowledge about structural and functional organization of the trigeminovascular system and its role in the pathogenesis of primary headaches. The dura mater and large cerebral vessels innervated by thin Ad and C trigeminal afferents are known to be the main sources of pain in headaches (Ray and Wolff, 1940;Goadsby et al., 2009;Messlinger, 2009). Consistent with this, electrical stimulation of the superior sagittal sinus in rats and cats resulted in both disturbance of cerebral blood flow and changes in plasma levels of neuropeptides similar to those observed in migraine or cluster headache attacks and was also shown to increase c-fos expression and spike activity in the STN, a segmental structure playing a prominent role in the modulation of nociceptive transmission from intracranial structures (Goadsby and Knight, 1997;Hoskin et al., 1999;Storer and Goadsby, 1999;Edvinsson and Uddman, 2005;Akerman et al., 2007;Goadsby et al., 2009). Therefore, electrical stimulation of the dura mater proves to be a valid method of activating the trigeminovascular system and mimicking nociceptive processes occurring during headache.

[3] 130w The model of craniovascular nociception is easily integrated into pre-clinical screening; it allows detection of drug action with proposed or clinically proven anticephalgic activity. Previously it has been shown that some antimigraine drugs of different pharmacological classes, such as triptans (Goadsby and Knight, 1997), non-steroidal anti-inflammatory drugs (Jakubowski et al., 2005(Jakubowski et al., , 2007;;Sokolov et al., 2010), and calcitonin gene-related peptide receptor antagonists (Storer et al., 2004a) produce considerable inhibition of spike activity of the spinal trigeminal neurons, indicating a key aspect of their pharmacodynamics. In the present study, the use of non-pharmacological but clinically proven effective physiotherapeutic technique resulted in analogous changes in trigeminal neuronal activity. These findings allow us to compare different headache treatment methods and, furthermore, suggest similar neurophysiological and neurochemical mechanisms of their anticephalgic effects.

[4] 149w In our study, almost half of the spinal trigeminal neuron responses to dural electrical stimulation were inhibited by the VNS. This finding is in agreement with the results of previous studies, which demonstrated the predominance of inhibitory effects of VNS on the spinal trigeminal neurons. In particular, it was shown that in cats continuous VNS inhibited rather than facilitated or did not affect responses of the spinal trigeminal neurons to noxious thermic or electrical stimulation of the facial skin or the tooth pulp (Bossut and Maixner, 1996). A similar VNSinduced attenuation of the tooth pulp stimulation-evoked activity of the spinal trigeminal neurons was observed in rats (Takeda et al., 1998;Tanimoto et al., 2002). Moreover, it has been demonstrated that in rats electrical stimulation of the vagus nerve significantly diminished the increase of c-fos expression induced in the STN by formalin injection into the vibrissa pad (Bohotin et al., 2003b).

[5] 181w The present study revealed for the first time that vagal stimulation altered the ongoing and evoked firings of the spinal trigeminal neurons in the same direction, mostly inhibited. However, while the VNS-induced suppression of neuronal responses was transient, inhibition of ongoing discharges lasted more than 5 min, suggesting different mechanisms of modulation of evoked and ongoing activity. Taking into account that ongoing activity of the trigeminal neurons depends primarily on sensory signaling from periphery (Roch et al., 2007), the observed reduction in ongoing neuronal firing may be considered as a sign of VNS-induced suppression of sensory inputs from various extra-and intracranial receptive fields, not only those located in the dura mater. Taken together, the ongoing and evoked activities are known to constitute an integral indicator of neuronal excitability; thus the VNS-induced decrease in both of them is thought to reflect the fact of neuronal inhibition. Thus, the present study provides the first evidence of VNS-induced inhibition of the spinal trigeminal neurons receiving meningeal inputs; as such it directly contributes to the understanding of mechanisms of the VNS therapeutic efficacy in headaches.

[6] 131w Although the predominant effect of VNS in our experiments was inhibitory, about one-third of the tested units demonstrated the excitatory effect of vagal afferent stimulation on their dural stimulation-evoked and ongoing firings. Similar VNS-induced excitation of spinal trigeminal neurons has been reported previously. It was shown that electrical stimulation of vagal afferents induced an enhancement of spike activity of these cells in monkeys (Chandler et al., 1999) and rats (Fu et al., 1992;Tanimoto et al., 2002;Mørch et al., 2007), as well as activated c-fos protein expression in the ipsilateral STN in rats (Bohotin et al., 2003b). Consistent with this, the unilateral vagotomy in rats resulted in a reduction of c-fos immunoreactivity observed in the bilateral STN following unilateral injection of complete Freund's adjuvant into the masseter muscle (Imbe et al., 1999).

[7] 162w The dual action of VNS on activity of spinal trigeminal neurons and, as a result, on pain sensation could probably be explained by the different threshold currents necessary to recruit myelinated and nonmyelinated vagal fibers, which have different effects on activity of central structures. It has been suggested that low-intensity stimulation of vagal afferents produces pronociceptive effects via activation of rapidly conducting pathways, whereas highintensity stimuli produce antinociceptive effects via activation of slowly conducting pathways (Ren et al., 1991;Ness et al., 2000). However, this relationship between VNS intensity and facilitatory/inhibitory effects appeared not to be consistent. For instance, in our experiments the intensity of VNS (100-300 lA) was below the threshold for nonmyelinated vagal fibers (per review, Ruffoli et al., 2011), suggesting that both inhibitory and facilitatory effects resulted from activation of myelinated fibers. Presumably, an explanation for this may be found in the delicate neurophysiological mechanisms of trigeminovagal convergence, which, in turn, may become affected by the descending pain control systems.

[8] 197w It seems possible that anatomical substrate enabling VNS effects on the spinal trigeminal neuronal activity may be represented by reciprocal connections between brainstem sensory nuclei of the vagus and trigeminal nerves. It has been shown that apart from integration to the dorsal horn of the upper cervical cord, the STN provides numerous bilateral inputs to the nucleus of the solitary tract, an integrative relay of facial (VII), glossopharyngeal (IX) and vagal (X) afferents (Ruggiero et al., 2000;Zerari-Mailly et al., 2005;Noseda et al., 2008). In addition to trigemino-solitarii projections, the solitarii-trigeminal pathway has been described, providing evidence for the existence of a reciprocal trigemino-solitarii-trigeminal feedback loop (Zerari-Mailly et al., 2005). Some trigeminal afferents were also reported to terminate directly on neurons of the nucleus of the solitary tract, whereas the spinal trigeminal neurons receive sensory input from the facial, glossopharyngeal and vagus nerves (Chandler et al., 1999;Imbe et al., 1999;Panneton et al., 2000;Sessle, 2000;Zhang and Ashwell, 2001;Bohotin et al., 2003b;Mørch et al., 2007). Thus, in addition to being innervated by trigeminal, facial, glossopharyngeal and vagal afferents, the STN and the nucleus of the solitary tract are highly interconnected and thus form an integrative neurosensory network, the trigemino-solitary complex.

[9] 115w In the present study, the short preconditioning vagal afferent stimulation, despite relatively high frequency (30 Hz), did not induce substantial changes in the dural stimulation-induced activity of the tested units. These results lead us to speculate that VNS effects on activity of the trigeminal neurons most likely depend on the involvement of the brainstem pain-modulating mechanisms rather than depending directly on the trigeminovagal interaction at the segmental level. We presume, that the short preconditioning VNS did not show efficacy due to its inability to engage supraspinal structures. In turn, the 10-Hz continuous stimulation, the duration of which could be long enough to activate the higher order brain levels, produced substantial effects on trigeminal neuronal firing.

[10] 141w Second-order neurons of the trigemino-solitary complex are known to provide substantial inputs to the rostroventral medial medulla, parabrachial complex, locus coeruleus, raphe nuclei, periaqueductal gray matter, and hypothalamus (Sessle, 2000;Benarroch, 2006;Noseda et al., 2008;Liu et al., 2009;Ruffoli et al., 2011), which, in turn, are involved in the descending modulatory control of trigemino-vascular nociception (Millan, 2002;Goadsby et al., 2009;Messlinger, 2009). It has been postulated that VNS, through disynaptic glutamatergic pathway via the nucleus paragigantocellularis, increased the firing activity of norepinephrine-producing neurons in the locus coeruleus and, subsequently, that of the dorsal raphe serotonergic cells, presumably resulting in the enhanced release of corresponding neurotransmitters (Groves et al., 2005;Dorr and Debonnel, 2006;Albert et al., 2009;Manta et al., 2009;Ruffoli et al., 2011). In addition, VNS has been shown to increase the level of free gamma-amino butyric acid (GABA) (Ben-Menachem et al., 1995;Albert et al., 2009).

[11] 141w All the discussed data allow us to speculate that VNSinduced reduction in responses of the STN cells to the dural electrical stimulation observed in our experiments may be the result of the activation of the brainstem serotonergic and/or noradrenergic descending inhibitory pathways and subsequent increase of GABA neurotransmission within the trigemino-solitary complex. It was shown previously in a similar experimental model that some GABA-positive drugs, such as midazolam, baclofen, muscimol, and GABA itself caused inhibition of glutamate-and dural electrical stimulation-evoked activity of the spinal trigeminal neurons (Storer et al., 2001(Storer et al., , 2004b)). At the same time, the VNS-induced facilitation of nociceptive transmission in the spinal trigeminal neurons may be mediated by the activation of the pain facilitatory mechanisms arising from the rostral ventromedial medulla (Thurston andRandich, 1992, 1995). However, further studies are needed to confirm or refute this assumption.

[12] 111w Thus, stimulation of vagal afferents can produce a wide range of biological effects resulting in dual action on nociceptive processing. This study demonstrates the predominantly inhibitory effect of the left cervical VNS on activity of the second-order sensory trigeminal neurons in the rat model of headache, thus providing evidence of VNS-induced attenuation of the trigeminovascular nociceptive transmission. The described vagal afferent modulation of the spinal trigeminal neurons may play a role in the ascending visceral regulation of processing of sensory information from the dura mater. We suppose that the data obtained can contribute to a deeper understanding of neurophysiological mechanisms underlying the therapeutic efficacy of the VNS for chronic drug-resistant headaches.

[1] 118w Twenty adult male Wistar rats (body weight 300-380 g) were used for the study. The animals were housed in the vivarium of the Pavlov Institute of Physiology (St. Petersburg, Russia) and maintained 2-5 animals per cage on a 12-h light/dark schedule with free access to food and water. All experiments were performed according to the Ethical Guidelines of the International Association for the Study of Pain and European Community Council Directive (86/609/EEC). The study protocol and experimental design were approved by the Institutional Animal Care and Use Committees of the Pavlov Institute of Physiology and the Saint-Petersburg Pavlov State Medical University. All efforts were made to reduce the number of animals used and to minimize any possible suffering.

[2] 233w Neuronal activity was recorded by varnish-insulated tungsten microelectrodes (Science Products, Hofheim, Germany) with a tip diameter of 5 lm and a resistance of 12 MX. The electrodes were lowered into the left STN at the level of C1 spinal cord in 4lm steps using a microdrive unit. The signals from the recording electrode were amplified and passed to the analogue input of the computer A/D converter by means of the multifunctional acquisition card (sampling period 25 ls). For online acquisition, processing and displaying the data, the custom-written software was used. To isolate the activity of single units from stimulus artifacts, adjacent cell potentials and noise, three-level amplitude discrimination was used online. Recordings of neuronal activity were analyzed as peristimulus time histograms, such that signals gated through the amplitude discrimination were collected in successive bins of 1 ms. The histograms had a sweep length of 500 ms and were created automatically from 50 recordings (one per 1 s). Ongoing activity (if analyzed) and electrically evoked responses of neurons were estimated within 250 ms before and 50 ms after stimulation of the dura mater, correspondingly. Apart from responses to the dural electrical stimulation, all recorded units were tested for responses to mechanical stimulation of their dural and facial cutaneous receptive fields by von Frey filaments (North Coast Medical, Morgan Hill, CA, USA). Only neurons demonstrating all three kinds of responses were selected for further testing.

[3] 87w Recordings of neuronal activity with simultaneous creation of peristimulus time histograms were performed before, under stimulation of the left vagus nerve and 5 min after the stimulation. As a rule, 2-3 neurons were tested in one animal. At the end of the experiment, rats were euthanized by an overdose of urethane (>3 g/kg, i.v.). The last recording sites within the spinal cord were marked by an electrolytic lesion through the recording electrode. After routine histological processing of the tissue, lesion sites were examined under a light microscope.

[4] 101w Using peristimulus histograms, ongoing neuronal activity and electrically evoked responses were expressed as a mean number of spikes per second (spikes/s) or a mean number of spikes per stimulus (spikes/stimulus), respectively. The electrically induced discharges were quantified by subtracting ongoing discharges from evoked activity. To estimate VNS-induced changes in neuronal activity, the ongoing firing and evoked responses under/after the vagal stimulation were normalized and expressed as percentages of the respective mean values prior to the vagal stimulation. If the changes relative to the baseline activity were more than 20% (in either direction), a given neuron was considered responsive to vagal stimulation.

[5] 77w Based on the results of the Shapiro-Wilk test of normality, the nonparametric Friedman, Wilcoxon signed rank and Kruskal-Wallis tests were used to determine the significance of changes in neuronal activity induced by the VNS. Statistical significance was set at P < 0.05. The data were expressed as the mean value ± SEM. The analysis was carried out using the Origin 7.5 (OriginLab, Northampton, MA, USA) and GraphPad InStat 3.02 (GraphPad Software, La Jolla, CA, USA) software packages.

[1] 123w Rats were anesthetized with urethane (1.5 g/kg, i.p; ICN Biomedicals, Aurora, OH, USA). Experimental procedures used were described in detail previously (Lyubashina and Panteleev, 2009;Sokolov et al., 2010Sokolov et al., , 2012)). Briefly, the rat under the surgical level of anesthesia was placed on a thermostatically controlled heating pad. Catheters were placed into the femoral vein for administration of anesthetics and myorelaxants, and into the femoral artery for continuous monitoring of blood pressure. The trachea was intubated and the head of the animal was fixed in a stereotaxic frame. The cervical portion of the left vagal nerve trunk was isolated and cut; its central end was displaced dorsally, put on silver bipolar stimulating electrodes and kept moist with warm paraffin oil (37 °C).

[2] 164w The neck muscles overlying the cisterna magna were separated along the midline and C1 laminectomy was performed. The dura mater was removed to expose the medulla and C1 spinal cord. A longitudinal parietal craniotomy was performed, and the bipolar stimulating electrodes were placed on the dura mater in close proximity to the superior sagittal sinus or visible blood vessels. The electrodes had resistance of 50 KO and consisted of two varnish-insulated silver wires with beads (0.3 mm in diameter) at the end. The animal was paralyzed using the pipecuronium bromide (i.v., 1.2 mg/kg initially, maintenance 0.6 mg/ kg as required; Gedeon Richter, Budapest, Hungary) and artificially ventilated with room air (75-100 cycles/min, 2-3 ml per cycle) using a small animal ventilator. Rectal temperature was maintained between 37 and 38 °C. The depth of anesthesia was assessed by monitoring blood pressure responses to noxious stimulation; supplementary anesthetic was administered when necessary to ensure the absence of gross (>20% from the baseline level) blood pressure fluctuations.

[3] 158w The dura mater was stimulated using single rectangular pulses of 300-800 lA (15-40 V) with a duration of 0.8 ms delivered by a computer-controlled stimulator. The stimulus intensity was 1.5 times the response threshold. The intensity of the current used to stimulate the vagus nerve was approximately 0.8 of its threshold to induce 10-15% changes (typically decrease) in arterial blood pressure and was not more than 350 lA. Most frequently, the rectangle current pulses of 100-300 lA (5-15 V) with a duration of 0.5 ms were used. In order to compare the post-effect of a short high-frequency stimulus train with the action of repetitive stimulation of a lower frequency, two corresponding settings of VNS were tested: (1) preconditioning stimulation-200-ms train of pulses at 30 Hz stopped 50 ms before each dural electrical stimulus-and (2) continuous single pulse stimulation with frequency of 10 Hz applied in parallel to dural stimuli. Only one VNS protocol was explored in each experiment.

[4] 158w Preconditioning VNS did not produce substantial changes in neuronal responses to electrical stimulation of the dura mater. Out of 8 units tested in these experiments, 5 showed a slight decrease in the mean discharge rate to 89 ± 4% of the baseline level in this group (4.5 ± 0.5 spikes/stimulus under the VNS versus 5.0 ± 0.6 spikes/ stimulus in baseline, N = 5). In the remaining 3 neurons, the mean number of evoked spikes increased up to 109 ± 4% of the value prior the vagal stimulation (4.9 ± 0.9 spikes/stimulus versus 4.5 ± 0.9 spikes/stimulus initially, N = 3). In the each group of cells, the difference did not exceed 15% and was not statistically significant (P > 0.05, Wilcoxon signed rank test). Five minutes after preconditioning VNS, the mean rates of evoked firing in the first and second groups of tested neurons were comparable to their baseline levels (P > 0.05, Wilcoxon signed rank test).