All experimental procedures were performed after approval from The University of Queensland Animal Ethics Anatomical Biosciences Committee (approval SBMS/142/15/NHMRC), and were in accordance with the Australian Code for the Care and Use of Animals for Scientific Purposes, 8th Edition (National Health and Medical Research Council, 2013), as well as the Queensland Government Animal Research Act 2001, and associated Animal Care and Protection Regulations (2002 and 2008).

Wistar rats (7–14 day old, either sex) were used. Sodium pentobarbitone (60–80 mg/kg, i.p.) was administered intraperitoneally to anaesthetize the animals; the skull was rapidly removed and the brainstem dissected out into ice-cold high Mg²⁺ artificial cerebrospinal fluid (aCSF), which contained (in mM) 130 NaCl, 26 NaHCO₃, 3 KCl, 5 MgCl₂, 1 CaCl₂, 1.25 NaH₂PO₄, 10 D-glucose (Bellingham and Berger, 1996; Bellingham, 2013). All solutions were bubbled with carbogen (95% O₂ + 5% CO₂) to maintain a pH of 7.4. The rostral brainstem end was fixed on a chuck using cyanoacrylate glue, with the ventral surface of the brainstem supported by an agar block glued to the chuck. The chuck with tissue was submerged in ice-cold high (5 mM) Mg²⁺ aCSF within a chamber, surrounded by an ice/water slurry. Transverse brainstem slices (300 μm thickness) were cut using a vibratome (Leica VT 1200 S, Leica Biosystems). Each rat yielded 3-4 slices. After cutting, slices were transferred to a holding chamber filled with the same high Mg²⁺ aCSF at 37°C, in a water bath for 45–60 min. Finally, slices were transferred to a new holding chamber containing recording aCSF (composition as above, except for 2 mM Ca²⁺, 1 mM Mg²⁺) and equilibrated for at least 30 min before recording at room temperature (21–22°C).

Individual slices were placed in a recording chamber (volume 200 μL) and continuously superfused (1–2 ml/min) with recording aCSF throughout the experiment. HMNs were identified using a 60X water-immersion objective, an infrared video camera (Hamamatsu, Japan) and Nomarski optics on a Nikon E600FN microscope (Tokyo, Japan); all HMNs were ventral and lateral to the central canal, within the hypoglossal motor nucleus, and had a large soma. Whole cell patch clamp recordings were made using an Axopatch 1D amplifier (Axon Instruments, Foster City, CA, USA). Synaptic current recordings were sampled at 10 kHz, and low-pass filtered at 2 kHz, while a higher sampling rate (50 kHz) and low-pass filter (10 kHz) setting were used for action potential recordings; data was sampled and stored on a Windows XP computer, using PClamp 10.2 software and a Digidata 1332A digitizer (Axon Instruments). After advancing the electrode to the cell body, so that electrode solution ejected by positive pressure produced a small dimple, positive pressure was released and mild suction was applied until a stable membrane seal (resistance >2 GΩ) was achieved. Stronger suction was then applied until the cell membrane was broken. Cells were voltage-clamped at membrane potentials of −60 mV.

The patch pipette internal solution for recording excitatory and inhibitory post-synaptic currents (EPSCs/IPSCs) contained (in mM) 120 CsCl, 4 NaCl, 4 MgCl₂, 0.001 CaCl₂, 10 Cs N-2-hydroxyethyl-piperazine-N'-2-ethanesulfonic acid (HEPES), 10 caesium ethylene glycol-bis(β-aminoethyl ether)-N,N,N,N-tetra-acetic acid (EGTA), pH adjusted to 7.2 with CsOH. The internal solution for recording action potentials contained (in mM) 135 K methyl sulfate, 8 NaCl, 10 HEPES, and 0.3 EGTA; pH 7.2 with KOH. Pipette solution osmolarity was 290–300 mOsM, measured using a vapor pressure osmometer (Wescor). 3 adenosine 5′- triphosphate (ATP-Mg) and 0.3 guanosine 5-triphosphate-tris(hydroxymethyl) aminomethane (GTP-Tris) were added to the internal solution just before use. For recording glutamatergic EPSCs, strychnine HCl (20 μM) was added to all external recording solutions to block both glycine and GABA_A receptor-mediated inhibitory synaptic currents in HMNs. Similarly, for recording glycinergic IPSCs, NBQX (10 μM), APV (50 μM), and bicuculline (5 μM) were added to the recording buffer. To isolate miniature synaptic currents, tetrodotoxin (TTX) at a final concentration of 1 μM was added to the recording buffer. For evoked EPSC recordings, a bipolar concentric stimulation electrode (Frederick Haer) was placed in the reticular formation near the ventrolateral border of the hypoglossal nucleus, and a pair of stimulus currents of 0.5–1.1 mA and 0.1 ms duration was applied to evoke an EPSC with consistent first pulse EPSC amplitudes; inter-stimulus interval was 150 ms (Bellingham, 2013). Action potential (AP) firing was achieved by repetitive depolarizing current steps (2–120 pA) of 400 ms duration from a membrane potential of −65 mV in current clamp mode (Bellingham, 2013). Input resistance was calculated from the steady state amplitude of a brief (40 ms) negative (−10 pA) current step prior to the depolarizing step.

All chemicals were from Sigma Aldrich (St Louis, MO, USA) except tetrodotoxin (TTX, Alomone Labs, Israel). The final concentrations of drugs were prepared just before each experiment. Capsaicin was dissolved in ethanol, capsazepine in methanol and SB366791 in DMSO to make stock solutions, which were diluted using recording aCSF to final concentrations of 10 μM (capsaicin and capsazepine) and 20 μM (SB366791). We chose these concentrations to achieve maximal capsaicin responses (Gavva et al., 2004) and for comparison to other electrophysiological studies of capsaicin effects. The reservoir containing SB366791 was protected from light by aluminum foil. Final concentration of all solvents was ≤ 0.2%.

Spontaneous and miniature synaptic currents were analyzed from one or more continuous data recordings of 2 min. Shape parameters (amplitude, 10–90% rise time, half-width) analysis was done in Clampfit 10.2 (Axon Instruments), using a template created by averaging visual detected individual synaptic events. Baseline holding current (I_holding) and frequency of EPSCs for spontaneous recordings and input resistance for evoked recordings were also recorded. Paired-pulse ratio (PPR) was calculated as the mean peak amplitude of the averaged second evoked EPSC divided by the mean peak amplitude of the averaged first evoked EPSC. For action potential analysis, action potentials were detected and their frequency and shape parameters measured using Clampfit.

Results are expressed as mean±SD, unless otherwise stated, and changes were determined to be statistically significant at P < 0.05 by a paired two-tailed t-test, except where indicated, using Prism 7 (GraphPad). The cumulative frequency distributions of synaptic currents was compared using a two-sample Kolmogorov-Smirnov test.

Earlier reports have suggested that capsaicin increases glutamatergic synaptic transmission in central neurons (Marinelli et al., 2002, 2003; Sikand and Premkumar, 2007). We therefore examined whether capsaicin altered spontaneous glutamatergic excitatory synaptic transmission (Figure 1A, sEPSCs) to HMNs. During capsaicin application (9–11 min, 10 μM), there was a significant increase in sEPSC frequency and amplitude (Figures 1B–G, Table 1). Increased sEPSC frequency is illustrated by the representative traces in Figure 1B. Figure 1C represents averaged spontaneous EPSCs from the three time periods shown, illustrating increased sEPSC amplitude with capsaicin and a partial washout of this effect. Mean sEPSC amplitude increased from −20.3 to −25.3 pA (+24% of control, p = 0.018, n = 7; Figure 1E). Mean sEPSC frequency was also increased from 1.34 to 2.89 Hz (+115% of control, p = 0.013; Figure 1D). A washout of 15 min partially restored sEPSC amplitude and frequency toward control values. The distribution of sEPSC amplitude showed a shift to higher amplitude events in the presence of capsaicin (Figure 1F). Similarly, the distribution of sEPSC inter-event intervals shows a significant shift toward smaller inter-event intervals (equating to higher frequency), which returned toward control values upon washout (Figure 1G). Other sEPSC shape parameters, such as half-width (Figure 1H) and 10–90% rise time (Figure 1I) remained unchanged during capsaicin application. Baseline holding current (I_holding), however, did show a significant inward shift from −95.75 to −116.6 pA (+21% of control, p = 0.0003, Figure 1J, Table 1), and this effect also partially recovered after washout.

Capsaicin-induced changes in sEPSCs can either be an effect on action potential-independent spontaneous release of glutamate, or may, in part, be mediated via action potential-mediated release of neurotransmitters. Hence, to further test the effect of capsaicin on excitatory synaptic transmission in the absence of spontaneous action potential generation, we analyzed miniature EPSCs (mEPSCs) where, in addition to strychnine, 1 μM tetrodotoxin (TTX) was added to completely block action potentials (Figure 2A). Mean mEPSC frequency was still significantly increased by capsaicin, from 2.65 to 7.5 Hz (+183% from control, p = 0.0026, n = 10, Figure 2B, Table 2). However, there was no change in mEPSC amplitude after capsaicin application (Figure 2E). Figures 2C,D show the distribution of mEPSC amplitudes and inter-event intervals in control and capsaicin conditions. The amplitude distribution shows no significant change (Figure 2C), however, inter-event interval distribution shows a significant shift toward higher frequency (Figure 2D, Kolmogorov-Smirnov test). As for sEPSCs, there were no significant changes in mEPSC half-width (Figure 2F) and 10–90% rise time (Figure 2G). Baseline holding current (I_holding) again showed a significant inward shift from −55.99 to −81.72 pA (+45% of control, p = 0.013, n = 10, Figure 2H, Table 2). Together, these effects of capsaicin on sEPSCs and mEPSCs show that capsaicin acts by increasing the release probability of pre-synaptic terminals or by modulating the excitability of presynaptic neurons making excitatory inputs to HMNs.

Next, we investigated if the actions of capsaicin are mediated via TRPV1 activation, by testing whether pre-application of the TRPV1 antagonist capsazepine (Urban and Dray, 1991) could block the effect of capsaicin on glutamatergic excitatory transmission to rat HMNs. Earlier reports have suggested that application of capsaicin in brain slice preparations (Marinelli et al., 2002) and cultured neurons (Sikand and Premkumar, 2007) increased excitatory transmission through activation of TRPV1 receptors and that this effect was abolished by capsazepine. Application of capsaicin in the presence of capsazepine (Figure 3A) still increased sEPSC frequency (+123% of control, p = 0.0038, n = 9; Figure 3D, Table 3) and I_holding (+25% of control, p = 0.0047, n = 9; Figure 3J, Table 3), but had no effect on amplitude (Figure 3H). Figure 3B represents the increase in frequency after addition of capsaicin in presence of capsazepine. Figure 3C shows averaged amplitude under 3 different conditions (control, capsazepine and capsazepine + capsaicin) respectively. It was interesting to note that capsaicin in presence of capsazepine increases the half-width of spontaneous currents (+30% of control, p = 0.032, n = 9; Figure 3E). The amplitude distribution of sEPSC showed a significant shift toward lower amplitude upon capsazepine application, but showed no further change after capsaicin application (Figure 3F). Also, the distribution of inter-event intervals showed no change between control and capsazepine, but showed a significant shift toward shorter inter-event intervals upon addition of capsaicin (Figure 3G). Capsazepine alone did not cause any significant effect on sEPSC frequency (Figure 3D), half-width (Figure 3E), rise time (Figure 3I) and I_holding (Figure 3J); however, there was a significant decrease in sEPSC amplitude (−19% of control, p = 0.028, n = 9, Figure 3H, Table 3), indicating a partial tonic effect of TRPV1 on sEPSC amplitude.

We therefore used a structurally different, selective, and potent TRPV1 antagonist, SB366791 (Gunthorpe et al., 2004; Varga et al., 2005), in order to further test the involvement of TRPV1 receptors in the actions of capsaicin. Earlier reports have shown that SB366791 antagonized capsaicin-induced changes in brain slice preparations (Gunthorpe et al., 2004; Varga et al., 2005). As for capsazepine, SB366791 (Figure 4A) did not alter the effect of capsaicin on sEPSC frequency (+94% of control, n = 6, p = 0.063; Figure 4C, Table 4), but did block the effect of capsaicin on sEPSC amplitude (Figure 4B) and I_holding (Figure 4H). The averaged sEPSC amplitude distribution showed no significant change between either control-SB366791 or control-capsaicin conditions (Figure 4D). Similarly, the averaged inter-event intervals distribution showed no change with SB366791, but showed a significant shift in events toward higher frequency when capsaicin was added (Figure 4E). No significant changes were observed in sEPSC half-width (Figure 4F) and 10–90% rise time (Figure 4G) under all conditions tested. By contrast to capsazepine, which decreased sEPSC amplitude, SB366791 by itself had no effect on sEPSC amplitude, indicating absence of any tonic SB366791-sensitive activity.

As capsaicin increased both sEPSCs and mEPSC frequency, capsaicin may act via a pre-synaptic mechanism that is not mediated via classic TRPV1 activation. To further differentiate between modulation of activity-dependent and -independent transmitter release, we analyzed the effects of capsaicin on evoked excitatory post-synaptic currents (eEPSCs). Two stimuli in the reticular formation lateral to the hypoglossal motor nucleus evoked a pair of short-latency multi-synapse excitatory currents (average first eEPSC amplitude = −237.9 ± 68.8 pA, n = 11) separated by 150 ms. As shown in Figure 5, the first eEPSC before and after application of capsaicin showed no change in amplitude (Figure 5B, Table 5), half-width (Figure 5C) and rise time (Figure 5D), suggesting that activity-dependent release is not altered by capsaicin. This interpretation was strengthened by a lack of effect of capsaicin on the PPR of the first and second eEPSCs (Figure 5F). As for our observations of sEPSCs and mEPSCs, there was a negative shift in I_holding while recording eEPSCs, but we found no associated change in HMN input resistance (Figure 5G).

The effects of capsaicin on sEPSCs and mEPSCs strongly suggests it acts by increasing the release probability of pre-synaptic neuron terminals, with little postsynaptic effect on HMN excitability, although capsaicin consistently evoked an inward current in all conditions tested. To directly measure the effect of capsaicin on post-synaptic HMN excitability, we evoked repetitive action potential (AP) firing from HMNs current clamped at −65 mV, using a series of progressively increasing depolarizing current steps (Figure 6A, Table 6). Capsaicin did not change AP firing frequency in HMNs (n = 9, Figures 6B,C), nor did capsaicin change the number of APs evoked by maximum current injected in individual cells (Figure 6E). Analysis of shape parameters of APs under control and capsaicin conditions showed no significant change in AP amplitude (Figure 6D), rise slope (Figure 6F), half-width (Figure 6G) and after-hyperpolarization amplitude (Figure 6H). Capsaicin also had no effect on the minimum amount of current required to generate the first AP under both conditions tested (Figure 6J). However, it was interesting to note that under the recording conditions used for APs (holding potential, −65 mV and potassium methyl sulfate-based internal solution), there was no significant shift in I_holding (Figure 6I).

Our observation that sEPSC and mEPSC frequency was consistently increased by capsaicin raised the question of whether this effect was specific to glutamatergic presynaptic terminals or whether glycinergic synaptic terminals were also modulated. To investigate this, HMNs were voltage-clamped at a membrane potential of −60 mV and were recorded in aCSF containing a cocktail of NBQX (10 μM), APV (50 μM) and bicuculline (5 μM) to isolate glycinergic sIPSCs. During capsaicin application (9–11 min, 10 μM), there was a significant decrease in sIPSC amplitude (Figures 7A,B, Table 7), without any effect on sIPSC frequency (Figure 7C). Decreased sIPSC amplitude is illustrated by representative continuous traces in Figure 7A. Figure 7B shows averaged sIPSCs from control and capsaicin conditions, illustrating the decreased amplitude with capsaicin (−62% of control, p = 0.0009, n = 11; Figure 7D). Mean sIPSC half-width was also significantly reduced (−45% of control, p = 0.0009, n = 11; Figure 7G, Table 7). Figure 7E shows the average distribution of sIPSC amplitude, indicating a shift to smaller amplitude events in the presence of capsaicin. The average distribution of sIPSC inter-event intervals (Figure 7F) showed no significant shift between control and capsaicin conditions consistent with an unchanged frequency. Other sIPSC parameters, such as rise time (Figure 7H) and baseline holding current (I_holding) (Figure 7I), were unchanged with capsaicin application, although I_holding, was shifted toward more positive levels, opposite to the change observed during sEPSC recordings.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.