Ayumu Inutsuka

Orexin neurons regulate physiological functions, such as energy homeostasis, wakefulness, and motivated behaviors. However, studies linking orexin neuron activity to behavior via selectively activating/inactivating these inputs in a temporally controlled manner in rats are scarce. Here, we examined the role that orexin neurons play in motivated behavior in transgenic rats using cell type–specific fiber photometry and optogenetic manipulation. Using chemogenetics, we found that motivation for a reward increased when orexin neurons were activated. Furthermore, during motivated behavior, orexin neuron activity changed dynamically: Activity increased during reward prediction and decreased after reward receipt. When an unexpected event occurred (i.e., not obtaining an expected reward), increased orexin activity was sustained. Notably, orexin activity strengthened with increasing effort. Optogenetic inhibition of orexin neuron activation during reward prediction and treatment with an orexin 1-receptor antagonist reduced reward-seeking behavior. Therefore, orexin is crucial for linking predicted expectations with motivated behavior. Moreover, optimization of orexin activity is necessary to overcome difficulties during motivated behaviors.

Orexin neurons regulate physiological functions, including not only energy homeostasis and wakefulness but also motivated behaviors. These neurons play important roles in linking metabolic requirements to such behaviors. However, it is unknown whether these neurons are involved in reward-based choice behavior. We recently developed a transgenic rat line that expresses Cre recombinase exclusively in orexin neurons. Here, we examined the roles of orexin neurons in reward choice behavior under conditions of uncertainty in a gambling test in rats and analyzed the changes in choice behavior when orexin neurons were manipulated using chemogenetic approaches. In the gambling test, chemogenetic activation of orexin neurons resulted in risky arm choice. Positive, but not negative, reward-prediction error contributed to reward-based risky choice when orexin neurons were activated. These risky choice behaviors were canceled by suvorexant, an orexin receptor antagonist. In addition, suvorexant administered alone led to conservative choice behavior. The computational approach revealed that orexin/orexin neurons modulate the learning process and motivational value of rewards. Thus, our data suggest that orexin/orexin neurons affect the processing of reward-prediction error and alter strategy in reward-based decision-making.

The cerebellum plays an important role in cognitive and social functioning. Childhood damage in the cerebellum increases the risk of autism spectrum disorder. Cerebellar inflammation induces social avoidance in mice. Oxytocin regulates social relationship and expression pattern of the oxytocin receptor in the brain is related to social behaviors. However, the expression patterns of the oxytocin receptor in the cerebellum remain controversial. Here, we report that the expression patterns of the oxytocin receptor in the cerebellum are highly variable among knock-in transgenic lines. We used Oxtr-Cre knock-in mice combined with a fluorescent reporter line and found that oxytocin receptor expression in Bergmann glia was more variable than that in Purkinje cells. We found that physical damage with inflammation induced the selective upregulation of the oxytocin receptor in Bergmann glia. Our findings indicate high variability in oxytocin receptor expression in the cerebellum and suggest that the oxytocin receptor can affect neural processing in pathological conditions, such as inflammation.

Transgenic animals expressing fluorescent proteins are widely used to label specific cells and proteins. By using a split Cre recombinase fused with mCherry-binding nanobodies or designed ankyrin repeat proteins, we created Cre recombinase dependent on red fluorescent protein (RFP) (Cre-DOR). Functional binding units for monomeric RFPs are different from those for polymeric RFPs. We confirmed selective target RFP-dependent gene expression in the mouse cerebral cortex using stereotaxic injection of adeno-associated virus vectors. In estrogen receptor-beta (Esr2)-mRFP1 mice and gastrin-releasing peptide receptor (Grpr)-mRFP1 rats, we confirmed that Cre-DOR can be used for selective tracing of the neural projection from RFP-expressing specific neurons. Cellular localization of RFPs affects recombination efficiency of Cre-DOR, and light and chemical-induced nuclear translocation of an RFP-fused protein can modulate Cre-DOR efficiency. Our results provide a method for manipulating gene expression in specific cells expressing RFPs and expand the repertory of nanobody-based genetic tools.

Transgenic animals expressing fluorescent proteins are widely used to label specific cells and proteins. GFP-dependent gene regulation utilizes these lines to manipulate gene expression; however, its application has been limited to fluorescent proteins derived from Aequorea jellyfish. By using a split Cre recombinase fused with mCherry-binding nanobodies or designed ankyrin repeat proteins, we created Cre recombinase dependent on red fluorescent protein (RFP) (Cre-DOR). Functional binding units for monomeric RFPs (mCherry, mRFP1) are different from those for dimeric RFP (tdTomato). We confirmed target RFP-dependent gene expression in the mouse cerebral cortex using stereotaxic injection of adeno-associated virus vectors including Cre-DOR, target RFP, and reporter GFP vector. We found highly selective GFP expression in RFP-positive cortical neurons with 93.5 ± 0.6% of GFP-positive cells being mRFP1-positive. In estrogen receptor-beta (Esr2)-mRFP1 mice, we confirmed that Cre-DOR can be used for selective expression of membrane-bound GFP in the paraventricular nucleus of the hypothalamus. The neural projection from Esr2-expressing neurons in the hypothalamic paraventricular nucleus to the posterior pituitary was visualized by Cre-DOR. In gastrin-releasing peptide receptor (Grpr)-mRFP1 rats, we similarly achieved anterograde tracing of Grpr-expressing neurons in the medial amygdala and found that they are projecting axons to the posterior bed nucleus of the stria terminalis. Cellular localization of RFPs affects recombination efficiency of Cre-DOR, and light and chemical-induced nuclear translocation of an RFP-fused protein can increase or decrease Cre-DOR efficiency. Our results provide a method for manipulating gene expression in specific cells expressing RFPs and expand the repertory of nanobody-based genetic tools.

Orexin neurons regulate physiological functions, including not only energy homeostasis and wakefulness, but also motivated behaviors. These neurons seem to play important roles in linking metabolic need to motivated behaviors via the dopaminergic system. Recently, we developed a rat model that expresses the Cre recombinase specifically in orexin neurons. Moreover, we established the gambling test for assessing reward motivation and decision-making under conditions of uncertainty, as well as a touch-screen system for assessing reward sensitivity and craving. Here, we examined the roles of orexin neurons in reward motivation and decision-making when orexin neurons are manipulated using pharmacogenetics approaches. In the gambling test for rodents, cell-specific excitatory manipulation of orexin neurons of rats using the DREADD technology resulted in risky arm-choice. Positive, but not negative, reward prediction error may contribute to reward-based risky choice when orexin neurons are activated. The motivational values of a large reward was increased when orexin neurons were activated by the DREADD technology. In the probability reversal learning test, cell-specific excitatory manipulation of orexin neurons resulted in low performance in rats. These results suggest that activated orexin neurons affect motivational processes, and may alter strategy in reward-based choice behavior, potentially leading to an action that fails to yield rewards and detect changes in reinforcement contingencies.

Uninterrupted arousal is important for survival during threatening situations. Activation of orexin/hypocretin neurons is implicated in sustained arousal. However, orexin neurons produce and release orexin as well as several co-transmitters including dynorphin and glutamate. To disambiguate orexin-dependent and -independent physiological functions of orexin neurons, we generated a novel Orexin-flippase (Flp) knock-in mouse line. Crossing with Flp-reporter or Cre-expressing mice showed gene expression exclusively in orexin neurons. Histological studies confirmed that orexin was knock-out in homozygous mice. Orexin neurons without orexin showed altered electrophysiological properties, as well as received decreased glutamatergic inputs. Selective chemogenetic activation revealed that both orexin and co-transmitters functioned to increase wakefulness, however, orexin was indispensable to promote sustained arousal. Surprisingly, such activation increased the total time spent in cataplexy. Taken together, orexin is essential to maintain basic membrane properties and input-output computation of orexin neurons, as well as to exert awake-sustaining aptitude of orexin neurons.

Diet affects health through ingested calories and macronutrients, and macronutrient balance affects health span. The mechanisms regulating macronutrient-based diet choices are poorly understood. Previous studies had shown that NAD-dependent deacetylase sirtuin-1 (SIRT1) in part influences the health-promoting effects of caloric restriction by boosting fat use in peripheral tissues. Here, we show that neuronal SIRT1 shifts diet choice from sucrose to fat in mice, matching the peripheral metabolic shift. SIRT1-mediated suppression of simple sugar preference requires oxytocin signalling, and SIRT1 in oxytocin neurons drives this effect. The hepatokine FGF21 acts as an endocrine signal to oxytocin neurons, promoting neuronal activation and Oxt transcription and suppressing the simple sugar preference. SIRT1 promotes FGF21 signalling in oxytocin neurons and stimulates Oxt transcription through NRF2. Thus, neuronal SIRT1 contributes to the homeostatic regulation of macronutrient-based diet selection in mice.