Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 May 21.
Published in final edited form as: Neuroscience. 2021 Apr 4;463:159–173. doi: 10.1016/j.neuroscience.2021.03.035

D2 Receptors in the Periaqueductal Gray/Dorsal Raphe Modulate Peripheral Inflammatory Hyperalgesia Via the Rostral Ventral Medulla

Luiz F Ferrari 1, JunZhu Pei 2, Michael Zickella 1, Charles Rey 1, Jacqueline Zickella 1, Anna Ramirez 1, Norman E Taylor 1,*
PMCID: PMC8106661  NIHMSID: NIHMS1692671  PMID: 33826955

Abstract

Dopamine neurons in the periaqueductal gray (PAG)/dorsal raphe are key modulators of antinociception with known supraspinal targets. However, no study has directly tested whether these neurons contribute to descending pain inhibition. We hypothesized that PAG dopamine neurons contribute to the analgesic effect of D-amphetamine via a mechanism that involves descending modulation via the rostral ventral medulla (RVM). Male C57BL/6 mice showed increased c-FOS expression in PAG dopamine neurons and a significant increase in paw withdrawal latency to thermal stimulation after receiving a systemic injection of D-amphetamine. Targeted microinfusion of D-amphetamine, L-DOPA, or the selective D2 agonist quinpirole into the PAG produced analgesia, while a D1 agonist, chloro APB, had no effect. In addition, inhibition of D2 receptors in the PAG by eticlopride prevented the systemic D-amphetamine analgesic effect. D-amphetamine and PAG D2 receptor-mediated analgesia were inhibited by intra-RVM injection of lidocaine or the GABAA receptor agonist muscimol, indicating a PAG-RVM signaling pathway in this model of analgesia. Finally, both systemic D-amphetamine and local PAG microinjection of quinpirole, inhibited inflammatory hyperalgesia induced by carrageenan. This hyperalgesia was transiently restored by intra-PAG injection of eticlopride, as well as RVM microinjection of muscimol. We conclude that D-amphetamine analgesia is partially mediated by descending inhibition and that D2 receptors in the PAG are responsible for this effect via modulating neurons that project to the RVM. These results further our understanding of the antinociceptive effects of dopamine and elucidate a mechanism by which clinically available dopamine modulators produce analgesia.

Keywords: Dopamine, Pain, Descending inhibition, Hyperalgesia, PAG, RVM

Introduction

The descending spinal pathway which projects from the periaqueductal gray (PAG) to the rostral ventral medulla (RVM) and terminates at the dorsal horn of the spinal cord is a powerful modulator of ascending nociceptive signals. The strength of this pathway was first demonstrated when electrical stimulation of the PAG produced sufficient analgesia to perform a laparotomy on rats (Reynolds, 1969). Since then, several studies have demonstrated that the PAG and RVM are primary sites of analgesic action by opioids (Pert and Yaksh, 1975; Manning et al., 1994; Lane et al., 2005) and cannabinoids (Hohmann et al., 2005).

It was long thought that antinociceptive actions of this pathway were mediated by opioid inhibition of GABAergic interneurons in the PAG (Vaughan and Christie, 1997; Vaughan et al., 1997) leading to activation of glutamatergic projections to the RVM (Wiklund et al., 1988; Beitz, 1990; Reichling and Basbaum, 1990). In turn, activation of RVM neurons with terminal projections in the dorsal horn of the spinal cord would result in inhibition of ascending nociceptive signals. However, recent studies have demonstrated that the circuit is considerably more complex. There are both antinociceptive “OFF-cells” and pro-nociceptive “ON-cells” within the RVM (Mason, 2012). Both excitatory glutamatergic projections from the PAG to RVM “OFF-cells,” as well as inhibitory GABAergic projections from the PAG to RVM “ON-cells,” contribute to the antinociceptive effects of opioid binding in the PAG (Morgan et al., 2008).

There is a less well studied population of dopaminergic (DA) neurons which reside within the ventrolateral PAG and dorsal raphe. They also modulate antinociception, although the mechanism by which they do so has not been well delineated (Suckow et al., 2013). This dopamine network is composed of two histologically distinct neuron types: small and large (Hasue and Shammah-Lagnado, 2002; Flores et al., 2004). Chemical lesion of these neurons attenuate opioid-induced antinociception (Flores et al., 2004) while local injection of the dopamine receptor agonist apomorphine into the ventrolateral PAG produces antinociception (Meyer et al., 2009; Schoo et al., 2017). Likewise, direct designer receptor exclusively activated by designer drugs (DREADD) activation of these neurons induced antinociception in the thermal paw-withdrawal test (Taylor et al., 2019b), and the D2-like receptor agonist piribedil induced antinociception in the mechanical paw-withdrawal test, but not in the tail-flick test (Tobaldini et al., 2018). Despite this progress, it is not known whether the analgesic effects are mediated via ascending supraspinal pathways or descending pathways via the RVM.

PAG DA neurons make up a full 50% of the DA projections to the central nucleus of the amygdala and stria terminalis, and have important but quantitatively fewer projections to the nucleus accumbens (NAc), caudate-putamen, lateral habenula, hippocampus, magnocellular basal forebrain, lateral septum and medial prefrontal cortex (Yetnikoff et al., 2014). Therefore, when Li et al demonstrated that optogenetic activation of PAG/dorsal raphe dopamine neurons result in dopamine release in the bed nucleus of the stria terminalis (BNST), they concluded that PAG dopamine neuron-mediated antinociception was due to supraspinal neural activity (Li et al., 2016). They also based this conclusion on the observation that DREADD activation of PAG dopamine neurons increased paw withdrawal latency in the hotplate test while producing no change in latency in the tail flick test, similar to the results seen by Tobaldini et al (2018).

However, neither study attempted to directly test the contribution of descending inhibition on PAG dopamine neuron antinociception. While there are no direct DA projections from the PAG to RVM (Li et al., 2016), there is evidence that dopamine may be able to modulate this circuit. Dopamine potently reduces GABA-mediated inhibitory postsynaptic currents (IPSCs) in the PAG, an effect blocked by α-flupenthixol (Meyer et al., 2009), and D2 receptors colocalize with opioid receptors on GABA neurons. Therefore, dopamine could indirectly modulate antinociception through the RVM either by increasing inhibition of “ON-cells” by GABAergic PAG projections, or by decreasing GABA interneuron inhibition of glutamatergic outputs to “OFF-cells” in the RVM. We hypothesized that increased hind paw withdrawal thresholds produced by dopamine receptor activation in the PAG/dorsal raphe could be prevented by inhibiting neural transmission through the RVM.

To test this hypothesis, we systemically treated mice with D-amphetamine and evaluated the hind paw withdrawal latency to thermal stimulation. D-amphetamine, L-DOPA, and selective D1 and D2 receptor agonists were infused into the PAG to evaluate the effect of dopaminergic activation at this site, and how treatment with the GABAA receptor agonist muscimol or lidocaine into the RVM would affect the dopaminergic effect triggered in the PAG. In sequence we investigated the effects of systemic D-amphetamine or direct dopaminergic activation at the PAG in the context of inflammatory hyperalgesia induced by carrageenan and if their effect would be affected by DA antagonist at the PAG or microinjection of lidocaine or muscimol in the RVM. The findings suggested that D-amphetamine-induced analgesia is in part mediated by a descending modulatory pathway that may include alteration of activity in the transmission system, and that D2 receptors in the PAG are responsible for the analgesic effect via interaction with neurons that project to the RVM.

Experimental Procedures

Experimental Animals:

All animal procedures were reviewed and approved by the authors’ Institutional Animal Care and Use Committee. Adult male C57bl/6J mice (Jackson Laboratory, stock number 00064), Slc6a3 (or DAT)-Cre mice (Jackson Laboratory, stock number 006660) and parvalbumin (Pvalb)-IRES-Cre mice (Jackson Laboratory, stock number 008069) weighing 25–30g were used in the experiments. Mice were kept on a 12:12 h light/dark cycle (lights on at 7am, lights off at 7pm) with ad libitum access to food and water. Mice had a minimum of 1 week to recover from the cannula implantation surgery before the behavioral experiments. Every effort was made to minimize the number of animals used and their suffering.

Anterograde Tracers:

Adeno-associated (rAAV) vectors were purchased from the Penn Vector Core (University of Pennsylvania, Philadelphia, PA). For anterograde mapping of PAG dopamine and parvalbumin neurons and projections, the Cre-dependent vector incorporated the flip-excision (FLEX) switch to control expression of EGFP from the CAG promoter (rAAV2/1.pCAG.FLEX.EGFP.WPRE.bGH). rAAV serotype 1 produced the most widespread tropism throughout diverse brain areas and was chosen for this study as previously described (Oh et al., 2014).

Drugs:

The following compounds were used in this study: the inflammatory agent λ-carrageenan plant mucopolysaccharide (CARR, 1%), the dopamine releaser D-amphetamine hemisulfate salt (6mg/kg, systemically, or 50ng, locally), the D1 dopamine receptor agonist chloro APB (Chloro-APB hydrobromide, 500ng), the D2 dopamine receptor agonist quinpirole [(−)-Quinpirole hydrochloride, 50ng], the D1 dopamine receptor antagonist SCH-23390 [R(+)-SCH-23390 hydrochloride, 50, 500ng, 5μg], the D2 dopamine receptor antagonist eticlopride [S-(−)-Eticlopride hydrochloride, 50, 500ng, 5μg], the non-selective glutamate receptor antagonist kynurenic acid (kynurenate, 5, 50, 500ng), the sodium channel blocker lidocaine hydrochloride monohydrate (2%) (all from Sigma-Aldrich), and the GABAA receptor agonist muscimol (0.5, 5, 50ng) (from Tocris). All drugs were dissolved in distilled water and diluted in saline. The starting concentration of each drug was determined from previous studies: CARR (Ferrari et al., 2010), lidocaine (Morgan and Fields, 1994; Burgess et al., 2002), dopamine receptor agonists and antagonists (Flores et al., 2004; Chiou et al., 2013; Zhao et al., 2015; Antunes et al., 2020), muscimol and kynurenate (Heinricher and McGaraughty, 1998; Gilbert and Franklin, 2001; Martenson et al., 2009; Tobaldini et al., 2019). CARR was injected subcutaneously in the plantar surface of the hind paw (50 μL); D-amphetamine was injected intraperitoneally (i.p., 6 mg/kg), except in the experiment shown in Figure 2, panel A. The other compounds were injected into the PAG or the RVM (50nL) through the implanted cannulas.

Figure 2: D-amphetamine action at the PAG produces analgesia by activating D2-dopaminergic receptors.

Figure 2:

Open in a new tab

Panels A, B and C: Paw withdrawal latency to thermal stimulation was evaluated after intra-PAG injection (50nL) of D-amphetamine (50ng, panel A), the dopamine precursor L-DOPA (500ng, panel B), the D1 receptor agonist chloro-APB (500ng, panel C, open symbols), or the D2 receptor agonist quinpirole (50ng, panel C, closed symbols). While significant increase in the thermal nociceptive threshold (analgesia) over time was observed in the D-amphetamine-, L-DOPA- and quinpirole-treated groups [F(1.441, 4.324) = 78.66, *p = 0.0488, ***p = 0.0004 for D-amphetamine; F(1.711, 5.133) = 21.30, **p = 0.0039 for L-DOPA; F(1.146, 3.438) = 109.6, #p = 0.0129, ****p < 0.0001 for quinpirole, when compared to baseline, one-way repeated measures analysis of variance followed by Bonferroni posttest], no effect was observed in the chloro-APB group [open symbols, F(1.898, 5.694) = 1.745, p = 0.2555, non-significant (NS)], indicating that activation of D2, but not D1, dopaminergic receptors in the PAG produces thermal analgesia; Panel D: Vehicle (control, closed circles) or eticlopride (500ng, open circles) was injected into the PAG (50 nL) 5 min before systemic injection of D-amphetamine (6mg/kg, i.p.). Evaluation of the thermal nociceptive threshold over time showed that the analgesic effect produced by D-amphetamine [intra-PAG vehicle-treated group, F(1.541, 4.622) = 23.72, p = 0.0043, when the paw withdrawal latency over time is compared to baseline, one-way repeated measures analysis of variance followed by Bonferroni posttest] was significantly attenuated by the intra-PAG injection of eticlopride [F(1, 10) = 48.39, **p = 0.0039, ##p =0.0155, ###p = 0.0011, two-way repeated measures analysis of variance followed by Bonferroni posttest, when both groups are compared]. Of note, the injection of eticlopride in the PAG did not produce significant effect on thermal threshold by itself (black triangles, F(2.425, 16.97) = 1.062, p = 0.3794, NS, when the paw withdrawal latency over time is compared to baseline), suggesting that D2 receptors in the PAG play a role in the analgesia produced by D-amphetamine. The schematics on the top shows the experimental protocol and the site of the treatments. (n=8 per group)

Stereotaxic surgery:

All mice were anesthetized with 2% isoflurane and placed in a stereotaxic frame (David Kopf Instruments, Tujunga, California). A skin incision followed by craniotomies were made above the target region. DAT-Cre mice received right-sided viral injections using glass pipettes (inner tip diameter of 10–20 μm) loaded with virus targeting the PAG (−4.36mm anterior/posterior, 0.18mm lateral, and −2.1mm dorsal/ventral to bregma) with 200nL of virus (rAAV2/1.pCAG.FLEX.EGFP.WPRE.bGH). Currents were applied for iontophoresis of rAAV particles. The majority of injections were done using 3 μA at 7 sec on/7 sec off cycle for 5 min total. Pvalb-IRES-Cre mice received right-sided injections targeting the PAG (−4.36mm anterior/posterior, 0.7mm lateral, and −2.2mm dorsal/ventral to bregma) with 500nL of virus.

Cannulas were implanted stereotaxically in C57bl/6J mice [PAG cannula, A/P: −4.6mm, M/L: 0.3mm, D/V: −2.5mm from bregma; RVM cannula, A/P: −5.8mm, M/L: 0mm, D/V: −5.7mm] and secured in place by dental acrylic (Taylor et al., 2019a). The incisions were then sutured closed and Carprofen (5mg/kg, from Putney, Portland, diluted in saline) was injected subcutaneously for post-operative care. Mice were allowed to recover for 1 week before the experiments were performed. After the behavioral tests were concluded, mice were perfused with phosphate buffered saline followed by neutral buffered formalin, and the brains harvested for tracer identification and histologic confirmation of correct cannula placement in the PAG or RVM. The brains were post-fixed in formalin overnight and sliced at 60 microns using a Leica VT1200 S vibratome (Leica Microsystems Inc., Buffalo Grove, IL). Correct location of the cannulas at the PAG and the RVM were confirmed by comparison of the images taken with a Zeiss Axio M2 microscope (Zeiss, Oberkochen, Germany) with images from a Mouse Brain Atlas (Paxinos et al., 2001) used as reference.

Evaluation of the thermal nociceptive threshold:

To evaluate nociceptive thresholds, paw withdrawal latencies to thermal stimulation were assayed by the Plantar Thermal Test (Hargreaves Apparatus, Ugo Basile, Model 37370), as previously described (Hargreaves et al., 1988). Briefly, a radiant heat was applied to the plantar surface of the mouse hind paw and the latency to evoke a withdrawal was measured. The protocol used in this study included a period for mouse habituation to the pain behavior assessment chambers. Specifically, mice were put in the chambers (acrylic enclosures that allowed the observation of the animal behavior) for 30 min, daily for 5 consecutive days, prior to the cannula implantation, and the paws received the thermal stimulation every 5 min, 3 to 5 times over 25 min. On the day before the surgeries, the baseline thermal thresholds were determined. After the recovery from the cannula implantation (at least a week later), a second baseline evaluation was performed, in order to determine the impact of the surgery on pain threshold. Of note, no significant difference in the thermal nociceptive sensitivity before and after the surgeries was observed (data not shown). The thermal tests started 5 min after the administration of the drugs in the PAG or RVM (or both), depending on the experiment.

Inflammatory pain model:

Inflammatory thermal hyperalgesia was produced by subcutaneous injection of CARR (1%, 50μL) in the plantar surface of the mouse hind paw under general anesthesia (2% isoflurane), using a 27-gauge hypodermic needle adapted to a Hamilton syringe (Reno, NV). Paw withdrawal latency was measured before and 3h after CARR administration, and a significant decrease in the latency was observed, in agreement with previous studies (Hargreaves et al., 1988).

cFOS Immunohistochemistry:

Immediately after the experiment shown in Figure 1, left panel, was completed, mice were administered sodium pentobarbitol (i.p.; Sigma-Aldrich, St. Louis, MO) and transcardially perfused with 0.1 M phosphate buffer (PB; pH 7.4) followed by neutral buffered formalin. The brains were post-fixed in formalin overnight and sliced at 50 microns using a Leica VT1200 S vibratome (Leica Microsystems Inc., Buffalo Grove, IL). Serial sections were collected from the PAG through the ventral tegmental area (VTA) (bregma: −0.5 to −2.5 mm). Sections were washed 5 × for 5 min each in 0.1 M PB. Following washing and blocking [0.1 M PB with 0.2% Triton X-100 containing 5% normal donkey serum (Jackson ImmunoResearch Laboratories Inc., West Grove, PA)], the sections were incubated in 0.1 M PB with mouse Anti-Tyrosine Hydroxylase (1:1000 dilution, Millipore Cat #MAB318, Temecula, CA) and rabbit anti-c-FOS (1:1000; Santa Cruz Biotechnology, Inc., Cat #SC-52, Santa Cruz, CA) antibodies overnight at 4°C. Sections were then washed 3 times with 0.1 M PB, followed by incubation with goat anti-mouse (Alexa 488, 1:200 dilution, catalog no. AB150113; ABCam) and goat anti-rabbit (Alexa 568,1:200 dilution, Cat #AB150077; ABCam) fluorescently-labeled secondary antibodies for 2h at room temperature. Cells were counterstained with 4’,6-diamidino-2-phenylindole (DAPI, catalog no. H-1200 Vectashield) for nuclear visualization. Images were taken with a Zeiss Axio M2 microscope (Zeiss, Oberkochen, Germany).

Figure 1: Systemic D-amphetamine produces analgesia and activates dopaminergic neurons in the PAG.

Figure 1:

Open in a new tab

Systemic injection of D-amphetamine (D-amph, 6mg/kg, i.p.) produced significant increase in paw withdrawal latency to thermal stimulation, evaluated 15 min after injection (on the left, t26 = 20.59, ****p < 0.0001, when compared to the saline group, unpaired Student’s t-test), indicating its analgesic effect. It also increased c-FOS (on the right, panel F) in PAG neurons. c-FOS activation co-localized with dopaminergic neurons in the PAG (arrows, panel H), suggesting these neurons might participate in D-amphetamine-induced analgesia. (n=14 per group)

Statistical Analysis:

In all behavior experiments, the dependent variable was hind paw-withdrawal latency to thermal stimulation, considered as the nociceptive threshold, and expressed in seconds. In the experiments shown in Figures 1 and 3, unpaired Student’s t-test was used to determine the magnitude of the analgesia produced by systemic treatment with D-amphetamine, compared to a control; paired Student’s t-test was used to determine the decrease in the thermal nociceptive threshold (hyperalgesia) produced by CARR, shown in Figures 6 (t39=20.92, p < 0.0001, when paw withdrawal latency before and 3h after CARR are compared) and 7 (t35=22.33, p < 0.0001). To analyze the impact of individual treatments on nociceptive threshold over time [Figure 2A, 2B, 2C and 2D (systemic D-amphetamine + vehicle group), Figure 5 (intra-PAG quinpirole, muscimol (RVM) control, lidocaine (RVM) control groups), and Figure 6 (eticlopride (PAG) or muscimol (RVM) control groups)], one-way repeated measures analysis of variance followed by Bonferroni posttest was performed, which was also used to analyze the effect of different doses/drugs injected into the RVM on systemic D-amphetamine-induced analgesia, shown in Figure 3. To evaluate the effect of different compounds injected into the PAG or the RVM, depending on the experiment, on the analgesia produced by systemic D-amphetamine [Figures 2D (systemic D-amphetamine + vehicle or eticlopride (PAG), and eticlopride (PAG) control, groups) and 6], or by intra-PAG quinpirole (Figures 5 and 7), two-way repeated measures analysis of variance followed by Bonferroni posttest was used. GraphPad Prism 8 (GraphPad Software, Inc, San Diego, CA) was used to plot the graphics and to perform statistical analysis; p < .05 was considered statistically significant. Of note, the results shown in Figures 2, 5, 6 and 7 represent the standard deviation of the mean of the readings (nociceptive thresholds) obtained from each experimental group at each time point; the results presented as box-plots (Figures 1 and 3) show the first and third quartiles (bottom and top of the boxes, respectively), and the 5–95 percentiles (whiskers below and above the boxes). The black horizontal line indicates the median.

Figure 3: Neural transmission through the RVM participates in the D-amphetamine-induced analgesia.

Figure 3:

Open in a new tab

The effect on systemic D-amphetamine (D-amph)-induced analgesia of different doses of compounds administered in the RVM was evaluated. 50 nL of the D1 antagonist SCH-23390 (50, 500ng, 5μg, middle section, gray boxes), the D2 antagonist eticlopride (50, 500ng, 5μg, middle section, black boxes on the right), the glutamate receptor antagonist kynunerate (5, 50, 500ng, lower section, gray boxes), or the GABAA receptor agonist muscimol (0.5, 5, 50ng, lower section, black boxes on the right) were injected into the RVM 5 min after systemic injection of D-amphetamine (6 mg/kg, i.p.), and the paw withdrawal latency to thermal stimulation was evaluated 20, 25 and 30 min later. The boxes represent the average of the three measures, for each dose. Injection of D-amphetamine only [D-amph (control), boxes with dotted pattern] produced robust increase in the paw withdrawal threshold (t33=39.68, p < 0.0001, when compared to the control, vehicle-treated group, unpaired Student’s t-test), that was not affected by intra-RVM injection of SCH-23390 [F(3, 28) = 1.516, p = 0.2320, NS, one-way repeated measures analysis of variance followed by Bonferroni posttest], eticlopride [F(3, 31) = 1.460, p = 0.2446, NS] or kynunerate [F(3, 28) = 0.1617, p = 0.9212, NS]. On the other hand, a dose-dependent decrease in the paw withdrawal latency was observed in the groups that received muscimol [F(3, 33) = 124.8, *p = 0.0172, ****p < 0.0001], indicating a role of RVM GABAA receptors in the analgesic effect of D-amphetamine. The schematics on the top shows the experimental protocol and the site of the treatments. (control group: n=21; D-amphetamine-only group: n=14; SCH-23390-treated and kynunerate-treated groups: n=6 per group; eticlopride-treated group: n=7; muscimol-treated group: n=8)

Figure 6: D-amphetamine produces anti-hyperalgesia by activating a PAG-RVM signaling pathway.

Figure 6:

Open in a new tab

Carrageenan (CARR, 1%, 50 μL) was injected subcutaneously into the plantar surface of the hind paw, and the thermal nociceptive threshold was evaluated before and 3h later. Mice were then divided into groups that received systemic (i.p.) injection of vehicle (open circles), only D-amphetamine (D-amph, 6mg/kg, i.p., closed circles), or D-amphetamine preceded by intra-PAG injection of eticlopride (500ng, closed squares in panel A), or by intra-RVM injection of muscimol (50ng, closed squares in panel B), both 50 nL. CARR produced robust decrease in paw withdrawal latency to thermal stimulation that was inhibited by D-amphetamine [F(6, 60) = 38.51, **p = 0.0026, ***p = 0.0002, ****p < 0.0001, when the time course of vehicle- and D-amphetamine-treated groups are compared, two-way repeated measures analysis of variance followed by Bonferroni posttest]. However, the effect of D-amphetamine was significantly attenuated in the groups that received eticlopride in the PAG [F(5, 50) = 3.136, p = 0.0154, when the D-amph alone and D-amph+eticlopride groups are compared] or muscimol in the RVM [F(5, 30) = 3.924, p = 0.0074, when the D-amph alone and D-amph+muscimol groups are compared], indicating that D-amphetamine attenuates thermal hyperalgesia produced by inflammation through a signaling pathway involving the PAG and the RVM. Of note, while the inhibition of the D-amphetamine anti-hyperalgesia by eticlopride was significant until 30 min (panel A: #p = 0.0108, ##p = 0.0074), the effect of muscimol was no longer present after 30 min (panel B: ****p < 0.0001). Evaluation over time of control groups that received CARR and only eticlopride at the PAG (panel A, open squares) or muscimol at the RVM (panel B, open squares), showed no impact of these treatments on CARR-induced thermal hyperalgesia (F(2.363, 16.54) = 2.894, p = 0.0766, NS, for eticlopride (PAG); F(2.544, 17.81) = 0.3296, p = 0.7724, NS, for muscimol (RVM), when the paw withdrawal threshold over time is compared to pre-injection values, one-way repeated measures analysis of variance followed by Bonferroni posttest). [all groups n=6, except the Eticlopride (PAG) and the Muscimol (RVM) groups (n=8)]

Figure 5: Analgesia produced by activation of D2 receptors in the PAG involves downstream signaling through the RVM.

Figure 5:

Open in a new tab

Quinpirole (50ng, 50 nL) was injected into the PAG 5 min after intra-RVM injection of muscimol (50ng, left panel, closed symbols) or lidocaine (2%, right panel, closed symbols). Significant increase in paw withdrawal latency, i.e., analgesia, was observed in the quinpirole-treated group [open circles, both panels, F(2.885, 31.74) = 31.90, p < 0.0001, when the withdrawal latency is compared to baseline (“0”) over time, one-way repeated measures analysis of variance followed by Bonferroni posttest]. Quinpirole-induced analgesia was significantly attenuated by the pretreatment with either muscimol [F(5, 75) = 18.25, ***p < 0.0004, ****p < 0.0001, **p = 0.0082, ##p = 0.0048, two-way repeated measures analysis of variance followed by Bonferroni posttest] or lidocaine [F(5, 75) = 5.155, ++p = 0.0090, +p = 0.0082, ###p = 0.0005, *p = 0.0122, +++p = 0.0012] into the RVM. Of note, while lidocaine only inhibited the effect of quinpirole, keeping the thermal nociceptive threshold near baseline levels [F(2.797, 19.58) = 0.6059, p = 0.6081, NS, one-way repeated measures analysis of variance followed by Bonferroni posttest], intra-RVM muscimol seemed to produce a small, although not statistically significant, thermal hyperalgesia [i.e., a decrease in paw withdrawal latency, F(2.385, 9.541) = 2.213, p = 0.1584]. In control groups (both panels, open squares), injection of muscimol or lidocaine alone in the RVM produced no significant impact on the nociceptive threshold over time [F(2.791, 13.95) = 3.177, p = 0.0601 and F(1.713, 8.567) = 0.6476, p = 0.5243, respectively]. [Quinpirole (PAG) group: n=12; Muscimol (RVM) and Lidocaine (RVM) groups: both n=6; Muscimol (RVM) + Quinpirole (PAG) and Lidocaine (RVM) + Quinpirole (PAG) groups: both n=5]

Figure 7: Direct activation of PAG D2 receptors signals to RVM to produce anti-hyperalgesia.

Figure 7:

Open in a new tab

Carrageenan (CARR, 1%, 50 μL) was injected subcutaneously into the plantar surface of the hind paw. Thermal nociceptive threshold evaluation before and 3h later showed robust decrease in the thermal paw withdrawal threshold (t35 = 22.33, p < 0.0001, unpaired Student’s t-test). Mice were then divided in 4 groups, including the CARR-only group (open circles), a group that received lidocaine (2%, 50 nL) into the RVM (open triangles), a group treated with quinpirole into the PAG (50ng, 50 nL, open squares), and a group that received quinpirole into the PAG 5 min after intra-RVM injection of lidocaine (closed squares). Injection of quinpirole into the PAG inhibited the thermal hyperalgesia induced by CARR, increasing the paw withdrawal latency to baseline levels [F(6, 96) = 7.932, ****p < 0.0001, when both groups are compared, two-way repeated measures analysis of variance followed by Bonferroni posttest]. However, the anti-hyperalgesia induced by quinpirole injection in the PAG was significantly attenuated by the pretreatment with lidocaine into the RVM [F(6, 108) = 9.572, **p = 0.0098, ***p = 0.0002, ****p < 0.0001, ##p = 0.0056, when the groups quinpirole alone and quinpirole + lidocaine are compared], confirming that the RVM is downstream in the effect produced by activation of PAG D2 receptors. Of note, no effect of the injection of lidocaine into the RVM on CARR-induced thermal hyperalgesia was observed (F(6, 84) = 1.208, p = 0.3103, NS, when the CARR and CARR + Lidocaine (RVM) groups are compared). (CARR-only group: n=6; CARR + Lidocaine (RVM): n=10; CARR + Quinpirole (PAG): n=12; CARR + Quinpirole (PAG) + Lidocaine (RVM): n=8)

Results

Systemic injection of D-amphetamine activates dopaminergic neurons in the PAG:

D-amphetamine was systemically administered by intraperitoneal (i.p.) injection to C57bl/6J mice and the paw withdrawal latency to thermal stimulation was measured 15 min later. An increase in the paw withdrawal threshold was observed in the group treated with D-amphetamine compared to a control group injected with vehicle (p < 0.0001, t26 = 20.59, n=14, Figure 1, left side). Mice were then immediately euthanized, and the brains harvested to investigate the neural locations of D-amphetamine action using immunohistochemical analysis of c-FOS expression. Along with previously reported effects in the VTA and NAc (Shi et al., 2000; Covey et al., 2016) (data not shown), we observed increased c-FOS expression in a group of neurons located in the ventrolateral PAG/dorsal raphe (Figure 1, right side, panel F). Saline-treated mice failed to show c-FOS expression (panel B). The c-FOS activation co-localized with dopaminergic neurons (arrows, panel H), stained by tyrosine hydroxylase (panels C, D, G, H).

D-amphetamine produces analgesia by activation of D2-dopaminergic receptors in the PAG:

Since systemic injection of D-amphetamine both activated PAG dopamine neurons and induced analgesia, we next investigated whether direct D-amphetamine injection into the PAG would produce the same effect. We observed a significant increase in paw withdrawal latency after injection of D-amphetamine into the PAG, with rapid onset and peak effect at 30 min after injection (p = 0.0005, F(1.441, 4.324) = 78.66, n=8, Figure 2, panel A). As the mechanism of action of D-amphetamine involves release of dopamine, we also evaluated if PAG injection of the dopamine precursor L-DOPA would similarly affect the paw nociceptive threshold. We saw a significant increase in paw withdrawal latency 45 min post L-DOPA treatment (p = 0.0039, F(1.711, 5.133) = 21.30, n=8, Figure 2, panel B). We next investigated whether activation of D1-like or D2-like receptors would produce analgesic effect. Following the same protocol, the D1-selective agonist chloro APB, or the D2-selective agonist quinpirole, were injected in the PAG and the paw withdrawal threshold to thermal stimulation was evaluated. We found that quinpirole (p = 0.0009, F(1.146, 3.438) = 109.6, n=8), but not chloro APB (p = 0.2555, F(1.898, 5.694) = 1.745, non-significant, n=8), produced analgesia almost immediately after injection (Figure 2, panel C). To determine if these receptors participate in the analgesic effect of D-amphetamine, we injected the D2 receptor antagonist eticlopride into the PAG of mice 5 min before systemic D-amphetamine injection. Eticlopride administered in the PAG attenuated the analgesic effect of D-amphetamine (Figure 2, panel D) when compared to the control group (D-amphetamine + vehicle, p < 0.0001, F(1, 10) = 48.39, n=8), suggesting that systemic D-amphetamine triggers D2 receptor activation at the PAG to produce analgesia. Of note, eticlopride by itself did not produce significant effect on the paw withdrawal latency (p = 0.3794, F(2.425, 16.97) = 1.062, non-significant, n=8) (Figure 2).

Neural transmission through the RVM participates in the D-amphetamine-induced analgesia:

We next sought to prevent D-amphetamine-induced analgesia by blocking neural transmission through the RVM. Following systemic injection of D-amphetamine, the D1 receptor antagonist SCH-23390, the D2 receptor antagonist eticlopride, a broad-spectrum excitatory amino acid antagonist kynurenate, or the GABAA receptor agonist muscimol were directly injected into the RVM in separate groups of mice, and paw withdrawal latency to thermal stimulation was evaluated 20, 25, and 30 min later. No effect was observed with increasing doses of SCH-23390 (p = 0.2320, F(3, 28) = 1.516, NS, n=6), eticlopride (p = 0.2446, F(3, 31) = 1.460, NS, n=7), or kynurenate (p = 0.9212, F(3, 28) = 0.1617, NS, n=6). In contrast, muscimol dose-dependently attenuated the analgesia produced by D-amphetamine (p = 0.0172 for the 0.5ng dose, and p < 0.0001 for the 5 and 50ng doses, F(3, 33) = 124.8, n=8), suggesting that transmission through the RVM could be mediating this process. Since the most pronounced effect of muscimol was observed with 50ng, we chose this dose – which does not produce significant effect on paw withdrawal threshold per se, as seen in Figure 5 - to be used in subsequent experiments. (Figure 3).

Antegrade Tracing of PAG neurons:

To provide anatomic confirmation for the efficacy of muscimol as well as the lack of effectiveness of dopamine antagonists infused in the RVM, antegrade tracing was used to locate cell bodies of dopaminergic- and parvalbumin-positive GABAergic neurons in the PAG and trace their axonal projections to determine if these cells project to the RVM (Oh et al., 2014). Adult male mice expressing Cre recombinase in dopaminergic cells carrying the dopamine transporter (DAT-Cre) underwent unilateral PAG injection of an rAAV carrying a Cre-dependent EGFP fluorescent tag. As seen in Figure 4, upper panel, dopaminergic cell bodies span from the dorsolateral to the ventrolateral columns and beyond to the dorsal raphe (http://connectivity.brainmap.org/projection/experiment/siv/160538548?imageId=160538936&imageType=TWO_PHOTON,SEGMENTATION&initImage=TWO_PHOTON&x=16981&y=12272&z=3). After completing all microinjection experiments, cannula tip locations were superimposed on the tracing image to verify that drug injections correctly targeted the immediate proximity of dopaminergic neurons.

Figure 4: Parvalbumin-positive PAG cells project to the RVM, while dopaminergic neurons do not.

Figure 4:

Open in a new tab

The sections on the left panels (upper, PAG; lower, RVM) show the antegrade tracing of dopaminergic- and parvalbumin-positive GABAergic neurons, with their cell bodies seen in the PAG, and their axonal projections in the RVM, indicating that PAG dopaminergic neurons do not project to the RVM, while parvalbumin-positive GABAergic neurons do. After the behavior experiments were concluded, the brains of the animals were harvested and prepared for the confirmation of the location of the infusion cannulas at the correct sites. The circles superimposed on the images represent the location of the cannulas targeting the PAG and the RVM. Image credit: Allen Institute.

To confirm the report that there was a population of GABAergic neurons with projections from the PAG to RVM (Morgan et al., 2008), mice expressing Cre recombinase in parvalbumin-positive GABAergic cells underwent unilateral PAG injection of an rAAV carrying a Cre-dependent EGFP fluorescent tag. Axonal staining was traced from the PAG down to the Magnocellular reticular nucleus of the RVM (Figure 4, lower panel) (http://connectivity.brainmap.org/projection/experiment/siv/267030155?imageId=267030611&imageType=TWO_PHOTON,SEGMENTATION&initImage=TWO_PHOTON&x=18788&y=10974&z=1). Cannula tip locations were subsequently superimposed on the tracing image to verify that drug injections correctly targeted the immediate proximity of these parvalbumin-positive GABAergic axons (Figure 4).

Activation of D2 receptors in the PAG produces analgesia by signaling downstream to the RVM:

We next investigated whether the analgesia produced by direct D2 activation at the PAG depends on a downstream signaling through the RVM. Mice were prepared with infusion cannulas targeting both the PAG and RVM. Infusion of the D2 agonist quinpirole in the PAG produced a significant increase in paw withdrawal latency to thermal stimulation (p < 0.0001, F(2.885, 31.74) = 31.90, n=12, Figure 5). The PAG D2-induced analgesia was totally inhibited by injection of muscimol in the RVM (p < 0.0001, F(5, 75) = 18.25, n=5, Figure 5, left panel), confirming that the RVM is downstream the PAG in this inhibitory pathway. Interestingly, the injection of muscimol in the RVM, in addition to inhibiting the effect of quinpirole, induced a small, although not significant, thermal hyperalgesia (p = 0.1584, F(2.385, 9.541) = 3.492, n=5). This occurred despite having no significant effect on paw withdrawal threshold with RVM injection in the absence of PAG stimulation [Muscimol (RVM) group, open squares, p = 0.0601, F(2.791, 13.95) = 3.177, NS, n=6]. For added rigor, we treated a separate group of mice with lidocaine in the RVM, and found that it also inhibited the analgesia produced by PAG injection of quinpirole (p = 0.0004, F(5, 75) = 5.155, n=5). Importantly, lidocaine injection in the RVM in the absence of PAG stimulation had no effect on paw withdrawal thresholds (p = 0.5243, F(1.713, 8.567) = 0.6476, NS, n=6, Figure 5).

D-amphetamine produces anti-hyperalgesia via PAG and RVM neurons in a model of inflammatory pain:

Since we observed that dopamine neurons in the PAG and neural transmission via the RVM participate in D-amphetamine-mediated increase in the thermal nociceptive threshold, we next evaluated if those neurons also modulate D-amphetamine effects on nociception during an inflammatory condition. We used the model of inflammatory hyperalgesia produced by subcutaneous injection of carrageenan (CARR, 1%) into the plantar surface of the hind paw. CARR produced a robust decrease in the paw withdrawal latency, evaluated 3h after injection (p < 0.0001, t39=20.92), that was significantly attenuated by systemic injection of D-amphetamine (p < 0.0001, F(6, 60) = 38.51, n=6). Eticlopride administered in the PAG, or muscimol microinjected in the RVM 5 min prior to systemic D-amphetamine injection significantly but transiently inhibited the anti-hyperalgesia (p = 0.0154, F(5, 50) = 3.136, for eticlopride + D-amphetamine; p = 0.0074, F(5, 30) = 3.924 for muscimol + D-amphetamine, when compared to the D-Amphetamine-only group, n=6, Figure 6).

D2 receptors in the PAG/Dorsal Raphe modulate peripheral inflammatory hyperalgesia via the RVM:

Because D-amphetamine was administered systemically in the previous experiment, the transient effect of the treatments with eticlopride or muscimol on D-amphetamine-induced anti-hyperalgesia could possibly involve compensatory actions of D-amphetamine at other targets. Therefore, the effects of direct PAG D2 activation and RVM inhibition were tested. CARR-induced thermal hyperalgesia was significantly inhibited by intra-PAG injection of quinpirole (p < 0.0001, F(6, 96) = 7.932, n=12). In contrast, RVM injection of lidocaine 5 min prior to quinpirole abolished the anti-hyperalgesic effect (p < 0.0001, F(6, 108) = 9.572, when the groups quinpirole alone and quinpirole + lidocaine are compared, n=8, Figure 7).

Discussion

Here we report that dopamine neurons in the PAG/dorsal raphe participate in descending pain inhibition, rather than the purely cortical role suggested by others (Li et al., 2016). Male C57BL/6 mice showed a significant increase in paw withdrawal latency to thermal stimulation after receiving a systemic (i.p.) injection of D-amphetamine and showed evidence of activated dopamine neurons in the PAG, as demonstrated by increased c-FOS expression. Targeted microinfusion of D-amphetamine, L-DOPA, or a selective D2 receptor agonist into the PAG increased paw-withdrawal threshold, while a D1 agonist had no effect. In addition, antagonizing D2 receptors in the PAG using eticlopride restored thermal nociception in the presence of systemic D-amphetamine. D-amphetamine and PAG D2 receptor-mediated antinociception were prevented when neural transmission through the RVM was inhibited by local pretreatment with the GABAA receptor agonist muscimol or with lidocaine. Finally, both systemic D-amphetamine and local PAG microinjection of the D2 agonist quinpirole, blocked the inflammatory hyperalgesia induced by CARR. This inflammatory hyperalgesia was transiently restored with intra-PAG injection of eticlopride, as well as microinjection of lidocaine or muscimol in the RVM. We conclude that D-amphetamine analgesia is partially mediated by descending inhibition and that D2 receptors in the PAG are responsible for this effect via modulating neurons that project to the RVM.

Neural dopamine effects are most commonly associated with arousal, locomotion and reward. However, neural dopamine is also an important pain modulator. The last major clinical study investigating the use of a dopamine releasing drug as an antinociceptive adjuvant used D-amphetamine (Forrest et al., 1977). We therefore administered D-amphetamine systemically by intraperitoneal (i.p.) injection to C57bl/6J mice and found that it produced anti-hyperalgesia. The claim that dopaminergic agonists are analgesic is controversial. Sensory processing is key to the pain experience, and some studies have shown that sensory thresholds in humans do not change with increasing dopamine activity (Becker et al., 2013). On the other hand, it is also well known that the experience of pain involves neural circuits that underlie emotion and motivation. Dopamine agonists produce analgesia in a subset of patients with Parkinson’s disease (Ha and Jankovic, 2012; Skogar and Lokk, 2016) and may, therefore, be considered pain modulators.

Descending modulatory pathways to the spinal cord comprise, among others, noradrenergic, serotonergic, γ-aminobutyric acid GABAergic, and dopaminergic fibers. (Millan, 2002; Lau and Vaughan, 2014; Bannister and Dickenson, 2017; Chen and Heinricher, 2019). While there has been recent progress in understanding the contributions of dopamine on descending pain modulation, much remains poorly understood (Li et al., 2016; Li et al., 2019; Yu et al., 2021). We observed increased c-FOS expression in response to D-amphetamine administration in a group of neurons located in the ventrolateral PAG/dorsal raphe. Because this is an important pain modulation region (Taylor et al., 2019b), we hypothesized that dopamine released in the PAG as consequence of D-amphetamine could have important effects in modulating pain. We therefore administered D-amphetamine directly to the PAG and observed the effects on thermal hyperalgesia. D-amphetamine causes the immediate release of endogenous stores of monoamines from nerve terminals into synapses, with released DA levels 10x higher than the other monoamines (Heal et al., 2013). Because it is known that other monoamines also have important modulating effects in descending inhibition, we also administered the dopamine precursor L-DOPA to more directly explore the role of dopamine in these systems.

In the present study, we demonstrate that both D-amphetamine and L-DOPA acting in the PAG produce significant analgesia. The effect onset is different in the two drugs, possibly due to differences in their pharmacodynamics. D-amphetamine causes the immediate release of monoamines and resulted in immediate antinociception. However, this antinociception did not appear to be sustained, perhaps as a result of rapid reuptake and scavenging. On the other hand, the onset of L-DOPA analgesia was delayed, reflecting the need for enzymatic conversion to dopamine. Unlike D-amphetamine and L-DOPA, the anti-nociceptive effects of the D2-like receptor agonist quinpirole was prolonged, suggesting it wasn’t metabolized as quickly as dopamine. L-DOPA has been tested as a non-traditional analgesic in preclinical pain models (Cobacho et al., 2010; Skinner et al., 2011), and clinically in bone cancer (Dickey and Minton, 1972; Nixon, 1975) and neuropathic pain (Kernbaum and Hauchecorne, 1981; Ertas et al., 1998). D-amphetamine produces antinociception when used alone (Morgan and Franklin, 1990) and potentiates opioid analgesia (Burrill et al., 1944; Goetzl et al., 1944; Ivy et al., 1944; Evans and Bergner, 1964; Evans and Smith, 1964; Evans, 1967; Webb et al., 1972). A large, multi-site, double-blind, prospective study concluded that the combination of 10 mg D-amphetamine and morphine was twice as potent as morphine alone in providing pain relief postoperatively while also offsetting the sedation and loss of alertness that accompanies morphine use (Forrest et al., 1977).

While dopamine modulating drugs have never been adopted clinically for the relief of pain, there is increasing evidence that dopamine has important effects in modulating pain (Yu et al., 2021). Dopamine modulation of ascending nociceptive inputs may decrease nociceptive signals in the spinal cord, and thus prevent nociceptive information from being relayed to cortical areas in the thalamus, amygdala and prefrontal cortex (Chen and Heinricher, 2019). Dopamine may also modulate the response to these inputs in the cortex by reducing the distress that normally accompanies noxious stimulation in a phenomenon termed “affective analgesia” (Franklin, 1989, 1998). Several studies have investigated the effects of dopamine modulation on analgesia in the circuit between the VTA and the NAc (Altier and Stewart, 1998; Schifirnet et al., 2014), but no studies have investigated dopaminergic PAG to RVM connections.

To determine if D-amphetamine-induced analgesia is mediated by descending inhibition, we set out to prevent the analgesic response by blocking signal transmission through the RVM. There are known glutamatergic projections to the RVM which release glutamate to activate antinociceptive “OFF-cells”. The inability of RVM-administered kynurenate, a glutamate receptor antagonist, to prevent the antinociception of systemically administered D-amphetamine suggests that glutamatergic projections to RVM “OFF-cells” are not involved. Alternatively, it could mean that D-amphetamine antinociceptive effects at non-PAG/dorsal raphe target sites could simply be stronger. In addition, administration of D1 and D2 receptor antagonists in the RVM had no effect on hyperalgesia. This is consistent with the observation of others that there are no known dopaminergic projections to the RVM (Li et al., 2016). We confirmed this using Cre-dependent EGFP expression in PAG dopaminergic neurons in DAT-Cre mice (Figure 4). It has been reported that numerous GABAergic neurons project from the PAG to the RVM, synapsing on both GABAergic interneurons and pronociceptive “ON-cells” (Morgan et al., 2008). We show that at least some of these are parvalbumin-positive GABAergic neurons that terminate in the nucleus raphe magnus, nucleus gigantocellularis pars alpha, and lateral paragigantocellular nucleus, all of which are considered part of the RVM. The cannula targeting the RVM collocate with staining from the parvalbumin-positive GABAergic neurons that project from the PAG. The net effect of the GABAA receptor agonist muscimol in the RVM appears to be inhibitory, as previous studies have used it to prevent all neural transmission through the RVM (Martenson et al., 2009). For added rigor, we also used lidocaine to inhibit transmission via a second, separate mechanism. Previous studies showed that microinjection of the sodium channel inhibitor lidocaine in the RVM abolished opioid-induced analgesia (Fields et al., 1976; Sandkuhler and Gebhart, 1984).

The PAG is often divided into regions including, dorsal, dorsolateral, ventral lateral, etc, based on observed differences in behavior caused by electrical stimulation or lesion studies. These techniques are nonspecific, activating or destroying all neurons in the vicinity, including resident cell bodies and traversing axons of all cell types including glutamatergic, GABAergic, or dopaminergic cells. The resulting behavioral response depends on the end target of the neurons themselves, a fact elegantly demonstrated by (Tovote et al., 2016). In contrast to electrical stimulation, dopamine will be active where ever dopamine receptors lie. We therefore inserted our infusion cannula with the purpose of targeting dopaminergic neurons, rather than an arbitrary dorsal or ventral topography. We have previously shown that dopamine neurons in the area specifically modulate antinociceptive pathways in contrast to glutamatergic neurons, which modulate a host of behaviors including anxiety, antinociception and locomotion (Taylor et al, 2019).

We demonstrated that systemic D-amphetamine inhibited inflammatory hyperalgesia produced by hind paw injection of CARR. Interestingly, eticlopride injected into the PAG persistently prevented the analgesic effect of systemic D-amphetamine (Figure 2) but only transiently blocked the effect in the presence of inflammatory hyperalgesia produced by hind paw injection of CARR (Figure 6). Blockade of neural transmission through the RVM with muscimol also prevented D-amphetamine analgesic action only transiently. In contrast, blockade of neural transmission through the RVM with lidocaine persistently prevented quinpirole-induced analgesia when this was administered directly into the PAG. This strongly suggests that in the presence of hyperalgesia caused by inflammation, systemic D-amphetamine induces analgesic actions via additional, alternative pathways, while the analgesic effects of selective PAG D2-like receptor activation is predominately mediated through the RVM. Dopamine neurons from the PAG/dorsal raphe project to the VTA, BNST, and NAc (Li et al., 2016). Many studies have shown antinociceptive effects mediated by each of these regions. Optogenetic activation of VTA dopaminergic neurons reverses pathological allodynia resulting from nerve injury or bone cancer (Watanabe et al., 2018), while selective DREADD inhibition of rostromedial tegmental nucleus GABAergic neurons or activation of VTA dopaminergic neurons attenuates thermal hyperalgesia in animals with persistent inflammatory (Taylor et al., 2019a) or neuropathic pain (Wakaizumi et al., 2016). Several studies implicate the BNST in modulating emotional aspects of pain (Morano et al., 2008; Tran et al., 2014; Minami and Ide, 2015) and dopamine neurons from the PAG/dorsal raphe release glutamate and dopamine in the BNST (Li et al., 2016; Yu et al., 2021). Finally, NAc medium spiny neurons integrate inputs from the anterior cingulate cortex, prefrontal cortex, amygdala, and thalamus, which in turn modulate output to the ventral pallidum, lateral hypothalamus, and VTA (Thompson and Neugebauer, 2019). Many of these input and output regions have been implicated in aversive and/or nociceptive processing (Wulff et al., 2019). Consequently, changing the integration of nociceptive or sensory information in the NAc or altering accumbal output to the ventral pallidum/lateral hypothalamus/VTA could all potentially alter pain perception.

In summary, we show that D2 receptors in the PAG/dorsal raphe modulate peripheral inflammatory hyperalgesia via the RVM and elucidate part of a novel pathway whereby D-amphetamine and other dopamine medications modulate pain. These results further our understanding of the antinociceptive effects of dopamine.

Highlights:

  • D-amphetamine produces analgesia by activating dopaminergic neurons in the PAG

  • Direct activation of D2 dopaminergic receptors in the PAG induces analgesia

  • PAG D2 dopaminergic neurons-triggered analgesia is mediated through the RVM

  • GABAergic modulation at the RVM is involved in PAG D2-dependent analgesia

Acknowledgements

We acknowledge Cell Imaging Core at the University of Utah for use of equipment (Nikon A1 Confocal Microscope) and thank Michael Bridge for his assistance in image acquisition.

Funding

This work was supported by the National Institutes of Health (grant numbers K08-GM121951 and R35-GM138168).

List of Abbreviations

BNST

bed nucleus of the stria terminalis

CARR

carrageenan

DA

dopaminergic

DREADD

direct designer receptor exclusively activated by designer drugs

NAc

nucleus accumbens

PAG

periaqueductal gray

RVM

rostral ventral medulla

VTA

ventral tegmental area

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Altier N, Stewart J (1998) Dopamine receptor antagonists in the nucleus accumbens attenuate analgesia induced by ventral tegmental area substance P or morphine and by nucleus accumbens amphetamine. J Pharmacol Exp Ther 285:208–215. [PubMed] [Google Scholar]
  2. Antunes GF, Gouveia FV, Rezende FS, Seno MDJ, de Carvalho MC, de Oliveira CC, Dos Santos LCT, de Castro MC, Kuroki MA, Teixeira MJ, Otoch JP, Brandao ML, Fonoff ET, Martinez RCR (2020) Dopamine modulates individual differences in avoidance behavior: A pharmacological, immunohistochemical, neurochemical and volumetric investigation. Neurobiol Stress 12:100219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bannister K, Dickenson AH (2017) The plasticity of descending controls in pain: translational probing. J Physiol 595:4159–4166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Becker S, Gandhi W, Elfassy NM, Schweinhardt P (2013) The role of dopamine in the perceptual modulation of nociceptive stimuli by monetary wins or losses. Eur J Neurosci 38:3080–3088. [DOI] [PubMed] [Google Scholar]
  5. Beitz AJ (1990) Relationship of glutamate and aspartate to the periaqueductal gray-raphe magnus projection: analysis using immunocytochemistry and microdialysis. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society 38:1755–1765. [DOI] [PubMed] [Google Scholar]
  6. Burgess SE, Gardell LR, Ossipov MH, Malan TP Jr., Vanderah TW, Lai J, Porreca F (2002) Time-dependent descending facilitation from the rostral ventromedial medulla maintains, but does not initiate, neuropathic pain. J Neurosci 22:5129–5136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Burrill DY, Goetzl FR, Ivy AC (1944) The pain threshold raising effects of amphetamine. Journal of Dental Research 23:337–344. [Google Scholar]
  8. Chen Q, Heinricher MM (2019) Descending Control Mechanisms and Chronic Pain. Curr Rheumatol Rep 21:13. [DOI] [PubMed] [Google Scholar]
  9. Chiou RJ, Chang CW, Kuo CC (2013) Involvement of the periaqueductal gray in the effect of motor cortex stimulation. Brain Res 1500:28–35. [DOI] [PubMed] [Google Scholar]
  10. Cobacho N, De la Calle JL, Gonzalez-Escalada JR, Paino CL (2010) Levodopa analgesia in experimental neuropathic pain. Brain Res Bull 83:304–309. [DOI] [PubMed] [Google Scholar]
  11. Covey DP, Bunner KD, Schuweiler DR, Cheer JF, Garris PA (2016) Amphetamine elevates nucleus accumbens dopamine via an action potential-dependent mechanism that is modulated by endocannabinoids. Eur J Neurosci 43:1661–1673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dickey RP, Minton JP (1972) Levodopa relief of bone pain from breast cancer. N Engl J Med 286:843. [DOI] [PubMed] [Google Scholar]
  13. Ertas M, Sagduyu A, Arac N, Uludag B, Ertekin C (1998) Use of levodopa to relieve pain from painful symmetrical diabetic polyneuropathy. Pain 75:257–259. [DOI] [PubMed] [Google Scholar]
  14. Evans WO (1967) The effect of stimulant drugs on opiate induced analgesia. Arch Biol Med Exp 4:144–149. [Google Scholar]
  15. Evans WO, Smith RP (1964) Some effects of morphine and amphetamine on intellectual functions and mood. Psychopharmacologia 6:49–56. [DOI] [PubMed] [Google Scholar]
  16. Evans WO, Bergner DP (1964) A Comparison of the Analgesic Potencies of Morphine, Pentazocine, and a Mixture of Methamphetamine and Pentazocine in the Rat. The Journal of new drugs 4:82–85. [DOI] [PubMed] [Google Scholar]
  17. Ferrari LF, Gear RW, Levine JD (2010) Attenuation of activity in an endogenous analgesia circuit by ongoing pain in the rat. J Neurosci 30:13699–13706. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Fields HL, Anderson SD, Clanton CH, Basbaum AI (1976) Nucleus raphe magnus: a common mediator of opiate- and stimulus-produced analgesia. Transactions of the American Neurological Association 101:208–210. [PubMed] [Google Scholar]
  19. Flores JA, El Banoua F, Galan-Rodriguez B, Fernandez-Espejo E (2004) Opiate antinociception is attenuated following lesion of large dopamine neurons of the periaqueductal grey: critical role for D1 (not D2) dopamine receptors. Pain 110:205–214. [DOI] [PubMed] [Google Scholar]
  20. Forrest WH Jr., Brown BW Jr., Brown CR, Defalque R, Gold M, Gordon HE, James KE, Katz J, Mahler DL, Schroff P, Teutsch G (1977) Dextroamphetamine with morphine for the treatment of postoperative pain. N Engl J Med 296:712–715. [DOI] [PubMed] [Google Scholar]
  21. Franklin KB (1989) Analgesia and the neural substrate of reward. Neurosci Biobehav Rev 13:149–154. [DOI] [PubMed] [Google Scholar]
  22. Franklin KB (1998) Analgesia and abuse potential: an accidental association or a common substrate? Pharmacol Biochem Behav 59:993–1002. [DOI] [PubMed] [Google Scholar]
  23. Gilbert AK, Franklin KB (2001) GABAergic modulation of descending inhibitory systems from the rostral ventromedial medulla (RVM). Dose-response analysis of nociception and neurological deficits. Pain 90:25–36. [DOI] [PubMed] [Google Scholar]
  24. Goetzl FR, Burrill DY, Ivy AC (1944) The analgesic effect of morphine alone and in combination with dextroamphetamine. Proceeding of the Society of Experimental Biology and Medicine 55:248–250. [Google Scholar]
  25. Ha AD, Jankovic J (2012) Pain in Parkinson’s disease. Mov Disord 27:485–491. [DOI] [PubMed] [Google Scholar]
  26. Hargreaves K, Dubner R, Brown F, Flores C, Joris J (1988) A new and sensitive method for measuring thermal nociception in cutaneous hyperalgesia. Pain 32:77–88. [DOI] [PubMed] [Google Scholar]
  27. Hasue RH, Shammah-Lagnado SJ (2002) Origin of the dopaminergic innervation of the central extended amygdala and accumbens shell: a combined retrograde tracing and immunohistochemical study in the rat. J Comp Neurol 454:15–33. [DOI] [PubMed] [Google Scholar]
  28. Heal DJ, Smith SL, Gosden J, Nutt DJ (2013) Amphetamine, past and present--a pharmacological and clinical perspective. J Psychopharmacol 27:479–496. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Heinricher MM, McGaraughty S (1998) Analysis of excitatory amino acid transmission within the rostral ventromedial medulla: implications for circuitry. Pain 75:247–255. [DOI] [PubMed] [Google Scholar]
  30. Hohmann AG, Suplita RL, Bolton NM, Neely MH, Fegley D, Mangieri R, Krey JF, Walker JM, Holmes PV, Crystal JD, Duranti A, Tontini A, Mor M, Tarzia G, Piomelli D (2005) An endocannabinoid mechanism for stress-induced analgesia. Nature 435:1108–1112. [DOI] [PubMed] [Google Scholar]
  31. Ivy AC, Goetzl FR, Burril DY (1944) Morphine-dextroamphetamine analgesia. War Med 6:67–71. [Google Scholar]
  32. Kernbaum S, Hauchecorne J (1981) Administration of levodopa for relief of herpes zoster pain. JAMA 246:132–134. [PubMed] [Google Scholar]
  33. Lane DA, Patel PA, Morgan MM (2005) Evidence for an intrinsic mechanism of antinociceptive tolerance within the ventrolateral periaqueductal gray of rats. Neuroscience 135:227–234. [DOI] [PubMed] [Google Scholar]
  34. Lau BK, Vaughan CW (2014) Descending modulation of pain: the GABA disinhibition hypothesis of analgesia. Current opinion in neurobiology 29:159–164. [DOI] [PubMed] [Google Scholar]
  35. Li C, Liu S, Lu X, Tao F (2019) Role of Descending Dopaminergic Pathways in Pain Modulation. Curr Neuropharmacol 17:1176–1182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Li C, Sugam JA, Lowery-Gionta EG, McElligott ZA, McCall NM, Lopez AJ, McKlveen JM, Pleil KE, Kash TL (2016) Mu Opioid Receptor Modulation of Dopamine Neurons in the Periaqueductal Gray/Dorsal Raphe: A Role in Regulation of Pain. Neuropsychopharmacology 41:2122–2132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Manning BH, Morgan MJ, Franklin KB (1994) Morphine analgesia in the formalin test: evidence for forebrain and midbrain sites of action. Neuroscience 63:289–294. [DOI] [PubMed] [Google Scholar]
  38. Martenson ME, Cetas JS, Heinricher MM (2009) A possible neural basis for stress-induced hyperalgesia. Pain 142:236–244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Mason P (2012) Medullary circuits for nociceptive modulation. Current opinion in neurobiology 22:640–645. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Meyer PJ, Morgan MM, Kozell LB, Ingram SL (2009) Contribution of dopamine receptors to periaqueductal gray-mediated antinociception. Psychopharmacology (Berl) 204:531–540. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Millan MJ (2002) Descending control of pain. Prog Neurobiol 66:355–474. [DOI] [PubMed] [Google Scholar]
  42. Minami M, Ide S (2015) How does pain induce negative emotion? Role of the bed nucleus of the stria terminalis in pain-induced place aversion. Curr Mol Med 15:184–190. [DOI] [PubMed] [Google Scholar]
  43. Morano TJ, Bailey NJ, Cahill CM, Dumont EC (2008) Nuclei-and condition-specific responses to pain in the bed nucleus of the stria terminalis. Prog Neuropsychopharmacol Biol Psychiatry 32:643–650. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Morgan MJ, Franklin KB (1990) 6-Hydroxydopamine lesions of the ventral tegmentum abolish D-amphetamine and morphine analgesia in the formalin test but not in the tail flick test. Brain Res 519:144–149. [DOI] [PubMed] [Google Scholar]
  45. Morgan MM, Fields HL (1994) Pronounced changes in the activity of nociceptive modulatory neurons in the rostral ventromedial medulla in response to prolonged thermal noxious stimuli. Journal of neurophysiology 72:1161–1170. [DOI] [PubMed] [Google Scholar]
  46. Morgan MM, Whittier KL, Hegarty DM, Aicher SA (2008) Periaqueductal gray neurons project to spinally projecting GABAergic neurons in the rostral ventromedial medulla. Pain 140:376–386. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Nixon DW (1975) Letter: Use of L-dopa to relieve pain from bone metastases. N Engl J Med 292:647. [PubMed] [Google Scholar]
  48. Oh SW et al. (2014) A mesoscale connectome of the mouse brain. Nature 508:207–214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Pert A, Yaksh T (1975) Localization of the antinociceptive action of morphine in primate brain. Pharmacol Biochem Behav 3:133–138. [DOI] [PubMed] [Google Scholar]
  50. Reichling DB, Basbaum AI (1990) Contribution of brainstem GABAergic circuitry to descending antinociceptive controls: II. Electron microscopic immunocytochemical evidence of GABAergic control over the projection from the periaqueductal gray to the nucleus raphe magnus in the rat. J Comp Neurol 302:378–393. [DOI] [PubMed] [Google Scholar]
  51. Reynolds DV (1969) Surgery in the rat during electrical analgesia induced by focal brain stimulation. Science 164:444–445. [DOI] [PubMed] [Google Scholar]
  52. Sandkuhler J, Gebhart GF (1984) Relative contributions of the nucleus raphe magnus and adjacent medullary reticular formation to the inhibition by stimulation in the periaqueductal gray of a spinal nociceptive reflex in the pentobarbital-anesthetized rat. Brain Res 305:77–87. [DOI] [PubMed] [Google Scholar]
  53. Schifirnet E, Bowen SE, Borszcz GS (2014) Separating analgesia from reward within the ventral tegmental area. Neuroscience 263:72–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Shi WX, Pun CL, Zhang XX, Jones MD, Bunney BS (2000) Dual effects of D-amphetamine on dopamine neurons mediated by dopamine and nondopamine receptors. J Neurosci 20:3504–3511. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Skinner GO, Damasceno F, Gomes A, de Almeida OM (2011) Increased pain perception and attenuated opioid antinociception in paradoxical sleep-deprived rats are associated with reduced tyrosine hydroxylase staining in the periaqueductal gray matter and are reversed by L-dopa. Pharmacol Biochem Behav 99:94–99. [DOI] [PubMed] [Google Scholar]
  56. Skogar O, Lokk J (2016) Pain management in patients with Parkinson’s disease: challenges and solutions. J Multidiscip Healthc 9:469–479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Taylor NE, Long H, Pei J, Kukutla P, Phero A, Hadaegh F, Abdelnabi A, Solt K, Brenner GJ (2019a) The rostromedial tegmental nucleus: a key modulator of pain and opioid analgesia. Pain 160:2524–2534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Taylor NE, Pei J, Zhang J, Vlasov KY, Davis T, Taylor E, Weng FJ, Van Dort CJ, Solt K, Brown EN (2019b) The Role of Glutamatergic and Dopaminergic Neurons in the Periaqueductal Gray/Dorsal Raphe: Separating Analgesia and Anxiety. eNeuro 6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Thompson JM, Neugebauer V (2019) Cortico-limbic pain mechanisms. Neurosci Lett 702:15–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Tobaldini G, Sardi NF, Guilhen VA, Fischer L (2019) Pain Inhibits Pain: an Ascending-Descending Pain Modulation Pathway Linking Mesolimbic and Classical Descending Mechanisms. Mol Neurobiol 56:1000–1013. [DOI] [PubMed] [Google Scholar]
  61. Tobaldini G, Reis RA, Sardi NF, Lazzarim MK, Tomim DH, Lima MMS, Fischer L (2018) Dopaminergic mechanisms in periaqueductal gray-mediated antinociception. Behavioural pharmacology 29:225–233. [DOI] [PubMed] [Google Scholar]
  62. Tovote P, Esposito MS, Botta P, Chaudun F, Fadok JP, Markovic M, Wolff SB, Ramakrishnan C, Fenno L, Deisseroth K, Herry C, Arber S, Luthi A (2016) Midbrain circuits for defensive behaviour. Nature 534:206–212. [DOI] [PubMed] [Google Scholar]
  63. Tran L, Schulkin J, Greenwood-Van Meerveld B (2014) Importance of CRF receptor-mediated mechanisms of the bed nucleus of the stria terminalis in the processing of anxiety and pain. Neuropsychopharmacology 39:2633–2645. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Vaughan CW, Christie MJ (1997) Presynaptic inhibitory action of opioids on synaptic transmission in the rat periaqueductal grey in vitro. J Physiol 498 ( Pt 2):463–472. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Vaughan CW, Ingram SL, Connor MA, Christie MJ (1997) How opioids inhibit GABA-mediated neurotransmission. Nature 390:611–614. [DOI] [PubMed] [Google Scholar]
  66. Wakaizumi K, Kondo T, Hamada Y, Narita M, Kawabe R, Narita H, Watanabe M, Kato S, Senba E, Kobayashi K, Kuzumaki N, Yamanaka A, Morisaki H, Narita M (2016) Involvement of mesolimbic dopaminergic network in neuropathic pain relief by treadmill exercise: A study for specific neural control with Gi-DREADD in mice. Mol Pain 12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Watanabe M, Narita M, Hamada Y, Yamashita A, Tamura H, Ikegami D, Kondo T, Shinzato T, Shimizu T, Fukuchi Y, Muto A, Okano H, Yamanaka A, Tawfik VL, Kuzumaki N, Navratilova E, Porreca F, Narita M (2018) Activation of ventral tegmental area dopaminergic neurons reverses pathological allodynia resulting from nerve injury or bone cancer. Mol Pain 14:1744806918756406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Webb SS, Smith GM, Evans WO (1972) Toward the development of an orally-effective, potent, non-sedating analgesic. Fed Proc 31:528. [Google Scholar]
  69. Wiklund L, Behzadi G, Kalen P, Headley PM, Nicolopoulos LS, Parsons CG, West DC (1988) Autoradiographic and electrophysiological evidence for excitatory amino acid transmission in the periaqueductal gray projection to nucleus raphe magnus in the rat. Neurosci Lett 93:158–163. [DOI] [PubMed] [Google Scholar]
  70. Wulff AB, Tooley J, Marconi LJ, Creed MC (2019) Ventral pallidal modulation of aversion processing. Brain Res 1713:62–69. [DOI] [PubMed] [Google Scholar]
  71. Yetnikoff L, Lavezzi HN, Reichard RA, Zahm DS (2014) An update on the connections of the ventral mesencephalic dopaminergic complex. Neuroscience 282:23–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Yu W, Pati D, Pina MM, Schmidt KT, Boyt KM, Hunker AC, Zweifel LS, McElligott ZA, Kash TL (2021) Periaqueductal gray/dorsal raphe dopamine neurons contribute to sex differences in pain-related behaviors. Neuron. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Zhao Z, Kim SC, Zhao R, Wu Y, Zhang J, Liu H, Kim YW, Zhu X, Gu C, Lee CW, Lee BH, Jang EY, Ko HL, Yang CH (2015) The tegmental-accumbal dopaminergic system mediates the anxiolytic effect of acupuncture during ethanol withdrawal. Neurosci Lett 597:143–148. [DOI] [PubMed] [Google Scholar]

RESOURCES