Skip to main content
NCBI home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2019 Feb 14;116(13):6379–6384. doi: 10.1073/pnas.1817391116

Insulin signaling in the hippocampus and amygdala regulates metabolism and neurobehavior

Marion Soto a,b,1, Weikang Cai a,b,1, Masahiro Konishi a,b, C Ronald Kahn a,b,2
PMCID: PMC6442573  PMID: 30765523

Significance

Loss of insulin receptors in the brain causes metabolic and behavioral abnormalities whereas loss of IGF-1 receptors in the brain leads to a developmental defect in the brain and periphery. However, less is known about the impact of brain insulin and IGF-1 receptor (IR/IGF1R) loss in adult mice, especially in higher neural processing regions. Here, we show that loss of IR/IGF1R in the hippocampus and central amygdala of adult mice results in decrease in glutamate receptors, accompanied by glucose intolerance, anxiety-like behavior, and impaired cognition. In addition, we identify an insulin/IGF-1 signaling-dependent neural circuit originating from the central amygdala, which regulates interscapular brown fat activity and thermogenesis. Thus, brain insulin/IGF-1 signaling is important for higher neural processing and systemic metabolism.

Keywords: insulin, hippocampus, amygdala, metabolism, cognition

Abstract

Previous studies have shown that insulin and IGF-1 signaling in the brain, especially the hypothalamus, is important for regulation of systemic metabolism. Here, we develop mice in which we have specifically inactivated both insulin receptors (IRs) and IGF-1 receptors (IGF1Rs) in the hippocampus (Hippo-DKO) or central amygdala (CeA-DKO) by stereotaxic delivery of AAV-Cre into IRlox/lox/IGF1Rlox/lox mice. Consequently, both Hippo-DKO and CeA-DKO mice have decreased levels of the GluA1 subunit of glutamate AMPA receptor and display increased anxiety-like behavior, impaired cognition, and metabolic abnormalities, including glucose intolerance. Hippo-DKO mice also display abnormal spatial learning and memory whereas CeA-DKO mice have impaired cold-induced thermogenesis. Thus, insulin/IGF-1 signaling has common roles in the hippocampus and central amygdala, affecting synaptic function, systemic glucose homeostasis, behavior, and cognition. In addition, in the hippocampus, insulin/IGF-1 signaling is important for spatial learning and memory whereas insulin/IGF-1 signaling in the central amygdala controls thermogenesis via regulation of neural circuits innervating interscapular brown adipose tissue.

Insulin, insulin-like growth factor 1 (IGF-1), and IGF-2 act through two cognate receptors to control distinct physiological processes throughout the body. Insulin binds with highest affinity to the insulin receptor (IR) and low affinity to the IGF-1 receptor (IGF1R) and thus regulates systemic metabolism (1, 2). In contrast, IGF-1 and IGF-2 bind with higher affinity to the IGF1R and act as growth factors important for tissue development and remodeling (3–5). Differences between insulin and IGF-1/2 are in part explained by the different expression patterns of these receptors, and in part by differences in signals generated by the IR and IGF1R (6).

In the central nervous system, both IR and IGF1R are widely expressed, and their actions have been implicated in the brain in control of metabolism and energy homeostasis, as well as brain development, injury repair, and higher neural processes, including cognition and mood (7, 8). At the disease level, altered insulin/IGF-1 signaling in the brain has been linked to increased risks for Alzheimer’s disease, premature cognitive decline, and dementia, as well as depression and anxiety (9–11). Patients with Alzheimer’s disease show decreased expression of both IR and IGF1R (12), as well as abnormal distribution and cellular localization of these receptors (13). These CNS phenotypes have also been observed in states in which there is insulin resistance in the brain, such as obesity and diabetes (14, 15).

Loss of IR in the whole brain causes hyperphagia, insulin resistance, central hypogonadism, impaired response to hypoglycemia, and increased depressive-like behaviors (16–18) whereas loss of IGF1R (or one of its major substrates, IRS-2) has been shown to impair brain development (19, 20). However, single IR or IGF1R knockout models do not completely eliminate insulin or IGF-1 signaling since both ligands can elicit signaling through the receptor that remains intact. Furthermore, up to now, most studies have focused on either the whole brain or hypothalamus (16, 19, 21, 22). Hence, the roles of IR/IGF1R in other nuclei controlling higher neural functions, including mood and cognition, are not known.

In the present study, we used stereotactic surgery and AAV-Cre to induce IR and IGF1R double knockout (DKO) in the hippocampus and central amygdala. We found that IR/IGF1R deletion specifically down-regulates the expression of an AMPA receptor subunit, glutamate receptor 1, in synaptosomes from both hippocampus and amygdala. This is accompanied by multiple metabolic and behavioral abnormalities, including glucose intolerance and increased anxiety-like behaviors. In addition, while deletion of both IR and IGF1R in the hippocampus and central amygdala leads to impaired recognition memory, IR/IGF1R loss in hippocampus also results in impaired spatial memory. Finally, deletion of IR and IGF1R in the central amygdala impairs cold-induced thermogenesis.

Results

Deletion of IR and IGF1R in the Hippocampus and Central Amygdala.

We used bilateral stereotaxic injection to deliver AAV encoding a Cre-GFP fusion protein into the hippocampus and central amygdala of IRlox/lox/IGF1Rlox/lox mice to delete both IR and IGF1R in these nuclei in the brain (Fig. 1 A and C). AAV encoding GFP alone was injected to generate control groups. The positions of injection and extent of coverage were confirmed by GFP expression (Fig. 1A). Western blotting of total tissue lysates taken 8 wk later revealed about 50% reduction in IR and IGF1R protein in the hippocampus from AAV-Cre-GFP–injected mice (Hippo-DKO) compared with AAV-GFP–injected control mice (Hippo-CTR) (Fig. 1B). A similar decrease was observed in synaptosomes isolated from the hippocampus (SI Appendix, Fig. S1A). Likewise, there was an ∼50% decrease in protein levels of IR and IGF1R in the central amygdala of CeA-DKO mice compared with CeA-CTR mice (Fig. 1D) and synaptosomes (SI Appendix, Fig. S1B). Thus, using stereotaxic approaches, we were able to significantly reduce IR and IGF1R in the hippocampus and central amygdala of adult mice.

Fig. 1.

Fig. 1.

Open in a new tab

AAV-Cre mediates efficient gene recombination in the hippocampus (Hippo-DKO) and central amygdala (CeA-DKO) of IR/IGF1Rlox/lox mice. (A) Bilateral AAV injection sites of the anterior and posterior hippocampus. Representative images of immunohistochemical staining of GFP and DAPI in the anterior and posterior hippocampus. (B) Immunoblotting of IR and IGF1R in the hippocampus of adult IR/IGF1Rlox/lox mice injected with AAV-GFP or AAV-Cre-GFP. Bottom, densitometry analysis. (C) Bilateral AAV injection sites of the central amygdala. Representative images of immunohistochemical staining of GFP and DAPI in the central amygdala. (D) Immunoblotting of IR and IGF1R in the central amygdala of adult IR/IGF1Rlox/lox mice injected with AAV-GFP or AAV-Cre-GFP. Bottom, densitometry analysis. *P < 0.05 by unpaired t test, n = 6. Data are presented as mean ± SEM.

Decreased Expression of Glutamate Receptor 1 in Synaptosomes of Hippo-DKO and CeA-DKO Mice.

Both insulin signaling and IGF-1 signaling modulate synaptic plasticity, especially in excitatory glutamatergic neurons (23, 24). Glutamate acts via AMPA and NMDA receptors, both of which are ionotropic transmembrane cation channels that mediate fast synaptic transmission (25, 26). Both Hippo-DKO and CeA-DKO mice displayed 60 to 70% reductions of the GluA1 (also known as GluR1) subunit of the AMPA receptor in the isolated synaptosomes from these regions, compared with their controls, whereas the expression of the other major AMPA receptor, subunit GluA2, was not affected by IR/IGF1R loss (Fig. 2). These occurred with no change in the phosphorylation of either GluA1 or GluA2 (SI Appendix, Fig. S1 C and D), suggesting that insulin/IGF-1 signaling was not important for the phosphorylation of the AMPA receptor subunits but did affect protein levels of GluA1. The NMDA receptor subunits NR2A and NR2B, on the other hand, were not affected by IR/IGF1R deletion (SI Appendix, Fig. S1 C and D). Expression of the synaptic marker PSD95 was similar in both the hippocampus and amygdala between control and DKO mice (SI Appendix, Fig. S1 E and F), indicating that the defect of GluA1 expression in the DKO mice was not likely due to the impairment of the structural integrity of the synapsis in these mice. Thus, IR signaling and IGF1R signaling modulate synaptic plasticity specifically through regulating the expression of GluA1 in both the hippocampus and central amygdala.

Fig. 2.

Fig. 2.

Open in a new tab

Decreased expression of glutamate receptor 1 in the synaptosome of Hippo-DKO and CeA-DKO mice. (A and B) Representative Western blots and quantification of glutamate receptor 1 and 2 (GluA1 and GluA2) in synaptosomes extracted from the hippocampus in Hippo-DKO and control mice (A) and the central amygdala in CeA-DKO and control mice (B). GAPDH serve as loading controls. *P < 0.05, **P < 0.01 by unpaired t test, n = 6. Quantitative data are presented as mean ± SEM.

Glucose Homeostasis Is Regulated by IR/IGF1R Signaling in the Hippocampus and Central Amygdala.

Central insulin signaling has been shown to affect systemic energy homeostasis (16, 21, 22). Following AAV injection, both controls and DKO mice showed similar body weight gain and food intake (SI Appendix, Fig. S2 A–D). However, 3 wk after injection, Hippo-DKO mice displayed significantly higher random-fed blood glucose levels (Fig. 3A) and a strong trend toward decreased plasma insulin levels 9 wk after injection (SI Appendix, Fig. S2G). This was accompanied with an impaired oral glucose tolerance test (Fig. 3B). IR/IGF1R deletion in the hippocampus also resulted in moderate impairment of the glucose lowering effect of i.p. insulin injection (SI Appendix, Fig. S2E) and glucose-stimulated insulin secretion (SI Appendix, Fig. S2F), both of which might contribute to the glucose intolerance in Hippo-DKO mice. AUC, area under the curve.

Fig. 3.

Fig. 3.

Open in a new tab

Glucose homeostasis is regulated by IR/IGF1R signaling in the hippocampus and central amygdala. (A) Blood glucose levels in the random-fed state of Hippo-DKO (n = 31) or Hippo-CTR mice (n = 27). (B) Oral glucose tolerance test (OGTT) performed in Hippo-DKO or Hippo-CTR mice (n =13). (C) Blood glucose levels in the random-fed state of CeA-DKO (n = 29) or CeA-CTR mice (n = 28). (D) OGTT performed in CeA-DKO or CeA-CTR mice (n = 6). *P < 0.05, **P < 0.01, ***P < 0.001 by unpaired t test. Data are presented as mean ± SEM. AUC, area under the curve.

CeA-DKO mice also displayed increased random-fed blood glucose levels compared with control mice (Fig. 3C) and glucose intolerance (Fig. 3D), but this did not appear until 3 wk after AAV injection. However, loss of IR/IGF1R in the central amygdala did not have any significant impact on plasma insulin levels, insulin sensitivity, or glucose-stimulated insulin secretion (SI Appendix, Fig. S2 H–J).

Both Hippo-DKO and CeA-DKO mice housed in the metabolic cages under the fed and fasted conditions displayed similar spontaneous activity, energy expenditure measured by oxygen consumption (VO2) and CO2 production (VCO2), and respiratory exchange ratio (RER) as their control mice (SI Appendix, Fig. S3), indicating no major role for hippocampal and amygdala IR and IGF1R signaling on circadian rhythm, energy expenditure, and substrate preference.

IR/IGF1R Signaling in the Amygdala Contributes to Cold-Induced Thermogenesis.

To investigate whether the IR/IGF1R in the hippocampus or amygdala is required for normal thermogenesis, mice were challenged to cold (∼6 °C), and their rectal temperature was measured every 30 min for 3 h. Core body temperature of Hippo-DKO mice dropped minimally from 38 °C to 37.6 °C following 3-h cold exposure, similar to the drop observed in control mice (Fig. 4A). In contrast, the core body temperature of the CeA-DKO mice decreased by ∼1.0 °C during the 3-h cold exposure, significantly more than control mice, whose temperature was decreased by ∼0.5 °C (Fig. 4 B and C). Since brown adipose tissue (BAT) plays an important role for thermoregulation in rodents by producing heat through uncoupling protein 1 (UCP-1) (27), we assessed the expression of UCP-1 in BAT and found that both messenger and protein levels of UCP-1 in CeA-DKO mice were slightly, but not significantly, reduced compared with control mice (SI Appendix, Fig. S4 A and B).

Fig. 4.

Fig. 4.

Open in a new tab

IR/IGF1R signaling in the amygdala contributes to cold-induced thermogenesis. (A) Rectal temperature in Hippo-CTR (n = 4) and Hippo-DKO mice (n = 5) during a 3-h exposure to a 6 °C environment. (B) Rectal temperature in CeA-CTR (n = 9) and CeA-DKO mice (n = 10) during a 3-h exposure to a 6 °C environment. *P < 0.05 by unpaired t test. (C) Thermal images using a FLIR T300 Infrared Camera showing surface temperature after 3 h at 6 °C between CeA-CTR and CeA-DKO mice. (D) Fluorescent images of brain sections of mice 7 d after injection of the PRV-765 virus in the BAT. BLA, basolateral amygdalar area; BMA, basomedial amygdalar nucleus; COA, cortical amygdalar area; ENTl, entorhinal area lateral part; PA, piriform-amygdalar area; PIA, piriform area; TR, postpiriform transition area. Data are presented as mean ± SEM.

Sympathetic outflow is a key regulator of thermogenic activation of brown fat upon cold exposure (28). The defective cold-induced thermogenesis of CeA-DKO mice led us to hypothesize that some neuronal population expressing IR and IGF1R in the central amygdala are connected to the neural circuitry that controls sympathetic nerves innervating the interscapular brown adipose tissue (iBAT). To test this, we injected pseudorabies virus PRV-765 encoding red fluorescent protein (RFP) into iBAT to perform retrograde tracing of neuronal connections from iBAT to the brain. Seven days after viral injection, RFP-expressing neurons were detected in the spinal cord, nucleus of the solitary tract (NTS) of the medulla, and parabrachial nucleus (PBN) of the pons, as well as several areas in the hypothalamus, including the paraventricular nucleus (PVN), dorsomedial hypothalamus (DMH), and lateral hypothalamic area (LHA) (SI Appendix, Fig. S4C and Table S1), highlighting the neuronal circuit from the hypothalamus to the iBAT. Interestingly, the amygdala was also highly RFP-labeled (Fig. 4D), demonstrating previously unrecognized input of the neural circuits in the amygdala for the regulation of iBAT.

Hippo-DKO and CeA-DKO Mice Display Increased Anxiety-Like Behaviors.

Behavior of the mice was assessed 8 wk after the deletion of IR/IGF1R in the hippocampus and central amygdala. During an open field test, Hippo-DKO mice exhibited a 40% reduction in the number of center zone entries and spent ∼75% less time in the center zone compared with controls (Fig. 5A), indicating increased anxiety in these mice. These mice also showed significantly greater marble-burying activity (Fig. 5B), another indicator for anxiety-like behavior in rodents. In the dark–light box test, however, both control and DKO mice showed similar time spent in the light zone (Fig. 5C). Loss of IR and IGF1R in the central amygdala also resulted in ∼40% reduction in entries and time in the center zone in the open field test (Fig. 5D) and an ∼25% increase in marble-burying activity (Fig. 5E). In these mice, there was also an ∼33% decrease in time spent in the light compartment of a dark–light box compared with control (Fig. 5F). Thus, IR/IGF1R signaling in both the hippocampus and amygdala is important for controlling anxiety-like behavior in mice.

Fig. 5.

Fig. 5.

Open in a new tab

Hippo-DKO and CeA-DKO mice display increased anxiety-like behaviors. (A) Time spent in the center zone and entries in center zone in open field test in Hippo-CTR (n = 12) and Hippo-DKO mice (n = 15). (B) Assessment of anxiety as number of buried marbles over 30 min during a marble-burying task in Hippo-CTR (n = 15) and Hippo-DKO mice (n = 17). (C) Time spent in light compartment during light/dark box test in Hippo-CTR (n = 11) and Hippo-DKO mice (n = 12). (D) Time spent in the center zone and entries in center zone in open field test in CeA-CTR (n = 13) and CeA-DKO mice (n = 14). (E) Assessment of anxiety as number of buried marbles over 30 min during a marble-burying task test in CeA-CTR (n = 9) and CeA-DKO mice (n = 9). (F) Time spent in light compartment during light/dark box test in CeA-CTR (n = 18) and CeA-DKO mice (n = 19). *P < 0.05, **P < 0.01 by unpaired t test. Data are presented as mean ± SEM.

Hippo-DKO and CeA-DKO Mice Display Impaired Cognition.

The effect of IR/IGF1R loss in the hippocampus and central amygdala showed differential effects on learning and memory. In the habituation stage of the novel object recognition test, both control and Hippo-DKO mice showed equal exploration time with each of the two identical objects (SI Appendix, Fig. S5A). When one of the familiar objects was replaced with a novel object in the test stage 6 h later, control mice spent ∼65% of the time exploring the novel object (Fig. 6A). By contrast, Hippo-DKO mice showed equal interacting time between the familiar and novel objects (Fig. 6A), suggesting impaired recognition memory. In the novel object location test, one of the familiar objects was moved to a new location in the testing stage 6 h after the habituation phase. Control mice explored the object in the new location ∼75% of the total exploration time whereas Hippo-DKO mice explored both objects equally regardless of the location of the objects (Fig. 6B and SI Appendix, Fig. S5B), which is a sign of a spatial memory deficit.

Fig. 6.

Fig. 6.

Open in a new tab

Hippo-DKO and CeA-DKO mice have impaired cognition. (A) Time spent exploring the novel object during the test session of the object recognition task in Hippo-CTR (n = 8) and Hippo-DKO mice (n = 9). (B) Time spent exploring the object that was moved during the test session of the object location task in Hippo-CTR (n = 4) and Hippo-DKO mice (n = 5). (C) Errors recorded during the 10 acquisition trials, 1-wk, and 1-mo memory testing of the Stone T maze in Hippo-CTR (n = 7) and Hipp-DKO mice (n = 8). (D) Time spent exploring the novel object during the test session of the object recognition task in CeA-CTR (n = 9) and CeA-DKO mice (n = 10). (E) Time spent exploring the object that was moved during the test session of the object location task in CeA-CTR (n = 6) and CeA-DKO mice (n = 6). (F) Errors recorded during the 10 acquisition trials, 1-wk, and 1-mo memory testing of the Stone T maze in CeA-CTR (n = 7) and CeA-DKO mice (n = 9). #P < 0.05, ##P < 0.01, ###P < 0.001 by unpaired t test. *P < 0.05, **P < 0.01 between DKO and control mice by unpaired t test. Data are presented as mean ± SEM.

We further analyzed spatial learning and memory using a Stone T maze. In this test, a mouse needs to learn the correct sequence of 13 left/right turns to successfully escape a water-filled maze and reach the goal box (SI Appendix, Fig. S5E). During the first 10 trials of learning conducted over 3 d, control mice displayed continuous improvement of learning, with fewer errors and shorter latency to reach the goal box (Fig. 6C and SI Appendix, Fig. S5F). In contrast, Hippo-DKO mice displayed significantly slower learning (Fig. 6C and SI Appendix, Fig. S5F). One week after the last learning trial, mice were exposed to the same maze to assess their memory. Control mice were able to finish the maze with low numbers of errors and short latency, similar to how they performed during the last trial of learning. By contrast, after a 1-wk hiatus, Hippo-DKO mice made significantly more errors and took a longer time to complete the maze (Fig. 6C and SI Appendix, Fig. S5F), and these mice performed even worse on this task after a 1-mo hiatus (Fig. 6C and SI Appendix, Fig. S5F). In addition, when the maze was rotated 180° from its original setting, only controls, but not Hippo-DKO mice, showed worsened performance in the T maze, indicating that control mice used spatial cues to finish the maze while Hippo-DKO mice did not (Fig. 6C and SI Appendix, Fig. S5F).

Effects on memory and learning were selective and dependent more on the specific task in CeA-DKO mice. Thus, CeA-DKO mice failed to recognize the novel object compared with controls in the novel object recognition test (Fig. 6D and SI Appendix, Fig. S5C). However, CeA-DKO mice were able to remember the location of the object as well as control mice (Fig. 6E and SI Appendix, Fig. S5D). In the Stone T maze test, CeA-DKO mice showed normal learning and only slightly impaired ability to remember the maze (Fig. 6F and SI Appendix, Fig. S5G). These data suggest that loss of IR and IGF1R in the central amygdala significantly impairs the recognition memory of the mice but has only minor effects on spatial memory.

Discussion

Insulin signaling and IGF-1 signaling produce a range of effects on cellular metabolism, proteostasis, growth, and many other functions (6, 29–32). Recent studies have shown that the brain is a major target of insulin/IGF-1 signaling. Insulin and IGF-1 are able to cross the blood–brain barrier (BBB), in part through receptor-mediated transcytosis, and act on brain tissues (33, 34). In addition, a significant amount of IGF-1 is produced in the brain. IR and IGF1R are widely distributed throughout the brain and are present on both neurons and glial cells (7). Thus, the brain is able to respond to locally and systemically produced insulin and IGF-1 and to modulate its activities accordingly. In the present study, we found that loss of IR and IGF1R in the hippocampus or central amygdala results in metabolic and behavioral abnormalities, including impaired glucose homeostasis and cognition and increased anxiety-like behavior. We also show that insulin/IGF-1 signaling in the central amygdala plays a specific role in thermogenesis while insulin/IGF-1 signaling in the hippocampus is important for spatial memory.

At a molecular level, loss of IR and IGF1R results in a significant reduction of the GluA1 subunit of the AMPA receptor present in the synaptosomal fraction. Unlike NMDA receptors, which are ligand-gated channels permeable to both calcium and sodium, AMPA receptors are ligand-gated sodium channels responsible for the rapid depolarization of neuron membrane potential (25, 26). The AMPA receptor is a tetrameric channel, composed mainly of GluA1/GluA2 or GluA2/GluA3 heterodimers (25). A large proportion of GluA1 subunit-containing AMPA receptors localize in the endosomes close to the postsynaptic membrane (35), which can undergo rapid recruitment to the synaptic membrane upon stimulation by insulin or by NMDA receptor-mediated calcium influx (36). This is a key molecular mechanism of long-term potentiation (LTP) (37) and is important for learning and memory. Consistent with this, mice with both IR and IGF1R deleted in the hippocampus display impaired recognition and spatial memory, possibly due to the limited GluA1 subunit-containing AMPA receptor pools in the synapses of these neurons. In agreement with this, lentiviral-mediated deletion of IR in the hippocampus results in impaired spatial memory due to impaired LTP (38). Also, knockout of SCAP, a cholesterol sensing protein downstream of insulin action, also results in impaired LTP (39). Intriguingly, IR/IGF1R deletion has no major effect on the phosphorylation of the GluA1 subunit of the AMPA receptor. Consistent with this, calmodulin kinase and cAMP/PKA pathways, which are responsible for GluA1 phosphorylation (40, 41), are not generally considered downstream signaling pathways regulated by IR/IGF1R. The mechanism by which IR/IGF1R regulates GluA1 protein levels will require further investigation but might include transcriptional or posttranscriptional alterations.

Insulin is the key hormone regulating glucose handling and energy homeostasis. Thus, it is not surprising that the majority of research on brain insulin action has focused on the hypothalamus, which controls many metabolic responses (21, 22, 42). Our study clearly shows that metabolic control by central insulin signaling is not limited to the hypothalamus since IR/IGF1R deletion in both hippocampus and amygdala leads to impaired glucose tolerance. In the hippocampus, this appears to be due to a combination of systemic insulin resistance and a decrease in insulin secretion. The causal factor for impaired glucose tolerance in mice lacking IR and IGF1R in the amygdala, on the other hand, is less clear. These mice show normal insulin sensitivity and glucose-stimulated glucose secretion, suggesting a defect in some peripheral insulin-independent pathway involved in glucose disposal. The sympathetic nervous system (SNS) innervates the liver and controls hepatic glucose production. Our data indicate that the amygdala may play a role in the regulation of hepatic glucose metabolism and, eventually, systemic glucose levels, directly or through other brain regions with which it has projections, such as the hypothalamus (43). Indeed, it has been shown that insulin injected in the central amygdala activates neuronal populations in regions of the hypothalamus (44) known to affect the autonomic output to the liver. Interestingly, it has been noted that mice with diet-induced obesity develop insulin resistance in the central amygdala (44) similar to that reported in the hypothalamus (45), and thus it is possible that insulin signaling in the central amygdala participates in abnormal glucose homeostasis in the development of obesity.

Interscapular brown adipose tissue (iBAT) is a major thermogenic tissue in rodents and is important in the regulation of core body temperature, as well as systemic glucose and lipid metabolism (27). Brown adipose tissue is also present in humans where it has both thermogenic and metabolic functions (46). Regulation of BAT is coordinated by the brain. Thus, when temperature changes, warm- and cold-sensitive neurons signal to the preoptic area of the hypothalamus (POA), which, through a circuit involving the dorsal medial hypothalamus (DMH) and the rostral raphe pallidus nucleus (rRPa), controls sympathetic nervous system (SNS) outflow to BAT and thus its thermogenic activity (47). Our retrograde transsynaptic tracing demonstrates that the amygdala is directly involved in the neural circuit innervating BAT. This neural pathway is important for cold-induced thermogenesis, and, more importantly, it is regulated by insulin/IGF-1 signaling since IR/IGF1R deletion in the central amygdala specifically leads to cold intolerance in mice. Several groups, including us, have previously observed the thermogenic effects of central insulin/IGF-1 signaling using whole brain IR knockout mice (7, 48, 49). The current study pinpoints the amygdala as a critical nucleus where insulin/IGF-1 signaling exerts a major role in thermogenesis. How this converges with the POA → DMH → rRPa pathway requires further studies.

We have also identified a crucial role for IR and IGF1R in the hippocampus and amygdala on mood and cognition. Both Hippo-DKO and CeA-DKO mice displayed significantly increased anxiety-like behavior and impaired cognition, indicating a beneficial role of insulin signaling on mood and cognition. Intranasal insulin delivery has been shown to improve mood and cognition in humans (50, 51). Our studies suggest that the ionotropic glutamate AMPA receptor and glutamate signaling may be a molecular link between brain insulin/IGF-1 signaling deficiency and altered neurobehavior. Other studies have shown that inhibition of metabotropic glutamate receptors mGluR2/3 can improve mood behavior, and this is accompanied by increased adult hippocampal neurogenesis (52). Both insulin/IGF-1 and BDNF have been shown to contribute to adult hippocampal neurogenesis and have potential beneficial effects in Alzheimer’s disease and psychotic disorders (53, 54). Whether insulin/IGF-1 signaling and BDNF signaling share a common mechanism for their antidepressive and antidementia effects awaits further investigation. In addition, future studies are needed to explore the electrophysiological mechanism of the amygdala GluA1-associated defects in cognition and mood in CeA-DKO mice.

Interestingly, the phenotypes of the region-specific IR/IGF1R double knockout mice are more severe than those of the mice with single IR-only knockout throughout the whole brain, since mice with Nestin-Cre-dependent IR deletion (NIRKO) display no apparent spatial memory deficit, and the anxiety-like behaviors develop only with aging (18, 55). On the other hand, IR knockout in astrocytes does have significant effects on mood and behavior in young mice (56). To what extent the phenotype of these region-specific IR/IGF1R knockouts involves astrocytes vs. neurons is uncertain. Combined IR and IGF1R deletion in peripheral tissues, like muscle and adipose tissue, leads to more severe phenotypes than IR-only knockout (29, 32), suggesting that IGF1R might partially compensate for IR in peripheral tissues, and this could be also true in the brain. Previous reports have shown higher expression of IGF1R than IR in the brain (7), indicating a potential important role for this insulin-IGF1R cross-activation in the brain. Further studies will be necessary to dissect the specific contributions of IR and IGF1R or the potential role of IR/IGF1R hybrids in these phenotypes.

In summary, we have shown important roles for hippocampal and amygdala IR and IGF1R signaling in systemic glucose homeostasis and thermogenesis. At least a part of this effect may be through modulation of an amygdala-dependent neural circuit to control sympathetic outflow to liver and brown adipose tissue, thus regulating peripheral glucose handling and cold-induced thermogenesis. In addition, we demonstrate that IR and IGF1R in the hippocampus and central amygdala are important for mood and cognition. These defects in IR/IGF1R-deficient mice may be the result of a reduction of AMPA receptor subunit GluA1 in the synapse, which impairs synaptic plasticity in DKO mice. Thus, both the hippocampus and amygdala are important brain regions for central insulin action and normal energy homeostasis and neural functions.

Methods

All animal studies were conducted in compliance with the regulations and ethics guidelines of the NIH and were approved by the Institutional Animal Care and Use Committee (IACUC) of the Joslin Diabetes Center. Detailed materials and methods are available in SI Appendix.

Supplementary Material

Supplementary File
pnas.1817391116.sapp.pdf (2.9MB, pdf)

Acknowledgments

We thank Dr. Lynn W. Enquist (Princeton University) for sharing the PRV-765 virus for the retrograde tracing studies. This work was supported by NIH Grants R01 DK031036 and R01 DK033201 (to C.R.K.). The Advanced Microscopy Core and Animal Physiology Core in the Joslin Diabetes Research Center (DRC) (P30 DK036836) also provided important help.

Footnotes

Conflict of interest statement: C.R.K. and S.G. are coauthors on a 2016 review article.

See Commentary on page 5852.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1817391116/-/DCSupplemental.

References

  • 1.Kahn CR, Freychet P, Roth J, Neville DM., Jr Quantitative aspects of the insulin-receptor interaction in liver plasma membranes. J Biol Chem. 1974;249:2249–2257. [PubMed] [Google Scholar]
  • 2.Accili D, et al. Early neonatal death in mice homozygous for a null allele of the insulin receptor gene. Nat Genet. 1996;12:106–109. doi: 10.1038/ng0196-106. [DOI] [PubMed] [Google Scholar]
  • 3.Steele-Perkins G, et al. Expression and characterization of a functional human insulin-like growth factor I receptor. J Biol Chem. 1988;263:11486–11492. [PubMed] [Google Scholar]
  • 4.Liu JP, Baker J, Perkins AS, Robertson EJ, Efstratiadis A. Mice carrying null mutations of the genes encoding insulin-like growth factor I (Igf-1) and type 1 IGF receptor (Igf1r) Cell. 1993;75:59–72. [PubMed] [Google Scholar]
  • 5.DeChiara TM, Efstratiadis A, Robertson EJ. A growth-deficiency phenotype in heterozygous mice carrying an insulin-like growth factor II gene disrupted by targeting. Nature. 1990;345:78–80. doi: 10.1038/345078a0. [DOI] [PubMed] [Google Scholar]
  • 6.Cai W, et al. Domain-dependent effects of insulin and IGF-1 receptors on signalling and gene expression. Nat Commun. 2017;8:14892. doi: 10.1038/ncomms14892. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Kleinridders A, Ferris HA, Cai W, Kahn CR. Insulin action in brain regulates systemic metabolism and brain function. Diabetes. 2014;63:2232–2243. doi: 10.2337/db14-0568. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Fernandez AM, Torres-Alemán I. The many faces of insulin-like peptide signalling in the brain. Nat Rev Neurosci. 2012;13:225–239. doi: 10.1038/nrn3209. [DOI] [PubMed] [Google Scholar]
  • 9.Kopf D, Frölich L. Risk of incident Alzheimer’s disease in diabetic patients: A systematic review of prospective trials. J Alzheimers Dis. 2009;16:677–685. doi: 10.3233/JAD-2009-1011. [DOI] [PubMed] [Google Scholar]
  • 10.Eaton WW, Armenian H, Gallo J, Pratt L, Ford DE. Depression and risk for onset of type II diabetes. A prospective population-based study. Diabetes Care. 1996;19:1097–1102. doi: 10.2337/diacare.19.10.1097. [DOI] [PubMed] [Google Scholar]
  • 11.Talbot K, et al. Demonstrated brain insulin resistance in Alzheimer’s disease patients is associated with IGF-1 resistance, IRS-1 dysregulation, and cognitive decline. J Clin Invest. 2012;122:1316–1338. doi: 10.1172/JCI59903. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Steen E, et al. Impaired insulin and insulin-like growth factor expression and signaling mechanisms in Alzheimer’s disease–Is this type 3 diabetes? J Alzheimers Dis. 2005;7:63–80. doi: 10.3233/jad-2005-7107. [DOI] [PubMed] [Google Scholar]
  • 13.Moloney AM, et al. Defects in IGF-1 receptor, insulin receptor and IRS-1/2 in Alzheimer’s disease indicate possible resistance to IGF-1 and insulin signalling. Neurobiol Aging. 2010;31:224–243. doi: 10.1016/j.neurobiolaging.2008.04.002. [DOI] [PubMed] [Google Scholar]
  • 14.Craft S. Insulin resistance syndrome and Alzheimer’s disease: Age- and obesity-related effects on memory, amyloid, and inflammation. Neurobiol Aging. 2005;26:65–69. doi: 10.1016/j.neurobiolaging.2005.08.021. [DOI] [PubMed] [Google Scholar]
  • 15.Blázquez E, Velázquez E, Hurtado-Carneiro V, Ruiz-Albusac JM. Insulin in the brain: Its pathophysiological implications for states related with central insulin resistance, type 2 diabetes and Alzheimer’s disease. Front Endocrinol (Lausanne) 2014;5:161. doi: 10.3389/fendo.2014.00161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Brüning JC, et al. Role of brain insulin receptor in control of body weight and reproduction. Science. 2000;289:2122–2125. doi: 10.1126/science.289.5487.2122. [DOI] [PubMed] [Google Scholar]
  • 17.Fisher SJ, Brüning JC, Lannon S, Kahn CR. Insulin signaling in the central nervous system is critical for the normal sympathoadrenal response to hypoglycemia. Diabetes. 2005;54:1447–1451. doi: 10.2337/diabetes.54.5.1447. [DOI] [PubMed] [Google Scholar]
  • 18.Kleinridders A, et al. Insulin resistance in brain alters dopamine turnover and causes behavioral disorders. Proc Natl Acad Sci USA. 2015;112:3463–3468. doi: 10.1073/pnas.1500877112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kappeler L, et al. Brain IGF-1 receptors control mammalian growth and lifespan through a neuroendocrine mechanism. PLoS Biol. 2008;6:e254. doi: 10.1371/journal.pbio.0060254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Taguchi A, Wartschow LM, White MF. Brain IRS2 signaling coordinates life span and nutrient homeostasis. Science. 2007;317:369–372. doi: 10.1126/science.1142179. [DOI] [PubMed] [Google Scholar]
  • 21.Könner AC, et al. Insulin action in AgRP-expressing neurons is required for suppression of hepatic glucose production. Cell Metab. 2007;5:438–449. doi: 10.1016/j.cmet.2007.05.004. [DOI] [PubMed] [Google Scholar]
  • 22.Shin AC, et al. Insulin receptor signaling in POMC, but not AgRP, neurons controls adipose tissue insulin action. Diabetes. 2017;66:1560–1571. doi: 10.2337/db16-1238. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Skeberdis VA, Lan J, Zheng X, Zukin RS, Bennett MV. Insulin promotes rapid delivery of N-methyl-D- aspartate receptors to the cell surface by exocytosis. Proc Natl Acad Sci USA. 2001;98:3561–3566. doi: 10.1073/pnas.051634698. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Beattie EC, et al. Regulation of AMPA receptor endocytosis by a signaling mechanism shared with LTD. Nat Neurosci. 2000;3:1291–1300. doi: 10.1038/81823. [DOI] [PubMed] [Google Scholar]
  • 25.Derkach VA, Oh MC, Guire ES, Soderling TR. Regulatory mechanisms of AMPA receptors in synaptic plasticity. Nat Rev Neurosci. 2007;8:101–113. doi: 10.1038/nrn2055. [DOI] [PubMed] [Google Scholar]
  • 26.Iacobucci GJ, Popescu GK. NMDA receptors: Linking physiological output to biophysical operation. Nat Rev Neurosci. 2017;18:236–249. doi: 10.1038/nrn.2017.24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Cannon B, Nedergaard J. Brown adipose tissue: Function and physiological significance. Physiol Rev. 2004;84:277–359. doi: 10.1152/physrev.00015.2003. [DOI] [PubMed] [Google Scholar]
  • 28.Lowell BB, Bachman ES. Beta-Adrenergic receptors, diet-induced thermogenesis, and obesity. J Biol Chem. 2003;278:29385–29388. doi: 10.1074/jbc.R300011200. [DOI] [PubMed] [Google Scholar]
  • 29.O’Neill BT, et al. Insulin and IGF-1 receptors regulate FoxO-mediated signaling in muscle proteostasis. J Clin Invest. 2016;126:3433–3446. doi: 10.1172/JCI86522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Boucher J, et al. Differential roles of insulin and IGF-1 receptors in adipose tissue development and function. Diabetes. 2016;65:2201–2213. doi: 10.2337/db16-0212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.O’Neill BT, et al. Differential role of insulin/IGF-1 receptor signaling in muscle growth and glucose homeostasis. Cell Rep. 2015;11:1220–1235. doi: 10.1016/j.celrep.2015.04.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Sakaguchi M, et al. Adipocyte dynamics and reversible metabolic syndrome in mice with an inducible adipocyte-specific deletion of the insulin receptor. Cell Metab. 2017;25:448–462. doi: 10.1016/j.cmet.2016.12.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Konishi M, et al. Endothelial insulin receptors differentially control insulin signaling kinetics in peripheral tissues and brain of mice. Proc Natl Acad Sci USA. 2017;114:E8478–E8487. doi: 10.1073/pnas.1710625114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Gray SM, Aylor KW, Barrett EJ. Unravelling the regulation of insulin transport across the brain endothelial cell. Diabetologia. 2017;60:1512–1521. doi: 10.1007/s00125-017-4285-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Park M, Penick EC, Edwards JG, Kauer JA, Ehlers MD. Recycling endosomes supply AMPA receptors for LTP. Science. 2004;305:1972–1975. doi: 10.1126/science.1102026. [DOI] [PubMed] [Google Scholar]
  • 36.Passafaro M, Piëch V, Sheng M. Subunit-specific temporal and spatial patterns of AMPA receptor exocytosis in hippocampal neurons. Nat Neurosci. 2001;4:917–926. doi: 10.1038/nn0901-917. [DOI] [PubMed] [Google Scholar]
  • 37.Zamanillo D, et al. Importance of AMPA receptors for hippocampal synaptic plasticity but not for spatial learning. Science. 1999;284:1805–1811. doi: 10.1126/science.284.5421.1805. [DOI] [PubMed] [Google Scholar]
  • 38.Grillo CA, et al. Hippocampal insulin resistance impairs spatial learning and synaptic plasticity. Diabetes. 2015;64:3927–3936. doi: 10.2337/db15-0596. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Suzuki R, Ferris HA, Chee MJ, Maratos-Flier E, Kahn CR. Reduction of the cholesterol sensor SCAP in the brains of mice causes impaired synaptic transmission and altered cognitive function. PLoS Biol. 2013;11:e1001532. doi: 10.1371/journal.pbio.1001532. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Kristensen AS, et al. Mechanism of Ca2+/calmodulin-dependent kinase II regulation of AMPA receptor gating. Nat Neurosci. 2011;14:727–735. doi: 10.1038/nn.2804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Esteban JA, et al. PKA phosphorylation of AMPA receptor subunits controls synaptic trafficking underlying plasticity. Nat Neurosci. 2003;6:136–143. doi: 10.1038/nn997. [DOI] [PubMed] [Google Scholar]
  • 42.Loh K, et al. Insulin controls food intake and energy balance via NPY neurons. Mol Metab. 2017;6:574–584. doi: 10.1016/j.molmet.2017.03.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Aizawa H, et al. Development of the amygdalohypothalamic projection in the mouse embryonic forebrain. Anat Embryol (Berl) 2004;208:249–264. doi: 10.1007/s00429-004-0399-9. [DOI] [PubMed] [Google Scholar]
  • 44.Boghossian S, Lemmon K, Park M, York DA. High-fat diets induce a rapid loss of the insulin anorectic response in the amygdala. Am J Physiol Regul Integr Comp Physiol. 2009;297:R1302–R1311. doi: 10.1152/ajpregu.00252.2009. [DOI] [PubMed] [Google Scholar]
  • 45.Posey KA, et al. Hypothalamic proinflammatory lipid accumulation, inflammation, and insulin resistance in rats fed a high-fat diet. Am J Physiol Endocrinol Metab. 2009;296:E1003–E1012. doi: 10.1152/ajpendo.90377.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Cypess AM, et al. Identification and importance of brown adipose tissue in adult humans. N Engl J Med. 2009;360:1509–1517. doi: 10.1056/NEJMoa0810780. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Tupone D, Madden CJ, Morrison SF. Autonomic regulation of brown adipose tissue thermogenesis in health and disease: Potential clinical applications for altering BAT thermogenesis. Front Neurosci. 2014;8:14. doi: 10.3389/fnins.2014.00014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Sanchez-Alavez M, et al. Insulin causes hyperthermia by direct inhibition of warm-sensitive neurons. Diabetes. 2010;59:43–50. doi: 10.2337/db09-1128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Sanchez-Alavez M, et al. Insulin-like growth factor 1-mediated hyperthermia involves anterior hypothalamic insulin receptors. J Biol Chem. 2011;286:14983–14990. doi: 10.1074/jbc.M110.188540. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Benedict C, et al. Intranasal insulin improves memory in humans. Psychoneuroendocrinology. 2004;29:1326–1334. doi: 10.1016/j.psyneuen.2004.04.003. [DOI] [PubMed] [Google Scholar]
  • 51.Craft S, et al. Intranasal insulin therapy for Alzheimer disease and amnestic mild cognitive impairment: A pilot clinical trial. Arch Neurol. 2012;69:29–38. doi: 10.1001/archneurol.2011.233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Perez-Garcia G, et al. PTSD-related behavioral traits in a rat model of blast-induced mTBI are reversed by the mGluR2/3 receptor antagonist BCI-838. eNeuro. 2018;5:ENEURO.0357-17.2018. doi: 10.1523/ENEURO.0357-17.2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Choi SH, et al. Combined adult neurogenesis and BDNF mimic exercise effects on cognition in an Alzheimer’s mouse model. Science. 2018;361:eaan8821. doi: 10.1126/science.aan8821. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Mir S, Cai W, Carlson SW, Saatman KE, Andres DA. IGF-1 mediated neurogenesis involves a novel RIT1/Akt/Sox2 cascade. Sci Rep. 2017;7:3283. doi: 10.1038/s41598-017-03641-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Schubert M, et al. Role for neuronal insulin resistance in neurodegenerative diseases. Proc Natl Acad Sci USA. 2004;101:3100–3105. doi: 10.1073/pnas.0308724101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Cai W, et al. Insulin regulates astrocyte gliotransmission and modulates behavior. J Clin Invest. 2018;128:2914–2926. doi: 10.1172/JCI99366. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary File
pnas.1817391116.sapp.pdf (2.9MB, pdf)

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES