Loss-of-function of GNAL dystonia gene impairs striatal dopamine receptors-mediated adenylyl cyclase/ cyclic AMP signaling pathway

Loss-of-function mutations in the GNAL gene are responsible for DYT-GNAL dystonia. However, how GNAL mutations contribute to synaptic dysfunction is still unclear. The GNAL gene encodes the G α olf protein, an isoform of stimulatory G α s enriched in the striatum, with a key role in the regulation of cAMP signaling. Here, we used a combined biochemical and electrophysiological approach to study GPCR-mediated AC-cAMP cascade in the striatum of the heterozygous GNAL (GNAL + / (cid:0) ) rat model. We first analyzed adenosine type 2 (A2AR)


Introduction
DYT-GNAL is an autosomal-dominant form of focal dystonia typically characterized by cranio-cervical distribution and adult onset.It is caused by loss-of-function mutations of the GNAL gene [guanine nucleotide binding protein (G protein), alpha activating activity polypeptide, olfactory type], on chromosome 18 (Fuchs et al., 2013), encoding Gα olf , an isoform of stimulatory G s protein.The G Protein Coupled Receptors (GPCRs) transduce their signals by activating heterotrimeric G proteins, that promote GTP binding to Gα subunit and release of Gβγ subunits, regulating several signaling complexes (Pierce et al., 2002).Gα olf is highly expressed both in olfactory bulb and striatum, where it couples to adenosine A2A (A2AR) and dopamine D1 (D1R) receptor, to activate adenylyl cyclase type 5 (AC5) and the cAMP cascade (Corvol et al., 2001).Therefore, a crucial role of Gα olf is the regulation of cAMP, a ubiquitous second messenger critical for many fundamental neuronal mechanisms, including synaptic plasticity and neuronal excitability (Devasani and Yao, 2022;Kheirbek et al., 2009).Specifically, Gα olf mediates dopamine signaling in direct pathway striatal projection neurons (dSPNs) by coupling to D1R, and adenosine signaling in indirect pathway SPNs (iSPNs), by coupling to A2AR (Goodchild et al., 2013).In addition, Gα olf is co-expressed with G s in striatal cholinergic interneurons (ChIs), where it contributes to both dopaminergic and purinergic signaling (Hervé et al., 2001).Pioneering studies have shown that complete deletion of Gα olf in mice results in hyperkinetic movements, indicating its involvement in motor function (Belluscio et al., 1998;Corvol et al., 2001;Xu et al., 2000;Zhuang et al., 2000).More recently, different DYT-GNAL rodent models have been generated (Pelosi et al., 2017;Yu-Taeger et al., 2020).In the heterozygous GNAL knockout (GNAL +/− ) mice no overt dystonic phenotype was observed in control conditions.However, these mice displayed abnormal movements following striatal administration of either cholinergic or dopamine D1R agonists indicating an important role of the cholinergic and dopaminergic system in DYT-GNAL dystonia (Belluscio et al., 1998;Corvol et al., 2001Corvol et al., , 2007;;Pelosi et al., 2017).Of relevance, >15 different mutations (nonsense, missense and frameshift) associated with DYT-GNAL have been identified, mostly affecting the isoform 2 of the gene (Fuchs et al., 2013).Here, we used the heterozygous GNAL +/− rat model of DYT-GNAL dystonia, characterized by a deletion of 13 base pairs in the first exon of isoform 2, the most expressed in the human and rodent striatum (Yu-Taeger et al., 2020;Vemula et al., 2013).In the GNAL +/− rat model an approximately 50% reduction of Gα olf expression is observed, together with an early-onset motor impairment (Yu-Taeger et al., 2020).Of interest, in GNAL +/− rats we previously reported an increased cell surface expression of AMPA receptors, coupled to an impairment of D2R-dependent corticostriatal synaptic plasticity, that was recovered by antagonizing the A2AR.This rescue effect was likely due to the A2AR-mediated antagonistic actions on striatal D2R, suggesting a possible involvement of D2R in the synaptic dysfunction observed in GNAL +/− rats.Of note, an impaired D2R signaling has also been reported in the striatum of dystonic patients (Carbon et al., 2009;Asanuma et al., 2005), as well as in several different animal models (Napolitano et al., 2010;Bonsi et al., 2019).Here, we investigated how the mutation causing DYT-GNAL dystonia affects the GPCR-cAMP pathway and striatal dopaminemediated transmission.In more detail, we focused on a specific neuronal cell type, the striatal cholinergic interneurons, which represent one of the main actors in the pathophysiology of different form of dystonia (Eskow Jaunarjas et al., 2015).

Animals
All experiments were approved by the ethics committee of IRCCS Fondazione Santa Lucia and authorized by the Italian Ministry of Health (authorizations nr.353/2020 and 240/2023).The procedures were carried out in accordance with Italian law (D.Lgs 26/2014) and the European Union Directive 2010/63/EU.All rats used on the Sprague-Dawley background were generated as described (Yu-Taeger et al., 2020).Mutant heterozygous GNAL +/− and their control littermates GNAL +/+ rats were maintained in our animal house with a 12/12 h light/dark cycle and constant temperature (22 • C), with food and water ad libitum.

cAMP ELISA-assay
For assessment of total cAMP level, after anesthesia, the sample was frozen in liquid nitrogen for 3 s, the brain was quickly removed and the dorsal striatal tissues collected in cold dry ice.The levels of cAMP were then determined using cAMP ELISA kit (ADI-900-066, ENZO Life Sciences direct).Samples were homogenized in 0.1 M HCl, at a ratio of 1:10 (w/v).The cAMP levels were then measured by diluting samples 1:50 (v/ v) in 0.1 M HCl and processed following the acetylated protocol in the manufacturer's guidelines.

Data collection and statistical analysis
Patch-clamp data were analyzed by Clampfit 10.2 (Molecular Devices) and Prism 5.0 (Graph-Pad, SanDiego, CA, USA).Normality tests were used to assess Gaussian distribution.Statistical significance was evaluated using a two-tailed unpaired or paired Student's test (t-test).For the forskolin dose-response curve comparison, we used Extra sumof-squares F, the confidence interval was 95%.The data are reported as mean ± SEM, and P was set at <0.05 *, <0.01 **, <0.001 ***."N" refers to the number of animals, whereas "n" indicates the value of independent observations.

GNAL loss-of-function affects the A2AR/D2R-AC-cAMP signaling
GNAL encodes for the Gα olf subunit of the heterotrimeric G proteins, a striatal protein with a key role in the production of cAMP, by directly coupling both A2AR and D1R to AC5.Therefore, in order to address the potential consequences of GNAL loss-of-function mutations on receptor signaling, we first measured striatal levels of A2AR and D1R.Our WB analysis revealed a significant increase in the A2AR protein level (GNAL +/− : 143% of control p < 0.05, Fig. 1A) in mutant GNAL +/− rats, whereas D1R receptor level was unchanged between genotypes (Fig. 1B).Then, in order to investigate the expression profile of other effectors along the cAMP-PKA pathway (Fig. 1C), we measured the expression level of AC5, the predominant AC striatal isoform directly stimulated by Gα olf (Matsuoka et al., 1997;Iwamoto et al., 2003;Kim et al., 2014).Of note, we found a significant reduction of AC5 levels in 30 μg of lysate was loaded.The graph shows the quantitative analysis of AC5 levels normalized to GNAL +/+ .The amount of AC5 was quantified relatively to β-actin.

Impaired D2R signaling in GNAL +/− rats
The GPCR signaling is finely modulated by the regulator of the Gprotein signaling (RGS) family (Lambert et al., 2010).In particular, RGS9-2 is specifically expressed in the striatum, where it regulates D2R function accelerating the termination of Gi (Anderson et al., 2010;Celver et al., 2010).Indeed, RGS9-2 binds the type 5 G-protein β (Gβ5), a member of the G protein beta subunit family, forming a complex that directly interacts with D2R, increasing its stability in the cell membrane (Celver et al., 2012;Kovoor et al., 2005).Based on this well-documented and direct relationship between D2R, RGS9-2 and Gβ5 proteins, we examined whether these D2R interactors were affected, as occurs in experimental models of DYT-TOR1A dystonia (Bonsi et al., 2019).In GNAL +/− rats, we found a significant reduction of both RGS9-2 and Gβ5 protein levels (Fig. 2 A, B).Then, to further elucidate the molecular mechanism of the D2R downregulation in GNAL +/− rats, we focused on the striatal level of the scaffolding proteins β-arrestin 2 and spinophilin.In particular, β-arrestin 2 is involved in the internalization and G protein-independent signaling of D2R, while spinophilin interacts with D2R and antagonizes the arrestin proteins (Del'guidice et al., 2011;Beaulieu and Beaulieu and Gainetdinov, 2011;Wang et al., 2004).Consistently, we found spinophilin level significantly decreased in GNAL +/− rats, whereas β-arrestin 2 was increased (Fig. 2 C, D).Taken together, our data demonstrate an unprecedented deficit of D2R signaling pathway in DYT-GNAL dystonia.

GNAL loss-of-function affects the cholinergic interneuron electrophysiological responses to D2R activation
Dopamine exerts a powerful control on striatal cholinergic transmission through D2R.Indeed, D2R activation reduces the firing discharge of cholinergic interneurons (ChIs), a small neuronal population with a key role in the control of striatal functions.An aberrant D2R function was found in ChIs from different rodent model of dystonia (Pisani et al., 2006;Sciamanna et al., 2011Sciamanna et al., , 2014;;Scarduzio et al., 2017;Bonsi et al., 2019;Eskow Jaunarjas et al., 2019;Caffall et al., 2021).Firstly, we investigated whether GNAL knockout affects the basal membrane properties of ChIs.The recorded neurons were identified by their typical morphology and peculiar electrophysiological properties (Fig. 3 A), including a depolarized resting membrane potential (RMP), a high input resistance (IR), a small rheobase value, and spontaneous firing activity (Fig. 3 B-E).No significant differences were observed between genotypes.Then, we evaluated the activation of D2R by two pharmacologically different agonists: quinpirole (3 μM, 2 min) and  bromocriptine (1 μM, 2 min), the latter showing a biased non-canonical signaling through β-arrestin recruitment biased (Brust et al., 2015).In GNAL +/+ ChIs the firing rate was reduced by bath-application of both D2R agonists, as expected (Pisani et al., 2000) (Fig. 4 A, B).Conversely, in GNAL +/− ChIs, D2R activation by both agonists failed to induce a significant effect on the firing frequency (Fig. 4 A, B).Furthermore, in order to test whether the altered responses to D2R activation exhibited a regional specificity, we recorded dopaminergic neurons from substantia nigra pars compacta, where Gα olf is also expressed (Hervé et al., 1993).Our recordings show that the dopaminergic neurons from both genotypes exhibit similar membrane properties, with a tonic firing activity and a prominent sag in response to hyperpolarizing current pulses (Fig. S1 A).Likewise, the responses of nigral neurons to D2R autoreceptor activation were comparable between genotypes.Specifically, bath-application of quinpirole (300 nM, 1 min) induced a typical membrane hyperpolarization, followed by a brief pause response of the firing activity in both genotypes (Fig. S1 B).In addition, we measured nigral levels of D2R protein, which resulted similar in the two genotypes (Fig. S1 C).

Discussion
Pre-clinical and clinical evidence supports the involvement of alterations in dopaminergic neurotransmission in different forms of dystonia, confirming the critical role of dopamine in the pathogenesis of this disabling disease (Goodchild et al., 2013;Ribot et al., 2019;Mencacci et al., 2020).To date, the discovery of Gα olf loss-of-function mutations linked to DYT-GNAL dystonia, points to dopamine signaling as potential pharmacological targets (Yu-Taeger et al., 2020).However, the mechanisms by which mutant Gα olf leads to specific pathogenic alterations in DYT-GNAL patients are poorly defined.Here, we used the heterozygous GNAL knockout rats (GNAL +/− ), characterized by a deletion of 13 base pairs in the first exon of the main isoform 2, resulting in an early stop and truncated protein.In this model, the mutant rats showed an earlyonset impairment in dopamine signaling and synaptic plasticity, providing a tool for a better understanding of the disease mechanisms and a valuable model to recapitulate the loss-of-function mutations causing DYT-GNAL dystonia (Yu-Taeger et al., 2020).Dopamine and adenosine modulate the AC5/cAMP pathway, which in turn represents a critical player in the control of the neurotransmission balance within the striatum (Corvol et al., 2001).
Our work shows that Gα olf loss-of-function mainly affects the dopaminergic system leading to a downregulation of D2R signaling.Our findings of a reduced AC5 striatal level in mutant rats are in line with previous work on GNAL +/− mice showing that the stability of AC5 in the striatum depends on Gα olf levels (Xie et al., 2015).Consistent with the reduction of striatal AC5 levels, activation of D1-like receptor failed to induce a significant increase of firing activity in GNAL +/− ChIs, differently from what typically occurs in wildtype neurons (Aosaki et al., 1998).In addition, we observed a shift in the dose-response curve of forskolin, indicating a reduced sensitivity to AC stimulation, downstream Gα olf .Overall, our data provide the first evidence that Gα olf lossof-function affects the AC-dependent cascade in a specific neuronal subtype.
Surprisingly, we did not observe any change in the basal level of either cAMP or in the main striatal target of the cAMP-PKA pathway, p (Th34) DARPP-32, in the whole striatum of the mutant rats compared to control littermates.Our data are consistent with previous work showing unchanged cAMP basal levels (Melis et al., 2021), although conflicting evidence emerged (Corvol et al., 2007).A possible explanation could be found in the variability of cAMP levels in basal conditions, compared to a stimulated response, which might be able to disclose an impairment of cAMP signaling.Although we found no difference in the D1R protein level, in agreement with previous work on GNAL +/− mice (Corvol et al., 2007), we previously observed that the impairment of Gα olf reduces the D1R-mediated AMPA receptor internalization, supporting an impairment of D1R function (Yu-Taeger et al., 2020).In GNAL +/− rats we found an increased striatal level of A2ARs coupled with a downregulation of D2Rs, which negatively modulate cAMP and selectively colocalize with A2ARs both in iSPNs and ChIs (Ferré et al., 1997;Ongini and Fredholm, 1996).Therefore, the increased A2AR expression may represent an adaptive compensation for the impairment of Gα olfdependent AC5 function.
Overall, it should also be noted that the D2R/A2AR signaling imbalance may affect other neurotransmitter systems in the GNAL +/− striatum.In particular, D2R activation induces the production of endocannabinoids, which then act presynaptically to modulate synaptic activity (Mathur and Lovinger, 2012).Furthermore, its physiological inhibitory effect on the tonic firing activity of striatal ChIs reduces acetylcholine release, thus reducing postsynaptic M1 muscarinic receptor activation and allowing the expression of long-term synaptic depression at corticostriatal synapses (Mathur and Lovinger, 2012).In our work, D2R activation in GNAL +/− ChIs did not produce the paradoxical effect observed in GNAL +/− mice and in models of different forms of dystonia (Eskow Jaunarjas et al., 2019;Bonsi et al., 2019;Sciamanna et al., 2011Sciamanna et al., , 2014;;Pisani et al., 2006), but rather failed to induce the inhibitory effect observed in GNAL +/+ ChIs.This inconsistency may be explained by the different species (mice vs rats) and/or the time of application of the D2R agonist quinpirole (Eskow Jaunarjas et al., 2019).Of note, we observed a D2R impairment selectively in the striatum.Indeed, in nigral neurons we found unaltered D2R expression levels, as well as a physiological response of dopaminergic neurons to autoreceptor activation.The regional difference may be explained, at least in part, by a lower expression of Gα olf in the substantia nigra (Hervé, 2011) and/or by a possible compensatory effect of Gα s .

Conclusions
Overall, our findings support the evidence of a significant alteration of cAMP-dependent striatal dopamine transmission in DYT-GNAL dystonia.Notably, we have demonstrated that in GNAL +/− rats the D2R downregulation causes an impairment of receptor function, and signaling, highlighting the relevance of D2R pathway in dystonia.

Declaration of competing interest
The authors report no conflicts of interest.
I. El Atiallah et al.