Neurosteroid dehydroepiandrosterone sulphate enhances pain transmission in rat spinal cord dorsal horn

Abstract

Background

The neurosteroid dehydroepiandrosterone sulphate (DHEAS) activates the sigma-1 receptor, inhibits gamma-aminobutyric acid A (GABA_A) and glycine receptors, and induces hyperalgesic effects. Although its effects have been studied in various tissues of the nervous system, its synaptic mechanisms in nociceptive pathways remain to be elucidated.

Methods

The threshold of mechanical hypersensitivity and spontaneous pain behaviour was assessed using the von Frey test in adult male Wistar rats after intrathecal administration of DHEAS. We also investigated the effects of DHEAS on synaptic transmission in the spinal dorsal horn using slice patch-clamp electrophysiology.

Results

Intrathecally administered DHEAS elicited dose-dependent mechanical hyperalgesia and spontaneous pain behaviours (withdrawal threshold: saline; 51.0 [20.1] g, 3 μg DHEAS; 14.0 [7.8] g, P<0.01, 10 μg DHEAS; 6.9 [5.2] g, 15 min after administration, P<0.001). DHEAS at 100 μM increased the frequency of miniature postsynaptic currents in the rat dorsal spinal horn; this increase was extracellular Ca²⁺-dependent but not sigma-1 and N-methyl-d-aspartate receptor-dependent. DHEAS suppressed the frequency of miniature inhibitory postsynaptic currents in a GABA_A receptor- and sigma-1 receptor-dependent manner.

Conclusions

These results suggest that DHEAS participates in the pathophysiology of nociceptive synaptic transmission in the spinal cord by potentiation of glutamate release and inhibition of the GABA_A receptor.

Editor's key points.

• •The neurosteroid dehydroepiandrosterone sulphate (DHEAS) induces hyperalgesic effects, but the receptor mechanisms for these effects are unknown.
• •In a rat model of neuropathic pain, intrathecally administered DHEAS elicited dose-dependent mechanical hyperalgesia and spontaneous pain behaviours in male rats.
• •The effects of DHEAS on synaptic transmission in the spinal dorsal horn slice showed increases in excitatory and decreases in inhibitory synaptic transmission.
• •These results suggest that DHEAS participates in the pathophysiology of nociceptive synaptic transmission in the spinal cord by potentiation of glutamate release and inhibition of gamma-aminobutyric acid A (GABA_A) receptors.

Neuroactive steroids (neurosteroids) are biosynthesised and metabolised in various tissues of the nervous system,¹ These steroids are known to have non-genomic effects and do not affect transcription and gene expression via steroid receptors on the nucleus membrane.² The non-genomic effects of neurosteroids involve various types of voltage- and ligand-gated ion channels.1, 3, 4, 5, 6, 7, 8, 9 Moreover, neurotransmitter release and synaptic transmission can be modulated by neurosteroids.8, 9, 10, 11, 12, 13, 14 Dehydroepiandrosterone (DHEA) and its sulphate derivative, dehydroepiandrosterone sulphate (DHEAS), are considered the most important neurosteroids and show multiple important memory-enhancing, antidepressant, and anxiolytic activities.7, 15, 16 A previous study showed that DHEAS increased the frequency of excitatory postsynaptic currents (EPSCs) in the prelimbic cortex¹⁷ and induced long-term potentiation-like effects in the hippocampus¹⁸; therefore, DHEAS may play important roles in synaptic plasticity.

DHEAS has also been studied extensively in relation to the pathophysiology of pain transduction and chronic pain modulation because of the abundance of enzymes that synthesise DHEAS (cytochrome P450 17α hydroxylase and sulphotransferase) in the spinal cord.19, 20, 21 DHEAS has pivotal effects on the pain state, including acute-phase pain-evoking effects19, 22, 23, 24 and late-phase pain-suppressive effects.23, 25 Yoon and colleagues24, 26 reported that intrathecally administered DHEAS decreased the threshold for mechanical and heat stimuli to the hindpaw of mice. These DHEAS-induced threshold reductions are inhibited by sigma-1 receptor antagonists and gamma-aminobutyric acid A (GABA_A) receptor agonists. This suggests that DHEAS can enhance pain, possibly via sigma-1 or GABA_A receptors in the dorsal horn of the spinal cord.

Sigma-1 receptors are putative receptors for several neurosteroids and are densely present in the spinal dorsal horn.²⁷ Intrathecally administered selective sigma-1 ligands can induce pain-like behaviours and can phosphorylate the GluN1 subunit of the N-methyl-d-aspartate (NMDA) receptor in the spinal dorsal horn.²² DHEAS-induced GluN1 phosphorylation can be suppressed by a sigma-1 receptor antagonist.26, 28 The NMDA receptor is well known as a key molecule that regulates pain sensation, and phosphorylation of GluN1 increases surface expression of NMDA receptors.²⁹ This evidence indicates that acutely administered DHEAS induces hyperalgesia because of the activation of the NMDA receptor through sigma-1 receptor activation.

However, the synaptic mechanisms of DHEAS in the nociceptive pathway remain to be elucidated. We re-analysed the effects of intrathecally administered DHEAS on the threshold of mechanical hypersensitivity and spontaneous pain behaviour and investigated the effects of DHEAS in synaptic transmission in the spinal dorsal horn in male rats.

Methods

Animals

All animal experiments were conducted in accordance with international guidelines on the ethical use of animals, including the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines, and all efforts were made to minimise pain or discomfort experienced by the animals. Animal housing and surgical procedures were approved by the Institutional Animal Care and Use Committee of Niigata University Graduate School of Medical and Dental Science (Niigata City, Japan; Approval No. H25-342-4). We used adult male Wistar rats (6–8 weeks old, 200–280 g) housed under a 12-h light/dark cycle with ad libitum access to food and water. We used only male rats to exclude the possibility of sexual cycle-modified behaviour because DHEAS is a precursor of sex steroids.

Behavioural testing

A catheter was placed in the spinal cord subarachnoid space for drug administration as described.³⁰ Briefly, 0.5% lidocaine was infiltrated at the surgical site after animals were anaesthetised with 2 vol% isoflurane. Next, the lumbar vertebrae were exposed, the spinal process of the fifth lumbar vertebrae was removed to access the dura, and ∼2 cm of the polyethylene catheter (PE-10; BD Biosciences, Franklin Lakes, NJ, USA) was inserted from the gap between the fourth and fifth lumbar vertebrae so that its tip was close to the lumbar enlargement of the spinal cord; the other end of the tube was closed by thermal melting. After 3 days, DHEAS (3, 10, 30 μg) or 20% dimethyl sulphoxide in saline (vehicle) was administered to the spinal cord subarachnoid space. The pain reaction threshold against mechanical stimuli was measured using the von Frey test (1, 1.4, 2, 4, 6, 8, 10, 15, 26, 60 g) before drug administration, and at 5, 15, 30, 60, and 120 min after administration. Eight stimuli were applied per filament on the left hindlimb, and the test was deemed positive if a reaction was observed more than twice.

Preparation of spinal cord slices

Cross-sectional slices of the spinal cord were prepared as previously described.³¹ Briefly, the lumbar enlargement of the spinal cord was extracted under urethane anaesthesia (1–1.5 g kg⁻¹ i.p.) and immersed in a cooled Krebs solution saturated with 95% O₂ and 5% CO₂. Except for the L4 dorsal root on the left side, all nerve roots were removed. Cross-sectional spinal cord slices, approximately 650 μm thick, were prepared with the left L4 dorsal root using a microslicer (Dosaka EM, Kyoto, Japan). Slices were placed on a recording chamber and perfused (15 ml min⁻¹) with a Krebs solution heated to 34–36°C. The composition of the Krebs solution (mM) was as follows: NaCl, 117; KCl, 3.6; CaCl₂, 2.5; MgCl₂, 1.2; NaH₂PO₄, 1.2; NaHCO₃, 25, and d-glucose, 11.5, pH=7.35. for Ca²⁺ measurements; and NaCl, 113.1; KCl, 3.6; MgCl₂, 5; NaH₂PO₄, 1.2; NaHCO₃, 25; and D-glucose, 11, pH=7.35. for Ca²⁺-free measurements.

Patch-clamp recording from dorsal horn neurones

Blind whole-cell patch-clamp recording was conducted under the stereomicroscope using glass microelectrodes with a tip electrode resistance of 5–10 MΩ from the substantia gelatinosa (SG) neurones. The composition of the glass microelectrode solution (mM) was as follows: K-gluconate, 135; KCl, 5; CaCl₂, 0.5; MgCl₂, 2; ethylene glycol-bis(β-aminoethyl ether)-N,N,N′,N′-tetraacetic acid (EGTA), 5; 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 5; tetraethylammonium (TEA), 5; and ATP-Mg, 5, pH=XX for the current clamp method; and Cs-sulphate, 110; CaCl₂, 0.5; MgCl₂, 2; EGTA, 5; HEPES, 5; TEA, 5; and ATP-Mg, 5, pH=XX for the voltage clamp method.

Drug application and experimental protocol

The volume of cerebrospinal fluid was assumed to be 300 μl; therefore, the predicted concentration was assumed to be 100 μM when 10 μg of DHEAS (Sigma-Aldrich, St. Louis, MO, USA) was administered to the spinal cord subarachnoid space. DHEAS was dissolved in Krebs solution before administration. A 60 pA current was applied for 1000 ms to measure the action potential using the current clamp method. The holding potential during measurements was –70 mV when observing EPSCs and 0 mV when observing inhibitory postsynaptic currents (IPSCs). Miniature EPSCs (mEPSCs), and miniature IPSCs (mIPSCs) were measured in the presence of 1 μM tetrodotoxin (TTX; Wako Pure Chemical Industries, Osaka, Japan). Evoked EPSCs, which were obtained by stimulating the dorsal root by a suction electrode, were measured after stimulation of C nerve fibres at 1 mA for 500 μs and stimulation of Aδ nerve fibres at 100 μA for 50 μs. The resulting current was amplified by Axopatch 200B (Molecular Devices, Sunnyvale, CA, USA), A/D converted using Digidata 1440A (Molecular Devices), recorded on a computer, and analysed with pCLAMP 10.4 (Molecular Devices) and Minianalysis 6.0 (Synaptosoft, Leonia, NJ, USA). All other drugs were form Sigma-Aldrich.

Statistical analysis

Data are expressed as mean (standard error of mean) for behavioural experiments and median (range) for electrophysiological experiments. For behavioural experiments, two-way analysis of variance (anova) followed by Bonferroni corrections were used, and the corresponding data obtained in electrophysiological experiments were tested using the Wilcoxon signed-rank test. Statistical significance was defined as P<0.05. GraphPad Prism (version 7, GraphPad Software, San Diego, CA, USA) was used for statistical analysis.

Results

Intrathecally administered DHEAS evoked dose-dependent mechanical hypersensitivity

Intrathecally administered DHEAS caused dose-dependent mechanical hyperalgesia 15 min after administration as compared with saline administration (saline: 51.0 [20.1] g; 3 μg DHEAS: 14.0 [7.8] g, P<0.01; 10 μg DHEAS: 6.9 [5.2] g, P<0.00; Fig. 1). The mechanical threshold returned to the pre-administration value 120 min after the drug was administered (Fig. 1). Intrathecally administered DHEAS (30 μg) elicited spontaneous pain behaviour (flinching, licking, and biting the hindlimb and tail); therefore, the mechanical threshold could not be measured in this group.

Fig. 1.

Paw withdrawal threshold in response to von Frey stimulation is reduced in a concentration-dependent manner by intrathecal injection of DHEAS (vehicle, DHEAS 3 μg, DHEAS 10 μg). The data are presented as mean (sem). n=5 in each group; *P<0.05, **P<0.01, ***P<0.001 by two-way anova. anova, analysis of variance; DHEAS, dehydroepiandrosterone sulphate; sem, standard error of the mean.

DHEAS potentiated excitability in spinal dorsal horn neurones

We investigated the mechanism of DHEAS action on SG neurones using the whole-cell patch-clamp technique. In the current clamp mode, several series of action potentials were observed with 60 pA of current injection for 1000 ms. The number of action potentials significantly increased after a 5-min superfusion of 100 μM DHEAS (control, 2.0 [1–8]; DHEAS, 6.5 [1–17], P=0.013; Fig. 2).

Fig. 2.

DHEAS potentiates neuronal excitability in the spinal dorsal horn. (a) Current injection (60 pA, 1000 ms) elicited action potential(s) in spinal dorsal horn neurones under current clamp mode. (b) DHEAS (100 μM, 5 min) increased the number of action potentials induced by current injection. (c) Summary of the data. DHEAS superfusion to spinal cord slices significantly increased the number of action potentials (n=9). Boxes show inter-quartile range (IQR), middle lines show medians, and whiskers show the range of values. DHEAS, dehydroepiandrosterone sulphate.

DHEAS increased both amplitude and frequency of miniature EPSCs

DHEAS (100 μM) increased cell excitability in SG neurones. To determine the effects of DHEAS on excitatory synapses in the spinal dorsal horn, we investigated the effects of DHEAS on mEPSCs. Superfusion of 100 μM DHEAS for 5 min increased mEPSC frequency and amplitude without the baseline inward current (frequency: control, 16.5 [1.6–36.8] Hz; DHEAS, 36.3 [13.8–48.6] Hz, P=0.008; amplitude: control, 9.3 [4.6–13.6] pA; DHEAS, 12.76 [5.0–20.5] pA, P=0.008; Fig. 3a–c).

Fig. 3.

DHEAS facilitates excitatory synaptic transmission mainly via the presynaptic mechanism in a sigma-1 receptor-, NMDA receptor-, and extracellular Ca²⁺-dependent manner. (a) Representative traces of mEPSCs in the spinal dorsal horn. Neurones in the spinal dorsal horn were voltage clamped at −70 mV. Downward arrows indicate outtakes of the top trace shown on an expanded timescale. Heavy horizontal bars show periods of drug application. (b, c) Bath-applied DHEAS (100 μM, 5 min) increased the frequency and amplitude of mEPSCs. Changes in frequency (b) and amplitude (c) of mEPSCs before and after DHEAS administration (n=9). (d) In the presence of AP5 (50 μM), an NMDA receptor antagonist, DHEAS increased mEPSC frequency. (e) DHEAS had no effect on mEPSC amplitude (n=9). (f) In the presence of BD1047 (100 μM), a sigma-1 receptor antagonist, DHEAS increased mEPSC frequency. (g) DHEAS had no effect on mEPSC amplitude (n=14). (h) Under a Ca²⁺-free Krebs solution environment, DHEAS had no effect on mEPSC frequency and amplitude (n=7). Boxes show inter-quartile range (IQR), middle lines show medians, and whiskers show the range of values. DHEAS, dehydroepiandrosterone sulphate; mEPSCs, miniature excitatory postsynaptic currents; NMDA, N-methyl-d-aspartate.

Excitatory synaptic currents were elicited via two glutamate receptors, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and NMDA receptors. Then, we investigated changes in amplitude and frequency of mEPSCs with an NMDA receptor inhibitor to determine if the NMDA receptor is influenced by DHEAS. In the presence of 50 μM AP5, an NMDA receptor antagonist, 100 μM DHEAS significantly increased the frequency, but not the amplitude, of mEPSCs (frequency: control, 11.3 [1.3–32.2] Hz; DHEAS, 18.1 [2.3–45.9] Hz, P=0.02; amplitude: control, 7.4 [4.8–12.7] pA; DHEAS, 7.3 [5.4–13.5] pA, P=0.13, Fig. 3d and e).

DHEAS is a putative ligand of the sigma-1 receptor.³² To determine the involvement of this receptor in the effects of DHEAS on mEPSCs, we investigated the effects of BD1047, a sigma-1 receptor antagonist, on mEPSCs. In the presence of 100 μM BD1047, 100 μM DHEAS increased the frequency, but not the amplitude, of mEPSCs (frequency: control, 10.3 [1.2–44.5]; DHEAS, 15.5 [2.4–82.7] Hz, P=0.03; amplitude: control, 8.0 [6.4–23.6]; DHEAS, 8.4 [5.0–24.0] pA, P=0.54; Fig. 3f and g).

The NMDA receptor is a Ca²⁺-permeable ion channel,³³ and Ca²⁺ is important for cell excitability and synaptic plasticity. To confirm the effect of DHEAS on excitatory synaptic transmission in the spinal dorsal horn, we investigated the effects of DHEAS on mEPSCs in a Ca²⁺-free environment. We found that 100 μM DHEAS did not elicit frequency or amplitude potentiation of mEPSCs (frequency: control, 5.1 [0.9–7.9); DHEAS, 4.2 (1.0–9.7) Hz, P=0.87; amplitude: control, 6.7 [3.9–7.3]; DHEAS, 6.1 [4.0–7.2] pA, P=0.11; Fig. 3h and i).

Evoked EPSCs were not influenced by DHEAS

In slice preparations, dorsal root stimulation elicited evoked EPSCs. Evoked EPSCs are thought to be induced by postsynaptic Ca²⁺ influx by voltage-dependent Ca²⁺ channels, and in this environment, the effects of presynaptic ligand-gated ion channels are negligible. DHEAS did not influence the evoked EPSCs by electrical stimulation of the dorsal root using a suction electrode in Aδ or C nerve fibres (Aδ fibres: control, 95.8 [51.7–120.3]; DHEAS, 93.8 [39.6–153.9] pA, P=0.63; C fibres: control, 90.3 (38.0–264.5); DHEAS, 81.8 [34.9–259.3] pA, P=0.13; Fig. 4).

Fig. 4.

DHEAS did not affect evoked excitatory synaptic currents in the spinal dorsal horn. (a) Representative traces of Aδ fibre-evoked monosynaptic EPSCs recorded during baseline and 5 min after DHEAS (100 μM) superfusion to spinal cord slices. (b) DHEAS had no effect on the amplitude of Aδ fibre-evoked monosynaptic EPSCs (n=5). (c) Representative traces of C fibre-evoked monosynaptic EPSCs recorded during baseline and 5 min after DHEAS superfusion. (e) DHEAS had no effect on the amplitude of C fibre-evoked monosynaptic EPSCs (n=5). All recordings were performed at a holding potential of –70 mV. Boxes show inter-quartile range (IQR), middle lines show medians, and whiskers show the range of values. DHEAS, dehydroepiandrosterone sulphate; EPSCs, excitatory postsynaptic currents.

DHEAS suppressed the frequency of mIPSCs

A previous study showed that DHEAS also inhibits GABA_A receptors and suppressed spontaneous inhibitory postsynaptic currents (sIPSCs) in cultured midbrain neurones of rats.³⁴ We investigated the effects of DHEAS on mIPSCs in the spinal dorsal horn. With superfusion of 1 μM TTX, 100 μM DHEAS decreased the frequency, but not the amplitude, of mIPSCs (frequency: control, 4.1 [1.4–10.3]; DHEAS, 2.8 [1.4–5.6] Hz, P=0.018; amplitude: control, 9.4 [7.1–12.7]; DHEAS, 8.7 [6.5–13.1] pA, P=0.16; Fig. 5a–c).

Fig. 5.

DHEAS suppressed inhibitory synaptic transmission in a GABA_A and sigma-1 receptor-dependent manner. (a) Representative traces of mIPSCs in spinal dorsal horn. Neurones in the spinal dorsal horn were voltage clamped at 0 mV. Downward arrows indicate outtakes of the top trace shown on an expanded timescale. Heavy horizontal bars show periods of drug application. (b, c) Bath-applied DHEAS (100 μM, 5 min) decreased mIPSC frequency. However, DHEAS had no effect on amplitude. Change in the frequency (b) and amplitude (c) of mEPSCs before and after DHEAS administration (n=7). (d, e) In the presence of bicuculline (10 μM), a GABA_A receptor antagonist, DHEAS had no effect on glycinergic mIPSC frequency and amplitude (n=14). (f, g) In the presence of strychnine (2 μM), a glycine receptor antagonist, DHEAS decreased GABAergic mIPSC frequency and amplitude (n=11). (h) In the presence of BD1047 (100 μM), a sigma-1 receptor antagonist, DHEAS had no effect on mIPSC frequency (n=7). (i) BD1047 suppressed the amplitude of mIPSCs. Boxes show inter-quartile range (IQR), middle lines show medians, and whiskers show the range of values. DHEAS, dehydroepiandrosterone sulphate; GABA_A, gamma-aminobutyric acid A; mEPSCs, miniature excitatory postsynaptic currents.

As both GABA_A and glycine receptors are involved in IPSCs, we attempted to clarify which of these receptors is involved in DHEAS inhibition of IPSCs. In the presence of 10 μM bicuculline, a GABA_A receptor antagonist, DHEAS did not show any influence on mIPSC frequency and amplitude (frequency: control, 1.5 [0.8–6.8]; DHEAS, 1.3 [0.4–6.4] Hz, P=0.20; amplitude: control, 6.6 [4.3–23.6]; DHEAS, 8.0 [3.8–19.2] pA, P=0.9; Fig. 5d and e). In contrast, DHEAS reduced both frequency and amplitude of mIPSCs in the presence of 2 μM strychnine, a glycine receptor antagonist (frequency: control, 2.6 [0.9–9.0]; DHEAS, 1.2 [0.2–3.4] Hz, P=0.003, amplitude: control, 8.2 [3.9–17.0]; DHEAS, 5.3 [2.6–12.8] pA, P=0.002; Fig. 5f and g). Taken together, bicuculine, but not strychnine, occluded the inhibitory effect on mIPSCs by DHEAS.

DHEAS is a sigma-1 receptor agonist, and activation of the sigma-1 receptor elicited neuronal excitability via activation of the NMDA receptor.²⁶ However, it is unclear if activation of the sigma-1 receptor influences inhibitory synaptic transmission. In the presence of 100 μM BD1047, DHEAS did not influence the frequency of mIPSCs. However, it suppressed the amplitude of mIPSCs (frequency: control, 5.2 [1.5–13.6]; DHEAS, 5.4 [1.4–8.5] Hz, P=0.67, amplitude: control, 10.5 [9.3–23.7]; DHEAS, 9.7 [6.9–15.8] pA, P=0.016; Fig. 5h and i).

Discussion

Intrathecally administered DHEAS induced dose-dependent mechanical hyperalgesia and spontaneous pain behaviours in a non-neuropathic pain model in rats. Results from in vitro patch-clamp recording experiments in rat spinal dorsal horn showed that DHEAS increased the number of action potentials with depolarisation in the current clamp mode, which indicates that DHEAS increases excitability in dorsal horn neurones. Under a voltage-clamp mode with tetrodotoxin to suppress spontaneous action potentials in recorded neurones, DHEAS increased mEPSC frequency and amplitude. This facilitation of mEPSC frequency by DHEAS may be mediated by NMDA- and sigma-1 receptor-independent presynaptic mechanisms because the frequency of mEPSCs induced by DHEAS superfusion was not suppressed by co-application of the NMDA receptor and sigma-1 receptor antagonists. Moreover, DHEAS did not affect the amplitude of mEPSCs in a Ca²⁺-free environment, and mEPSC frequency was strongly suppressed and summation of mEPSCs did not occur. NMDA receptors are known to allow permeation of Ca²⁺, and this Ca²⁺ is thought to be crucial to synaptic plasticity by both pre- and postsynaptic mechanisms. Hence, facilitation of mEPSC frequency by DHEAS is thought to be Ca²⁺-independent because of NMDA receptor activation.

DHEAS is known to be hyperalgesic when intrathecally22, 24, 26 and intraplantarly³⁵ administered. A previous study showed that DHEAS, but not its non-sulphated form DHEA, produced mechanical allodynia in a sigma-1- and GABA_A-receptor-dependent manner. The hyperalgesic effect of DHEAS administered intrathecally is caused by postsynaptic NMDA receptor activation via sigma-1 receptor-mediated phosphorylation of the GluN122, 26 subunit and direct inhibition of GABA_A receptors.24, 34, 36 Similar to our results, another study showed that DHEAS potentiated both synaptic potentials and cell membrane excitability.³⁷ These studies show that direct inhibition of GABA_A receptors by DHEAS should elicit postsynaptic neurone excitability. Our results are also compatible with evidence showing that muscimol, a GABA_A receptor agonist, attenuates the hyperalgesic effects of DHEAS.²⁴

The nociceptive properties of DHEAS in the spinal dorsal horn have been explained by alterations of NMDA receptor function via the sigma-1 receptor. Activation of the sigma-1 receptor is known to initiate activation of protein kinase C (PKC) that phosphorylates the GluN1 subunit and enhances cell surface expression of NMDA receptors,³⁸ and is involved in the pathophysiology of hypersensitivity.³⁹ Our findings are partially compatible with those of previous studies. The amplitude of mEPSCs was inhibited by co-application of a sigma-1 receptor and NMDA receptor antagonist. However, the frequency of mEPSCs was not affected by either NMDA or sigma-1 receptor antagonists. Our results suggest that DHEAS predominantly enhances nociceptive synaptic transmission at presynaptic sites. Previous studies have shown that DHEAS increased glutamate release from presynaptic terminals12, 40 and potentiated excitatory synaptic transmission¹⁷ via sigma-1 receptor activation. Moreover, presynaptic NMDA receptor activation is thought to be associated with the pathophysiology of neuropathic pain.⁴¹ However, we also found that extracellular Ca²⁺, but not NMDA receptor or sigma-1 receptor activation, is crucial for presynaptic facilitation of glutamate release by DHEAS. Although our findings were not compatible with those of previous studies,12, 17, 40 DHEAS is also known to activate the Ca²⁺-activated K channel (KCa)⁴² and Ca²⁺-permeable ion channels, such as P2X purinergic receptors,5, 43 which may account for these discrepancies.

Dorsal root-evoked EPSCs in the spinal dorsal horn were not influenced by DHEAS. During evoked EPSCs in the dorsal horn, glutamate release is thought to be elicited by Ca²⁺ via voltage-gated Ca²⁺ channels (VGCCs), and its amplitude should be mainly determined by presynaptic Ca²⁺ influx (which is equal to VGCC activity) or the conductance and amount of surface expression of postsynaptic AMPA receptors. Our results indicated that both presynaptic VGCCs and postsynaptic AMPA receptors were not modulated by DHEAS via sigma-1 receptor activation. Although sigma-1 receptor activation decreases VGCC current⁴⁴ and expression of AMPA receptors,⁴⁵ these previous studies were carried out using isolated cardiac or sympathetic neurones and cultured cortical neurones, and the properties of the expressed receptors and intracellular signal transduction mechanisms may differ from the sensory nervous system. This may explain why the results of this study and previous studies are inconsistent.

During inhibitory synaptic transmission, DHEAS can inhibit both GABA_A receptors34, 36 and glycine receptors46, 47; however, the effects of DHEAS on inhibitory neurotransmitter release are still unclear. This study showed that DHEAS suppressed GABA_A-receptor-mediated mIPSCs via sigma-1 activation. Previous studies showed that 10 μM of DHEAS suppressed spontaneous IPSCs in midbrain neurones, and sigma-1 receptor activation suppressed the frequency of mIPSCs in medullary dorsal horn neurones⁴⁸; our results were compatible with these findings. However, our results also indicate that DHEAS suppressed GABAergic mIPSCs, not glycinergic mIPSCs. The frequency of glycinergic mIPSCs increased after administration of pregnenolone sulphate, another neuroactive steroid that activates the sigma-1 receptor.⁶ In that study, the frequency of glycinergic mIPSCs was not sigma-1 receptor-dependent, and the authors suggested that transient receptor potential (TRP) channel-mediated intracellular Ca²⁺ may be involved. Involvement of a TRP channel for neurotransmitter release by DHEAS is a likely mechanism because intracellular Ca²⁺ elevation is crucial for enhancing both excitatory and inhibitory neurotransmitter release from presynaptic terminals. Our results also indicate that a Ca²⁺-free environment strongly suppressed the frequency of mEPSCs induced by administration of DHEAS.

These studies have several limitations. First, experiments were performed only in male rats. However, DHEAS is a type of androgen and it may affect sexual cycle in female rats. we want to exclude the possibility of sexual cycle-modified behavior in this study. Second, Ca²⁺ influx and neurotransmitter release were not measured directly. However, many studies have performed previously clearly showed that electrophysiological assessment was used for evaluating neurotransmitter release and influence of extracellular Ca²⁺ concentration for neurotransmitter release, and our data were also showed that DHEAS increases release of excitatory neurotransmitter (glutamate) in extracellular Ca²⁺-dependent manner. Third, pharmacological concentration of DHEAS were used, which may not translate to physiological conditions. Further study should be needed to clarify the physiological role of DHEAS in neurotransmission of pain.

Considering these findings, we propose the following model circuit for the underlying mechanism of DHEAS in the spinal dorsal horn (Fig. 6): 1) DHEAS inhibits GABA_A receptors located on postsynaptic sites of the recorded SG neurones and inhibits presynaptic GABA release on GABAergic neurones via sigma-1 receptors, and 2) DHEAS leads to increased glutamate release from presynaptic excitatory interneurones to SG neurones via indirect activation of neither NMDA nor sigma-1 receptors. We propose that these mechanisms could produce hyperalgesia.

Fig. 6.

Model of the spinal dorsal horn circuit underlying the mechanism of DHEAS-produced hyperalgesia. DHEAS potentiates glutamate release from the presynaptic site to the recorded substantia gelatinosa (SG) neurone via neither sigma-1 receptor- nor NMDA-receptor-dependent manners and also inhibits postsynaptic GABA_A receptors. This increases neuronal excitability of SG neurones to elicit mechanical hyperalgesia and spontaneous pain behaviours. SG, substantia gelatinosa; DHEAS, dehydroepiandrosterone sulphate; GABA_A, gamma-aminobutyric acid A; NMDA, N-methyl-d-aspartate.

Conclusions

We showed that intrathecally administered DHEAS induced mechanical hyperalgesia and spontaneous pain behaviours in rats and facilitated excitatory synaptic transmission and inhibition of inhibitory synaptic transmission in the spinal dorsal horn. Synthetic enzymes of neurosteroids are abundant in the nervous system, and some of them, including DHEAS, facilitate synaptic transmission. Our results suggest that DHEAS may be a potential therapeutic target for prevention of plastic changes of nociceptive transmission in the spinal cord.

Authors' contributions

Conception and design of study: GY, MI, YK.

Conduct of experiments: GY, MS.

Analysis of data: GY.

Drafting of the manuscript: GY, YK, TK, HB.

Editing and revision of the manuscript: YK.

All authors read and approved the final version of the manuscript.

Declaration of interest

The authors declare that they have no conflicts of interest.

Funding

Japan Society for the Promotion of Science, Tokyo, Japan (26861224 and 16K20082 to GY; 24592330 to MI; 18H02897 to YK).

Handling editor: H.C. Hemmings Jr

Editorial decision : 21 March 2019