Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2015 Mar 24.
Published in final edited form as: Learn Mem. 2011 May 19;18(6):375–383. doi: 10.1101/lm.2111311

Insulin receptor substrate 2 is a negative regulator of memory formation

Elaine E Irvine 1,2,3, Laura Drinkwater 1, Kasia Radwanska 4, Hind Al-Qassab 2, Mark A Smith 2,4, Melissa O’Brien 4, Catherine Kielar 4, Agharul I Choudhury 2,3, Stefan Krauss 5, Jonathan D Cooper 4, Dominic J Withers 2,4, K Peter Giese 1,4
PMCID: PMC4371579  EMSID: EMS62703  PMID: 21597043

Abstract

Insulin has been shown to impact on learning and memory in both humans and animals, but the downstream signaling mechanisms involved are poorly characterized. Insulin receptor substrate-2 (Irs2) is an adaptor protein that couples activation of insulin- and insulin-like growth factor-1- receptors to downstream signaling pathways. Here, we have deleted Irs2, either in the whole brain or selectively in the forebrain, using the nestin Cre- or D6 Cre- deleter mouse lines respectively. We show that brain- and forebrain-specific Irs2 knockout mice have enhanced hippocampal spatial reference memory. Furthermore, NesCreIrs2KO mice have enhanced spatial working memory and contextual- and cued-fear memory. Deletion of Irs2 in the brain also increases PSD-95 expression and the density of dendritic spines in hippocampal area CA1, possibly reflecting an increase in the number of excitatory synapses per neuron in the hippocampus that can become activated during memory formation. This increase in activated excitatory synapses might underlie the improved hippocampal memory formation observed in NesCreIrs2KO mice. Overall, these results suggest that Irs2 acts as a negative regulator on memory formation by restricting dendritic spine generation.

Keywords: insulin receptor substrate 2, memory, hippocampus, synapses

Introduction

Insulin is best known for its action on peripheral insulin target tissues such as muscle, liver and adipocytes where it acts to regulate glucose homeostasis (Kahn and Flier 2000). In the central nervous system insulin signaling plays a much more diverse role, including modulation of food intake, weight control, regulation of synaptic plasticity, and involvement in neuronal survival, life span, neurological disorders and learning and memory (Burks et al. 2000; Park 2001; Schubert et al. 2004; Zhao et al. 2004; Choudhury et al. 2005; Plum et al. 2005; Steen et al. 2005; Selman et al. 2008; Taguchi and White 2008). In both human and animal experiments insulin has been reported to enhance memory formation (Park et al. 2000; Benedict et al. 2004; Moosavi et al. 2006; Babri et al. 2007; Moosavi et al. 2007; Benedict et al. 2008), whereas pharmacological disruption of the blood insulin supply by acute administration of streptozotocin (STZ) results in learning and memory deficits (Biessels et al. 1996). Furthermore, insulin deficiency has also been linked to development of aging-associated neurodegenerative disorders, such as Alzheimer’s disease where memory impairments are an obvious symptom (de la Monte 2009). The memory impairments observed in patients and STZ rodents with insulin deficiency can be reversed with insulin administration (Biessels et al. 1998; Reger et al. 2006; Dhamoon et al. 2009). However, there are a small number of studies where insulin has been shown to impair memory and these suggest that the behavioral response to insulin is dependent on dose and time of testing after administration (Schwarzberg et al. 1989; Kopf and Baratti 1996; Kopf and Baratti 1999). Moreover, recent studies in lower organisms have shown that abrogation of insulin signaling can either impair or enhance memory formation (Kauffman et al. 2010; Lin et al. 2010). These contradictory results highlight the further clarification needed regarding the role of insulin signaling in learning and memory.

Insulin receptor substrate (Irs) adaptor proteins act downstream of insulin and insulin-like growth factor-1 (Igf-1) receptors, and when phosphorylated bind via intracellular domains to specific signaling molecules. This results in the engagement of various downstream signaling cascades, such as the phosphatidylinositol-3 kinase and MAP kinase pathways (White 2002; Taniguchi et al. 2006). Insulin- and Igf-1-receptors are widely expressed throughout the CNS and are especially concentrated in brain regions that are crucially involved in learning and memory, such as the hippocampus and cerebral cortex (Zhao and Alkon 2001). Consistently, high expression of Irs1 and Irs2 are observed in both these areas (Allen Mouse Brain Atlas, Lein et al. 2007). Studies investigating the role of insulin signalling in the context of energy homeostasis regulation have suggested that much of its CNS actions are largely mediated via Irs2 (Withers et al. 1998; Burks et al. 2000; Choudhury et al. 2005; Taguchi et al. 2007). However, the role of Irs2 in leaning and memory is unknown.

To determine the contribution of Irs2 to memory formation we utilized two Irs2 brain-deleted mouse models, the NesCreIrs2KO (global-brain knockout) and D6CreIrs2KO (forebrain-specific knockout) mice. We measured the performance of mice generated from these lines in multiple learning and memory tasks. We also analyzed dendritic spine density and the expression of a marker of excitatory synapses, PSD-95 protein, in the stratum radiatum of hippocampal area CA1, which receives Schaffer collateral glutamatergic inputs. Our results suggest that Irs2 is a negative regulator of memory formation as both NesCreIrs2KO and D6CreIrs2KO mice have enhanced memory. Immunohistochemical analysis revealed that NesCreIrs2KO mice have increased PSD-95 expression and increased dendritic spine density, which may underlie the enhancement in memory formation.

Results

The NesCreIrs2KO mice are obese, hyperphagic and mildly glucose intolerant, which are all factors that could affect the general activity and well being of these mice (Choudhury et al. 2005). Therefore, motor activity and anxiety levels of young adult NesCreIrs2KO mice were assessed using the open field task. In this task, mice explored a brightly lit arena for a period of 10 min while their movements were monitored by a tracking system. The NesCreIrs2KO mice had normal activity in this task as they travelled the same distance, made the same number of entries and spent the same amount of time moving around the arena as their wild-type (WT) littermates [Distance, t43=0.5, not significant (NS); Entries, t43=0.4, NS; % time moving, t43=0.6, NS; Table 1]. Another aspect of this task is to assess anxiety levels, as mice that are more anxious will spend significantly less time in the bright centre of the box compared to the darker, safer periphery of the box. We found that NesCreIrs2KO mice had normal anxiety levels, as both genotypes spent the same percentage of time in the centre of the arena (t43=0.2, NS; Table 1). We next tested motor coordination using the accelerating rotarod. Mice were trained with 2 trials per day for 4 days and the latency to fall from the rod was recorded. Both genotypes improved over the course of the training period but there was no difference between genotypes (effect of day, F7,217=42.6, P<0.001; effect of genotype, F1,43=2.3, NS; genotype × day interaction, F7,217=0.8, NS; Table 1). These results suggested that the NesCreIrs2KO mice had normal motor function and anxiety levels making them a suitable model for investigating the role of Irs2 in learning and memory.

Table l.

Movement and anxiety behaviour in NesCreIrs2KO mice

WT (n = 20) NesCreIrs2KO (n = 25)
Open field behaviour
 Path length (m) 38.3 ± 2.6 36.3 ± 2.8
 Entries 357 ± 27 341 ± 30
 % Time moving 78.2 ± 1.5 77.0 ± 1.2
 % Time in centre 15.9 ± 1.1 16.3 ± 1.3
Rotarod performance
 Day 1 213 ± 13 239 ± 11
 Day 2 249 ± 11 267 ± 13
 Day 3 256 ± 10 286 ± 11
 Day 4 269 ± 9 281 ± 7
Open in a new tab

In order to assess the role of Irs2 in learning and memory we began by examining hippocampus-dependent spatial reference memory using the hidden-platform version of the Morris water maze (MWM) (Angelo et al. 2003; Antunes-Martins et al. 2007). Mice were trained for 8 days with 2 trials per day. Both genotypes improved in their latency to find the platform over the course of the training period but the NesCreIrs2KO mice had significantly improved acquisition compared to their WT littermates (effect of genotype, F1,40=6.7, P=0.01; effect of day, F7,280=33.3, P<0.001; genotype × day interaction, F7,280=0.6, NS; Fig. 1A). However, latency to find the platform during training is not the optimal measure of spatial memory as the mice may use alternative, hippocampus-independent strategies to locate the platform. Therefore to discriminate the spatial and non-spatial strategies a probe trial was given immediately after training on days 5 and 8. During the probe trial the platform was removed from the pool and the swim pattern of the mice was monitored with a tracking system. Mice that had formed a spatial memory spent most of their time swimming in the area where the platform was previously located. The probe trial given at the end of training on day 5 revealed that NesCreIrs2KO mice had improved spatial memory (Figs. 1B & C). Analysis of the quadrant search times showed that NesCreIrs2KO mice spent significantly more time in the target quadrant (TQ) compared to all other quadrants (F3,84=27.8, P<0.001), and compared to the time that their WT littermates spent in the TQ (F1,40=9.7, P<0.005; Fig. 1B). Analysis of the proximity measure confirmed that NesCreIrs2KO mice searched selectively (F3,84=25.1, P<0.001), and that they spent more time in close proximity to the former platform position than their WT littermates (F1,40=9.3, P<0.005 Fig. 1C). The improvement in NesCreIrs2KO mice was not due to improved task performance as there was no difference in swim speed (F1,40=1.6, NS; Fig. 1D) or visible platform escape latency (H1,33=2.1, NS; Fig. 1E) between the genotypes. After additional training both genotypes searched selectively in the TQ according to percentage search time (WT, F3,76=11.9, P<0.001; NesCreIrs2KO, F3,84=33.1, P<0.001; Fig. 1F) and proximity (WT, F3,476=11.5, P<0.001; NesCreIrs2KO, F3,84=27.0, P<0.001; Fig. 1G). The NesCreIrs2KO mice showed a trend towards an increased amount of time in the TQ and in close proximity to the former platform position than their WT littermates, but the difference did not reach significance (% Time, p=0.07; Proximity, p=0.09). Thus, brain-restricted loss of Irs2 enhances hippocampus-dependent spatial reference memory formation.

Figure 1.

Figure 1

Open in a new tab

NesCreIrs2KO mice have improved spatial memory in the Morris water maze. (A) NesCreIrs2KO mice (n=22) and their WT (n=20) littermates were trained in the hidden version of the water maze with 2 trials per day for 8 days. The NesCreIrs2KO mice were significantly improved in their ability to locate the platform during training. (B) A probe trial given on day 5 showed that NesCreIrs2KO mice had enhanced spatial memory as they searched selectively in the TQ, whereas their WT littermates did not. Knockout mice spent significantly more time searching in the TQ than their WT littermates. TQ, target quadrant, AR, adjacent right quadrant; AL, adjacent left quadrant; OP, opposite quadrant. (C) Furthermore, NesCreIrs2KO mice also searched in close proximity to the TQ platform position compared to all other platform positions, whereas their WT littermates did not. NesCreIrs2KO mice also searched in closer proximity to TQ than their WT littermates. (D) There was no significant difference in swim speed between the NesCreIrs2KO mice and their WT littermates. (E) After hidden-platform training the latency to locate a visible platform was similar between the genotypes (WT, n=17; NesCreIrs2KO, n=18). (F&G) During the probe trial on day 8 both genotypes searched selectively in the TQ and searched equally close to the platform position in the TQ. *** P<0.005; * P<0.05.

In order to overcome the hypothalamic body weight and glucose homeostasis phenotype observed in NesCreIrs2KO mice we generated a forebrain-restricted Irs2 knockout mouse using the D6 Cre deleter mouse that expresses Cre recombinase only in the forebrain (van den Bout et al. 2002). In situ hybridisation showed that Irs2 deletion in the D6CreIrs2KO mouse brain was restricted to the neocortex and areas CA1 and CA3 of the hippocampus while dentate gyrus expression was preserved (Fig. 2A). Consistent with this pattern of deletion, both male and female D6CreIrs2KO mice had normal body weight (Figs. 2B & C). Furthermore, glucose homeostasis was normal in these mice as shown by their normal fasting blood glucose levels and normal response to a bolus injection of glucose in a glucose tolerance test (Figs. 2D & E). When tested in the MWM, D6CreIrs2KO mice displayed normal acquisition (effect of genotype, F1,31=0.8, NS; effect of day, F7,217=42.6, P<0.001; genotype × day interaction, F7,217=0.8, NS; Fig. 3A) but a probe trial on day 5 revealed that they had improved spatial memory formation (Figs. 3B & C). Analysis of the quadrant search times showed that the D6CreIrs2KO mice spent significantly more time in the TQ compared to all other quadrants (F3,68=12.5, P<0.001), and compared to the time that their WT littermates spent in the TQ (F1,31=4.2, P<0.05; Fig. 3B). Analysis of the proximity measure confirmed that the D6CreIrs2KO mice searched selectively (F3,68=8.2, P<0.001), and that they spent more time in close proximity to the former platform position than their WT littermates (F1,31=4.8, P<0.05 Fig. 3C). Performance and visual acuity were unchanged in these mice as shown by swim speed and visible platform training (Swim speed, F1,31=2.2, NS; Visible platform, F1,8=2.7, NS; Figs. 3D & E). Both groups of mice searched equally in the TQ after additional training according to percentage search time (WT, F3,56=8.5, P<0.001; D6CreIrs2KO, F3,68=22.7, P<0.001; Fig. 3F) and proximity (WT, F3,56=4.2, P<0.01; D6CreIrs2KO, F3,68=16.8, P<0.001; Fig. 3G). In summary, these results show that total brain- and forebrain-restricted loss of Irs2 enhances spatial reference memory formation. While the D6CreIrs2KO mice dissociated behavioural from energy homeostasis phenotypes variable gene deletion occurred in this line (Methods and data not shown). Therefore, subsequent studies to refine the behavioral effects of loss of CNS Irs2 were performed using NesCreIrs2KO mice.

Figure 2.

Figure 2

Open in a new tab

Demonstration of forebrain-restriction of the Irs2 knockout in D6CreIrs2KO mice, and their body weight and energy homeostasis phenotype. (A) In situ hybridizations with adult sagittal brain sections showed that IRS2 expression is reduced in the forebrain of D6CreIrs2KO mice. Body weights for (B) male D6CreIrs2KO mice (n=13) and WT littermates (n=11) and (C) female D6CreIrs2KO mice (n=16) and WT littermates (n=16) were normal. (D) Fasting blood glucose levels (mmol/l) of 4-6 month old D6CreIrs2KO mice (n=6) and their WT littermates (n=6) following a 16 h fast were normal. (E) A glucose tolerance test performed on 4-6 month old D6CreIrs2KO mice (n=6) and their WT littermates (n=5) was normal.

Figure 3.

Figure 3

Open in a new tab

D6CreIrs2KO mice have improved spatial memory in the Morris water maze. (A) D6CreIrs2KO (n=18) and their WT (n=15) littermates were trained in the hidden-platform version of the water maze with 4 trials per day for 8 days. There was no significant difference in escape latency between the genotypes. (B) A probe trial given on day 5 showed that the D6CreIrs2KO mice had enhanced spatial memory as they searched selectively in the TQ, whereas their WT littermates did not. Knockout mice spent significantly more time searching in the TQ than their WT littermates. (C) Furthermore, D6CreIrs2KO mice also searched in close proximity to the TQ platform position compared to all other platforms, whereas their WT littermates did not. The D6CreIrs2KO mice also searched in closer proximity to the TQ than their WT littermates. (D) There was no significant difference in swim speed between the D6CreIrs2KO mice and their WT littermates. (E) After hidden-platform training the latency to locate a visible platform was not significantly different between the D6CreIrs2KO mice (n=6) and their WT littermates (n=4). (F&G) During the probe trial on day 8 D6CreIrs2KO mice and their WT littermates searched selectively in the TQ, and searched equally close to the platform position in the TQ. *** P<0.005; * P<0.05.

The cellular mechanisms and anatomical substrates underlying spatial reference memory and spatial working memory differ (Reisel et al. 2002). Therefore, to investigate whether loss of Irs2 also improves spatial working memory the NesCreIrs2KO mice were tested in spontaneous alternation in the T-maze. Spontaneous alternation is a hippocampus-dependent working memory task that relies on the mouse’s natural preference to spontaneously and flexibly shift between alternative spatial responses (Deacon and Rawlins 2006). The NesCreIrs2KO mice performed significantly better on this task than their WT littermates (H1,35=5.2, p<0.05; Fig. 4)

Figure 4.

Figure 4

Open in a new tab

Enhanced spatial working memory in NesCreIrs2KO mice. NesCreIrsKO mice (n=21) and their WT littermates (n=16) were tested for spontaneous alternation in the T-maze. * P<0.05.

We next studied NesCreIrs2KO mice in hippocampus- and amygdala-dependent contextual fear conditioning and amygdala-dependent cued fear conditioning (Anagnostaras et al. 2001). In this task, animals learn to associate an aversive stimulus (electrical foot shock) with the location in which it occurred (contextual fear), or with a tone presentation (cued fear), and memory is scored as a freezing response to re-presentation of the context or cue. The mice were trained with three tone-shock pairings and then tested for contextual fear memory at 24 hours, and cued fear memory at 48 hours, after training. The NesCreIrs2KO mice had improved contextual fear memory as they froze significantly more than their WT littermates when re-exposed to the contextual chamber (F1,18=8.2, P<0.01; Fig. 5A). Furthermore, NesCreIrs2KO mice had improved cued fear memory, as they froze significantly more on tone presentation than their WT littermates (effect of genotype, F1,18=15.9, P<0.001; effect of time, F1,18=76.7, P<0.001; genotype × time interaction, F1,39=11.6, P<0.005; Fig. 5B). Importantly, the NesCreIrs2KO mutation did not alter footshock sensitivity as demonstrated by the similar awareness (H1,10=0.3, NS), vocalization (F1,10=0.03, NS) and jumping (F1,10=0.2, NS) thresholds to their WT littermates (Fig. 5C). Taken together, all these behavioral data show that neuronal loss of Irs2 throughout the brain enhances memory formation.

Figure 5.

Figure 5

Open in a new tab

NesCreIrs2KO mice have improved contextual and cued fear memory. Mice were trained with three tone/shock pairings. (A) NesCreIrs2KO mice were improved in contextual fear conditioning as they froze significantly more than their WT littermates when exposed to the conditioning chamber 24 h after training (WT, n=10; NesCreIrs2KO, n=10). (B) NesCreIrs2KO mice were improved in cued fear conditioning as they froze significantly more to the tone presentation when tested 48 h after training. (C) This enhancement in contextual and tone fear conditioning was not due to an alteration in pain sensitivity as NesCreIrs2KO mice had normal shock reactivity (WT, n=6; NesCreIrs2KO, n=6). *** P<0.005 and ** P<0.01.

It is a widely held belief that the cellular basis of learning and memory occurs at the synapse (Bhatt et al. 2009), and both insulin signaling and hippocampal memory formation have been implicated in synaptic modifications that regulate synaptic strength and plasticity (Leuner and Shors 2004; Zhao et al. 2004). PSD-95 is located in dendritic spines and is a key postsynaptic scaffold protein at excitatory synapses that directly promotes synapse maturation and exerts a major influence on synaptic strength and plasticity (Keith and El-Husseini 2008). In order to elucidate the neurobiological underpinnings of the enhancement of memory formation in NesCreIrs2KO mice we initially used immunohistochemical analysis to determine the expression levels of PSD-95 in the stratum radiatum of hippocampal area CA1. We observed that both naïve and trained NesCreIrs2KO mice had significantly more PSD-95 puncta in CA1 stratum radiatum than WT mice (effect of genotype, P<0.01; Fig. 6A). However, the density of PSD-95 puncta did not change after conextual fear conditioning for both genotypes (Fig. 6A), consistent with an earlier study showing that spine density in hippocampal area CA1 does not change after contextual fear conditioning (Matsuo et al. 2008). Next, we studied the average size of the PSD-95 puncta after contextual fear conditioning (Fig. 6B). There was no difference between naïve NesCreIrs2KO and WT mice, but for both genotypes the size of the PSD-95 puncta increased with training (effect of training, P<0.001. The observed increase in PSD-95 puncta in NesCreIrs2KO mice was not due to an increase in neuronal number in area CA1 of the hippocampus as neuronal number was comparable between the two genotypes (Fig. 6C). This result suggested that there may be a change in the number and/or density of spines on the dendrites of the CA1 pyramidal neurons. Dendritic spines are the major sites of synapse formation where more than 90% of the excitatory synapses in the central nervous system are generated (Bhatt et al. 2009). It has been suggested that changes in spine density can regulate memory formation by determining the number of synapses contributing to the memory formation (Leuner and Shors 2004). Accordingly, mice with increased spine density have been shown to have enhanced memory (Li et al. 2004; Liu et al. 2008). Whole-cell electrophysiology recordings were used to fill single hippocampal CA1 pyramidal neurons, from naïve WT and NesCreIrs2KO mice, with Lucifer yellow in order to visualize the dendritic arbor and spine distribution. Sholl analysis of the dendritic arbor revealed that the number of dendritic branches was not altered in naïve NesCreIrs2KO mice compared to their WT littermates (Figs. 6D & 6E). However, the density of dendritic spines along individual dendrites of hippocampal CA1 pyramidal neurons was significantly increased in the NesCreIrs2KO mice (Fig. 6F), suggesting that Irs2 suppresses spine formation in this area..

Figure 6.

Figure 6

Open in a new tab

NesCreIrs2KO mice have increased PSD-95 expression and dendritic spines density in the stratum radiatum of the hippocampal CA1 subfield. (A) NesCreIrs2KO mice (Naïve, n=6; Trained, n=5) have an increased density of PSD-95 puncta in the stratum radiatum of the hippocampal CA1 subfield compared to their WT littermates (Naïve, n=5; Trained, n=6). (B) The size of the PSD-95 puncta was increased after training in both groups. (C) The CA1 neuronal number was not altered in the NesCreIrs2KO mice. (D) Hippocampal pyramidal neurons of WT and NesCreIrs2KO mice were filled with lucifer yellow and stacked confocal images were complied. Scale bar = 60 μm. Panels on the right show a magnified view of the apical oblique dendritic segments that are highlighted for each genotype on the left. Scale bar = 5 μm (E) NesCreIrs2KO mice have normal dendritic branching (N=6 mice per genotype; 9 neurons were quantified). (F) NesCreIrs2KO mice have an increased density of dendritic spines (N=6 mice per genotype; 9 neurons were quantified). ** P< 0.01; * P<0.05.

Discussion

There is an increasing body of evidence suggesting that insulin and insulin/Igf-1 signaling contribute to cognitive function in humans and animals (Park et al. 2000; Park 2001; Zhao and Alkon 2001; Benedict et al. 2004; Zhao et al. 2004; Benedict et al. 2008). Based on these reports we predicted that deletion of Irs2 in the brain would impair learning and memory. However, our behavioral studies show that loss of Irs2 in whole brain or selectively in the forebrain enhances memory formation in a number of hippocampus-dependent tasks. We also found that whole brain deletion of Irs2 increases dendritic spine density and PSD-95 puncta in hippocampal area CA1 of naïve mice, whilst the total neuronal number in this region was unaltered. Together, these findings suggest that an increase in activated excitatory synapses might underlie the improved hippocampal memory formation observed in mice with brain deletion of Irs2.

These findings might appear at odds with the prevailing view that insulin/Igf-1 signaling primarily plays a positive role in memory function. However, it is increasingly recognized in lower organisms and rodents that abrogation of insulin/Igf-1 signaling can have beneficial effects on age-related phenotypes including motor and memory function as well as increasing longevity (Taguchi et al. 2007; Selman et al. 2008; Kauffman et al. 2010). In C. elegans the insulin/Igf-1 receptor is encoded by the daf-2 gene and worms with this gene deleted have improved short- and long-term memory formation in early adulthood and they retain the ability to learn better with age (Kauffman et al. 2010). The improved memory formation seen in our studies is consistent with this observation and suggests that specific targeting of insulin signaling events may have beneficial effects.

Previous studies have suggested that the number and size of PSD-95 puncta correlate with the number and size of postsynaptic densities in spines (Aoki et al. 2001; Fedulov et al. 2007). We found that neuronal loss of Irs2 increases PSD-95 puncta as well as spine density in stratum radiatum of area CA1 in the hippocampus. Various conditions, such as environmental enrichment, stress and exposure to estrogen, are known to alter spine density in the stratum radiatum (Leuner and Shors 2004). Furthermore, an increase in spine density has been shown to correlate with enhanced hippocampal memory formation (Leuner and Shors 2004). For example, Li and colleagues (2004) showed that estrogen administered to C57BL/6J mice increased spine density and PSD-95 expression throughout the hippocampus and enhanced performance of object placement, a hippocampal memory task. Our PSD-95 immunohistochemical and dendritic spine analyses substantiate these data as NesCreIrs2KO mice have an increase in PSD-95 and spine density with a concomitant enhancement in hippocampal memory formation. It is therefore possible that this increased spine density in the mutants may reflect enhanced synaptic connectivity which could serve to facilitate memory formation.

A possible explanation for the improvement in memory and increase in spine density could be due to a compensatory mechanism of another signaling pathway. The NesCreIrsKO mice are hyperleptinaemic yet retain leptin sensitivity (Choudhury et al. 2005), which suggests that CNS Irs2 pathways are not required for leptin action. Leptin is known to circulate in the plasma and is transported to the brain via saturable transport across the blood–brain barrier (Banks 2008a; Banks 2008b). Leptin receptors have been found in several brain regions including all areas of the hippocampus and direct administration of leptin into this area improves memory formation in rodents (Farr et al. 2006). It is therefore possible that the increased levels of leptin circulating in NesCreIrs2KO mice contribute to the enhancement in memory formation observed in these mice. Furthermore, there is evidence that leptin increases both the motility and density of dendritic filipodia in hippocampal neurons (O’Malley et al. 2007). However, follow-up studies are required to address this hypothesis.

Taken together, our findings show that in mice Irs2 is a negative regulator of memory formation which may be due to a suppressive effect on synaptogenesis. Since the vast majority of mammalian studies have shown that insulin enhances memory formation, we conclude that other components of the insulin signaling pathway enhance memory formation and that they compete with Irs2 for the efficiency of memory formation.

Materials and Methods

Animals

The subjects were housed in groups of two to eight and maintained on a 12 h light/dark cycle with food and water ad libitum. The NesCreIrs2KO mice were generated and PCR genotyped as previously described (Choudhury et al. 2005). The D6CreIrs2KO mice were generated by crossing D6 Cre deleter mice (van den Bout et al. 2002) with Irs2 floxed mice. Genotyping was carried out by PCR analysis: To genotype for floxed Irs2 alleles primers around the 3′ loxP site were used: forward 5′ ACT TGA AGG AAG CCA CAG TCG 3′ reverse 5′AGT CCA CTT TCC TGA CAA GC 3′. To genotype for the presence of the Cre transgene the primers were: forward Cre31 5′ TAC ATT GGT CCA GCC ACC 3′ and reverse Cre51: 5′CGA TGC AAC GAG TGA TGA G 3′. All of the mice for the behavioral studies were 3-7 months at the time of training. All experiments used approximately equal numbers of male and female mice, and were undertaken in accordance with the UK Animals (Scientific Procedures) Act 1986.

Open Field

Open field activity was assessed at 3-4 months of age using the HVS Image tracking software (HVS Image 2100, Hampton, UK). Mice were individually placed in a wooden grey arena (45 × 45 × 30 cm) that had a base covered in sawdust. Each mouse was released into a corner of the box and was allowed to explore for 10 min. The tracking system recorded path length, % time moving and % time spent in the center.

Rotarod

The rotarod apparatus (Accelerating Model, Ugo Basile, Italy) was used to measure fore- and hind-limb motor coordination and balance. Mice were trained at 4 months of age and received three trials per day with an inter-trial interval of 1 h for three consecutive days. The rod accelerated from 5 to 60 rpm over a period of 600 s and the latency to fall was recorded.

Morris Water Maze

Mice were studied in a set-up that has been shown to be hippocampus-dependent for male and female WT mice (Angelo et al. 2003; Antunes-Martins et al. 2007). Before training, the mice were handled each day for 2 min daily for 5-7 days, in order to reduce their anxiety levels. Mice were acclimatised to the dim light conditions for 30-60 min before testing began. The pool was 1.5 m in diameter, with a platform of 0.1 m diameter positioned 0.5 cm below the surface of the water. The water was maintained at 24-28°C throughout the trials, and made opaque by adding non-toxic white paint so that the platform was not visible. Mice were trained with either 2 (for NesCreIrs2KO mice) or 4 (for D6CreIrs2KO mice) training trials per day for 8 days. Each trial started from a position of the pool selected pseudorandomly. The maximal trial length was 90 s, and the inter-trial interval was 60 s. In order to assess spatial memory formation probe trials were given at the end of days 5 and 8. For the probe trial the platform was removed from the pool and the mouse was put in the swimming pool at the opposite quadrant and allowed to swim for 60 s. The movement of the mice was recorded by a video tracking system (HVS Image, Hampton, UK). After completion of the hidden-platform training some of the mice were tested in the visible platform version of this task in order to test visual acuity. White curtains to hide cues surrounded the swimming pool and the platform was made visible with a flag. Two 90 s trials were given with an inter-trial interval of 60 s.

Spontaneous Alternation in the T-maze

Each mouse was placed individually in the start-arm of the T-maze and allowed to explore. Once the mouse entered a goal-arm a sliding door was closed preventing the mouse leaving the goal-arm. The arm entered (left or right) was recorded and the mouse was allowed to explore that goal-arm for 30 s. The mouse was then taken out of the maze and 10 s later was returned to the start-arm, with all doors open again, and allowed to explore again. The goal-arm entered on the second run was recorded and the mouse was returned to its home cage. Each mouse was tested twice each day for a total of 8 trials. The percentage of trials on which the mouse entered a different goal-arm on the second run (i.e. alternated) was calculated.

Contextual and Cued Fear Conditioning

Background fear conditioning has two components that can be tested for, hippocampus- and amygdala-dependent contextual fear conditioning and amygdala-dependent cued fear conditioning (Anagnostaras et al. 2001). Conditioning took place in a conditioning chamber (Med Associates Inc, St Albans, USA) that was situated in a soundproof box. The conditioning chamber floor was made up of stainless steel rods that were used for shock delivery. A speaker was mounted on one side of the chamber for delivery of the tone (80 dB, 3.0 kHz). In order to camouflage any noise in the behavioral room background noise was supplied to the chamber by a white noise generator positioned in the side of the soundproof box. Prior to training the chamber was cleaned with 70% ethanol and a paper towel soaked in ethanol was placed under the grid floor. The tone testing was conducted in a novel chamber that was structurally different from the conditioning chamber. This chamber was semi-circular, had a plastic floor and was lit by red light. Prior to tone test, the chamber was cleaned with a lemon-scented solution.

On the conditioning day, the mice were brought from the housing room into a holding room where they were allowed to acclimatize for 30 min before training. Mice were then placed individually in the conditioning chamber and after a 120 s introductory period a tone (80 dB, 3.0 kHz) was presented for 30 s, the last 2 s of which coincided with a footshock (0.7 mA). Mice received a further two tone/shock pairings, each separated by a 60 s inter-trial interval (ITI). The mice were removed from the conditioning chamber 60 s after the last shock presentation and returned to their home cage. Once all mice had been trained they were returned to the housing room.

Contextual fear memory was tested 24 h after training by re-exposing to the conditioning chamber for 5 min. Cued fear memory was tested 48 h after training by re-exposing the mice to a novel chamber for 3 min without tone presentation, followed by 3 min with a tone presentation.

A video camera was fixed inside the door of the soundproof box, which allowed the behavior to be observed and scored by an experimenter blind to the genotype of the mice. Freezing behaviour (defined as complete lack of movement, except for respiration) was scored for 2 s in every 5 s.

In order to test for sensory and pain thresholds mice were placed into the conditioning chamber and the shock intensity was increased from 0.05 mA, in 0.05 mA intervals, until the mice showed the first signs of becoming aware of the shock (moving backwards and flicking with the hind legs) and “pain” (vocalization and jumping). A recording of the shock intensity was taken when each mouse became aware of the shock and vocalized or jumped in response to the shock.

In Situ Hybridisation

Immediately after dissection brains were frozen in isopentane cooled to −20 to −30°C, then transferred to dry ice. Coronal brain sections (15 μm) were cut on a cryostat and mounted on Superfrost slides (VWR). Sections were fixed in 4% paraformaldehyde/phosphate-buffered saline (PBS; pH 7.4) for 5 min on ice, and stored at 4°C in 95% ethanol.

A 40-mer oligodeoxynucleotide complementary to the Irs2 mRNA, 3′-AGCTTGGAGCCACACCACATTCGCATGTACCCACTGTCTT-5′ was synthesised by Invitrogen (Paisley, UK) and was end-labeled with [α-35S] dATP (12.5 μCi/μl; Amersham Pharmacia Ltd) using terminal deoxytransferase (Promega). Air-dried slides were hybridized at 42°C for 16-18 h in a humidifying chamber with 65 μl of hybridization buffer [50% (v/v) formamide, 4× SSC, pH 7.0, 25 mM sodium phosphate, pH 7.0, 1 mM sodium pyrophosphate, 20 mM DTT, 2× Denhardt’s solution, 200 μg/ml heat-denatured salmon sperm DNA, 10% (w/v) dextran sulfate, and 100,000-300,000 cpm 35S-labeled probe] per slide. Sections were washed twice in 1× SSC at 55°C for 30 min, transferred through 0.1× SSC, 70% ethanol, and 95% ethanol, and air dried. The sections were exposed to a 35S-sensitive film for autoradiography together with 14C microscale standards (Amersham Biosciences) for 30 days at room temperature. The autoradiograph of every brain section was imaged with a monochrome camera. The resultant images were calibrated (nCi/g) with reference to the 14C standards [nCi/g of tissue equivalent (TE)], and the intensity of the signal was quantified/analyzed using the MCID M5+ image analysis system (Imaging Research, St. Catherines, Ontario, Canada).

PSD-95 Immunostaining

Anaesthetised mice were perfused with ice-cold saline followed by 4% PFA (Sigma) in saline. Brains were stored overnight in the same fixing solution, and then for 3 days in 30% sucrose in PBS. Coronal brain sections (40 μm) were cut on a cryostat (Microm HM560) and collected into cryoprotectant solution (30% sucrose, 30% ethylene glycol in TBS). The brain sections were stored at −20°C until used for immunohistochemical staining.

Brain sections were washed three times in PBS/0.3% Triton X-100 for 6 min each time. Next, the sections were incubated for 2 h at room temperature in a blocking solution containing 5% normal goat serum (Vector Lab. Ltd.) in PBS/0.3% Triton X-100. After blocking, the tissue was placed in the same solution to which the PSD-95 antibody (Abcam, ab18258) was added at a final concentration of 1:200 and incubated overnight at 4°C. After washing three times in PBS/0.3% Triton X-100, the sections were incubated with the secondary antibody (biotinylated anti-rabbit, Vector Lab. Ltd.) in a PBS/0.3% Triton solution at an antibody concentration of 1:500 for 1 h at room temperature, followed by incubation with streptavidin conjugated Alexa647 (1:200; Invitrogen, UK) for 1 h at room temperature. After washing in PBS the sections were mounted on slides and covered with Vectashield HardSet Mounting Medium with DAPI (Vector Lab. Ltd).

Immunostaining was analyzed using an AxioImager Z1 microscope, with Plan-Apochromat 63× objective and ApoTome (Zeiss). 4 Z-stack microphotographs (10 pictures per stack, every 0.5 μm) were taken with AxioCamMT3 M27 per brain section; every 6th section through dorsal hippocampus (stratum radiatum of CA1 field; Bregma −1.58 to −2.18 mm) was analyzed (on average 5 sections per animal). Z-stacks were reconstructed with ImageJ (NIH), and processed with ‘find edges’ and ‘max projection’ (‘grouped ZProjector’) functions. Density and size of stained puncta (threshold function: 90-255) were analyzed with ImageJ. The analysis excluded puncta, which were on the border of visibility. Five to six animals per experimental group were analyzed.

Counting of CA1 neurons

We used StereoInvestigator software (MicroBrightField Bioscience, Williston, VT) to obtain unbiased optical fractionator estimates of cell numbers from Nissl stained sections. Cells were sampled using a series of counting frames distributed over a grid superimposed onto the section, with a random starting section chosen, followed by every 6th section through dorsal hippocampus thereafter. The sizes of the ‘sampling grid’ and of the ‘dissector frame’ used in this study were 250 × 100 μm2 and 30 × 30 μm2, respectively. Only clearly identifiable cells that fell within the dissector frame were counted, using a ×100 oil objective. The mean coefficient of error (CE) for all individual optical fractionator estimates was calculated according to the method of Gundersen and Jensen (1987) and was less than 0.08 in this analysis.

Dendritic Branching and Spine Morphology

Six month old female NesCreIrs2KO mice and WT littermates were killed by cervical dislocation and brains rapidly transferred to an ice-cold slicing solution containing 2.5 mM KCl, 1.25 mM NaH2PO4, 28 mM NaHCO3, 0.5 mM CaCl2, 7 mM MgCl2, 7 mM D-glucose, 235 mM sucrose and perfused with 95% O2/5% CO2, pH 7.4. Sagital brain slices (350-μm) containing the hippocampus were prepared using a Vibratome Series 1000, as previously described (Smith et al. 2007). Slices were maintained at 33-35°C in an external solution containing in mM: 125 NaCl, 2.5 KCl, 1.25 NaH2PO4, 25 NaHCO3, 2 CaCl2, 1 MgCl2, 10 D-glucose, 15 D-mannitol. Hippocampal CA1 pyramidal neurons were identified by video-enhanced differential-interference contrast using an upright Zeiss Axioskop 2 FS plus microscope. Whole-cell voltage-clamp recordings were made with patch pipettes (5–8 MΩ) containing 130 mM potassium gluconate, 10 mM KCl, 0.5 mM EGTA, 10 mM HEPES, 1 mM NaCl, 0.28 mM CaCl2, 3 mM MgCl2, 3 mM Na2ATP, 0.3 mM tris-GTP, 14 mM phosphocreatine and 0.5 mg ml−1 Lucifer yellow (pH 7.2). Following 10-15 minutes of intracellular dialysis of healthy neurons (membrane potential between −65 and −75 mV), patch electrodes were removed and slices fixed prior to immunohistochemistry for Lucifer yellow (1:500 dilution of rabbit anti-Lucifer, Invitrogen) followed by a rabbit secondary antibody conjugated to Alexa Fluor 488, as previously described (Karadottir and Attwell 2006). A Leica SPE inverted confocal microscope was used to excite the Alexa Fluor 488 and generate z-stacked (0.25-0.5 mm steps) images. Spine density was counted from apical trunk and radial oblique dendrites that were located ~80-160 μm distal from the soma in stratum radiatum. Mean spine density was calculated as a function of dendritic length (>100 spines per filled neuron). Differences in dendritic arborisation were examined by Sholl analysis by which the numbers of apical or basal dendrites that intercrossed concentric circles radiating from the soma (15 mm increments) were determined in ImageJ.

Statistical Analysis

All data were analysed using SigmaStat statistics software. For statistical analysis Student’s t-test, one-way analysis of variance, one-way analysis of variance on ranks (if data were not distributed normally), two-way analysis of variance with repeated measures, and Student-Newman-Keuls multiple comparison post hoc tests were used when appropriate.

Acknowledgements

We thank Drs Jeff Vernon and Florian Plattner for helpful advice during the course of these studies. The work was funded by the Biotechnology and Biological Sciences Research Council IABB Initiative (to DJW and KPG) and an MRC grant (to KPG).

References

  1. Anagnostaras SG, Gale GD, Fanselow MS. Hippocampus and contextual fear conditioning: recent controversies and advances. Hippocampus. 2001;11:8–17. doi: 10.1002/1098-1063(2001)11:1<8::AID-HIPO1015>3.0.CO;2-7. [DOI] [PubMed] [Google Scholar]
  2. Angelo M, Plattner F, Irvine EE, Giese KP. Improved reversal learning and altered fear conditioning in transgenic mice with regionally restricted p25 expression. Eur J Neurosci. 2003;18:423–431. doi: 10.1046/j.1460-9568.2003.02746.x. [DOI] [PubMed] [Google Scholar]
  3. Antunes-Martins A, Mizuno K, Irvine EE, Lepicard EM, Giese KP. Sex-dependent up-regulation of two splicing factors, Psf and Srp20, during hippocampal memory formation. Learn Mem. 2007;14:693–702. doi: 10.1101/lm.640307. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Aoki C, Miko I, Oviedo H, Mikeladze-Dvali T, Alexandre L, Sweeney N, Bredt DS. Electron microscopic immunocytochemical detection of PSD-95, PSD-93, SAP-102, and SAP-97 at postsynaptic, presynaptic, and nonsynaptic sites of adult and neonatal rat visual cortex. Synapse. 2001;40:239–257. doi: 10.1002/syn.1047. [DOI] [PubMed] [Google Scholar]
  5. Babri S, Badie HG, Khamenei S, Seyedlar MO. Intrahippocampal insulin improves memory in a passive-avoidance task in male wistar rats. Brain Cogn. 2007;64:86–91. doi: 10.1016/j.bandc.2007.01.002. [DOI] [PubMed] [Google Scholar]
  6. Banks WA. The blood-brain barrier as a cause of disease. Curr Pharm Des. 2008a;14:1553–1554. doi: 10.2174/138161208784705478. [DOI] [PubMed] [Google Scholar]
  7. Banks WA. The blood-brain barrier: connecting the gut and the brain. Regul Pept. 2008b;149:11–14. doi: 10.1016/j.regpep.2007.08.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Benedict C, Hallschmid M, Hatke A, Schultes B, Fehm HL, Born J, Kern W. Intranasal insulin improves memory in humans. Psychoneuroendocrinology. 2004;29:1326–1334. doi: 10.1016/j.psyneuen.2004.04.003. [DOI] [PubMed] [Google Scholar]
  9. Benedict C, Kern W, Schultes B, Born J, Hallschmid M. Differential sensitivity of men and women to anorexigenic and memory-improving effects of intranasal insulin. J Clin Endocrinol Metab. 2008;93:1339–1344. doi: 10.1210/jc.2007-2606. [DOI] [PubMed] [Google Scholar]
  10. Bhatt DH, Zhang S, Gan W-B. Dendritic spine dynamics. Annu Rev Physiol. 2009;71:261–282. doi: 10.1146/annurev.physiol.010908.163140. [DOI] [PubMed] [Google Scholar]
  11. Biessels GJ, Kamal A, Ramakers GM, Urban IJ, Spruijt BM, Erkelens DW, Gispen WH. Place learning and hippocampal synaptic plasticity in streptozotocin-induced diabetic rats. Diabetes. 1996;45:1259–1266. doi: 10.2337/diab.45.9.1259. [DOI] [PubMed] [Google Scholar]
  12. Biessels GJ, Kamal A, Urban IJ, Spruijt BM, Erkelens DW, Gispen WH. Water maze learning and hippocampal synaptic plasticity in streptozotocin-diabetic rats: effects of insulin treatment. Brain Res. 1998;800:125–135. doi: 10.1016/s0006-8993(98)00510-1. [DOI] [PubMed] [Google Scholar]
  13. Burks DJ, Font de Mora J, Schubert M, Withers DJ, Myers MG, Towery HH, Altamuro SL, Flint CL, White MF. IRS-2 pathways integrate female reproduction and energy homeostasis. Nature. 2000;407:377–382. doi: 10.1038/35030105. [DOI] [PubMed] [Google Scholar]
  14. Choudhury AI, Heffron H, Smith MA, Al-Qassab H, Xu AW, Selman C, Simmgen M, Clements M, Claret M, Maccoll G, et al. The role of insulin receptor substrate 2 in hypothalamic and beta cell function. J Clin Invest. 2005;115:940–950. doi: 10.1172/JCI24445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. de la Monte SM. Insulin resistance and Alzheimer’s disease. BMB Rep. 2009;42:475–481. doi: 10.5483/bmbrep.2009.42.8.475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Deacon RMJ, Rawlins JNP. T-maze alternation in the rodent. Nat Protoc. 2006;1:7–12. doi: 10.1038/nprot.2006.2. [DOI] [PubMed] [Google Scholar]
  17. Dhamoon MS, Noble JM, Craft S. Intranasal insulin improves cognition and modulates beta-amyloid in early AD. Neurology. 2009;72:292–293. doi: 10.1212/01.wnl.0000344246.91081.2c. author reply 293-294. [DOI] [PubMed] [Google Scholar]
  18. Farr SA, Banks WA, Morley JE. Effects of leptin on memory processing. Peptides. 2006;27:1420–1425. doi: 10.1016/j.peptides.2005.10.006. [DOI] [PubMed] [Google Scholar]
  19. Fedulov V, Rex CS, Simmons DA, Palmer L, Gall CM, Lynch G. Evidence that long-term potentiation occurs within individual hippocampal synapses during learning. J Neurosci. 2007;27:8031–8039. doi: 10.1523/JNEUROSCI.2003-07.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gundersen HJ, Jensen EB. The efficiency of systematic sampling in stereology and its prediction. J Microsc. 1987;147:229–263. doi: 10.1111/j.1365-2818.1987.tb02837.x. [DOI] [PubMed] [Google Scholar]
  21. Kahn BB, Flier JS. Obesity and insulin resistance. J Clin Invest. 2000;106:473–481. doi: 10.1172/JCI10842. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Karadottir R, Attwell D. Combining patch-clamping of cells in brain slices with immunocytochemical labeling to define cell type and developmental stage. Nat Protoc. 2006;1:1977–1986. doi: 10.1038/nprot.2006.261. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kauffman AL, Ashraf JM, Corces-Zimmerman MR, Landis JN, Murphy CT. Insulin signaling and dietary restriction differentially influence the decline of learning and memory with age. PLoS Biol. 2010;8:e1000372. doi: 10.1371/journal.pbio.1000372. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Keith D, El-Husseini A. Excitation Control: Balancing PSD-95 Function at the Synapse. Front Mol Neurosci. 2008;1:4. doi: 10.3389/neuro.02.004.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kopf SR, Baratti CM. Memory modulation by post-training glucose or insulin remains evident at long retention intervals. Neurobiol Learn Mem. 1996;65:189–191. doi: 10.1006/nlme.1996.0020. [DOI] [PubMed] [Google Scholar]
  26. Kopf SR, Baratti CM. Effects of posttraining administration of insulin on retention of a habituation response in mice: participation of a central cholinergic mechanism. Neurobiol Learn Mem. 1999;71:50–61. doi: 10.1006/nlme.1998.3831. [DOI] [PubMed] [Google Scholar]
  27. Lein ES, Hawrylycz MJ, Ao N, Ayres M, Bensinger A, Bernard A, Boe AF, Boguski MS, Brockway KS, Byrnes EJ, et al. Genome-wide atlas of gene expression in the adult mouse brain. Nature. 2007;445:168–176. doi: 10.1038/nature05453. [DOI] [PubMed] [Google Scholar]
  28. Leuner B, Shors TJ. New spines, new memories. Mol Neurobiol. 2004;29:117–130. doi: 10.1385/MN:29:2:117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Li C, Brake WG, Romeo RD, Dunlop JC, Gordon M, Buzescu R, Magarinos AM, Allen PB, Greengard P, Luine V, et al. Estrogen alters hippocampal dendritic spine shape and enhances synaptic protein immunoreactivity and spatial memory in female mice. Proc Natl Acad Sci USA. 2004;101:2185–2190. doi: 10.1073/pnas.0307313101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lin CHA, Tomioka M, Pereira S, Sellings L, Iino Y, van der Kooy D. Insulin signaling plays a dual role in Caenorhabditis elegans memory acquisition and memory retrieval. J Neurosci. 2010;30:8001–8011. doi: 10.1523/JNEUROSCI.4636-09.2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Liu F, Day M, Muniz LC, Bitran D, Arias R, Revilla-Sanchez R, Grauer S, Zhang G, Kelley C, Pulito V, et al. Activation of estrogen receptor-beta regulates hippocampal synaptic plasticity and improves memory. Nat Neurosci. 2008;11:334–343. doi: 10.1038/nn2057. [DOI] [PubMed] [Google Scholar]
  32. Matsuo N, Reijmers L, Mayford M. Spine-type-specific recruitment of newly synthesized AMPA receptors with learning. Science. 2008;319:1104–1107. doi: 10.1126/science.1149967. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Moosavi M, Naghdi N, Maghsoudi N, Zahedi Asl S. The effect of intrahippocampal insulin microinjection on spatial learning and memory. Hormones and behavior. 2006;50:748–752. doi: 10.1016/j.yhbeh.2006.06.025. [DOI] [PubMed] [Google Scholar]
  34. Moosavi M, Naghdi N, Maghsoudi N, Zahedi Asl S. Insulin protects against stress-induced impairments in water maze performance. Behav Brain Res. 2007;176:230–236. doi: 10.1016/j.bbr.2006.10.011. [DOI] [PubMed] [Google Scholar]
  35. O’Malley D, MacDonald N, Mizielinska S, Connolly CN, Irving AJ, Harvey J. Leptin promotes rapid dynamic changes in hippocampal dendritic morphology. Mol Cell Neurosci. 2007;35:559–572. doi: 10.1016/j.mcn.2007.05.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Park CR. Cognitive effects of insulin in the central nervous system. Neuroscience and biobehavioral reviews. 2001;25:311–323. doi: 10.1016/s0149-7634(01)00016-1. [DOI] [PubMed] [Google Scholar]
  37. Park CR, Seeley RJ, Craft S, Woods SC. Intracerebroventricular insulin enhances memory in a passive-avoidance task. Physiol Behav. 2000;68:509–514. doi: 10.1016/s0031-9384(99)00220-6. [DOI] [PubMed] [Google Scholar]
  38. Plum L, Schubert M, Brüning JC. The role of insulin receptor signaling in the brain. Trends Endocrinol Metab. 2005;16:59–65. doi: 10.1016/j.tem.2005.01.008. [DOI] [PubMed] [Google Scholar]
  39. Reger MA, Watson GS, Frey WH, Baker LD, Cholerton B, Keeling ML, Belongia DA, Fishel MA, Plymate SR, Schellenberg GD, et al. Effects of intranasal insulin on cognition in memory-impaired older adults: modulation by APOE genotype. Neurobiol Aging. 2006;27:451–458. doi: 10.1016/j.neurobiolaging.2005.03.016. [DOI] [PubMed] [Google Scholar]
  40. Reisel D, Bannerman DM, Schmitt WB, Deacon RMJ, Flint J, Borchardt T, Seeburg PH, Rawlins JNP. Spatial memory dissociations in mice lacking GluR1. Nat Neurosci. 2002;5:868–873. doi: 10.1038/nn910. [DOI] [PubMed] [Google Scholar]
  41. Schubert M, Gautam D, Surjo D, Ueki K, Baudler S, Schubert D, Kondo T, Alber J, Galldiks N, Küstermann E, 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]
  42. Schwarzberg H, Bernstein HG, Reiser M, Günther O. Intracerebroventricular administration of insulin attenuates retrieval of a passive avoidance response in rats. Neuropeptides. 1989;13:79–81. doi: 10.1016/0143-4179(89)90002-4. [DOI] [PubMed] [Google Scholar]
  43. Selman C, Lingard S, Choudhury AI, Batterham RL, Claret M, Clements M, Ramadani F, Okkenhaug K, Schuster E, Blanc E, et al. Evidence for lifespan extension and delayed age-related biomarkers in insulin receptor substrate 1 null mice. FASEB J. 2008;22:807–818. doi: 10.1096/fj.07-9261com. [DOI] [PubMed] [Google Scholar]
  44. Smith MA, Hisadome K, Al-Qassab H, Heffron H, Withers DJ, Ashford MLJ. Melanocortins and agouti-related protein modulate the excitability of two arcuate nucleus neuron populations by alteration of resting potassium conductances. J Physiol (Lond) 2007;578:425–438. doi: 10.1113/jphysiol.2006.119479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Steen E, Terry BM, Rivera EJ, Cannon JL, Neely TR, Tavares R, Xu XJ, Wands JR, de la Monte SM. 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]
  46. 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]
  47. Taguchi A, White MF. Insulin-like signaling, nutrient homeostasis, and life span. Annu Rev Physiol. 2008;70:191–212. doi: 10.1146/annurev.physiol.70.113006.100533. [DOI] [PubMed] [Google Scholar]
  48. Taniguchi CM, Emanuelli B, Kahn CR. Critical nodes in signalling pathways: insights into insulin action. Nat Rev Mol Cell Biol. 2006;7:85–96. doi: 10.1038/nrm1837. [DOI] [PubMed] [Google Scholar]
  49. van den Bout CJ, Machon O, Røsok Ø , Backman M, Krauss S. The mouse enhancer element D6 directs Cre recombinase activity in the neocortex and the hippocampus. Mech Dev. 2002;110:179–182. doi: 10.1016/s0925-4773(01)00597-4. [DOI] [PubMed] [Google Scholar]
  50. White MF. IRS proteins and the common path to diabetes. Am J Physiol Endocrinol Metab. 2002;283:E413–422. doi: 10.1152/ajpendo.00514.2001. [DOI] [PubMed] [Google Scholar]
  51. Withers DJ, Gutierrez JS, Towery H, Burks DJ, Ren JM, Previs S, Zhang Y, Bernal D, Pons S, Shulman GI, et al. Disruption of IRS-2 causes type 2 diabetes in mice. Nature. 1998;391:900–904. doi: 10.1038/36116. [DOI] [PubMed] [Google Scholar]
  52. Zhao W-Q, Chen H, Quon MJ, Alkon DL. Insulin and the insulin receptor in experimental models of learning and memory. Eur J Pharmacol. 2004;490:71–81. doi: 10.1016/j.ejphar.2004.02.045. [DOI] [PubMed] [Google Scholar]
  53. Zhao WQ, Alkon DL. Role of insulin and insulin receptor in learning and memory. Mol Cell Endocrinol. 2001;177:125–134. doi: 10.1016/s0303-7207(01)00455-5. [DOI] [PubMed] [Google Scholar]

RESOURCES