Dual Dopaminergic Regulation of Corticostriatal Plasticity by Cholinergic Interneurons and Indirect Pathway Medium Spiny Neurons

Abstract

Endocannabinoid (eCB)-mediated long-term depression (LTD) requires dopamine (DA) D2 receptors (D2Rs) for eCB mobilization. The cellular locus of the D2Rs involved in LTD induction remains highly debated. We directly examined the role in LTD induction of D2Rs expressed by striatal cholinergic interneurons (Chls) and indirect pathway medium spiny neurons (iMSNs) using neuron-specific targeted deletion of D2Rs. Deletion of Chl-D2Rs (Chl-Drd2KO) impaired LTD induction in both subtypes of MSNs. LTD induction was restored in the Chl-Drd2KO mice by an M1-selective muscarinic acetylcholine receptor antagonist. In contrast, after the deletion of iMSN-D2Rs (iMSN-Drd2KO), LTD induction was intact in MSNs. Separate interrogation of direct pathway and iMSNs revealed a deficit in LTD induction only at synapses onto iMSNs that lack D2Rs. LTD induction in iMSNs was restored by D2R agonist application. Our findings suggest that Chl D2Rs strongly modulate LTD induction in MSNs, with iMSN-D2Rs having a weaker, iMSN-specific, modulatory effect.

Keywords: D2 receptors; cholinergic interneurons; dMSNs; iMSNs; long-term depression.

Figure 1.. Selective Deletion of D2Rs in iMSNs and Chls

(A) Schematic representation of Cre-mediated recombination depicting the removal of exon 2, which is flanked by loxP sites, of the Drd2 gene. (B and C) Diagram of breeding strategy used to generate (B) Chl-Drd2KO and (C) iMSN-Drd2KO mice. (D) Data reported as mean ± SEM. There was no significant change in D2R mRNA expression levels after D2R deletion from Chls (n = 8–10 mice). (E) iMSN D2R deletion resulted in an 90% reduction in D2R mRNA levels in the iMSN-Drd2KO mice (n = 3–4 mice).

Figure 2.. D2R-Mediated Pausing in Chls Is Lost in Chl-Drd2KO Mice

(A) Example cell-attached recordings from Chls from wild-type (ChAT Cre, left) or knockout (Chl-Drd2KO, right) mice. Top: 1 s of UV light (385 nm, violet line) had no effect on the tonic firing rate of Chls in ACSF. Middle: In the presence of 100 µM NPEC-caged dopamine, UV light caused a pause in the tonic firing of Chls in wildtype but not knockout animals. Bottom: This pause was blocked by 5 µM sulpiride. (B) Data reported as mean ± SEM. The dopamine-mediated pause, or latency to fire, following the light pulse in ChAT Cre mice was 0.97 ± 0.11 s, which was reduced to 0.49 ± 0.08 s in sulpiride (p < 0.001, paired t test). This pause was absent in Chl-Drd2KO animals (latency to fire baseline: 0.74 ± 0.08 s; in sulpiride: 0.7 ± 0.11; p = 0.75, paired t test). (C) Sulpiride increased the firing rate of Chls from ChAT Cre animals but not in Chl-Drd2KO cells. Data from individual cells are in gray, with averages overlaid in black for the ChAT Cre and the Chl-Drd2KO in blue. N = 6 ChAT Cre and 8 Chl-Drd2KO mice.

Figure 3.. Intact D2 Autoreceptor Function in iMSN and Chl-Drd2KO Mice

(A) Top: Representative DA voltammograms of Drd2^loxP/loxP (gray trace) and iMSN-Drd2KO (orange trace) mice. Bottom: Representative traces of evoked DA release in the DLS and corresponding color plots depicting the data with time on the x axis, applied scan potential (Eapp) on the y axis, and background subtracted faradaic current on the z axis in pseudocolor. (B) Representative traces of DA release under control conditions (no drug application) and after 25 min of 30 nM quinpirole followed by 10–15 mins of 300 nM quinpirole bath application in the Drd2^loxP/loxP and iMSN-Drd2KO mice (n = 4 slices/3 mice for each genotypes). (C) The effects of quinpirole inhibition were similar in both genotypes. (D) Representative DA traces from DLS and corresponding color plots with DA voltammograms (top) of Drd2^loxP/loxP (gray trace) and Chl-Drd2KO (blue trace) mice. (E) Representative traces (Drd2^loxP/loxP and Chl-Drd2KO) of DA transients of the following experimental conditions administered sequentially: (1) after 20 min bath application of DHβE, (2) after 25 mins of 30 nM quinpirole co-applied with DHβE, and (3) after 10–15 min 300 nM quinpirole applied in the presence of DHβE. (F) The effects of co-application of DHβE and quinpirole was similar in both genotypes (Drd2^loxP/loxP n = 6 slices/5 mice; Chl-Drd2KO n = 5 slices/3 mice). All data presented as mean ± SEM.

Figure 4.. Loss of LTD in Chl-Drd2KO Mice in Field Potential Recordings

PS amplitudes were normalized to baseline, averaged (mean ± SEM), and plotted as a function of time. Insets: Representative traces before (solid line) and after HFS (dashed line). Scale bars, 0.2 mV, 1 ms. Symbol representation: Drd2^loxP/loxP (gray circles), Chl-Drd2KO mice (blue circles), iMSN-Drd2KO mice (orange circles). (A) HFS failed to induce LTD in Chl-Drd2KO mice (n = 8 slices/5 mice). However, HFS induced LTD in controls, Drd2^loxP/loxP (n = 6 slices/5 mice). (B) LTD was blocked during the bath application of AM251 (n = 6 slices/4 mice). (C) Bath application of sulpiride prevented LTD induction in Drd2^loxP/loxP (n = 5 slices/3 mice). (D) HFS induced LTD in iMSN-Drd2KO (n = 9 slices/6 mice) and Drd2^loxP/loxP mice (n = 9 slices/5 mice). (E) HFS-LTD was blocked in the presence of AM251 in both genotypes (Drd2^loxP/loxP n = 5 slices/4 mice; iMSN-Drd2KO n = 5 slices/3 mice). (F) HFS-LTD was blocked in the presence of sulpiride (Drd2^loxP/loxP n = 7 slices/6 mice; iMSN-Drd2KO n = 6 slices/5 mice).

Figure 5.. LTD Induction at Synapses onto MSNs Is Dependent on Chl D2Rs

The average of the first evoked EPSC amplitudes plotted as a function of time. Insets: Representative traces showing first evoked EPSC before (1) and after (2) LTD induction protocol. Scale bars, 100 pA, 10 ms. Symbol representation: Drd2^loxP/loxP (gray circles), Chl-Drd2KO mice (blue circles). Scatterplot of individual cell responses (averaged EPSCs during last 10 min of recording) after HFS. Data reported as mean ± SEM. (A) HFS paired with depolarization induced LTD in Drd2^loxP/loxP mice (n = 6 cells/6 mice). In contrast, no change in synaptic transmission was observed in the Chl-Drd2KO mice after HFS (n = 7 cells/6 mice). (B) The magnitude of LTD was variable in Drd2^loxP/loxP after LTD induction. There was little change in synaptic transmission after HFS in Chl-Drd2KO mice. (C and D) LTD was lost in both the (C) dMSNs (n = 7 cells/5 mice) and (D) iMSNs (n = 6 cells/4 mice) of the Chl-Drd2KO mice, but it was intact in their littermate controls (dMSN n = 8 cells/7 mice; iMSN n = 7 cells/5 mice). (E and F) LTD was blocked in the presence of (E) AM251 (n = 7 cells/5 mice) and (F) sulpiride in Drd2^loxP/loxP mice (n = 6 cells/5 mice). (G) Perfusion of quinpirole did not restore LTD in Chl-Drd2KO mice (n = 7 cells/5 mice). HFS in the presence of quinpirole induced LTD in Drd2^loxP/loxP mice (8 cells/5 mice). (H) Bath application of DHPG induced LTD in both genotypes (Drd2^loxP/loxP n = 7 cells/6 mice; Chl-Drd2KO n = 6 cells/6 mice). (I) DHPG-induced LTD was blocked in the presence of AM251 in both genotypes (n = 5 cells/4 mice for each genotype). (J) HFS in the presence of VU 0255035 restored LTD in the Chl-Drd2KO mice (n = 6 cells/4 mice). (K) LTD was blocked in the presence of AM251 (n = 5 cells/3 mice).

Figure 6.. HFS Can Induce LTD at MSN Synapses in iMSN-Drd2KO Mice

Average evoked EPSC amplitudes plotted as a function of time. Insets: Representative traces showing first evoked EPSC before (1) and after (2) LTD induction protocol. Scale bars, 100 pA, 10 ms. Symbol representation: Drd2^loxP/loxP (gray circles), iMSN-Drd2KO mice (orange circles). Scatterplot of individual cell responses (averaged EPSCs during last 10 min of recording) after HFS. Data reported as mean ± SEM. (A) HFS induced LTD in Drd2^loxP/loxP (n = 12 cells/10 mice), but not significantly in the iMSN-Drd2KO mice (n = 7 cells/7 mice). (B) The magnitude and type of synaptic plasticity varied after HFS in both genotypes. (C) In the presence of AM251, HFS failed to induce a change in the average EPSC amplitude in both genotypes (n = 5 cells/3 mice for each genotype). (D) Sulpiride also blocked the induction of LTD in Drd2^loxP/loxP (n = 6 cells/6 mice) and iMSN-Drd2KO (n = 5 cells/4 mice) mice. (E) Pre-application of quinpirole enhanced the magnitude of LTD induction in both genotypes (n = 5 cells/4 mice for each genotypes). (F) Bath application of DHPG induced LTD in both genotypes (Drd2^loxP/loxP n = 5 cells/3 mice; iMSN-Drd2KO n = 6 cells/4 mice). (G) LTD induction after DHPG bath application was blocked in the presence of AM251 in both genotypes (Drd2^loxP/loxP n = 6 cells/5 mice; iMSN-Drd2KO n = 5 cells/4 mice).

Figure 7.. LTD Is Lost at Inputs onto iMSNs but not dMSNs in the iMSN-Drd2KO Mice

Insets: Representative traces showing first evoked EPSC before shown in black (1) and after the LTD induction protocol shown in color (2); Drd2^loxP/loxP EPSC traces (gray), iMSN-Drd2KO traces (orange). The dMSNs are filled color circles; iMSNs are open circles. Scale bars, 100 pA, 10 ms. Data reported as mean ± SEM. (A) HFS induced LTD in dMSNs from both genotypes (Drd2^loxP/loxP n = 10 cells/6 mice; iMSN-Drd2KO n = 11 cells/7 mice). (B) HFS induced LTD in the iMSNs of the Drd2^loxP/loxP (n = 11 cells/8 mice), but not in the iMSNs of the iMSN-Drd2KO mice (n = 10 cells/7 mice). (C and D) Quinpirole paired with HFS induced LTD in both (C) dMSNs (n = 6 cells/4 mice for each genotypes) and (D) iMSNs (Drd2^loxP/loxP n = 5 cells/3 mice; iMSN-Drd2KO n = 8 cells/7 mice) in both genotypes, thus restoring LTD in the iMSNs of the iMSN-Drd2KO mice.