Inputs to the locus coeruleus from the periaqueductal gray and rostroventral medulla shape opioid-mediated descending pain modulation

Abstract

The supraspinal descending pain modulatory system (DPMS) shapes pain perception via monoaminergic modulation of sensory information in the spinal cord. However, the role and synaptic mechanisms of descending noradrenergic signaling remain unclear. Here, we establish that noradrenergic neurons of the locus coeruleus (LC) are essential for supraspinal opioid antinociception. Unexpectedly, given prior emphasis on descending serotonergic pathways, we find that opioid antinociception is primarily driven by excitatory output from the ventrolateral periaqueductal gray (vlPAG) to the LC. Furthermore, we identify a previously unknown opioid-sensitive inhibitory input from the rostroventromedial medulla (RVM), the suppression of which disinhibits LC neurons to drive spinal noradrenergic antinociception. We also report the presence of prominent bifurcating outputs from the vlPAG to the LC and the RVM. Our findings significantly revise current models of the DPMS and establish a novel supraspinal antinociceptive pathway that may contribute to multiple forms of descending pain modulation.

Figure 1.. Intrathecal noradrenergic and opioidergic antagonists attenuate systemic morphine antinociception.

A. Schematic of intrathecal and systemic injection combinations and morphine doses. B. Hot plate (52°C) withdrawal latencies resulting from increasing doses (0, 5, 12, 20 mg/kg) of subcutaneous morphine grouped by intrathecal antagonist administered (saline vs. 5 μg phentolamine vs. 5 μg naltrexone; ordinary two-way ANOVA with post hoc Tukey’s multiple comparisons test; intrathecal drug effect, p<0.0001, F(2,81)=95.35; morphine dose effect, p<0.0001, F(3,81)=56.61; interaction, p<0.0001, F(6,81)=16.75). C. Same data as in B from intrathecal saline and phentolamine groups organized by systemic morphine dose to facilitate comparisons at each dose (ordinary two-way ANOVA with post hoc Sidak’s multiple comparisons test; intrathecal drug effect, p<0.0001, F(1,52)=63.07; morphine dose effect, p<0.0001, F(3,52)=50.55; interaction, p=0.0002, F(3,52)=8.037). D. Schematic of intrathecal coadministration of morphine (2.5 μg) or saline with saline, phentolamine (5 μg) or naltrexone (5 μg). E. Hot plate withdrawal latencies before (white) and after (blue: saline, red: phentolamine, green: naltrexone) intrathecal injection of morphine and antagonist (pre- vs. post-intrathecal injection, n=9 pairs each, two-sided Wilcoxon matched-pairs signed rank test; post-intrathecal morphine + saline vs. post-intrathecal morphine + phentolamine, n=9 saline, n=9 phentolamine, two-sided Mann-Whitney test). F. Hot plate withdrawal latencies before (white) and after (light blue: saline, pink: phentolamine, light green: naltrexone) intrathecal injection of saline and antagonist (pre- vs. post-intrathecal injection, n=6 pairs each, two-sided Wilcoxon matched-pairs signed rank test). All bar graphs depict mean ± SEM and include individual experimental replicate values.

Figure 2.. LC activity is required for systemic morphine antinociception.

A. Subcutaneous saline, 10 mg/kg morphine, and 10 mg/kg morphine + 10 mg/kg naloxone injections and representative images of resulting c-Fos (green) immunohistochemistry in DBH-cre::tdTom mice. Scale bars 150 μm. B. Percentage of DBH-positive LC neurons that colocalize with green c-Fos signal (5–8 LC images analyzed per mouse; n=6 saline, n=6 morphine, n=4 morphine + naloxone; saline = 29.9 ± 5.1%, morphine = 71.0 ± 4.5%, morphine + naloxone = 47.9 ± 4.2%; Ordinary One-way ANOVA with post hoc Tukey’s multiple comparisons test, p<0.0001, F(2,13) = 20.97)). C. Left: representative traces of CNV-Y-DAMGO light-evoked uncaging (purple arrows: 50 ms, 365 nm LED, 84 mW) during current clamp (top) and voltage clamp (bottom) recordings in whole cell patch clamp from LC neurons in wild type mice. Right: LC neuron opioid response characterized by change in membrane potential after light flash, pause in firing calculated as latency to first spike after light flash, tonic firing rate during the 10 seconds before and 10 seconds after light flash (p<0.0001, two-sided Wilcoxon matched-pairs signed rank test), and evoked outward current amplitude (n=16 cells each). D. f-I curves recorded from LC neurons in wild type mice before and after bath application of 1 μM DAMGO (n=7 cells; two-sided Wilcoxon matched-pairs signed rank test at 50 pA, 100 pA, 150 pA, and 200 pA). E. Injection schematic, representative images and quantification of neuronal ablation of mice after bilateral injection of AAV-DIO-Casp3 in LC (n=11 DBHcre:TdTom, n=15 TdTom control; DBH-cre = 79 ± 15 cells , control = 2376 ± 137 cells; two-sided Mann-Whitney test). F. Left, hot plate paw withdrawal latencies of control and DBH-cre mice following subcutaneous administration of 0, 5, and 10 mg/kg morphine (Two-way repeated measures ANOVA with post hoc Sidak’s multiple comparisons test; n=11 DBHcre:TdTom n=15 TdTom control; morphine dose effect, p<0.0001, F(1.365,32.75)= 78.80, genotype effect, p=0.0006, F(1,24)=15.67, morphine dose x genotype interaction, p=0.0004, F(2,48)=9.328). Right, Response to 10 hind paw pin pricks following morphine administration (Two-way repeated measures ANOVA with post hoc Sidak’s multiple comparisons test; n=11 DBHcre:TdTom n=15 TdTom control; morphine dose effect, p<0.0001, F(1.690,40.56)=72.20, genotype effect, p=0.0006, F(1,24)=15.52, morphine dose x genotype interaction, p=0.0037, F(2,48)=6.294). G. Top, Slice electrophysiology recordings from Kir2.1-positive and -negative LC neurons. Bottom, representative traces of tonic AP firing in control vs. Kir2.1-expressing neurons. H. Left, resting membrane potential of control vs. Kir2.1-expressing neurons (n=10 Kir2.1-positive, n=9 control; Kir2.1 RMP = −70.1 ± 3.0 mV, control RMP = −46.9 ± 2.4mV, two-sided Mann-Whitney test). Right, tonic AP firing rate of control vs. Kir2.1-expressing neurons (Kir2.1 tonic firing rate = 0.006 ± 0.006 Hz, control tonic firing rate = 1.61 ± 0.25 Hz, two-sided Mann-Whitney test). I. Left, f-I curves of control vs. Kir2.1-expressing neurons for 7 current steps (n=10 Kir2.1-positive, n=9 control; two-sided Mann-Whitney test). Right, representative traces of evoked AP firing from a control (grey) and a Kir2.1-expressing (green) neuron at 50 pA (top) and 200 pA (bottom) current steps. J. Injection schematic, representative image of zsGreen expression in the LC of a DBH-cre::tdTom mouse (scale bar = 150 μm) and quantification of viral coverage given as % tdTom LC neurons labeled by zsGreen (n=12 DBHcre:TdTom, n=13 TdTom control; control = 1.2 ± 0.3%, DBH-cre = 79.0 ± 2.0%; two-sided Mann-Whitney test). K. Left, hot plate paw withdrawal latencies of control and DBH-cre mice following 0, 5, and 10 mg/kg morphine, s.c. (Two-way repeated measures ANOVA with post hoc Sidak’s multiple comparisons test; n= 12 DBHcre:TdTom, n=13 TdTom control; hot plate: morphine dose effect, p<0.0001, F(1.092,25.11)=46.27, genotype effect, p=0.001, F(1,23)=14.30, morphine dose x genotype interaction, p=0.013, F(2,46)=4.790). Right, Response to 10 hind paw pin pricks following morphine (Two-way repeated measures ANOVA with post hoc Sidak’s multiple comparisons test; n= 12 DBHcre:TdTom, n=13 TdTom control; morphine dose effect, p<0.0001, F(1.695,38.99)=58.23, genotype effect, p=0.0004, F(1,23)=17.20, morphine dose x genotype interaction, p=0.0044, F(2,46)=6.131). Data in each graph reported as mean ± SEM.

Figure 3.. vlPAG activity is required for systemic morphine antinociception and drives spinal NA-dependent antinociception.

A. Bilateral AAV-DIO-HM4Di-mCherry injections in vlPAG of VGLUT2-cre mice with representative image of viral expression. Scale bar 500 μm. B. Left, hot plate withdrawal latencies of vlPAG^VGLUT2-cre::HM4D mice administered 3 mg/kg CNO, i.p. vs. saline without (light green bars) and with 5 mg/kg morphine, s.c. (dark green bars) (Two-way repeated measures ANOVA with post hoc Tukey’s multiple comparisons test; n=13 mice; sal vs. CNO effect, p<0.0001, F(1,12)=56.93, morphine effect, p<0.0001, F(1,12)=43.25, morphine x CNO interaction, p=0.048, F(1,12)=4.837). Right, hot plate withdrawal latencies of non-virus injected C57 controls (Two-way repeated measures ANOVA with post hoc Tukey’s multiple comparisons test; n=9 mice; sal vs. CNO effect, p=0.32, F(1,8)=1.143, morphine effect, p<0.0001, F(1,8)=75.02). C. Bilateral AAV-DIO-HM3Dq-mCherry injections in vlPAG of VGLUT2-cre mice and combinations of systemic CNO (3 mg/kg, i.p.) with intrathecal antagonists and representative image of viral expression. Scale bar 500 μm. D. Left, hot plate withdrawal latencies after systemic saline or CNO and intrathecal saline, phentolamine (5 μg), or naltrexone (5 μg) (Friedman test with post hoc Dunn’s multiple comparisons test, n=12 subjects, p<0.0001, Friedman statistic = 26.40). Right, von Frey mechanical thresholds (Friedman test with post hoc Dunn’s multiple comparisons test, n=12 subjects, p=0.0001, Friedman statistic = 21.00). E. Representative images of c-Fos immunohistochemistry in TH-positive LC neurons of vlPAG^VGLUT2-cre::HM3D mice after systemic injection of saline (top) or CNO (bottom). Right, % of TH-positive LC neurons that colocalize with green c-Fos signal (5–8 images analyzed per mouse; n= 6 saline, n=5 CNO; saline = 4.8 ± 1.1%, CNO = 79.6 ± 3.6%, two-sided Mann-Whitney test). Data in each graph reported as mean ± SEM.

Figure 4.. Anatomical characterization of inputs to LC from vlPAG and RVM.

A. Injection of AAV-DIO-tdTom in vlPAG of VGLUT2- or VGAT-cre mice. B. Injection site in VGLUT2-cre mouse. C. vlPAG glutamatergic terminals (red) in pericoerulear region defined by TH immunohistochemistry (green). Delineations taken from the Paxinos & Franklin Brain Atlas show that the pericoerulear terminals are largely targeted to the medial Barrington’s nucleus. D. Injection site in VGAT-cre mouse. E. vlPAG GABAergic/glycinergic terminals in pericoerulear region. B-E. Scale bars 300 μm. F. Quantification of red (vlPAG glutamatergic terminals) and green (LC TH immunohistochemistry) pixel intensity in the LC and pericoerulear region normalized by z-score to account for fluorescence differences between animals and during imaging (n=6 LC slices from n=3 mice each). G. Same as F for vlPAG GABAergic/glycinergic terminals. H. Left: injections of AAVretro-cre in LC and AAV-DIO-mCherry in vlPAG of wild type mice to capture LC-projecting vlPAG neurons. Center: image of mCherry+ vlPAG neurons captured. Right: resulting terminals captured in LC and RVM. Scale bars 500 μm. I. Same as H for vlPAG neurons that project to RVM. Scale bars 500 μm. J. Left: Orthogonal recombinase strategy to label vlPAG neurons that project to RVM and LC. Center: representative image of mCherry (red) and YFP (green) viral labeling in vlPAG with neurons co-expressing both fluorophores appearing yellow. White arrows indicate examples of double labeled neurons. Scale bar 300 μm. Right: quantification of mCherry and YFP labeling in vlPAG. K. Injection of AAV-DIO-tdTom in RVM of VGLUT2- or VGAT-cre mice. L. Injection site in VGLUT2-cre mouse. M. RVM glutamatergic terminals (red) in LC and pericoerulear region defined by TH immunohistochemistry (green). N. Injection site in VGAT-cre mouse. O. RVM GABAergic/glycinergic terminals in LC and pericoerulear region. L-O. Scale bars 300 μm. P. Quantification of RVM glutamatergic terminals by pixel intensity z-score similar to F (n=6 LC slices from n=3 mice each). Q. Same as P for RVM GABAergic/glycinergic terminals. R. Quantification of red pixel intensity normalized by z-score across the dorsal to ventral axis of LC for all four projection origin and cell type combinations (n=6 LC slices from n=3 mice each). S. Injections of AAVretro-cre in LC and AAV-DIO-tdTom in RVM of wild type mice to capture LC-projecting RVM neurons. T. tdTom+ RVM neurons captured. U. Resulting terminals (red) captured in LC (green) and pericoerulear region. Scale bar 300 μm. V. tdTom+ fibers located in bilateral parafascicular nucleus of the thalamus. Scale bar 1mm.

Figure 5.. Electrophysiological characterization of inputs from vlPAG and RVM to LC.

A. Top, viral injection of ChR2 into vlPAG for LC slice electrophysiology. Bottom, representative trace of tonic spiking before, during, and after a 2 second blue LED stimulus (470 nm, 50×2 ms pulses at 25 Hz, 18 mW). B. Left, firing rate before and during the light stimulus for 20 LC neurons (baseline = 2.4 ± 0.4 Hz, light on = 9.1 ± 1.0 Hz; two-sided Wilcoxon matched-pairs signed rank test). Right, recorded neurons were categorized as “excited” (a z-score of spiking during vs. before light >2), “inhibited” (a z-score of spiking during vs. before light < −2), or “not modulated.” C. Top, viral injection of ChR2 into RVM for LC slice electrophysiology. Bottom, representative trace of tonic spiking before, during, and after the blue LED stimulus. D. Left, firing rate before and during the blue light stimulus for 20 LC neurons (baseline = 2.3 ± 0.2 Hz, light on = 1.3 ± 0.3 Hz; two-sided Wilcoxon matched-pairs signed rank test). Right, categorization of recorded LC neurons as “excited”, “inhibited”, or “not modulated.” E. Left, proportions of LC neurons in which optically-evoked EPSCs, EPSCs and IPSCs or no evoked currents were present when stimulating vlPAG terminals (470nm, 1 × 5 ms pulse, 18 mW). Middle, representative example of oEPSC and oIPSC recorded in a single LC neuron via electrical isolation. Right, peak amplitude of oEPSCs and oIPSCs driven by vlPAG terminal stimulation (oEPSCs = −177.8 ± 40.6 pA, oIPSCs = −0.52 ± 12.7 pA, n=12 neurons). F. Left, proportions of LC neurons in which optically-evoked EPSCs, IPSCs, both EPSCs and IPSCs or no evoked currents were present when stimulating RVM terminals. Middle, representative example of oEPSC and oIPSC recorded in a single LC neuron via electrical isolation. Right, peak amplitude of oEPSCs and oIPSCs driven by RVM terminal stimulation (oEPSCs = −117.7 ± 39.9 pA, oIPSCs = 466.6 ± 79.5 pA, n=12 neurons). G-K. Top, summary bar graphs of oEPSC/IPSC amplitude. Bottom, representative examples. G. vlPAG-driven oEPSC amplitude before and after bath application of NBQX (10 μM) (ACSF = −337.1 ± 88.0 pA, +NBQX = −4.9 ± 4.4 pA; two-sided paired t-test: t=3.950, n=4 pairs). H. vlPAG-driven oEPSC amplitude before and after bath application of TTX (1 μM) and subsequent application of 4-AP (100 μM) (+TTX = −2.9 ± 1.5 pA, +TTX & 4-AP = −288.5 ± 98.4 pA; two-sided paired t-test: t=2.926, n=5 pairs). I. RVM-driven oIPSC amplitude before and after bath application of GABAzine (20 μM) and additional application of strychnine (10 μM) (ACSF = 636.3 ± 154.9 pA, +GABAzine = 286.6 ± 100.8 pA, +GABAzine & strychnine = −2.5 ± 1.7 pA; Repeated Measures One-way ANOVA with Dunnett’s multiple comparisons test, p=0.013, F(1.087,4.348)=16.32, n=5 cells). J. RVM-driven oIPSC amplitude before and after bath application of TTX (1 μM) and subsequent application of 4-AP (100 μM) (+TTX = 13.1 ± 9.8 pA, +TTX & 4-AP = 443.9 ± 164.6 pA; two-sided Wilcoxon matched-pairs signed rank test, n=6 pairs). K. RVM-driven oIPSC amplitude before and after bath application of NBQX (10 μM) + CPP (10 μM) (ACSF = 309.6 ± 36.1 pA, +NBQX/CPP = 207.6 ± 35.8 pA; two-sided paired t-test: t=6.976, n=7 pairs). L. Top, example traces of RVM oIPSCs (blue) and vlPAG oEPSCs (red) before and after bath application of DAMGO (1 μM; black). Bottom, opioid sensitivity reported as % suppression of amplitude by DAMGO (RVM oIPSC = 47.6 ± 7.0%, PAG oEPSC = 11.4 ± 4.2%; two-sided unpaired t-test: t=3.785, n=8 RVM IPSC, n=5 PAG EPSC). All summary data reported as mean ± SEM.

Figure 6.. Pathway specific modulation of vlPAG and RVM terminals in LC modulates nociceptive behavior.

A. Left: cannula placement over bilateral LC of uninjected C57 control mice. Right: hot plate withdrawal latencies of control mice microinfused in LC with saline (150nl) vs. CNO (3 μM 150nl) without (light grey) and with 5 mg/kg morphine, s.c. (dark grey; Two-way repeated measures ANOVA with post hoc Tukey’s multiple comparisons test; n=8 mice; sal vs. CNO effect, p=0.4178, F(1,7)=0.7412, morphine effect, p=0.0001, F(1,7)=57.98; morphine x CNO interaction, p=0.58, F(1,7)=0.3434). B. Left: viral injection of AAV-DIO-HM4Di-mCherry in bilateral vlPAG of VGLUT2-cre mice with cannula placement over bilateral LC. Right: hot plate withdrawal latencies after microinfusion with saline vs. CNO without (light green) and with 5m/kg morphine, s.c. (dark green; Two-way repeated measures ANOVA with post hoc Sidak’s multiple comparisons test; n=12 mice; sal vs. CNO effect, p=0.0022, F(1,11)=15.72, morphine effect, p=0.0003, F(1,11)=27.60; morphine x CNO interaction, p=0.0106, F(1,11)=9.456). C. Left: viral injection of AAV-DIO-HM4Di-mCherry in bilateral RVM of VGAT-cre mice with cannula placement over bilateral LC. Right: hot plate withdrawal latencies after microinfusion of saline vs. CNO (blue bars n=12 mice), microinfusion of CNO with intrathecal injections of saline vs. phentolamine (5ug, light blue bars, n=9 mice), and microinfusion of saline vs. CNO with 5 mg/kg morphine, s.c. (dark blue, n=12 mice; Mixed effects analysis with matching across row and post hoc Tukey’s multiple comparisons test, p<0.0001, F(2.560,25.09)=27.36). D. Circuit diagram of DPMS inputs to LC and their opioid sensitivity. Data in each graph reported as mean ± SEM.