Male Sprague-Dawley rats (Orient Bio, Korea), aging 6-8 weeks, were used in this study. All the experiments were approved by Institutional Animal Care and Use Committee at Seoul National University (SNU-160216-4) and Institutional Animal Care and Use Committee at Korea University College of Medicine (KOREA-2016-0022). Animals were housed under a 12-h light/dark cycle. Food and water were available ad libitum.

Oxytocin, atosiban, and capsaicin were purchased from Sigma-Aldrich (USA). Selective oxytocin receptor antagonist (desGly-NH₂-d(CH₂)₅[D-Tyr²,Thr⁴]OVT, OTRA) and V1a receptor antagonist (d(CH₂)₅[Tyr(Me)²,Dab⁵]AVP, V1aRA) were generously donated from Dr. M. Manning at the University of Toledo College of Medicine (Toledo, OH, USA). All drugs were diluted to the final concentrations from stock solutions on the day of the experiment. The solutions used in behavioral test were dissolved in normal saline. The solutions used in electrophysiological recordings and calcium imaging were dissolved in external solutions of each experiment except capsaicin which were dissolved in ethanol as a stock solution and further diluted in the extracellular solution when applied.

We adopted Hargreaves' test for thermal nociception assay [14]. As previously described [15], thermal pain threshold was measured using radiant heat test. Rats were placed on a glass surface at 23-25℃. A heat source was focused on the mid-third hindpaw of a rat standing on the glass plate with all four feet aground, and paw withdrawal latency was assessed using Series 8, Model 390 (IITC Life Science, US).

Modified from the previous study [16], rats were anesthetized with isoflurane and then decapitated. The DRGs were taken out from L1-L6 and placed immediately in 4℃ HEPES buffered HBSS at pH 7.4. After removing the surrounding connective tissues, the DRGs were minced with fine spring scissors under a microscope and the ganglions were transferred to a 15 ml tube containing 2 ml HEPES buffered HBSS in which 100 mg/ml collagenase and 2.4 U/ml dispase had been dissolved, and incubated at 37℃ in a water bath for 40 min. The cells were washed in DMEM containing 10% FBS and 0.5% Penicillin and Streptomycin and triturated with a flame-polished Pasteur pipette to separate cells and remove processes. The dissociated cells were placed on 1 mg/ml poly-Dlysine (Sigma-Aldrich, US)-coated glass cover slips (25 mm in diameter), then maintained at 37℃ in an incubator. The dissociated neurons were kept for at least another 4 h before electrophysiological recordings or calcium imaging, and used for the experiments within 36 h. The neurons selected for electrophysiological experiments and single cell RT-PCR were 15-35 µm in diameter. More than 3 rats were sacrificed for the collection of dissociated DRG neurons in each experiment.

The cells were harvested into a micropipette with a tip diameter of about 20 µm, put into a tube containing reagents solution, and reverse-transcription was performed using EasyScript Two-Step RT-PCR Kit (ABM Inc, Canada) according to the manufacturer's instruction. Subsequent cDNA product was divided into aliquots of 2 µl and the each was used in separate reactions. All PCR amplifications were performed with nested primers (Table 1) using 2X PCR Taq Plus MasterMix (ABM Inc, Canada) according to manufacturer's protocol. Briefly, the first round of PCR was performed in 50 µl of reaction buffer containing 25 µl MasterMix, 2 µl cDNA, 2 µl “outer” primers. The second round reaction buffer (25 µl) contained 12.5 µl MasterMix, 2 µl product from the first round, and 1 µl “inner” primers. The protocol included a 5 min initial denaturation step at 94℃ followed by 40 cycles for the first round or 30 cycles for the second round of 30 s denaturation at 94℃, 30 s annealing at 55-60℃ depending on primers, 60 s elongation at 72℃. The reaction was completed with 5 min of final elongation. A negative control was acquired from the pipette that did not collect any cell contents, but were submerged in the extracellular solution. The PCR products were displayed on Loading-STAR (DYNEBIO, Korea) stained, 2% agarose gels. Primers were shown in Table 1.

Coverslips cultured 24 h with rat DRG neurons were fixed in 4% formaldehyde solution for 20 min. The DRG neurons were washed with 0.01 M phosphate-buffered saline (PBS) and immunostained with following antibodies. The primary antibodies were rabbit anti-V1a receptor polyclonal antibody (1:400; Alomone Labs, UK) and guinea pig anti-TRPV1 polyclonal antibody (1:500; Novus biologicals, USA), diluted in PBS containing 0.3% Triton X-100, 1% normal donkey serum (Jackson ImmunoResearch Co, USA), incubated overnight at 4℃ then incubated for 1 h with donkey anti-guinea pig IgG conjugated with cyanine dye 3 (Cy3, 1:200; Jackson ImmunoResearch Co, USA) and donkey anti-rabbit IgG conjugated with fluorescein-isothiocyanate (FITC, 1:200; Jackson ImmunoResearch Co, USA) diluted in the same buffer as the primary antibody. All rinses before and after the incubations were performed with PBS. Coverslips were photographed using confocal microscope (LSM-700; Zeiss, Germany).

Modified from the previous study [17], whole-cell currentand voltage-clamp recordings were performed to measure action potentials and voltage-gated potassium current respectively. The signals from neurons amplified by Axoclamp-700A amplifier (bandwidth filter set at 10 kHz for current-clamp and 1 kHz for voltage-clamp recordings) were digitized and sampled at 50 µs intervals (Digidata 1440, pClamp 10.3; Molecular Devices). Patch pipettes were pulled from borosilicate capillaries (P-97; Sutter Instruments, USA). The resistance of the pipettes was 3-5 MΩ when filled with the intracellular solution composed of (mM): 140 K-gluconate, 10 HEPES, 2 MgCl₂, 2 CaCl₂, 11 EGTA, 4 K₂-ATP and 0.3 Na₂-GTP adjusted to pH 7.2-7.3 with KOH, osmolarity 280-285 mOsm. Recording chamber was continuously perfused (1-2 ml/min) with the warm (32-35℃) extracellular solution composed of (mM): 150 NaCl, 4 KCl, 2.5 CaCl₂, 10 glucose, 2 MgCl₂, 10 HEPES, and 0.5 tetrodotoxin (TTX, applied for voltage-clamp experiments), adjusted to pH 7.4 with NaOH, osmolarity 290-300 mOsm. Na⁺-free extracellular solution contained 150 mM Tris instead of Na⁺ and was adjusted to pH 7.4 with HCl [18]. Cadmium chloride (300 µM) was added to the Na⁺-free extracellular solution for blockade of calcium channels [19]. A small patch of cell membrane underneath the tip of the pipette was aspirated to form a gigaseal. After gigasealing, negative pressure was applied to rupture the membrane for a whole-cell configuration. In voltage-clamp experiments, the membrane voltage was maintained at –60 mV. At the holding potential, the membrane potential was stepped to –110 mV for 200 ms and then a depolarizing ramp pulse to 0 mV at a rate of 110 mV/s (step-and-ramp pulse) was applied. In current-clamp experiments, action potentials were generated by 1000 ms depolarizing current with 100-500 pA. Only cells with a stable resting potential (more negative than –40 mV) were used in this study.

As described in the previous study [16], DRG neurons were loaded with 3 µM of the fluorescent calcium indicator dye Fura-2-AM (dissolved in DMSO, F1201, Thermo Fisher Scientific) for 40 min in an incubator at 37℃ and then washed. After washing, the cells on a coverslip were placed on a chamber with a perfusion system (4 channel gravity ALA-VM4, Charles Austen Pump, Dynamax 30) in dark surroundings at room temperature. Regions of interest were defined on a computer connected to a CCD camera and the recording area of which were viewed through an inverted microscope. Fura-2 fluorescence of the cells was recorded at 510 nm during shifting excitation at 340 and 380 nm at 1Hz using Lamda DG4. The ratio of emission at 510 nm from excitation at 340 nm to at 380 nm was analyzed. Only cells with a 340/380 fluorescence ratio difference higher than 0.3 were analyzed. Since our experimental protocol of repeated treatment of capsaicin produced TPRV1 desensitization, Phorbol 12-myristate 13-acetate (PMA) was mixed to the 200 nM capsaicin to the final concentration of 10 µM before the experiment to reduce capsaicin-induced calcium-dependent desensitization of TPRV1 [20]. The cells whose 340/380 ratio increased more than 0.3 in response to capsaicin were considered as capsaicin responsive cells. The cells showing 20% decrease in the second capsaicin response in comparison to the first capsaicin response were considered as oxytocin responsive cells.

Data were expressed as mean±SEM. Behavioral and electrophysiological data were analyzed using ANOVA followed by posthoc Bonferroni's test. Differences with a p<0.05 were considered significant.

We aimed to determine the specific receptor that mediates the oxytocin-induced thermal analgesia in vivo by using Hargreaves' test. Thus, we examined whether the analgesic effect produced by IP injection of oxytocin (1 mg/kg), would be affected by coadministration with OTRA (1 mg/kg), atosiban (1 mg/kg) and V1aRA (1 mg/kg). Paw withdrawal latency after thermal stimuli was significantly increased at 30 and 60 min after the injection of oxytocin (Fig. 1A). This thermal analgesia induced by oxytocin was significantly reversed by co-injection of 1 mg/kg of V1aRA, vasopressin-1a receptor antagonist, but not with 1 mg/kg of OTRA, oxytocin receptor antagonist (Figs. 1A and B). Co-application of atosiban, oxytocin and vasopressin receptor antagonist, also showed trends for the reversing effect although not statistically significant (Fig. 1B; p>0.01, posthoc Bonferroni's test). These results suggest that oxytocin induces thermal analgesia via V1a receptor.

Given that TRPV1 is the primary transducer of thermal nociception in DRG neurons [21], we then examined mRNA expression pattern for oxytocin binding receptors in relation with TRPV1 by using scRT-PCR. We only collected small to medium-sized DRG neurons for scRT-PCR (Fig. 2A, Supplementary Fig. 1A). Amongst 54 GAPDH expressing neurons, mRNA transcripts of V1a receptor (n=35/54, 65%), vasopressin-1b receptor (V1b, n=3/54, 6%) and oxytocin receptor (n=1/54, 2%) were expressed in DRG neurons. Vasopressin-2 (V2) receptor mRNA was not detected (n=0/54, 0%) (Fig. 2B). To examine whether the samples were contaminated by glial cells, we investigated the expression of glial fibrillary acidic protein (GFAP) mRNA and GFAP mRNA was not detected in the V1a receptor expressing neurons (n=0/14, 0%) (Supplementary Fig. 1B). We also found that 67% (n=23/34) of TRPV1-expressing DRG neurons were also co-expressed with V1a receptor. (Fig. 2C). Our result also demonstrated that the co-expression of TRPV1 and V1a receptor proteins in the DRG neurons (Supplementary Fig. 2). These findings suggest that V1a receptor activated by oxytocin might transact with TRPV1 to induce thermal analgesia.

To verify the functional interaction between oxytocin and TRPV1, we examined effects of oxytocin on capsaicin-induced calcium transients in DRG neurons. A series of repetitive capsaicin application elicited reproducible calcium transients in DRG neurons with PMA although we observed small desensitization (Fig. 3A). Our results demonstrated that bath application of 5 µM oxytocin significantly decreased calcium transients induced by capsaicin (Figs. 3B and F). The co-application of 5 µM oxytocin with either 20 µM atosiban (Figs. 3C and F) or 5 µM V1aRA (Figs. 3D and F) reversed the reducing effect of oxytocin on capsaicin-induced calcium transients. We found that 78% (55/70) of capsaicin responsive DRG neurons also responded to oxytocin application (Fig. 3E). These results suggested that oxytocin-induced thermal analgesia might be attributed to the reduction of TRPV1 activity by oxytocin via V1a receptor.

We also investigated effects of oxytocin on the membrane excitability in the small to medium-sized DRG neurons. While oxytocin (5 µM) for 60 s lowered the number of action potential in response to a 300 pA and 1000 ms current injection, the number of action potentials evoked by the current injection returned almost to the control value after 5 min-washout (Fig. 4A). It is noteworthy that action potentials were observed only at the onset and before the offset of the current pulse, suggesting activation of voltage-gated potassium channels rather than other voltage-gated ion channels. Consistent with our results above, inhibitory effects of oxytocin on the action potential firings were also reversed by atosiban (20 µM, Fig. 4B) and V1aRA (5 µM, Fig. 4C). These results again suggest that oxytocin reduce membrane excitability of DRG neurons via V1a receptor.

As reduction of the number of action potentials indicated regulation of voltage-gated potassium channels, we performed voltage clamp recordings with the small to medium-sized DRG neurons in the presence of 0.5 µM TTX. At a holding potential of –60 mV, the membrane potential was stepped to –110 mV for 200 ms and then a depolarizing ramp pulse to 0 mV at a rate of 110 mV/s (step-and-ramp pulse) was applied (Fig. 5A). Effects of oxytocin on the membrane current response to the ramp pulse were evaluated as a change in the I-V relationship that was obtained by subtraction of the control response from the response obtained after oxytocin application. Application of oxytocin (5 µM) for 60 s markedly enhanced the outward current activated at membrane potentials more positive than –40 mV (Fig. 5B, left and middle) while it caused a slight increase in hyperpolarization-activated inward current at –110 mV (Fig. 5B, left and middle). The outward current was recorded in Cd²⁺ containing Na⁺-free external solution (Supplementary Fig. 3B). The mean peak current density of the DRG neurons at 0 mV also significantly increased by oxytocin (Fig. 5B, right). In agreement with our previous results, not only atosiban (20 µM, Fig. 5C) but also V1aRA (5 µM, Fig. 5D) inhibited the currents elicited by oxytocin, which would indicate that oxytocin induce the prominent outward potassium current and modest hyperpolarization-induced inward current via V1a receptor.