Advance of age is associated with insulin resistance and glucose intolerance (Barbieri et al. 2001). The causes of age-related insulin resistance include changes in body composition (increase in fat mass and decrease in lean mass), decline in physical activity, and increase in oxidative stress in the elderly. Moreover, abnormal endoplasmic reticulum (ER) stress response is promoted during aging (Brown and Naidoo 2012). ER stress response serves as an adaptive mechanism to cope with protein misfolding, but chronic ER stress causes deterioration of cellular functions, such as insulin resistance (Yalcin and Hotamisligil 2013). Excessive ER stress during aging is at least partly due to sleep disturbance, because quality of sleep declines along with aging, and because ER stress is augmented by disruption of normal sleep (Brown and Naidoo 2012). In addition, aging causes ER stress in orexin-producing neurons, leading to instability of wakefulness in mice (Naidoo et al. 2011). Orexin release and receptor activation in the brain also decrease during aging (Zink et al. 2014). Thus, a vicious cycle between perturbation of orexin system, sleep disturbance, and ER stress could promote the development of insulin resistance with aging. It should be noted, however, that orexin expression in the hypothalamus mainly declines during maturation (from adolescence to adulthood), compared to normal aging (from adulthood to old age) (Zink et al. 2014; Hunt et al. 2015). Moreover, daily rhythm of CSF orexin levels were largely maintained in aged rodents, compared to young controls (Desarnaud et al. 2004). Therefore, we consider that the endogenous orexin still contributes to various circadian regulations even in the elderly.

To know the significance of orexin on maintenance of glucose homeostasis during aging, we investigated age-related changes in glucose metabolism, using Orexin ^−/− mice at 3, 6, and 9 months old (Tsuneki et al. 2008, 2015). Impairment of insulin sensitivity occurred in the hypothalamus at 3 months old, in the liver at 6 months old, and in the skeletal muscle at 9 months old in male Orexin ^−/− mice, whereas there were no changes in the body weights. Consistently, hepatic glucose production was abnormally increased in Orexin ^−/− mice at 6 and 9 months old, whereas systemic insulin resistance and glucose intolerance occurred at 9 months old, when compared to age-matched wild-type mice. These indicate that orexin deficiency preferentially causes hepatic insulin resistance, leading to promotion of systemic insulin resistance. Thus, endogenous orexin action is required for preventing insulin resistance with aging, especially in the liver.

We further explored the mechanism underlying severe insulin resistance in the liver of Orexin ^−/− mice along with aging. In the liver, ER stress response serves to support metabolic reprogramming during transition from fasting to feeding state, in which protein synthesis is intensively reduced to adapt to metabolic changes (Deng et al. 2013). In general, these stress responses are rapidly terminated at postprandial state; however, when ER homeostasis is disrupted under the deleterious conditions, such as continuous nutrient overload, ER stress response is prolonged and causes insulin resistance and type 2 diabetes (Hotamisligil 2010; Lee and Ozcan 2014). In our study, Orexin ^−/− mice showed the disruption of daily blood glucose oscillation via hepatic glucose production at 3 months old, and thereafter at 6 months old, they exhibited abnormal ER stress responses in the liver upon refeeding after 24-h fasting, including an increase in phosphorylation levels of IRE1α and JNK, a major ER stress-related signaling. Thus, maintaining daily glucose rhythm by orexin appears to be beneficial for preventing ER dyshomeostasis. It has been reported that ‘resting time’ for ER is required for successful recovery from the state of ER stress (Eizirik et al. 2008). In fact, we observed that hepatic glucose production is daily suppressed at resting period under the control of endogenous orexin. This suppression may provide the resting time for ER, thereby preventing hepatic insulin resistance (Fig. 1).

It is well known that energy expenditure is reduced with aging, resulting in increase in body weight and fat volume. These abnormalities are considerably due to age-related impairment of thermogenesis in brown adipose tissue (Sellayah and Sikder 2014). Orexin promotes thermogenic function of brown adipose tissue through central pathways, including the rostral raphe pallidus-spinal intermediolateral nuclei-sympathetic nerve route (Morrison et al. 2014; Perez-Leighton et al. 2014). In addition, peripheral orexin action can directly promote the differentiation from committed brown preadipocyte to mature brown adipocyte (Sellayah et al. 2011). In fact, the cold-induced and diet-induced thermogenesis was impaired in Orexin ^−/− mice (Sellayah et al. 2011; Sellayah and Sikder 2012). Orexin injection daily for 14 days in aged mice (24-month-old male C57BL6 mice) increased the expression levels of uncoupling protein-1, a main regulator of thermogenesis in brown adipose tissue, and improved cold tolerance. Interestingly, glucose tolerance was also improved under this condition (Sellayah et al. 2011). Therefore, orexin is a crucial factor for preventing age-dependent deterioration of glucose and energy homeostasis.

Increased sympathetic activity predisposes obesity and metabolic syndrome (Lambert et al. 2010; Licht et al. 2013). Type 2 diabetic db/db mice showed a higher sympathetic tone with a reduced parasympathetic tone (Goncalves et al. 2009) and disrupted circadian rhythms of heart rate and locomotor activity (Su et al. 2008). In addition, type 2 diabetic subjects exhibited increased resting sympathetic nerve activity and blunted sympathetic response to oral glucose loading (Straznicky et al. 2012). Importantly, either hepatic sympathetic or parasympathetic denervation abolished daily rhythm of blood glucose, while no such effect was produced by complete denervation of both branches (Cailotto et al. 2008). Thus, the unbalanced autonomic nervous system in obese and diabetic states appears to deteriorate daily glucose homeodynamics. Moreover, since orexin-producing neurons belong to the glucose-inhibited neurons in the hypothalamus, the activity of orexin neurons and the expression levels of orexin were down-regulated by hyperglycemia in obese and diabetic animals, such as ob/ob and db/db mice (Yamamoto et al. 1999; Yamanaka et al. 2003). These raise a question whether or not orexin could induce the improvement of glucose metabolism impaired in obese and type 2 diabetic states. The answer is yes, because CAG/orexin transgenic mice on a high fat diet showed improvement of insulin sensitivity by an OX2R-dependent mechanism (Funato et al. 2009), whereas female Orexin ^−/− mice fed on a high fat diet exhibited remarkable obesity, glucose intolerance, and insulin resistance (Tsuneki et al. 2008). We further examined the effect of daily intracerebroventricular administration of orexin A on blood glucose in db/db mice (Tsuneki et al. 2015). Interestingly, nighttime administration of orexin A for 3 days, mimicking a physiological orexin secretion pattern, amplified daily glucose oscillation, leading to gradual reduction of blood glucose levels at daytime resting period. The glucose lowering effect in db/db mice was much greater than that in control mice. In contrast, neither daytime administration nor continuous administration of orexin A showed limited impacts on blood glucose levels in db/db mice. The mechanism underlying the glucose-lowering effect of orexin A administered at nighttime active period can be explained by suppression of hepatic glucose production at the subsequent daytime resting period in db/db mice. Therefore, we thought that the amplification of daily endogenous orexin action by the timely supplementation of orexin agonist effectively reconstitutes normal daily rhythm of hepatic glucose production, thereby ameliorating metabolic disorders related to obesity and type 2 diabetes (Fig. 1).

What about the timely inhibition of orexin system by orexin antagonist? Currently, an orexin receptor antagonist, suvorexant, is available for clinical use to treat insomnia; however the effect of this drug on glucose metabolism remains unknown. An early study reported that the other orexin receptor antagonist SB-334867-A exerted anti-obese and anti-diabetic effects when repetitively injected into genetically obese ob/ob mice at daytime for 2 weeks (Haynes et al. 2002). The mechanisms behind these beneficial effects have been explained by reduced adiposity and increased energy expenditure through the improvement of brown adipose tissue thermogenesis. We also assume, from the viewpoint of circadian rhythm, that administration of SB-334867-A at resting state may strengthen normal sleep/wake and feeding cycles, which can improve glucose metabolism in the liver and skeletal muscle.

Although high-fat feeding promotes the development of obesity and insulin resistance, time-restricted feeding strongly prevented the metabolic disorders at least through improvement of circadian clock gene expressions in the liver of high-fat-fed C57BL/6 mice (Hatori et al. 2012; Sherman et al. 2012) and db/db mice (Kudo et al. 2004). It should be reminded that feeding cycle is a robust entrainer of the daily orexinergic activation rhythm, as mentioned above. These raise a possibility that time-restricted feeding is a non-pharmacological regimen to alleviate obesity and type 2 diabetes via the activation of hypothalamic orexin system.

Taken together, impairment of daily orexin action under hyperglycemic conditions is an exacerbating factor for obesity and type 2 diabetes. Therefore, chronopharmacological approach to the treatment of such metabolic diseases, using orexin agonist and/or antagonist, appears to be promising as a new therapeutic strategy.

Several meta-analyses indicate that major depressive disorders are associated with insulin resistance and type 2 diabetes (Mezuk et al. 2008; Kan et al. 2013). This notion has been supported by animal studies. For instance, high fat diet-induced obese mice exhibited depressive behavior (Yamada et al. 2011). Acute psychological stress by inescapable foot shock rapidly caused impaired glucose tolerance and hepatic insulin resistance in mice (Li et al. 2013). Chronic social defeat stress, a mouse model of major depression, caused insulin resistance and lipid dysregulation in mice (Chuang et al. 2010). Thus, it appears that there is a synergistic relationship between major depression and metabolic disorders, including insulin resistance.

Most of the orexin-elicited behavioral changes are triggered by increased motivation under various physiological and psychological conditions, including hunger and reward-associated stimuli (Mahler et al. 2014; Sakurai 2014). The orexin levels in the human amygdala increased during positive emotion, social interaction and anger in addition to arousal (Blouin et al. 2013). Although many clinical observations demonstrate the implication of dysfunctional orexinergic activity in depression, the pathophysiological relationship between them is complicated and remains unclear (Nollet and Leman 2013). In fact, a behavioral test using OX1R ^−/− and OX2R ^−/− mice indicated that OX1R and OX2R signalings produced pro-depressant and anti-depressant effects, respectively (Scott et al. 2011), and SB-334867, a selective OX1R antagonist, exerted depressant or anti-depressant effects in mice, depending on the experimental conditions (Scott et al. 2011; Deats et al. 2014). Moreover, both hypoactivity and hyperactivity of orexin system have been reported to predispose depression (Nollet and Leman 2013). Therefore, there may be optimal levels of orexinergic activation for alleviating depression.

We asked whether or not central orexin system has some role in the regulation of glucose metabolism in depressive state. To address this question, we employed Orexin ^−/− mice subjected to chronic social defeat stress, and compared the efficacy of glucose metabolism with wild-type controls that received the same psychological stress (Tsuneki et al. 2013). The chronic social defeat stress caused depression-like behavior in both wild-type and Orexin ^−/− mice, suggesting that orexin deficiency alone does not affect the onset of depression. Instead, the defeat stress caused hyperinsulinemia without changing fasting blood glucose levels in Orexin ^−/− mice, and as a result, the HOMA-IR (homeostasis model assessment for insulin resistance) was increased. In pyruvate tolerance test to evaluate hepatic gluconeogenic activity reflecting hepatic insulin sensitivity, the defeat stress caused excessive glucose elevation in Orexin ^−/− mice compared to wild-type mice. Consistently, hepatic insulin signaling was severely impaired in Orexin ^−/− mice after the defeat stress exposure. These results indicate that endogenous orexin action is required for maintaining hepatic insulin sensitivity under prolonged stressful condition. Therefore, we propose that it is hypothalamic orexin system that prevents the development of type 2 diabetes under depressive condition.