Background
Orofacial somatosensory signals are transmitted to the first central relay site within the trigeminal sensory nuclear complex of brainstem via afferent component of the trigeminal nerve. The trigeminal complex consists of the principal nucleus and the spinal trigeminal nucleus. The latter is subdivided into oralis (Vo), interpolaris (Vi) and caudalis (Vc) in a rostro-caudal direction. The Vo receives somatosensory inputs from various oral and perioral structures including tooth [
1,
2], periodontal tissue[
3], lip [
4] and skin [
5,
6]. Recently, it has been suggested that the Vo is implicated in the central processing of nociceptive information owing to containing convergent and nociceptive-specific neurons, the neurons changing their properties by intramuscular mustard oil [
5] and subcutaneous formalin [
7]. Moreover, some Vo neurons transmit the processed somatosensory signals into other nociception-related brainstem areas, such as the parabrachial nucleus [
8] and the thalamus [
9,
10]. Therefore, Vo is an important functional brainstem area participating in the processing and transmission of nociceptive information, resembling the deep laminae of the Vc and the spinal cord dorsal horn [
6,
8].
In the central nervous system (CNS), glutamate mediates fast excitatory synaptic transmission or exerts slow synaptic effects by binding its ionotropic (iGluR) or metabotropic (mGluR) glutamate receptors [
11]. The mGluRs are one of G-protein coupled receptor families, enclosing eight subtypes (mGluR1-8) that can be divided into three groups according to sequence homology, pharmacology and signaling mechanisms [
12].
Group I mGluRs (mGluR1 and 5) activate phospholipase C (PLC) via Gq-protein, resulting in phosphoinositide hydrolysis, Ca
2+ release from inositol 1,4,5-triphosphate (IP
3)-sensitive intracellular stores, and protein kinase C (PKC) activation by diacylglycerol (DAG) [
12]. On the contrary, group II (mGluR2 and 3) and III (mGluR4, 6-8) mGluRs are negatively coupled to cAMP production pathway via Gi/Go-protein [
12]. Group I mGluRs are expressed in perisynaptic region of postsynaptic dendrites, and can mediate slow excitatory synaptic transmission in the CNS, including the spinal cord [
13,
14]. A previous study showed the expression of mGluR1 and mGluR5 in the trigeminal system including the subnucleus Vo [
15]. mGluR1 was immunostained in the neuropil of all the trigeminal nuclei, and mGluR5-immunoreactive neurons are distinguishable in Vo. To date, although the precise synaptic localization and synaptic function of mGluR1 and 5 have been widely studied in various brain regions, including the spinal cord dorsal horn [
16‐
21], the functions of group I mGluRs in the regulation of glutamatergic synaptic transmission in the Vo [
15] have not been studied yet. Hence, we here attempted to study the functional roles of group I mGluRs in regulating glutamate release and modulating an α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptor-mediated synaptic responses by recording spontaneous excitatory postsynaptic currents (sEPSCs) and trigeminal tract (Vt) stimulation-evoked EPSCs from the Vo neurons in the horizontal brainstem slices. In addition, we examined signaling pathways responsible for the group I mGluRs-mediated synaptic regulation.
Discussion
The present study demonstrated that the bath application of DHPG markedly increased the frequency of sEPSCs recorded in Vo neurons, consequently indicating a facilitation of glutamate release by the activation of group I mGluRs from presynaptic terminals. Even though inhibitory effects of group I mGluRs on glutamate release have been widely discovered in the various brain regions [
16], including the hippocampus [
27] and the spinal cord [
28], there have been few studies reporting the facilitating effects [
16,
29]. Therefore, the present study provides a new example of the facilitating effect of group I mGluRs on glutamate release in the CNS. Although the sources of glutamate were not clear since sEPSCs recorded from the Vo are mainly occurred by spontaneous unsynchronized glutamate release from terminals of other brainstem neurons, local Vo interneurons or trigeminal ganglion neurons, we found partial but significant mediation of mGluR1 or mGluR5 for the presynaptic facilitating effect of DHPG. In addition, the DHPG-induced facilitating effect was not dependent on the activation of NMDA receptors but voltage-dependent Na+ channels and CaMKII. The dependence of voltage-dependent Na+ channels indicates that an action potential-dependent mechanism for glutamate release is for the most part responsible for the DHPG-induced facilitating effect. Further, the interesting finding shown in the present study is that PLC and CaMKII mediate the facilitating effect of DHPG on the sEPSC frequency, whereas PKC negatively mediates it because its inhibition markedly enhances the facilitating effect of DHPG. The enhancement of DHPG effect caused by PKC inhibition involves mGluR1, rather than mGluR5. Our data together reveal that spontaneous glutamate release in Vo is strongly regulated by activation of group I mGluRs depending on activated signal transduction pathways.
AMPA receptors, iGluRs, consist of heteromeric assemblies of four different subunits GluR1-4, and mediate fast excitatory synaptic transmission at glutamatergic synapses in the CNS [
30]. Diverse electrophysiological studies have indicated that the AMPA receptors are modulated by activation of group I mGluRs [
16]. Group I mGluRs-mediated modulation of AMPA receptors includes both potentiation and depression, and is achieved by various intracellular signal transduction molecules downstream to group I mGluRs [
12,
31]. In this study, we demonstrated that DHPG-induced long-lasting potentiation of AMPA receptors-mediated sEPSC amplitude was completely blocked only when both mGluR1 and mGluR5 were blocked, indicating that a single subtype of group I mGluR is enough for the postsynaptic potentiating effect. These results has been demonstrated in the spinal dorsal horn neurons [
29], and are comparable with earlier studies showing potentiaiton of AMPA, an agonist of AMPA receptors, -induced responses by group I mGluR activation [
32‐
35]. Interestingly, albeit the potentiation during DHPG application was not dependent on NMDA receptors and PKC but on PLC, IP
3 receptors, NOS, CaMKII and ERK, the long-lasting property of the potentiation was disappeared by the blockades of NMDA receptors, as well as individual antagonists used to block signal transduction pathways. This result indicates that the long-lasting potentiation requires a full set of signal transduction molecules recruited after the activation of group I mGluRs.
Although the synaptic localization of mGluR1 and mGluR5 has been known in the spinal cord dorsal horn [
20,
21], it has not been clearly demonstrated in the Vo. Compelling anatomical and electrophysiological studies have indicated the presence of mGluR1 and mGluR5 in the neuropil of all the trigeminal nuclei [
15]. In the present study, the immunohistochemical data demonstrate cell body-like immunostaining of mGluR1 and diffuse staining of mGluR5 in the Vo area. In addition, the electrophysiological data put forward the expression of both subtypes on postsynaptic membrane and at presynaptic terminals in the Vo area, respectively, because of no significant blockade of sEPSC amplitude and significant but partial inhibition of the DHPG-induced increase of sEPSC frequency by either LY367385 or MPEP. On the other hand, a previous western blot analysis revealed the abundant expression of mGluR5 in the trigeminal ganglion but almost no expression of mGluR1, suggesting that the subtype of group I mGluR expressing at central terminals of trigeminal primary afferents in the Vo is exclusively mGluR5 [
36]. Therefore, the DHPG-induced increase of sEPSC frequency, found in the present study, is probably originated from the activation of both mGluR1 and mGluR5 at terminals of other brainstem neurons or local Vo neurons, as well as the activation of mGluR5 at trigeminal afferent terminals in Vo [
36]. This interpretation was also supported by the result that MPEP even more effectively inhibited the DHPG effect on sEPSC frequency (Fig.
2D). Accordingly, mGluR5 possibly mediates the regulatory effect of DHPG on the synchronized glutamate release from central trigeminal primary afferents, although the effect of MPEP on the DHPG-induced increase of PPR has not been tested in the present study.
Although other signaling molecules or channels, which were not studied here, may be involved in the DHPG effects [
16], some signal transduction pathways relating to synaptic regulation by mGluR1 and mGluR5 are sequentially represented in the present study. Particularly, the present study demonstrates CaMKII as a positive regulator and PKC as a negative regulator in the presynaptic terminals of Vo region. Because PKC inhibition caused enhancement of the DHPG-induced facilitation of glutamate release, it could be postulated that the activation of PKC might render feedback inhibition to the group I mGluR activation by certain mechanisms, for instance, a desensitization [
17,
37‐
39] which terminates further activation of mGluR1 and/or 5 and thereby prevents the facilitation of glutamate release. In the postsynaptic dendritic spine, the activation of both mGluR1 and mGluR5 elevates intracellular Ca
2+ concentration that is due to the IP
3-mediated intracellular Ca
2+ release [
40], which may lead to the activation of the Ca
2+-sensitive enzyme NOS [
22] and CaMKII, via the binding with Ca
2+-bound protein calmodulin, and then the positive modulation of AMPA receptor channels. The activation of NOS can activate NO-sensitive GC, converting GTP to cGMP [
41]. The cGMP may play a role in the long-lasting increase of AMPA receptor-mediated synaptic responses (Fig.
5E). On the other hand, in the postsynaptic dendritic spine, ERK may be activated by Gβγ or Homer proteins [
42], rather than a calcium-dependent manner, potentiating channel function of AMPA receptors responded to glutamate.
Previous studies have indicated significant roles of Vo brainstem area in the processing and the integration of somatosensory signals, including nociception [
8,
43‐
45]. It has been known that the Vo area contains convergent Vo neurons [
46], the neurons also called wide-dynamic range neurons and responded to both non-noxious and noxious stimuli. The convergent Vo neurons are likely to be processors of somatosensory signals as well as mediators of wind-up phenomenon [
47]. In addition to the convergent Vo neurons, neurons projecting their axons to the diencephalons have been found in Vo [
9,
10]. The diencephalic projections of Vo neurons were contralaterally or bilaterally reached to the various subnuclei of thalamus, such as ventral posteromedial nucleus and posterior thalamic nuclei, which involve the transmission of pain information into the higher brain areas. Therefore, the modulation of orofacial nociceptive signals in the Vo has functional significance in these contexts. In this study, we demonstrated the modulation of glutamate release and synaptic responses by the group I mGluR activation in the Vo. The modulation by group I mGluRs has been demonstrated in the central pain modulation area, such as the spinal cord dorsal horn. The activation of group I mGluR in the spinal cord dorsal horn induced long-lasting potentiation of the polysynaptic response [
18] and, to the contrary, long-lasting depression of the monosynaptic response [
18,
19]. These studies
in vitro have been correlated with other studies
in vivo demonstrating group I mGluRs-mediated nociceptive sensitization [
23,
24,
48,
49]. Thus, our present finding, i.e., group I mGluRs-mediated augmentation of synaptic inputs into the Vo neurons, may support the notion that the amplification of somatosensory signals from the periphery by central pain transmission neurons underlies persistent pain [
50].
In summary, we provided strong evidences that the activation of group I mGluR subtypes, mGluR1 and mGluR5, and their signal transduction pathways, differentially regulates glutamate release and AMPA receptor-mediated synaptic responses in the Vo region. These data will contribute to our understanding regarding the mode of the group I mGluR action adjusting brain functions such as orofacial normal sensation and pain.
Materials and methods
Experiments were approved by Institutional Animal Care and Use Committee of Kyungpook National University, and were carried out in accordance with the National Institute of Health guidelines for the Care and Use of Laboratory Animals.
Preparation of horizontal brain stem slices
Horizontal brainstem slices (400-450 μm) were prepared from 6-14 day-old Sprague-Dawley rats of either sex as described previously [
51]. Under deep urethane anaesthesia (1.5 g/kg, i.p.), the brain and part of the cervical spinal cord were removed after decapitation, and then immediately transferred into an ice-cold Krebs' solution (composition in mM: NaCl 117, KCl 3.6, CaCl
2 2.5, MgCl
2 1.2, NaH
2PO
4 1.2, NaHCO
3 25, and glucose 11; pre-oxygenated with a mixture of 95% O
2 and 5% CO
2; pH 7.4; CaCl
2 was substituted with MgCl
2 in case of the Ca
2+-free medium). In the pre-oxygenated ice-cold Krebs' solution, other brain parts except for the brainstem were eliminated. The brainstem was glued upside down on the flat top of a hard mounting cube block that was prefixed to the bottom of a slice chamber. Then, horizontal slices were cut using a Vibratome 1000
+ (Vibratome, St. Louis, MO, USA). Typically, the first cut of the ventral part of the brainstem was discarded, and two horizontal brainstem slices were obtained. The slices prepared were transferred into a fresh oxygenated Krebs' solution at room temperature, and incubated at least an hour for an incubation that may wash hazardous molecules that occurred during the preparation processes. Either the right or the left side of the slice was moved to a recording chamber, and submerged and fixed with nylon strands drawn taut across a C-shaped sliver wire (~0.5 mm o.d.).
Blind whole-cell voltage-clamp recordings
Blind whole-cell recordings with patch pipettes (borosilicate glass, TW150F; WPI, Sarasota, FL, USA) were made from Vo neurons. When viewed under a microscope (BX51WI, Olympus, Tokyo, Japan) with transmitted illumination (40×), the caudal border of the Vo area was distinguishable with the Vi area although it was difficult to discern with the rostral border of the Vo area [
51,
52]. Under visual guidance, the tip of the patch pipette was positioned in the Vo area above the caudal border with the Vi. The resistance of patch pipettes was typically 8-12 MΩ when filled with Cs-based (composition in mM: Cs
2SO
4 110, CaCl
2 0.5, MgCl
2 2, EGTA 5, HEPES 5, TEA chloride 5, ATP-Mg salt 5; Figs.
1,
2 and
3) or K-based (composition in mM: 145 K-gluconate, 5 NaCl, 1 MgCl
2, 0.2 EGTA, 10 HEPES, 2 Mg-ATP, and 0.1 Na
3-GTP; Figs.
3,
4,
5,
6 and
7) internal solutions (pH 7.2).
All recordings were made under a continuous perfusion of the preoxygenated Krebs' solution (2-3 ml/min), the same as the slice preparation solution, at room temperature (23-25°C). BMI (5 μM) and strychnine (1-2 μM) were always added to all the recordings to block inhibitory synaptic responses mediated by γ-aminobutyric acid type A (GABA
A) and glycine receptors, respectively. After whole-cell formation was identified by appearance of the capacitance transients upon voltage pulses (-5 mV), sEPSCs were recorded at a holding potential of -70 mV. In this recording condition, sEPSCs were completely blocked by 10 μM NBQX or 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), AMPA/kainate receptor antagonists (Fig.
1B), indicating that the synaptic responses are mediated by the AMPA/kainate receptors. In addition, EPSCs were evoked in Vo neurons by two electrical stimuli (spaced at 50 ms) of Vt, and analyzed in their amplitudes and in PPRs. PPR was calculated by dividing the second EPSC by the first EPSC in amplitude. Recordings were amplified with a Multiclamp 700A amplifier (MDS Inc., Toronto, Canada), and sampled at 5-10 KHz and filtered at 1-2 KHz. Series resistances are occasionally monitored in the beginning, throughout and at the end of experiments. The recordings were terminated or discarded if series resistance (about 8-20 MΩ) changed by more than 20%. Data acquisition was performed using pClamp software (version 10; MDS Inc., Toronto, Canada). After a stable baseline recording for 5-10 min, DHPG (10 μM), known as a selective group I mGluR agonist, was bath-applied for 5 min. Antagonists were added to the perfusing and also the DHPG-containing Krebs' solutions. Frequencies and amplitudes of sEPSC before and after bath-application of DHPG were analyzed using a template-matching method (Clampfit of pClamp software) followed by cut-off filtering of amplitude threshold (typically 2 - 3 pA). The analyzed data were represented as mean ± SEM (%) of the baseline (before DHPG) over time (bin, 30 sec). Statistical comparisons were made using Student's
t-test. Statistical significance, P < 0.05 or P < 0.01.
Immunohistochemical staining
Similar aged rats used in whole-cell recordings were perfused with 4% paraformaldehyde in 0.1 M phosphate buffered saline (PBS; pH 7.4), and their brainstems were collected. The brainstems were post-fixed overnight at 4°C, and cryoprotected overnight in 30% sucrose solution (in 0.1 M PBS) at 4°C. Cryostat sections (40 μm) were made in a similar way as the horizontal brainstem slice preparation for whole-cell recordings (see above). For 3,3'-diaminobenzidine (DAB) development, sections were incubated overnight at 4°C with either goat polyclonal anti-mGluR1a/b (1:50; sc-47131, Santa Cruz Inc., Santa Cruz, CA, USA) or rabbit polyclonal anti-mGluR5 (1:50; 06-451, Upstate, Lake Placid, NY, USA) as primary antibodies, followed by anti-goat IgG for 1 hr or anti-rabbit IgG for 2 hrs, respectively. Nuclei were stained with methyl green. For fluorescent staining, sections were incubated with anti-goat IgG-FITC for 2 hrs and anti-rabbit IgG-Cy3 for 2 hrs after incubating at room temperature with both primary antibodies (1:200 for both). The primary antibodies were not incubated in case of negative controls. Nuclei were stained with 4',6'-diamidino-2-phenylindole (DAPI) (data not shown).
Chemicals
Drugs were dissolved as a stock in distilled water or dimethylsulfoxide (DMSO), and diluted more than 1000 times to the final concentrations in the oxygenated Krebs' solution. Drugs and their sources were as follows: 2-APB, D-AP5, BMI, CNQX, DHPG, LY367385, MPEP, GF109203X, KN-62, L-NAME, PD98059, PTIO, ODQ, strychnine and U73122 from Tocris Cookson (Ellisville, MO, USA); TTX from Sigma (St. Louis, MO, USA).
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
All electrophysiological recordings were performed by JHS and SMH. JHS and ESP analyzed the electrophysiological records. ESP, SRH and DKA performed immunohistochemical experiments and analysis, or helped interpreting data. DHY designed the experiments, participated in data analysis and interpretation, and wrote this manuscript. All authors read and approved the final manuscript.