Synaptic mechanism underlying serotonin modulation of transition to cocaine addiction

Abstract

Compulsive drug use despite adverse consequences defines addiction. While mesolimbic dopamine signaling is sufficient to drive compulsion, psychostimulants such as cocaine also boost extracellular serotonin (5-HT) by inhibiting reuptake. We used SERT Met172 knockin (SertKI) mice carrying a transporter that no longer binds cocaine to abolish 5-HT transients during drug self-administration. SertKI mice showed an enhanced transition to compulsion. Conversely, pharmacologically elevating 5-HT reversed the inherently high rate of compulsion transition with optogenetic dopamine self-stimulation. The bidirectional effect on behavior is explained by presynaptic depression of orbitofrontal cortex–to–dorsal striatum synapses induced by 5-HT via 5-HT_1B receptors. Consequently, in projection-specific 5-HT_1B receptor knockout mice, the fraction of individuals compulsively self-administering cocaine was elevated.

Figure 1.. SertKI animals were more compulsive in cocaine self-administration.

(A) Schedule of saline and cocaine injections. (B) GRAB 5-HT sensor expression indicated with GFP staining in the DS. Scale bars, 1 mm. (C) 5-HT transients in the dorsal striatum induced by saline / cocaine (15 mg/kg) i.p. injection in WT and SertKI mice were detected with a GRAB 5-HT sensor (n = 3 mice for WT and SertKI group, data from saline injected WT and SertKI are pooled). (D) Timeline of cocaine self-administration. (E) Left, number of active (A) and inactive (IA) lever presses of WT (black) and SertKI (red) animals (n = 26 and 25 for WT and SertKI group) in acquisition sessions. Right, cocaine infusions obtained from WT (black) and SertKI (red) mice in acquisition sessions (two way ANOVA; F_1,588 = 1.996, P = 0.1583; n = 26 and 25 for WT and SertKI group). (F) Raster plots for infusions (blue lines) and punishments (red lines) in baseline and punishment sessions of a renouncer (upper, WT mouse) and a perseverer (lower, SertKI mouse). (G) Hierarchical clustering based on tSNE projection of different parameters of punishment sessions 3 and 4 (P3 and P4) of cocaine self-administration. Blue and green arrow heads indicate examples presented in F. Genotype, break point, baseline rate and perseverance were not used for clustering. (H) tSNE three dimensional representation of clusters of perseverers (cluster 1) and renouncers (cluster 2) in cocaine self- administration. (I) Percentage of perseverers and renouncers among WT and SertKI groups (Fisher’s exact test; P = 0.001). (J) Perseverance rate as a function of baseline rate (Pearson r = −0.08; P = 0.55). (K) The success rate of performance as a function of the last progressive ratio value achieved by the mice (logrank test; P = 0.94). Abbreviations, PR, progressive ratio; FR, fixed ratio; A, active lever presses; IA, inactive lever presses; R, renouncer; P, perseverer; sal, saline; coc, cocaine; WT, wildtype; SertKI, Sert-knockin.

Figure 2.. Citalopram decreased compulsive dopamine self-stimulation.

(A) Timeline of oDASS and saline / citalopram treatments. (B) Schematic of virus injection sites and optic fiber implantation sites. (C) ChR2-EYFP expression in the VTA from a sagittal slice (upper) and a coronal slice co-labeled with TH (lower). Scale bars, 1 mm. (D) Left, number of active (A) and inactive (IA) lever presses of saline (black) and citalopram (blue) treated mice (n = 25 and 26 for saline and citalopram group) in acquisition sessions. Right, laser stimulation per minute obtained from saline (black) and citalopram (blue) treated mice in acquisition sessions (two way ANOVA; F_1,588 = 0.73, P = 0.39; n = 25 and 26 for saline and citalopram group). (E) Raster plots for laser stimulations (blue lines) and punishments (red lines) in baseline and punishment sessions of a perseverer (upper, from saline group) and a renouncer (lower, from citalopram group). (F) Hierarchical clustering based on different parameters of punishment sessions 3 and 4 (P3 and P4) of oDASS. Red and green arrow heads indicate examples presented in E. Treatment, break point, baseline rate and perseverance were not used for clustering. (G) tSNE three dimensional representation of clusters of perseverers (cluster 1) and renouncers (cluster 2) in oDASS. (H) Percentage of perseverers and renouncers among saline and citalopram treated groups (Fisher’s exact test; P = 0.001). (I) Perseverance rate as a function of baseline rate (Pearson r = 0.05; P = 0.71). (J) The success rate of performance as a function of the last progressive ratio value achieved by the mice (logrank test; P = 0.95). Abbreviations, PR, progressive ratio; FR, fixed ratio; LS, laser stimulation; A, active lever presses; IA, inactive lever presses; R, renouncer; P, perseverer; sal, saline; cita, citalopram.

Figure 3.. The OFC-DS pathway was modulated by 5-HT.

(A) Upper, Schematic of virus injection and recording sites. Lower, Chrimson-Tdtomato expressing cell bodies in the OFC (left) and terminals in the DS (right). Scale bars, 1 mm. (B) Representative traces of AMPA and NMDA currents recorded at +40 mV of a renouncer (left) and a perseverer (right) from WT (upper) and SertKI (lower) group after cocaine self-administration. Scale bars, 200 pA, 15 ms. (C) Average A/N of WT and KI renouncers and perseverers after cocaine self-administration (two tailed t test; t₅₉ = 4.26, P < 0.0001; n = 53 and 8 cells from 10 and 2 mice for renouncer and perseverer in WT group; t₆₅ = 3.48, P = 0.0009; n = 37 and 30 cells from 7 and 6 mice for renouncer and perseverer in SertKI group). (D) Representative traces of AMPA and NMDA currents recorded at +40 mV of a renouncer (left) and a perseverer (right) from saline (upper) and citalopram (lower) treated group after oDASS. Scale bars, 200 pA, 15 ms. (E) Average A/N of saline and citalopram treated renouncers and perseverers after oDASS (two tailed t test; t₅₄ = 2.54, P = 0.01; n = 21 and 35 cells from 4 and 7 mice for renouncer and perseverer in saline treated group; t₅₅ = 2.38, P = 0.02; n = 34 and 23 cells from 7 and 4 mice for renouncer and perseverer in citalopram treated group). (F) Representative traces before and after bath application of 5-HT in the presence of ACSF (red), NAS181 (gray), and WAY100635 (blue). Scale bars, 200 pA, 10 ms. (G) Average traces of EPSC before, during and after bath application of 5-HT in the presence of ACSF (red), NAS181 (gray), and WAY100635 (blue) (n = 13, 14 and 7 cells from 3 mice for ACSF, NAS181, and WAY100635 group). (H) Normalized coefficient of variation (1/CV²) versus normalized EPSC (one way ANOVA; F_2,31 = 3.682, P = 0.0367 for Norm. 1/CV²; F_2,31 = 7.948, P = 0.0016 for Norm. EPSC; n = 13, 14 and 7 cells from 3 mice for ACSF, NAS181, and WAY100636 group). (I) Normalized pair pulse ratio (PPR) versus normalized EPSC (one way ANOVA; F_2,26 = 2.807, P = 0.08 for Norm. PPR; F_2,31 = 7.948, P = 0.0016 for Norm. EPSC; n = 13, 14 and 7 cells from 3 mice for ACSF, NAS181, and WAY100636 group). Abbreviations, SA, self-administration; R, renouncer; P, perseverer; sal, saline; cita, citalopram; WT, wildtype; KI, Sert-knockin.

Figure 4.. Knockout 5-HT_1B receptors promoted compulsive cocaine self-administration.

(A) Schematic of virus injections for cocaine self-administration. (B) Left and middle, GFP co-expressing with mCherry in the OFC. Right, mCherry seeding in the DS. Scale bars, 1 mm (left and right), and 20 μm (middle). (C) Schematic of virus injections for patch clamp recording. (D) Average traces of EPSC before, during and after bath application of 5-HT_1B receptor agonist CP39129 in control (CT, black) and pathway specific knockout 5-HT_1B receptor (KO, violet) groups (compared Norm. EPSC recorded on last 5 minutes; two tailed t test; t₂₅ = 2.86, P = 0.008; n = 15 and 12 cells from 3 mice for CT and KO group). Scale bars, 200 pA, 10 ms. (E) Timeline of virus injection and cocaine self-administration experiments. (F) Left, number of active (A) and inactive (IA) lever presses of CT (black) and KO (violet) mice in acquisition sessions. Right, cocaine infusions obtained by CT (black) and KO (violet) mice in acquisition sessions (two way ANOVA; F_1,288 = 2.53, P = 0.06; n = 10 and 11 for CT and KO group). (G) Hierarchical clustering based on different parameters of punishment sessions 3 and 4 (P3 and P4) of cocaine self-administration. Treatment, break point, baseline rate and perseverance were not used for clustering. (H) tSNE three dimensional representation of clusters of perseverers (cluster 1) and renouncers (cluster 2) in cocaine self-administration. (I) Percentage of perseverers and renouncers among CT and KO groups (Fisher’s exact test; P = 0.02). (J) Perseverance rate as a function of baseline rate (Pearson r = 0.36; P = 0.11). (K) The success rate of performance as a function of the last progressive ratio value achieved by the mice (logrank test; P = 0.6). Abbreviations, PR, progressive ratio; FR, fixed ratio; A, active lever presses; IA, inactive lever presses; R, renouncer; P, perseverer; CT, control; KO, pathway specific knockout.