Dystonia is a neurological movement disorder characterized by involuntary and intermittent or sustained muscle contractions leading to abnormal, repetitive twisting movements and/or abnormal postures [1]. This condition can have diverse manifestations, affecting specific muscle groups or the entire body, leading to a loss of coordinated movements [2]. It is a highly heterogeneous neurological movement disorder both clinically and genetically and in recent years many important genetic as well as molecular insights have suggested several therapeutic drug targets [3]. However, the translation of such knowledge into new therapies is yet to emerge as developing effective drugs involves in-depth research on identified genes, requiring significant resources and time. The genetically inherited monogenic dystonia manifests in various forms; each one characterized by distinct features [2]. Focal Dystonia targets specific body regions, such as the neck (cervical dystonia), eyelids (blepharospasm), hand (writer’s cramp), or vocal cords (spasmodic dysphonia). In contrast, segmental dystonia impacts adjacent body parts, potentially combining areas like cervical and oromandibular dystonia. Generalized dystonia extends its reach across multiple or all body parts, exerting a notable influence on both upper and lower extremities, thereby affecting mobility and posture. Hemidystonia uniquely affects one side of the body, inducing muscle contractions and abnormal movements. Multifocal dystonia involves multiple non-contiguous body parts, presenting a diverse clinical picture. Task-Specific dystonia emerges during specific activities, such as musician’s dystonia or writer’s cramp and paroxysmal dystonia is marked by intermittent episodes of dystonia. This spectrum highlights the complex nature of dystonia and the various ways it can manifest in affected individuals.

DYT-PRKRA is caused by mutations in the PRKRA gene (OMIM: 612067), which encodes the protein activator (PACT) of the interferon-induced protein kinase PKR [4]. The characteristics of DYT-PRKRA patients have been summarized in a recent review [2] and in Table 1. The vast majority of PRKRA mutation carriers show generalized dystonia, but some patients with segmental/multifocal dystonia or focal dystonia have been noted. DYT-PRKRA most often starts in the limbs (upper > lower), sometimes cervical or laryngeal, and rarely in the neck. Tremor was reported in some patients, myoclonus in none of them, and Parkinsonism was described in about half the patients. Information on psychiatric signs and other nonmotor symptoms is rarely indicated but cognitive impairment and global developmental delay especially after a childhood febrile illness has been noted [5–9]. The age of onset was reported to be early during childhood in most cases but later onset during adulthood has been observed indicating environmental or other modifying genetic factors. Abnormalities and degeneration in striatal region have been noted in a few patients but this information was not available for most patients [6–8]. Investigations into structural brain changes in DYT-PRKRA patients remain ongoing and a possible neurodegenerative classification of DYT-PRKRA can be considered after such analysis in additional DYT-PRKRA patients. The globus pallidus internus (GPi) region has evolved as a potential target for deep brain stimulation (DBS) and GPi-DBS is used as a therapeutic intervention for several types of dystonia [10]. However, GPi-DBS has not shown much benefit in several DYT-PRKRA patients and other established treatments including botulinum toxin injections, baclofen, and benzodiazepines were shown not to be beneficial [2]. In one case, DYT-PRKRA was reported to improve after thiamine therapy [11], but this has not been reported in other cases. Thus, understanding the molecular mechanisms responsible for DYT-PRKRA is a priority of significant importance for developing novel and effective treatment options.

The most studied function of PACT is its role in catalytic activation of the interferon-induced protein kinase PKR (protein kinase, RNA activated) via a direct interaction. PKR (aka EIF2AK2) is a serine threonine protein kinase that was originally discovered in the context of antiviral innate immune response [12]. PKR is ubiquitously expressed at low constitutive levels and its kinase activity stays latent until bound by an activator. Upon binding to one of its two activators: i) double-stranded (ds) RNA [13], or ii) protein activator PACT [4] PKR undergoes autophosphorylation and enzymatic activation. The dsRNA-dependent PKR activation occurs mainly during viral infections [14], and in uninfected cells PACT activates PKR in response to oxidative stress, endoplasmic reticulum (ER) stress, and serum deprivation [15, 16]. Patel and Sen cloned and identified PACT as a stress-modulated protein activator of PKR in 1998 [4]. Since then, the functional involvement of PACT-mediated PKR activation in regulating cellular response to diverse types of stress signals has been studied extensively [15, 17–19]. The functional domains of PKR and PACT have been characterized in detail and both PACT and PKR have the evolutionarily conserved dsRNA binding motifs (dsRBMs) [20–22] that also mediate the dsRNA independent protein-protein interactions between them and with other proteins that contain dsRBMs [23–25] (Figure 1). Upon binding dsRNA or PACT via the dsRBMs, PKR undergoes a conformational change which results in the autophosphorylation and activation of PKR [26, 27]. PACT is a stress-modulated activator of PKR that acts via a dsRNA-independent interaction in response to ER stress, oxidative stress, and serum deprivation [15, 16, 28]. Of the three dsRBMs present in PACT, the two amino terminal motifs dsRBM1 and 2, are critical for dsRNA binding and PACT-PKR interaction and a carboxy terminal dsRBM3 motif that does not bind dsRNA is essential for PKR activation [4, 23, 24]. Within dsRBM3, serines 246 and 287 serve as phosphorylation sites to enhance PACT-PACT homomeric interaction and the heteromeric interaction of PACT’s dsRBM3 with PKR’s catalytic domain that takes place only after PACT undergoes a stress-induced phosphorylation of serine 287 [19, 29] (Figure 2). In the absence of stress, PACT is constitutively phosphorylated on S246 [29], associates with PKR weakly [30] and is unable to activate PKR. Once phosphorylated on serine 287 in response to cellular stress, PACT’s affinity tor PACT-PACT and PACT-PKR interactions increases, thereby leading to efficient PKR association and catalytic activation [17, 30].

In last few years, several mutations have been identified (Figure 3) in PRKRA gene (OMIM: DYT16, 612067) leading to DYT-PRKRA [5–9, 31–40]. Although DYT-PRKRA was originally described to have an autosomal recessive inheritance pattern [31] but dominantly inherited variants of DYT-PRKRA have also been reported [32, 38]. Most mutations reported in DYT-PRKRA are substitution mutations that map within either the dsRBM1 and 2 or in the linker region between dsRBM2 and dsRBM3. One frameshift mutation reported in a single patient produces an early stop codon and truncates the PACT protein within dsRBM1 [32]. It is unclear if such a truncated protein would be present in the patient as no study has been conducted on patient cells. However, this truncated protein if present, will be unable to activate PKR via a direct interaction as dsRBM3 is essential for PKR activation. It is important to note that in several of DYT-PRKRA cases, developmental regression and dystonia was first noted after a febrile illness in the childhood [5–9]. This detail becomes relevant in the context of the cellular functions of PACT discussed in this review. The effects of one frameshift and several substitution mutations on PACT’s functional contribution to various cellular pathways has been studied and is discussed in the next section of this review [41–44].

PACT impacts cellular regulation via its participation in several pathways relevant to dystonia and Figure 4 summarizes these pathways as well as how they are altered in DYT-PRKRA to affect cellular responses and function.

Soon after cloning and characterization of human PACT as a PKR activator [4], the murine homolog of PACT was identified and termed RAX [16]. Human and murine proteins are highly homologous differing only in 6 amino acids, 4 of which are conservative changes [4, 16]. Like human PACT, murine PACT activates PKR by a direct interaction in response to cellular stress and regulates cellular fate [16, 28, 86]. A targeted disruption of murine Prkra gene demonstrated its functional contribution to craniofacial and postnatal pituitary development [87]. PACT null mouse had reduced size, severe microtia, hearing loss, reproductive issues, and diminished pituitary function. Surprisingly, these effects on the pituitary growth and function were dependent on activation of PKR and revealed that PACT functions as a PKR inhibitor in pituitary cells [88]. Such a role reversal of PACT’s function has also been observed in the context of human immunodeficiency virus (HIV) replication [89, 90]. A missense mutation S130P in the second dsRBM of murine PACT resulted in defects in ear development, growth, craniofacial development, and ovarian structure [91]. Another study reported that deletion of the entire Prkra gene in mice is embryonic lethal at a preimplantation stage of development [92]. Interestingly, the same study also reported that Drosophila carrying a mutation in loquacious, a Prkra homolog, have a severe defect in nervous system coordination or neuromuscular function resulting in significantly reduced locomotion.

The most relevant for the topic of dystonia is a recent study of a recessively inherited spontaneously arisen frameshift mutation (Figure 9), Prkra ^lear-5J [93]. Mice homozygous for this mutation exhibit craniofacial developmental abnormalities, reduced body size, kinked tails, and progressive dystonia with altered gait beginning at 2 weeks of age and continuing until death at about 3 weeks of age. Some neurons in the dorsal root ganglia and the trigeminal ganglion were apoptotic, consistent with the observed neurodegenerative phenotype. Basic neurological testing on mutant mice showed that the mutant mice had an elongated step/push gait, no grasping reflex with the hind paws, a weak grasping reflex with the forepaws, kinked tails and gnarled wrists. The kinked tail and gnarled wrist phenotypes were determined to result from dystonia as the bone structure of the tail and wrists was normal. The biochemical and developmental consequences of the Prkra ^lear-5J mutation were investigated recently [94]. The truncated PACT protein produced due to the frameshift mutation retained its ability to interact with PKR, however as it lacked the dsRBM3 required for PKR activation, it inhibited PKR activation. Furthermore, mice homozygous for the mutation had abnormalities in the cerebellar development as well as a severe lack of dendritic arborization of Purkinje neurons. Reduced eIF2α phosphorylation was noted in the cerebellums and Purkinje neurons of the homozygous Prkra ^lear-5J mice indicating that PACT mediated regulation of PKR activity and eIF2α phosphorylation plays a role in cerebellar development and may contribute to the dystonia phenotype resulting from this Prkra mutation.

Cellular stress response and dysregulated eIF2α phosphorylation has emerged as a major area of functional convergence [3] among various monogenic dystonia types (Figure 10). Research on DYT-PRKRA established that enhanced PKR activation and dysregulated eIF2α signaling caused increased sensitivity to apoptosis in DYT-PRKRA patient cells after endoplasmic reticulum (ER) stress [42–44]. Following this initial report for DYT-PRKRA, several other dystonia types also reported dysregulated eIF2α pathway as a possible pathomechanism. DYT-TOR1A is a childhood-onset autosomal-dominant disease caused by a single amino acid deletion in the ER-resident torsinA protein. DYT-TOR1A patient cells exhibit activated ER stress response and eIF2α signaling is dysregulated [95]. Remarkably, in case of DYT-TOR1A when the eIF2α phosphorylation status was restored to normal levels, the dystonia symptoms were alleviated [96]. DYT-THAP1 is caused by mutations in THAP1 [97] and a transcriptomic analysis in neonatal mouse striatum and cerebellum indicated eIF2α signaling pathway dysregulation and the neuronal plasticity defects could be partially corrected by salubrinal, which inhibits eIF2α dephosphorylation [98]. DYT-SGCE is caused by mutations in ε-sarcoglycan (ε-SG), and PKR is upregulated in a DYT-SGCE mouse model [99]. Sporadic cervical dystonia patients have several mutations in ATF4, a downstream effector protein of the ISR response pathway [95]. Additionally, traumatic brain and spinal-cord injuries lead to injury-induced dystonia and activation of ISR and eIF2α signaling is noted in response to the injuries in animal models [100]. Finally, a growing list of dystonia genes are related to calcium physiology and may also have altered ISR and eIF2α signaling [101].

In view of the dysregulated eIF2α phosphorylation emerging as a convergent mechanism for several dystonia types, it is crucial to characterize the changes in eIF2α phosphorylation status and ultimately the regulation of ISR in each individual form of dystonia. Both increased as well as decreased eIF2α phosphorylation has been reported in various forms of monogenic dystonia. In case of DYT-PRKRA, there is a reduction in the basal eIF2α phosphorylation levels in Prkra ^lear-5J mice [94], which is in contrast to the increased phosphorylation of eIF2α, heightened PKR kinase activity and enhanced sensitivity to ER stress in DYT-PRKRA patient cells [42, 44]. Additionally, increase in PKR activity and eIF2α phosphorylation is reported in DYT-EIF2AK2 (DYT33) patients carrying PKR missense variants with early onset generalized dystonia [102]. For DYT-EIF2AK2 (DYT33), PKR inactivating mutations were reported in some patients [103], thereby suggesting that a reduction in PKR activity and consequently reduced eIF2α phosphorylation may also lead to dystonia pathophysiology. The most compelling evidence of reduced eIF2α phosphorylation in dystonia comes from studies on DYT-TOR1A (DYT1), where a genome-wide RNAi screen suggested a pathogenic role of deficient eIF2α signaling [95]. The HIV protease inhibitor ritonavir, which boosts eIF2α phosphorylation corrected the mutant TOR1A protein mislocalization in vitro and when administered during an early postnatal period, showed therapeutic effects in a mouse model of DYT-TOR1A, restoring brain abnormalities and ameliorating the dystonia phenotype [96]. Additionally, there is similar eIF2α pathway impairment in patients with sporadic cervical dystonia, due to rare inactivating mutations in ATF4 [95]. There are no current or past clinical trials for drugs targeting the eIF2α pathway to treat dystonia patients and in future, a few important points should be considered for conducting such trials. Although the studies on eIF2α and dystonia are encouraging for therapeutic interventions, such manipulations must be controlled carefully. Based on the available evidence, a precise regulation of the extent and duration of eIF2α phosphorylation may be essential for optimal neuronal regulation of motor control and either a reduction or elevation of the ISR response both could lead to lack of motor coordination. Thus, any future treatments that target eIF2α phosphorylation would need to be developed with caution keeping in mind not to overcorrect the underlying pathology using drugs to either boost or inhibit eIF2α phosphorylation throughout the body under all physiological scenarios. For example, it was recently reported that the cholinergic neurons constitutively engage the ISR for dopamine modulation and skill learning [104]. Such specific use of transient eIF2α phosphorylation to regulate neuronal functions will be disturbed by drugs globally targeting eIF2α pathway and thus can have detrimental off target effects.

Although phosphorylation of eIF2α has classically been viewed as a stress response, eIF2α phosphorylation mediated regulation of protein synthesis is utilized by neurons for mechanisms besides stress response that include behavior, memory consolidation, neuronal development, and motor control [105]. Future research using targeted mutations in specific neuronal subtypes to test the exact contribution of ISR and specifically eIF2α phosphorylation for neuronal control of muscle movement will be valuable.

In addition to the characterization of molecular pathways, it is also crucial to explore the specific regions of the brain affected by dystonia. Although dystonia is considered traditionally as a disorder of the basal ganglia [106], increasing evidence suggests that other brain areas may also play a role [107–112]. In this regard, mouse models could provide important clues to understand how alterations in the eIF2α signaling can affect neuronal function in specific regions of the brain to ultimately influence coordinated muscle movements. The dysfunction of cholinergic neurons which engage the eIF2α pathway for constitutive neuronal functionality [96] is one of the convergent mechanisms in dystonia etiology [113, 114]. Future studies can address the effects of manipulating the eIF2α pathway using several drugs currently available [98, 115–118]. It is possible to either measure physiologic dynamic changes in eIF2α phosphorylation or manipulate eIF2α signaling using genetic tools in a specific subset of neurons to understand how it influences muscle movement.

Given the functional role of PACT in the RNAi pathway, it would also be valuable to examine if there are any changes in miRNA profiles in DYT-PRKRA patient cells. Although it would be most meaningful to investigate the changes in miRNA expression profiles in induced pluripotent stem cell (iPSC) derived neurons from DYT-PRKRA patients, the miRNA profiles of patient lymphoblasts or fibroblasts can offer initial assessment if the DYT-PRKRA mutations can affect the miRNA biogenesis. Additionally, based on initial studies indicating the role of IFNs in DYT-PRKRA, it remains to be investigated if additional DYT-PRKRA mutations also enhance IFN production in response to dsRNA. Several DYT-PRKRA and DYT-EIF2AK2 patients developed dystonia symptoms after a febrile illness [5–9, 119], thus If DYT-PRKRA mutations lead to IFN production at higher levels or for a longer duration during viral infections, it can explain the neurologic regression and motor problems arising after a childhood illness. Based on such future studies the treatment for DYT-PRKRA can be significantly different based on the specific effects seen with various mutations, underscoring the urgency and importance of undertaking such basic mechanistic studies.