How have model systems of dystonia contributed to our understanding of pathogenesis over the past 10-years?
A recurrent term that emerges when describing dystonia is heterogeneity, as it encompasses a wide group of movement disorders, with heterogeneous etiology, classification, anatomical, and physiological substrates. In the last decade, the development of several models has contributed to our knowledge on the pathophysiology of dystonia, allowing us to better explore the molecular pathways and neuronal networks underlying the disease. The development of better therapeutic strategies is still limited, although significant steps forward have been made.
Genetics has been highly successful in the identification of the numerous causative genes of different forms of monogenic dystonia, thereby providing the possibility to generate multiple animal models. At the system level, evidence from clinical, neuroimaging, and neurophysiology studies support the concept of dystonia as a network disorder, including the cortical-thalamic-basal ganglia pathway, and more recently, cerebellar pathways, although the precise role and the relevance of each component have not been elucidated yet. At the circuit and cellular level, altered synaptic plasticity, caused by an abnormal neurochemical balance between striatal cholinergic and dopaminergic signalling, is consistently implicated in many different forms of dystonia. At the molecular level, several novel genes encode proteins with distinct biological functions, including dopamine signalling mitochondrial function, heavy metal accumulation, endoplasmic reticulum and nuclear envelope dysfunction, and lipid metabolism. Of note, despite the heterogeneity in gene mutations, these often converge on shared molecular pathways.
Dystonia is a heterogenous group of disorders. It is likely that there are many routes of dysfunction that result in the uncontrolled muscle contractions and abnormal postures that constitute dystonia. Over the last 20 years, multiple rodent models of dystonia have been analyzed. More recently, cellular models derived from humans with genetic forms of dystonia have also been studied. Both animal and cell models provide data regarding intracellular abnormalities in dystonia.
With the identification of multiple gene mutations that can cause dystonia in humans, rodent models with these gene mutations have been created. While the rodent models do no exhibit evident uncontrolled muscle contraction and thus do not display overt dystonia, there has been evidence of abnormalities in cholinergic neurotransmitter release in several, distinct TOR1A mouse models and abnormalities in myelination in cells from mice that do not express THAP1.
A rat model of blepharospasm has been established, in which treatment with a drug that interferes with striatal dopaminergic transmission combined with a facial nerve lesion produces increased eyelid closure. This model demonstrates that relatively mild defects in two essential physiologic functions (neurotransmitter function and nerve function) can produce symptoms consistent with an isolated dystonia.
Taken together, genetic models of dystonia and animal models that display overt signs of dystonia both give information regarding pathophysiology. These diverse models indicate that dysfunction at one or multiple levels within the nervous system (intracellular defects, neurotransmitter signaling defects and/or in combination with abnormal nerve function) contribute to the development of dystonia.
The most commonly used experimental models exploited to advance our understanding of the dystonias have been animal models. For animals, we have both rodent and primate models, as well as some other species. The animal models fall into two broad categories. They include phenotypic models that replicate the abnormal movements seen in humans with dystonia, and etiologic models that replicate a known cause for dystonia. Both have contributed substantially to our understanding of the pathogenesis of dystonia.
The phenotypic models have transformed our understanding of the anatomical underpinnings of dystonia. Traditionally, most of our information came from lesion studies of humans, and most of the human data suggested that dystonia was a disorder of the basal ganglia. The phenotypic animal models confirmed this concept, but also demonstrated that the anatomical basis was a bit more complex. The phenotypic animal models provided strong evidence for our modern conceptualization of dystonia as a disorder of a broader motor network, where dystonia may sometimes arise from a defect in the basal ganglia, and other times the defect arises in the cerebellum or elsewhere. There are phenotypic animal models with dystonia where the initial defect is unequivocally originating in the basal ganglia, and others where it is unequivocally arising from the cerebellum.
The etiologic models are often based on genes known to have dystonia in humans. Like most other genetic models for other movement disorders, such as Parkinson disease or Huntington disease, the genetic models often do not replicate the overt motor features. Despite the absence of overt dystonia, the genetic models have been instrumental in advancing our understanding of the molecular pathogenesis of numerous inherited forms of dystonia, such dopa-responsive dystonia and dystonia associated with the TOR1A and THAP1 genes. Although the genes are all quite different, converging evidence has suggested a role for common mechanisms, such as defects in synaptic signaling, defects in ion channels, and a developmental window of vulnerability for many types.
2. Can these models directly contribute to our understanding of human disease? Or are they too far removed?
Investigations of animals provide results that do not fully reproduce the clinical features of human disease. Indeed, most of the available models do not exhibit overt symptoms although they do have subtle behavioral abnormalities and neurochemical and neurophysiological alterations. Despite such limitations and apparent contradictory evidence, a useful model must be judged by how reliably and effectively it can be used to explore novel aspects of pathophysiology and potential treatments. Selection of a particular model system largely depends on the specific hypotheses and overall goals of the experiment. In vitro studies are often devoted to assessing biochemical and cell biology issues, whereas mammalian models are essential for exploring circuits and evaluating the efficacy of candidate drugs. Altogether, models have been certainly useful to confirm what was hypothesized by clinical observations and provided an excellent platform to explore dystonia pathophysiology from a causal and mechanistic viewpoint. More recently, evidence emerged in favor of a gene-environment interaction in dystonia, which should be considered in the continued quest to better model human disease, and gain more insights in dystonia pathophysiology. To translate such advances into clinical benefit for patients, we need comprehensive models that better reproduce symptoms in order to identify new treatment targets. To this aim, novel experimental tools are expected to provide relevant novel information (see my answer to question 4 below).
The cellular and animal models do provide some information regarding the pathogenesis of human dystonia. Specifically, each model elucidates a possible pathway that leads to dystonia. The identification of malfunctions in neurotransmitter release and intracellular pathology, discussed in greater detail in my answer to question 3 below, provides a blueprint for the development of drugs that may normalize these pathways.
This question feels a bit leading because it implies that models may not be useful. Let us first consider what we mean by “directly” here. Studies of humans are almost never direct. The types of studies required to understand the biology of dystonia often require direct manipulations of the brain, and such manipulations are not technically or ethically feasible in human research subjects. For example, experiments that directly establish mechanism may require making a lesion in a specific brain region to silence it, or stimulating a specific brain region to see what happens. Other examples include determining what happens when a specific gene or biochemical process is manipulated. We cannot do any of these things in humans. There is an obvious limit to what can be done with post-mortem human brains, so the best we can do with human studies is non-invasive imaging or look at downstream effects of DBS targeting a single area. The results from these types of human studies provide only very indirect information and distinguishing cause versus effect is almost never possible. In contrast, direct studies of the brain are feasible in animals. What makes the findings from animals less “direct” or maybe even “too far removed” is that they come from a different species. While different mammals certainly have brains that look different macroscopically, the underlying circuits and neurotransmitters are strikingly conserved. Animals have guided our understanding of the majority human neurological disorders, so there is no reason to suspect they cannot help us with dystonia.
We have always advocated an interactive strategy for human and animal studies. Questions that arise in humans can be evaluated in animals. Conversely, findings in animals can often be verified in humans. When animal studies produce results that were not expected from our existing viewpoint of human dystonia, we have two options. We can dismiss the findings from animals as “incorrect” and “too far removed”. Or we can ask ourselves if our viewpoint of human dystonia deserves to be reconsidered.
3. From the work to date, what key mechanisms/pathways have emerged as central in dystonia?
Several novel aspects of dystonia pathophysiology emerged in the past decade. Clinical and experimental evidence emerged to suggest that despite being a heterogeneous disorder, different forms of dystonia indeed share commonalities commonalities. A relevant contribution to such idea came from genetics, which led to a significant growth in the number of genes associated with dystonia. These genes encode proteins with distinct biological functions including mitochondrial dysfunction, heavy metals accumulation, endoplasmic reticulum (ER) and nuclear envelope dysfunction, and lipid metabolism. Of interest, among the shared biological pathways are defects in dopamine signalling. Indeed, multiple mutations causing dystonia to converge, to affect dopamine signal transduction pathways, including GTPCH1, GNAL, and ADCY5, to name a few. The role of dopamine emerges also from clinical observations since dystonia may appear in early-onset inherited Parkinson disease and in different dopaminergic states. Similarly, dopamine receptor-blocking drugs may cause dystonia. Several lines of observations emphasize the centrality of a dysregulation of cholinergic transmission at the striatal level, particularly in DYT1 dystonia, with a hypercholinergic tone resulting in an aberrant striatal network activity. Elucidation of these and other shared pathways is relevant for understanding the biological basis of dystonia and for designing novel therapeutics that may have a broad potential for distinct types of dystonia.
While many genes and genetic loci have been associated with dystonia, the cellular pathways that become disrupted and lead to dystonia remain poorly understood. The best-defined pathways are those that affect the dopamine biosynthetic pathway. Defects in dopamine biosynthesis result in dystonia in childhood, and reduced dopamine production can result in dystonia as a treatment-related complication in Parkinson disease.
The intracellular stress response pathway appears to be affected in cellular studies of TOR1A and PRKRA gene mutations. Thus, cells with a defective ER-mediated stress response appear to be more likely to undergo apoptosis and contribute to the development of dystonia.
In addition, data from both gene expression studies conducted in neural stem cells derived from humans with different THAP1 mutations and mouse models of THAP1 dystonia, point to abnormalities in expression of genes related to myelination. Thus, defects in myelin production and the glial cells that produce it may play a role in the pathogenesis of some forms of dystonia.
In summary, defects in dopamine synthesis, the ER stress response, and myelin production are implicated in the pathogenesis of dystonia.
There have been many advances at many levels. At the anatomical level, the traditional view of dystonia as a basal ganglia disorder has been replaced by the modern view that dystonia is a disorder of a motor network. This new view has opened our eyes to other regions of the brain that also need to be studied. The new view started with evidence from animal models but has now been replicated in humans with dystonia.
At the genetic level, we used to believe we had only a few genes responsible for dystonia. Now we have hundreds. While only a handful of genes cause isolated dystonia, online databases such as OMIM now list more than 300 different genes that cause dystonia combined with other neurological problems. This longer list of genes has enabled us to begin to lump and split them into subgroups that may be biologically related. Some of the shared molecular pathways that are the best recognized include impairments in dopaminergic transmission, defects in energy production, alterations in ion channels that alter synaptic activity, and changes in protein trafficking or quality control. These advances have allowed us to begin to consider common themes for molecular pathogenesis, and they lead to ideas for common molecular targets for therapy.
4. How might these findings advance future research?
Although distinct modelling approaches present different levels of validity, they allow to improve our knowledge of dystonia pathophysiology. To this purpose, the availability of multiple models is welcome, and might certainly be beneficial. Advances in experimental tools will help to answer fundamental questions.
Novel techniques, including optogenetics, designer receptors exclusively activated by designer drugs (DREADDs), and CRISPR, offer unprecedented abilities to alter specific targets to address their role in motor function and dysfunction. By means of optogenetic tools and DREADDs, we are now able to modulate the activity of a single type of neuron in a given brain region. Moreover, the novel and highly accurate genome-editing technology based on the bacterial CRISPR/Cas9 system will allow faster DNA editing yet opening new opportunities for modelling human diseases. In addition, the generation of protocols for directed differentiation of human pluripotent stem cells into basal ganglia neurons would be a relevant tool to characterize human neurons and eventually confirm data from animal models or gain novel insights into disease pathophysiology. Finally, novel imaging and machine-learning-based protocols may provide unbiased results on brain regions and circuitries involved.
In conclusion, we should be confident that exciting findings will be available in the near future, as these techniques are becoming widely used and applied to the study of movement disorders, specifically dystonia. Overall, the findings on the mechanistic underpinnings uncovered so far should help to promote the design of novel agents for these molecular targets, such as dopamine and muscarinic acetylcholine receptors.
The identification of abnormalities in myelination and the ER stress response point to specific cellular pathways that may be targeted in drug development. In addition, animal models such as the rat model of blepharospasm may help to identify the triggers that push non-manifesting carriers of a TOR1A or THAP1 mutation into manifesting dystonia.
The advances in our understanding of the underlying anatomy of dystonia have opened our eyes to other brain regions that need to be studied. These other regions are now being studied in humans using functional imaging such as fMRI and PET, non-invasive physiology such as transcranial magnetic stimulation, and even post-mortem pathological studies. Some have even begun to exploit the new information to explore novel treatment strategies. We have had deep brain stimulation of the basal ganglia as a treatment for dystonia for years, and some investigators are now looking into stimulation paradigms for the human cerebellum.
The advances in our understanding of the molecular basis of dystonia have been coming so fast it has been hard to keep up with them. The most obvious impact of these findings has been related to diagnostic tools. Several recent studies have now shown that a definitive molecular diagnosis can be achieved in 20-30% of dystonia cohorts. We still cannot find answers for every patient, but even this amount is far better than we were doing 10 years ago, when less than 2% of patients could get a molecular diagnosis. The molecular diagnosis is becoming increasing important, because several specific subtypes now have highly effective interventions, and early treatment produces a better outcome.
 Albanese A, Bhatia K, Bressman SB, Delong MR, Fahn S, Fung VSC, et al. Phenomenology and classification of dystonia: A consensus update. Mov Disord 2013;28:863–73. https://doi.org/10.1002/mds.25475.
 Peall KJ, Kuiper A, de Koning TJ, Tijssen MAJ. Non-motor symptoms in genetically defined dystonia: Homogenous groups require systematic assessment. Park Relat Disord 2015;21:1031–40. https://doi.org/10.1016/j.parkreldis.2015.07.003.
 Wadon ME, Fenner E, Kendall KM, Bailey GA, Sandor C, Rees E, et al. Clinical and genotypic analysis in determining dystonia non-motor phenotypic heterogeneity: a UK Biobank study. J Neurol 2022. https://doi.org/10.1007/s00415-022-11307-4.
 Keller Sarmiento IJ, Mencacci NE. Genetic Dystonias: Update on Classification and New Genetic Discoveries. Curr Neurol Neurosci Rep 2021;21. https://doi.org/10.1007/s11910-021-01095-1.
 Li J, Liang CC, Pappas SS, Dauer WT. TorsinB overexpression prevents abnormal twisting in DYT1 dystonia mouse models. Elife 2020;9:1–20. https://doi.org/10.7554/eLife.54285.
 Aïssa HB, Sala RW, Georgescu Margarint EL, Frontera JL, Varani AP, Menardy F, et al. Functional abnormalities in the cerebellothalamic pathways in a mouse model of DYT25 dystonia. Elife 2022;11. https://doi.org/10.7554/eLife.79135.
 Balint B, Mencacci NE, Valente EM, Pisani A, Rothwell J, Jankovic J, et al. Dystonia. Nat Rev Dis Prim 2018;4. https://doi.org/10.1038/s41572-018-0023-6.
 Rauschenberger L, Knorr S, Pisani A, Hallett M, Volkmann J, Ip CW. Second hit hypothesis in dystonia: Dysfunctional cross talk between neuroplasticity and environment? Neurobiol Dis 2021;159. https://doi.org/10.1016/j.nbd.2021.105511.
 Jinnah HA, Sun Y V. Dystonia genes and their biological pathways. Neurobiol Dis 2019;129:159–68. https://doi.org/10.1016/j.nbd.2019.05.014.
 Goodchild RE, Grundmann K, Pisani A. New genetic insights highlight “old” ideas on motor dysfunction in dystonia. Trends Neurosci 2013;36:717–25. https://doi.org/10.1016/j.tins.2013.09.003.
 Eskow Jaunarajs, KL; Bonsi, P; Chesselet, MF; Standaert, DG; Pisani A. Striatal cholinergic dysfunction as a unifying theme in the pathophysiology of dystonia. Prog Neurobiol 2015;127–128:91–107.