Imagine one day you notice a tremor and can’t walk easily anymore. Your doctor tells you that you have Parkinson’s disease and gives you levodopa, a 60-year-old treatment that temporarily relieves your symptoms. But over time, the medication makes you impulsive, and after a few years, your symptoms return. The doctor tells you that you have developed resistance to your medicine and that nothing else is likely to work—your condition continues to degenerate. Sadly, this is the reality for millions of Americans suffering from Parkinson’s disease.
This need not be the case. Despite half a century in which huge strides have been made in our understanding of the causes of Parkinson’s disease, our treatments are still largely based on the oversimplified idea that augmenting dopamine is all that matters. For the last two decades, we have known that Parkinson’s disease, as well as a number of other movement disorders, result from an imbalance in the activity of two opposing neural pathways in a brain region called the striatum (1–3). Overactivity of the “direct pathway” promotes disorders of excessive movement, such as Huntington’s disease, dystonia, and Tourette syndrome; overactivity of the “indirect pathway” promotes disorders of insufficient movement, such as Parkinson’s disease. Why have we not transformed this model into effective therapies for movement disorders?
Most therapeutic interventions fail because they do not target the direct or indirect pathways with sufficient precision in human patients. Better therapies will require a deeper understanding of the organizing principles of striatal circuits. Can we find neurons that selectively tune output of the direct and indirect pathways? I believe the answer is yes, if we target fast-spiking interneurons (FSIs).
FSIs belong to a family of neurons called inhibitory interneurons, which specialize in controlling firing rate and timing of activity across ensembles of neurons. Because neurons in the striatum of patients with movement disorders show abnormalities in both rate and timing, I hypothesized that dysfunction of FSIs would be a common theme across a broad array of human diseases.
To determine how FSIs communicate with direct and indirect pathway neurons, I performed recordings in transgenic mice, bred to express different fluorophores in each cell type (4). These experiments revealed that FSIs preferentially target direct pathway neurons. Based on this observation, I hypothesized that loss of local inhibition from FSIs would create an imbalance between the direct and indirect pathways. To test this idea, I developed a pharmacological approach to reduce FSI activity acutely and reversibly in the striatum. The strategy took advantage of the fact that the glutamate receptors expressed by FSIs lack a molecular subunit commonly found in other striatal neurons, called GluA2 (5, 6). Without GluA2, the glutamate receptors in FSIs are blocked by a drug called IEM-1460 (7). Without glutamate, FSIs cannot become activated, so IEM-1460 effectively silences these neurons. I confirmed this approach synaptically in acute brain slices and functionally with extracellular recordings in vivo through a collaboration with the Berke laboratory (5).
Would temporarily silencing FSIs impair movement in freely moving mice? Indeed, infusions of IEM-1460 into the striatum produced motor symptoms resembling dystonia seen in human patients. Dystonia is characterized by excessive muscle contractions and twisted postures and is thought to reflect disproportionately high activity of the direct pathway (8). The finding that many symptoms of dystonia are reproduced by suppressing striatal FSIs argues that these neurons are central to the etiology of the disease and that silencing these neurons destabilizes the balance between the direct and indirect pathways.
Does altered activity of FSIs also destabilize the balance between the direct and indirect pathways in Parkinson’s disease? To address this question, I turned to a mouse model of Parkinson’s disease, in which chemical ablation of dopamine neurons with the toxin 6-hydroxydopamine (6-OHDA) produces a well-documented increase in indirect pathway activity and reduced locomotion. Because FSIs are inhibitory, I hypothesized that loss of FSI innervation onto indirect pathway neurons would contribute to overactivity of this pathway in Parkinson’s disease. To my surprise, I found exactly the opposite. Dopamine depletion induced a rapid sprouting of FSI axons and a doubling of their synaptic contacts selectively onto indirect pathway neurons. At the single-cell level, this result seemed paradoxical. But at the network level, this result made sense if increased innervation from FSIs worked to promote synchrony across indirect pathway neurons. Synchrony, or the simultaneous firing of large groups of neurons, is often promoted by FSIs, because single FSIs inhibit hundreds of neurons—and when hundreds of neurons are turned off at the same time, they tend to come back online at the same time. In fact, a hallmark of striatal dysfunction in Parkinson’s disease is the onset of pathological synchrony in the movement-suppressing indirect pathway (9). High levels of synchrony are thought to be more destructive to movement than changes in neuronal firing rate, just as a group of people are much louder if they speak in unison, rather than individually. To test whether plasticity of FSIs caused pathological synchrony across indirect pathway neurons, I generated a computer model of the striatal circuit. When FSI connectivity was increased in a pathway-specific manner, as in my experimental data, I found that this caused pathological synchrony across indirect pathway neurons (10). This experiment illustrated that FSIs act in a pathway-selective manner to alter the balance of direct and indirect pathway output in an animal model of Parkinson’s disease. …Continue