Tourette’s syndrome (TS) is a neurodevelopmental tic disorder occurring before the age of 18 and is characterized by motor and vocal tics that wax and wane over time (DSM-V). Tics in TS usually follow a remitting course, improving symptomatology in late adolescence or early adulthood. Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental disorder characterized by a persistent pattern of inattention and/or hyperactivity-impulsivity occurring before the age of 12 that interferes with overall functioning and/or development (DSM-V). In almost half of TS cases, the two pathologies co-occur in children (1). Comorbid ADHD commonly antecedes first tics and, during TS natural course was shown to be more impairing on patients’ quality of life then the tics themselves (2–5). This emphasizes the need for comprehensive in vivo approaches for studying TS, i.e., animal models able to resemble TS comorbidities together with the tic-like phenotype. A tic is a sudden, rapid, recurrent, non-rhythmic motor movement, or vocalization that can vary in frequency, intensity, duration, and anatomical localization. Tics are typically preceded by an uncomfortable phenomenon called “premonitory urge” and can be voluntarily suppressed by most patients for a short period of time (DSM-5). In patients, tics and stereotypies coexist and can be easily confused, leading to a delayed TS diagnosis or to a wrong treatment approach (6). In animal models, stereotypies belong to the spontaneous repetitive movement repertoire and their overexpression (hyper-grooming/-sniffing/-climbing) following genetic or pharmacological manipulation has been used as tic-like phenotype in modeling TS (7–10). A phenotype characterized by bousts of repeated and somatotopically organized movements have also been reached in different species applying bicuculline directly to the dorsolateral striatal region (11–14).

At the moment, a lack of homogeneity in the terms employed to identify motor phenomena in animal models of TS exists, mostly due to the variety of approaches used to achieve the ticcing phenotype.

In this study, we aim to overcome some of the actual TS model limitation choosing the striatal disinhibition paradigm applied first to a small cohort of Wistar Kyoto (WKY) rats and then to young spontaneous hypertensive rats (SHRs).

Among the existing variety of ADHD animal models, SHRs represent the most commonly used model showing all of the core disorder’s symptoms (inattention, impulsiveness, and aggressiveness) even though its predictive validity remains limited (15, 16). However, SHRs demonstrating tic-like movements as well as tic-expressing animals showing comorbid ADHD symptoms have not been documented so far. Validity for the single features were reported: hyperactivity was associated with specific bicuculline injections sites in the dorsal striatum and dorsal GPe of primates (12, 17) or in the ventral striatum of the adult rat (13); and attention deficit occurred after injections in associative regions of the GPe (17). In our model, the unilateral bicuculline injection within the anterior dorsolateral striatum resulted in a juvenile spontaneously hyperactive animal model that displayed transient tics, recapitulating the core-symptoms of both TS and comorbid ADHD. The phenotype was stable during the entire adolescence, a period of time that in the rat is placed between postnatal days (PNDs) 35 and 50 (18).

To test the predictive validity of our model, we tested aripiprazole, which was recently shown to effectively modulate the total tic score (45.9 and 54.2% with low and high dose, respectively) in a large multinational study (19). Due to its better-tolerated side effect profile, aripiprazole has been introduced as first-line treatment for tic management, overcoming the use of other typical antipsychotics.

In our model, aripiprazole dramatically reduced the number of tics and their latency immediately after the first administration and during consequent administrations. Over the same timespan hyperactivity, stereotypies and spontaneous USVs did not change their pattern following neither focal bicuculline injection (leading to tic generation) nor systemic aripiprazole application (leading to tic reduction), suggesting a unique neurological pathway for tic generation that is separate from those related to the other motor or vocal comorbid phenomena.

Finally, we used our prior established proton magnetic resonance spectroscopy (1H-MRS) protocol to investigate changes in the neurochemical composition of the bilateral striatum with the intention to identify a neurochemical signature, if present, of the “just occurred” bicuculline-induced tic as well as of the effective action of the aripiprazole treatment (20). Surprisingly, our results revealed an “unaltered striatum,” i.e., only few changes occurred following the tic session or of the effective drug treatment, raising questions on the role of the striatum within the supposed dysfunctional cortical–striatal–thalamic–cortico circuitry in TS.

Data from human imaging studies, as well as electrophysiological and postmortem anatomical studies (28–31) have contributed to our understanding of the pathophysiology of TS. Dataset obtained over the last decades from in vivo studies investigating the underlying pathology are heterogeneous and largely dependent on the ample variety of TS animal models available (5). In this study, we introduce a novel animal model that combines core features of TS with important comorbidity traits. The tic session was similar in the hyperactive/hypertensive and normotensive cohorts (Video S1 in Supplementary Material). Strain-related differences could be qualitatively noticed: in SHRs tic latency was reduced compared with WKY and tics appeared to reach faster a high intensity, giving to SHR tic session a “prone to tic” shape. SHRs are known to have a basal dopamine efflux within the striatum that is significantly higher than in long evans (LE) (32), Sprague–Dawley (SD) (33), and WKY (34). Dopamine is reported to directly inhibit GABA_A receptors that are immediately adjacent to dopamine release sites in the striatum (35). The existing higher basal efflux of dopamine in the SHR strain (32–34) may enhance the disinhibition exerted by bicuculline injection resulting in a stronger and slightly shorter tic session compared with the WKY, as we observed in our experiments. Tic latency and the total tic-session duration were shorter compared with the SD and LE adult model, but might arise due to the difference in age, strain, and sex.

Despite the invasive nature of the surgery and experiments, the procedures used to achieve the tic phenotype in our model did not affect the whole body and brain weight (Figure S1 in Supplementary Material) throughout the duration of the experiments. The procedures also did not alter the typical hyperactivity of the SHR strain (Figure S2 in Supplementary Material), maintaining the ADHD-comorbid characteristic. Bicuculline injection exerted tic movements but did not influence the frequency or latency of any other behavior within the stereotypic repertoire (Figures S3 and S4 in Supplementary Material). This establishes a clear-cut division between the neurological mechanism responsible for the tic generation and the one responsible for stereotypies.

Underpinning the circuit underlying such a result was not the focus of this study, but it connects with the call for standardization of the preclinical model for TS due to the still remaining confusion between these two motor phenomena (tics and stereotypies) often used in the tic-model literature with an interchangeable meaning (36). Tic-like movements in animal models are often confused with stereotypies, though they are separated clinical entities. Especially in young TS patients, however, comorbidities like ADHD are also commonly seen.

We tested aripiprazole acute treatment efficacy on our model. Aripiprazole reduced by threefold bicuculline induced tics immediately after the first administration and this effect persisted over consequent treatments (Figure 3A). Aripiprazole efficacy can also be appreciated through the reduced latency of tics (Figure 3B), even though its powerful effect at early timepoint is reduced toward the end of the adolescent period. Aripiprazole significantly ameliorated tics, but did not exert any notable effect on the frequency or latency of repetitive behaviors or on the numbers of USVs, as well as on the hyperactivity. This evidence strengthens our hypothesis that these three features, even though all existing in TS clinical presentation, at least in rats belong to different neuronal circuits. In a non-human primate proposed-model for TS, distinct but inter-digitated neuronal circuits of the basal ganglia were associated with different symptoms. In particular, tic-like movements were associated with the medial and lateral premotor cortex and sensorimotor parts of the basal ganglia; hyperactivity was associated with the prefrontal dorsolateral cortex; and associative territories of the basal ganglia and stereotyped behaviors were linked to the orbitofrontal cortex and limbic part of the basal ganglia (17, 37, 38).

Our results showed for the first time that repeated bicuculline injections, despite its invasiveness, did not compromise the ability of our rats to emit the whole range of spontaneous USVs (Figure 4B). When comparing the number of calls pre- and post-bicuculline injection, our results did not demonstrate statistically significant changes in USVs. There is evidence that nucleus accumbens would have been an alternative site of injection putative for evoking vocal tics as demonstrated in non-human primates (39).

We also performed for the first time in vivo 1H-MRS longitudinally in the rat model of striatal disinhibition with the aim of detecting neurochemical changes within the striatum that could be connected with the observed bicuculline-induced phenotype. Unexpectedly, comparing both hemispheres within each cohort, no significant differences were detected among the pool of investigated metabolites. When differentially analyzing the same brain hemisphere among the different cohorts, only PE resulted in a significant increase after saline injection and its levels dramatically dropped after each bicuculline injection. The same phenomenon, even though to a smaller extent, was observed in the contralateral intact side of the brain. PE content in the mammalian brain is 45% of the total phospholipids and serves multiple important cellular functions. It is a precursor of anandamide, the ligand for cannabinoid receptors in the brain (40). The dramatically decreased level of PE observed in our study after bicuculline injection could actively contribute to the tic phonotype. A study using an indirect cannabinoid agonist that enhances endogenous anandamide levels was conducted on another proposed animal model of TS, the DOI-induced head twitch responses mice in four different strains resulting in ameliorating of the tics without suppressing other behaviors (41). Among the variety of TS etiological theories available (31, 42), a pathophysiological role for cannabinoids in TS was also hypothesized following the observation of patients who reported that prior use of marijuana led to a reduction or complete remission of motor and vocal tics as well as of other comorbid symptoms (43).

On the neurochemical level, we could not detect any significant change within the analyzed metabolites as a readout of the aripiprazole treatment. In our previous work, we were able to detect even minimal changes of the neurochemicals in the intact brain of WKY and SHR under subchronic aripiprazole treatment condition using the same dosage and administration route (20). A potential hypothesis is that the effect of the injection procedure itself produced a stressful impact significantly bigger than the one exerted by the administered neuroactive compounds (bicuculline, aripiprazole). Thus, if a neurochemical effect is present, it probably lies below the detectable cutoff of our established spectroscopy protocol. More importantly, an interesting explanation for the “unaltered striatum” we revealed, despite an obvious bicuculline-induced phenotype and its modulation using aripiprazole, could be that the striatum itself is able per se to cope with local chemical/mechanical alterations. This would result in a protection of its neurometabolites homeostasis, maintaining a preserved functional output to the cortex that, at least in our model, is hypothesized to be dysfunctional but that needs to be further investigated. A recent support for this direction has been given by a study in which TS participants showed distributed differences in the activation of multiple cortical regions with differences in functional reactivity and that may account for heterogeneity in the symptomatic expression of TS and its comorbidities (44).

The lack of standardization procedures and the absence of a rigorous evaluation tool led us to stick to the traditional idea of testing and discussing the three aspects of model validity: construct (the rationale behind the model matches the pathological hypothesis), face (the model shows symptoms similar to the patients’ ones), and predictive (the model and patients response to the same treatment is similar) (45–47). In terms of construct validity, the rat model of striatal disinhibition achieved via bicuculline injection fits Mink’s tic generation hypothesis. The loss of inhibition within a specific neuronal population of the striatum results in the inability to inhibit a movement that should be prevented (48). Such a loss is experimentally induced in the rat by the GABA_A antagonist bicuculline intrastriatal injection that evokes a motor phenotype characterized by bouts of movements that fluctuate in frequency and duration over time, resulting in a tic session comparable with some extents to those experienced by TS patients (13). Construct and face validity, both reasonably achieved in this model, convinced us to use this paradigm for testing aripiprazole efficacy in tic management (predictive validity).

This study is the first of its kind and potentially a starting point for future proliferation of non-invasive in vivo spectroscopic brain investigations using our 1H-MRS protocol and the newly proposed animal model for TS and comorbid ADHD. We are aware that the main limitation of this study, the “acute” nature of the tic session, has been recently overcome by a chronic tic model achieved so far only in the adult female LE rats (49). Our belief is that such a method would perfectly fit our TS-ADHD model at young age and would be feasible to our 1H-MRS protocol. Focusing on different regions of interest, we are optimistic that such a study-design in the future would reveal the neurochemical signature of a tic and/or its pharmacological treatment, adding striking new information to prior TS pathophysiological hypothesis (49–52).

To the best of our knowledge, this is the first animal model able to provide high degrees of construct, face, and predictive validity as a model for TS and comorbid ADHD. We believe that our model together with our 1H-MRS protocol represents a valid tool for TS pathophysiological investigation as well as for testing novel treatment approaches.

All the experiments were approved by the Committee for Animal Experimentation of the University of Ulm and the regional administrative authority in Tübingen (TVA No. 1200).

FR was primarily involved in the writing of the ethical allowance and execution of the research project, as well as the design and execution of the statistical analysis and the writing of the manuscript. EN contribution was fundamental in the establishment of the surgery and behavioral protocols. DD and AA established the 1H-MRS in vivo protocol, AA was responsible for the spectra quality control and data screening. IBG expertise was essential for the research project rationale, experimental design and the critical review of the manuscript from the preclinical point of view. VR and JF critically reviewed the manuscript contributing to the clinical importance of MRI in psychiatric diseases and to the bench-to-bed importance of this study respectively. BH supported the study and helped in the experimental drug testing design. TMB conceived, supported and organized the research project and critically contributed to the writing and reviewing process of the manuscript.

During the last 5 years Professor JF received research funding from the EU, DFG (German Research Foundation), BMG (Federal Ministry of Health), BMBF (Federal Ministry of Education and Research), BMFSFJ (Federal Ministry of Family, Senior Citizens, Women and Youth), German Armed Forces, several state ministries of social affairs, State Foundation Baden-Württemberg, Volkswagen Foundation, European Academy, Pontifical Gregorian University, RAZ, CJD, Caritas, and Diocese of Rottenburg-Stuttgart. Moreover, he received travel grants, honoraria, and sponsoring for conferences and medical educational purposes from DFG, AACAP, NIMH/NIH, EU, Pro Helvetia, Shire, several universities, professional associations, political foundations, and German federal and state ministries. Clinical trials for Lundbeck, BMBF, and Servier. Professor Fegert holds no stocks of pharmaceutical companies. The remaining authors declare no conflict of interest.