Changes in synaptic function and structure caused by drugs of abuse

项目1：可卡因对直接和间接伏隔-被盖通路中棘神经元谷氨酸受体和钙信号的差异影响。本研究延续了我们之前的发现，即可卡因引起的脊柱密度增加和谷氨酸能传递增强，虽然需要时间发展，但它们是由几次重复的可卡因暴露引发的。接下来的实验试图通过研究每天注射7次可卡因（15 mg/kg）对表达d1r和表达d2r的msn上的突触的影响来阐明触发增强的事件。我们使用电生理记录结合双光子激光扫描成像和谷氨酸释放来探测谷氨酸受体的密度和组成，以及生理盐水对照和可卡因处理的D1-GFP小鼠的msn中单个脊柱的局部钙信号。虽然有一些证据表明可卡因可以通过增加NMDA-Rs的表面表达来产生沉默突触，但我们的实验表明，可卡因处理后，NMDA-R介导的解包裹诱导的脊髓兴奋性突触后电流（uEPSC）的大小在D1(+) msn中没有改变。同样，AMPA-R介导的uEPSCs在这些细胞中没有改变。在不改变突触电流的情况下，古柯碱组D1(+)- msn棘上谷氨酸引起的钙信号更大。通过NMDA-Rs的钙信号不受影响，这表明AMPA-Rs和/或电压敏感的钙通道在这种可塑性中起作用。进一步的实验旨在确定这种变化的机制，并研究D1(-) msn的变化。

Project 1: Differential effect of cocaine on glutamate receptors and calcium signals on spines medium spiny neurons of the direct and indirect accumbal-tegmental pathway. This study follows up on our previous findings that the cocaine-induced increase in spine density and the potentiation of glutamatergic transmission, while they take time to develop, are triggered by a few repeated cocaine exposures. The next experiments attempt to elucidate the events that trigger the potentiation by studying the effects of 7 daily injections of cocaine (15 mg/kg) on synapses made onto D1R-expressing and D2R-expressing MSNs. We use electrophysiological recordings in combination with 2-photon laser scanning imaging and uncaging of glutamate to probe glutamate receptor density and composition, as well as local calcium signals at individual spines in MSNs from saline controls and cocaine treated D1-GFP mice. While there are some evidences that cocaine can generate silent synapses by increasing surface expression of NMDA-Rs, our experiments showed that the size of the NMDA-R mediated uncaging-induced excitatory postsynaptic currents (uEPSC) at spines was unchanged in D1(+)MSNs by the cocaine treatment. Similarly, AMPA-R mediated uEPSCs in these cells were not altered. While synaptic currents were not changed, calcium signals evoked by glutamate at spines of D1(+)-MSNs were larger in the cocaine group. Calcium signals through NMDA-Rs were unaffected, suggesting a role for AMPA-Rs and/or voltage-sensitive calcium channels in this plasticity. Further experiments are aimed at identifying the mechanisms for this change as well as to investigate changes in D1(-)MSNs. Project 2: Development of addictive-like behaviors in mice and associated changes in synaptic transmission. We implemented a mouse model of cocaine self-administration (SA) that identifies addictive-like behaviors by evaluating 3 criteria described in the DSM-IV to define drug dependence: (1) perseverance, shown as difficulty in stopping drug use (2) motivation of cocaine seeking, shown by the effort exerted to obtain the drug, and (3) continued cocaine SA despite adverse consequences. Naive adult mice were trained to nose poke to gain an intravenous infusion of cocaine and allowed to self-administer cocaine for 30-40 days. A wide range of daily cocaine intake (5 - 23 mg/kg/day) and addictive-like behaviors were found. The behaviors were combined to create an addiction score for each mouse. While most animals with high cocaine intake displayed positive addiction scores, many managed to keep negative scores even while taking high doses for many weeks. Electrophysiological recordings were made from the two subtypes of MSNs: D1R-expressing (GFP positive) and from putative D2R-expressing (GFP negative) neurons in slices from sham surgery controls and cocaine animals. The AMPA/NMDA ratio of the eEPSC was significantly increased only in D1(+)MSN in cocaine animals. Interestingly, AMPAR mediated currents showed increased rectification mainly in D1(-)MSNs after cocaine. This indicated that cocaine self-administration had differential effects on transmission onto the two subclasses of MSNs. Using 2-photon imaging and glutamate uncaging, we next explored the mechanisms underlying the increased AMPA/NMDA ratio in D1(+)MSNs. The results showed that the AMPAR mediated uEPSC and the size of the spine heads were larger in D1(+)MSNs in cocaine animals. These strongly suggested that the cocaine-induced potentiation is caused by insertion of AMPA-R at synapses in direct pathway neurons. Resembling the variability observed in intake and addiction scores, the physiological data revealed a degree of changes across animals. Mice with low cocaine intake showed no increase in the AMPA/NMDA ratio in D1(+)MSNs suggesting that there is a threshold dose for this potentiation to develop. In mice with high cocaine intake, some, but not all, showed increased ratio but there was not correlated between plasticity and the addiction scores. Surprisingly, we found that plasticity in D1(-)MSNs was inversely correlated with the addiction scores. Mice with high intake that showed potentiation in D1(-)MSNs had negative addiction scores which were statistically different from the scores of high takers with no potentiation. In conclusion, cocaine self-administration triggered different changes at glutamategic synapses made onto direct and indirect accumbal-tegmental neurons. However, while plasticity in direct neurons was not correlated with addiction scores, potentiation of glutamatergic synapses onto indirect pathway neurons may provide protection from the development of addictive-like behaviors towards cocaine. Project 3: Properties and modulation of glutamate release from dopaminergic neurons in the nucleus accumbens. It has been shown that a subset of dopaminergic neurons in the VTA express the type 2 vesicular glutamate transporter (vGluT2) and they are able to release glutamate, in addition to dopamine, from terminal in the NAc. It is still unclear whether dopaminergic terminals co-release glutamate from the same terminals or even vesicles that contain dopamine. In this study, we investigated the properties and modulation of the glutamatergic transmission from the VTA. We reasoned that if glutamate and dopamine are co-released from the same terminals, glutamatergic transmission should be subject to the same presynaptic modulation than dopamine release. Channelrhodopsin 2 (ChR2) was selectively expressed in midbrain dopaminergic neurons and light pulses in the NAc generated an excitatory postsynaptic current (L-EPSC) in MSNs. The L-EPSC were mediated by both AMPA and MNDA glutamate receptors and, similarly to the dopaminergic transmission, they were depressed by the D2R agonist quinpirole. Glutamatergic L-EPSCs showed marked paired pulse depression that took minutes to recover, just like the slow recovery of the light evoked dopamine release measured with fast scanning cyclic voltammetry in these slices. A component of this paired pulse depression was dependent on D2R activation as it was minimized by D2R antagonist and enhanced by cocaine. Thus, similar to the dopamine release, glutamate release is regulated by activation of D2 dopamine presynaptic autoreceptors and shows an activity dependent recovery after paired pulse depression. Project 4: Changes in dendritic spine in medium spiny neurons after binge-like ethanol drinking in mice. This study examines the effects of repeated binge-like drinking on the synaptic morphology in the NAc and dorsolateral striatum (DLS). Adult mice were given short (2-4 hours) access to 20% ethanol solution for 4 days/week during 6 weeks. Control mice received water. Blood ethanol concentrations (BEC) were in average 90-100 mg/dl (BECs greater than 80 mg/dl defined as binge in humans by NIAAA). A lickometer was used to study drinking patterns and it revealed an early peak of drinking in ethanol mice. There was a weak correlation between BEC and ethanol intake in volume. However, a stronger correlation emerged between BECs and drinking time using the lickometer. This indicated that lickometer data may be a better predictor of BEC, given its improved temporal resolution. Two and 30 days after the final binge, brains were collected and labeled using DiOlitistics. Morphological analysis of spines showed increased density of spines mainly in the DLS after 30 days and some changes in spine morphology in NAc and DLS. In conclusion, in this mouse model of repeated binge-like ethanol drinking, we identified morphological changes at glutamatergic synapses mainly the DLS. These changes are similar to those we recently described in a monkey model of prolonged drinking (Cuzon et. al., 2011) and they may be associated with ethanol-triggered changes in connectivity in this brain region.