Jens C Brüning - C 6

Variations in the human FTO gene have been linked to obesity, altered connectivity and function of the dopaminergic neurocircuitry. We report that FTO in D2 medium spiny neurons (MSNs) of mice regulates the excitability of these cells in vitroand in vivo. Lack of FTO in D2 MSNs translates into increased locomotor activity to novelty in particular, without impairing the ability to control actions or affecting reward-driven and conditioned behaviour.

Introduction

The fat mass and obesity-associated protein (FTO)is the first RNA demethylase identified, but the physiological role of FTO-dependent methyl-adenosine regulation remains only partially defined. In mice, ubiquitous inactivation of Fto reduces adiposity and protects from diet-induced obesity, and overexpression of FTO increases fat and body mass. However, whether FTO polymorphisms affect obesity development through altered regulation of FTO or other genes is not fully resolved. FTO polymorphisms are also associated with traits that are not restricted to overweight. Along these lines, Fto-deficient mice exhibit altered spontaneous and cocaine-induced locomotor activity, which is accompanied by impaired activation of the ventral tegmental area-substantia nigra (VTA-SN), and the striatum, brain areas critically involved in motor planning, decision-making, reward perception and judgment of time. Moreover, deletion of Fto in dopamine (DA) midbrain neurons impairs dopamine-2/3 (D2/3R) receptor-mediated inhibition of DA-signalling, resulting in enhanced sensitivity to the locomotor- and reward-stimulatory actions of cocaine. 

Consistent with a role for FTO to regulate dopaminergic signalling, human FTO variants have been linked to changes in reward sensitivity, impulsivity, food reinforcement, preference for energy-dense food and augmented cortical activation in response to visual food cues. Collectively, all these data point toward a regulatory role for FTO in the integrated control of several behaviours governed by dopaminergic signalling. Nevertheless, while prior studies revealed that FTO controls D2/D3R-controlled auto-inhibition of DA neurons and that specific FTO genotypes alter midbrain responses to rewards, it is unclear whether FTO regulates the activity of basal ganglia neurons extending beyond the midbrain. Therefore, in the current study we investigated a potential role of FTO in controlling dopaminoceptive medium spiny neurons (MSNs).

FTO in D2R-expressing neurons controls locomotor behaviour

To assess the potential physiological role of FTO through its action in neurons of the direct or indirect pathway, we generated mice lacking Fto selectively in D1 or D2 MSNs, i.e. Fto^∆D1aor Fto^∆D2mice. Given FTO’s role in regulating energy homeostasis, we assessed the metabolic phenotype of mice lacking FTO in D1 or D2 MSNs and found unaltered food intake, energy expenditure and body weight compared to control mice. 

In light of the MSNs’ well-defined role in motor control, we also investigated diurnal activity patterns and found no alteration in circadian locomotor activity in both knockout lines in a domestic environment. To provoke ambulatory movement, we next examined the acute locomotor response to a novel environment. While Fto^∆D1amice covered the same distance as their corresponding controls, Fto^∆D2mice travelled substantially longer and were faster than wild-type littermates (Fig. 1). Further analyses revealed that Fto^∆D2mice move faster and cover larger areas, have reduced stand, longer stride length, and higher swing speed. Combined, these results demonstrate that Fto^∆D2mice exhibited hyperlocomotion acutely when exposed to a novel environment, whereas Fto^∆D1amice show no motor changes. 

Role of FTO in D2-expressing neurons for novelty-induced locomotion

To further investigate the hypothesis that the increase in activity was linked to novelty, we examined an array of behaviours during novel object recognition tests. While our results so far show that FTO’s actions in neurons harbouring D2Rs is involved in controlling explorative behaviour, they do not distinguish between pre- versus post-synaptic effects. Therefore, to differentiate between these two classes of D2R neurons and since FTO modulates D2R-like dependent control of DAT-neuron activity, we complemented the analysis with mice in which we ablated Fto in dopamine neurons (Fto^∆DAT), i.e. neurons pre-synaptic to MSNs. Our data show that Fto^∆D2mice had a higher preference for the novel object over the familiar, and to a larger extent than controls. Collectively, these results demonstrate that the behavior emerging from inactivating Ftoin D2 MSNs is induced upon exposure to novelty. These observations also indicate that the critical nexus underlying the observed behavior is post-synaptic D2 MSNs, as Fto^∆DATmice - lacking FTO pre-synaptically - failed to replicate the same behavior as Fto^∆D2mutants - lacking FTO pre- and post-synaptically.

FTO in D2 MSNs regulates cell excitability

In light of these results, we next examined the electrical properties of D2 MSNs ex vivoin order to gain insights into the consequences of removing Fto from D2 MSNs at the cellular level. We used genetically marked D2 MSNs by expressing tdTomato in D2R-expressing neurons in a Cre-dependent manner. Electrophysiological recordings revealed that the resting membrane potentials and the action potential (AP) thresholds for Fto^∆D2MSNs were unaltered compared to Fto^wtMSNs. Furthermore, several AP parameters were unaltered by the lack of FTO in these neurons. However, when provoking APs with step and current injection pulses, pronounced changes between genotypes were observed in the voltage response, and in the current-voltage and current-frequency relationships of Fto^∆D2and Fto^wtMSNs. Strikingly, less current was needed to depolarize the membrane and evoke an AP of Fto^∆D2neurons compared to their wild-type counterparts, which is consistent with increased excitability (Fig. 2A). 

To examine whether these differences in excitability would accompany a similar change in neuronal activity in vivo and to concurrently define D2 MSN activity dynamics during behaviour, we next performed fiber-photometry-assisted Ca2⁺imaging in awake mice based on viral- and Cre-induced expression of the calcium indicator GCaMP6s. We found that exposure to a novel environment led to a stronger increase in fluorescence intensity ratios in Fto^Ko-GCaMP6s+compared to Fto^Wt-GCaMP6s+mice, thus demonstrating augmented D2 MSN activation in the intact brain of mice lacking FTO in D2 MSNs upon novelty (Fig. 2B).