Table 1 Evidence for aberrant corticostriatal synaptic plasticity in diseases from clinical studies and animal models.

Loss of control

L-DOPA-induced dyskinesia (LID)

In L-DOPA-induced dyskinetic rats, no corticostriatal synaptic depotentiation was observed. Furthermore, activation of dopamine D1 receptors prevented depotentiation.

[95]

Activation of a subpopulation of D1 neurons caused dyskinesia in the absence of L-DOPA. Inhibition of these neurons ameliorated LID.

[97]

Dopamine depletion led to increased firing of D2 neurons during immobility in mice, while D1 neuron firing decreased. Moreover, a subpopulation of D1 neurons exhibited abnormally high firing rates that correlated with LID.

[96]

Obsessive-Compulsive Disorder (OCD)

In cognitive tasks, OCD patients were more prone to slips of action, suggesting a deficit in goal-directed control and an excessive reliance on habits.

[110]

Sapap3 knockout mice exhibited impaired behavioral response inhibition and reduced activity of striatal projection neurons. Optogenetic stimulation of the lateral orbitofrontal cortex and its striatal terminals restored both functions.

[113]

In Sapap3 knockout mice exhibiting OCD-like behaviors, corticostriatal postsynaptic potentials were impaired. Viral-mediated restoration of Sapap3 expression in the striatum rescued both synaptic and behavioral deficits.

[111, 112]

Clinical studies found that the anterior limb of the internal capsule, ventral capsule, ventral striatum, nucleus accumbens, ventral caudate, subthalamic nucleus (STN) and the inferior thalamic peduncle were all effective for DBS treatment of OCD. Side effects were mild, transient and reversible.

[207, 209]

Repeated orbitofrontal cortex-ventromedial striatum (VMS) stimulation in mice led to increased grooming that persisted after stimulation, and was associated with increased evoked firing of postsynaptic VMS cells. Both were reversed by chronic SSRI treatment.

[114]

DBS is effective in patients with severe OCD who are resistant to conventional pharmacological treatments

[208]

Loss of control

Substance use disorders

Optogenetic self-stimulation of VTA dopamine neurons facilitated cue-induced relapse of drug seeking in mice, associated with potentiation of excitatory afferents onto nucleus accumbens D1 neurons.

[128]

A review of 25 animal and 22 human studies (1974–2021) found that DBS of nucleus accumbens, insula, and STN reduced drug use and seeking in animal models. Human studies mostly targeted the nucleus accumbens and generally showed positive outcomes.

[194]

Repeated cocaine exposure reduced GABAergic inhibition, increasing VTA dopamine neurons’ susceptibility to long-term potentiation (LTP). The chronic exposure to cocaine disrupted the relationship between synaptic enhancement and behavior.

[135, 136]

5 patients received DBS of the nucleus accumbens for treatment-resistant alcohol addiction. All reported a complete absence of craving for alcohol, with no severe or long-standing side effects.

[196]

Cocaine potentiated excitatory transmission onto D1 neurons in the nucleus accumbens in mice, in parallel with locomotor sensitization. Depotentiation of cortical inputs to the nucleus accumbens by optogenetic stimulation in vivo restored normal transmission and abolished cocaine-induced locomotor sensitization.

[142]

3 patients received DBS of the nucleus accumbens to treat alcohol dependence. Craving was greatly reduced, and all patients achieved prolonged abstinence.

[197]

Cocaine induced sensitization and potentiation of excitatory inputs onto nucleus accumbens D1 neurons in mice were normalized by acute nucleus accumbens DBS combined with D1 receptor blockade. This intervention produced a long-lasting abolishment of behavioral sensitization.

[143]

DBS targeting the posterior nucleus accumbens and bed nucleus of the stria terminalis in a patient with severe addiction reduced drug craving and consumption.

[198]

Repeated morphine exposure potentiated excitatory transmission and increased GluA2-lacking AMPA receptor expression in nucleus accumbens D1 neurons in mice. In vivo optogenetic stimulation of infralimbic cortex-nucleus accumbens shell inputs blocked the reinstatement of morphine-induced conditioned place preference.

[144]

Reversing aberrant enhancements of excitatory synaptic inputs onto striatal D1 neurons is therapeutic in animal models of drug addiction (see left column for details).

[142,143,144,145, 190]

Experimentally induced LTP in D1 neurons in vivo caused a long-lasting increase in alcohol-seeking behavior in rats, while experimentally induced LTD decreased alcohol-seeking behavior.

[145]

Propranolol administration before memory reactivation disrupted the reconsolidation of smoking-related memories in smokers and decreased craving.

[228]

Repeated cocaine administration upregulated the NAc-cAMP system, contributing to tolerance of cocaine reinforcement. Acute inhibition of this system enhanced drug craving and relapse.

[149]

The NMDA receptor antagonist MK-801 impaired drug-seeking related memory reconsolidation, thereby reducing relapse in animal models.

[229,230,231]

Repeated alcohol consumption in mice strengthened glutamatergic transmission onto D1 neurons in the striatum. Chemogenetic excitation of these neurons in vivo promoted alcohol consumption behavior.

[190]

Lack of movement or motivation

Parkinson’s disease (PD)

In mouse models with dopamine neuron lesions, the lack of corticostriatal LTD in the indirect pathway was observed. Pharmacological restoration of this LTD improved motor control.

[150]

A gradual loss of long-duration response (LDR) was observed after cessation of dopamine replacement therapy in PD patients.

[200]

Discontinuation of L-DOPA treatment in a PD mouse model did not immediately reduce motor task performance. Instead, performance declined gradually and was dependent on continued task exposure.

[153]

LDR to dopamine replacement therapy in PD patients.

[201]

In mice with dopamine neuron lesions, motor task performance gradually worsened with continued task exposure. Repeated L-DOPA treatment rescued performance, and this improvement persisted despite treatment withdrawal. Both worsening and rescue were task-specific.

[155]

LDR to dopamine replacement therapy in PD patients was prolonged by A2A antagonists.

[202,203,204]

Drug-induced parkinsonism (DIP)

Chronic haloperidol treatment promotes dopamine D2 receptor-dependent corticostriatal LTP in mice.

[166]

A2A antagonism protects against haloperidol-induced catalepsy in rats.

[133, 134]

Sensitization of haloperidol-induced catalepsy in rats is a context-dependent learning phenomenon

[167]

D2 receptor blockade induced gradual and long-lasting motor performance impairment associated with corticostriatal LTP in mice. A2A antagonism protected against this “inhibitory motor learning”.

[152]

Haloperidol-induced catalepsy is absent in dopamine D2 receptor knockout mice

[168]

Chronic nicotine treatment mitigated the D2 receptor blockade-induced gradual and long-lasting motor impairment in mice.

[154]

Adenosine receptor antagonists reduced haloperidol-induced catalepsy in rats.

[169, 170]

Depression

A meta-analysis of reward-learning studies in patients with major depressive disorder (MDD) and controls found that MDD and anhedonia reduced reward sensitivity more than they affected the learning rate.

[174]

Studies have shown that DBS targeting the subcallosal cingulate gyrus, ventral capsule/ventral striatum, medial forebrain bundle, nucleus accumbens, and the anterior limb of the internal capsule is effective in treating treatment-resistant depression.

[212, 213]

A comprehensive review of the literature found that depression occurs in approximately 40% of PD patients.

[175]

Chronic stress decreased the strength of hippocampus–accumbens synapses and impaired LTP, antidepressant treatment reversed the stress-induced changes.

[214]

A review of the literature suggested that depression in PD patients seemed associated with decreased limbic CSTC circuit activity, whereas impulse control disorders seemed associated with increased limbic CSTC circuit activity due to dopamine replacement therapy.

[176]

Chronic stress decreased the excitatory synapses strength on D1 neurons in the nucleus accumbens in mice.

[177]

Social defeat stress and anhedonia increased excitatory transmission onto D2 neurons, while reducing it on D1 neurons and decreasing their dendritic complexity in the nucleus accumbens.

[178, 179]

Depressed patients showed a greater influence of aversive Pavlovian cues on both approach and withdrawal actions over instrumental behavior, suggesting disrupted neural circuits that modulate emotional and motivational processing.

[180]

In mice, serotonin neurons respond to both reward-predicting cues and rewards, whereas dopamine neurons shift from reward to cue responses during learning. Both neuron types are modulated by reward value and suppressed by stressors in the nucleus accumbens.

[183, 184]