Modulation of Long-term Synaptic Depression by D2 Dopamine Receptors Expressed by Different Striatal Neurons One of the long-term goals of this project is to understand the mechanisms underlying synaptic plasticity in the striatum, with an eye to determining how these mechanisms contribute to striatal-based learning and pathology in this brain region. We have characterized many of the mechanisms involved in long-term synaptic depression (LTD) at corticostriatal synapses. Among these is activation of D2-type dopamine receptors (D2Rs)that contributes to LTD induction, as indicated by previous pharmacological experiments with selective antagonists and constitutive knockout of this receptor in all cells throughout the body and brain. However, D2Rs are expressed by several striatal neuronal subtypes, including the medium spiny projection neurons (MSNs) and the cholinergic interneurons (CINs), and the cellular locus of the receptors involved in LTD induction is unclear. One possibility is that D2Rs on indirect pathway medium spiny projection neurons (iMSNs) that express D2Rs but not D1 receptors participate in LTD. However, this effect would likely occur only at synapses onto iMSNs, and not direct pathway neurons (dMSNs), despite the fact that LTD has been observed at synapses onto both MSN subtypes. There is also evidence that D2Rs on cholinergic interneurons (CINs) participate in LTD induction by inhibiting release of acetylcholine which normally suppresses LTD. To differentiate the roles in LTD of D2 on iMSNs and CINs we used a Cre x lox-based breeding scheme in mice to knock D2Rs out in iMSNs (ADORA2-Cre x D2-Flox) or CINs (ChAT-IRES-Cre x D2-Flox). In striatal slices from CIN-D2KO mice, LTD could not be induced by high-frequency stimulation (HFS) as measured either in field potential or whole-cell recordings, while LTD was intact in littermate controls. In contrast, LTD was intact in iMSN-D2KO mouse slices in field potential recordings, and in whole-cell recordings from a mixed iMSN/dMSN sample. The magnitude of LTD was reduced in iMSNs, but not dMSNs, in the iMSN-D2KO mice, assessed with whole cell recordings where the subtypes were identified by expressing a fluorescent protein in the dMSNs. Robust LTD could be restored in iMSNs in the iMSN-D2KO mice by bath application of the D2 agonist quinpirole, but this treatment did not restore LTD in the CIN-D2KO mouse slices. When HFS-LTD was observed it was always blocked by antagonists of D2Rs and the cannabinoid receptor type 1 (CB1)receptor. LTD induced by activation of group I metabotropic glutamate receptors was intact in all genotypes, supporting previous findings that this form of induction is independent of D2 receptors. In behavioral experiments, iMSN-D2KO mice showed lower rates of lever pressing compared to littermate controls in a self-paced instrumental learning task. However, the knockout mice were capable of learning this task. The iMSN-D2KO mice showed learning and performance equivalent to controls in an appetitive pavlovian conditioning paradigm. The CIN-D2KO mice show impaired learning and performance on a self-paced operant sequence-learning task. Our findings indicate that D2Rs modulate LTD induction, with a strong role for receptors expressed by CINs, and a weaker, iMSN-selective effect of those expressed by MSNs. The D2Rs on both neuronal subtypes contribute to performance of striatal-based instrumental performance tasks, and it will be interesting to determine if these impairments involve alterations in LTD. Endocannabinoids and CB1 Receptors Modulate Dopamine Release in the Nucleus Accumbens via Actions at Glutamatergic Afferents from Prefrontal Cortex Dopamine (DA) release from presynaptic terminals in the ventral striatum/Nucleus accumbens (NAc) can be driven by activation of CINs, and we have discovered that activation of CB1 receptors inhibits DA release induced in this manner both in brain slices and in vivo. However, CB1 receptors are not expressed by either CINs or dopaminergic neurons. These findings raised the question of the location in the circuit of the CB1 receptors that control DA release. The two major sites of CB1 expression in NAc are on the presynaptic terminals of glutamatergic cortical afferents, and on terminals of locally-connected GABAergic neurons. We found that CB1 agonist effects persisted in the presence of GABA receptor antagonists, and thus we focused our attention on the role of CB1 on cortical-NAc inputs, in particular those from the prefrontal cortex (PFC). Using optogenetics to specifically stimulate PFC afferents in NAc we were able to evoke DA release, and this DA release was reduced by CB1 agonist application. Thus, receptors on PFC terminals are a reasonable candidate to underlie the agonist actions we have observed. Evidence from other laboratories indicates that cortical activation leads to excitation of CINs, and activation of these neurons contributes to the DA release induced by cortical stimulation. We found evidence for involvement of a similar mechanism in PFC-induced DA release. Thus, one of the predominant effects of CB1 is to reduce PFC glutamatergic activation of CINs. Indeed, PFC-driven glutamatergic responses in CINs are inhibited by CB1 agonist. To determine if CB1 receptors on PFC terminals are necessary for the agonist-induced depression of DA release, we knocked the receptors out only in the PFC. This CB1 knockout eliminated the CB1 agonist modulation of DA release induced by either CIN or PFC optogenetic activation, supporting the hypothesis that these receptors have important roles in modulation. It was not clear how PFC afferents are activated by optogenetic stimulation of CINs to participate in DA release. Indeed, CIN activation stimulates DA release directly through activation of nicotinic ACh receptors (nAChRs) present on DA terminals. However, we had already found that antagonists of the AMPA-type glutamate receptor reduce CIN-induced DA release and occluded the CB1 agonist modulatory action. Thus, our findings indicate a strong glutamatergic contribution to DA release that is CB1-modulated. We then found that activation of alpha7 containing nAChRs increases glutamatergic transmission onto CINs, including glutamate release induced by optogenetic PFC stimulation. Thus, CIN activation recruits CB1-containing PFC terminals via a sequence events involving ACh release activation of presynaptic alpha7 nAChRs increased glutamate release. We also found that AMPA receptors are expressed on dopaminergic axons and varicosities in NAc, and that AMPA application induced DA release. These receptors may participate in the glutamatergic mechanisms that enhance DA release after CIN activation. CIN-induced DA release can be modulated by the endogenous cannabinoids (eCBs) that are the biological CB1 agonists. Activation of CINs with bursts of optogenetic stimulation resulted in DA release that was augmented by a CB1 antagonist. Application of a blocker of the enzyme that degrades the eCB arachidonoylglycerol (2-AG), monoacylglyceol lipase (MGL), inhibited DA release induced by a single CIN stimulus, and this inhibition was blocked by a CB1 antagonist. Thus, enhancing tissue 2-AG levels leads to CB1 activation and suppression of DA release. Thus, eCB release can induce CB1-mediated reduction in DA release under certain conditions. Intracranial self-stimulation (ICSS) induced by optogenetic PFC activation is reduced by MGL inhibition, and this inhibition is prevented by a CB1 antagonist or by knocking out CB1 expression in PFC. Lever-pressing rates in this task were enhanced in mice lacking CB1 in PFC. This ICSS was inhibited by intra-NAc injection of either an nAChR or D1 DA receptor antagonist. These findings provide evidence that eCBs modulate reward-driven behavior that involves ACh and DA in the NAc.
Showing the most recent 10 out of 72 publications
