Insulin action in the brain regulates both central and peripheral functions

Abstract

The brain has been traditionally thought to be insensitive to insulin, primarily because insulin does not stimulate glucose uptake/metabolism in the brain (as it does in classic insulin-sensitive tissues such as muscle, liver, and fat). However, over the past 20 years, research in this field has identified unique actions of insulin in the brain. There is accumulating evidence that insulin crosses into the brain and regulates central nervous system functions such as feeding, depression, and cognitive behavior. In addition, insulin acts in the brain to regulate systemic functions such as hepatic glucose production, lipolysis, lipogenesis, reproductive competence, and the sympathoadrenal response to hypoglycemia. Decrements in brain insulin action (or brain insulin resistance) can be observed in obesity, type 2 diabetes (T2DM), aging, and Alzheimer’s disease (AD), indicating a possible link between metabolic and cognitive health. Here, we describe recent findings on the pleiotropic actions of insulin in the brain and highlight the precise sites, specific neuronal population, and roles for supportive astrocytic cells through which insulin acts in the brain. In addition, we also discuss how boosting brain insulin action could be a therapeutic option for people at an increased risk of developing metabolic and cognitive diseases such as AD and T2DM. Overall, this perspective article serves to highlight some of these key scientific findings, identify unresolved issues, and indicate future directions of research in this field that would serve to improve the lives of people with metabolic and cognitive dysfunctions.

HOW INSULIN ENTERS THE BRAIN

After being secreted from pancreatic β-cells, insulin crosses into the brain by either, 1) circumventing the blood-brain barrier (BBB) crossing the median eminence, or 2) crossing the vascular endothelium via transport proteins (1–3). Transport of insulin into the brain is highly regulated and may be altered by states such as obesity, diabetes mellitus, fasting, and Alzheimer’s disease (AD) (1). The mechanism(s) of altered insulin transport by these disease states is incompletely understood, but in diabetes, BBB disruption may be mediated by altered insulin transport proteins, pericytes loss, and altered cerebral microvessel expression of tight junction proteins (2, 4, 5). Insulin actions on the cerebral blood flow is region specific and may also be altered in insulin-resistant states (6–10). Regional-specific actions of insulin in the brain may be based on sites of increased insulin receptor expression within the olfactory bulb, hypothalamus, hippocampus, cerebral cortex, and cerebellum (11, 12). Further specificity of brain insulin actions are likely dependent on the regional distribution of insulin receptor isoforms and cell-type differences in receptor density (i.e., neurons vs. glia) (13).

Early studies in rodents (and subsequent human studies) identified insulin mRNA in the central nervous system (CNS), thus hinting at the possibility that a small amount of insulin may be synthesized, de novo, within the brain (14–20). Physiological roles for localized insulin synthesized in the brain have been identified (21–23). Recently, Lee et al. (21) detected insulin mRNA in the paraventricular nucleus and identified its role in the regulation of pituitary growth hormone production.

To specifically study insulin action in the brain, animal studies often make use of (albeit nonphysiological) insulin infusion into the brain or targeted genetic manipulations of the insulin receptor, whereas human studies make use of intranasal insulin dosing to selectively increase brain insulin levels. The pros and cons of these techniques have been reported (and include issues with possible supratherapeutic dosing, regional-specific effects, and issues of systemic spillage); however, the consistency of results using different techniques and across species strengthen the body of work demonstrating insulin action in the brain (24–30).

BRAIN INSULIN ACTION REGULATES CNS FUNCTIONS

Neuronal Development and Neuronal Survival

Insulin action is critically important in the developing nervous system, directing differentiation, proliferation, and neurite growth (31–33). Insulin acts via its receptor to stimulate brain growth, as noted by increased expression and activity in developing brain tissues, both within neurons and surrounding glial cells (34, 35). Insulin also acts as a neuroprotector, preventing damage induced by ischemia, β-amyloid toxicity, oxidative stress, and apoptosis (36–39). Thus, brain insulin action may play a role in preventing the development of dementia. Conversely, these protective actions of insulin may be diminished in the setting of impaired brain insulin signaling (40), thereby increasing the risk for both neuropsychiatric and neurogenerative disorders (40–42), and thereby linking brain insulin resistance (i.e., lack of insulin action in brain) with cognitive dysfunction and depression.

Insulin Effects on Anxiety and Depression

Studies in animals support the concept that insulin action in the brain can regulate anxiety and depression. The knockdown of insulin receptors in the hypothalamus in animals has shown to trigger depressive and anxiety-like behaviors (43). Similarly, neuronal-specific insulin receptor knockout (NIRKO) mice demonstrated depression-like behaviors (42). Indeed, the knockdown of insulin receptors, specifically in astrocytes, also results in increased anxiety and depressive-like behavior in mice, via decreasing dopamine release (44). Consistent with these rodent findings, postmortem brain tissue of patients with mood and psychotic disorders demonstrated a lower expression of insulin receptor-related signaling genes, again associating decreased insulin action with a downregulation of dopaminergic signaling (45). Clinically, intranasal administration of insulin (which specifically increases brain insulin levels) has been shown to be beneficial in blunting stress in healthy subjects (46, 47). Collectively, these studies support the notion that insulin action in the brain may be an important component of mood regulation and neuropsychiatric disorders.

Insulin Action in the Brain Regulates Cognition/Alzheimer’s Disease

Evidence from human and animal studies indicate that decreased insulin action in the brain (i.e., brain insulin resistance) and/or deficiency of brain insulin is a pathological feature of metabolic and cognitive dysfunctions, including obesity, type 2 diabetes (T2DM), and Alzheimer’s disease (AD) (48, 49). Specifically, the impairment in brain insulin signaling has shown to increase hyperphosphorylation of tau protein (50) and β-amyloid (Aβ) accumulation (51), which are the neuropathological hallmarks of AD. People with T2DM are at a higher risk of developing AD and cognitive decline (52, 53). Brain insulin resistance has been documented in patients with AD (54) and the term “type 3 diabetes” has been used to identify this brain type of T2DM (55). Given the beneficial action of insulin in the brain, intranasal insulin administration has been shown to provide cognitive benefits in healthy humans as well as in patients with T2DM and AD (56–62).

Insulin Action on Brain Cholesterol Synthesis

Cholesterol is an important constituent of cell membrane, which plays an important role in cellular physiology and signaling (63, 64). Changes in brain cholesterol metabolism have been implicated in diabetes (65) and AD (66), suggesting that altered brain cholesterol metabolism is another possible link between diabetes and AD. The impaired brain cholesterol biosynthesis that occurs in diabetes can be restored with increased insulin action in the brain (65). The action of insulin on brain cholesterol metabolism is not only restricted to neurons but also occur in glial cells, especially astrocytes, and plays an important role in brain development and behavioral functions (65, 67).

Body Weight and Food Intake

Of all insulin’s actions in the brain, insulin’s action in regulating body weight and feeding behavior has been most convincingly established (68). Insulin action in the brain to reduce food intake and body weight have been linked to insulin’s ability to alter the expression of critical body weight regulating neuropeptides [e.g., neuropeptide Y, agouti-related peptide, pro-opiomelanocortin (POMC), and amphetamine-regulated transcript] in the arcuate nucleus of the hypothalamus (69–72).

Following intranasal administration, insulin has been shown to decrease appetite, food palatability, and reduce food intake and body weight (73, 74). Imaging studies have shown that intranasal insulin administration in the presence of food-related images leads to a reduction of activity in areas of the brain involved in memory and object processing, suggesting that insulin reduces the wanting of food (70).

Although insulin action in the brain has been shown to reduce body weight, treatment with insulin often leads to weight gain (75), indicating that when delivered peripherally, insulin’s anabolic actions on peripheral tissues (e.g., adipocytes) are dominant.

BRAIN INSULIN ACTION REGULATES PERIPHERAL FUNCTIONS

Hepatic Glucose Production

Insulin acts to lower blood glucose levels by its suppressive effects on hepatic glucose production (HGP). In addition to direct effects on the liver, there is increasing evidence that insulin acts in the brain to indirectly suppress HGP, as recently reviewed (76, 77).

Studies in rodents consistently demonstrated that insulin acts in the brain to suppress HGP (27, 78) via several mechanisms including action on pro-opiomelanocortin neurons (79), agouti-related protein neurons (80), and hepatic vagal efferents (81). In contrast to rodent studies, an elegant series of studies in dogs demonstrate that the insulin actions to acutely suppress HGP are attributable to insulin’s direct effect on the liver and that indirect effects of insulin are redundant (28, 29, 82, 83). Thus, there is ongoing debate regarding the physiological relevance of brain insulin action in regulating HGP (84, 85). In humans, some (74, 86) but not all (87, 88) studies demonstrated an effect of intranasal insulin to augment suppression of HGP, although early studies were possibly confounded by spillover of insulin into the systemic circulation. More recent studies that controlled for insulin spillover did demonstrate an effect of intranasal insulin to suppress HGP (24, 25). Particularly interesting was the evidence for brain insulin resistance in obese and type 2 diabetic subjects, as these subjects failed to demonstrate an effect of increased brain insulin in suppressing HGP, which was noted in healthy controls (25, 87). These findings support the concept that the brain can sense insulin and alter hepatic glucose production, but whether this represents a physiological role of insulin remains to be established.

Lipolysis and Lipogenesis

In rodents, intracerebroventricular infusion of insulin induces lipogenesis (89), and insulin infusion into the mediobasal hypothalamus suppresses systemic lipolysis by dampening of sympathetic nervous system outflow to white adipose tissue (WAT) (90). This suppressive effect on lipolysis may be mediated by insulin action on POMC neurons (91). Consistent with this effect of brain insulin action, NIRKO mice lacking the neuronal insulin receptor showed increased WAT lipolysis and disrupted lipogenesis (90). Corroborating these animal studies, intranasal insulin administration in humans was shown to suppress systemic lipolysis and reduce levels of circulating free fatty acids (92).

Reproductive Competence

In addition to regulating glucose homeostasis and adiposity, insulin action in the brain regulates reproductive competence. Studies in NIRKO mice have demonstrated that the lack of insulin action in the brain leads to hypothalamic luteinizing hormone (LH) dysregulation that results in reduced fertility due to decreased spermatogenesis in males and impaired ovarian follicle maturation in females (93). Consistent with these observations, intracerebroventricular insulin infusion in rats with diabetes has been shown to restore and normalize LH surges that are necessary for ovulation (94). Another recent study in mice demonstrated that impaired insulin signaling in astrocytes dramatically reduces adult reproductive competence due to dysfunction of the hypothalamic-pituitary-gonadotropin axis (95). Thus, independent of neuronal insulin signaling, insulin signaling in astrocytes appears to play an important role in regulating reproductive competence.

Role of Insulin in Mediating the Counterregulatory Response to Hypoglycemia and CNS Glucose Sensing

Many animal and human studies (96–99), but not all (100–102), have shown that insulin acts acutely in the brain to augment the counterregulatory response to hypoglycemia. Consistent with the positive effect of central insulin action in augmenting counterregulation, NIRKO mice have impaired CNS glucose sensing and a blunted sympathoadrenal response to insulin-induced hypoglycemia (99). This blunted CNS glucose sensing may be mediated by insulin action on neuronal glucose transporter 4 (99, 103, 104) or astrocyte insulin signaling (105, 106). Thus, in addition to acute actions of insulin in the brain to augment the counterregulatory response to hypoglycemia, insulin also acts chronically in the brain to regulate CNS glucose sensing, and thus, the counterregulatory response to hypoglycemia. To support this notion, we recently demonstrated that in rats with diabetes, insulin deficiency impairs the sympathoadrenal response to hypoglycemia and that chronic insulin infusion into the ventromedial hypothalamus alone is sufficient to normalize the sympathoadrenal response to hypoglycemia (107). These studies suggest that insulin action, particularly in hypothalamic region, is required for brain glucose sensing. Thus, insulin acts in a paradoxical fashion by both causing hypoglycemia via dominant glucose lowering actions on peripheral tissues (e.g., liver, muscle, and fat), yet also acts to protect against hypoglycemia by augmenting the brain’s ability to detect and mount a counterregulatory response.

PERSPECTIVE AND FUTURE DIRECTIONS

Scientists no longer question whether insulin acts in the brain but now seek to delineate specific actions of insulin in the brain (Fig. 1, Table 1). Identifying mechanisms of insulin transport into the brain and specific areas of insulin action will help to enhance our understanding in the field. Future novel transgenic models will continue to advance the field by precisely identifying sites (i.e., neuronal and/or astrocytic) and mechanisms by which insulin exerts its actions in the brain. Though studies in humans have been limited, targeting brain insulin action via intranasal insulin delivery has validated earlier preclinical findings regarding brain insulin action. Unfortunately, the role of insulin action in the brain has been less well characterized under pathophysiological settings in humans (i.e., insulin-resistant T2DM). Since clinical studies are often confounded by comorbid conditions, a clearer link between brain insulin resistance and cognitive dysfunction/AD is needed. Another unanswered question is whether interventions known to increase peripheral insulin sensitivity (exercise, weight loss, and insulin-sensitizing drugs) act similarly in the brain to reap the benefits of improved CNS insulin signaling. In addition to further studies with intranasal insulin delivery, future efforts to create insulin analogs that preferentially target brain action are needed. Future strategies to exploit the multiple unique and beneficial actions of insulin in the brain seem warranted.

Figure 1.

In addition to regulating CNS (via neurons and/or astrocytes) functions (such as cognition, depression, and food intake), insulin acts in the brain to regulate peripheral functions via the autonomic nervous system (ANS) and the hypothalamic pituitary axis (HPA). Specifically, insulin acts in the prefrontal cortex and hippocampus to improve cognitive function and reduce depressive symptoms. Insulin acts in the hypothalamic nuclei to decrease food intake and reduce body weight. Via the efferent autonomic nervous system to target organs, insulin acts in the brain to decrease hepatic glucose production, increase lipogenesis, decrease lipolysis, and increase the sympathoadrenal response to hypoglycemia. Insulin acts via the hypothalamic-pituitary-gonadal axis to improve reproductive competence. CNS, central nervous system.

Table 1.

Contrasting effects of insulin action on peripheral organs versus action in the central nervous system

Function	Peripheral Action	Central Action
Metabolism	Anabolic (75, 108)	Catabolic (109)
Muscle glucose uptake	↑ (110,111)	–
Lipolysis	↓ (112)	↓ (90–92)
Lipogenesis	↑ (113)	↑ (89, 90)
Hypoglycemia	↑ (114)	↓*(97–99, 107)
Hepatic glucose production	↓(82)	↓ (24, 25, 27, 74, 76–79, 84–86, 115, 116)
Reproductive competence	↑ (117)	↑ (93–95, 118)
Neuroprotective	–	↑ (36–39, 119, 120)
Cognition	–	↑ (56–59, 61, 73, 121)
Depression	–	↓ (42–46, 122)
Food intake	↑ (123, 124)	↓ (70, 71, 73, 74, 125)

In addition to insulin’s direct effects on classic insulin sensitive tissues (muscle, fat, and liver), insulin acts in the brain to regulate peripheral tissues indirectly (via the autonomic nervous system and the hypothalamic pituitary axis). Some of insulin’s actions in the brain parallel its peripheral actions, but, perhaps acting in a homeostatic fashion, some of insulin’s action in the brain oppose its peripheral action.

*Insulin acts centrally to augment the counterregulatory responses, thus acting to reduce hypoglycemia.

GRANTS

This work was supported by funding from the National Institutes of Health Grants NS070235 and DK118082 (to S.J.F), by Juvenile Diabetes Research Foundation Grant 1-FAC-2020-984-A-N (to C.M.R.), and by the University of Utah’s Diabetes and Metabolism Research Center.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

R.A. conceived and designed research; R.A. and S.J.F. prepared figures; R.A., C.M.R., S.S., Y.H., and S.J.F. drafted manuscript; R.A., C.M.R., S.S., C.C., Y.H., and S.J.F. edited and revised manuscript; R.A., C.M.R., S.S., C.C., Y.H., and S.J.F. approved final version of manuscript.