A glutamatergic DRN-VTA pathway modulates neuropathic pain and comorbid anhedonia-like behavior in mice

Abstract

Chronic pain causes both physical suffering and comorbid mental symptoms such as anhedonia. However, the neural circuits and molecular mechanisms underlying these maladaptive behaviors remain elusive. Here using a mouse model, we report a pathway from vesicular glutamate transporter 3 neurons in the dorsal raphe nucleus to dopamine neurons in the ventral tegmental area (VGluT3^DRN→DA^VTA) wherein population-level activity in response to innocuous mechanical stimuli and sucrose consumption is inhibited by chronic neuropathic pain. Mechanistically, neuropathic pain dampens VGluT3^DRN → DA^VTA glutamatergic transmission and DA^VTA neural excitability. VGluT3^DRN → DA^VTA activation alleviates neuropathic pain and comorbid anhedonia-like behavior (CAB) by releasing glutamate, which subsequently promotes DA release in the nucleus accumbens medial shell (NAcMed) and produces analgesic and anti-anhedonia effects via D2 and D1 receptors, respectively. In addition, VGluT3^DRN → DA^VTA inhibition produces pain-like reflexive hypersensitivity and anhedonia-like behavior in intact mice. These findings reveal a crucial role for VGluT3^DRN → DA^VTA → D2/D1^NAcMed pathway in establishing and modulating chronic pain and CAB.

Fig. 1. Dampened activity of VGluT3^DRN → DA^VTA circuit in mice with chronic pain.

a, b, f, g, k Schematic of the experimental design. c, h Representative images and percentage of GCaMP6m-expressing neurons that expressed TH (c) or VGluT3 (h), n = 9 sections. Scale bars, 100 μm. d, e, i, j Averaged Ca²⁺ responses, heatmaps, and area under the curve (AUC) evoked by von Frey stimulation (d: P = 0.0004; i: P = 0.096) and sucrose licking (e: P = 0.0264; j: P = 0.072). n = 6 mice. l–n Representative traces (l) and statistics of amplitude (m: Sham vs SNI 2W P = 0.0305; Sham vs SNI 6W P = 0.0429) and PPR (n: Sham vs SNI 2W P = 0.0389; Sham vs SNI 6W P = 0.0210). o–r Representative traces (o) and statistics of AMPA/NMDA ratio (p: Sham vs SNI 2W P = 0.0005; Sham vs SNI 6W P = 0.0387), AMPA-EPSCs amplitude (q: Sham vs SNI 2W P = 0.0105; Sham vs SNI 6W P = 0.0017), and NMDA-EPSCs amplitude (r: Sham vs SNI 2W P = 0.7065; Sham vs SNI 6W P = 0.3992). s–v Statistics of firing rate (s: Sham vs SNI 2W P = 0.0167; Sham vs SNI 6W P = 0.0173), rheobase (t: Sham vs SNI 2W P = 0.3780; Sham vs SNI 6W P = 0.0071), input resistance (u: Sham vs SNI 2W P = 0.0042; Sham vs SNI 6W P = 0.0439), and I_h (v: Sham vs SNI 2W P > 0.9999; Sham vs SNI 6W P = 0.0404; SNI 2W vs SNI 6W P = 0.0103). The groups in panels (c, h, l–v) were all from 3 mice. Significance was assessed by two-tailed paired Student’s t-test in (d, e, i, j), one-way ANOVA followed by Bonferroni’s multiple comparisons test in (m, n, p–r, t–v), and two-way ANOVA followed by Bonferroni’s multiple comparisons test in (s). All data are presented as the mean ± s.e.m except for (c, h) shown as box and whisker plots (medians, quartiles (boxes) and ranges minimum to maximum (whiskers)). *P < 0.05, **P < 0.01, ***P < 0.001, not significant (ns). Detailed statistics are presented in Supplementary Data 1. Created with BioRender.com (b, d, e, g, i, j, k).

Fig. 2. Glutamate is essential for the analgesic and anti-anhedonia effects conferred by VGluT3^DRN→VTA neural excitation.

a, g Schematic of the experimental design. b Schematic diagram of viral injection and optical fiber implantation (left). Representative images (middle) and summary data (right) for the percentage of mCherry-expressing neurons co-localized with VGluT3 immunofluorescence, n = 9 sections from 3 mice. Scale bar, 50 μm. c, d Mechanical paw withdrawal threshold (c: ChR2&ON vs ChR2&OFF P < 0.0001; ChR2&ON vs mCherry&ON P < 0.0001) and thermal paw withdrawal latency (d: ChR2&ON vs ChR2&OFF P = 0.0005; ChR2&ON vs mCherry&ON P = 0.0153) with (on) or without (off) optogenetic stimulation. Sham&ChR2, n = 8; Sham&mCherry, n = 9; SNI&ChR2, n = 9; SNI&mCherry, n = 6. e Experimental design of conditional place-preference (CPP) test (left) and quantification of CPP training (right). Sham&ChR2, n = 6; Sham&mCherry, n = 6; SNI&ChR2, n = 10; SNI&mCherry, n = 5. Sham&ChR2 P = 0.0470; SNI&ChR2 P = 0.0118. f Preference for sucrose in the SPT. Sham&ChR2, n = 6; Sham&mCherry, n = 6; SNI&ChR2, n = 7; SNI&mCherry, n = 5. P = 0.0094. h Schematic diagram of viral injection and drug delivery cannula implantation (left), and a representative trace showing depolarization of the hM3Dq-expressing neuron by CNO (right). i, j Effects of chemogenetic activation of VGluT3^DRN neurons on the von Frey test in post-SNI 2W mice (i: P < 0.0001) and SPT in post-SNI 6W mice (j: ACSF&hM3Dq vs ACSF&mCherry P = 0.0035; Ketanserin&hM3Dq vs Ketanserin&mCherry P = 0.0473). hM3Dq, n = 7; mCherry, n = 6. Significance was assessed by two-way ANOVA followed by Bonferroni’s multiple comparisons test in (c, d, f, i, j) and two-tailed paired Student’s t-test in (e). All data are presented as the mean ± s.e.m except for (b) shown as box and whisker plots (medians, quartiles (boxes) and ranges minimum to maximum (whiskers)). *P < 0.05, **P < 0.01, ***P < 0.001, not significant (ns). Details of the statistical analyses are presented in Supplementary Data 1. Created with BioRender.com (b, h).

Fig. 3. Inactivation of VGluT3^DRN→VTA terminals is sufficient to induce reflexive hypersensitivity and comorbid anhedonia.

a, g, l Schematic of the experimental design. b Schematic diagram of viral injection and optical fiber implantation (left), and a representative trace of action potential firing of an eNpHR-expressing neuron during light photostimulation (right). c Representative images (left) and summary data (right) for the percentage of EYFP-expressing neurons co-localized with VGluT3 immunofluorescence in the DRN, n = 9 sections from 3 mice. Scale bar, 50 μm. d, e Mechanical paw withdrawal threshold (d: eNpHR, n = 9; EYFP, n = 8. P < 0.0001) and thermal paw withdrawal latency (e: eNpHR, n = 11; EYFP, n = 8. P < 0.0001) with (on) or without (off) optogenetic stimulation. f Experimental design of conditional place aversion (CPA) test (left) and quantification of CPA training (right). eNpHR, n = 8; EYFP, n = 8. P = 0.0003. h Schematic diagram of viral injection and optical fiber implantation (left) and a representative trace showing hyperpolarization of the hM4Di-expressing neuron by CNO (right). i Experimental design of CPA test (left) and quantification of training (right). hM4Di, n = 5; mCherry, n = 5. P = 0.0112. j Time-course of mechanical paw withdrawal threshold changes. P = 0.0012. k Preference for sucrose in the SPT, P = 0.0442. m–o Light-evoked EPSCs (m: P = 0.0493), firing rate (n: P = 0.0332) and I_h (o: P = 0.0002) recorded from VGluT3^DRN-targeted postsynaptic DA^VTA neurons. hM4Di, n = 16 cells from 3 mice; mCherry, n = 18 cells from 3 mice. Significance was assessed by two-way ANOVA followed by Bonferroni’s multiple comparisons test in (d, e, j, n), two-tailed paired Student’s t-test in (f, i), two-tailed unpaired Student’s t-test in (k), and two-tailed Mann–Whitney U test in (m, o). All data are presented as the mean ± s.e.m except for (c), shown as box and whisker plots (medians, quartiles (boxes) and ranges from minimum to maximum (whiskers)). *P < 0.05, **P < 0.01, ***P < 0.001, not significant (ns). Details of the statistical analyses are presented in Supplementary Data 1. Created with BioRender.com (b, h, l).

Fig. 4. The VGluT3^DRN→DA^VTA circuit for neuropathic pain-induced inhibition of DA release outputs to the medial NAc.

a, i Schematic of the experimental design (top) and schematic diagram of viral injection, VTA laser stimulation, and NAc fiber photometry of DA sensor in VGluT3-Cre mice (bottom). b Representative images showing optical fiber track for fiber photometry recordings in the NAcMed (left) and NAcLat (right). Scale bars, 500 μm. c–h Averaged responses (left), heatmaps (middle), and AUC during 0–5 s (right) showing DA2m signals evoked by optogenetic activation of VGluT3^DRN→VTA terminals (c: P = 0.0027, d: P = 0.7926), 0.4 g von Frey stimulation (e: P = 0.0001, f: P = 0.3838) in pre- and post-SNI 2W mice and sucrose licking in pre- and post-SNI 6W mice compared with pre-SNI mice (g: P = 0.0202, h: P = 0.9069). NAcMed group, n = 6; NAcLat group, n = 5. j Schematic diagram of DRN viral injection, VTA laser stimulation, and NAc drug delivery cannula implantation. k–n Effects of optogenetic activation of VGluT3^DRN→VTA terminals on the von Frey test and SPT with drug infusion into the NAcMed (k: ACSF&ON vs Eticlopride&ON P < 0.0001; Eticlopride&ON vs Eticlopride&OFF P > 0.9999, l: ACSF vs SCH23390 P = 0.037, ACSF vs Eticlopride P = 0.8775. n = 5 mice) and NAcLat (m: ACSF&ON vs Eticlopride&ON P > 0.9999; Eticlopride&ON vs Eticlopride&OFF P < 0.0001, n: ACSF vs SCH23390 P = 0.9983, ACSF vs Eticlopride P = 0.8684. n = 6 mice). Significance was assessed by two-tailed paired Student’s t-test in (c–h), two-way ANOVA followed by Bonferroni’s multiple comparisons test in (k, m), and one-way ANOVA followed by Bonferroni’s multiple comparisons test in (l, n). All data are presented as the mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001, not significant (ns). Details of the statistical analyses are presented in Supplementary Data 1. Created with BioRender.com (a, e–h, j).