Minireview: Food for Thought: Regulation of Synaptic Function by Metabolic Hormones

Abstract

The peripheral actions of the metabolic hormones, leptin and insulin, are well documented. However, the functions of these hormones are not restricted to the periphery because evidence is growing that both leptin and insulin can readily cross the blood-brain barrier and have widespread central actions. The hippocampus in particular expresses high levels of both insulin and leptin receptors as well as key components of their associated signaling cascades. Moreover, recent studies indicate that both hormones are potential cognitive enhancers. Indeed, it has been demonstrated that both leptin and insulin markedly influence key cellular events that underlie hippocampal learning and memory including activity-dependent synaptic plasticity and the trafficking of glutamate receptors to and away from hippocampal synapses. The hippocampal formation is also a prime site for the neurodegenerative processes that occur during Alzheimer's disease, and impairments in either leptin or insulin function have been linked to central nervous system-driven diseases like Alzheimer's disease. Thus, the capacity of the metabolic hormones, leptin and insulin, to regulate hippocampal synaptic function has significant implications for normal brain function and also central nervous system-driven disease.

The endocrine hormone leptin is the product of the obese (ob) gene that is primarily made and secreted by white adipose tissue and circulates in amounts proportional to body fat content (1, 2). Leptin uses a saturable transporter to cross the blood-brain barrier and enter the brain. One of its key central targets is the hypothalamus and in particular the arcuate nucleus in which it plays a pivotal role in regulating food intake and body weight. However, the neuronal actions of leptin are not restricted to controlling feeding behavior because leptin receptors are widely expressed throughout the central nervous system (CNS), and evidence is growing that leptin is a pleiotropic signal that regulates diverse neuronal functions.

Materials and Methods

Leptin

The diabetes (db) gene encodes the leptin receptor (ObR), which shares the greatest homology with the class I cytokine superfamily of receptors (3). Six splice variants of ObR have been identified that are denoted ObRa to ObRf. The short isoforms (ObRa, ObRc, ObRd, and ObRf) are involved in the internalization and breakdown of leptin, whereas ObRe that lacks a transmembrane domain is thought to buffer circulating leptin levels. In contrast, the long form, ObRb, has the longest intracellular C-terminal domain (302 residues) and is the predominant signaling competent isoform. ObRs signal via association with and activation of Janus tyrosine kinases, and in particular Janus tyrosine kinase-2. Phosphinositide 3-kinase (PI3-kinase), MAPK, and signal transducers and activators of transcription 3 are the principal signaling pathways activated by ObRs in neurons (4, 5).

Insulin

Insulin, a 6-kDa protein is made and secreted by pancreatic β-cells, and once secreted, it controls blood glucose levels by regulating the uptake of glucose from the circulation. In a manner similar to leptin, the actions of insulin extend beyond the periphery. Indeed, it is well documented that insulin readily enters the brain via a saturable transport mechanism (6) and that insulin receptors are highly expressed in neurons (7). Numerous studies indicate that insulin has an impact on various central processes including hypothalamic feeding behavior and neuronal survival. Moreover, like leptin, insulin has a marked influence on cognitive processes, and impaired insulin signaling is associated with CNS-driven diseases like Alzheimer's disease (AD).

The insulin receptor is a transmembrane receptor that belongs to the tyrosine kinase receptor superfamily. The insulin receptor is encoded by a single gene, INSR, and alternate splicing of this gene results in the generation of either α- or β-subunits. Insulin binding to the α-chains induces structural changes within the receptor, leading to the autophosphorylation of tyrosine residues within the intracellular domain of the β-chain. This in turn leads to either intracellular recruitment of insulin receptor substrate (IRS) proteins and subsequent activation of the PI3-kinase signaling cascade or the recruitment of shc, an adaptor protein, resulting in ERK1/2 stimulation (8).

Regulation of neuronal development

Leptin

The discovery of high levels of leptin receptor expression in placenta was the first indication that leptin is involved in developmental processes (9). This was further supported by the identification of leptin receptors in both fetal tissues and human umbilical cord (10). Studies in rodent brains have implicated leptin in neuronal development as the expression and localization of leptin receptors markedly alters at different developmental stages (11–13). Age-related alterations in the expression of leptin mRNA has been observed in rodent brain, suggesting developmental regulation of the leptin (ob) gene. Moreover, there is a significant increase in the circulating levels of leptin during the postnatal period, with a distinct surge in leptin occurring between postnatal day (P) 8 and P12 in rodents (14). Additionally, leptin-deficient (ob/ob) or leptin-insensitive (db/db) mice display various CNS abnormalities, including reduced brain weight and altered levels of neuronal and glial proteins (15). Attenuated levels of key synaptic proteins have also been detected in the hippocampus and cortex of ob/ob and db/db mice, and these deficits are alleviated by leptin treatment (15). More recent studies have shown that leptin promotes the development of key pathways in the arcuate nucleus that are critically involved in the hypothalamic regulation of food intake and body weight (16), an action likely to reflect the ability of leptin to promote neurite outgrowth from arcuate neurons (16) and also alter the efficacy of excitatory and inhibitory synaptic contacts onto neuropeptide Y and proopiomelanocortin neurons (17).

Insulin

Insulin levels are detectable in embryonic tissue as early as embryonic day (E) 8 (18), with embryonic levels peaking around E9. In the CNS, expression of insulin has also been detected in mouse embryos at E9, which correlates well with elevated insulin levels in embryonic peripheral tissues at this age (19). Insulin receptor levels are also developmentally regulated, with higher levels detected at early developmental stages but this declines with age. Because insulin receptors are found at synapses and concentrate at the postsynaptic density (20), it is likely that insulin receptors play a pivotal role in regulating synaptic function during initial neuronal development. Indeed, insulin receptor signaling promotes stabilization of excitatory synaptic transmission by association with synaptic scaffolding proteins. Overexpression of IRSp53, a novel insulin receptor substrate protein that is enriched in the brain, also enhances the density of spines in hippocampal neurons (21). Insulin also regulates hypothalamic development by influencing the density of specific hypothalamic neuronal populations. For instance, streptozotocin-induced hypoinsulinemia increases the density of the neuropeptide Y- and proopiomelanocortin-containing neurons in the arcuate nucleus (22). The ablation of the insulin receptors in mice also increases the number of proopiomelanocortin neurons in adulthood (23).

Regulation of hippocampal synaptic plasticity by leptin and insulin

Leptin

It is well documented that one of the key functions of the hippocampus is learning and memory. Moreover, activity-dependent forms of excitatory synaptic plasticity, namely long-term potentiation (LTP) and long-term depression (LTD), are important cellular correlates of hippocampal spatial learning, memory, and habituation (24). Several studies have demonstrated high levels of leptin receptor mRNA and immunoreactivity expressed throughout the hippocampal formation (25–27). In addition, dual-labeling immunocytochemical studies have revealed that ObR colocalizes with the synaptic marker synapsin 1, indicating its distribution at synapses (27). In support of a synaptic locus, several studies have found that leptin potently regulates excitatory synaptic transmission at hippocampal synapses (28–30) (see Figure 1). In addition, leptin modifies various forms of hippocampal synaptic plasticity (31). Indeed, leptin-insensitive rodents (db/db mice; fa/fa rats) display impairments in hippocampal synaptic plasticity (32), whereas administration of leptin directly into the hippocampus augments hippocampal LTP (33). Furthermore, application of leptin to acute hippocampal slices promotes the conversion of short-term potentiation to LTP (28).

Figure 1. Age-dependent effects of leptin on excitatory synaptic transmission at hippocampal CA1 synapses.

Exposure of acute hippocampal slices to leptin (25 nM) for 15 minutes resulted in distinct age-dependent effects on excitatory synaptic transmission (Ai–Di). At P5–8 leptin induced a novel form of LTD (Ai), whereas a transient synaptic depression was observed at P11–18 (Bi). Conversely, a persistent increase in synaptic strength was evoked by leptin in adult (12–16 wk; Ci) and aged (12–14 mo; Di) hippocampus, respectively. Aii–Dii, The ability of leptin to modulate synaptic transmission in the developing and adult hippocampus were not linked to alterations in the paired pulse facilitation ratio (PPR) or coefficient of variation (CV), suggesting the likely involvement of a postsynaptic expression mechanism. [Reproduced with permission from P. R. Moult and J. Harvey: Neuropharmacology. 61:924–936, 2011 (29). NMDA receptor subunit composition determines the polarity of leptin-induced synaptic plasticity. ©Elsevier.]

Although N-methyl-D-aspartate (NMDA) receptors contribute little to basal excitatory synaptic transmission, synaptic activation of NMDA receptors and the subsequent postsynaptic rise in intracellular Ca²⁺ are critical for the induction of LTP and LTD at hippocampal CA1 synapses (34). Numerous hormones and growth factors modulate NMDA receptor function and in turn influence hippocampal synaptic plasticity. Likewise, leptin facilitates NMDA-evoked responses in native hippocampal neurons and heterologous expression systems (35), an effect that is likely to underlie the ability of this hormone to modulate hippocampal synaptic plasticity. Indeed, leptin-driven facilitation of hippocampal LTP (28) reversal of established LTP (36) and induction of de novo LTD (37) are all NMDA receptor-dependent processes. Synaptic activation of NMDA receptors is also crucial for leptin-driven increases in synaptic efficacy (leptin induced LTP) at adult hippocampal CA1 synapses (38).

Previous studies indicate that divergent signaling cascades couple leptin receptors to increased NMDA receptor (GluN2B) function in cerebellar granule cells (39). Distinct GluN2 subunits are also implicated in leptin's effects on hippocampal excitatory synaptic transmission at different stages of postnatal development and aging (29) and that divergent signaling pathways are involved at different ages. Thus, in juvenile hippocampus, leptin facilitates GluN2B-mediated responses via an ERK-dependent mechanism resulting in synaptic depression. Conversely, leptin-driven stimulation of PI3-kinase signaling promotes an increase in GluN2A-mediated responses that in turn results in a persistent increase in synaptic efficacy in adult (29).

It is known that the activation of NMDA receptors accelerates the movement of 2-amino-3-hydroxy-5-methyl-4-isoxazol propionic acid (AMPA) receptors to synapses during LTP (40) and that the molecular identity of synaptic AMPA receptors can change after LTP induction (41–43). Similarly, increased synaptic insertion of the AMPA receptor subunit, GluA1 accompanies leptin-induced LTP in adult hippocampus (38). Thus, enhanced AMPA receptor rectification was observed after the induction of leptin-induced LTP, and leptin-induced LTP was reversed by the blockade of GluA2-lacking AMPA receptors, suggesting that an increase in the synaptic density of GluA2-lacking AMPA receptors is involved (38).

Insulin and synaptic plasticity

A number of studies indicate that insulin, like leptin, can markedly influence activity-dependent synaptic plasticity in the hippocampus. For instance, brief exposure of hippocampal slices to insulin results in a persistent depression (LTD) of excitatory synaptic transmission (44–46), a novel form of LTD that requires the activation of NMDA receptors and is expressed postsynaptically. In a manner similar to LTD induced by low frequency stimulation, phosphorylation, and subsequent endocytosis of AMPA receptor subunits, and in particular GluA2, is reported to underlie insulin-induced LTD (44). Insulin also regulates AMPA receptor trafficking processes as the surface expression of GluA1 is elevated after treatment of hippocampal neurons with insulin (44, 47). However, the insertion of GluA2 is not affected by treatment with insulin (47). The trafficking of NMDA receptors is also modulated by insulin because NMDA receptor surface expression and NMDA-evoked currents are enhanced by insulin (48, 49). Exposure of rat hippocampal slices to insulin also leads to enhanced expression of postsynaptic density protein-95 (PSD-95), a scaffold protein closely involved in regulating NMDA receptor assembly. Consequently, the ability of insulin to modulate pleckstrin and Sec7 domain protein-95 function may be involved in the regulation of hippocampal synaptic plasticity by insulin. In addition, a leftward shift in the frequency response function for the induction of both LTD and LTP has been observed in hippocampal slices exposed to insulin (46). This is not surprising because insulin-driven enhancement of NMDA receptor function will lower the threshold for the induction of synaptic plasticity.

Structural changes induced by leptin and insulin

Several studies have reported significant changes in the structure of hippocampal dendrites and spines after activity-dependent synaptic plasticity (50). Rapid alterations in neuronal morphology have been observed in response to various hormones, which is thought to enable modification of neuronal connectivity and synaptic strength. Indeed, leptin-treated hippocampal neurons display marked alterations in dendritic filopodia compared with control neurons (51). Rapid increases in filopodial motility and density have been observed after short exposures to leptin in real-time imaging studies on enhanced green fluorescent protein-transfected hippocampal neurons (51). Previous studies indicate that structural changes in the actin cytoskeleton accompany hippocampal synaptic plasticity (52, 53). Similarly, leptin-driven hippocampal dendritic morphogenesis is associated with attenuated polymerized actin levels in proximal dendrites, suggesting that leptin promotes the redistribution of actin filaments to dendritic filopodia. In accordance with other studies (50, 54), NMDA receptor activation is required for the leptin-dependent structural changes because the effects were absent following NMDA receptor blockade. However, inhibition of GluN2B subunits failed to alter the leptin-driven structural changes, indicating the likely involvement of GluN2A-containing NMDA receptors (51). It is known that the activation of two major leptin-driven signaling pathways (31) mediates the enhancement of hippocampal NMDA receptor function by leptin (28). In contrast, however, only the MAPK (ERK) signaling cascade is required for the rapid structural changes induced in hippocampal neurons by leptin (51).

It is well documented that increased motility and extension of dendritic filopodia occurs during synaptogenesis (55, 56) and dendritic filpodia are actively involved in the development of new synaptic contacts (57). Thus, it is feasible that leptin-driven structural alterations in dendritic filopodia ultimately results in an increase in synaptic density. Indeed, leptin-treated hippocampal cultures display elevated levels of actin-rich spines and synapsin-1-positive puncta, consistent with an increase in the density of hippocampal synapses (51, 58) (see Figure 2). Leptin also promotes structural changes in other regions of the brain, including hypothalamus (16), cortex (59), and cerebellum (60). However, in contrast to the hippocampus, the morphological changes induced by leptin in hypothalamic, cortical, and cerebellar neurons are observed only after several hours of exposure to leptin (16, 59, 60).

Figure 2. The density of hippocampal synapses is increased by leptin.

Exposure of cultured hippocampal neurons (7DIC) to leptin (50 nM) for 30 minutes increased the density (A) and intensity (B) of synapsin-1 puncta (red). C, In older hippocampal neurons (14 DIC), 30 minutes of treatment with leptin (50 nM) not only increased the number of synapses (synapsin-1 positive immunostaining; red) but also the density of spines (arrows). [Modified from O'Malley, MacDonald N, Mizielinska S, Connolly CN, Irving AJ, Harvey J: Leptin promotes rapid dynamic changes in hippocampal dendritic morphology. Mol Cell Neurosci. 35:559–572, 2007 (51).]

Insulin

Numerous studies indicate that insulin has neurotrophic properties because widespread changes in neuronal survival and morphology are evident in response to insulin. For instance, treatment of organotypic hypothalamic explants with insulin markedly increased axonal growth (61). The density of neurites is also increased in sympathetic and sensory neurons after exposure to insulin and IGF-2 (62). In addition, endogenous insulin promotes axonal length in rat-cultured neurones via a MAPK-dependent process (63). However, exogenous insulin failed to alter axonal length, suggesting that endogenous and exogenous insulin may have different central roles (64). More recent studies have examined the role of insulin signaling cascades in altering neuronal structure. Overexpression of the insulin receptor substrate protein, IRSp53, leads to an increase in dendritic spine density in hippocampal neurons, whereas IRSp53 knockdown attenuates spine density (21). Dendritic spine density and complexity is also significantly reduced in cortical neurons from IGF-1 null mice (65). Although a growing body of evidence supports a neurotrophic role for insulin in the developing CNS, there is limited evidence of neurotrophic actions of insulin in adult brain.

Metabolic dysfunction and neurodegenerative disease

It is well documented that the functionality of metabolic hormonal systems decline with age. Deficiencies in metabolic processes are also associated with faster aging and an increased risk of developing neurodegenerative disease (66). Several studies indicate that, like other metabolic systems, alterations in leptin receptor-driven signaling occur during the aging process. Thus, a reduction in signal transducers and activators of transcription-3 activation is associated with reduced leptin responsiveness in aged rats (67). Leptin uptake into the hypothalamus is also reduced with age, an effect that is linked to diminished expression of leptin receptors (68). Recent evidence also suggests reduced neuronal leptin receptor function with age as the ability of leptin to enhance hippocampal excitatory synaptic strength is markedly attenuated at 12–14 months compared with 3–4 months of age (29) (see Figure 1).

It is well documented that age is a key risk factor for developing neurodegenerative diseases like AD. Consequently, the incidence of AD is growing rapidly because people are living longer. Diet and lifestyle have also been identified as important contributory factors, with midlife obesity associated with a significantly greater risk of developing AD. Because resistance to leptin is one of the main causes of midlife obesity, it is likely that resistance and/or impaired leptin function plays a significant role in AD. In support of a link between leptin levels and AD incidence, clinical studies have found that AD patients have markedly reduced circulating leptin levels (69). Moreover, APP_Swe and CRND8 rodent AD models display attenuated leptin levels (70), thereby supporting a correlation between leptin and neurodegeneration.

Accumulation of toxic amyloid-β_1–42 (Aβ) and subsequent plaque formation is a key pathological feature of AD, and build-up of Aβ has a detrimental effect during the early and late stages of this disease. Thus, acute exposure to Aβ triggers aberrant changes in hippocampal synaptic function that mirror the early synaptic impairments in AD. Neuronal viability is also drastically reduced after exposure to longer duration exposure to Aβ, which reflects the neuronal cell death occurring at later stages in the disease. Recent evidence indicates that leptin not only prevents the ability of Aβ to block hippocampal LTP, but it also reverses Aβ-induced facilitation of LTD (71). It is well established that trafficking of AMPA receptors to and away from synapses is pivotal for activity-dependent synaptic plasticity (72), and several studies have shown that exposure to Aβ interferes with AMPA receptor trafficking processes because Aβ promotes internalization of the AMPA receptor subunit, GluA1 (73). Treatment with the hormone leptin also influences this effect because the ability of Aβ to trigger removal of AMPA receptors from hippocampal synapses is significantly reduced in leptin-treated neurons (71). There is also evidence that leptin prevents the aberrant effects of Aβ on neuronal viability because exposure to this hormone increases the survival of cortical neurons treated with Aβ (71). In addition, leptin can reduce the generation and deposition of Aβ via direct inhibition of β-secretase activity and can enhance neuronal uptake of Aβ (74).

Another key pathological feature of AD is the formation of neurofibrillary tangles consisting of hyperphosphorylated-τ. Inappropriate activation of serine/threonine protein kinases such as glycogen synthase kinase (GSK)-3β and cyclin-dependent kinase 5 (cdk-5) (75) promote hyperphosphorylation of τ. As a consequence, misfolding and aggregation of τ occurs, resulting ultimately in the generation neurofibrillary tangles. Recent evidence indicates that the levels of phosphorylated τ are also controlled by leptin because exposure to leptin significantly attenuates build-up of τ in neurons (76), an effect that involves the leptin receptor-driven inhibition of GSK-3β (77). Phosphorylated τ levels are also markedly decreased in cortical neurons treated with leptin, whereas cortical levels of phosphorylated τ are enhanced in leptin-insensitive Zucker fa/fa rats (71).

In addition to attenuating AD pathology, treatment with leptin promotes significant cognitive improvements. In SAMP8 mice, with elevated levels of toxic Aβ, memory processing is improved after leptin treatment (70). Similarly, treatment of CRND8 mice with leptin results in better performance in novel object recognition paradigms as well as cued fear conditioning tests (70). Clinical evidence also supports a cognitive enhancing role for leptin in humans. Indeed, three adults with congenital leptin deficiency who were treated with leptin displayed marked increases in gray matter volume (78). A study by Paz-Filho et al (79) indicated a role for leptin in promoting cognitive development in children because leptin treatment of a 5-year-old boy with congenital leptin deficiency led to a significant improvement in neurocognitive skills.

Insulin and neurodegenerative disorders

Links between insulin and the development of neurodegenerative disorders such as AD are also well documented. Clinical studies have identified type 2 diabetes as an important risk factor for AD (80–82). Because a key feature of type 2 diabetes is resistance to insulin, impaired insulin signaling is thought to play a role in the development of neurodegenerative processes. Indeed, AD and insulin resistance share many features including attenuated insulin signaling, defective mitochondrial function, increased oxidative stress, and metabolic dysfunction (83). Moreover, studies of human postmortem tissue have identified insulin resistance as well as attenuated insulin receptor levels in the AD brain (84, 85). Central resistance to insulin has also been directly linked to cognitive decline in the early stages of AD (86).

Key pathological features of AD have also been linked to the insulin system. Indeed, both insulin and IGF are able to regulate τ gene expression and the levels of phosphorylated τ, whereas defective insulin signaling promotes activation of GSK-3β and subsequent hyperphosphorylation of τ (87). In addition, impaired insulin/IGF signaling interferes with amyloid precursor protein (APP) processing, leading to increased accumulation of Aβ (88). Central resistance to insulin is also reported to enhance deposition of toxic Aβ (89). At the cellular level, insulin increases extracellular secretion of Aβ by stimulating Aβ trafficking between the Golgi and plasma membrane (90) and also limiting intracellular breakdown by the insulin-degrading enzyme (91). In addition, Lee et al (92) demonstrated that insulin attenuates the aberrant effects of Aβ on synaptic function and neuron viability because treatment with insulin prevents the toxic effects of Aβ, and it rescues Aβ-induced impairment of hippocampal long-term potentiation (92, 93). Moreover, overexpression of IGF-2 in the hippocampus reverses the cognitive and synaptic deficits in APP transgenic mice (94).

Further evidence supporting a role for insulin-driven signaling in AD has come from studies examining IRS-1 signaling. Indeed, elevated levels of phosphorylated IRS-1 have been detected in brain tissue from AD patients, whereas defective hippocampal IRS-1 signaling is evident in murine models of AD (95). Moreover transgenic mice overexpressing active GSK-3 exhibit AD-like pathology including hyperphosphorylated τ and neuronal loss as well as significant cognitive deficits (96). The molecular parallels between diabetes and dysfunctional insulin-driven signaling in AD has provided strong rationale for developing therapeutic strategies that boost insulin signaling in the brain, and this has been the basis for testing insulin and antidiabetes drugs in AD. Indeed, clinical trials indicate that intranasal insulin treatment results in enhanced memory performance as well as metabolic integrity in AD patients (97–99). Moreover, recent studies have explored using the glucagon-like peptide 1 analog liraglutide, a common antidiabetic agent, in AD. Treatment with liraglutide not only reduces neuropathology but also improves cognition in a mouse model of AD (100).

Conclusions

It is well established that the central actions of the endocrine hormones, leptin and insulin, extend beyond their roles in regulating metabolic function. There is now extensive evidence from clinical and laboratory studies that leptin and insulin have cognitive-enhancing properties. In particular, both hormones have a marked influence on hippocampal synaptic function, with prominent effects on neuronal morphology, activity-dependent synaptic plasticity, and glutamate trafficking processes (Figure 3 and also summarized in Tables 1 and 2). The ability of both hormones to regulate key cellular events at synapses has important consequences for early stages of postnatal development and also during the aging process. Indeed, dysfunctions in the leptin or insulin systems are associated with not only pronounced neurodevelopmental abnormalities but also an increased risk of neurodegenerative disorders. Greater knowledge of the impact endocrine hormones like leptin and insulin have on hippocampal synaptic function is key to understanding their role in CNS-driven disease.

Figure 3. Regulation of hippocampal synaptic function by leptin and insulin.

Schematic representation of the main signaling pathways activated by leptin receptors (ObRs) and insulin receptors (IRs) and the key downstream cellular processes (synaptic plasticity, glutamate receptor trafficking, and neuronal structure) that are regulated by these hormonal systems, which in turn influence hippocampal synaptic function. Jak, Janus kinase; MEK, MAPK kinase.

Table 1.

Summary of the Key Reported Effects of the Metabolic Hormone, Leptin, on Hippocampal Synaptic Function

Effect on Hippocampal Synaptic Function	Leptin	Key References
Hippocampal synaptic plasticity	Facilitates LTP	Wayner at al (2004) (33)
	Converts STP to LTP	Shanley et al (2001) (28)
	Induces novel form of LTD at P14–21	Durakoglugil et al (2005) (37)
	Induces LTD at P5–8	Moult and Harvey (2011) (11)
	Induces LTP in adult and aged hippocampus	Moult et al (2010 (38); Moult and Harvey (2011) (11)
Glutamate receptor trafficking	Promotes synaptic incorporation of GluA1	Moult et al (2010) (38)
	Increases NMDA receptor trafficking	Harvey et al (2006) (35)
Neuronal morphology	Increase motility and density of dendritic filopodia	O'Malley et al (2007) (51)
	Increase density of hippocampal synapses and spines	O'Malley et al (2007) (51)
Neuronal development	Surge in leptin during early postnatal development (P8-P12)	Ahima et al (1998) (14)
	Leptin deficiency or insensitivity early in development causes brain abnormalities	Ahima et al (1999) (15)
AD	Interferes with Aβ production and uptake	Fewlass et al (2004) (74)
	Protects against neurotoxic effects of Aβ	Doherty et al (2013) (71)
	Prevents detrimental effects of Aβ on LTP and AMPA receptor trafficking	Doherty et al (2013) (71)
	Reduces levels of phosphorylated tau	Greco et al (2010) (76)
	Enhances cognition in murine AD models	Farr et al (2006) (70)

Abbreviation: STP, short-term potentiation.

Table 2.

Summary of the Key Reported Effects of the Metabolic Hormone, Insulin, on Hippocampal Synaptic Function

Effect on Hippocampal Synaptic Function	Insulin	Key References
Hippocampal synaptic plasticity	Enhances NMDA receptor function	Liao and Leonard (1999) (49)
	Induces novel form of LTD	Huang et al (2004) (45)
	Shifts the frequency response function for induction of LTD and LTP	van der Heide et al (2005) (46)
Glutamate receptor trafficking	Promotes synaptic removal of GluA2	Passafaro et al (2001) (47)
	Increases surface expression of GluA1	Man et al (2000) (44)
	Enhances exocytosis of NMDA receptors	Skeberdis et al (2001) (48)
Neuronal morphology	Overexpression of IRSp53 increases spine density	Choi et al (2005) (21)
	Knockdown of IRSp53 reduces hippocampal spine density
Neuronal development	High levels of IR detected early in development but this declines with age	Abbott et al (1999) (20)
	Stabilizes excitatory synaptic transmission during development	Choi et al (2005) (21)
AD	Type 2 diabetes and insulin resistance linked to AD	de la Monte et al (2009) (83); Talbot et al (2012) (86)
	Defects in IRS-1 linked to AD	Bomfim et al (2012) (95)
	Increases extracellular secretion of Aβ and limits intracellular breakdown	Watson et al (2003) (90); Gasparini et al (2002) (91)
	Reduces aberrant effects of Aβ on synaptic function and neuron viability	Lee et al (2009) (92)
	Intranasal insulin as an AD therapy	Reger et al (2006) (99); Reger et al (2008) (98)

Abbreviation: IR, insulin receptor.

Funding Statement

This work was supported by grants from The Cunningham Trust and the Anonymous Trust (to J.H.). G.M. is supported by a Biotechnology and Biological Sciences Research Council Eastbio studentship.